2008 JASN IMPACT FACTOR 7.505	HOME   AUTHOR INFO   EDITORIAL BOARD   SUBSCRIBE   FEEDBACK   ALERTS   HELP
advanced	CURRENT ISSUE	ARCHIVES	JASN Express	ONLINE SUBMISSION

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by KASISKE, B. L. Articles by SINGER, G. G. Search for Related Content PubMed PubMed Citation Articles by KASISKE, B. L. Articles by SINGER, G. G.

Recommendations for the Outpatient Surveillance of Renal Transplant Recipients

BERTRAM L. KASISKE^*, MIGUEL A. VAZQUEZ, WILLIAM E. HARMON, ROBERT S. BROWN, GABRIEL M. DANOVITCH^||, ROBERT S. GASTON^¶, DAVID ROTH, JOHN D. SCANDLING, JR.^|||| and GARY G. SINGER^¶¶ THE AMERICAN SOCIETY OF TRANSPLANTATION

^* Division of Nephrology, Hennepin County Medical Center, University of Minnesota, Minneapolis, Minnesota
Division of Nephrology, Southwestern Medical Center, University of Texas, Dallas, Texas
Children's Hospital, Harvard University, Boston, Massachusetts
Renal Unit, Beth Israel Hospital, Harvard University, Boston, Massachusetts
^|| Division of Nephrology, Department of Medicine, UCLA Medical Center, University of California, Los Angeles, California
^¶ Division of Nephrology, UAB Medical Center, University of Alabama at Birmingham, Birmingham, Alabama
Division of Nephrology, University of Florida, Miami, Florida
^|||| Transplantation, Division of Nephrology, Department of Medicine, Stanford University Medical Center, Stanford University, Palo Alto, California
^¶¶ Washington University School of Medicine, St. Louis, Missouri.

Correspondence to Dr. Bertram L. Kasiske, Division of Nephrology, Hennepin County Medical Center, 701 Park Avenue, Minneapolis, MN 55415. Phone: 612-347-5871; Fax: 612-347-2003; E-mail: kasis001{at}tc.umn.edu

	Abstract

Top Abstract Introduction Scope Intended Users Materials and Methods Organizational Scheme I. Frequency and Timing... II. Graft Function III. Immunosuppressive... IV. Cardiovascular Disease V. Bone and Bone... VI. Nutrition and Metabolism VII. Cancers VIII. Infections Endorsements References

 
Abstract. Many complications after renal transplantation can be prevented if they are detected early. Guidelines have been developed for the prevention of diseases in the general population, but there are no comprehensive guidelines for the prevention of diseases and complications after renal transplantation. Therefore, the Clinical Practice Guidelines Committee of the American Society of Transplantation developed these guidelines to help physicians and other health care workers provide optimal care for renal transplant recipients. The guidelines are also intended to indirectly help patients receive the access to care that they need to ensure long-term allograft survival, by attempting to systematically define what that care encompasses. The guidelines are applicable to all adult and pediatric renal transplant recipients, and they cover the outpatient screening for and prevention of diseases and complications that commonly occur after renal transplantation. They do not cover the diagnosis and treatment of diseases and complications after they become manifest, and they do not cover the pretransplant evaluation of renal transplant candidates. The guidelines are comprehensive, but they do not pretend to cover every aspect of care. As much as possible, the guidelines are evidence-based, and each recommendation has been given a subjective grade to indicate the strength of evidence that supports the recommendation. It is hoped that these guidelines will provide a framework for additional discussion and research that will improve the care of renal transplant recipients.

	Introduction

Top Abstract Introduction Scope Intended Users Materials and Methods Organizational Scheme I. Frequency and Timing... II. Graft Function III. Immunosuppressive... IV. Cardiovascular Disease V. Bone and Bone... VI. Nutrition and Metabolism VII. Cancers VIII. Infections Endorsements References

 
The morbidity and mortality rates associated with renal transplantation and the use of immunosuppressive medications are high. However, many posttransplant complications can be prevented, or at least more effectively treated, if they are detected earlier, rather than later. Guidelines have been developed for the prevention of diseases in the general population, but there are no comprehensive guidelines for the prevention of diseases and complications after renal transplantation. In addition, whether general disease prevention strategies and guidelines developed for the general population are applicable to transplant recipients has not been addressed in a systematic manner.

The recent emphasis on cost-cutting in medicine has created an impetus for health plans to develop guidelines designed to reduce expenditures. Unfortunately, guidelines developed by payers are rarely evidence-based and may not be primarily focused on optimizing patient outcomes. The American Society of Transplantation (AST) conducted a survey of medical and surgical directors of United Network for Organ Sharing (UNOS) renal transplant centers. Of the 117 respondents, 97% agreed that there was a need for recommendations guiding the surveillance of renal transplant recipients (R.S. Gaston, B.L. Kasiske, R.J. Tesi, G.M. Danovitch, and M.J. Bia, unpublished observations). The AST Board of Directors also concluded that there is a need for evidence-based recommendations designed to reduce the burden of disease after renal transplantation, in a cost-effective manner. It was also recognized that guidelines could indirectly help patients receive the access to care that they need to ensure long-term allograft survival. Therefore, the Board asked the AST Clinical Practice Guidelines Committee to develop recommendations for the outpatient surveillance of renal transplant recipients.

	Scope

Top Abstract Introduction Scope Intended Users Materials and Methods Organizational Scheme I. Frequency and Timing... II. Graft Function III. Immunosuppressive... IV. Cardiovascular Disease V. Bone and Bone... VI. Nutrition and Metabolism VII. Cancers VIII. Infections Endorsements References

 
These guidelines are applicable to all (adult and pediatric) renal transplant recipients. They cover outpatient screening for and prevention of diseases and complications that commonly occur after renal transplantation. They do not cover the diagnosis and treatment of diseases and complications after they become manifest. They do not cover the pretransplant period, and guidelines for the evaluation of renal transplant candidates have been published (1). They do not address the choice of immunosuppressive medications and cover only outpatient monitoring for and prevention of complications. They do not cover the treatment or prevention of renal allograft rejection. In addition, the guidelines do not cover every aspect of screening and prevention, and omissions should not be construed as recommendations for avoiding particular screening measures or preventive strategies.

	Intended Users

Top Abstract Introduction Scope Intended Users Materials and Methods Organizational Scheme I. Frequency and Timing... II. Graft Function III. Immunosuppressive... IV. Cardiovascular Disease V. Bone and Bone... VI. Nutrition and Metabolism VII. Cancers VIII. Infections Endorsements References

 
These guidelines were designed to be used by physicians and health care workers who care for renal transplant recipients in the outpatient clinics of kidney transplant centers. They were not specifically developed for primary care physicians; however, local referring physicians may find them useful. In addition, these guidelines may be helpful for trainees at all levels who wish to learn about renal transplantation. To the extent that these guidelines are evidence-based, they may also be useful for those seeking to identify future research needs. Finally, we recognize that some health care planners and providers may find that the guidelines help them to understand what is involved in the optimal treatment of renal transplant recipients.

	Materials and Methods

Top Abstract Introduction Scope Intended Users Materials and Methods Organizational Scheme I. Frequency and Timing... II. Graft Function III. Immunosuppressive... IV. Cardiovascular Disease V. Bone and Bone... VI. Nutrition and Metabolism VII. Cancers VIII. Infections Endorsements References

 
The committee searched the medical literature and reviewed pertinent publications. Searches were conducted using Medline and recent pertinent bibliographies. An electronic database was used to collate references, but no systematic data extraction or synthesis was performed. A draft of recommendations was developed using the expertise of committee members and the results of the literature review. This draft was sent to societies and individuals for review and suggested modifications. In particular, the draft was sent to the Council of American Kidney Societies and its non-AST member organizations, including the American Society of Nephrology, the Renal Physicians Association, the American Society of Pediatric Nephrology, and the National Kidney Foundation. The draft was also sent to individuals who are recognized experts in areas covered by the recommendations; the experts were specifically asked for their input. Taking into account the suggestions made by these organizations and individuals, a final draft was written and submitted to the AST Board for approval. It is anticipated that the reliability of these recommendations will diminish after 3 yr, at which time they will need to be updated.

We graded the strength of the evidence supporting each recommendation using the system developed by the Canadian Task Force on the Periodic Health Examination (2) and adopted by the United States Preventive Services Task Force (3). Accordingly, recommendations were graded A, B, C, D, or E as follows: A, there is good evidence to support the recommendation that the condition be specifically considered in a periodic health examination; B, there is fair evidence to support the recommendation that the condition be specifically considered in a periodic health examination; C, there is poor evidence regarding the inclusion of the condition in a periodic health examination, but recommendations may be made on other grounds; D, there is fair evidence to support the recommendation that the condition be excluded from consideration in a periodic health examination; E, there is good evidence to support the recommendation that the condition be excluded from consideration in a periodic health examination.

	Organizational Scheme

Top Abstract Introduction Scope Intended Users Materials and Methods Organizational Scheme I. Frequency and Timing... II. Graft Function III. Immunosuppressive... IV. Cardiovascular Disease V. Bone and Bone... VI. Nutrition and Metabolism VII. Cancers VIII. Infections Endorsements References

 
The guidelines are organized in sections and tables. Most tables indicate the incidence of the disease or complication being screened, the consequences of the disease or complication, the rationale for screening and/or prophylaxis, and specific recommendations. Each table is followed by a detailed discussion, which may include a definition of the disease or complication and discussions of the incidence, the consequences, and the rationale for screening or prophylaxis. Table 1 is a table of contents.

	I. Frequency and Timing of Outpatient Visits (Table 2)

Top Abstract Introduction Scope Intended Users Materials and Methods Organizational Scheme I. Frequency and Timing... II. Graft Function III. Immunosuppressive... IV. Cardiovascular Disease V. Bone and Bone... VI. Nutrition and Metabolism VII. Cancers VIII. Infections Endorsements References

 
There are virtually no scientific data on which to base decisions regarding the optimal frequency or type of contact between renal transplant recipients and transplant centers. Most outpatient encounters occur on the basis of circumstances and experiences that may be unique to each patient population and individual transplant center. To better understand current practices, we conducted a survey of medical and surgical directors of UNOS renal transplant centers.

Respondents to the survey reported that three-quarters of adult renal allograft recipients leave the hospital within 8 d after transplantation. Although the issue was not addressed in the survey, pediatric patients generally require longer hospitalizations (4). Virtually all patients return home, although some centers maintain nearby facilities for patients living a long distance away. In the AST survey, 80% of centers reported monitoring patients two or three times each week during the first 1 mo after transplantation. Between 1 and 3 mo after transplantation, 86% of centers continue to monitor recipients on at least a weekly basis. During this early posttransplant period, the risk of acute rejection and graft loss is at its peak (5). Accordingly, immunosuppression is most intense during this time. The overall risk of adverse immunologic and non-immunologic events during this period, combined with the relatively asymptomatic nature of most rejection episodes, suggests that frequent monitoring is warranted. In 75% of centers responding to the AST survey, a transplant physician or surgeon provides care during the first 3 mo.

After the first 3 mo, when and by whom patients are monitored vary widely among transplant centers. The role of transplant surgeons diminishes, in terms of the amount of outpatient care provided. Transplant physicians become the primary care-givers at some transplant centers, whereas the majority of care is provided elsewhere, usually by referring physicians, at other centers. Most recipients (4 to 12 mo after transplantation) are examined at least monthly, with 65% of outpatient visits occurring at the transplant center. The primary focus of visits during this period is ongoing surveillance for acute rejection, infection (the highest risk of opportunistic infections occurs between 1 and 6 mo after transplantation), and immunosuppressive agent toxicity (6,7,8).

For patients with functioning allografts after 12 mo, issues and practices change. The risk of acute rejection, in the absence of therapy nonadherence, is small. Immunosuppressive medication dosing for most patients requires only infrequent monitoring and adjustment. Common practice is for patients to undergo outpatient visits every 3 to 6 mo (as reported by 62% of the centers responding to the AST survey); however, 25% of centers continue to monitor patients monthly. Only approximately one-half of these visits occur at the transplant center. The primary focus of these "late" encounters is ongoing surveillance of allograft function (with careful attention to allograft function and urinary protein excretion), as well as re-evaluation of the immunosuppressive regimen in terms of efficacy, toxicity, and cost (9,10,11). If patients experience problems, prompt attention is most likely to yield effective results. Problems detected by transplant physicians or other primary providers usually result in visits to the transplant center.

Pediatric patients require different schedules of surveillance. These differences are based on the unique complications and outcomes of pediatric renal transplant recipients. The incidence of acute rejection episodes is higher among children than among adults, and rejection may be more difficult to diagnose, particularly if there is a large difference in size between the donor and the recipient (12, 13). Also, the metabolism of many immunosuppressive medications is substantially different in young children, compared with adults, and drug metabolism changes as children grow and develop. Furthermore, the incidence and severity of nonadherence are thought to be higher among adolescents than among adults. The growth and development of children are adversely affected by graft dysfunction and by some of the commonly used immunosuppressive medications, necessitating frequent monitoring and adjustment. Children frequently exhibit urologic abnormalities as the cause of end-stage renal disease (ESRD), which may require reconstructive surgery before or after renal transplantation (4). Therefore, they require careful assessment for urinary tract infections and monitoring of bladder and genitourinary function. Some reports have demonstrated better outcomes if the care of children is provided by specialized pediatric renal transplant centers (14,15,16,17).

Recent improvements in short-term graft survival rates have resulted in a larger number of patients being exposed to the risks of late complications (5, 18). Maximization of long-term patient survival rates requires the ongoing management of cardiovascular disease (CVD) risk factors, such as hyperlipidemia and hypertension, as well as screening for cancer and other complications that are directly or indirectly linked to immunosuppression. Because of the complexity of current immunosuppressive regimens and the high incidence of adverse drug effects and drug interactions, an experienced transplant physician should be involved in the long-term care of transplant recipients.

There are a number of reasons why relatively frequent visits to the transplant center may improve long-term outcomes. Chronic allograft failure, death with a functioning allograft, and nonadherence account for the great majority of late graft losses. The effects of each of these factors may be diminished by frequent visits to the transplant center. With respect to nonadherence, effective immunosuppressive regimens have greatly reduced early graft loss and have enabled a growing number of recipients to achieve long-term graft survival, despite the fact that they remain at high risk for allograft rejection. It is ironic that the effectiveness of immunosuppressive medications administered to prevent early rejections has resulted in more recipients being dependent on these potent drugs to maintain long-term graft function. Consequently, the relative effects of nonadherence on graft outcomes is becoming greater (19,20,21). Data gathered in other fields of medicine confirm the importance of frequent encounters between patients and providers in facilitating adherence to therapy. Therefore, efforts to decrease the frequency of outpatient visits to the transplant center or other providers may be counterproductive. Some data indicate that patients actually dislike long periods between visits; such gaps tend to foster uncertainty regarding the status of the allograft and worries about potential adverse outcomes.

	II. Graft Function

Top Abstract Introduction Scope Intended Users Materials and Methods Organizational Scheme I. Frequency and Timing... II. Graft Function III. Immunosuppressive... IV. Cardiovascular Disease V. Bone and Bone... VI. Nutrition and Metabolism VII. Cancers VIII. Infections Endorsements References

 
Graft Dysfunction (Table 3)
Definition. Most but not all of the adverse consequences of renal failure can be directly attributed to insufficient GFR. Although other functions of the kidney are important, failure of these other functions is rarely observed in the absence of reduced GFR.

Incidence. Epidemiologic studies using large transplant registries have unequivocally demonstrated a progressive loss of renal allografts in the late posttransplant period. Therefore, only patients who die with a functioning allograft will likely escape a decrease in GFR at some time after transplantation.

Consequences. The function of the transplant needs to be estimated repeatedly during the life of the graft. Episodes of acute allograft rejection have been documented in the third decade after transplantation, and allografts are constantly threatened by nonimmunologic injures that may or may not be reversible. At no time after transplantation can function stability be taken for granted. Clinical signs of rejection and dysfunction are notoriously unreliable, and repeated laboratory testing is required.

Rationale. There are a number of circumstances in which screening for graft dysfunction may alter therapy and benefit the patient. The first is the detection of acute rejection. Generally, the only practical way to detect acute rejection in its early treatable stage is to screen at regular intervals with some direct or indirect measurement of renal function. Acute rejection often occurs without signs or symptoms. Frequent repetitive allograft biopsies or fine-needle aspiration biopsies are theoretically possible but are less practical. The early recognition of acute rejection and its complete reversal may minimize its unfavorable effects on long-term graft function. The completeness of reversal of graft dysfunction after an episode of acute rejection, as measured by the serum creatinine level, is a prognostic indicator.

The second case is the detection of nephrotoxicity resulting from immunosuppressive medications. Cyclosporine A (CsA) and tacrolimus have become mainstays of immunosuppression after renal transplantation. Both can cause renal toxicity. As in the case of acute rejection, nephrotoxicity can be detected only by screening for changes in renal function. The detection of nephrotoxicity may prompt changes in immunosuppressive medications. Even small changes in the doses of immunosuppressive medications may cause changes in graft function.

The third case is the detection of changes in renal function resulting from obstruction, renal artery stenosis, or other reversible causes of graft dysfunction. Other causes of graft dysfunction may be reversible if detected early and may be detected only if screening tests of renal function are routinely performed.

In the fourth case, chronic renal allograft failure is often evident only from subtle changes in renal function. With time, all kidney transplants are subject to chronic allograft failure, which may occur as early as 6 mo after transplantation (22). Although there is no proof from randomized controlled trials, an increasing amount of evidence suggests that aggressive risk-factor management may help slow the rate of progression of chronic allograft failure (23,24,25,26,27,28,29,30,31). Controlling BP, reducing urinary protein excretion, treating hyperlipidemia, and maximizing immunosuppression may help slow the rate of progression. Repetitive measurements of renal function may indicate patients with early chronic allograft failure and thereby identify patients who require close management and supervision. In addition, measurements of the rate of progression may help patients and caregivers plan for the eventual need for renal (allograft) replacement therapy.

Finally, it is likely that measuring renal function is a helpful reminder to patients of the possibility of allograft dysfunction and regular measurements of renal function may help encourage patients and remind them to adhere to therapy. At least one study noted that the failure of patients to adhere to prescribed serum creatinine screening was a predictor of late renal allograft failure (32).

Measurement of serum creatinine levels is a simple and inexpensive but invaluable indirect method to detect decreases, especially acute decreases, in GFR. The universal availability of laboratories that can measure serum creatinine levels makes this test particularly useful. The serum creatinine level is also the most valuable prognostic marker of subsequent graft function at all times after transplantation. The interval for measuring serum creatinine levels can be based on the frequency of expected graft dysfunction at different times after transplantation. Most cases of acute rejection and episodes of graft dysfunction occur in the first 2 mo after transplantation.

Pediatric renal transplant recipients experience more acute rejection episodes than do adults (12,13). This may be attributable to increased metabolism of immunosuppressive medications, heightened immune responses, or medication nonadherence. The incidence of late acute rejections may also be higher. As a result, children require more frequent surveillance. The normal range for serum creatinine levels changes with age and body size.

Notwithstanding the practical advantages of using the serum creatinine level as a marker of renal function, it is a relatively inaccurate indicator of the true GFR for renal transplant recipients (33,34,35). With formulas that take into account clinical parameters that are correlated with muscle mass, serum creatinine levels may more accurately reflect GFR (36,37,38,39,40). However, GFR estimated by using formulas that include serum creatinine levels are also subject to substantial variability (41,42,43). In addition, many of the variables used in formulas to estimate GFR (e.g., gender, age, height, and body weight) adjust more for differences between patients than for differences that occur in the same patients with time. The latter differences are most critical in the monitoring of renal transplant recipients. However, during long periods of time, changes in muscle mass and the resulting changes in creatinine production make the serum creatinine level a less sensitive marker for detecting chronic decreases in GFR (44). Indeed, studies have demonstrated that the serum creatinine level is relatively unreliable for detecting chronic changes in renal function (33,34,45). Other endogenous markers of GFR, such as the serum cystatin C level (46), may ultimately prove to be superior to the serum creatinine level for detecting acute and chronic changes in renal function, but these markers require additional testing before they are adopted in routine clinical practice (47).

Urinary creatinine clearance determinations may be even less accurate than serum creatinine measurements. This is largely attributable to the difficulty of obtaining accurate timed urine collections. In addition, tubular secretion of creatinine results in systematic overestimation of the true GFR for renal transplant recipients (42,48). This overestimation of the GFR can be reduced by blocking tubular secretion of creatinine with cimetidine (49,50,51,52). Nevertheless, the usefulness of urinary creatinine clearance measurements is limited.

Urinary and plasma clearance methods that use isotopic or nonisotopic markers of GFR are most accurate. Plasma clearance tests are relatively simple to perform and do not require accurate (and difficult-to-obtain) urine collections. Even methods that use only a single plasma sampling time after injection are reasonably accurate and generally acceptable for clinical practice (48,53,54). Although clearance of isotopic agents can also be measured with a gamma camera positioned over the allograft (55,56), this approach is more cumbersome and expensive.

It can be argued that the detection of chronic decreases in GFR is not critical for the treatment of transplant recipients, because few therapeutic options are available. This line of reasoning holds that it is generally not necessary to routinely measure renal function by more accurate techniques that use isotopic or nonisotopic markers of GFR. However, these tests may be valuable to assess the significance of changes in serum creatinine levels with time and to provide more accurate baseline levels for follow-up monitoring. A growing body of evidence (albeit largely circumstantial) suggests that at least some therapies may delay the progression of chronic allograft failure (23,24,25,26,27,28,29,30,31). Therefore, detecting even small, gradual, progressive decreases in GFR may prompt patients and caregivers to be even more diligent in managing immunologic and nonimmunologic risks for graft failure. Finally, these more accurate techniques are often needed to assess the true GFR in planning for subsequent dialysis or repeat transplantation. UNOS currently mandates that patients cannot begin to accrue waiting time points on the cadaveric transplant waiting list until their measured or estimated creatinine clearance values are 20 ml/min.

Proteinuria (Table 4)
Definition. Transient proteinuria is common and is often observed in association with episodes of acute allograft rejection. Transient proteinuria may not affect allograft or patient survival independent of its underlying cause. Persistent proteinuria is usually defined as protein excretion of >0.5 to 1.0 g/24 h for at least 3 to 6 mo.

Incidence. A number of epidemiologic studies have examined the incidence of persistent proteinuria among renal transplant recipients. In those studies, the incidence of proteinuria (variously defined) ranged from 10 to 25% (57,58,59,60,61,62,63,64,65,66,67).

