Proceedings of the Fourth Genoa Meeting on Hypertension, Diabetes, and Renal Diseases
Randomized Clinical Trials on Surrogate End Points: Are They Useful for Evaluating Cardiovascular and Renal Disease Protection in Hypertension? The Case for Yes
Alberto Morganti
Cattedra di Medicina Interna and Centro Ipertensione Arteriosa Ospedale San Paolo and Centro Fisiologia Clinica e Ipertensione, IRCCS Ospedale Maggiore, University of Milan, Milan, Italy
Address correspondence to: Prof. Alberto Morganti, Cattedra Medicina Interna, Centro Ipertensione Arteriosa, Ospedale San Paolo, Via Di Rudinì 8, 20142 Milano, Italy. Phone/Fax: +39-02-50323080; E-mail alberto.morganti{at}unimi.it
Hard end point studies represent the best available evidencefor demonstrating the cardiovascular and renal protection thatis achievable with a given treatment in hypertensive patients,yet properly designed end point studies require large cohortsof patients and long follow-up, are expensive, and do not provideany insight on the mechanisms that lead to the clinical manifestations.Studies that are based on the incidence of preclinical alterations,i.e., the surrogate end points, may circumvent these limitationsprovided that their relationship with the major cardiovascularevents is scientifically proved. In this respect, among themany surrogate end points that are under investigation, leftventricular hypertrophy, microalbuminuria, and treatment-induceddiabetes seem most promising for replacing the hard end pointsin view of their undisputed mechanistic relationship with theclinical events and of the mounting evidence indicating thatfrom their changes it is possible to predict the clinical outcomeof patients. In addition, the limited resources that are requiredto carry out this kind of investigations make them preferableto hard end point studies for anticipating the cardiovascularand renal benefit associated with the use of antihypertensiveagents.
Hard end point studies, i.e., those that are based on the incidenceof major cardiovascular (CV) events such as myocardial infarction,stroke, and heart failure or renal events such as the developmentof ESRD, are considered the gold standard for comparing thecardiac and renal protection that is achievable with two differenttreatments. However, this kind of study has a number of logistic,mechanistic, and economic limitations (Table 1), and this ledto the idea of replacing them with the surrogate, or intermediate,end point studies, i.e., those that are based on the incidenceof subclinical alterations, some of which are listed in Table 2.The use of these alternative end points is rationally conceivablebecause many of them represent the logic transition phase betweenthe exposure to CV risk factors and the manifestation of theclinical events. However, before whichever surrogate end pointis accepted as a valid alternative of the hard end point, twoprerequisites need to be satisfied: (1) Proof of its biologicrelevance for causing the event and (2) a convincing demonstrationof the relationship between the progression/regression of thesurrogate end point and the increase/decrease of the relevantclinical events.
Herein I address these issues taking advantage of the resultsof some recent intervention trials and of some observationalstudies to demonstrate that at least some of these surrogateend points, namely left ventricular hypertrophy (LVH), microalbuminuria(MA), and new-onset diabetes, do indeed fulfil these requirementsand, therefore, can and should be used as an alternative tothe major clinical events for predicting CV and renal protection.
There is plenty of evidence that LVH is the initial step towardthe development of major CV events such as congestive heartfailure, cardiac ischemia and arrhythmias, and stroke. Suchevidence goes back to more than 30 yr, when the Framingham Studyclearly demonstrated that LVH, recognized by electrocardiogram,is a strong predictor of CV events (1). Subsequent echocardiographicstudies confirmed these early observations (2). In agreementwith these findings, Verdecchia et al. (3) showed that in hypertensivepatients with a left ventricular mass (LVM) >125 g/m2, therate of CV events is almost three times higher than in patientswith LVM <125 g/m2. Similar results supporting the valueof electrocardiographic LVH in predicting CV morbidity wereobtained by Levy et al. (4). More recent studies have shownalso that differences in the geometry of the left ventriclecan affect the CV prognosis in hypertensive patients, in thatfor similar values of LVM, patients with concentric LVH havea significantly greater incidence of CV events than those withthe eccentric geometry (5,6). The combination for LVH with electrocardiographicST depression also is relevant for patient’s prognosis;indeed, the Strong Heart Study that was conducted with AmericanIndians has shown that the concomitant presence of both of thesecardiac alterations is associated with a four-fold increasein CV mortality with respect to patients who harbor just oneof them (7). Moreover, echocardiographic studies have shownthat the antihypertensive treatment can induce the reversalof LVH and that the reduction of LVM is associated with a reductionof the risk for subsequent CV disease (8,9). In this respect,the study of Verdecchia et al. (3) has shown that during a follow-upof several years, the event rate in patients with the regressionof LVH was four-fold lower than that observed in patients whohad no regression with treatment.
