Roca-Cusachs, Alejandro*; Torres, Ferran†‡; Horas, Manuel†; Ríos, José†; Calvo, Gonzalo†‡; Delgadillo, Joaquín§; Terán, Maite§; Spanish Nitrendipine/Enalapril Collaborative Study Group
The goal of hypertension therapy is to reduce morbidity and cardiovascular mortality by the least intrusive means possible. For this purpose it is important to decrease systolic and diastolic blood pressure (SBP and DBP) to < 140/90 mm Hg or lower if tolerated and to control other modifiable risk factors for cardiovascular disease. The National Health and Nutrition Examination Survey (NHANES) revealed an increase in awareness of hypertension from 51%–73%, but among persons with hypertension, the rate of those treated achieving goal blood pressure (< 140/90 mm Hg) had only improved from 10% in NHANES-II to 29% in NHANES-III. Thus, > 50% of hypertensive patients have inadquately controlled disease (1). Other studies have shown similar results (2–6).
Pharmacologic treatment should begin with monotherapy for most patients; it is sometimes necessary to add a second or more drugs to achieve a good blood pressure control (7). Therefore, in some cases the use of the fixed-dose combinations may be advantageous (8,9).
Angiotensin-converting enzyme (ACE) inhibitors and calcium antagonists are commonly used as monotherapy in the management of hypertension because they have a favorable side effect profile and minimal metabolic consequences compared with older drugs, especially diuretics. The combination of drugs from both groups has been shown not only to be more effective than either drug alone in lowering blood pressure but also to have a better tolerability than each drug separately (10–15).
Clinical trials of combination therapy for hypertension typically have either provided information on the additive effects of combination drug therapy in patients unresponsive to monotherapy or have used a design with three groups (A, B, A + B) that can compare only one, or at most two, dose combinations. Neither of these clinical trial designs gives sufficient data to characterize the dose-response relationship of each drug in the presence of the other drug. The factorial trial, if well designed and using as wide a dose range for each drug as feasible, generates data that allow construction of a mathematical model to characterize the relationship of the two drugs. This information on dose response is of considerable value to the thorough understanding of the combination product in the development of dose selection and ultimate regulatory approval.
A factorial design was applied in a multicenter trial of the ACE inhibitor enalapril in combination with the calcium antagonist nitrendipine. The objectives of the trial were to evaluate the dose-response relationship with regard to efficacy and safety and to help to define the optimal dose ratio of the combination for maximum efficacy and minimum side effects in the treatment of mild to moderate essential hypertension.
Men and women aged 18–70 years with mild to moderate essential hypertension defined as sitting DBP ≥ 90 mm Hg and < 110 mm Hg were eligible for the study. Patients with secondary or malignant hypertension, grade III or IV hypertensive retinopathy, history of repeated hypertensive emergencies, hypertension refractory to treatment, drug hypersensitivity or unsatisfactory response to ACE inhibitors or calcium antagonists, acute myocardial infarction or stroke within the previous 6 months, chest pain, New York Heart Association class III or IV chronic heart failure, known renal or hepatic impairment, diabetes, and obesity (brachial perimeter > 40 cm) were excluded. Women who were pregnant or breast-feeding and those not using adequate contraceptive measures were also excluded. Concomitant anti-hypertensive medications, nitrates, steroids, and nonsteroidal anti-inflammatory drugs were not permitted. The study protocol was reviewed by the local Ethics Committees and approved by the Spanish Health Authorities. The clinical trial was conducted in agreement with all local regulations and with the Declaration of Helsinki and its updates and in accordance with Good Clinical Practices. Informed consent was obtained from all subjects prior to their participation in the clinical trial.
Following a 1-week washout period (only for patients previously treated with anti-hypertensive therapy), patients were enrolled into a 2-week single-blind, placebo run-in period. At the end of the run-in period, qualifying patients (those with consistently elevated DBP within the range of 90–109 mm Hg, which differed < 10 mm Hg from that observed in the previous visit), were randomly assigned to a 6-week double-blind treatment period with one of 16 parallel treatments: placebo, one of three doses of nitrendipine (5, 10, or 20 mg), one of three doses of enalapril (5, 10, or 20 mg), or one of nine possible combinations of nitrendipine and enalapril doses. All patients received one capsule, once daily in the morning. As a safety measure, patients assigned to receive either 10 or 20 mg of any or both drugs received invariably 5 mg/d for the first week of treatment before titration to the dose assigned.