Consequences. The capacity to excrete urine that is virtually free of filtered protein is a critical function of healthy kidneys. Proteinuria is a manifestation of renal dysfunction and, when heavy, has important consequences for extracellular fluid volume regulation. The degree of proteinuria also has prognostic and diagnostic implications. Heavy proteinuria is often associated with more rapid deterioration of renal function and is more likely to be associated with pathologic glomerular lesions than with interstitial lesions or anatomic or hemodynamic causes of renal dysfunction (10,59,60,62,63,64,65,66). The most common causes are chronic allograft nephropathy and recurrent or de novo glomerulonephritis (57,58,59,60,61,62,63,65). Recurrent diabetic nephropathy is likely to become an increasingly common cause of proteinuria as the number of diabetic patients with long-term graft survival increases. Recently, experimental data suggested that proteinuria itself may cause renal injury (68,69). Therefore, it is possible that the often-reported clinical association between proteinuria and the rate of renal function decrease (70) could be, in part, a direct result of injurious effects of proteinuria. A number of studies in the general population have suggested that proteinuria is an important risk factor for CVD (71,72,73).

Rationale. Indications for and methods of screening for proteinuria were recently reviewed in a position paper by the National Kidney Foundation (74). The prevalence of proteinuria among renal transplant recipients is high enough to give screening tests sufficient positive and negative predictive values. The consequences of new-onset persistent proteinuria are of sufficient magnitude to make screening a useful prognostic test. Treatment that may diminish the adverse consequences of proteinuria is available (25,75,76,77,78,79). Many of the reasons for screening in high-risk populations that were outlined in the National Kidney Foundation position paper can be extrapolated to renal transplant recipients. However, it should be kept in mind that diseased native kidneys may often contribute to low levels of proteinuria and may occasionally account for large amounts of urinary protein excretion.

Screening for proteinuria is probably best performed with a spot, first-voided, morning urine sample, to negate effects of postural proteinuria, but this may not be practical. The urine dipstick test is a convenient, inexpensive, readily available screening test. It provides an estimate of the urine total protein concentration, which is an indirect measure of urinary protein excretion. The urine protein concentration can be influenced not only by the protein excretion rate but also by the urine concentration (80). Therefore, the interpretation of dipstick results should take into account the urine volume, which may dilute the protein concentration. A 1 + or trace result for a dilute urine sample (e.g., specific gravity of 1.010) likely indicates more protein excretion than does a 1 + or trace result for a concentrated urine sample (e.g., specific gravity of 1.030) (80). Although the standard dipstick test provides only a semiquantitative estimate of protein excretion, a result of 1 + or greater typically reflects clinically significant proteinuria.

The urine protein/creatinine ratio provides a more accurate assessment of total protein excretion, largely because it takes into account errors attributable to differences in urine concentrations. Measurement of the urine protein/creatinine ratio is convenient, because it requires only a spot sample. A value of 200 mg protein/g creatinine is generally considered abnormal (74). Studies have indicated that protein/creatinine ratio measurements provide results equivalent in value to 24-h collection data in most cases. However, the definitive measurement of total protein excretion is the 24-h urine collection test. Protein excretion of >200 mg/24 h is considered abnormal (74). This test should be accompanied by urine and serum creatinine level measurements, to permit estimation of GFR and determination of the completeness of urine collection. The principle disadvantages of the 24-h collection test are its inconvenience and its potential inaccuracy because of poor adherence to correct collection techniques. Patients must be trained in the proper method for collection and must be motivated to adhere to instructions. Results are delayed by the time required for patients to collect the samples.

Patients with glomerular disease should undergo proteinuria measurements just before transplantation, if possible. The possibility that posttransplant proteinuria is attributable to the native kidneys should be considered, especially in the early posttransplant period. After proteinuria is detected, how often proteinuria needs to be measured is dependent on the cause, amount, and management of the proteinuria.

Measurements of urine albumin excretion have not been demonstrated to provide diagnostic or prognostic information that is not provided by measurements of total protein excretion for transplant recipients. Measuring urine albumin excretion may prove to be unnecessary, because the prevalence of clinically significant glomerular proteinuria is very high in the transplant population. How the presence of native kidneys might confound the interpretation of microalbuminuria after renal transplantation is unknown. Additional studies are needed to better define the role, if any, of measuring urine albumin excretion after renal transplantation.

Clinically Silent Allograft Rejection (Table 5)
Definition. The histologic findings of acute allograft rejection, i.e., interstitial infiltrates and mild tubulitis that usually occur in association with an acute deterioration in renal function, may be observed in the absence of renal functional changes.

Incidence. The prevalence of acute rejection in protocol biopsies obtained in the first days to weeks during delayed graft function is 15 to 30% (81,82,83). The prevalence of clinically silent acute rejection at 3 mo has been reported to be 4% (84), 17% (85), and 27% (86). At 3 mo, borderline acute rejection changes were observed in 31% (84), 71% (85), and 21% (86) of cases. The incidences of subclinical and borderline acute rejection at 6 mo were 24 and 25%, respectively, in one study (86). The prevalence of subclinical acute rejection at 2 yr has been reported to be 9% (87) and 12% (85). Evidence of chronic allograft nephropathy in protocol biopsies performed at 3 mo has been reported to be present in 3% (Banff chronic grade, 2) (86), 20% (85), 24% (Banff chronic grade, 2) (88), and 38.3% (84) of cases. The prevalence of chronic allograft nephropathy in protocol biopsies increases to 50 to 70% by 2 yr (85,87).

Consequences. Preliminary data suggest that clinically silent acute rejection may be associated with an increased incidence of graft dysfunction (86).

Rationale. Renal allograft biopsies are often obtained for the diagnosis and management of decreased allograft function, and biopsies performed in that setting are not considered further in these guidelines for posttransplant screening and surveillance. However, periodic examinations of renal histologic features in the absence of changes in renal function may reveal silent rejection that could jeopardize long-term graft survival. Core needle biopsies performed under ultrasonographic guidance, using disposable, spring-loaded biopsy needles, are generally safe and provide adequate tissue for the detection of acute rejection in adult and pediatric transplant recipients (89,90,91,92). Standard histologic techniques are adequate for most cases of acute graft dysfunction.

In a small, randomized, controlled trial, Rush et al. (86) found that treatment of subclinical rejection detected by protocol biopsies led to better graft function, compared with standard management without protocol biopsies. They also reported uncontrolled data indicating that additional immuno-suppression may decrease the incidence of clinically apparent rejection, without affecting the incidence of subclinical rejection (93). These data suggest that protocol biopsies may be necessary to maximally decrease the incidence of rejection, even with newer, more potent, immunosuppressive regimens. Confirmation of these results may elevate protocol biopsies from a research tool to standard clinical practice (94). Additional studies are needed to confirm the utility and cost-effectiveness of protocol biopsies before their widespread adoption in clinical practice. Performing biopsies in high-risk situations, e.g., before major reductions in immunosuppression, may also be warranted if their utility can be confirmed in well-designed clinical trials.

Fine-needle aspiration cytologic examinations have been proposed as an alternative, less invasive strategy to accomplish the same purpose (95,96,97,98,99,100,101). Aspiration biopsy is a technically simple and safe procedure that can rapidly provide valuable information regarding causes of acute graft dysfunction in the early posttransplant period. It is limited by an approximately 20 to 30% incidence of inadequate samples and by the necessity for skilled cytologic evaluation. Special stains may permit the diagnosis of cytomegalovirus (CMV) infection, and experimental techniques may suggest preclinical diagnoses of acute rejection. Fine-needle aspiration cytologic examinations are not widely available. Noninvasive experimental techniques for detecting silent rejection, using plasma or urine samples, have not yet entered routine clinical practice.

	III. Immunosuppressive Medications

Top Abstract Introduction Scope Intended Users Materials and Methods Organizational Scheme I. Frequency and Timing... II. Graft Function III. Immunosuppressive... IV. Cardiovascular Disease V. Bone and Bone... VI. Nutrition and Metabolism VII. Cancers VIII. Infections Endorsements References

 
Efficacy and Toxicity of CsA (Table 6)
Incidence. Compared with conventional immunosuppression with azathioprine, CsA reduced the incidence of acute rejection and prolonged graft survival in randomized controlled trials (102,103,104,105,106,107). Early on, it was appreciated that CsA caused nephrotoxicity. The incidence of CsA nephrotoxicity is variable and poorly defined. Reversible decreases in renal blood flow and GFR are observed for most patients. Acute toxicity characterized by endotheliosis and arteriopathy is less common. Chronic, nonspecific, tubulointerstitial atrophy and fibrosis are probably common, but it is difficult to distinguish these from chronic allograft nephropathy attributable to other causes. Hypertension occurs in 41 to 82% (102,106,107), hypercholesterolemia in 37% (107), hyperuricemia in 35 to 52% (106,107), hyperkalemia in 55% (106), tremor in 12 to 39% (102,104,106,107), gingival hyperplasia in 7 to 43% (102,104,107), diabetes mellitus in 2 to 13% (102,107), and hirsutism in 29 to 44% (102,104) of cases.

Consequences. Failure to prevent acute rejection can lead to allograft failure. The role of acute and chronic CsA nephrotoxicity in causing graft failure is less clear. Some adverse effects, such as hypercholesterolemia, hypertension, and diabetes mellitus, increase the risk of CVD.

Rationale. The bioavailability of CsA is quite variable. In addition, the therapeutic window for CsA (the range of blood levels at which CsA is efficacious but not toxic) seems to be very narrow. As a result, it is generally not possible to determine the correct dose of CsA for individual patients without measuring blood levels. Numerous studies have demonstrated that low CsA blood levels are correlated with subsequent episodes of allograft rejection and, ultimately, allograft failure (108,109,110,111,112,113,114,115,116,117,118,119,120,121,122,123). Studies have also demonstrated that high CsA blood levels tend to be correlated with decreased renal allograft function, presumably as a result of nephrotoxicity (108,110,115,116,117,124,125). These correlations tend to confirm the biologically plausible notion that efficacy and toxicity are linked to blood levels. The association between blood levels and CsA efficacy and toxicity is inexact. Very high levels are more likely to be associated with nephrotoxicity, whereas low levels are more likely to be associated with acute rejection. However, levels in the relatively wide range between these two extremes are less predictive of efficacy and toxicity. A few studies have even suggested that variability in CsA absorption itself is correlated with chronic allograft rejection and decreased graft survival (126).

A number of factors affect CsA blood levels. These include hemoglobin levels (127,128), serum lipid levels (127,128,129), age (121,124,128,130), gender (124), hepatic dysfunction (124), time of day (131), and race (128). A large number of medications have also been demonstrated to affect CsA blood levels, and it is extremely important to measure CsA blood levels whenever medications that can interact with CsA are prescribed. The inherent intra- and interpatient variability in CsA absorption and metabolism make it necessary to use blood levels to monitor and adjust CsA doses. In general, children metabolize CsA faster than adults, and young children generally require three times/d dosing (132). As children grow, the dose of CsA required to maintain the same therapeutic levels increases (132). Because of these differences in metabolic rates and doses, children usually require more frequent monitoring than adults.

Pharmacokinetic studies are less variable (133), are more closely correlated with dose (109,133), and more accurately predict efficacy and toxicity than do trough blood levels (109,134,135,136). However, pharmacokinetic studies are inconvenient for patients and expensive to perform. As a result, few clinics routinely perform pharmacokinetic studies. A number of investigators have attempted to render pharmacokinetic studies more practical for day-to-day patient care by demonstrating that the number of blood samples required can be reduced (115,132,137,138,139). However, these abbreviated pharmacokinetic studies have not yet gained widespread acceptance in clinical practice. Instead, most transplant centers use trough CsA levels for dose adjustments. There are few data to suggest how frequently CsA levels need to be determined. However, because it is well known that absorption and metabolism change rapidly soon after transplantation, levels should be measured more frequently during the first few months after transplantation. Blood levels should also be measured whenever there is a change in medications or other factors that may affect CsA metabolism.

Monitoring of CsA efficacy could theoretically be achieved by developing appropriate pharmacodynamic assays, e.g., by monitoring CsA-suppressible immunologic parameters. Several investigators have attempted to develop pharmacodynamic assays to monitor CsA activity (116,140,141,142,143,144,145,146,147). Although this approach holds great promise, pharmacodynamic monitoring of CsA therapy is generally considered experimental at this time.

A number of adverse effects are attributable to CsA. These include hypertension, hyperlipidemia, glucose intolerance, hypomagnesemia, hyperkalemia, hyperuricemia, gout, gingival hyperplasia, hirsutism, central nervous system toxicity, peripheral neuropathy, and possibly metabolic bone disease. Renal dysfunction, hypertension, hyperlipidemia, glucose intolerance, hypomagnesemia, hyperkalemia, hyperuricemia, gout, and metabolic bone disease are multifactoral, and appropriate screening tests are covered elsewhere in these guidelines. Gingival hyperplasia, hirsutism, central nervous system toxicity (especially tremor), and peripheral neuropathy are all detected by physical examinations. It is essential that patients being treated with CsA be examined periodically by a transplant physician familiar with these and other CsA-related complications. Although the optimal frequency of examination has not been well defined, patients should probably be examined more frequently in the early posttransplant period (Table 2).

There are several different formulations of CsA. The pharmacokinetic characteristics of these formulations vary, and the interpretation of CsA trough levels, for example, may be quite different for different formulations. A detailed discussion of the specific pharmacokinetic characteristics of each CsA formulation is beyond the scope of these guidelines. However, a number of studies suggest that the microemulsion formulation of CsA, e.g., Neoral (Novartis, Basel, Switzerland), is better absorbed and is associated with reduced intra- and interpatient variability in CsA blood levels, compared with standard CsA preparations, e.g., Sandimmune (Novartis) (148,149,150,151,152,153,154). It is possible that microemulsion CsA could be monitored with fewer blood level measurements, compared with the standard CsA formulation; however, there have been no studies correlating outcomes with the frequency of blood level monitoring for any CsA preparation.

Studies suggest that the number of acute rejections is higher, whereas the incidence of adverse events may be no different or even lower, with Neoral versus Sandimmune (155). However, a recent meta-analysis suggested that there were differences in the results of blinded, randomized, controlled trials, compared with those of open-label trials (156). In open-label studies, Sandimmune was associated with more adverse events than was Neoral. However, in blinded, randomized, controlled trials, the incidence of adverse events was higher with Neoral, compared with Sandimmune (156). Therefore, the results of that analysis suggest that the lower rate of acute rejection with Neoral, compared with Sandimmune, is accompanied by a higher incidence of adverse events. These results raise the question of whether the reported differences between the two preparations are entirely attributable to higher blood levels with Neoral versus Sandimmune, resulting in both lower rejection rates and greater toxicity.

Efficacy and Toxicity of Tacrolimus (Table 7)
Incidence. The efficacy and toxicity of tacrolimus were compared with those of CsA in five randomized controlled trials (157,158,159,160,161) and in a meta-analysis of four of those trials (162). The quality of the studies was poor, with none using blinding and none reporting methods of randomization (162). In those trials patients also received corticosteroids and azathioprine for maintenance of immunosuppression. Tacrolimus had no effect on graft loss or mortality rates at 1 yr (162). The incidence of acute rejection at 1 yr was lower among patients treated with tacrolimus, compared with CsA (odds ratio, 0.52; 95% confidence interval, 0.36 to 0.75) (162). In the two largest trials (159,160), detailed information was provided on the incidences of adverse effects with tacrolimus, i.e., decreased renal function in 35 to 42%, diarrhea in 22 to 44%, constipation in 31 to 35%, vomiting in 13 to 29%, hypertension in 37 to 50%, infections in 72 to 76%, and CMV infection in 14 to 20% of cases. In 3-yr follow-up data, the incidences of CMV infection were similar for patients treated with tacrolimus (19.5%) and those treated with CsA (19.3%) (163). Tremor was more common with tacrolimus (35 to 54%), compared with CsA (12 to 34%) (159,160). However, gingival hyperplasia was more common with CsA (5.3 to 6.2%) than with tacrolimus (0.5 to 1.3%) (159,160). Similarly, hirsutism was more common with CsA (8.7 to 9.7%) than with tacrolimus (0 to 0.5%) (159,160). Three trials reported the prevalence of posttransplant diabetes mellitus at 1 yr to be substantially higher with tacrolimus (12 to 20%), compared with CsA (2 to 4%) (158,159,160). The odds ratio for posttransplant diabetes mellitus at 1 yr was 5.03 (95% confidence interval, 2.04 to 12.36) (162). In a more recent trial, the incidence of new-onset, posttransplant diabetes mellitus at 1 yr was 14.0% among patients treated with tacrolimus and azathioprine, 6.5% among patients treated with tacrolimus and mycophenolate mofetil (MMF), and 6.5% among patients treated with CsA and MMF (161). The incidence of hypercholesterolemia was higher among patients treated with CsA (14.5 to 31.9%), compared with tacrolimus (7.8 to 17.5%) (159,161). The incidence of tacrolimus nephrotoxicity was variable and dependent on the definition of toxicity (transient or sustained decreases in renal function, arteriopathy, or nonspecific tubulointerstitial atrophy and fibrosis), as well as the dose and target blood levels used (158,159,160,164,165,166,167,168,169,170). Small uncontrolled reports suggested that tacrolimus can be safely used for children (171).

Consequences. Failure to prevent acute rejection can lead to allograft failure. The role of acute and chronic tacrolimus nephrotoxicity in causing graft failure is unclear. However, the therapeutic window for tacrolimus is narrow, and the incidence of renal and nonrenal toxicity is roughly proportional to doses and blood levels.

Rationale. Whole-blood tacrolimus levels can be measured with a variety of analytical techniques (172,173,174,175,176,177,178). The bioavailability of tacrolimus is quite variable, being influenced by such factors as age, race, hepatic function, and concomitant medications (179,180,181,182). In addition, the range of blood levels at which tacrolimus is efficacious but not toxic appears to be very narrow. Therefore, it is generally not possible to determine the correct dose of tacrolimus for individual patients without measuring blood levels. Studies have demonstrated that low tacrolimus blood levels are correlated with subsequent episodes of allograft rejection (161,164,168,169). High tacrolimus blood levels tend to be correlated with decreased renal allograft function (presumably as a result of nephrotoxicity) and other adverse effects (164,165,167,168,169,183). These correlations tend to confirm the biologically plausible notion that efficacy and toxicity are linked to drug exposure. However, the association between blood levels and tacrolimus efficacy and toxicity is inexact.

A number of factors affect tacrolimus blood levels. Like CsA, tacrolimus is metabolized by the cytochrome P-450 3A enzyme system. Therefore, as is the case with CsA, medications that inhibit or induce this enzyme pathway can increase or decrease tacrolimus levels, respectively (184). The intra- and interpatient variability in absorption and metabolism necessitates the use of blood levels for monitoring and adjustment of tacrolimus doses (178,180,185,186). Although pharmacokinetic monitoring is likely more predictive of efficacy and toxicity than are trough level measurements, complete pharmacokinetic studies are inconvenient and expensive. Therefore, most transplant centers use trough tacrolimus levels for dose adjustments. Abbreviated (one- or two-point) area under the plasma concentration-time curve (AUC) determinations may be a reasonable alternative to trough level measurements or complete pharmacokinetic studies (187). There are few data to suggest how frequently levels need to be determined. However, levels should be measured more frequently during the first few months after transplantation. Levels should also be measured whenever there is a change in medications or other factors that may affect tacrolimus metabolism.

Efficacy and Toxicity of Sirolimus (Table 8)
Incidence. In a multicenter, randomized, open-label trial, 42 patients treated with CsA were compared with 41 patients treated with sirolimus (188). All patients received corticosteroids and azathioprine. The incidences of biopsy-confirmed acute rejection at 12 mo were similar for sirolimus (41%) and CsA (38%). Several adverse effects were significantly more common among patients treated with sirolimus versus CsA, including hypertriglyceridemia (51% versus 12%), hypercholesterolemia (44% versus 14%), thrombocytopenia (37% versus 0%), leukopenia (39% versus 14%), elevated alanine aminotransferase levels (17% versus 0%), and hypokalemia (34% versus 0%). Hyperglycemia was also more common (20% versus 7%), but this difference was not statistically significant, and the incidences of diabetes mellitus were the same for the two groups (1%). There were no statistically significant differences between sirolimus- and CsA-treated patients with respect to the incidences of CMV infection (14% versus 12%), hypertension (17% versus 33%), tremor (2% versus 14%), or gingival hyperplasia (0% versus 10%). Serum creatinine and uric acid levels were lower for patients treated with sirolimus, compared with CsA (188).

A similar efficacy and toxicity profile was demonstrated for sirolimus in a phase II trial (n = 149) (189). In that trial, sirolimus appeared to have a CsA-sparing effect, in that the rates of acute rejection among Caucasian patients treated with sirolimus were similar for patients treated with either full or reduced doses of CsA (189). However, sirolimus failed to reduce the rate of acute rejection among African American patients treated with a reduced dose CsA (189).

In another phase II, randomized, open-label trial, sirolimus (n = 40) was compared with CsA (n = 38) among patients who also received corticosteroids and MMF (190). At 12 mo, the incidences of biopsy-proven acute rejections were not significantly different for sirolimus (27.5%) and CsA (18.4%). Adverse effects that were significantly more frequent in the sirolimus-treated group included thrombocytopenia (45% versus 8%) and diarrhea (38% versus 11%). Increases in serum creatinine levels were significantly less common with sirolimus (18%), compared with CsA (39%). Tremor (5% versus 21%) and CMV viremia (5% versus 21%) were also significantly less common with sirolimus, compared with CsA (190). There were no differences in the incidences of hyperlipidemia, hyperglycemia, hyperuricemia, or hypertension (190).

Consequences. Sirolimus seems to be efficacious in preventing acute rejection when used in place of, or in combination with, CsA. However, most patients experience adverse effects.

Rationale. Methods that have been used to measure sirolimus blood levels include HPLC with ultraviolet light detection (188,191,192,193,194), HPLC-mass spectrometry (195), electrospray-HPLC-mass spectrometry (196), and immunophilin radioreceptor assays (197,198). Very few studies have been conducted to determine the relationship between blood levels of sirolimus and either acute rejection or toxicity. However, drug concentrations of sirolimus exhibit considerable intra- and interpatient variability, and monitoring therapy by measuring blood levels could be expected to improve efficacy and reduce toxicity. In a pilot, randomized, open-label, multicenter trial, steady-state whole-blood concentrations were used to adjust doses (188). Trough concentrations of "approximately 30 ng/ml" in the first 2 mo and 15 ng/ml thereafter were targeted (188). In a series of 150 patients who were monitored for 4 yr, trough levels of >15 ng/ml were correlated with hypertriglyceridemia, thrombocytopenia, and leukopenia, whereas levels of <5 ng/ml were associated with acute rejections (192). These and other preliminary data suggest that monitoring of sirolimus blood levels may be useful. Whether pharmacokinetic monitoring is superior to trough level measurements is unclear. In at least one study, trough levels were well correlated with AUC levels (192).