MA has been recognized as an independent risk factor in therecent European Society of Hypertension/European Society ofCardiology guidelines on hypertension (10); this occurred asa result of the increasing evidence indicating that an excessof urinary albumin excretion (UAE) rather than being a simplemarker of renal damage is associated with an increased rateof CV events not only in hypertensive patients and patientswith diabetes but also in the general population. In a cohortof almost 10,000 individuals who were 55 yr or older and hada history of previous CV disease or diabetes, Gerstein et al.(11) found that MA significantly increased the adjusted relativerisk for major CV events, all-cause deaths, and hospitalizationfor congestive heart failure to 1.8, 2.1, and 3.2, respectively,these increments being similar for patients with and withoutdiabetes. Moreover, these authors observed that compared withpatients of the lowest quartile of the urinary albumin/creatinineratio (UACR), those in the highest quartile had a two-fold increasein the relative risk for the primary aggregate end point; theyalso noticed that there is not a clear-cut threshold value ofMA for indicating an increased CV risk, because for every 0.4mg/mmol increase in UACR, the adjusted hazard of major CV eventsincreased by 5.9%.
Along this line of research, in the Prevention of Renal andVascular End Stage Disease (PREVEND) Study that was conductedin the Netherlands in a cohort of 85,421 patients who were aged28 to 75 yr, Hillege et al. (12) found a positive dose-responserelationship between increasing UAE and CV and non-CV mortalityeven after adjustment for conventional CV risk factors; moreover,a two-fold increase in UAE was associated with a 29% increasein the relative risk for CV mortality. Results of these studiesled several authorities in the field to recommend a redefinitionof the normal threshold for MA, suggesting that the levels ofMA that are associated with an increased CV risk may be muchlower than was previously appreciated (13). MA also increasesthe mortality risk associated with electrocardiographic ST changes;in fact, it has been shown that in patients with the combinationof these two surrogate end points, the hazard ratio for CV deathsis four times higher than in those with ST segment changes alone(14). Further evidence in favor of MA as an independent riskfactor comes from studies in hypertensive patients. Agrawallet al. (15) reported that in a large cohort of hypertensivepatients with and without diabetes, those with MA at baselinehad a significantly greater prevalence of coronary artery disease,LVH, stroke, and peripheral vascular disease. These findingswere reinforced by a subanalysis of the recent Losartan Interventionfor End Point Reduction (LIFE) Study (16) in which among the8206 hypertensive patients with LVH, those in the lowest decileof UACR (<0.26) had a rate of primary composite end pointsfour times lower than those in the highest decile (>12) irrespectiveof whether they were treated with losartan or atenolol. In thesame study, when patients were stratified by time-varying UACR,the rate of CV end points was significantly related to the degreeof albumin excretion throughout the 5 yr of follow-up. In addition,it has been calculated that 17% of the benefit of losartan relativeto atenolol in preventing CV events was explained by its effecton albuminuria.
A retrospective analysis of the Reduction in End-Point in NIDDMwith the Angiotensin II Antagonist Losartan (RENAAL) Study (17)that was conducted in hypertensive patients who had type 2 diabetesand mild to severe renal insufficiency and were treated withlosartan-based versus conventional drug–based therapyhas shown that the rate of the renal end points was relateddirectly to the levels of proteinuria at entry and inverselyto its changes from baseline to 6 mo of treatment; moreover,after adjustment for other risk factors, the same stratificationof baseline and in treatment UAE was significantly related tothe incidence of CV end points, providing evidence, for thefirst time, that the reduction of protein excretion can predictboth the renal and the CV outcomes.
There is mounting evidence from recent clinical trials thatsome antihypertensive agents, namely blockers and diuretics,may augment the propensity of hypertensive patients to developtype 2 diabetes (18). This is relevant because it is widelyknown that hypertensive patients are already exposed to a higherrisk for developing type 2 diabetes with the attendant burdenof CV disease; in addition, the prevalence of diabetes and hypertensionincreases in parallel with the increases in obesity and agingof the populations throughout the world (19). Therefore, thetreatment-induced onset of diabetes can be considered as a plausiblesurrogate end point of subsequent CV events. This concept hasbeen challenged by the result of the Antihypertensive and LipidLowering Treatment to Prevent Heart Attack Trial (ALLHAT) Studyin which the greater incidence of new diabetes in patients ofthe chlorthalidone group compared with the amlodipine and lisinoprilgroups was not accompanied by a greater incidence of CV mortalityand morbidity (20).