Safety was assessed at each visit through recording adverse events. Routine hematologic, biochemical, and urinary investigations were undertaken at the end of run-in and active treatment periods.
Sitting blood pressure was measured in duplicate, after resting for 5 min, with a standard calibrated mercury sphygmomanometer using Korotkoff phases I and V as SBP and DBP indicators, respectively. The mean value of both readings was used in the efficacy analysis. If DBP differed by > 5 mm Hg between measures, a third measurement was obtained and the mean of the two closest readings was used in the efficacy analysis. All blood pressure recordings were done at trough and after 24 h (range, 20–28 h) after dose administration. Cuff size was appropriate for the size of the patient's arm and measurements were taken from the same arm at each visit. Heart rate was measured in each visit.
Due to the exploratory nature of the study, accurate sample size estimation was not deemed justified. So, sample size projection was aimed to ensure a minimum number of patients per treatment group that would allow us to set accurately mean DBP reductions and therefore ensure a correct description of the observed data. Using this approach and estimating a coefficient of variation of DBP not > 15% (based on published studies with similar design, with mean ± SD DBP of 100 ± 15 mm Hg), 25 patients (range, 23–28) per treatment group would be necessary to achieve 90% CI of mean DBP values in each cell with an accuracy of 5%. Therefore, the total sample size was fixed to enroll 500 patients to obtain valid data on 400 patients for the main efficacy analysis (25% predicted dropout rate). One year after the study initiation, an external statistician performed a blind interim analysis to assess whether the initial assumption on DBP variability was correct. The observed variability was about 10%, lower than assumed, allowing for a reduction in the number of patients with valid data required per cell in the range of 16–22.
The change in sitting DBP from baseline (defined as the end of the run-in period) to the last available measurement during the active treatment period was the primary efficacy variable. The change in sitting SBP and the rate of responders were assessed as secondary efficacy variables. The rate of responders was defined as the percentage of patients whose SBP and/or DBP was < 140/90 mm Hg or the reduction from baseline was > 20/10 mm Hg at the end of the active treatment period.
Results were analyzed assuming a strictly additive model for the effects through the approximation of the generalized linear models to test linear effects of each drug and interaction (16). A stepped analysis plan was used to analyze the primary efficacy parameter. Such an approach allows for testing if the combination treatment is superior to its individual components while also it may establish a dose-response relationship (16).
A two-way analysis of variance (ANOVA) (one for each drug), considering the effect of the interaction only when it was found to be statistically significant, was performed using a linear model for the primary efficacy analysis. This was used to test for overall effects of nitrendipine and enalapril adjusted for each other.
The superiority of combination therapy over monotherapy was assessed by means of hierarchical comparisons. Four kinds of superiority were defined: overall superiority of the combination versus monotherapy; wide superiority; weak superiority; and strict-sense superiority. The first three cases were assessed respectively by comparing all combination groups versus all monotherapy groups; all combination groups versus either monotherapy group; and all combination groups containing a fixed dose of one of the monocomponents versus all monotherapies of the other. Strict-sense superiority was established only if the combination therapy at a specific dose produced a significantly greater reduction in blood pressure than each individual agent at the same dose.
In addition, response surface modeling as regression analysis for gaining best estimators than that used in the ANOVA approach and to minimize over- or underestimation of dose-response relationships were executed (17). The efficacy variables were analyzed using polynomial functions in the doses of the two components accounting for the effects, with linear and quadratic terms for each drug and one term of the interaction between both linear terms. These estimators were used to search for differences between combination and monotherapy and to investigate dose-response relationships of the combination. The same general plan was used for SBP.
The rate of responders was analyzed by means of Fisher's exact test. Descriptive analysis was used for safety data.