It is not yet clear whether sirolimus affects the pharmacokinetics of CsA (191,195,196,199), but blood levels of sirolimus were observed to be increased by CsA in at least one study (200). Sirolimus may cause modest increases in corticosteroid levels in patients receiving prednisone (201). Other preliminary studies suggested that blood levels of sirolimus vary with respect to race (black versus nonblack) but not gender (199). Dietary fat seems to increase sirolimus absorption (202). Very few studies have evaluated the potential utility of pharmacodynamic monitoring of sirolimus (203).

Efficacy and Toxicity of MMF (Table 9)
Screening tests. White blood cell counts and hemoglobin levels (or hematocrit values) may be used to detect leukopenia and/or anemia. Clinical manifestations of nonimmune toxicities of MMF are usually detectable in routine history assessments and physical examinations.

Incidence. The recommended starting dose of MMF is 1.0 g twice daily for adults. The dose for children is 600 mg/m² twice daily (204). The authors of one study recommended that the dose for children should be 600 mg/m² twice daily when MMF is used in combination with CsA, 300 mg/m² twice daily when MMF is used with tacrolimus, and 500 mg/m² twice daily when MMF is used without a calcineurin inhibitor (205). Capsules, liquid, and intravenous preparations are available. Higher doses are associated with increased toxicity but little added efficacy (206,207,208,209,210), except perhaps for blacks. The incidence of clinical adverse reactions to MMF seems to be high. However, most of the large, randomized, controlled trials compared MMF with azathioprine rather than placebo, making it difficult to determine the true incidence of adverse effects attributable to MMF (207,208,209,210). The principle nonimmune toxicities of MMF are gastrointestinal and hematologic. At the 2 g/d dose, diarrhea was reported in 13% (206), 28% (208), and 35% (210) of cases in three large, randomized, controlled trials. In these trials, with 2 g/d MMF, abdominal pain occurred in 12% (206), 26% (208), and 30% (210) of cases, respectively. Vomiting occurred in 2% (206), 12% (208), and 16% (210), leukopenia occurred in 11% (206), 19% (208), and 20% (210), anemia occurred in 4% (206), 15% (208), and 16% (210), and thrombocytopenia occurred in 4.2% (206), 9% (208), and 8.8% (210) of cases, respectively. The principal short-term, nonspecific, immune-related toxicity was an increased incidence of infection. For example, at the dose of 2 g/d, CMV infection occurred in 12 to 16% of cases (206,207,208,210). In a case-control study, the severity of CMV infection was higher with MMF, compared with azathioprine (211). Similarly, the initiation of MMF administration was associated with an increase in hepatitis C virus (HCV) viremia (212).

Consequences. Although MMF seems to be efficacious in preventing acute rejection, many patients experience adverse effects. Treatment of gastrointestinal, hematologic, or other adverse effects is generally with dose reductions, dosing interval changes, or discontinuation of MMF administration. For example, in one large, randomized controlled trial, the 2 g/d dose of MMF was reduced because of leukopenia in 15% of cases and MMF administration was discontinued because of leukopenia in 2 to 3% of cases (208).

Rationale. Techniques for determining mycophenolic acid (MPA) blood concentrations seem to be reliable (213,214,215,216). MMF is a prodrug that is highly bioavailable, with 94% absorption after an oral dose (217). MMF is rapidly hydrolyzed to MPA, the active compound, which then undergoes enterohepatic recirculation (218,219). This latter property may complicate the interpretation of trough blood levels. MPA is highly protein-bound (214), and levels of MPA are correlated with serum albumin levels (219). MPA is metabolized to MPA glucuronide (MPAG). Although MPAG is not thought to have pharmacologic activity, it may contribute to toxicity. Because MPAG is renally excreted, patients with reduced renal function may have high levels of MPAG; this may help explain an increased incidence of gastrointestinal toxicity among patients with reduced renal function (220).

Pharmacokinetic studies of MPA demonstrated substantial intra- and interindividual variations for both adults and children (218,219,221,222,223,224). Levels in children may be higher than levels in adults treated with a comparable dose (600 mg/m² twice daily) (224). The total AUC values for MPA were similar with oral or intravenous administration, with intraindividual and interindividual coefficients of variation of 25% and 10%, respectively (218). A number of factors affect MPA levels. The maximal concentration of MPA and the AUC value determined immediately after transplantation were only 30 to 50% of those measured for patients 3 mo after transplantation (218,222). However, free levels remained unchanged, suggesting that the increase in MPA levels was largely attributable to increased levels of MPA bound to plasma proteins (222). The pharmacologic activity of MMF is thought to be correlated with free and not bound MPA levels. Therefore, reductions in plasma protein levels (as may occur in liver disease, for example) may increase the free levels of MPA without altering whole-blood concentrations. The MPA maximal concentration and AUC may be reduced by concomitant administration of antacids containing magnesium and aluminum (225) and by cholestyramine (218). Tacrolimus seems to increase MPA levels (181,205,226). However, in most (205,227) but not all studies (228), CsA reduced the blood concentrations of MPA.

The variability in MPA levels suggests that measurements of plasma MPA levels could be beneficial. Indeed, blood levels of MPA have been demonstrated to correlate with the incidence of acute rejection in adults and children (204,213,223,229). A recent, blinded, randomized trial demonstrated that targeting of different MPA AUC values resulted in different incidences of acute rejection and toxicity (230). MPA trough levels have also been found to be correlated with hemoglobin levels (231). It has been suggested that MMF levels should be monitored in children when MMF is used as rescue therapy after acute rejection (232). Although the case for monitoring of MMF blood levels seems to be compelling, additional data are needed to establish whether monitoring will improve outcomes.

Monitoring the pharmacodynamic activity of MMF, by monitoring the activity of the enzyme inosine monophosphate dehydrogenase, which is inhibited by MPA, has also been suggested as a potential method to monitor the efficacy of MMF and adjust its dose (233). Because this is a key enzyme in the de novo biosynthesis of purines, measurement of its activity could be used to effectively measure the desired pharmacodynamic effect of MMF, thereby offering a rational approach to the monitoring of MMF therapy (233,234). Additional studies are needed to determine whether pharmacodynamic monitoring improves outcomes.

Efficacy and Toxicity of Azathioprine (Table 10)
Incidence. There have been few randomized controlled trials demonstrating the efficacy of azathioprine, but indirect evidence suggests that azathioprine reduces the incidence of acute rejection (235,236). Leukopenia occurs in approximately 35% of renal transplant cases when initial azathioprine doses are at least 3 mg/kg daily, but white blood cell counts are usually not lower than 2500/µl (237). When azathioprine doses are initiated at 1 to 2 mg/kg per d, leukopenia occurs in approximately 10 to 25% of cases (237,238). The presence or absence of a spleen influences the incidence of leukopenia (238). Thrombocytopenia occurs in approximately 13% of cases and, although macrocytic blood cells are a common finding (239,240), anemia is not frequent. Hepatoxicity, including rare cases of veno-occlusive disease, occurs in a small percentage of patients, most commonly in the first 6 mo after transplantation (241,242). Gastrointestinal toxicity (usually nausea and vomiting) occurs in approximately 10% of cases, but rare patients develop severe gastrointestinal hypersensitivity, which mimics infectious gastroenteritis (243,244).

Consequences. Hematologic and gastrointestinal toxicities are usually dose-related and respond to dose reductions (245). However, the gastrointestinal hypersensitivity reaction or hepatic veno-occlusive disease usually necessitates the withdrawal of azathioprine. Some early uncontrolled reports linked azathipprine use with pancreatitis in renal transplant recipients, but this association was not confirmed in a randomized controlled trial (246).

Rationale. Azathioprine is metabolized to 6-mercaptopurine. The metabolites of azathioprine can be readily measured by HPLC (247). The bioavailability of azathioprine is low and highly variable (range, 5 to 24%) (248). Both azathioprine and 6-mercaptopurine are rapidly cleared, so that blood levels are low and would not be expected to be correlated with clinical efficacy or toxicity (249). The final metabolic end products, i.e., 6-thioguanine nucleotides, are cleared very slowly (249). The dose of azathioprine is correlated with levels of 6-thioguanine (250). Interestingly, 6-thioguanine nucleotides appear to be concentrated in granulocytes, which suggests that these metabolites could be important in the myelotoxicity of azathioprine (251). In some cross-sectional studies, leukopenia has been observed to be correlated with 6-thioguanine levels (250,252). In one prospective study, 360 patients were randomized to receive high doses of azathioprine (monitored and adjusted on the basis of 6-thioguanine levels) and were compared with patients treated with lower doses of azathioprine (253). Acute rejection episodes were less frequent for the high-dose, monitored group, but leukopenia was also more common. Unfortunately, the trial design did not permit the authors to determine the role of therapeutic monitoring (compared with the higher dose of azathioprine) in altering the efficacy or toxicity of azathioprine (253).

Thiopurine methyltransferase is an important enzyme in the metabolism of 6-thioguanine nucleotides, and this enzyme exhibits considerable genetic variability in the general population (254,255). In theory, individuals with very low levels of this enzyme may be susceptible to myelotoxicity because of high levels of 6-guanine nucleotides, and individuals with low levels may be more susceptible to acute rejection (255,256,257). Therefore, measurements of the activity of this enzyme could theoretically predict the efficacy and toxicity of azathioprine. Thiopurine methyltransferase activity increases after transplantation in some patients, and in some studies the incidence of acute rejection was inversely correlated with the increase in thiopurine methyltransferase activity (255,258). In another study, however, the authors failed to observe the predicted correlation between baseline thiopurine methyltransferase activity and 6-thioguanine levels, and neither enzyme activity nor substrate levels predicted the incidence of acute rejection among 82 patients (259). Negative results were also reported by others (260). Therefore, there is currently insufficient evidence to suggest that therapeutic drug monitoring would improve the efficacy or toxicity of azathioprine.

Monitoring hematologic toxicity using blood counts, monitoring hepatic toxicity using liver enzyme and serum bilirubin levels (261), and assessing other toxicities using clinical symptoms are the only practical means for minimizing azathioprine toxicity. Because myelosuppression can occur with azathioprine at any time after transplantation, monitoring of white blood cell counts must be continued throughout the course of treatment. Allopurinol can alter azathioprine metabolism and thereby cause life-threatening bone marrow suppression, and it is generally best to avoid using allopurinol and azathioprine in combination unless it is absolutely necessary. If allopurinol is used for azathioprine-treated patients, the dose of azathioprine should be reduced. The amount of dose reduction may be dictated by the dose of allopurinol. Thus, the dose of azathioprine may need to be drastically reduced, e.g., to 25 mg/d, when the dose of allopurinol is 300 mg/d, whereas less marked reductions of the azathioprine dose may be appropriate when the dose of allopurinol is 100 mg/d. Co-trimoxazole has been reported to cause leukopenia in azathioprine-treated patients (257,262). Despite newer drug regimens, it seems likely that azathioprine will continue to be used for many years for patients with successful grafts who are already receiving azathioprine, for patients who are intolerant of other agents, for some patients because of economic reasons, and for pregnant patients.

Efficacy and Toxicity of Corticosteroids (Table 11)
Incidence. Clinical signs of corticosteroid toxicity, which are observed relatively soon after the initiation of prednisone treatment, include cushingoid facial and body habitus changes, changes in mood or mentation, acne and other skin changes, hypertension, peptic ulcer disease, and myopathy. Bone and eye toxicities generally occur later. Avascular necrosis of the hips exhibits a peak incidence toward the end of the first year and in the second year after transplantation, whereas osteoporosis, cataracts, glaucoma, and growth retardation of children can occur even later.

The incidence of many steroid-induced complications is influenced by the concomitant use of other immunosuppressive agents, such as CsA and tacrolimus. The overall incidence of hypertension among prednisone-treated renal transplant recipients is 75 to 85% (263,264,265,266,267,268,269,270,271). Hypercholesterolemia occurs by 3 to 12 mo in 38 to 68% of patients treated with prednisone and CsA (264,267,269,272,273). The incidence of hypercholesterolemia may be as low as 13% among patients in stable condition for whom the dose of prednisone can be reduced to 10 mg/d in the late posttransplant period (272). Hypercholesterolemia was observed for only 7 to 30% of patients treated with tacrolimus and prednisone (273,274). Hypertriglyceridemia is less common than hypercholesterolemia and occurs in 15 to 35% of patients treated with prednisone and CsA (264,272). HDL cholesterol levels are generally normal in prednisonetreated renal transplant recipients and decrease as the dose of prednisone is reduced.

The incidence of new-onset diabetes mellitus is affected both by the prednisone dose and by the concomitant diabetogenic effects of CsA (159,275,276) or tacrolimus (159,277,278,279). The overall incidence of new-onset diabetes mellitus among renal transplant recipients treated with prednisone and CsA, with or without azathioprine, is generally reported to be 3 to 17% (263,265,266,269,270,275,280,281,282), with black (283), older (282), and more obese (282) patients experiencing higher incidences of posttransplant diabetes mellitus. The incidence of new-onset diabetes mellitus among tacrolimustreated renal transplant recipients has been reported to be higher than that among CsA-treated patients, ranging from 10 to 20% (159,284,285).

Bone complications are common among prednisone-treated patients, and bone mineral density is influenced by many factors (286). Avascular necrosis, however, occurs in only 1.1 to 5.5% of cases (263,266,269,270). The incidence of symptomatic cataracts has been reported to be 9 to 21% (265,269,270). As detected in annual systematic screening, cataracts occurred in 21 of 38 patients (53%) treated with high doses of corticosteroids, 33 of 117 patients (28%) treated with low doses of steroids, and 1 of 16 patients (6%) who received no steroids (287).

The growth of children is adversely affected by the daily administration of corticosteroids, even in very small doses. The onset of puberty is also delayed in children who receive corticosteroids, further delaying growth. Growth rates are higher for children who receive corticosteroids on an alternate-day schedule (288,289,290), but <30% of children are receiving alternate-day prednisone treatments by 4 yr after transplantation (4).

It is difficult to accurately assess the incidence of adverse effects caused by corticosteroids, because most renal transplant recipients receive at least some corticosteroid therapy. However, several randomized, controlled, steroid-withdrawal trials compared the incidences of adverse effects (263,264,265,266,267,268,269). Patients in the withdrawal arms of these trials usually received steroids early after transplantation, and many who experienced failure of withdrawal attempts were returned to steroid therapy. Nevertheless, these trials provide a clearer view of the incidence of adverse effects attributable to long-term corticosteroid use. Differences in the incidences of hypertension (variously defined) were generally small, e.g., 2 to 6% lower in the withdrawal groups (265,266,267). Hypercholesterolemia was 10 to 38% lower in the withdrawal groups (264,267,269); however, HDL cholesterol levels were proportionately reduced. The incidence of diabetes mellitus was 2 to 10% lower in the withdrawal groups (263,265,266,267,269). Cataracts were 10 to 20% less frequent in the withdrawal groups (265,269), and the incidence of avascular necrosis was 0.1 to 3.0% lower in the corticosteroid withdrawal groups (263,266,269). The best growth rates for children are observed for those from whom corticosteroids have been withdrawn; these children frequently grow at a normal rate (291,292).

A systematic review of controlled and uncontrolled trials examined the adverse effects of steroids in steroid withdrawal or steroid avoidance CsA-based immunosuppressive regimens (293). Included were 10 trials that examined either hypertension, diabetes mellitus, fractures, avascular necrosis of the hips, or cataracts (263,264,265,266,269,286,287,294,295,296). The estimated incidences of late, steroid-related, side effects were as follows: hypertension, 15%; diabetes mellitus, 10%; fractures, 2%/yr; avascular necrosis of the hip, 8%; cataracts, 22% (293).

Consequences. A number of randomized controlled trials (263,264,265,266,267,268,269,297,298,299,300) and a meta-analysis (301) have examined the feasibility of prednisone withdrawal or avoidance after renal transplantation. However, in most of those trials the incidence of acute rejection was increased in the steroid withdrawal/avoidance group, and in at least one large study graft survival was also reduced in long-term follow-up monitoring (266). In uncontrolled trials involving children, approximately 50% of patients underwent prednisone withdrawal without immediate acute rejection, but for almost 50% of that group steroid administration was reinitiated because of late rejections (302,303). As a result, the long-term use of corticosteroids continues to be a mainstay of immunosuppression for both children and adults. Whether newer immunosuppressive agents, such as tacrolimus or MMF, could allow prednisone to be withdrawn is being actively investigated. It is currently necessary to monitor the adverse effects of corticosteroid therapy.

Rationale. Blood levels of prednisone and its major metabolite prednisolone can be measured with HPLC. Endogenous cortisol levels, which can also be measured with HPLC, may reflect the degree of adrenal suppression. However, the biologic effects of the adrenal corticosteroids are diverse, and no laboratory measurements have been proven to be reliable in monitoring either efficacy or toxicity. Prednisone and methylprednisolone pharmacokinetics exhibit considerable inter- and intraindividual variability (304,305,306,307,308,309). The bioavailabilities of prednisone and prednisolone exceed 85% (310). Clearance tends to decrease with time after transplantation (307). Clearance also varies with race. Black subjects exhibit lower methylprednisolone clearance rates, compared with white subjects, and this may predispose black patients to a higher incidence of steroid-related side effects, such as diabetes mellitus (311,312). The metabolic clearance of prednisone is inhibited by CsA (313), oral contraceptives (314,315), and ketoconazole (316,317). Indeed, ketoconazole has been demonstrated to cause increased weight gain and more bone loss among prednisone-treated patients (318). Conversely, the clearance of prednisone may be increased by the concomitant use of drugs such as phenytoin, phenobarbital, or rifampin (319).

There are virtually no studies correlating the rate of acute or chronic allograft nephropathy with pharmacokinetic or pharmacodynamic parameters for corticosteroids. A few studies attempted to correlate pharmacokinetic parameters with the adverse effects of steroids. However, the results of those studies were variable. For example, some (320,321), but not all investigators (322) were able to correlate the development of cushingoid features with differences in prednisone pharmacokinetic parameters. Other investigators also attempted to use cortisol levels as a pharmacodynamic measure of corticosteroid activity. In one study, for example, it was demonstrated that the degree of cortisol suppression by methylprednisolone was greater in older men, compared with younger men, and this greater age-related cortisol suppression was correlated with a greater methylprednisolone AUC value (323). As is the case for prednisone and prednisolone pharmacokinetics, there is considerable variability in the pharmacodynamic responses to exogenous steroids, as measured by cortisol levels (324). It has been suggested that single timed measurements of prednisone blood levels could be used to guide therapy (305,306), but there are no clinical data to suggest that these measurements would reliably predict efficacy or toxicity.

	IV. Cardiovascular Disease

Top Abstract Introduction Scope Intended Users Materials and Methods Organizational Scheme I. Frequency and Timing... II. Graft Function III. Immunosuppressive... IV. Cardiovascular Disease V. Bone and Bone... VI. Nutrition and Metabolism VII. Cancers VIII. Infections Endorsements References

 
Cardiovascular Disease (Table 12)
Definition. CVD is defined as ischemic heart disease, cerebral vascular disease, or peripheral vascular disease.

Incidence. In one study, the prevalence of ischemic heart disease (detected by routine clinical measures) was 9.5% at the time of transplantation (325). The incidence of new ischemic heart disease events in 46 ± 36 mo of follow-up monitoring after transplantation was 11.1% among patients without histories of prior ischemic heart disease and 15.1% among all patients. The prevalence of cerebral vascular disease was 3.7% at the time of transplantation. The incidence of cerebral vascular disease events was 6.0% among patients without prior cerebral vascular disease and 7.3% among all patients. The prevalence of all CVD types at the time of transplantation was 12.9%. The incidence of new events was 15.8% among patients without prior CVD and 21.3% among all patients, with or without prior CVD (325). The incidence of CVD was approximately fivefold greater than predicted from Framingham Heart Study data for patients of comparable age and gender (325). In a more recent study, the cumulative (actuarial) incidence of primary and secondary ischemic heart disease events was 23% by 15 yr after transplantation (326). The cumulative incidence of cerebral vascular disease was 15% by 15 yr. The cumulative incidence of peripheral vascular disease was also 15% by 15 yr after transplantation (326). In another cross-sectional study (n = 406), the incidence of ischemic heart disease was 14% (24% for diabetic patients and 12% for nondiabetic patients), that of cerebral vascular disease was 3%, and that of peripheral vascular disease was 4% (327). Compared with the general population, odds ratios for angina pectoris were 12 and 16 for men and women 40 to 49 yr of age, 6 and 4 for men and women 50 to 59 yr of age, and 3 and 4 for men and women 60 to 69 yr of age, respectively (327). Similar data have been reported by others (325,326,328,329,330,331,332,333).

Consequences. CVD is a major cause of morbidity and death among renal transplant recipients (334,335). Recently, the National Kidney Foundation convened a task force on CVD and reviewed the available evidence linking risk factors to CVD in patients with renal disease, including renal transplant recipients (336). The task force concluded that renal transplant recipients were at high risk for CVD and that a number of potentially modifiable risk factors could be targeted for intervention (336).

Rationale. There are two possible reasons to screen for CVD, i.e., (1) detection of symptomatic CVD could lead to the relief of symptoms, and (2) detection of asymptomatic disease could lead to the prevention of CVD-related morbidity and death. Routine history assessments and physical examinations seem to be appropriate tools for detecting symptomatic CVD for the relief of symptoms; however, the sensitivity and specificity of routine history assessments and physical examinations to detect CVD have never been rigorously studied.

Few studies have defined the sensitivity and specificity of screening tests for ischemic heart disease in renal transplant recipients. However, studies of the general population suggest that the sensitivity and specificity of exercise electrocardiography (ECG), exercise thallium testing, dipyridamole thallium (or sestamibi) testing, adenosine thallium (or sestamibi) testing, exercise echocardiography, dobutamine echocardiography, and dipyridamole echocardiography are roughly 70 to 80% and 85 to 95%, respectively. Therefore, these tests exhibit only fair sensitivity and specificity. A number of new tests have recently been developed to detect coronary atherosclerosis, e.g., magnetic resonance imaging (MRI), electron-beam computed tomography (CT), and contrast-enhanced CT. However, inadequate data are available to allow the recommendation of any of these newer tests for the screening of renal transplant recipients. In general, the positive and negative predictive values of a particular screening test are dependent on the underlying prevalence of the disease. Therefore, the high prevalence of ischemic heart disease after renal transplantation should theoretically increase the predictive value of screening tests for transplant recipients, compared with the general population. However, these predictive values may still be too low to make screening tests cost-effective in many cases. The United States Preventive Services Task Force concluded that there is insufficient evidence to recommend for or against screening middle-age and older men and women for asymptomatic coronary artery disease with resting ECG, ambulatory ECG, or exercise ECG (category C evidence) (337).