However, several previous studies have shown that the treatment-induceddiabetes is far from being a benign condition. Indeed, Aldermanet al. (21) showed that after adjustment for age and gender,the rate of CV diseases in hypertensive patients who developeddiabetes while on antihypertensive treatment was similar tothat of patients who already had diabetes at the time of entryin the study. Along this line of observation, Dunder et al.(22) reported that in middle-aged treated hypertensive patients,the increase in fasting blood glucose levels was the most powerfulpredictive index of subsequent myocardial infarction. Moreover,Verdecchia et al. (23) in a long-term follow-up study confirmedthat the rate of CV events in treated hypertensive patientswas similar in those with new or previously known diabetes andsignificantly higher than in those without diabetes despitesimilar levels of BP; in addition, in that study, the relativerisk for CV events was roughly doubled for all three groupsof patients with diabetes when LVH also was present.
It seems that at least some surrogate end points do fulfillthe criteria for reliably replacing the major events as primaryend point in clinical trials that aim to establish the cardio/renalprotection of treatments. The surrogate end points have theadditional advantages listed in Table 3; among these, the reductionin duration and cost is particularity attractive in view ofthe increasing limitation in financial resources facing eventhe most affluent societies.
Kannel WB, Gordon T, Offut T: Left ventricular hypertrophy by electrocardiogram. Prevalence, incidence and mortality in the Framingham Study.
Ann Intern Med 71
: 89
–105, 1969[Medline]
Casale PN, Devereux RB, Milner M, Zullo G, Harshfield GA, Pickering TG, Laragh JH: Value of echocardiographic left ventricular mass in predicting cardiovascular morbid events in hypertensive men.
Ann Intern Med 105
: 173
–178, 1986[Medline]
Verdecchia P, Schillaci G, Borgioni C, Ciucci A, Gattobigio R, Zampi I, Reboldi G, Porcellati C: Prognostic significance of serial changes in left ventricular mass in essential hypertension.
Circulation 97
: 48
–54, 1998[Abstract/Free Full Text]
Levy D, Salomon M, D’Agostino R, Belanger A, Kannel WB: Prognostic implications of baseline echocardiographic features and their serial changes in subjects with left ventricular hypertrophy.
Circulation 90
: 1786
–1793, 1994[Abstract/Free Full Text]
Muiesan ML, Salvetti M, Monteduro C, Bonzi B, Paini A, Viola S, Poisa P, Rizzoni D, Castellano M, Agabiti-Rosei E: Left ventricular concentric geometry during treatment adversely affects cardiovascular prognosis in hypertensive patients.
Hypertension 43
: 731
–738, 2004[Abstract/Free Full Text]
Wachtell K, Bella JN, Liebson PR, Gerdts E, Dahlof B, Aalto T, Roman MJ, Papademetriou V, Ibsen H, Rokkerdal J, Devereux RB: Impact of different partition values on prevalence of left ventricular hypertrophy and concentric geometry in large hypertensive population. The LIFE Study.
Hypertension 35
: 6
–12, 2000[Abstract/Free Full Text]
Okin PM, Roman MJ, Lee ET, Galloway JM, Howard BV, Devereux RB: Combined echocardiographic left ventricular hypertrophy and electrocardiographic ST depression improve prediction of mortality in American Indians. The Strong Heart Study.
Hypertension 43
: 769
–774, 2004[Abstract/Free Full Text]
Muiesan ML, Salvetti M, Rizzoni D, Castellano M, Donato F, Agabiti-Rosei E: Association of changes in left ventricular mass with prognosis during long-term antihypertensive treatment.
J Hypertens 13
: 1091
–1097, 1995[CrossRef][Medline]
Verdecchia P, Angeli F, Borgioni I, Gattobigio R, de Simone G, Devereux RB, Porcellati C: Changes in cardiovascular risk by reduction of left ventricular mass in hypertension: A meta-analysis.
Am J Hypertens 16
: 895
–899, 2003[CrossRef][Medline]
Guidelines Committee: 2003 European Society of Hypertension–European Society of Cardiology guidelines for the management of arterial hypertension.
J Hypertens 21
: 1011
–1053, 2003[CrossRef][Medline]
Gerstein HC, Mann JFE, Yi Q, Zinman B, Dinneen SF, Hoogwerf B, Halle JP, Young J, Rashkow A, Joyce C, Nawaz S, Yusuf S: Albuminuria and risk of cardiovascular events, death and heart failure in diabetic and non diabetic individuals.
