Renal Function and Intensive Lowering of Blood Pressure in Hypertensive Participants of the Hypertension Optimal Treatment (HOT) Study

Trial Design

The HOT Study design and organization has been described in detail in a previous publication, reporting the principal results of the trial (8). In brief, 18,790 hypertensive patients from 26 countries were randomly assigned to one of three diastolic BP target groups (≤90 mmHg, ≤85 mmHg, ≤80 mmHg) and to either ASA (75 mg once daily) or placebo. They were followed for an average period of 3.8 yr (range, 3.3 to 4.9 yr). Patients were randomized on the basis of the following baseline variables: age, gender, previous antihypertensive therapy, smoking, previous myocardial infarction, other coronary heart disease, stroke, and diabetes mellitus; serum cholesterol and creatinine were not included in the randomization procedure. Antihypertensive therapy with the long-acting calcium antagonist felodipine at a dose of 5 mg once a day was given to all patients. Additional therapy and dose increments in four further steps were prescribed to reach the randomized target BP. ACE inhibitors or β-blockers were added at step 2, and dose titrations were used at step 3 (felodipine 10 mg once a day) or 4 (doubling the dose of either the ACE inhibitor or the β-blocker), with the possibility of adding a diuretic or another antihypertensive drug at step 5. BP was measured with an oscillometric device (Visomat OZ, D₂, International, Hestia, Germany) with the patient in the seated position at each prerandomization visit, at randomization, 3 and 6 mo after randomization, and twice a year thereafter. Major cardiovascular events (acute myocardial infarction, stroke, cardiovascular death) as well as noncardiovascular deaths were determined by the Independent Clinical Event Committee (blind to the study group to which patients were assigned) according to criteria defined in the study protocol.

Renal Function Measurement

According to the protocol, serum creatinine had to be measured, at each recruiting center, by standard laboratory techniques in every patient twice, at baseline and at the final study visit. A baseline serum creatinine value >3.0 mg/dl was considered as an exclusion criterion. Creatinine clearance was estimated using the Cockroft and Gault formula (9). For the final analysis, the presence of a serum creatinine concentration above 1.5 mg/dl (132.6 μmol/L) was considered as indicative of the presence of chronic renal failure, and an estimated creatinine clearance ≤60 ml/min was arbitrarily taken as indicative of a clinically significant reduction in renal function, as described previously in the literature (10,11). To keep the large trial as simple as possible, the protocol did not require systematic measurement of proteinuria, and therefore data on proteinuria are not reported.

Statistical Analyses

Event rates were evaluated separately for patients with serum creatinine concentrations >1.5 and ≤1.5 mg/dl and for patients with estimated creatinine clearance ≤60 and >60 ml/min at baseline. Risk ratios (with 95% confidence intervals [CI] and P values) were adjusted for the following baseline variables: achieved systolic and diastolic BP, age, gender, smoking habits, previous cardiovascular disease, diabetes, and total serum cholesterol. The adjustments were performed by Poisson regression (12). A sensitivity analysis to correct for heterogeneity among different countries was not performed because both center and nationality were controlled for by the randomization procedure.

In the analysis of trends for differences between diastolic BP target groups and the effects of ASA compared with placebo, a Cox proportional-hazards model based on a factorial design was used to calculate relative risks. Comparisons between groups were made on a pair-wise basis with one degree of freedom, using t test. For comparisons of proportions, Fisher's exact test was used. Correlation analyses were performed calculating Pearson correlation coefficients.

Cardiovascular Events in Patients with Reduced Renal Function at Baseline

Serum creatinine values were available in a total of 18,597 patients at baseline. Among them, 470 subjects had a serum creatinine value higher than 1.5 mg/dl and 2821 had an estimated creatinine clearance ≤60 ml/min.

As shown in Table 1, at randomization, the HOT patients with high serum creatinine predominantly were male and were significantly older. They also more frequently had a history of myocardial infarction or other sequelae of coronary heart disease, of stroke, and of diabetes mellitus than those a with serum creatinine value ≤1.5 mg/dl. The two groups did not differ significantly as far as current smoking and serum cholesterol concentration were concerned. Patients with low estimated creatinine clearance (Table 2) were older with a greater prevalence of a history of previous cardiovascular events and predominantly were women. The predominance of women in the group with lower estimated creatinine clearance could be explained by the fact that this parameter is usually lower in women and also because the Cockroft and Gault formula does not include correction for body surface area.