Even fewer studies are available for assessment of the utility of screening tests for cerebral vascular disease and peripheral vascular disease after renal transplantation. Possible screening tests for cerebral vascular disease include carotid artery auscultation during physical examinations, as well as Doppler flow and ultrasonographic examinations of the carotid arteries. The United States Preventive Services Task Force concluded that there is insufficient evidence to recommend for or against screening asymptomatic individuals for carotid artery stenosis, using physical examinations or carotid artery ultrasonography (category C evidence) (338). However, they suggested that a recommendation could be made on the basis of other grounds for high-risk patients, provided that the quality of available vascular surgical care was high. Possible screening methods for peripheral vascular disease could include history assessments and physical examinations, segmental arterial BP determinations (usually using Doppler ultrasonography), and cutaneous oxygen saturation testing. However, the role, if any, for screening of asymptomatic transplant recipients with any of these tests is unclear. The United States Preventive Services Task Force recommended against screening for peripheral vascular disease in asymptomatic patients in the general population (category D evidence) (339).

Pretransplant CVD is an important risk factor for CVD after renal transplantation (326,340). Therefore, CVD screening should be part of the pretransplant evaluation. Pretransplant screening for CVD has been covered in other guidelines (1). Any CVD screening strategy for the posttransplant period should take into account the results of pretransplant screening. For example, patients who have already been identified as being at high risk for posttransplant CVD, as a result of positive pretransplant screening, may benefit from additional periodic screening after transplantation more than do patients who exhibited negative pretransplant screening results.

Preoperative Screening for CVD (Table 13)
Definition. CVD was defined as ischemic heart disease, cerebral vascular disease, or peripheral vascular disease.

Consequences. The risk for ischemic heart disease events and stroke is increased by surgery. Because the risk for CVD is high among renal transplant recipients, it is likely that the risk of surgery for renal transplant recipients is even greater than that for the general population.

Rationale. Guidelines have been developed by the American College of Physicians for assessment and management of the perioperative risk for coronary artery disease (341,342). These guidelines are generally applicable to renal transplant recipients who are being considered for surgery. The guidelines classify patients as having low, intermediate, or high risk, on the basis of a modified cardiac risk index. The cardiac risk index assigns points as follows: myocardial infarction <6 mo earlier, 10; myocardial infarction >6 mo earlier, 5; angina with walking one to two level blocks or climbing one flight of stairs or less at a normal pace, 10; inability to perform any physical activity without developing angina, 20; alveolar pulmonary edema within 1 wk, 10; alveolar pulmonary edema ever, 5; suspected critical aortic stenosis, 20; ECG rhythm other than sinus or sinus plus atrial premature beats, 5; more than five premature ventricular contractions, 5; PO₂ of <60 mmHg, PCO₂ of >50 mmHg, potassium level of <3 mM, blood urea nitrogen level of >50 mM, serum creatinine level of >260 µM, or bedridden, 5; age >70 yr, 5; emergency surgery, 10. The points are added and classes are assigned as follows: class I, 0 to 15 points; class II, 20 to 30 points; class III, >30 points.

Class II or III in the modified cardiac risk index system predicts a high risk for perioperative cardiac events (10 to 15%). Low cardiac risk index scores (class I) do not reliably identify patients who have low risks for perioperative cardiac events, and additional information on "low-risk" variables should be collected for these patients. Low risk (<3%) is predicted by the presence of no cardiac risk factors or one cardiac risk factor, as defined by the low-risk variables. These risk factors include age of >70 yr, history of angina, diabetes mellitus, Q waves in electrocardiograms, history of myocardial infarction or ventricular ectopy, ST-segment ischemic abnormalities during resting ECG, hypertension with severe left ventricular hypertrophy, and history of congestive heart failure.

Low-risk patients may proceed directly to surgery without further testing. For patients who are at intermediate risk, consideration should be given to noninvasive cardiac stress testing. Exercise testing is not suitable for many transplant recipients and has not reliably predicted patients at high risk for surgery. Dipyridamole thallium (or sestamibi) imaging and dobutamine stress echocardiography provide useful stratification information for intermediate-risk patients undergoing vascular surgery, especially if the test results are normal. For patients undergoing nonvascular surgery, neither dipyridamole thallium imaging nor dobutamine echocardiography reliably predicts adverse perioperative cardiac events, as indicated in studies of the general population. However, because the probability of CVD is high after renal transplantation, it is possible that these tests are more reliable for renal transplant recipients than for the general population.

Asymptomatic patients who have recently undergone coronary angiography can be considered at low risk and can proceed with surgery. Patients with unstable angina should undergo cardiac catheterization and should be treated according to usual clinical standards. For high-risk patients who are not candidates for revascularization, consideration should be given to canceling, or at least postponing, surgery if the risk posed by not proceeding with surgery is lower than the perceived cardiac risk. The role of coronary revascularization in preventing perioperative coronary events has not been adequately studied. In deciding whether intermediate- or high-risk patients should undergo revascularization, the risk of angiography and the risk of the revascularization procedure(s) should be taken into account. Of course, if revascularization is clinically indicated on the basis of probable long-term outcomes, then this may best be performed before the planned surgery, if the condition of the patient allows this approach to be taken. Finally, studies have suggested that perioperative monitoring with a pulmonary artery catheter may be useful for some patients.

There are virtually no data on the risk of noncardiac vascular disease events among patients in the general population who undergo noncardiac surgery. Therefore, screening for asymptomatic carotid artery disease as part of preoperative assessments is probably not warranted. If possible, elective surgery should be postponed for patients who have experienced a stroke or transient ischemic attack within the previous 6 mo. Such patients should be evaluated with carotid artery ultrasonography, and guidelines for the management of carotid artery disease have been developed by the American Heart Association (343,344). Several studies, but no randomized controlled trials, have addressed the risk of noncardiac vascular disease complications in the perioperative period for cardiac surgery. Identified risk factors include a previous stroke, the presence of carotid artery bruit, a history of hypertension, advanced age, diabetes mellitus, and the severity of carotid artery stenosis in ultrasonograms (345,346,347). Some investigators have advocated simultaneous carotid endarterectomy and coronary artery bypass grafting (348,349,350,351,352), whereas others have advocated a more conservative approach (353,354). The American Heart Association guidelines categorize simultaneous bypass grafting and endarterectomy as an uncertain indication (344). In the absence of data, it is reasonable to follow the approach suggested in the American Heart Association guidelines.

Aspirin Prophylaxis for CVD (Table 14)
Prophylaxis. Aspirin is safe. It is an effective antiplatelet agent, and it is inexpensive.

Rationale. The effectiveness of aspirin prophylaxis in preventing CVD has been extensively studied in nontransplant patients. Long-term aspirin use has been demonstrated to reduce CVD events and overall mortality rates among patients with histories of myocardial infarction, stroke, or transient ischemic attacks or other evidence of CVD (355). On the strength of the evidence, a number of guidelines have recommended the prophylactic use of aspirin for patients with CVD (355,356,357,358,359). A low dose of aspirin (75 mg) appears to be as effective as higher doses (355,360). Other antiplatelet drugs do not appear to be more beneficial than aspirin (355). Aspirin may increase the risk of hemorrhagic stroke, but the reduction in myocardial infarction and ischemic stroke rates appears to outweigh this risk for the general population (355,361).

Some studies indicate that aspirin may also be effective as primary prevention, i.e., when used for patients without known CVD (355). However, the strength of the evidence for the use of aspirin for patients without CVD is less certain. The United States Preventive Services Task Force concluded that there was no evidence for or against the use of aspirin for primary prevention (362). The American Diabetes Association recommends the use of aspirin for type I and type II diabetic patients who have additional risk factors for CVD (358).

Hypertension (Table 15)
Definition. For adults 18 yr of age who are not taking antihypertensive medications, BP values of 140 mmHg (systolic) or 90 mmHg (diastolic) (based on the average of two or more readings) are considered hypertension. BP values of <120 mmHg (systolic) and <80 mmHg (diastolic) are considered to be optimal (363).

Incidence. The prevalence of hypertension varies with the type of immunosuppressive medications, the time after transplantation, and other factors. Among patients treated with prednisone and azathioprine, the prevalence of hypertension (variously defined) was 53% (364), 42% (365), 60% (366), and 72% (367) at 1 yr, 38% (365) and 58% (368) at 2 yr, and 46% (364), 58% (366), and 71% (367) at 5 yr after transplantation. Among patients treated with CsA, the prevalence of hypertension was 63% (365), 64% (367), 66% (369), 71% (366), and 78% (370) at 1 yr, 61% (365) and 73% (355) at 2 yr, 77% (371) at a mean of 3.75 yr, and 70% (366), 83% (367), and 85% (369) at 5 yr. Among 29,751 cadaveric transplant recipients at 1 yr, 44.5% were normotensive (systolic BP of <140 mmHg), 37.4% exhibited systolic BP of 140 to 159 mmHg, 13.9% exhibited systolic BP of 160 to 179 mmHg, and 4.2% exhibited systolic BP of 180 mmHg. Among 14,351 cadaveric transplant recipients at 5 yr, 46.1% were normotensive (systolic BP of <140 mmHg), 37.5% exhibited systolic BP of 140 to 159 mmHg, 12.7% exhibited systolic BP of 170 to 179 mmHg, and 3.7% exhibited systolic BP of 180 mmHg (372). Among children, the prevalence of hypertension (defined as the use of antihypertensive medication) was 79% at 1 mo after transplantation; the prevalence decreased to 59% at 24 mo (373).

Common causes of posttransplant hypertension include (1) allograft dysfunction resulting from acute rejection or chronic allograft nephropathy, (2) treatment with corticosteroids or calcineurin inhibitors, (3) diseased native kidneys, (4) renal artery stenosis, and (5) essential hypertension (364,374). The incidence of radiographically evident renal artery stenosis varies substantially and depends on which patients are selected for screening. Among adult transplant recipients, the incidence of renal artery stenosis was 2.0 to 6.6% in most studies (375,376,377,378). In one study in which angiography was routinely performed for hypertensive patients, the incidence was 11.6% (379). The incidence may be higher among cadaveric versus living donor transplants (375). Among children, the incidence has been reported to be as high as 15% (375,380).

Consequences. Many studies in the general population have demonstrated that hypertension causes CVD and that its detection and treatment can reduce morbidity and mortality rates (359,363,381,382). Hypertension is also thought to contribute to renal disease in the general population (383,384,385). There is little reason to think that hypertension would not have similar adverse consequences for renal transplant recipients (386). An investigation of 29,751 cadaveric renal transplant recipients found that BP was a strong predictor of subsequent graft survival, whether or not death with a functioning graft was censored (372). Single-center studies have also linked adverse outcomes and posttransplant CVD to hypertension (327,366,387,388,389,390). However, there have been no large, randomized, controlled trials demonstrating that BP reduction improves outcomes after renal transplantation. Nevertheless, interventional trials have established the benefits of BP reduction in the general population, and BP is correlated with adverse outcomes after transplantation. Therefore, it is unlikely that trials comparing treated and untreated hypertension will be conducted among transplant recipients, because of ethical concerns.

Rationale. Standard measurement of BP in the clinic setting, as described by the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure (363), is probably most cost-effective. As a minimum, BP should be checked during each office visit, using a mercury sphygmomanometer or recently calibrated aneroid manometer. Care should be taken to use BP cuffs of the appropriate size for children, because inaccurate results are obtained with cuffs that are either too large or too small. Automated methods of BP measurement may be necessary to obtain accurate results for infants. Self-measurement of BP provides valuable information. It helps to (1) distinguish sustained hypertension from "white-coat hypertension" (elevated BP in the clinic but not at other times), (2) assess responses to treatment, (3) improve adherence to treatment, and (4) reduce costs (363).

Ambulatory BP monitoring may also help distinguish white-coat hypertension. In addition, ambulatory BP monitoring may help detect nocturnal hypertension. Studies demonstrated that nocturnal hypertension is more common among dialysis patients than in the general population and that transplantation reduces the prevalence of nocturnal hypertension (391, 392). However, other studies found that the prevalence of nocturnal hypertension is increased among transplant recipients, compared with the general population (393), and that CsA may contribute to posttransplant nocturnal hypertension (394). Few studies of the general population (363) and no studies of transplant recipients have investigated whether ambulatory BP monitoring predicts outcomes better than casual BP measurements. However, mean 24-h ambulatory BP values were found to be correlated more closely with left ventricular mass than were casual BP measurements in small studies of both pediatric and adult renal transplant recipients (395, 396).

The prevalence of hypertension after renal transplantation is very high. Data from many observational studies and interventional trials have firmly established that hypertension contributes to CVD in the general population. Although there have been no large interventional trials proving that BP reductions are beneficial for renal transplant recipients, there are no compelling reasons to think that the relationship between hypertension and CVD would be different among renal transplant recipients, compared with the general population. Indeed, observational studies have linked hypertension to poor outcomes after renal transplantation. Guidelines for the treatment of hypertension in the general population are available (258, 363, 381, 382, 397,398,399). The National Kidney Foundation Task Force on CVD recommended that the goal for therapy should probably be 135/85 mmHg for renal transplant recipients without proteinuria and should possibly be 125/75 mmHg for patients with proteinuria (386).

The specific treatment of posttransplant hypertension is not addressed in these guidelines. Most standard therapies have been demonstrated to be safe and effective after renal transplantation; however, there are a number of management issues that are unique to transplant recipients and are probably best addressed by experienced transplant physicians. For example, transplant recipients may be more prone to decreased renal function resulting from diuretic use than are hypertensive patients in the general population. Some calcium channel blockers alter the levels of calcineurin inhibitors. Patients may occasionally develop decreased renal function after angiotensin-converting enzyme inhibitor therapy, especially if the patients exhibit renal artery stenosis or chronic allograft nephropathy. Anemia and hyperkalemia may also be associated with the use of angiotensin-converting enzyme inhibitors and angiotensin II receptor antagonists. Angiotensin-converting enzyme inhibitors have been reported to cause graft thrombosis and acute tubular necrosis (400, 401), but these complications are probably rare.

A complete discussion of the screening and treatment of possible secondary causes of hypertension is also beyond the scope of these guidelines; however, when to screen for renal artery stenosis deserves some comment. Although renal artery stenosis may contribute to hypertension in many patients, treatment is invasive (375,376,377, 402, 403). Unless hypertension is difficult to control and/or renal function is compromised, it is probably not necessary to screen for allograft renal artery stenosis. Screening tests include color duplex sonography (376, 404,405,406,407,408,409), captopril renography (376, 404, 410), magnetic resonance angiography (409, 411), and spiral CT (412). Color duplex sonography appears to be the most cost-effective of these tests. The standard method for diagnosis is angiography.

Hyperlipidemia (Table 16)
Definition. According to the National Cholesterol Education Program (NCEP) guidelines, low-risk total and LDL cholesterol levels are <200 mg/dl and <130 mg/dl, respectively (413). High-risk total and LDL cholesterol levels are >240 mg/dl and >160 mg/dl, respectively. Low HDL levels are considered to be <35 mg/dl, whereas high fasting triglyceride levels are >200 mg/dl (413).

Incidence. The prevalence of hyperlipidemia varies with the type of immunosuppressive medications used and other factors. In a report by the National Kidney Foundation Task Force on CVD, the prevalence of hyperlipidemia was estimated by combining the results of studies reporting the proportions of patients with lipoprotein levels above defined levels (414). In the combined results from five studies, 63% of 549 patients exhibited total cholesterol levels of >240 mg/dl (415,416,417,418,419). LDL cholesterol levels were >130 mg/dl for 60% of 769 patients (415, 420, 421). In contrast, HDL cholesterol levels were <35 mg/dl for only 12% of 777 patients (415, 416, 421, 422). Triglyceride levels were >200 mg/dl for 36% of 1309 patients (415, 416, 418, 422,423,424). Lipoprotein(a) measurements may help to identify patients who are at increased risk for CVD, although no therapy has been proven to be effective in reducing lipoprotein(a) levels. Lipoprotein(a) levels were >30 mg/dl for 23% of 468 patients (419, 420, 425,426,427). Information regarding lipid levels among pediatric renal transplant recipients is sparse. The North American Pediatric Renal Transplant Cooperative Study reviewed results for patients who were treated using a common immunosuppression protocol. Those investigators reported that, at 1 yr after transplantation, children did not exhibit elevated levels of total and VLDL cholesterol, compared with normal control subjects, and cholesterol levels (mean 213 mg/dl) were not high enough to require lipid-lowering agents (428). In contrast, a recent, long-term, retrospective study of 62 pediatric renal transplant recipients indicated that 52% of the patients exhibited elevated serum total cholesterol levels and 46% exhibited high LDL cholesterol levels (429). Additional studies are clearly warranted.

Consequences. It is generally accepted that total and LDL cholesterol cause ischemic heart disease in the general population. This conclusion is based on both observational studies and interventional trials (430). Increasing amounts of evidence implicate cholesterol in the pathogenesis of other CVD types as well. There have been no large-scale interventional trials among renal transplant recipients, but observational studies identified correlations between hyperlipidemia and the high incidence of CVD after transplantation (325,326,327, 431, 432). Correlations between hyperlipidemia and chronic renal allograft dysfunction have also been reported. For example, Dimény et al. (433) observed that, for 151 patients, pretransplant cholesterol and triglyceride levels were correlated with chronic allograft nephropathy in biopsies obtained 6 mo after renal transplantation. In addition, cholesterol and triglyceride levels measured 4 yr after transplantation were associated with biopsy findings typical of chronic allograft nephropathy (434). Isoniemi et al. (435) reported that cholesterol and triglyceride levels measured 2 yr after transplantation predicted chronic allograft nephropathy at 4 yr after transplantation. Massy et al. (10) found the triglyceride level to be an independent risk factor for graft loss resulting from chronic allograft nephropathy. In a recent retrospective study, the serum cholesterol level was an independent predictor of death and death-censored allograft failure (436). Therefore, the clinical evidence is very strong, although not conclusive, that hyperlipidemia contributes to the incidence of CVD, death, and possibly death-censored allograft failure after renal transplantation (414).

Rationale. The National Kidney Foundation Task Force on CVD concluded that the high incidence of CVD, the strong association between hyperlipidemia and CVD, and the availability of effective lipid-lowering therapies make screening for hyperlipidemia after renal transplantation worthwhile (414). The NCEP recommends screening using nonfasting total and HDL cholesterol levels (413) and obtaining follow-up fasting lipid profiles for individuals with abnormal nonfasting levels. Although this strategy is reasonable, most transplant recipients have blood samples drawn after fasting for other reasons, e.g., blood glucose screening for diabetes mellitus (see below). Therefore, the rationale for screening using nonfasting lipid measurements is less applicable to renal transplant recipients. The most useful screening tests for hyperlipidemia are fasting total cholesterol, LDL, HDL, and triglyceride measurements.

Lipid levels may change rapidly during the first 1 yr after transplantation. Therefore, screening at least twice during the first 1 yr after transplantation seems prudent. Patients with pretransplant hyperlipidemia or patients at risk for hyperlipidemia because of the use of sirolimus may need to undergo screening more often. Screening relatively soon after transplantation may allow patients to receive adequate instruction and to begin lipid-lowering therapy before they return to caregivers who may not be comfortable with initiating lipid-lowering therapy for renal transplant recipients. The immunosuppressive medications, graft function, and proteinuria that influence plasma lipid levels often change during the posttransplant period. Therefore, periodic screening for hyperlipidemia beyond the first posttransplant year is warranted, particularly for individuals who are at high risk for CVD. Treatment should be based on the NCEP guidelines (413). Long-term studies of hyperlipidemia and its treatment in children are needed to provide evidence for recommendations.

Hyperhomocysteinemia (Table 17)
Definition. Fasting plasma total homocysteine (tHcy) levels of >10 µM are generally considered to be abnormal (437, 438). In the third National Health and Nutrition Examination Survey (1991 to 1994), high levels (exceeding gender-specific 95th percentiles) were 11.4 µM for men and 10.4 µM for women (439). What constitutes an abnormal increase in tHcy levels after methionine loading is unclear, but some have considered increases to levels that are >2 SD above the mean to be abnormal (437).

Incidence. Fasting tHcy levels are elevated in patients undergoing hemodialysis and decrease after transplantation (440, 441). However, levels remain elevated, compared with the general population (441,442,443,444,445,446). Levels are inversely correlated with plasma folate and vitamin B6 and B12 levels (443, 447,448,449) and even more so with renal function (high levels with decreased renal function) (441,442,443, 448, 449). Levels also tend to be higher for CsA-treated patients (442, 447), but this association may not be as strong when the reduced renal function attributable to CsA is accounted for (447, 448).

Consequences. The association between tHcy levels and CVD has been demonstrated by retrospective and prospective epidemiologic studies in the general population (437, 450, 451), and the evidence was recently critically reviewed (452). Ischemic heart disease, cerebral vascular disease, and peripheral vascular disease are all more common among nontransplant patients with elevated fasting tHcy levels. Less clear is whether there is an additional association between post-methionine load increases in tHcy levels and CVD among nontransplnat patients. A case-controlled study demonstrated that renal transplant recipients with CVD exhibited higher tHcy levels, compared with renal transplant recipients without CVD (453). Cross-sectional studies of renal transplant recipients reported a similar association between tHcy levels and CVD (447, 448, 454). Recently, a prospective observational study also found the tHcy level to be an independent risk factor for CVD (446). Although the association between tHcy levels and CVD in the general population is very strong and preliminary data suggest that a similar association might exist for renal transplant recipients, there are no interventional trials proving that lowering tHcy levels decreases the incidence of CVD.

Rationale. tHcy can be safely and effectively reduced by folate, vitamin B6, and vitamin B12 (455, 456). Because CVD is associated with increased tHcy levels, therapy could be beneficial. However, the lack of well-designed interventional trials for any patient population makes it difficult to recommend routine screening.

Posttransplant Diabetes Mellitus (Table 18)
Incidence. The incidence of posttransplant diabetes mellitus is high but varies according to how diabetes mellitus is defined, the type of immunosuppressive medications used, age, and other factors. Posttransplant diabetes mellitus is caused primarily by glucocorticoids, CsA, and tacrolimus. In large prospective and retrospective studies, the incidences of post-transplant diabetes mellitus were 32 of 901 patients (3.6%) (280), 39 of 337 patients (11.6%) (457), 26 of 222 patients (11.7%) (458), and 30 of 167 patients (18.0%) (459). The incidence was reported to be higher among patients treated with CsA and prednisone, compared with azathioprine and prednisone (275, 460); however, other investigators failed to confirm that finding (280). Tacrolimus has also been linked to posttransplant diabetes mellitus (461, 462). Most cases of new onset diabetes mellitus occur within the first few months after transplantation (280, 283, 460). Older individuals (270, 459, 460, 463) and black or Hispanic patients are most susceptible (270, 283, 312, 457, 461). One case-controlled study suggested that patients with adult polycystic kidney disease were more likely to develop posttransplant diabetes mellitus (464).