JAMA 286
: 421
–426, 2001[Abstract/Free Full Text]
Hillege HL, Fidler V, Diercks GFH, van Gilst WH, de Zeeuw D, van Valdhuisen DJ, Gans ROB, Janssen WMT, Grobbee DE, de Jong PE: Urinary albumin excretion predicts cardiovascular and noncardiovascular mortality in general population.
Circulation 106
: 1777
–1782, 2002[Abstract/Free Full Text]
Redon J, Williams B: Microalbuminuria in essential hypertension: Redefining the threshold.
J Hypertens 20
: 353
–355, 2002[CrossRef][Medline]
Diercks GF, Hillige HL, van Boven AJ, Kors JA, Crijns HJ, Grobbee DE, de Jong PE, van Gilst WH: Microalbuminuria modifies the mortality risk associated with electrocardiographic ST-T segment changes.
J Am Coll Cardiol 40
: 1401
–1407, 2002[Abstract/Free Full Text]
Agrawal B, Berger A, Wolf K, Luft FC: Microalbuminuria screening by reagent strips predicts cardiovascular risk in hypertension.
J Hypertens 14
: 223
–228, 1996[Medline]
Ibsen H, Wachtell K, Olsen MH, Borch-Johnsen K, Lindholm LH, Mogensen CE, Dahlof B, Devereux RB, de Faire U, Fyrquist F, Julius S, Kjeldsen SE, Lederballe-Pedersen O, Nieminen M, Omvik P, Oparil S, Wan Y: Does albuminuria predict cardiovascular outcome on treatment with losartan versus atenolol in hypertension with left ventricular hypertrophy? A LIFE substudy.
J Hypertens 22
: 1805
–1811, 2004[CrossRef][Medline]
de Zeeuw D, Remuzzi G, Parving HH, Keane WF, Zhang Z, Shahinfar S, Snapinn S, Cooper ME, Mitch WE, Brenner BM: Proteinuria, a target for renoprotection in patients with type 2 diabetic nephropathy.
Kidney Int 65
: 2309
–2320, 2004[CrossRef][Medline]
Bakris GL, Sowers JR: When does new onset diabetes resulting from antihypertensive therapy increase cardiovascular risk.
Hypertension 43
: 941
–942, 2004[Free Full Text]
McFarlane SI, Banerji M, Sowers JR: Insulin resistance and cardiovascular disease.
J Clin Endocrinol Metab 86
: 713
–718, 2001[Free Full Text]
ALLHAT Officers and Coordinators for the ALLHAT Collaborative Research Group; The Antihypertensive and Lipid Lowering Treatment to Prevent Heart Attack Trial: Major outcomes in high risk hypertensive patients randomized to angiotensin converting enzyme inhibitor or calcium channel blocker vs. diuretic: The Antihypertensive and Lipid Lowering Treatment to Prevent Heart Attack Trial (ALLHAT).
JAMA 288
: 2981
–2997, 2002[Abstract/Free Full Text]
Alderman M, Cohen H, Madhaven S: Diabetes and cardiovascular events in hypertensive patients.
Hypertension 33
: 1130
–1134, 1999[Abstract/Free Full Text]
Dunder K, Lind L, Zethelins B, Berglund L, Lithell H: Increase in blood glucose concentration during antihypertensive treatment as a predictor of myocardial infarction: Population based cohort study.
BMJ 326
: 681
, 2003[Abstract/Free Full Text]
Verdecchia P, Reboldi G, Angeli F, Borgioni C, Gattobigio R, Filippucci L, Morgiolini S, Bracco C, Porcellati C: Adverse prognostic significance of new diabetes in treated hypertensive subjects.
Hypertension 43
: 963
–969, 2004[Abstract/Free Full Text]
Top
Abstract
Introduction
blockers and diuretics, may augment the propensity of hypertensive patients to develop type 2 diabetes (18). This is relevant because it is widely known that hypertensive patients are already exposed to a higher risk for developing type 2 diabetes with the attendant burden of CV disease; in addition, the prevalence of diabetes and hypertension increases in parallel with the increases in obesity and aging of the populations throughout the world (19). Therefore, the treatment-induced onset of diabetes can be considered as a plausible surrogate end point of subsequent CV events. This concept has been challenged by the result of the Antihypertensive and Lipid Lowering Treatment to Prevent Heart Attack Trial (ALLHAT) Study in which the greater incidence of new diabetes in patients of the chlorthalidone group compared with the amlodipine and lisinopril groups was not accompanied by a greater incidence of CV mortality and morbidity (20). 