A global 5%, two-sided level of significance was fixed for the ANOVA and the Fisher's exact test, as well as to determine the significance of the effects in the ANOVA and in the regression tests. For the superiority analyses, the significance level was set at 5%, one sided.
Two patient populations were defined for the efficacy analysis: the intention-to-treat population included patients receiving the assigned treatment and with blood pressure assessments (done between 8 am and 2 pm ) available at baseline and at the end of the 6-week active treatment period; and the per protocol population, a subset of the intention-to-treat population that included only patients without major protocol violations. The main efficacy analyses were based on the per protocol population. Safety analysis was done in all randomized patients who received at least one dose of the study medications.
Patient population and disposition
A total of 496 patients were enrolled in the study between September 1996 and April 1998; 82 of them did not qualify for randomization mainly because hypertension was not confirmed at the end of the run-in period (46 patients), due to voluntary withdrawal (15 patients), or due to adverse events (9 patients). Consequently, 414 patients were randomly assigned to active treatment. Of these, 36 patients were excluded because blood pressure measurements were not available after randomization (33 patients), blood pressure was measured outside the valid interval (2 patients), or error occurred in treatment assignment (1 patient). Thus the intention-to-treat population consisted of 378 patients (individual treatment cells ranging from 17–27). The per protocol population was 342 patients (individual treatment cells ranging from 17–25) because 36 patients were excluded.
Global exploratory comparisons among all groups were performed by means of an ANOVA for age, weight, and height and a Fisher's test for sex and race. No statistical significance was reached in these tests. Besides, per protocol and intent-to-treat populations were compared at baseline and no differences were observed (Table 1).
Unless otherwise indicated, all efficacy results are referred to the per protocol population of patients. The highest mean DBP reduction from baseline to last visit at the end of the 6-week double-blind period was achieved in the group receiving nitrendipine 20 mg plus enalapril 10 mg. The highest mean SBP reduction was achieved in the group receiving nitrendipine 20 mg plus enalapril 5 mg (Fig. 1A and B). The ANOVA model showed a significant dose-effect linear relationship for both drugs (linear terms: enalapril: p = 0.0032 for DBP, p = 0.0039 for SBP; nitrendipine: p = 0.0106 for DBP, p = 0.0025 for SBP), but the interaction was not statistically significant (cross-product term: p = 0.874 for DBP, p = 0.1515 for SBP), indicating that combination therapy yielded additive anti-hypertensive effect compared with monotherapy.
This additive effect was further explored in the superiority analyses. Overall superiority analyses showed significant superiority of the combination over monotherapy for both DBP and SBP. Wide superiority analyses showed significant superiority of the combination versus both monotherapies for SBP but failed to show superiority of the combination versus the monotherapy with enalapril for DBP. However, the weak superiority analyses confirmed significant differences in efficacy favoring combinations containing the highest doses of nitrendipine versus monotherapy with enalapril (mean DBP reductions with the combinations containing nitrendipine 20 mg: −13.2 mm Hg; with enalapril monotherapy: −9.2 mm Hg, p = 0.0054 for the comparison); and all combinations grouped by the dose of enalapril versus nitrendipine monotherapy (mean DBP reduction with all combinations containing enalapril 5 mg: −10.0 mm Hg, all combinations containing enalapril 10 mg: −11.3 mm Hg, all combinations containing enalapril 20 mg: −10.8 mm Hg; with nitrendipine monotherapy: −7.2 mm Hg, p = 0.0297, 0.0068, and 0.0138 for the respective comparisons). Significant differences were also found in these comparisons regarding SBP. From these results, strict superiority analyses were performed, and the combination nitrendipine 20 mg plus enalapril 5 mg was found to be significantly superior to either component monotherapy in the per protocol as in the intent-to-treat population in terms of both DBP and SBP reduction, as well as the combination of nitrendipine 20 mg plus enalapril 10 mg in terms of DBP. Combination of nitrendipine 10 mg plus enalapril 10 mg was also found to be significantly superior to either component monotherapy in terms of SBP in the per protocol population and nitrendipine 5 mg plus enalapril 10 mg in terms of SBP in the intent-to-treat population. The results of these superiority analyses for the per protocol population are shown in Tables 2 and 3.