Consequences. Diabetes mellitus is a major cause of morbidity and death. There is no reason to think that posttransplant diabetes mellitus produces fewer complications than diabetes mellitus in other settings, and at least one study has indicated that the consequences of posttransplant diabetes mellitus are similar to those of pretransplant diabetes mellitus (465). Some studies have reported that patients with posttransplant diabetes mellitus experience decreased allograft survival (282, 460).

Rationale. The most readily available and cost-effective screening test is fasting blood glucose level measurement. Glycosylated hemoglobin level measurement is also proving to be an effective test for screening (466). The oral glucose tolerance test is still the standard test for diagnosing diabetes mellitus but is less than optimal for routine screening because of its cost and inconvenience. The American Diabetes Association recommends screening by measuring fasting plasma glucose levels (467). A plasma glucose level of >126 mg/dl is an indication for diagnostic testing (467). Strict glycemic control decreases retinopathy, nephropathy, neuropathy, and possibly CVD in nontransplant patients with diabetes mellitus (468). By inference, the early detection, prevention, and treatment of diabetes mellitus may reduce the frequency of these complications in renal transplant recipients. It is likely that the screening and early detection of posttransplant diabetes mellitus will help control and prevent the complications of diabetes mellitus and thereby decrease morbidity and mortality rates. The National Kidney Foundation Task Force on CVD concluded that strict control of posttransplant diabetes mellitus in the absence of advanced age or comorbid conditions was warranted (469).

Cigarette Smoking (Table 19)
Incidence. The prevalence of cigarette smoking after renal transplantation is comparable to that found in the general population (470).

Consequences. Retrospective observational studies have linked smoking to CVD (325, 470), decreased patient survival (470,471,472), and graft failure (473).

Rationale. Evidence from studies in the general population strongly suggests that cigarette smoking increases the risk for CVD and malignancies. The morbidity and mortality rates for both of these complications are higher after renal transplantation, compared with the general population. Moreover, some observational studies have linked smoking to adverse outcomes after renal transplantation. Because intervention is often effective, cigarette smoking should be screened for and treated in renal transplant recipients. Guidelines have been developed for smoking cessation (474,475,476).

	V. Bone and Bone Marrow

Top Abstract Introduction Scope Intended Users Materials and Methods Organizational Scheme I. Frequency and Timing... II. Graft Function III. Immunosuppressive... IV. Cardiovascular Disease V. Bone and Bone... VI. Nutrition and Metabolism VII. Cancers VIII. Infections Endorsements References

 
Erythrocytosis (Table 20)
Definition. Erythrocytosis is usually defined as a hemoglobin level of >17 to 18 g/dl or a hematocrit value of >51 to 52%.

Incidence. The reported incidence is variable, but most studies report that it is between 10 and 20%. Specifically, studies have reported that the incidence of erythrocytosis after renal transplantation is 8.1% (477), 9.8% (478), 11% (479), 12.4% (480), 13.5% (481), 17.3% (482), and 21.6% (483). The incidence seems to be somewhat higher in the first 1 yr after transplantation (477).

Consequences. Some retrospective epidemiologic studies suggested that higher hematocrit values are associated with an increased risk of thromboembolic events after renal transplantation (325, 479, 482), but at least one study failed to confirm that finding (484). Among nontransplant patients, epidemiologic studies clearly demonstrated an increased incidence of CVD complications for patients with palycythemia rubra vera (485,486,487,488). Although there have been few large, randomized, controlled trials, interventions that reduce hematocrit values seem to reduce the incidence of CVD events among patients with polycythemia rubra vera (485, 487).

Rationale. The incidence of erythrocytosis among renal transplant recipients seems to be high, and the consequences may be life threatening. In addition, effective therapies are available for patients with marked erythrocytosis. In particular, angiotensin-converting enzyme inhibitors (489,490,491,492,493,494,495,496) or angiotensin II receptor antagonists (497, 498) have been demonstrated to consistently reduce elevated hematocrit values. Theophylline may also be effective (499,500,501). If these therapies are not sufficient, phlebotomy can effectively correct erythrocytosis and help prevent its complications (502). Surgical removal of the native kidneys may also be a treatment option in some cases (503).

Anemia (Table 21)
Definition. Anemia is usually defined as hemoglobin levels of <13 g/dl (hematocrit of 42%) for male patients and <12 g/dl (hematocrit of 37%) for female patients (504). Normal hemoglobin levels in children are age-related and generally are lower than normal levels for adults.

Incidence. The incidence of anemia after renal transplantation is not well documented. However, most authors report that anemia is relatively common early after transplantation. It is also common late after transplantation among patients with chronic graft dysfunction. The incidence was 12% (hematocrit of <33%) in one study (505).

Consequences. Common, potentially reversible causes of anemia include bone marrow suppression resulting from immunosuppressive agents (especially but not exclusively MMF and azathioprine) (506, 507), iron deficiency (505, 508,509,510,511), and the use of angiotensin-converting enzyme inhibitors (512,513,514,515) or angiotensin II receptor antagonists (497, 516). Allograft dysfunction is a frequent contributing factor to anemia early and late after renal transplantation (505, 509, 517). Erythropoietin deficiencies and erythropoietin resistance that are not associated with decreased renal function may occur in some patients (518). The differential diagnosis of anemia also includes many potentially life-threatening disorders that are amenable to treatment and need to be detected as quickly as possible. These include (but are not limited to) gastrointestinal bleeding, malignancies, autoimmune hemolytic anemia (519,520,521), and some rare disorders, such as B19 parvovirus infection (522,523,524). CMV infection is frequently associated with anemia but is usually symptomatic and therefore is not a reason to screen for anemia itself. Folate and vitamin B12 deficiencies may cause or contribute to anemia.

Rationale. Because removal of its underlying cause can often reverse anemia, screening for anemia is particularly important. When there is no reversible underlying cause, subcutaneous treatment with human recombinant erythropoietin is both safe and effective (525,526,527,528,529). Studies in nontransplant patient populations, e.g., patients undergoing hemodialysis, have unequivocally demonstrated that correction of anemia improves the quality of life. At least one study has confirmed that correction of anemia improves the quality of life for renal transplant recipients (528). Although the true incidence of anemia is poorly defined, it appears to be high enough to make screening an effective tool after renal transplantation.

Osteoporosis (Table 22)
Definition. Osteoporosis is defined as bone mineral density >2.5 SD below the young adult mean value (t-score).

Incidence. As many as 60% of kidney transplant recipients treated with corticosteroids may lose sufficient bone mineral density to meet the definition of osteoporosis in the first 18 mo after transplantation (530,531,532,533,534,535,536,537,538,539,540,541,542,543). Bone demineralization may then improve or persist at a slower rate (286,540,544,545,546,547,548,549,550). In a recent cross-sectional study, 55 patients underwent bone densitometric examinations and bone biopsies a mean of 10 yr after transplantation (551). None exhibited decreased axial bone mineral density, but most exhibited decreased femoral neck bone mineral density (551). In biopsies, 46.5% of patients exhibited mixed uremic osteodystrophy, 23.2% exhibited adynamic bone disease, 13.9% exhibited hyperparathyroid disease, and only 16.3% exhibited normal bone (551). Bone fractures are late complications in the current era, occurring in approximately 10% of recipients an average of 8 yr after kidney transplantation (286,552,553,554). In contrast, bone fractures are common early complications among patients with insulin-dependent diabetes mellitus who undergo combined kidney-pancreas transplantation, occurring in approximately 50% of recipients within the first 5 yr after transplantation (555).

Consequences. Bone pain and fractures occur both in vertebral bone and in peripheral bone (286). In a recent retrospective cross-sectional study of 193 renal transplant recipients, 17% developed fractures attributable to osteoporosis after 0.5 to 23 yr of follow-up monitoring (556). Fractures were more common among patients with type 1 diabetes mellitus (40%), compared with nondiabetic patients (11%) (556). The foot has been the most common fracture site in combined kidneypancreas transplant recipients (555). The bone mineral density fracture risk threshold for kidney transplant recipients has not been defined (534). Threshold values in the general population are 0.900 g/cm² for the lumbar sacral spine and 0.600 g/cm² for the femoral neck.

Rationale. Lumbar spine and hip bone mineral density measurements by dual x-ray absorptiometry are readily available in most areas. There are few data on the value of serum biochemical markers of bone metabolism after renal transplantation (557,558). Bone loss after kidney transplantation is predominately a consequence of corticosteroid-induced decreases in bone formation. CsA caused high-turnover bone disease in an experimental animal model (559), but firm clinical evidence of this disorder is lacking (560). Secondary hyperparathyroidism and aluminum bone disease (which is becoming increasingly uncommon) may be contributing factors (533).

Corticosteroid-induced osteoporosis is time- and dose-dependent. Bone demineralization is rapid in the early months after transplantation, when corticosteroid doses are highest (535,536,537,538,539,540,541,542,543). Bone loss may stabilize or persist >2 yr after transplantation (286,540,544,545,546,547,548,549). Effective treatment of corticosteroid-induced bone loss is now available, but there are few randomized controlled trials examining preventive strategies for osteoporosis among renal transplant recipients (561,562).

Secondary Hyperparathyroidism (Table 23)
Incidence. Up to 10 to 20% of recipients may develop hypercalcemia within the first 1 to 2 yr after transplantation (563,564,565,566). The incidence of refractory posttransplant hyperparathyroidism may be decreasing with better pretransplant management of secondary hyperparathyroidism (567).

Consequences. Symptomatic hypercalcemia, renal allograft dysfunction, hypertension, soft-tissue calcification, and bone demineralization attributable to high-turnover bone disease (568,569) may all result from secondary hyperparathyroidism. It is theoretically possible that secondary hyperparathyroidism may contribute to the high incidence of CVD among patients with renal disease.

Rationale. Secondary hyperparathyroidism persists after transplantation, because involution of parathyroid gland hyperplasia is slow and may never occur, despite the re-establishment of renal function (569). Glandular involution and the return to basal parathyroid hormone secretion may take years. Hypercalcemia is usually evident within the first 1 yr after transplantation (563,566). Up to 5% of transplant recipients may require a parathyroidectomy for control of progressive bone demineralization, symptomatic hypercalcemia, or asymptomatic moderate hypercalcemia (total serum calcium levels of 12 to 12.5 mg/dl) (565,566,570,571). Calcium and vitamin D therapy may accelerate the return to basal parathyroid hormone secretion (572,573).

	VI. Nutrition and Metabolism

Top Abstract Introduction Scope Intended Users Materials and Methods Organizational Scheme I. Frequency and Timing... II. Graft Function III. Immunosuppressive... IV. Cardiovascular Disease V. Bone and Bone... VI. Nutrition and Metabolism VII. Cancers VIII. Infections Endorsements References

 
Hypophosphatemia (Table 24)
Definition. Hypophosphatemia is defined by serum phosphorus levels of <2.6 mg/dl (504).

Incidence. Hypophosphatemia is very common in the early weeks after transplantation (574). The decrease in serum phosphorus levels is less among transplant recipients who receive corticosteroid-free immunosuppressive therapy (574). Its prevalence decreases with time after transplantation, but mild hypophosphatemia may persist indefinitely for >50% of recipients (575,576,577,578).

Consequences. Muscle weakness and possibly osteomalacia may occur (578,579,580). Most recipients are asymptomatic.

Rationale. Hypophosphatemia is primarily a result of phosphaturia resulting from persistent hyperparathyroidism, corticosteroid effects on renal phosphate reabsorption (576), and a parathyroid hormone-independent renal "leak" of phosphorus (575,577). Intestinal malabsorption of dietary phosphorus may also occur (581,582). The use of phosphorous-binding antacids may also interfere with the intestinal absorption of phosphorous. A small (n = 28), randomized, controlled trial demonstrated that replacement therapy with orally administered neutral phosphate for 12 wk restored muscular phosphorous content, increased tissue ATP levels, and improved metabolic acidosis (through increased urinary titratable acidity), without adverse effects (583).

Hypomagnesemia (Table 25)
Definition. Hypomagnesemia is defined by serum total magnesium levels of <1.5 mg/dl (504).

Incidence. Up to 25% of long-term CSA-treated recipients manifest hypomagnesemia, which is usually mild (584). The prevalence decreases with time after transplantation, possibly because of decreasing CsA blood levels (585,586,587). Markell et al. (587) found that ionized magnesium levels were low, even when total magnesium levels were near normal, and ionized magnesium levels were especially low in patients with high levels of CsA. Other investigators have confirmed these findings (588). The prevalence of CsA-induced hypomagnesemia may be higher among diabetic patients (588).

Consequences. Muscle weakness, hypokalemia, hypocalcemia, and (rarely) seizures may result from hypomagnesemia. Low serum magnesium levels may be associated with hypertension in CsA-treated patients (586). Low magnesium levels have also been linked to hyperlipidemia. Indeed, magnesium replacement improved the lipid profiles of nontransplant patients with ischemic heart disease in a double-blind, randomized, controlled trial (589). Low magnesium levels have also been linked to hyperlipidemia in renal transplant recipients and, in a small uncontrolled trial, magnesium replacement was demonstrated to reduce elevated total and LDL cholesterol levels (590). It has been suggested that low magnesium levels could contribute to the toxicity of CsA (587).

Rationale. Hypomagnesemia can result from CsA- or tacrolimus-induced renal magnesium leaks (584,591). The use of thiazide diuretics is another common cause of renal magnesium loss. Hypomagnesemia may have several adverse consequences. Oral magnesium supplements are readily available.

Hyperuricemia (Table 26)
Definition. Hyperuricemia is defined by serum uric acid levels of >6.6 mg/dl in women and >8.5 mg/dl in men (504).

Incidence. In a randomized controlled trial, 80% of 131 patients allocated to receive CsA and prednisone developed serum uric acid levels of >8.0 mg/dl, compared with 55% of 115 patients allocated to receive azathioprine, prednisone, and antilymphocyte globulin (592). In a retrospective study, 84% of 129 CsA-treated patients developed hyperuricemia (uric acid levels of >6.7 mg/dl in women and >7.9 mg/dl in men), compared with 30% of 168 azathioprine-treated transplant recipients (593). In another retrospective study, after exclusion of patients with gout, 67% of 211 CsA-treated patients developed hyperuricemia (uric acid levels of >7.0 mg/dl in women and >8.0 mg/dl in men), compared with 32% of 32 patients treated with azathioprine (594). In that study, the incidence of hyperuricemia was higher among patients treated with diuretic agents (594). Hyperuricemia is also common among pediatric renal transplant recipients, being found in 39% of 81 patients at 6 mo after transplantation (595). Most authors have reported that elevated uric acid levels result from decreased renal excretion (596,597,598,599), and serum uric acid levels are inversely correlated with renal function (595,597,598,600).

Consequences. Gout was observed for 6 of 131 patients (4.6%) randomly allocated to receive CsA, compared with 0 of 115 patients randomly allocated to receive azathioprine (592). In a retrospective study, 9 of 129 CsA-treated patients (7%) developed gout, compared with 0 of 168 patients treated with azathioprine (593). In another retrospective study, 25 of 211 CsA-treated patients (11.8%) developed gout, compared with 0 of 32 patients treated with azathioprine (594). In a similar retrospective study, gout developed in 13 of 55 CsA-treated patients (24%), compared with 0 of 23 azathioprine-treated patients. In another retrospective study, 13 of 133 CsA-treated patients (9.9%) developed gout, compared with 0 of 81 patients treated with azathioprine (597). In a randomized trial of CsA withdrawal, gout developed in 9 of 68 patients (13.2%) who continued to receive CsA, compared with 1 of 60 azathioprine-treated patients (1.6%) (601). Hyperuricemia has been linked to adverse outcomes in one (602) but not in another (592) retrospective study. In a case report, renal allograft failure was attributed to hyperuricemia (603). Other authors reported urolithiasis associated with hyperuricemia after renal transplantation (604).

Rationale. Except for gout, the incidence of adverse consequences associated with hyperuricemia appears to be very low. Therefore, it can be argued that serum uric acid levels need only be measured for patients suspected of having gout or other potential complications of hyperuricemia. Indeed, mild to moderate hyperuricemia in asymptomatic patients may not warrant treatment, because gout is uncommon among those patients. Markedly elevated levels of serum uric acid could lead to nephrolithiasis as well as gout. Intervention may be appropriate for asymptomatic patients with markedly elevated serum uric acid levels. Therefore, it is reasonable to screen transplant recipients to detect the occasional patients with very high uric acid levels. These high levels are most likely to occur early after transplantation in patients treated with CsA and diuretics.

Malnutrition and Obesity (Table 27)
Incidence. Approximately 10% of patients exhibit hypoalbuminemia at 1 yr after transplantation (605). The prevalence of low serum albumin levels increases in the late posttransplant period, to >20% by 10 yr after transplantation. In a study of 232 kidney-pancreas transplant recipients, hypoalbuminemia was present in 15% at 1 yr and 8% at 3 yr (606). Many studies report the prevalence of obesity at the time of transplantation but few report its prevalence after transplantation. In two small studies, both diabetic and nondiabetic renal transplant recipients gained an average of 6 to 8 kg by 2 yr and an average of 9 to 10 kg by 3 yr after transplantation (607,608). One retrospective study of 115 renal transplant recipients found that only 21% were obese at the time of transplantation but 43% were overweight 1 yr after transplantation (609).

Consequences. At least one cross-sectional study of renal transplant recipients found that low serum albumin levels were associated with poor long-term outcomes after renal transplantation (605). Similar results were reported for kidney-pancreas transplant recipients (606). Obesity at the time of transplantation was associated with worse outcomes in most (609,610,611,612,613,614) but not all (615,616) studies. However, few studies have examined the effects of posttransplant weight gains on outcomes.

Rationale. Serum albumin levels can be easily measured in most laboratories. Low serum albumin levels may be the result of decreased production and/or increased catabolism, both of which are frequently associated with malnutrition. Increased urinary protein excretion may also result in low serum albumin levels. Hepatic synthesis of prealbumin (transthyretin) is very sensitive to the adequacy of protein and energy intake, and serum prealbumin levels may be a good measure of nutrition (617,618). However, levels of prealbumin, like other plasma proteins, may be affected by inflammation and other factors (619). There are virtually no studies examining prealbumin levels in renal transplant recipients, but prealbumin levels appear to predict outcomes in other populations (620), including patients with ESRD (621). Weight loss, evidence of muscle wasting, and the clinical setting may also suggest malnutrition.

Corticosteroids accelerate the protein catabolic rate and frequently create a negative nitrogen balance (622). Studies have documented significant increases in the protein catabolic rate, accompanied by decreases in serum albumin levels, in the immediate posttransplant period (622,623). Even low-dose maintenance corticosteroid therapy increases protein catabolism and muscle wasting. In one study, renal transplant recipients with stable kidney function exhibited mid-arm muscle circumference values below the 5th percentile (608). Chronic allograft nephropathy is associated with decreases in muscle mass, which are reflected in decreases in urinary creatinine excretion (44).

Increased caloric intake may also occur after transplantation (624), primarily because of enhanced appetite associated with corticosteroid use. Obesity may contribute to cardiovascular risks and other complications. Obesity can be detected by physical appearance during examinations, as well as by measurements of height, weight, and body mass index (weight in kilograms divided by height in meters squared). Guidelines for identifying and managing obesity have been developed by the National Heart, Lung, and Blood Institute (625). These guidelines suggest that body mass index values be measured to assess obesity (category C evidence). Weight loss is recommended to improve lipid profiles, lower BP, and improve glycemic control for patients with type II diabetes mellitus (each category A evidence) (625). A low-fat, low-calorie diet is recommended (category A), with exercise (category A) (625). The United States Preventive Services Task Force recommended that periodic height and weight measurements, with body mass index values or published tables for normal ranges, be used to screen for obesity (category B evidence) (626).

Both malnutrition and obesity are common enough to suggest that screening could be effective. Although there have been no studies demonstrating that intervention can improve nutrition and outcomes after renal transplantation, studies in the general population provide sufficient evidence to suggest that interventions may be effective. Few studies have examined the success of diet alterations in correcting posttransplant weight gains (627).

Growth and Development of Children (Table 28)
Definition. Children of different ages grow at different rates. The fastest growth occurs in the first 2 yr of life and during puberty. Growth failure is defined as a height >2 SD below the normal value for that age, a height velocity SD score (SDS) of -2.0, or an absolute growth rate of <4 cm/yr. The development of young children is assessed as the achievement of certain landmarks of fine motor, gross motor, social, and communication skills, compared with age-specific normal findings. School function, again compared with age-matched peers, is the best measure of development for older children.

Incidence. Children typically develop growth failure early in the course of ESRD. A survey of the European Dialysis and Transplantation Registry demonstrated than 50% of patients exhibited a final adult height below the 3rd percentile, with some advantage being observed for the children who received renal transplants (628,629). In a North American Pediatric Renal Transplant Cooperative Study analysis of 1768 children with GFR of <75/ml per min per m², more than one-third exhibited a height deficit of >2 SDS. It has been amply demonstrated that chronic renal insufficiency beginning in infancy leads to permanent reductions in growth potential (630). Growth retardation continues in children undergoing dialysis (peritoneal dialysis or hemodialysis). It has been suggested that a functioning transplant should enable the child to achieve catch-up growth (631,632). Unfortunately, long-term registry data on growth have been disappointing. The North American Pediatric Renal Transplant Cooperative Study tracked growth after transplantation longitudinally, using the same cohort, for at least 5 yr. The height deficit was -2.41 SD at 2 yr, -2.46 SD at 3 yr, and -2.29 SD at 54 mo. Children experienced improvements in height SDS of +0.18 at 2 yr, +0.16 at 3 yr, and +0.11 at 54 mo. When improvements in height deficits were evaluated according to donor source, no differences were noted between living-related and cadaver donor recipients. Analysis of height SDS according to race revealed that, whereas steady improvements were noted at 2 yr and 54 mo for Caucasian children, there was an actual deceleration of growth for African American and Hispanic children. Only the initial height deficit and recipient age were independent predictors of improved height after transplantation. Catch-up growth, defined as an improvement of 1 SDS, was observed only for children 0 to 1 yr of age. Overall, catch-up growth was observed for only 47% of children between the ages of 2 and 5 yr. For children >5 yr of age (72% of the study cohort), little catch-up growth was noted (292).

During the first 2 yr of life, the brain more than doubles in volume and attains 80% of its final size. Children who develop renal failure early in life are at serious risk for neurologic dysfunction (633,634). In children, verbal performance and memory function are decreased (635,636). Test performance generally improves after transplantation (635,637). In general, the proportion of patients who achieve success in higher education is less than that in the general population or that of patients with other chronic diseases, such as diabetes mellitus (638). There is some indication that the long-term results of rehabilitation after kidney transplantation are improving (639,640,641). However, all ages report problems with usual daily activities (642).