Table 3 compares the rate of events that occurred in the group of patients with a serum creatinine value >1.5 mg/dl and ≤1.5 mg/dl. Risk ratios adjusted for all other baseline variables except serum creatinine indicate a markedly increased risk for all types of events considered. The increase in risk ratios for major cardiovascular events, cardiovascular, and total mortality was statistically highly significant. When considering patients with less markedly reduced renal function (estimated creatinine clearance ≤60 ml/min versus patients with values >60 ml/min) adjusted risk ratios for all types of events were less markedly elevated but still statistically significant with the exception of the risk ratio for myocardial infarction (Table 4).

Effects of Intensive BP Lowering in Patients with Reduced Renal Function

Control of BP. During follow-up, the achieved values of systolic and diastolic BP were not different between patients with a baseline serum creatinine value >1.5 mg/dl and ≤1.5 mg/dl (Table 5). The percentages of patients who achieved the diastolic BP target to which they had been randomized did not differ significantly between patients with higher or lower baseline serum creatinine concentrations. More intensive therapy was, however, necessary for patients with than those without reduced renal function (Table 6).

Cardiovascular Events According to Randomized Diastolic BP Target. The incidence of major cardiovascular events in patients with a serum creatinine value >1.5 mg/dl was very similar in the three groups randomized to different diastolic BP targets (26.5, 26.7, and 27.9 per 1000 patient years). There were on the whole too few events to analyze interactions between BP target and type of event. There were more events in the larger group of patients with an estimated creatinine clearance ≤60 ml/min (but the absolute rate was lower than in patients with a serum creatinine concentration >1.5 mg/dl). Cardiovascular events tended to be less frequent in patients with the lower diastolic BP target, but this was not statistically significant (Figure 1).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.

Major cardiovascular events in the diastolic BP (DBP) target groups in patients with estimated creatinine clearance ≤60 ml/min.

Effects on Renal Function. Baseline and final serum creatinine concentrations were available for 15,601 patients (83.0% of the entire study population). Of these, 15,237 patients had baseline serum creatinine ≤1.5 mg/dl and 364 had baseline values >1.5 mg/dl. Of the 3048 patients whose final serum creatinine was not available, 491 (1.6%) had been lost to the study follow-up, 589 (1.9%) had died, and 1968 (6.5%) specifically failed to have final serum creatinine measurement. Although the rate of death was significantly higher among patients with higher serum creatinine, as shown in Table 3, loss to follow-up and failure to remeasure serum creatinine at the final visit were similar among patients with higher and lower values at baseline (10.8 and 11.1%, respectively).

Baseline serum creatinine concentrations, mean 1.00 mg/dl (SD 0.25), were unrelated to baseline diastolic BP both in men (r = 0.01) and in women (r = -0.004), as well as in patients older than 65 yr (r = -0.003) or younger than 65 yr (r = 0.03). No significant changes were seen in serum creatinine values at the end of the 3.8-yr treatment period, in all HOT Study participants considered as a single group or in the three diastolic BP target subgroups (Table 7). Table 7 also shows that no significant changes in serum creatinine had occurred at the end of the study in any of the diastolic BP target subgroups. This was true in patients with higher or lower baseline serum creatinine concentrations as well as in patients with higher or lower estimated creatinine clearance. Baseline and final serum creatinine concentrations were not influenced by diabetes mellitus, ischemic heart disease, or age. Despite that mean initial and final serum creatinine values were not different, an increase in the final value of serum creatinine concentration by 30% or more of the baseline values was seen in 1220 patients (7.8%), corresponding to 1185 (7.6%) patients with initial serum creatinine ≤1.5 mg/dl and 35 patients (0.22%) with higher initial values. A decrease in serum creatinine by a similar amount was observed in 652 patients (4.18% of the total sample). Ninety patients with a serum creatinine increase ≥30% reached final values of 2.0 mg/dl. This amounted to only 0.58% of the total study population. Among patients with initial serum creatinine >1.5 mg/dl, 30 additional patients reached a final creatinine ≥2.0 mg/dl, despite an increment <30%, or already had a baseline creatinine ≥2.0 mg/dl, so that on the whole 120 (0.8%) of 15,601 patients in the HOT Study had serum creatinine ≥2.0 mg/dl at the final visit (17.9% of patients with baseline creatinine >1.5 mg/dl and 0.4% of patients with baseline creatinine ≤1.5 mg/dl).

Effects of ASA

Table 8 provides the data on the effect of ASA according to baseline serum creatinine concentration. Patients with baseline serum creatinine values >1.5 mg/dl and receiving ASA tended to have lower rates of all types of events. Presumably because of the relatively small number of patients in this group, the ASA effects did not reach the conventional level of statistical significance.