The response surface analysis to assess the dose-response relationship showed significant linear and nonsignificant quadratic components both for nitrendipine and enalapril respect to DBP (p = 0.0102 and p = 0.0033, respectively, Fig. 2A), and SBP (p = 0.025 and p = 0.0043, respectively), suggesting progressively increasing effects with increasing doses of the two drugs and also suggesting that the maximum response was achieved with the highest doses. Conversely, in the intention-to-treat population, a significant quadratic component was found with enalapril (p = 0.0436) for DBP, with a maximum effect achieved with the combination of nitrendipine 20 mg plus enalapril 10 mg (Fig. 2B), supporting the results of the superiority analyses that indicated a higher benefit with the combination containing 20 mg of nitrendipine and 10 mg of enalapril than the combination containing 20 mg of both drugs. The product term was not statistically significant, supporting the assumption of an additive model for the effects on blood pressure reduction.
Regarding the response rate, the higher the dose administered, the higher the rate of responders; monotherapy with enalapril 20 mg those yielded the closest response to that achieved with the combinations containing higher doses of both drugs. The highest response rate was achieved with the combination containing the highest doses of nitrendipine with 10 mg enalapril (Table 4). When the response rate was considered only in terms of DBP, the response rate was still high, ranging from 29%– 76% with placebo and the combination containing the highest doses of each drug, respectively (Table 5). However, the analysis of the response rate failed to provide evidence for decision making, due to the heterogeneity of results, although a greater effect was observed with the combinations containing highest doses of both drugs.
None of the groups showed any clinically significant mean change in heart rate from baseline to the last visit.
Overall, the incidence of adverse events was similar between treatment groups. During the active treatment period, at least one treatment-related adverse event was reported in 7% of the patients on placebo, 17% on enalapril monotherapy, 21% on nitrendipine monotherapy, and 22% on combination treatments. The differences among all treatment groups were not statistically significant. No serious adverse events were reported. The study was discontinued because of adverse events in 10 patients during the run-in period (mainly headache) and in 7 patients during randomized treatment (mainly flushing).
Headache and flushing were more frequently reported in patients receiving nitrendipine, whereas nonproductive cough was reported more frequently in patients who received enalapril. These differences were not statistically significant. A trend toward additive effect of the combination therapy to cause any specific adverse event was not observed.
There was no evidence of clinically relevant trends for changes in any of the routine laboratory safety parameters assessed through the study.
In this study, the combination of enalapril and nitrendipine effectively lowered blood pressure and was generally well tolerated. Moreover, efficacy was enhanced when these drugs were used in combination compared with their monotherapies.
The factorial design of the study, combined with the modeling of the response through ANOVA and response surface methodologies, allows us to describe the dose-response relationships for a range of doses of nitrendipine and enalapril when administered once daily for the treatment of mild to moderate essential hypertension. It has been particularly useful in demonstrating an additive anti-hypertensive effect of enalapril and nitrendipine when used in combination.
In addition, dose-related DBP and SBP reductions have been found with all active treatments, including each agent used as monotherapy. These results indicate that each drug contributed within the combination to reduce the blood pressure. Even the smallest doses of one agent can enhance the anti-hypertensive efficacy of the other. Blood pressure reductions achieved with the combination of 10 mg of each drug were comparable to those gained with highest doses of each single agent as monotherapy. In addition, the combinations containing 20 mg of nitrendipine and 5–10 mg of enalapril were found to be significantly superior to any monotherapy, defining the trend showed by the general tests. Enalapril monotherapy was more efficacious that nitrendipine monotherapy, although the latter did markedly enhance the efficacy of enalapril. However, our results suggest that no therapeutic advantage is gained by co-administering enalapril doses > 10 mg daily with nitrendipine.
In the response surface analysis, the maximum effective doses of both agents were not reached in the per protocol population. A plateau was found in the enalapril effect beyond 10 mg using the intent-to-treat population while keeping a strict linear dose-response effect for nitrendipine, providing further support for the suitability of the combinations containing 20 mg of nitrendipine and 5–10 mg of enalapril.