Consequences. Although transplantation can temporarily restore normal renal function to children with ESRD, residual somatic and neurologic damage and the requirements for longterm medication administration prevent complete and longlasting rehabilitation. Therefore, the majority of children do not attain normal height. Furthermore, cognitive functioning is affected and the problems associated with chronic disease management interfere with complete rehabilitation.

Rationale. There are several possible approaches to treating growth failure after renal transplantation, such as discontinuation of prednisone administration (292), alternate-day steroid therapy (289,290,643), or the use of recombinant human growth hormone (644,645), but none are universally accepted. However, close monitoring of growth for all children is mandatory, so that treatment may be provided if necessary. Furthermore, formal psychologic and developmental testing is necessary to provide support for rehabilitation that is as complete as possible.

	VII. Cancers

Top Abstract Introduction Scope Intended Users Materials and Methods Organizational Scheme I. Frequency and Timing... II. Graft Function III. Immunosuppressive... IV. Cardiovascular Disease V. Bone and Bone... VI. Nutrition and Metabolism VII. Cancers VIII. Infections Endorsements References

 
Cancers of the Skin and Lip (Table 29)
Definition. Malignant skin lesions common among renal transplant recipients include squamous cell carcinomas, basal cell carcinomas, malignant melanomas, and Merkel cell tumors (646). Kaposi's sarcomas (KS) frequently involve the skin but are discussed below.

Incidence. Carcinomas of the skin and lip are the most common malignancies among adult renal transplant recipients (647,648,649,650). As many as 40 to 53% of all malignancies among transplant recipients are skin and lip cancers (649,651). The incidence of skin cancer varies with the amount of sun exposure and the length of follow-up monitoring after transplantation (646). In regions with limited sun exposure, such as the Netherlands, the risk of a first skin tumor is 10% at 10 yr after transplantation and 40% at 20 yr (652). In regions with increased sun exposure, such as Australia, the incidence is 66% by 23 yr after transplantation (653).

In the general population, basal cell carcinoma is the most common malignant skin tumor (654). Basal cell carcinomas outnumber squamous cell carcinomas by ratios of 5:1 to 8:1 in the general population, but squamous cell carcinomas outnumber basal cell carcinomas by ratios of 1.1:1 to 1.8:1 among renal transplant recipients (647,652). The overall incidence among renal transplant recipients is 250 times higher for squamous cell carcinoma and 10 times higher for basal cell carcinoma, compared with the general population (652). Patients treated with dialysis do not exhibit a higher incidence of skin cancer, compared with the general population, and cohorts of patients for whom immunosuppression is withdrawn after allograft failure exhibit a much lower incidence of skin cancer, compared with cohorts of patients with functioning grafts (655).

Malignant melanomas also occur with increased frequency among renal transplant recipients (654,656). They represent 5% of posttransplant skin cancers, compared with 2.7% of skin cancers in the general population (657). Most malignant melanomas in transplant recipients originate in the skin (657). Malignant melanomas are the most important tumors that are inadvertently transmitted with transplanted organs (657). Malignant melanomas can recur after transplantation among patients who were treated for the disease several years before transplantation (657).

The risk of lip cancer is increased up to 29-fold, compared with the general population (650). In a study of 160 renal transplant recipients, 13% exhibited leukoplakia; of those with leukoplakia who underwent biopsies, 62% exhibited dysplastic lesions and 10% squamous cell carcinoma of the lip (658). Sun exposure and smoking were identified as risk factors.

In one long-term study of pediatric organ transplant recipients, skin cancer was the second most common malignancy (659). Malignant melanoma and lip cancer are more common among pediatric transplant recipients than adult recipients.

Immunosuppressive therapy, ultraviolet radiation, and infection with human papillomavirus (HPV) are risk factors for skin cancer (646,652). Skin cancer among transplant recipients is more common in regions with abundant sun exposure, especially for fair-skinned individuals. The association between total sun exposure (before and after transplantation) and the risk of developing skin cancer (648,660) may be linked to mutations in the p53 gene (661). There is a high prevalence of p53 immunoreactivity in premalignant and malignant skin lesions in renal transplant recipients, suggesting a role for p53 protein in skin cancers (662).

Immunosuppression contributes to the development of skin malignancies (647,663,664). Renal transplant recipients with cutaneous malignant melanomas are unable to mount appropriate cellular immune responses to neoplastic cells (654). The role of the immune system in skin cancer is further emphasized by the noted association between MHC-DR1 and basal cell carcinomas in the general population (665). For renal transplant recipients, an association between MHC-B mismatching and the development of skin cancer that was reported for Dutch patients (666,667) could not be confirmed in a study of Australian patients (668). MHC-B27, MHC-DR7, and MHC-DQw2 are associated with an increased risk of cancer among renal transplant recipients, whereas there has been a negative association between MHC-DA11 and skin cancer (667,669).

Infection by HPV has also been postulated to be an important factor in the development of skin cancer among renal transplant recipients (646). Warts have been noted in up to 75% of renal transplant recipients, and HPV has been isolated from 43% of warts (660). HPV has been detected in close to 60% of nonmelanoma skin cancers in renal transplant recipients (670). A broad spectrum of HPV types and multiple viral types in the same specimen have been noted in biopsies of skin lesions from transplant recipients (660,670,671,672). Nevertheless, it is not clear to what extent HPV contributes to the development of skin cancer among renal transplant recipients (673,674).

Consequences. Skin cancers are more aggressive in transplant recipients than in the general population (647). The incidence of multiple cancers is high, and many patients have several different types of skin malignancies that are prone to recurrence and metastasis (646,648). The onset of squamous cell and basal cell carcinomas occurs at a younger age among transplant recipients, compared with the general population (675,676).

Lymph node metastases have been reported to occur in 5.8 to 7% of transplant recipients with squamous cell carcinoma of the skin (647,663), and approximately 5% die as a result. Lymph node metastases occur in 20% of patients with malignant melanoma, and 30% ultimately die as a result of their malignancies. Overall, approximately 5% of transplant recipients with skin carcinomas that were reported to the Cincinnati Transplant Tumor Registry died as a result of their skin malignancies, compared with 1 to 2% of cancer-related deaths in the general population (647). Basal cell carcinomas rarely metastasize but do cause local tissue destruction, which may lead to disfigurement or functional impairment (654).

Rationale. The incidence of skin cancer among renal transplant recipients is high, and morbidity and mortality rates are much higher than those for the general population. Skin cancers can be detected clinically and can often be cured by excision. Most renal transplant recipients with cutaneous malignant melanomas that are diagnosed early survive and remain free of disease (654,657). Early diagnosis is currently the best method to alter the course of malignant melanomas among renal transplant recipients (677). Early diagnosis is also the best method to alter the course of other skin malignancies.

Examination of the skin is the principal screening test for skin cancer (678). The sensitivity and specificity of skin examinations is unknown (679). Total-body skin examinations are safe and may increase the detection of skin lesions, which may be present in unexposed areas. Early detection of skin cancer requires clinical recognition of both precancerous and malignant skin lesions (680). Self-examinations of the skin by patients are probably less accurate than physician examinations (678). Dermatologists have the greatest expertise in correctly identifying premalignant and malignant skin lesions (680).

The American Academy of Dermatology and the National Institutes of Health Consensus Panel recommend regular visits for skin cancer screening and patient education concerning periodic skin self-examinations (681,682). The American Cancer Society recommends monthly self-examinations for all adults and physician skin examinations every 3 yr for individuals between the ages of 20 and 39 yr and annually for individuals >40 yr of age (683). The United States Preventive Services Task Force does not recommend for or against skin cancer screening for the general population but states that a recommendation may be made to consider referring patients at increased risk for skin cancer to specialists for evaluation and surveillance (678).

Avoiding sun exposure and using protective clothing are recommended for the general population (678). Sunscreens may offer additional protection. It is prudent to make these recommendations to all renal transplant recipients (660). In a recent survey of 122 patients attending a clinic an average of 3.1 yr after renal transplantation, 41% were unable to recall specific skin cancer education and only 14% had ever visited a dermatologist (684).

Anogenital Carcinomas (Table 30)
Definition. Carcinomas of the anogenital region are responsible for substantial morbidity and deaths (646). The most common sites affected by anogenital malignancies reported to the Cincinnati Transplant Tumor Registry include the vulva, anus, perianal region, penis, perineum, and scrotum (646).

Incidence. Carcinomas of the anogenital area represent as much as 2.5 to 2.8% of cancers in the renal transplant population (685,686). Cancers of the female genital tract are the most frequent nonskin cancers in some reports (648). Epidemiologic studies revealed a 100-fold increased risk for cancer of the vulva and anus, compared with the general population (650). Analysis of data from the Australia and New Zealand Dialysis and Transplant Registry demonstrates that, although dialysis patients are not at increased risk for anogenital malignancies, transplant recipients have risks 10 times higher than those of the general population for these malignancies (687). Anogenital malignancies are more common among women than men (2.5:1), in contrast to other neoplasms among transplant recipients, for which men outnumber women (646). The average time between transplantation and the appearance of anogenital malignancies is 114 mo (646). Transplant recipients are generally much younger (41 yr) when anogenital cancers are diagnosed, compared with patients in the general population, who are generally 50 to 70 yr of age when cancers are diagnosed (646,688). Carcinoma of the vulva and perineum is the fourth most common tumor among children who have undergone transplantation, representing approximately 4% of all neoplasms in this group (659). Pediatric renal transplant recipients present with anogenital malignancies at an average age of 27 yr (659).

Immunosuppression and oncogenic viruses seem to be important in the pathogenesis of anogenital cancer (686). Some anogenital lesions remain refractory to therapy during immunosuppression (689). Renal transplant recipients exhibit a higher incidence of HPV (690), and HPV is associated with anogenital cancers (691). Many transplant recipients have anogenital warts, and transplant recipients with anogenital malignancies commonly have histories of condyloma acuminatum (646,689). Infection with HPV (including oncogenic strains) has been identified in anogenital lesions of renal transplant recipients (689,690). There appears to be an association between the prevalence of HPV DNA and the susceptibility to potentially lethal anogenital squamous cell cancers (692).

Consequences. Although one-third of transplant recipients have in situ lesions (646,686), multiple extensive lesions are common (686). Some patients have involvement not only of the anogenital area but also of the uterine cervix, vagina, or urethra by squamous cell carcinoma or severe dysplasia (686). Lymph node extension occurs in at least 11% of patients (686). Treatment may require local excisions, vulvectomies, abdominoperineal resections, and inguinal lymph node resections (646). Some patients require chemotherapy or radiation therapy (646). In some instances, immunosuppression has been discontinued (689). In a recent report from the Cincinnati Transplant Tumor Registry, 11.6% of transplant recipients with anogenital malignancies died as a result of their malignant neoplasms (646).

Rationale. The incidence of anogenital lesions is high, and the clinical presentation is often misleading. Regular surveillance with biopsies is often necessary to determine the presence of dysplasia or malignant transformation (657). Lesions can be very destructive locally, can be associated with coexistent vaginal and uterine carcinomas, and can cause metastasis and death. Some renal transplant recipients manifest a high susceptibility for the development of potentially lethal malignancies, including anogenital malignancies (692). Neoplastic lesions associated with HPV in immunosuppressed patients have the potential for exceedingly rapid progression (693).

Early diagnosis of anogenital lesions permits early treatment (646). Early lesions can be treated with less destructive methods, whereas extensive disease may be refractory to therapy (689). Lesions that appear clinically benign may be malignant, and lesions that appear to be benign in biopsies may subsequently become malignant (689). Therefore, repeated surveillance, with biopsies of lesions and treatment of condylomas or warts, is recommended (646).

Women should undergo regular pelvic examinations and cytologic studies (686). Women may experience a "field effect," with multifocal lesions; coexistent uterine cervical carcinomas are observed for up to 17% of patients and vaginal carcinomas for 7.6% of patients (646). All women with external anogenital lesions should undergo regular pelvic examinations and cytologic studies (646,686). Anoscopy and biopsies of the anal epithelium can lead to early histologic diagnosis of anal intraepithelial neoplasia and/or anal HPV infection, but further studies are required to determine whether these tests should be recommended as screening tests for renal transplant recipients (690). The optimal frequency of surveillance for anogenital cancer among renal transplant recipients has not been established, but it is prudent to recommend at least annual examinations, with prompt evaluation of any new lesions.

KS and Other Sarcomas (Table 31)
Definitions. KS is rare in the general population and was first described in association with renal transplantation in 1969 (694). KS has also been described in several other clinical settings, including classic KS (in individuals of Mediterranean or Jewish ancestry), endemic KS (in Africa), and epidemic KS (in patients with AIDS) (695). Transplant recipients with KS can present with a combination of nodular and lymphadenopathic forms (696). In the nodular form, reddish-blue to purple nodules or plaques appear on the skin and/or oropharyngeal mucosa. In the lymphadenopathic form, there can be involvement of lymph nodes and the mucosa of the gastrointestinal tract, tracheobronchial tree, and lung parenchyma.

Many other sarcomas have been observed in renal transplant recipients, with the majority involving visceral organs or soft tissues (697). Bone and cartilage involvement by sarcomas in renal transplant recipients is rare (688). The most common varieties of sarcomas (excluding KS) in transplant recipients are fibrous histiocytoma, leiomyosarcoma, fibrosarcoma, rhabdomyosarcoma, hemangiosarcoma, undifferentiated sarcoma, and mesothelioma (697).

Incidence. The incidence of KS is markedly increased among renal transplant recipients, compared with the general population, and ranges from 0.4% for patients in northern and western countries to as high as 4 to 5% for Arab, Jewish, or Mediterranean patients (698,699). The incidence of KS among transplant recipients of Jewish or Mediterranean ancestry may be 500-fold higher than that for the same ethnic groups in the general population (700). KS represents only 0.01 to 0.06% of all tumors in the general population but 3% of tumors reported to the Australian and New Zealand Combined Dialysis and Transplant Registry and 4% of tumors reported to the Cincinnati Transplant Tumor Registry (648,651,701). In Saudi Arabia, KS is responsible for 70% of de novo tumors after transplantation (699). On average, KS appears 21 mo after transplantation, with the largest number of tumors being diagnosed within 5 yr after transplantation (695). The male/female ratio for KS among transplant recipients is nearly 3:1 (695).

Several factors seem to be important in the development of KS, including immunosuppression, viral infections, genetic factors, and perhaps environmental factors (695,699). Immunosuppression has been associated with KS, and reduction or discontinuation of immunosuppressive treatment can lead to remissions (701,702). Reinstitution of more intense immunosuppressive therapy is usually followed by recurrence and progression of KS lesions (701,702). Infection with human herpes virus 8 (HHV-8) has been detected in renal transplant recipients with KS (703,704,705), and pretransplant HHV-8 seropositivity is a strong risk factor for posttransplant KS (706,707). In a recent case-controlled study, 17 of 25 renal transplant recipients with antibodies to HHV-8 developed KS, compared with 1 of 33 patients with no antibodies to HHV-8 (705). Infection with HHV-8 (or KS-associated herpes virus) could explain the susceptibility of some renal transplant recipients to KS (708). Genetic factors are probably also important, because most transplant recipients with KS are of Arab, Italian, Greek, or Jewish background (705,709). There is an increased frequency of certain MHC phenotypes among patients with KS, but these MHC types correspond to those expected on the basis of the ethnic backgrounds of the affected patients (709). A possible role for environmental factors is suggested by the extremely high incidence of KS among renal transplant recipients in Saudi Arabia (699).

Sarcomas other than KS are also observed more frequently for transplant recipients, compared with the general population; they represent 1.7% of neoplasms reported to the Cincinnati Transplant Tumor Registry, compared with 0.5% in the general population (697). For pediatric organ transplant recipients, sarcomas are the third most common malignancy, comprising 4% of the tumors in children (659).

Consequences. Sixty percent of patients with KS reported to the Cincinnati Transplant Tumor Registry had nonvisceral KS confined to the skin, conjunctiva, or oropharyngeal mucosa (697). Fifty-three percent of these patients experienced complete remissions with treatment, but in some cases the disease progressed even with therapy (697). Most of the deaths in this group were the result of infections or rejection (695,697).

Forty percent of patients with KS reported to the Cincinnati Transplant Tumor Registry exhibited visceral involvement, including the gastrointestinal tract, lungs, lymph nodes, and other organs (65). Skin and oral involvement was also present in most of these cases (697). Only 27% of patients with KS visceral involvement experienced remissions with therapy, and 57% of those patients died in a short time, in most cases as a direct result of KS (695,697). A report from a single institution that has cared for large numbers of patients with KS also noted high mortality rates for patients with KS involving several organs (699).

Multiple therapies have been used for renal transplant recipients with KS, including changes in immunosuppressive therapy, surgery, radiation therapy, chemotherapy, and immune therapy, among others (697). Reduction/discontinuation of immunosuppressive treatment is usually one of the first therapeutic interventions for KS but can lead to allograft dysfunction and/or irreversible graft loss in more than one-half of renal transplant recipients (695,697). Increasing cumulative doses of immunosuppressive medications used for graft dysfunction or repeat transplantation (even after many years) usually lead to the recurrence of KS and the progression of lesions (701,702).

Transplant recipients with KS frequently have other malignancies, and 6% of patients reported to the Cincinnati Transplant Tumor Registry exhibited a second primary malignancy (697). KS is a less common tumor in the pediatric transplant population but is associated with a very high mortality rate for children with visceral KS (659).

Sarcomas (other than KS) are associated with high morbidity and mortality rates for transplant recipients, with at least 61% of transplant recipients with other sarcomas in the Cincinnati Transplant Tumor Registry being reported as dead, in most cases (44%) as a direct result of the sarcomas (697). A relatively large number of sarcomas (6%) occur adjacent to or in the renal allograft itself (697). Twenty percent of patients with sarcomas also exhibit other neoplasms (697).

Rationale. The detection of reddish-blue macules or plaques in physical examinations of the skin and oropharyngeal mucosa of renal transplant recipients should engender concern regarding KS (695). Cutaneous lesions that appear like infected granulomas that fail to heal are also suspect (695). Histologic examinations of KS lesions typically reveal a mixture of endothelium-lined vascular spaces and spindle-shaped cells (710).

More than 70% of patients with visceral KS also have skin lesions, but some patients do not have skin or oropharyngeal mucosal lesions (695). There are no data in support of or against screening for visceral KS in patients without skin, conjunctival, or oropharyngeal mucosal lesions. Although KS involvement of the gastrointestinal tract can be detected with endoscopy (695) and CT or MRI may reveal pulmonary and visceral involvement (711), these tests are not suitable for routine screening. Abnormalities attributable to pulmonary KS can be noted in chest radiographs (712). Similarly, CT and MRI may detect other sarcomas (711,713) but are not suitable for routine screening.

Complete remissions from KS are much more common when the disease is diagnosed and treated early, before visceral involvement is present (695,697,701). Once visceral involvement develops, the mortality rate is substantial and exceeds 50% (695,699). Patients with other sarcomas may respond to single or combined treatment modalities, including excision, amputation, radiation therapy, chemotherapy, alteration of immunosuppressive therapy, and administration of interferon, among others. Metastatic sarcoma, even in the general population, is associated with very poor survival rates (714). Early diagnosis of other sarcomas requires attention to any areas of bone abnormalities or soft-tissue swelling, new masses, or unexplained pain.

Posttransplant Lymphoproliferative Disorders (Table 32)
Definition. Posttransplant lymphoproliferative disorders (PTLD) are abnormal proliferations of lymphoid cells that result from immunosuppression (715,716). The increased risk for PTLD and lymphomas after transplantation has been appreciated for more than three decades (656,717,718). Non-Hodgkin's lymphomas constitute 94% of PTLD (647). Myelomas comprise 4% and Hodgkin's disease 2.5% of the lymphomas reported to the Cincinnati Transplant Tumor Registry (647). Studies of non-Hodgkin's lymphomas occurring after transplantation revealed that 87% arise from B lymphocytes, 13% are of T cell origin, and <0.5% are of null cell origin (647). Genetic studies have revealed that most PTLD in solid organ transplant recipients are of recipient B cell origin (719). There is substantial evidence that Epstein-Barr virus (EBV) plays an important role in most PTLD (716, 720, 721). PTLD can range from benign, polymorphic, polyclonal hyperplasias to highly malignant monoclonal tumors (716, 722). PTLD are classified using morphologic characteristics, immunohistochemical analyses, cytogenetic findings, Ig gene rearrangement studies, and EBV clonal studies (716, 722,723,724). There is a strong correlation between morphologic and molecular genetic categories of PTLD and clinical outcomes (723, 724).

Incidence. PTLD occur in 1 to 5% of renal transplant recipients (722, 725,726,727,728). The incidence appears to be higher (up to 10%) for pediatric recipients (727, 729). The incidence of PTLD is highest during the first posttransplant year (725). Among 45,141 cadaveric renal transplant recipients from Europe and North America, PTLD occurred in 0.2% during the first 1 yr after kidney transplantation and in 0.04%/yr thereafter (725). Lymphomas represent approximately 16% of all tumors reported to the Cincinnati Transplant Tumor Registry and 11% of the nonskin malignancies in the Australia and New Zealand Dialysis and Transplant Registry (663).

EBV plays a major role in the development of lymphomas after transplantation (720, 721, 730), and primary infections are associated with the greatest risk (715, 721). T cell-mediated immunity controls but does not eradicate EBV infection, which results in clinically latent infection (731). With immunosuppression, the number of EBV-specific cytotoxic T lymphocytes is decreased and there is an increase in the number of circulating EBV-infected B cells. With immunosuppression, EBV-infected B cells can proliferate into polyclonal hyperplasias or even monoclonal lymphomas (715, 731). The type and intensity of immunosuppression are important in the development of PTLD, and the use of antilymphocyte antibodies greatly increases the risk (725, 732). Other factors, such as MHC mismatching and CMV disease have also been associated with PTLD (663, 715). In one pediatric study, Caucasian race and cadaver donor source were identified as risk factors for PTLD (733). The incidence of PTLD among children treated with tacrolimus was substantially higher than that among children treated with CsA (715, 733).

Consequences. The clinical presentation of PTLD in transplant recipients is variable, and extranodal involvement is more common than in the general population (715, 721). Some patients experience a syndrome similar to infectious mononucleosis, characterized by fever, malaise, and lymphadenopathy, with or without pharyngitis. Some experience a febrile illness with leukopenia and elevations in liver enzyme levels. Others present with focal involvement of the central nervous system, renal allograft, gastrointestinal tract, liver, bone marrow, lungs, or other isolated organs (715, 720). Finally, >50% of patients present with disseminated disease affecting multiple organs (647, 688, 715).

