 |
||||
| 2007 JASN IMPACT FACTOR 7.111 | HOME AUTHOR INFO EDITORIAL BOARD SUBSCRIBE FEEDBACK ALERTS HELP | |||
| CURRENT ISSUE | ARCHIVES | JASN Express | ONLINE SUBMISSION | |
|
||||
| 2007 JASN IMPACT FACTOR 7.111 | HOME AUTHOR INFO EDITORIAL BOARD SUBSCRIBE FEEDBACK ALERTS HELP | |||
| CURRENT ISSUE | ARCHIVES | JASN Express | ONLINE SUBMISSION | |
Nephrology beyond JASN |
It had been known since the early days of Franz Volhard (1) and Arthur Fishberg (2) that renal disease and renal failure occur commonly in hypertensive patients. In the study of Perera (3), at a time when antihypertensive medication had not yet become available, a large proportion of patients with essential hypertension wound up in renal failure. This was later ascribed to the occurrence of malignant hypertension. With the advent of effective antihypertensive medication, malignant hypertension has become much rarer. In relatively short-term trials that clearly documented the cardiovascular benefit from antihypertensive medication, few if any cases with renal failure were observed. This led to an as yet unresolved dilemma.
There are authors who made strong statements such as: there are "no reported cases of benign essential hypertensive patients with normal serum creatinine levels and no proteinuria who subsequently went on to develop renal failure" (4). On the other hand, in the US (5,6) and elsewhere, but with remarkable differences between countries (7), a high proportion of cases reaching end-stage renal disease with hypertension and a nondiagnostic clinical course are given the diagnosis of "hypertensive nephropathy" in the absence of renal biopsy and other more in-depth investigations.
What was the evidence available so far and what were its shortcomings?
In two long-term studies in the US, Perry (8) and Klag (9) had followed patients in a VA study and the MRFIT study without known renal disease for more than a decade. Both authors noted that progressively higher BP values at baseline were significantly associated with a graded increase of the risk of end-stage renal disease. A study in Okinawa also found that over 18 yr a similar relationship between baseline BP and late onset of end-stage renal disease extending down into the range of values that before the reports VI and VII of the Joint National Committee was regarded as "normotension" (10). Finally Fox, in a small, population-based, 12-yr study, observed that in hypertensive as compared with normotensive patients the risk of reaching a more advanced stage of renal disease was higher by 76% (11).
All these studies, however, had one great shortcoming, i.e., it was not known whether the study participants had not some renal disease at the start of the study, as frankly acknowledged by Klag (9) who stated we "do not know whether renal insufficency was already present in those in whom end-stage renal disease later developed" and "lack of information on renal function, both at baseline and during follow-up, does mean...that we cannot say definitely whether the strong association between BP and the incidence of end-stage renal disease was due to the initiation of renal disease or to the accelerated progression of preexisting renal disease.."
It is here that the study of Hsu, based on the inexhaustible source of the Kaiser Permanente data, is a most valuable step forward. The authors studied a total of 316,675 voluntary participants in the Multiphasic Health Testing Service Program between 1964 and 1985 whose estimated GFR was >60 ml/min per 1.73 m2 and in whom urine testing for proteinuria and hematuria had been negative. Analyzable data included measured BP, serum creatinine concentration, and urine dipstick analysis, which were available for 3 out of 4 of the program’s check-ups. Up to December 31, 2000, 1149 participants had reached end-stage renal disease, i.e., renal transplantation or dialysis dependency—a rare event, but strikingly dependent on baseline BP in these individuals in whom renal disease at baseline had been excluded with accepted methodology (although tests for albuminuria had not been performed). The age-adjusted relationship between BP and risk of end-stage renal disease was seen in all subgroups, although the absolute risk—not surprisingly—varied widely. In individuals with a BP <120/80 mmHg, the age-adjusted rate of end-stage renal disease was only 2.8 per 100,000 person-years for white and 14 for black subjects; as anticipated even in this low BP group the rate was higher in diabetic (12.7) as compared with nondiabetic individuals(3.8). Of note, the risk persisted when adjustments were made for items such as age, gender, race, smoking status, diabetes mellitus, weight, or history of myocardial infarction. Compared with that in individuals with BP <120/80 mmHg, the relative risk of end-stage renal disease rose from a 1.6-fold increase (for individuals with BP 120 to 129/80 to 84 mmHg) to a 4.2 fold increase (for individuals with a BP >210/120 mmHg).
This is the best evidence available to date that high BP precedes the onset of renal disease other than malignant hypertension. It is superior to preceding studies (8–11) by providing information on renal findings at baseline excluding pre-existent renal disease. Similar to the finding reported by Tozawa (10), the study was sufficiently powered to document that the relationship between renal risk and BP extended down into the range of what was previously considered "normotensive" and today is somewhat unfortunately called "prehypertensive" BP values.