The response rate, considered only in terms of DBP, was higher in combination than in monotherapy groups, being as high as 76% in the combination containing 20 mg of either drug. Nonetheless, it still remained as high as 60% when the response was considered in terms of both SBP and DBP (9). This finding is consistent with other reports in which the combination therapy was necessary in as many as 50% of cases to achieve the target blood pressure levels (2,6).
There are several pharmacologic rationales for combining an ACE inhibitor with a calcium antagonist. Calcium blockers induce a negative sodium imbalance that persists with long-term treatment beyond the initial natriuretic effect observed with the first doses. The acute calcium blockade provokes the activation of the renin-angiotensin system, which in turn limits their anti-hypertensive efficacy. However, the activation of the renin-angiotensin system apparently does not accompany the long-term use of calcium blockers. It can not be concluded that this negative sodium imbalance, persisting after the initial natriuretic effect, is by itself a pharmacodynamic rationale for combining these agents with ACE inhibitors in long-term treatments, because it has not been determined whether the potentiation of the blood pressure effect by ACE inhibitors is due to the negative sodium imbalance per se or to the induced activation of the renin-angiotensin-aldosterone system (2). Conversely, the activation of the sympathetic nervous system secondary to acute calcium blockade is most likely a limiting factor for the fall in the blood pressure due to the subsequent increase in heart rate. Experiments have shown that the larger the increase in the heart rate after one dose, the smaller the reduction achieved in the blood pressure, and pretreatment with an ACE inhibitor prevents changes in heart rate despite the presence of a fall in blood pressure. To ascertain whether the counterregulation mediated by the sympathetic nervous or renin-angiotensin systems after the acute administration of a calcium antagonist is involved throughout a long-term treatment, a study was completed with patients taking β-blockers and felodipine. Persisting sympathetic hyperactivity was found after 1 year of continuous treatment, and this could explain the modest regression of left ventricular hypertrophy despite decreases in SBP (18). Results of this study are encouraging, because ACE inhibitors combined with calcium antagonists would be the anti-hypertensive agents most able to induce a regression of hypertensive left ventricular hypertrophy.
Several studies of the free combination of ACE inhibitors with calcium antagonists have indicated that the anti-hypertensive efficacy of each component is additive in patients with varying severity of blood pressure elevation as well as in high-risk patients. In addition, the deleterious effects on renal function, peripheral edema, or headache associated with calcium antagonists have not been shown when these agents are co-administered with ACE inhibitors, suggesting a side effect counteraction. Moreover, among anti-hypertensive agents, the postsynaptic α-blockers, the calcium antagonists, and the ACE inhibitors have been reported to be either neutral or beneficial with regard to the overall metabolic risk factor profile (19) and electrolyte imbalance, supporting the interest drawn by this combination. Beyond these findings, extensive documentation available about the cardiovascular effects of the individual components, such as the reduction of recurrent myocardial infarction, congestive heart failure, cardiac mortality in post–myocardial infarction patients with depressed left ventricular ejection fraction, or heart remodeling after myocardial infarction, have suggested a superior efficacy for alleviating disease in target organs by the combination in addition to that expected by the blood pressure reduction itself (20). This suggests that pharmacologic management of hypertension could be directed toward blood pressure reduction as well as toward the treatment of target organs.
The anti-hypertensive effects of enalapril seem to plateau at a dose of 10 mg once daily. Because a priori we could expect that the higher the dose, the higher the effect, it is difficult to give a clinical meaning to this finding with the intent-to-treat population; the finding is probably related to chance imposed by the low number of cases per cell.
Overall, the tolerability was good in all groups. There were no statistically significant differences regarding the safety profile among them. No serious adverse event was reported and only seven of the randomly allocated patients were withdrawn because of treatment side effects. The combination therapy was not found to cause any specific adverse event or to foster the side effect profile of any of the individual agents. The profile of adverse events with the combination was consistent with the adverse event profile of each monocomponent. Edema and flushing occurred more frequently in patients receiving nitrendipine, whereas cough was noted more often in enalapril-treated patients. The incidence of cough seemed similar in patients receiving the combination compared with those receiving enalapril alone, and there was a lower incidence of edema in patients receiving combination than in those receiving nitrendipine monotherapy, but no conclusion could be drawn because of the small size of the groups. Earlier reports showed that combinations of ACE inhibitors and calcium blockers were associated with less drug-induced peripheral edema than calcium channel blockade alone (14,21).