Central nervous system involvement is common among patients with PTLD and was noted in 22% of the cases reported to the Cincinnati Transplant Tumor Registry (734). Central nervous system lesions frequently involve the brain and are multicentric in distribution (734). The mortality rate associated with central nervous system involvement is extremely high, and only 20% of patients achieve remissions with therapy (734). The allograft itself is involved in up to 20% of patients (688). Renal allograft dysfunction may occur as a consequence of involvement by PTLD (735, 736). Gastrointestinal tract involvement is common (737). Presentations may include abdominal pain, gastrointestinal bleeding, obstruction, or bowel perforation (715). Young patients tend to present within the first 1 yr after transplantation and tend to exhibit more diffuse polymorphic lesions, with mortality rates of 38 to 50% (738, 739). Older patients present several years after transplantation and exhibit more localized extranodal disease, monomorphic lesions, and mortality rates of 66 to 91% in some series (738, 739).

Although some patients achieve complete remissions, overall survival rates may be as low as 36%, with a median survival time of 5 mo (739). Patients with PTLD of T cell origin have very poor prognoses (739). Of 1061 patients in the Cincinnati Transplant Tumor Registry who were treated, 399 (38%) experienced complete remissions (740). Of 1345 treated and untreated patients, 474 (35%) died as a result of PTLD, 216 (16%) died as a result of other causes (but PTLD may have contributed to some of the deaths), and 655 (49%) were still alive (740).

Rationale. Mortality rates exceed 50% for PTLD (715), and symptoms may not occur until late in the disease. Early diagnosis could improve outcomes. Serologic evidence of EBV infection or reactivation is not reliable for immunosuppressed transplant recipients (715). Viral reactivation and excretion may occur without changes in antibody titers (721, 741). Measurements of the EBV burden in the peripheral blood may be useful for identifying patients who will eventually develop PTLD (742,743,744). Although an early study noted that the detection of EBV genomic material in circulating B cells was not predictive of lymphoma development (745), subsequent studies demonstrated that the results of semiquantitative and quantitative PCR assays for EBV DNA in peripheral blood were correlated with the risk of developing PTLD (743, 744, 746). However, the utility of measuring the EBV viral load to screen for PTLD has not been systematically evaluated.

Determination of oropharyngeal shedding of EBV does not appear to be useful for the identification of transplant recipients at risk for EBV-associated morbidity (741). Quantitative oropharyngeal EBV shedding, as determined using molecular probes, is almost universally observed for seropositive renal transplant recipients in the first posttransplant year. Many allograft recipients who shed very high levels of EBV in the oropharynx do not subsequently develop clinically apparent PTLD (741).

In a study of liver transplant recipients that used in situ hybridization for detection of the EBER1 gene during latent EBV infection, it was noted that expression of this gene in liver tissue preceded the occurrence of PTLD (747). However, others failed to confirm the association of EBV gene expression in liver tissue with subsequent progression to PTLD (748). Therefore, additional studies are needed to determine whether organ-specific diagnoses of EBV can be used to screen for PTLD.

Sonography (749), CT (750, 751), and MRI (752,753,754,755,756) are useful in diagnosing PTLD but have little benefit for routine screening of asymptomatic patients. Upper and lower endoscopic evaluation of the gastrointestinal tract is the diagnostic procedure of choice for early recognition of abnormal lesions involving the bowel and for obtaining biopsy samples to establish the diagnosis of PTLD (757). Involvement of the gastrointestinal tract by PTLD is common, but symptoms may not occur until late in the disease (757). There are no data for or against routine endoscopic evaluation of the gastrointestinal tract in asymptomatic renal transplant recipients as a screening test for PTLD. Renal transplant recipients with unexplained gastrointestinal manifestations should be promptly referred for endoscopic evaluation of the gastrointestinal tract.

Uroepithelial Malignancies and Renal Carcinomas (Table 33)
Definition. Renal carcinomas reported to the Cincinnati Transplant Tumor Registry include renal cell carcinomas (74%), transitional cell carcinomas/urothelial carcinomas (12%), and miscellaneous carcinomas (14%) (647). Most renal carcinomas occur in the native kidneys, but renal cell carcinomas have been accidentally transplanted with allografts (758). Renal cell carcinoma has also been reported to develop in renal allografts many years after transplantation (759).

Incidence. Renal carcinoma occurs in 0.5 to 3.9% of renal transplant recipients (760,761,762). The risk of renal carcinoma is 10 to 100 times higher for renal transplant recipients than for the general population (648, 761). Several risk factors have been associated with renal carcinoma in renal transplant recipients, including previous analgesic abuse, previous history of renal cell carcinoma, and acquired renal cystic disease (758, 762,763,764). Analgesic abuse is associated with an increased risk for urothelial cancers, specifically tumors of the renal pelvis and the bladder (763). In one study, urothelial tumors were reported for 17% of patients who received renal transplants because of analgesic nephropathy, compared with only 0.1% of patients who received transplants because of other causes of ESRD (763). Data from the Cincinnati Transplant Tumor Registry indicated that patients with prior histories of symptomatic renal carcinomas who underwent renal transplantation after surgical removal of their lesions experienced a 25% rate of recurrence of renal cell carcinoma. More than one-half of the renal carcinoma recurrences occurred during the first 2 yr after renal transplantation (758). Acquired renal cysts occur in 40 to 80% of patients who undergo long-term dialysis (764, 765), and acquired cysts are an important risk factor for the development of renal cell carcinomas among dialysis patients (762, 764). A longer duration of hemodialysis before transplantation, male gender, and older age are associated with acquired renal cystic disease and the development of renal carcinoma after renal transplantation.

Consequences. Renal transplant recipients with renal cell carcinomas may present with fever, weight loss, early satiety, anorexia, and hematuria, among other manifestations (765). However, many patients do not present with symptoms or abnormal findings until their disease is advanced. Renal carcinomas can be particularly aggressive in some renal transplant recipients (765). Many patients exhibit widespread metastases involving the lymph nodes, liver, and lungs, as well as invasion of the renal veins and inferior vena cava. Approximately 40% of transplant recipients with renal cell carcinomas die as a consequence of their malignancies (758). The clinical course of urothelial tumors is particularly aggressive. More than one-half of the patients with urothelial cancers already have extensive disease at the time of presentation, and the median survival time for this group is only 17 mo (763). In contrast, 24% of renal cell carcinomas in the Cincinnati Transplant Tumor Registry were discovered incidentally (740).

Rationale. Most renal transplant recipients exhibit no clinical manifestations early in the course of renal carcinomas. Surgical resection at an early stage is the only established cure for renal carcinomas. The course of renal tumors appears to be more aggressive in renal transplant recipients than in the dialysis population (765). Tumors grow and metastasize, leading to morbidity and high mortality rates. Disease-free survival rates of >40% for transplant recipients who undergo successful resections of renal carcinomas have been reported to the Cincinnati Transplant Tumor Registry (758). Therefore, early detection of renal carcinomas could improve outcomes.

Urine cytologic analyses are not very useful for detecting urothelial neoplasms that originate in the upper urinary tract (763). An additional limitation in renal transplant recipients is that most urine originates in the transplant kidney and not in the native kidneys (763). Urine cytologic analyses may useful for detecting lower-tract tumors in patients with histories of analgesic abuse (758). Hematuria, detected by urinalysis for asymptomatic patients, can be caused by renal malignancies, extrarenal malignancies, and nonmalignant disorders (766). The use of urinalyses to screen asymptomatic renal transplant recipients for renal carcinomas has not been evaluated.

Renal sonography may be useful to screen for renal cell carcinomas in renal transplant recipients (760,761,762, 764, 767). Renal ultrasonography, however, has limited sensitivity in detecting small masses (750). The optimal frequency for screening has not been established. Several investigators have suggested yearly sonographic examinations for asymptomatic renal transplant recipients (762). Others have recommended twice-yearly sonographic screening for renal transplant recipients at high risk (760) or annual testing for patients with acquired renal cystic disease (764). A study that included repeated sonographic testing revealed a growth rate of 0.5 to 1.0 cm/yr for renal carcinomas in two allograft recipients (762). On the basis of this growth rate of 1 cm/yr and the fact that patients with tumors of <3 cm have limited disease, a group of investigators suggested that ultrasonographic examinations be performed every 3 yr to diagnose renal cell carcinoma of the native kidneys at an early stage (761). There are no prospective data to indicate how cost-effective these screening strategies might be for renal transplant recipients (758).

Contrast-enhanced CT is much more sensitive than sonography for detecting small renal masses (752, 762). MRI is at least as effective as contrast-enhanced CT in detecting small renal lesions (752, 754). MRI avoids the risk of severe reactions to iodinated contrast material (754). MRI using gadolinium for enhancement does not cause nephrotoxicity (768). Given the sensitivity of CT and MRI for detecting lesions of <1 cm, it could be possible to use these methods as screening tools with less frequent testing, compared with renal sonography. Nevertheless, there are no data for or against the routine use of CT or MRI to screen for renal carcinomas. In the selection of one modality over the other, factors such as the safety of administration of contrast agents, local expertise, and costs should be considered.

Hepatobiliary Carcinomas (Table 34)
Definition. Hepatocellular carcinomas are associated with chronic liver disease, especially chronic hepatitis B virus (HBV) and HCV infections, and are the most important hepatobiliary tumors in renal transplant recipients (647, 769).

Incidence. The incidence of hepatocellular carcinoma demonstrates marked geographic variation. The presence of chronic liver disease, especially chronic HBV or HCV infection, greatly increases the risk of hepatocellular carcinoma among renal transplant recipients (647, 769). The incidence of hepatocellular carcinoma among renal transplant recipients in Europe is <0.1%, with older male patients exhibiting the highest incidence (655). The incidence of hepatocellular carcinoma is as high as 3% in Taiwan, where there is a very high incidence of chronic hepatitis associated with HBV and HCV infection (769). The relative risk of hepatocellular carcinoma among renal transplant recipients, compared with the general population, is increased 38-fold in Europe and 100-fold in Taiwan. Hepatocellular carcinomas represent approximately 1.7% of all malignancies reported to the Cincinnati Transplant Tumor Registry (647), and primary liver tumors represent approximately 2% of all malignancies among childhood renal transplant recipients (659).

Consequences. Patients with hepatocellular carcinoma usually present with clinical manifestations late in the course of their disease. Most patients have advanced disease at the time of presentation, which precludes effective therapy. Surgical resection is the only curative treatment for hepatocellular carcinoma, and most tumors are unresectable at the time of presentation (770, 771). Most patients survive for only a few weeks or months. Survival rates for patients with hepatocellular carcinoma in the general population are 30% at 1 yr and 5% at 5 yr. Patients who have limited disease and can undergo resection exhibit 5-yr survival rates that range between 25 and 39% (770).

Rationale. Clinical manifestations associated with hepatocellular carcinoma occur only with advanced disease. In nonrenal transplant recipients, hepatocellular carcinomas have been reported to grow slowly and remain confined to the liver for years (770). Because tumor stage is the best predictor of survival, early diagnosis of hepatocellular carcinoma could lead to improvements in outcomes. Patients who undergo surgical resection of small hepatocellular carcinomas exhibit 5-yr survival rates that may approach 40% (770). Screening for hepatocellular carcinoma in nonrenal transplant recipients is controversial (772). Limiting screening to patients at high risk for carcinoma, with a strong likelihood of improvements in life expectancy, has been suggested (772). The indiscriminate application of screening for hepatocellular carcinoma may not benefit most patients with chronic liver disease and cirrhosis (772).

Serum -fetoprotein can be detected in some patients with early stages of hepatocellular carcinoma (770). The sensitivity of serum -fetoprotein analyses for the detection of small hepatocellular carcinomas is low (between 20 and 60%) (770, 772). Serum -fetoprotein analyses have high specificity for the detection of hepatocellular carcinoma, exceeding 90% in most series (771, 772). False-positive results are caused by the activity of underlying liver diseases (770, 771).

Hepatic ultrasonography is a more sensitive test than serum -fetoprotein measurements for the early diagnosis of hepatocellular carcinoma. The sensitivity of liver sonography exceeds 80 to 85% for the detection of small hepatocellular carcinomas (ranging in size from 1 to 5 cm) (772). The specificity of liver ultrasonography for the detection of liver carcinomas is close to 90% (771). The predictive value of liver sonography could be improved by limiting screening to patients who are at the highest risk for hepatocellular carcinoma, such as those with longstanding hepatitis and chronic liver disease. Unfortunately, the accuracy of sonography is reduced among patients with cirrhosis, because of the presence of cirrhotic nodules (770). Nevertheless, liver sonography performed every 6 mo for patients with longstanding hepatitis and chronic liver disease is currently the most widely accepted screening method for hepatocellular carcinoma.

CT with contrast enhancement is the primary method for detection and characterization of liver lesions (773). The detection of hepatocellular carcinomas in patients with liver cirrhosis is limited because of distorted liver morphologic features and the difficulty of differentiating nodular masses from malignant tumors (770). The use of helical (spiral) CT improves the ability to characterize small hepatic lesions (774). MRI has been reported to have greater diagnostic accuracy for small hepatic lesions, compared with other radiologic techniques (755, 770). MRI using newer contrast agents with reticuloendothelial cell specificity, such as supermagnetic iron oxides, may become useful for the detection and characterization of liver lesions (775). However, there are currently no data on the routine screening of renal transplant recipients for hepatocellular carcinoma using any CT or MRI techniques.

There has been interest in the identification of some other possible serum markers for hepatocellular carcinoma. Lectin-reactive -fetoprotein, plasma urokinase-type plasminogen activator, serum des--carboxyprothrombin, and prothrombin induced by vitamin K absence or antagonist II are potential serum markers of hepatocellular carcinoma but are not now being used clinically (771).

Carcinomas of the Uterine Cervix (Table 35)
Incidence. Abnormal cytologic features, with dysplastic lesions and cervical intraepithelial neoplasias, have been noted in up to 9% of renal transplant recipients in some studies (776, 777). The risk for development of cervical carcinoma has been estimated to be 3 to 16 times higher for renal transplant recipients than for the general population; however, 70% of these patients have in situ lesions (685, 687, 777). The incidence of cervical carcinoma appears to be particularly increased among female patients with functioning grafts before menopause (655). Cervical carcinomas represent approximately 3% of all malignancies among transplant recipients (651). Some recent data suggested that there has been a decrease in the incidence of cervical carcinoma among renal transplant recipients (778).

Several factors have been associated with an increased risk of cervical carcinoma for renal transplant recipients. The number of previous sexual partners has also been identified as a risk factor for cervical carcinoma among renal transplant recipients, similar to observations for the general population (777). The potential contribution of other risk factors for cervical carcinoma that have been noted in the general population has not been determined for renal transplant recipients. The presence of lower anogenital malignancies greatly increases the risk for cervical carcinoma among renal transplant recipients (686).

Consequences. Low-grade cervical neoplasias in renal transplant recipients may respond well to conservative therapies or simple hysterectomies (776). Therapies used for preinvasive lesions include laser vaporization or excision, cryosurgery, cold-knife conization, electrosurgical excision, and simple hysterectomy, among other therapies (779). Advanced-stage disease may not respond even to aggressive therapeutic interventions. In the general population, advanced cervical cancer is associated with a 5-yr survival rate of only 14% (776).

Rationale. There is considerable evidence that regular screening for cervical cancer reduces the incidence of invasive cancer and is associated with reductions in the mortality rate for cervical carcinoma in the general population (776, 780). Most cervical neoplasias in transplant recipients respond well to conventional therapy when treated at an early stage (776, 781).

Several organizations recommend annual pelvic examinations and Pap smear testing for sexually active women who have reached the age of 18 yr (776, 780, 782, 783). The optimal frequency of Pap smear testing has not been established for the general population. Women who are undergoing immunosuppression (such as renal transplant recipients) have a higher risk of cervical carcinoma and should undergo pelvic examinations, with Pap smear testing, annually (782). The upper age limit for screening with Pap smear testing is not well established. Although cytologic evaluations of the vagina after total hysterectomy to treat benign disease are generally not recommended (784), it may be prudent to continue periodic testing for renal transplant recipients, because of their higher risk for carcinoma (782).

Regular colposcopic examinations of the cervix have been suggested in some transplant series (785). Colposcopy for detection of suspect lesions in the uterine cervix has poor sensitivity and specificity when used as a screening test for cervical neoplasia among asymptomatic women (776). There is no evidence to recommend for or against routine colposcopic screening for cervical cancer in the general population (776) or among renal transplant recipients. Cervicography has a sensitivity comparable to that of Pap smears but exhibits much lower specificity in the detection of cervical carcinoma (776). Cervicography is not recommended as a screening test for cervical carcinoma in the general population (776). There is no basis to recommend cervicography as a screening test for cervical carcinoma among renal transplant recipients.

Screening for HPV can be performed by visual inspection, Pap smear testing, colposcopy, cervicography, in situ DNA hybridization, the dot blot technique, the Southern blot technique, and PCR, among other techniques (786). Several oncogenic strains of HPV (such as types 16 and 18) demonstrate strong epidemiologic associations with cervical cancer. Nevertheless, the positive predictive value of HPV for cervical intraepithelial neoplasia is low (776, 778). The incidence of HPV infection among renal transplant recipients is variable (777, 778, 781). Some authors reported that HPV (especially oncogenic strains) is more common among renal transplant recipients, compared with the general population (787), but others did not (778). Screening for HPV as a screening test for cervical carcinoma is not recommended for the general population (26). There are currently no data for or against HPV screening as a screening test for cervical carcinoma among renal transplant recipients.

Breast Cancer (Table 36)
Incidence. Among malignancies reported to the European Dialysis and Transplant Association—European Renal Association Registry, the breast was the most common cancer location for women, although the incidence of breast cancer was not increased, compared with the general population (655). A recent report on >23,000 female cadaveric renal transplant recipients from North America and Europe noted a breast cancer incidence of 0.3% during the first 1 yr after transplantation. This represented a relative risk of 0.49, compared with the general population (788). The relative risk for breast cancer for these transplant recipients increased to 0.84 in the subsequent years of follow-up monitoring in the study (788). Other groups observed the incidence of breast cancer among female renal transplant recipients to be approximately 0.6%, and relative risks for breast cancer have been noted to be as high as 1.1 to 1.3, compared with the general population (648, 685, 788).

Breast cancer represents approximately 3% of all malignancies reported to the Cincinnati Transplant Tumor Registry and approximately 5% of the nonskin cancers reported to the Australia and New Zealand Combined Dialysis and Transplant Registry (647, 663). In the United States, the lifetime risk for women to develop breast cancer is approximately 1 in 8 (789). The estimated lifetime risk for women to die as a result of breast cancer is approximately 3.6% (790). Several explanations have been proposed for the lower incidence of breast cancer shortly after renal transplantation. It has been suggested that immunosuppression during a premalignant phase in breast neoplasia may reduce the incidence of subsequent development of breast cancer (791). Other groups have suggested that the lower incidence of breast cancer noted early after renal transplantation may be a direct consequence of increased examinations and screening before renal transplantation (788).

Consequences. Breast cancer is associated with high morbidity and mortality rates. In the general population, breast cancer is the leading contributor to cancer-related deaths for female patients in the 15- to 54-yr age group (790). The incidence of breast cancer increases with age, and more than one-half of all deaths attributable to breast cancer occur in women 65 yr of age (790). Breast cancer advances by local extension and also by metastasis to multiple sites, including the lymph nodes, bone, lungs, liver, and central nervous system, among other sites (792). Treatments include surgery, radiation therapy, chemotherapy, and hormonal therapy. Extensive disease is not curable with conventional therapy, and interventions are usually limited to palliation of symptoms (792). Renal transplant recipients with advanced breast carcinoma have very poor prognoses (793).

Rationale. Randomized controlled trials in the general population have demonstrated that screening reduces the mortality rate for breast cancer (789, 790). Screening mammography is associated with 20 to 30% reductions in the mortality rate for women 50 to 69 yr of age (789, 790). Screening mammography for women 40 to 49 yr of age also reduces the mortality rate, and the effect is statistically significant 15 yr after the beginning of screening (789, 794). Renal transplant recipients may develop aggressive breast cancer, and there is no reason to think that screening for breast cancer would not reduce the mortality rate for this population as well.

Breast self-examinations have limited sensitivity for the detection of breast carcinoma (790). Breast self-examinations are considered a supplement to but not a substitute for other screening tools available for the detection of breast carcinoma (789). Clinical breast examinations are useful for the detection of breast cancer when they are performed in addition to mammography (789). Some palpable breast cancers may not be visible in mammograms and may be detected early only in breast examinations (789).

The best test for the early detection of breast cancer is x-ray mammography. The sensitivity of mammography alone as a screening test for breast cancer is approximately 75%, and the sensitivity increases to 88% when mammography is used in combination with clinical breast examinations (790). The development of more advanced techniques for mammography has improved its sensitivity and specificity for the detection of breast carcinoma. There is a risk of false-positive results with breast cancer screening (790). A significant proportion of women without breast cancer require additional evaluation because of abnormal breast cancer screening test results (795).

Women with family histories of premenopausal breast cancer in a first-degree relative (brother, sister, or parent) and women with histories of breast and/or gynecologic cancers are at high risk for breast cancer. It is logical that such women warrant more aggressive screening, although there are few data to substantiate this. Several organizations have made detailed recommendations regarding screening for breast cancer (789, 790, 794). Applying these guidelines to the screening of women after renal transplantation seems to be prudent, although there have been no studies evaluating the effects of breast cancer screening in this population.

Colorectal Carcinomas (Table 37)
Definition. Adenocarcinomas constitute the majority of colorectal malignancies.

Incidence. Carcinomas of the colon and rectum occur in up to 0.7% of renal transplant recipients, with risks ranging from two- to threefold higher than those for the general population (685). Data from the EDTA-ERA Registry indicate that the incidence of colon cancer is not increased in the first 10 yr after renal transplantation. Nevertheless, after the first 10 yr after renal transplantation, there is a two- to fourfold increase in the risk of colon cancer for male and female renal transplant recipients (655). The incidence of carcinomas of the rectum does not seem to be increased among the renal transplant recipients monitored in the EDTA-ERA Registry. A recent report of a large number of transplant recipients also noted an increase in the incidence of colon cancer but a reduction in the incidence of rectal cancer among transplant recipients, compared with the general population (796). Carcinomas of the colon and rectum represent approximately 3.6% of the malignancies among transplant recipients reported to the Cincinnati Transplant Tumor Registry (647).

Consequences. Colorectal cancer is the second most common form of cancer in the general population in the United States, and it exhibits the second highest mortality rate (797). The lifetime risk of dying as a result of colorectal cancer for an individual in the United States is approximately 2.6% (797). Colon and rectal carcinomas can present with bleeding, abdominal pain, changes in bowel habits, or other manifestations. Cancers of the colon may spread directly from the bowel to regional lymph nodes and through the portal circulation to the liver. Metastases to the peritoneal cavity, lungs, and other distant sites can also occur (798). Renal transplant recipients with metastatic colon carcinomas have very poor prognoses (793). In some studies, the prognoses for renal transplant recipients with colorectal cancers appear to be similar to those for individuals in the general population with the same diagnosis (655).