By studying survivors, the study may have underestimated, however, the renal risk because cardiovascular mortality of renal patients is excessive even in early stages of chronic kidney disease before end-stage renal disease has been reached (12). This has recently also been confirmed in the Kaiser Permanente program (13).
Of course, the gold standard study would be a prospective, randomized intervention trial—but that type of investigation will never be done for logistic reasons. It is remarkable that one intervention study failed to observe a reduction of serum creatinine increase despite adequate BP control in treated (primarily black) patients with essential hypertension with initially normal serum creatinine concentration (14). Similarly a recent meta-analysis of several small and underwhelming studies had not shown definite benefit from antihypertensive treatment (15). There may be many explanations: the BP targets may not have been sufficiently low, the antihypertensive agents selected were not always state of the art, treatment may have started late. For instance, when hypertension-associated "benign nephrosclerosis" has developed, renal dysfunction tends to progress inexorably as suggested by the Norwegian experience (16). Nevertheless, these negative findings remain irritating and justify further investigation. Certainly they do not justify therapeutic nihilism, however, because one decisive, large, well-controlled study had clearly documented the benefit of antihypertensive treatment on progression (17,18), particularly when the renin-angiotensin-system is blocked (19).
The issue remains unresolved why end-stage renal disease supervenes in a very small minority of patients with very moderate elevation of BP. It is easier to explain why high BP is injurious to the kidney than to explain why only such a minute proportion of hypertensive patients develop end-stage renal disease. There are sufficient potential pathomechanisms; to mention just one, hypertension causes vascular lesions in the kidney (20,21), and as a result activates the renin-angiotensin system (21). Nevertheless it appears that hypertension must meet an organism that is somehow predisposed to develop renal lesions, perhaps because of genetic factors. This hypothesis would provide a potential explanation why hypertension is more common in families of patients with IgA-glomerulonephritis (22) or diabetic nephropathy (23). A conceptual framework, which would link hypertension to a predisposition of glomeruli to injury, has been proposed by Brenner. He postulated that a low number of nephrons predisposes not only to hypertension, but also to a greater susceptibility to renal injury (24). At least as far as the postulated lower nephron endowment in essential hypertension is concerned, this hypothesis has been confirmed (25).
At the end of the day we still don’t have a clear understanding why even minor elevations of BP increase the risk of end-stage renal disease. This fascinating problem with considerable public health implications will undoubtedly keep nephrologists busy in the years to come.
|
 |
Footnotes |
|---|
|
References |
|---|
 Top References References References |
|---|
Cardiac hypertrophy is a frequent complication of renal malfunction. In essential hypertension left ventricular hypertrophy (LVH) predicts cardiovascular death (1) BP independently. In dialysed uremic patients, Silberberg was the first to show that LVH predicted survival (2) and this has been confirmed in virtually all subsequent studies (3). In patients with renal disease, LV remodelling and diastolic malfunction are seen even before GFR is decreased (4). The prevalence of LVH increases progressively as renal function declines (5,6). At the start of dialysis a normal echocardiogram was found in no more than 16%, concentric hypertrophy in 41%, and LV dilation in 28% (7). Median time to the development of heart failure, with dire prognostic implications (8), was 48 mo in patients with concentric LVH compared with >66 mo in patients with normal echocardiograms (7). Routine echcardiograms may underestimate the functional impact, and the midwall fractional shortening/end systolic stress relation is a more sensitive and common indicator of systolic malfunction in concentric LVH (9).
The genesis of cardiac hypertrophy in uremia is certainly multifactorial. Pressure overload (increased peripheral resistance associated with hypertension and vascular remodelling) and impedance (from increased stiffness of central arteries) as well as volume overload (salt-water retention, anemia, arteriovenous shunt) play a role (10). Undoubtedly, however, systemic factors independent of pressure and volume overload must also play a role, presumably sensitizing the heart to the impact of the classic LVH risk factors. We concluded this from the fact that experimentally one finds hypertrophy not only of the left, but also of the right ventricle and that interventions including abrogation of hypertension and volume overload failed to completely prevent LVH (11). Such aggravating systemic factors presumably include among others neurohumoral activation, particularly sympathetic overactivity, hyperparathyroidism, and oxidative stress.