In conclusion, the combination of nitrendipine and enalapril has been shown to be additive in blood pressure reducing effects compared with each of the monotherapies over a wide range of doses. This combination may be particularly useful for patients who respond insufficiently to monotherapy with nitrendipine or enalapril. The dose combination with the greater pronounced DBP-lowering effect was nitrendipine 20 mg plus enalapril 10 mg.
This study was funded by VITA INVEST, S. A. Barcelona, Spain. The study was coordinated by Alejandro Roca-Cusachs Coll, MD, from the Department of Internal Medicine of the Hospital de la Santa Creu i Sant Pau, Barcelona, Spain.
The following investigators participated in the study: María Dolors Berengué Iglesias, MD; Raquel Adroer Martori, MD; José Luis Ballvé Moreno, MD; Silvia Cañadas Crespo, MD; Xavier Monteverde Curto, MD; Luis Miguel Ruilope Urioste, MD; Carlos Campo Sien, MD; Luisa Fernández López, MD; Olga García Vallejo, MD; Lucía Guerrero Llamas, MD; Alejandro Roca-Cusachs Coll, MD; Teresa Benet Justi, MD; Pedro Aranda Lara, MD; Javier Aranda Lara, MD; José Manuel Logroño González, MD; Consuelo Simón Muela, MD; Isabel Duaso Allue, MD; Alberto Sanjuan Hernández-Franch, MD; Antonio Dalfó Baqué, MD; J. M. Bordas Julve, MD; Betlem Salvador González, MD; Joan J. Mascort Roca, MD; Josep M. Pérez Santos, MD; Xavier Puigdengolas, MD; Montserrat Tubau Grau, MD; Luis Revert Torrellas, MD; Jordi Calls Ginesta, MD; Anna Oliveras Serrano, MD; Vicente Masip Marza, MD; Teodoro Martín Jiménez, MD. Joan Llibre Bombardo, MD; Angel Rodríguez Jornet, MD; Matilde González Solanellas, MD; Núria Bastida Bastús, MD; Dolors Roselló Farràs, MD; Josep M a Segura Noguera, MD; Montserrat Pujol Anglada, MD; Xavier Mundet Tuduri, MD; Pedro Tomás Santos, MD; Jaime Merino Sánchez, MD; Fernando Quirze Amores, MD; Remedios Alarcón Barbero, MD; Marisol Botella Rodríguez, MD; José Ignacio Sánchez García, MD; Francisco Vicente Gil, MD; José M a Pascual Izuel, MD; Juan Gómez Octavio, MD; Mercedes Salcedo Torán, MD; Santiago Zelich Martínez, MD; Rosa Monteserín, MD; Marisa Galán Díez, MD; Elena Jordi Casas, MD; Rodrigo Munné, MD; M a Luisa Rodríguez Morató, MD; M a Eulalia Teixidó Fontanillas, MD; Antonio Ribera Castellano, MD; Eulalia Almendro Almanedro, MD; M a Pilar Esteras Iguacel, MD; Dolors Lumbreras Garluz, MD; Carmen Martínez Altarriba, MD; Emilio Márquez Contreras, MD; José L. Martín de Pablos, MD; Francisco Atienza Martín, MD; Ferran Cordón Granados, MD; Pedro J. Ferrer, MD; Pilar Font Roura, MD; J. Carlos González Pastor, MD; Pascual Solanas, MD; Carlos Rodríguez, MD; M a Luisa Rubio Montanés, MD; Narcís Salleras, MD; Ángel Cano Romera, MD; Roser Espona Barris, MD; Francesc López Expósito, MD; Ana López Plana, MD; Luis Llosa Dessy, MD; Carmen Pous Ruiz, MD.
© 2001 Lippincott Williams & Wilkins, Inc.