Cardiovascular Events

Renal insufficiency at baseline, as indicated by a serum creatinine concentration >1.5 mg/dl or an estimated creatinine clearance <60 ml/min, was accompanied by a significant elevation of cardiovascular risk in the present study. The highest adjusted relative risk was 3.24 for cardiovascular death. It is widely known that the cardiovascular system is affected profoundly by the presence of advanced renal failure. In fact, patients who undergo maintenance hemodialysis have a cardiovascular mortality approximately threefold higher than that of age-matched nonuremic patients (13). The increased mortality is associated with a higher prevalence of atherosclerotic heart disease with myocardial infarction, left ventricular hypertrophy, and congestive heart failure.

An increase of serum creatinine in patients with essential hypertension is commonly attributed to the presence of nephrosclerosis. This condition is characterized by hyalinization of arterioles and by fibroplastic intimal thickening of small arteries. It has been reported that hyalinization of renal arterioles is more marked in patients with coronary heart disease than in matched control subjects (14). Furthermore, autopsy studies (15) of otherwise asymptomatic young people also showed that hyalinization in the renal arterioles was a marker of advanced coronary atherosclerosis.

The Hypertension Detection and Follow-up Program trial (5) documented that the baseline serum creatinine concentrations have a significant prognostic value for 8-yr mortality. For people with a serum creatinine concentration >1.7 mg/dl, mortality (after adjustment for other cardiovascular risk factors) was 2.22 times higher than that of all other participants. In agreement with these data, we found a risk ratio of 2.86 in patients with serum creatinine >1.5 mg/dl.

The recent data of the Framingham Study (11) document that mild renal insufficiency, defined as serum creatinine >1.5 mg/dl, is common (8.7% in men and 8.0% in women) and is associated with a high prevalence of cardiovascular disease. In the HOT Study, assignment to different diastolic BP targets did not significantly influence the rate of cardiovascular events, but the differences in achieved BP among groups were small. Furthermore, serum creatinine concentrations had not been considered in the randomization procedure.

Effects on BP

As described previously (16), target diastolic BP could be reached even in patients with diminished renal function at baseline. Good control of BP necessitated more intensive antihypertensive treatment, however. Following the new guidelines for the management of hypertension (17, 18), the control of diastolic BP obtained in the HOT Study was adequate both for patients initially presenting with normal renal function and for those with impaired renal function. The mean diastolic BP achieved throughout the trial was 82 mmHg, in line with the recommendation to achieve values lower than 85 mmHg. However, the control of systolic BP remained more than 10 mmHg (mean) above the goal of <130 mmHg, which has been recommended for patients with elevated levels of serum creatinine. Control of systolic BP was not the aim of this study, and when it was designed the new goals had not been established. Whether incomplete control of systolic BP was responsible for the high incidence of all events in patients with diminished renal function cannot be decided on the basis of available data.

Effects on Renal Function

The assessment of the effects of intensive lowering of BP on renal function was partly limited by the fact that serum creatinine was measured only at baseline and at the final visit. Consequently, we do not have the information on the 3.3% of patients who died during the study. Furthermore, approximately 10% of the patients failed to have the final serum creatinine measurement. Despite these reservations, it can be concluded that the overall good BP control achieved in this study did not have an adverse effect on serum creatinine values, at least in the great majority of the patients. These results are in agreement with those published by Mandhavan et al. (3, who treated and followed 2125 men with mild and moderate hypertension over an average of 5 yr with good control of BP. Conversely, in the HDFP study, serum creatinine increased by 0.14 to 0.2 mg/dl at the end of 5 yr, whereas in the HOT study, patients we observed had an increase of only 0.01 mg/dl over 3.8 yr.

In the HOT Study, not only was renal function well preserved, but also the rate of cardiovascular events and death was low. In fact, the event rate in this study was much lower than the average rate calculated in the meta-analysis of all previous antihypertensive treatment trials (19). There is a small group of patients in whom renal function deteriorates progressively despite an adequate BP control, which is in agreement with what has been reported in the literature (20,21). To take account of regression toward the mean, we adopted a conservative definition of deterioration of renal function, i.e., an increase in serum creatinine exceeding 30% with a final creatinine concentration of ≥2.0 mg/dl. Whereas the proportion of patients who showed a creatinine increase >30% was 7.8%, only 90 patients fulfilled both criteria of renal deterioration, yielding an overall rate of approximately 0.5%. On the whole, at the end of the study, 120 patients had serum creatinine levels ≥2.0 mg/dl. Because of the small size of this group, it is inappropriate to analyze their baseline characteristics in detail, but patients whose renal function deteriorated had mostly baseline serum creatinine >1.5 mg/dl and were equally distributed among the three groups randomized to different BP targets

Effects of ASA

The benefit of adding a low dose of ASA to antihypertensive therapy was difficult to prove in patients with elevated values of serum creatinine, because the numbers were small. Nevertheless, the risk ratios favoring aspirin versus placebo were low and approached, though did not reach, statistical significance.