Why oxidative stress? In recent years there has been increasing evidence that reactive oxygen species (ROS) play a role in the genesis of cardiac hypertrophy and the progression of heart failure (12,13). Furthermore, interventions such as administration of vitamin E (14) have been shown to ameliorate LVH. Conversely, disruption of the potent antioxidant thioredoxin-1 system by creating a dominant negative mouse mutant caused oxidative stress in the heart and more severe cardiac hypertrophy in response to aortic banding. This was abrogated by administration of an antioxidant (15).
It had been known that generation of ROS may be the result of agonists such as catecholamines (16) or angiotensin II (17). It was also known that ROS mediate the hypertrophic response to mechanical stretch (18), induce reexpression of the fetal gene program (19), and trigger cardiac remodelling by activating metalloproteinases (19). What had not been defined, however, was the exact pathway through which ROS were generated in hearts developing hypertrophy. Candidates were the mitochondrial electron transport system (20), NADPH oxidase (21), xanthin oxidase (22), and nitric oxide synthase (NOS) (23).
This field has now been carried further by the important study of Takimoto (24), who identified NOS as the crucial pathway for the generation of oxidative stress in the hypertrophying heart and who also found an intervention which might provide novel perspectives for intervention.
The authors generated knockout mice lacking the endothelial isoform NOS-3 (previously called eNOS) of NOS. In response to transverse aortic constriction wild-type mice developed progressive dilatory LV remodelling, an increase in cardiomyocyte size and fibrosis; in the NOS-3 knockout mice, aortic banding caused only modest initial nonprogressive concentric hypertrophy with little fibrosis. In wild-type mice, impaired systolic (dP/dt) and diastolic (rate of pressure decline) function was noted, but there was little change in the NOS-3 knockout mice. Striking differences between these groups of animals were also seen with respect to the expression of fetal genes in the heart. The role of ROS was proven by measuring NO production via a luminal chemiluminescence assay and by staining for and measurement of nitrotyrosine, a downstream ROS reaction product. Nitrotyrosine was minimally elevated in the NOS-3 knockout mice, but strikingly so in the wild-type mice.
The finding of next to no ROS production in animals deficient in NOS identified this enzyme as the mediator of ROS generation in the model of LV hypertrophy in response to aortic banding.
But how does aortic banding increase the activity of NOS-3 in the wild-type mice? NOS is normally present as a dimer that transforms L-arginine into the products NO and L-citrulline. If it is present as a monomer and is uncoupled, however, the enzyme generates superoxide (O2–) instead of NO. This transition from the monomeric to the uncoupled dimeric form is seen when the enzyme is exposed to peroxynitrite (ONOO–) or deprived of the cofactor tetrahydrobiopterin (BH4) or the substrate, i.e., L-arginine, respectively.
Therefore the authors studied immuneprecipitated NOS3 by nondenaturing gel electrophoresis. They were able to show that after aortic banding the enzyme was present in the monomeric form. They also found the reason why: They measured the concentration of the cofactor BH4 and it was low.
The logical next step was to administer BH4. This blunted the concentric hypertrophic response to aortic banding. If the same is true in humans, administration of BH4 might well provide a novel form of intervention.
These findings are of obvious interest against the background of the epidemiology of LVH in renal patients (2–9). There is, however, an added consideration. Uremia is a state of increased oxidative stress (25–27), as reflected by a number of indicators such as plasma aminothiol oxidation (28), lipid peroxidation (29), appearance of advanced oxidation protein products (30), or formation of the DNA damage product 8-hydroxy 2'-deoxyguanosine (31). The relevance of ROS is underlined by observations that advanced oxidation protein products predict cardiovascular events in patients with advanced renal failure (30) and that interventions to reduce oxidative stress such as vitamin E (d,l-
-tocopherol) (32) or N-acetylcysteine (33) reduce cardiac and atherosclerotic abnormalities in experimental uremia.
This raises the issue whether the uremic organism might not also be particularly susceptible to the cardiac effects of uncoupled NO synthase and furthermore whether evidence of cardiac hypertrophy in the absence of pressure and volume overload (11) might not be the consequence of uncoupled NO synthase. Finally tantalizingly incomplete evidence had pointed to potential cardiovascular benefit from administration of vitamin E (34) and N-acetycysteine (35) in hemodialysed patients. Controlled trials in this field are obviously needed. A point that should now also be considered, however, is whether BH4 should be added as another candidate. Obviously further experimental and pharmacologic information is needed—but the potential prospect is fascinating.
|
References |
|---|
|
TopReferencesReferences References |
|---|
Despite hyperglycemia, not all diabetic patients develop nephropathy. The risk of developing nephropathy clusters within families, both in type 1 (1) and type 2 diabetes (2–5). Clustering of cardiovascular events and higher cardiovascular mortality have been observed in families with diabetes and nephropathy (6,7). Many genes have been considered as candidate genes for the development of nephropathy in diabetes: An association was observed between parental hypertension and nephropathy in offspring with type 1 diabetes (8–10) and higher BP were found in offspring of type 2 diabetic parents with nephropathy (5). Great interest had initially been focused on the D/I polymorphism of the angiotensin-converting enzyme (ACE) gene (11), based also on suggestive animal experiments (12), but a meta-analysis showed at best a modest correlation to the risk of nephropathy (13). Further interest had been generated by the sodium-lithium countertransport system (14), which predicts onset of diabetic nephropathy (15), but again, no evidence of genetic linkage had been forthcoming. More recent candidate genes were adiponectin (16), PPAR
(17) and others.