Rationale. The estimated 5-yr survival rate for patients in the general population with localized colorectal carcinoma is 91%, but the rate decreases to <6% for patients with metastatic disease (797). Screening for colorectal carcinoma in the general population has been demonstrated to lead to the diagnosis of lesions that are smaller and more localized (799). Screening for rectal carcinoma using fecal occult blood testing and sigmoidoscopy has been demonstrated to lead to reductions in colorectal cancer-associated mortality rates. There are groups of individuals, such as those with some hereditary conditions, personal or family histories of other cancers, or inflammatory bowel disorders, who are at higher risk for colorectal carcinoma and benefit from screening for colorectal cancer (799). Renal transplant recipients may experience a higher incidence of colorectal carcinomas and should benefit from screening.

The digital rectal examination is of limited value as a screening test for colorectal cancer. Fewer than 10% of colorectal cancers can be palpated by the examining finger (797). Annual fecal occult blood testing for asymptomatic individuals decreases the mortality rate for colorectal cancer (797,799,800). The sensitivity and specificity of fecal occult blood testing for the detection of colorectal cancer in asymptomatic individuals are influenced by the protocols for collecting and testing samples (800,801). Flexible sigmoidoscopic screening of asymptomatic individuals has also been associated with reductions in the mortality rates for cancer of the rectum and distal colon (797,799,801,802). Sigmoidoscopy with a 60-cm endoscope can detect 65 to 75% of polyps and 40 to 65% of colorectal cancers (801,802).

Barium enema examinations allow observation of the entire colon and are associated with detection of 90 to 95% of polyps of 1 cm (803). Colonoscopy permits observation of the entire colon and has an estimated sensitivity of 75 to 95% for the detection of lesions within its reach (797). Colonoscopy has been associated with reductions in the incidence of colorectal cancer among cohorts of patients with adenomatous polyps (800). There have been no studies evaluating whether barium enema or colonoscopic examinations reduce mortality rates for individuals of average risk. Costs and potential complications are associated with barium enema and colonoscopic examinations (797,800).

There is no reason to think that evidence-based recommendations for the general population would be any less effective for renal transplant recipients. Guidelines for screening and surveillance for the early detection of colorectal cancer have been developed by the National Cancer Institute, the United States Preventive Services Task Force, the American Cancer Society, and the American College of Gastroenterology (797,799,800,801).

Prostate Cancer (Table 38)
Definition. Prostate cancer is now the second most common cause of cancer-related deaths among American men (804). Although it has been underdiagnosed for many years, there has been a recent appreciation of the importance of prostate cancer among renal transplant recipients (805).

Incidence. Prostate cancer is the second most common cause of cancer-related deaths among men in the general population in the United States (806). Prostate cancer has likely been underdiagnosed among male renal transplant recipients. The incidences of prostate cancer among renal transplant recipients range from 0.3 to 1.9% (685,805,807). As the number of older men with functioning renal allografts continues to grow, it is expected that the incidence of prostate cancer will exhibit similar increases. Prostate cancer represents approximately 1.9% of the total number of malignancies reported to the Cincinnati Transplant Tumor Registry (805).

Consequences. Prostate cancer is associated with morbidity and death in the general population. Clinical manifestations include urinary frequency, dysuria, hematuria, and urinary retention. As the disease progresses, local spreading and then metastatic disease, with bone pain and widespread dissemination, occur (807). In the general population, reported 10-yr survival rates for patients with prostate cancer are 75% for patients whose cancers are confined to the prostate, 55% for patients with regional extension, and 15% for patients with distant metastases (804). In a recent report on renal transplant recipients, the mortality rates were 13% for patients with localized prostate cancer at the time of presentation and 33% for patients with metastatic disease at the time of diagnosis. There was a median follow-up period of 33 mo for renal transplant recipients in that study (805).

The treatment modalities that have been used for renal transplant recipients with prostate cancer include radical prostatectomy, radiation therapy, and hormonal therapy. Therapeutic limitations include the difficulty of performing pelvic lymph node dissections on the side of the renal allograft. Radiation therapy could also be associated with nephritis and damage to the renal allograft, and shielding of the allograft is recommended.

Rationale. Prostate cancer is one of the most important causes of cancer-related deaths in the general population. Prostate cancers that are localized exhibit better responses to therapy and are associated with lower cancer-related mortality rates than are extensive cancers. In a recent report, it was noted that 84% of renal transplant recipients diagnosed with prostate cancer had localized disease and 16% of patients had metastatic disease at the time their prostate cancers were diagnosed (805). Patients with localized disease exhibited favorable responses to therapy, but patients with advanced metastatic disease exhibited rates of disease progression and treatment failures that appeared to be higher than those in the general population (805).

Digital rectal examinations have been reported to have sensitivities ranging from as low as 18% to as high as 68% for the detection of prostate cancer among asymptomatic men (804). The positive predictive value of digital rectal examinations is higher when the examinations are performed by urologists (804).

Serum prostate-specific antigen testing has a sensitivity of approximately 70% for the detection of prostate cancer among asymptomatic men (806). It is a simple, noninvasive, reproducible test. Several conditions, such as prostatitis and benign prostatic hypertrophy, can lead to elevations in levels. The positive predictive values reported for the measurement of prostate-specific antigen levels as a screening test for prostate cancer range between 26 and 52% (804). The prostate-specific antigen test appears to be valid for the early detection of prostate cancer after renal transplantation (808).

Transrectal ultrasonography has a reported sensitivity of 57 to 68% for the detection of prostate cancer among asymptomatic men (804). Transrectal ultrasonography cannot distinguish between benign and malignant nodules and has a lower positive predictive value for prostate cancer, compared with prostate-specific antigen measurements (804). Transrectal ultrasonography is usually reserved for patients with digital rectal examination abnormalities and/or elevations in serum prostate-specific antigen levels. Transrectal needle biopsy of the prostate is an invasive test that is reserved for patients with digital rectal examination abnormalities and elevations in prostate-specific antigen levels. It is not an accepted screening test for prostate cancer. CT scanning and MRI are probably not cost-effective screening tests for prostate cancer.

There is no consensus on whether to recommend routine screening for prostate cancer for men in the general population. Several organizations and expert panels have provided different recommendations regarding the potential benefits, risks, and cost-effectiveness of screening for prostate cancer (804,806,809,810). There have been insufficient data to conclusively demonstrate decreases in the mortality rate for prostate cancer as a result of screening (804,806,810). Nevertheless, patients with estimated life expectancies of 10 yr may benefit from screening (809).

Given the potential effects of prostate cancer among renal transplant recipients, it seems prudent to screen for prostate cancer, using digital rectal examinations and prostate-specific antigen measurements, yearly starting at age 50 yr for men whose life expectancy is 10 yr. It is also reasonable to initiate screening at an earlier age (such as 40 yr) for men with special risk factors, such as black race and family histories of prostate cancer. Physicians and health care professionals should discuss with each patient the potential benefits and limitations of the screening, diagnosis, and treatments for prostate cancer before screening is initiated (810).

Lung Cancer (Table 39)
Definition. Lung cancer is a leading cause of cancer-related deaths for both men and women in the United States. It is associated with very poor prognoses.

Incidence. The incidence of lung cancer reported for renal transplant recipients has ranged from <0.5% to approximately 0.7% in some studies (685,811). In the EDTA-ERA Registry, the lung was the site most commonly affected by cancer in male patients, although the overall incidence of lung cancer was lower among renal transplant recipients than in the general population (655). It was thought that exclusion of smokers with advanced chronic obstructive lung disease from renal transplantation could partly explain these differences (655).

Lung cancers represent approximately 5.5% of all malignancies among organ transplant recipients that were reported to the Cincinnati Transplant Tumor Registry and 9% of nonskin cancers among renal allograft recipients that were reported to the Australia and New Zealand Combined Dialysis and Transplant Registry (647,663). Lung cancers seem to develop at a younger age for transplant recipients, compared with the general population (811). Small cell cancers were also more common in that study (811). Smoking increases the risk of lung cancer among renal transplant recipients (812).

Consequences. Patients with lung cancer usually present with symptoms related to involvement of the respiratory tract (813). Metastases to distant sites and paraneoplastic syndromes also develop in many patients (813). Cancer of the lung is the leading cause of cancer-related deaths for men and women in the United States. The 5-yr survival rate for patients with lung cancer in the general population is <13% (814). Renal transplant recipients in the EDTA-ERA Registry who developed lung cancer exhibited a 5-yr survival rate of <15% (655).

Rationale. Lung cancer is associated with very high mortality rates for the general population and for renal transplant recipients. Lung cancer has one of the poorest prognoses of all cancers and can be cured only when detected and surgically resected in its early stages (814). Lung cancer usually remains asymptomatic until its advanced stages (814). However, screening for lung cancer among asymptomatic individuals, using chest radiographic examinations or sputum cytologic analyses, has not been demonstrated to reduce mortality rates and is not recommended for the general population (814). Chest x-rays have been extensively studied as a screening test for lung cancer in the general population. By the time lung cancer is indicated in chest x-rays, metastatic dissemination has often occurred, which limits the effectiveness of chest x-rays as a tool for the early detection for lung cancer (814). Sputum cytologic analysis has also been studied as a screening test in the general population and appears to be even less effective than chest x-rays. The sensitivity of sputum cytologic analysis for the detection of lung cancer may be as low as 10% (814). A recent study noted that the use of low-radiation dose chest CT for people in the general population at high risk for lung cancer greatly improved the likelihood of detecting lung cancer at an early, potentially curable stage (815). That study led to renewed controversy regarding previous guidelines, which generally recommended against any screening for lung cancer (816,817). Whether low-dose CT may be effective in screening high-risk renal transplant recipients, e.g., cigarette smokers, is unclear.

	VIII. Infections

Top Abstract Introduction Scope Intended Users Materials and Methods Organizational Scheme I. Frequency and Timing... II. Graft Function III. Immunosuppressive... IV. Cardiovascular Disease V. Bone and Bone... VI. Nutrition and Metabolism VII. Cancers VIII. Infections Endorsements References

 
Cytomegalovirus (Table 40)
Incidence. The incidence of CMV disease is generally <5% for recipients who exhibited no prior serologic evidence of CMV (antibodies) [R(-)] and who received kidneys from donors who were also antibody-negative [D(-)] (818). The incidence of primary CMV disease among R(-)/D(+) recipients is approximately 50 to 75% (818). The incidence among R(+)/D(+) or R(+)/D(-) recipients is approximately 25 to 40% (818). The incidence of CMV disease is also influenced by the level of immunosuppressive therapy used. In one prospective study, 37 R(-)/D(-) recipients exhibited no CMV disease, regardless of whether anti-CD3 antibody (Orthoclone OKT3; Ortho Biotec, Raritan, NJ) was used. However, 16 of 45 R(+)/D(±) recipients (36%) developed CMV, and in that group the risk of CMV disease was fivefold greater (odds ratio, 5.2; 95% confidence interval, 1.4 to 17.5) for patients who received OKT3 (819). High doses of other immunosuppressive medications likely also increase the risk for CMV. For example, in the European Mycophenolate Mofetil Cooperative Study, the incidence of CMV was 36% (10 of 28 patients) for patients who received a high dose of MMF (3 g/d), compared with 8% for the 54 patients who received either placebo or 2 g/d MMF (820).

Consequences. CMV also causes considerable morbidity and expense. Acute CMV infection is often manifest as fever, leukopenia, thrombocytopenia, myalgias, and flu-like symptoms. End-organ involvement may cause nephritis, retinitis, hepatitis, gastrointestinal bleeding, and/or pneumonia. CMV may also predispose recipients to other opportunistic superinfections and graft rejection. Indeed, CMV infection is associated with reduced patient and graft survival rates (821). Among 47,146 patients in the UNOS registry, kidneys from CMV-positive donors demonstrated approximately 4% lower graft survival rates at 3 yr, compared with kidneys from CMV-negative donors (822).

Rationale. The prevention of CMV remains a major goal in kidney transplantation, and guidelines have been developed for CMV prophylaxis in the early posttransplant period (818). In theory, screening should reliably identify patients destined to develop symptomatic CMV, thereby allowing the selective use of antiviral therapy to prevent CMV (823). Selective treatment of high-risk patients, rather than indiscriminate treatment of all transplant recipients, could reduce cost, avoid adverse drug reactions, and help prevent the emergence of resistant CMV strains (824). Antibody titers are useful in assessing the risk of CMV infection at the time of transplantation, by establishing whether the donor and/or recipient were previously infected with CMV. Antibody titers and traditional viral cultures are useful for diagnosing CMV infection after its development but do not change quickly enough to be used to screen for impending CMV infection (825,826).

The use of a monoclonal antibody against CMV early antigen in a shell vial culture assay was developed for rapid detection of CMV viremia. However, this assay was supplanted by assays using antibodies against viral immunodominant matrix protein pp65 (detected in peripheral blood leukocytes), which not only is more rapid but also is quantitative (825,827). For example, in a prospective study of 64 patients, CMV pp65 antigenemia testing exhibited a sensitivity of 87.5% and a specificity of 92.5% and detected antigenemia 8 d before the onset of symptoms (825). Recently, flow cytometry has been used to quantify the number of pp65-positive cells in peripheral blood and thereby detect early antigenemia (828).

PCR techniques to detect CMV DNA in blood are extremely sensitive (829,830,831,832). The PCR assay has been further enhanced by reverse transcription and other techniques that quantify CMV RNA and distinguish latent from active infection. The nearly 100% sensitivity and specificity of PCR assays using leukocytes made it possible to detect CMV in the blood 8 to 13 d before the onset of symptoms for 37 recipients who were seronegative at the time of transplantation (829). These assays were less predictive for recipients who were seropositive at the time of transplantation (829). Non-PCR methods are very specific but are not quite as sensitive as PCR for the detection of viremia before the onset of symptoms (829).

Therefore, techniques that use specific monoclonal antibodies against CMV and techniques that quantify CMV DNA levels in peripheral blood leukocytes appear to be sensitive and specific. However, they have not yet been compared in large prospective trials that accurately assess their positive and negative predictive values for populations with different risks for CMV. In addition, whether these tests will be cost-effective in identifying patients for preemptive therapy has not been adequately studied. In short, there are currently insufficient data to justify the widespread clinical application of any posttransplant CMV screening.

Chemoprophylaxis for CMV has been demonstrated to be effective in large, randomized, controlled trials. Agents that have been studied include acyclovir, ganciclovir, valacyclovir, and intravenously administered Ig. Chemoprophylaxis should be used according to the risk for CMV, as recommended by existing guidelines (818).

Influenza A and B (Table 41)
Incidence. Few studies have reported the incidence of influenza among renal transplant recipients, but there is no reason to think that the incidence is less than that in the general population and there are many reasons to suspect that it may be higher. The level of immunosuppressive therapy that patients receive may increase the risk for influenza. In one study of 54 patients who received antilymphocyte antibody therapy, 2 subsequently developed influenza (833).

Consequences. It is likely that influenza is more severe when it occurs among immunosuppressed renal transplant recipients, although systematic studies documenting this finding are lacking. Influenza can be fatal. Bacterial superinfections are common. In a study of 12 pediatric organ transplant recipients with influenza B (5 of whom were kidney recipients), 5 exhibited neurologic involvement and 1 died; 10 were hospitalized and 2 required mechanical ventilation (834).

Rationale. It is likely (albeit unproven) that influenza infection is at least as common and probably more severe among renal transplant recipients, compared with the general population. For the general population, studies have clearly demonstrated that influenza vaccination reduces infection and its consequences (835). Several studies examined the safety and efficacy (based on increases in antibody titers) of influenza vaccination among adult and pediatric renal transplant recipients. Most studies reported that 50 to 100% of the recipients responded to vaccination. The response rate was often less than that of control subjects but was nevertheless adequate to justify the use of vaccination for renal transplant recipients (836,837,838,839,840,841,842,843,844,845,846,847). In one study, no benefit of a second dose was observed (848). In most of the studies, the numbers of patients were too small to reliably indicate whether some patients were more likely to respond than others. However, one study reported that patients receiving CsA were less likely to respond than patients receiving azathioprine (840), and another study reported that patients with decreased renal function were less likely to respond (846). In none of the studies was there an increase in acute rejections or other major adverse effects. Data on the efficacy of vaccination in studies including other organ transplant recipients corroborate the efficacy and safety of vaccination (848,849,850,851,852,853,854). One study of 68 (kidney and other) organ transplant recipients confirmed that a second or third injection did not increase the response rate (848). In a recent survey of pediatric renal transplant centers, almost 90% recommended influenza vaccination (855).

A recent randomized controlled trial among 20 hospitals demonstrated that offering influenza vaccinations to health care workers increased the rate of vaccination from 4.9% in control sites to 50.9% in treatment sites (856). In hospitals where health care workers received vaccinations, the mortality rate among patients was significantly reduced (856). A similar trial also reported reduced mortality rates with the vaccination of health care workers (857). Therefore, it is reasonable to conclude that the vaccination of health care workers who are in close contact with renal transplant recipients may reduce the incidence of influenza-related morbidity and death not only among the health care workers but also among the patients for whom they care.

Amantadine and rimantadine hydrochloride are approximately 70 to 90% effective against influenza A in the general population. Neither is effective against influenza B, and neither should be considered a substitute for vaccination (858). Rimantadine appears to be associated with fewer central nervous system side effects than amantadine, but it is more expensive. These drugs may also be effective when administered within 48 h after illness onset and can reduce the severity and duration of symptoms without inhibiting the development of immunity. A recent randomized controlled trial also demonstrated that the neuraminidase inhibitor zanamivir is efficacious and well tolerated for the prevention of influenza in the general population (859). Zanamivir, administered once daily by inhalation, has the advantage of being effective against both influenza A and B. Like amantadine and rimantadine, zanamivir can also reduce the duration and severity of illness if therapy is initiated early (within 36 to 48 h) after the onset of symptoms (860,861,862). A 5-d course of zanamivir therapy was also effective in preventing infection among patients with presumed exposure to influenza in the community (863,864,865). Finally, in two randomized controlled trials, a 6-wk course of oseltamivir (an oral neuraminidase inhibitor that is active against influenza A and B) therapy was effective in preventing influenza in the community (866). Oseltamivir was also effective if administered within 36 h after the onset of influenza symptoms (867) or after experimental influenza A inoculation (868). None of the four antiviral agents discussed above has been demonstrated to prevent serious influenza-related complications.

The Advisory Committee on Immunization Practices recommends that amantadine or rimantadine be considered for patients at high risk who are vaccinated after influenza activity has begun and that therapy be continued for approximately 2 wk, until immunity has developed (858). In the United States, zanamivir and oseltamivir are currently licensed for treatment but not for prophylaxis.) Chemoprophylaxis should also be considered for patients who cannot be treated with the influenza vaccine, e.g., patients with histories of egg allergies and individuals at high risk (including patients receiving immunosuppressive medications) who are expected to exhibit an inadequate antibody response to influenza vaccine. Chemoprophylaxis should be administered only during the influenza season. There are no data for renal transplant recipients to suggest which patients should receive chemoprophylaxis. A reasonable strategy would be to limit the use of amantadine, rimantadine, zanamivir, or oseltamivir to the influenza season and to patients who have not yet been vaccinated or are receiving high doses of immunosuppressive medications, i.e., patients who may not develop immunity despite vaccination (858). The inappropriate use of antiviral agents could contribute to the emergence of resistant strains and should be discouraged (869). Effective timing of the use of neuraminidase inhibitors requires physicians to be aware of outbreaks of influenza in their communities. This information is reported by the Centers for Disease Control and Prevention and is updated weekly (http://www.cdc.gov/ncidod/diseases/flu/fluvirus.htm).

Tuberculosis (Table 42)
Incidence. Few studies have documented the incidence of tuberculosis (TB) among renal transplant recipients in North America. Among 565 renal allograft recipients in the United States, primary TB developed in 5 (0.9%), whereas none of 14 high-risk patients who were prophylactically treated with isoniazid developed TB (870). In England, a retrospective study reported TB for 11 of 633 renal transplant recipients (1.7%) (871). A similar study in England reported TB for 5 of 400 renal transplant recipients (1.3%) (872). A retrospective study of organ transplant recipients in Spain demonstrated that the incidence of TB was 0.8% (873). Of the 51 cases described, 63% involved pulmonary, 25% disseminated, and 12% extrapulmonary TB. Infection developed at any time after transplantation, but the mean time to diagnosis was 23 mo (range, 15 d to 13 yr). In another study of renal transplant recipients in Spain, 22 of 525 patients (4.2%) with new cases of TB represented a rate of 259/100,000, which was sevenfold greater than the rate of 35/100,000 for the general Spanish population (874). In a retrospective study from South Africa, there were 21 of 487 renal transplant recipients (4.3%) with confirmed TB (875). The median time to diagnosis was 14 mo (range, 2 to 74 mo) (875). In Turkey, TB was reported for 36 of 880 patients (4.1%) (876). A retrospective review from Saudi Arabia reported TB for 14 of 403 renal transplant recipients (3.5%), which yields an annual incidence 50-fold higher than that in the general population (877). Infection appears to be more common after the intensification of immunosuppressive therapy (873,875). Nosocomial outbreaks may also occur in transplant units (878). Transmission of TB with allografts has been reported (879,880).

Consequences. TB is often fatal. In one large retrospective study, 16 of 51 patients (31%) died (873). In another study, 4 of 21 patients (19%) died while receiving active treatment, but only one of these deaths was directly attributable to TB (875). Disseminated disease often involves the liver, spleen, and bone marrow.

Rationale. For the treatment of latent infection to be effective, the risk of active TB should be high, compared with the risk of treatment. There are no adequately powered, randomized, controlled trials among transplant recipients to determine the efficacy of treatment for latent infection. In one double-blind controlled trial from India, 184 patients undergoing dialysis were randomly allocated to receive isoniazid or placebo, and the administration of drug or placebo was continued for 1 yr after transplantation (881). Hepatitis developed in 33 of 92 placebo-treated patients (36%) and in 32 of 92 isoniazidtreated patients (35%). Only 39 of 92 patients (42%) in the placebo-treated group and 34 of 92 patients (37%) in the isoniazid-treated group completed therapy. During the first 1 yr of follow-up monitoring, 7 patients in the isoniazid-treated group and 10 in the placebo-treated group developed TB. However, only three of the patients in the isoniazid-treated group who developed TB had completed treatment. During the second year, four patients in the placebo-treated group and three in the isoniazid-treated group developed TB (881). Because of the large number of patients who failed to complete therapy, the results of this trial were inconclusive. In one retrospective study, there were no cases of TB among high-risk patients who received treatment, but 6 of 27 high-risk patients (22%) who did not receive treatment developed TB (871). In another retrospective study, none of the 14 high-risk patients treated with isoniazid developed TB (870). In a similar study, none of 23 high-risk patients who received isoniazid developed TB, whereas 1 of 13 patients who did not receive isoniazid developed TB (876).