Apart from looking at "candidate genes," a strategy which is based on considerations of a priori plausibility and pathomechanistic hypotheses, another approach is to scan the genome for evidence that certain loci are associated with a given disease. In 2002 Vardarli et al. had reported that a locus on the long arm of chromosome 18 (18q22.3-23) was associated with higher susceptibility for diabetic nephropathy in 18 Turkish families (18), but the relevant gene had not been identified. Although other studies found, apart from several other loci, also a high logarithm of the odds (lod) score (an index of association) for this specific locus (19), the findings were not consistent in all studies (20).
The study by Janssen et al. now showed, however, not only linkage to the locus on chromosome 18, but provided also plausible evidence for the involvement of one specific enzyme as well as a presumed pathomechanism explaining the link to diabetic nephropathy via two well-known culprits in the genesis of this condition, i.e., oxidative stress and advanced glycation end products (AGE).
The authors identified carnosinase, i.e., the enzyme that degrades the dipeptide 
-alanyl-L-histidine, as the culprit. Two isoforms of this enzyme exist, and they are coded for at this locus on chromosome 18 in a head-to-tail configuration: one isoenzyme is secreted and demonstrable in the circulation, it has 12 exons, and it is coded for by the gene CNDP1; the other is located intracellularly, it also has 12 exons, and it is coded for by the gene CNDP2.
The authors screened 5500 diabetic patients and selected for genetic analysis 135 patients with long-lasting diabetes and diabetic nephropathy according to stringent criteria, and 107 diabetic patients without diabetic nephropathy.
The first step was to look for gross truncating deletions by complete sequencing analysis in a limited number of patients, but none were found.
In a second step, an exploratory analysis, they looked for associations with polymorphisms and identified a trinucleotide sequence in the 5' region, i.e., exon 2 of CNDP1. This trinucleotide sequence, coding for leucine in the leader sequence, yielded a string of 5, 6, or 7 leucines. The allele with 5 leucines was more frequent in diabetic patients without nephropathy (88%) than in those with nephropathy (59%).
In a further step, the authors investigated whether homozygosity for the apparently protective allele was more frequent in diabetic patients without nephropathy, and this was indeed the case: Homozygosity was found in 25 of 63 type 2 diabetic patients without diabetic nephropathy, but only in 20 of 77 diabetic patients with nephropathy, and the same was found in patients with type 1 diabetes.
The next step was to measure serum carnosinase activity, and it was indeed lowest in patients homozygous for the carriers of the gene with 5 trinucleotides. The serum carnosine concentrations were all extremely low, but this may not reflect the local concentrations at target organs, specifically the kidney, where the enzyme is also expressed, particularly in podocytes as documented in orientating studies.
The findings suggest the following sequence: Low carnosinase activity in carriers of the favorable low trinucleotide sequence, particularly in the homozygous constellation, causes high carnosine levels, and the latter are protective. If this is true, addition of carnosine to renal cells should prevent processes thought to be relevant in the genesis of diabetic nephropathy. To prove this point, the authors added carnosine to mesangial cells or podocytes incubated in the presence of a hyperglycemic medium, and indeed this reduced the expression of TGF-2 by mesangial cells and of collagen IV and fibronectin by podocytes.
The findings suggest that a higher number of trinucleotides acts functionally as a gain of function trait consistent with a dominant mode of inheritance.
What is known about carnosine? It prevents cellular senescence (21) and is said to have anti-aging properties (22). It scavenges reactive oxygen species (23) and reacts with macromolecular carbonyls (22). It also inhibits ACE (24). These are all properties that would interfere with initiation and progression of diabetic nephropathy and conceivably also with other micro- and macrovasular complications of diabetes. We shall certainly hear about this interesting substance in the future.
|
References |
|---|
|
TopReferencesReferences References |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
HOME
CURRENT ISSUE
ARCHIVES
JASN Express
ONLINE SUBMISSION
AUTHOR INFO
EDITORIAL BOARD SUBSCRIBE FEEDBACK ALERTS HELP |
