Different Effects of Antihypertensive Regimens Based on Fosinopril or Hydrochlorothiazide With or Without Lipid Lowering by Pravastatin on Progression of Asymptomatic Carotid Atherosclerosis
Principal Results of PHYLLIS—A Randomized Double-Blind Trial
Background and Purpose— The Plaque Hypertension Lipid-Lowering Italian Study (PHYLLIS) tested whether (1) the angiotensin-converting enzyme (ACE) inhibitor fosinopril (20 mg per day) was more effective on carotid atherosclerosis progression than the diuretic hydrochlorothiazide (25 mg per day), (2) pravastatin (40 mg per day) was more effective than placebo when added to either hydrochlorothiazide or fosinopril, and (3) there were additive effects of ACE inhibitor and lipid-lowering therapies.
Methods— A total of 508 hypertensive, hypercholesterolemic patients with asymptomatic carotid atherosclerosis were randomized to: (A) hydrochlorothiazide; (B) fosinopril; (C) hydrochlorothiazide plus pravastatin; and (D) fosinopril plus pravastatin, and followed up blindly for 2.6 years. B-Mode carotid scans were performed yearly by certified sonographers in 13 hospitals and read centrally. Corrections for drift were calculated from readings repeated at study end. Primary outcome was change in mean maximum intima-media thickness of far and near walls of common carotids and bifurcations bilaterally (CBMmax).
Results— CBMmax significantly progressed (0.010±0.004 mm per year; P=0.01) in group A (hydrochlorothiazide alone) but not in groups B, C, and D. CBMmax changes in groups B, C, and D were significantly different from changes in group A. Changes in group A were concentrated at the bifurcations. “Clinic” and “ambulatory” blood pressure reductions were not significantly different between groups, but total and low-density lipoprotein cholesterol decreased by ≈1 mmol/L in groups C and D.
Conclusions— Progression of carotid atherosclerosis occurred with hydrochlorothiazide but not with fosinopril. Progression could also be avoided by associating pravastatin with hydrochlorothiazide.
Ultrasound measurement of carotid intima-media thickness (IMT) is used to measure asymptomatic atherosclerosis progression. Studies in hypertensive patients showed a significant slowing down with calcium antagonists compared with diuretics1,2 or β-blockers.3 Two trials4,5 explored the effects of angiotensin-converting enzyme (ACE) inhibitors versus placebo, with contradictory results. A number of studies tested the effects of lipid lowering by statins on carotid IMT,6–12 showing significant retardation of IMT progression, but none was conducted in hypertensives.
The Plaque Hypertension Lipid-Lowering Italian Study (PHYLLIS) aimed to test in hypertensive patients with hypercholesterolemia and asymptomatic atherosclerosis whether (1) antihypertensive therapy with the ACE inhibitor fosinopril was more effective on carotid atherosclerosis progression than antihypertensive therapy with hydrochlorothiazide; (2) lipid lowering with pravastatin was more effective than placebo when associated with either hydrochlorothiazide or fosinopril; and (3) there were additive effects of ACE inhibitor and lipid-lowering therapies on carotid atherosclerosis.
Participants and Interventions
In 13 Italian hospitals, men and postmenopausal women aged 45 to 70 years with untreated or uncontrolled hypertension, hypercholesterolemia, asymptomatic carotid atherosclerosis (maximum carotid IMT, Tmax, 1.3 to 4.0 mm), no previous cardiovascular events, and giving informed consent underwent 6-week washout under triple placebo and American Heart Association low-lipid diet. At washout end, those maintaining systolic blood pressure (SBP) 150 to 210 mm Hg, diastolic blood pressure (DBP) 95 to 115 mm Hg, serum low-density lipoprotein (LDL) cholesterol 4.14 to 5.17 mmol/L (160 to 200 mg/dL), and triglycerides ≤3.39 mmol/L (≤300 mg/dL) were randomized to: (A) hydrochlorothiazide, 25 mg QD plus fosinopril placebo and pravastatin placebo; (B) fosinopril, 20 mg QD plus hydrochlorothiazide placebo and pravastatin placebo; (C) hydrochlorothiazide, 25 mg QD, and pravastatin, 40 mg QD plus fosinopril placebo; (D) fosinopril, 20 mg QD, and pravastatin, 40 mg QD plus hydrochlorothiazide placebo. Randomization was computer generated with a block size of 4. If DBP was not <90 mm Hg or <95 mm Hg with a fall of ≥10 mm Hg, open-label nifedipine gastrointestinal therapeutic system (GITS), 30 mg QD, was added after 3 months to be eventually increased to 60 mg after 6 months. Low-lipid diet was maintained throughout the study. Patients and study personnel were blinded to treatment assignment. The protocol was approved by ethics committees of all institutions involved.
Certified sonographers at 13 participating units performed duplicate B-mode carotid scans at randomization and subsequently at yearly intervals during 3 years by a Biosound 2000IISA (Biosound) with an 8-MHz annular array transducer. All scans were read at the Ultrasound Coordinating Center. Three mercury manometer measurements of sitting clinic blood pressure were taken at baseline and every 3 months thereafter. Ambulatory blood pressure monitoring was done at baseline, yearly intervals, and study end, and read centrally. Blood chemistry was performed at baseline and every 6 months in certified laboratories. All data were handled by an independent statistical analysis center.
Primary outcome was rate of change in mean maximum IMT of the 8 far and near walls in distal common carotids and bifurcations bilaterally (CBMmax). Duplicate scans were read blindly during study. Readers’ drift was checked from rereading single scans of 4 far walls at study end, with time sequence randomized. Differences between old and new readings were estimated by linear regression against treatment time and original values corrected by subtracting fitted values at corresponding dates from original values. Slopes were calculated from a mean (SD) of 3.6 (0.6) time points in duplicate and from a mean (SD) of 7.73 (0.72) walls. Secondary outcomes were: (1) changes in mean maximum IMT of the 4 far and near walls in distal common carotids (CC-IMT) and separately in carotid bifurcations (Bif-IMT); (2) changes in clinic and ambulatory blood pressure; and (3) changes in serum total, LDL, and high-density lipoprotein (HDL) cholesterol and other laboratory variables.
A sample size of 500 (125 for treatment group) with a 20% dropout rate was calculated to provide 2-sided 5% significance and 80% power to detect a mean (±SD) difference in CBMmax changes of 0.0150 (±0.0373) mm per year (assumption derived from Verapamil in Hypertension and Atherosclerosis Study [VHAS]1 and Carotid Atherosclerosis Italian Ultrasound Study [CAIUS]10).
Treatment-related IMT changes were calculated as means±SE and 95% CIs of yearly progressions (slopes), and of differences between final and baseline scans and significance tested by ANOVA. The analysis plan specified pairwise comparisons of groups B, C, and D with group A. Because each pairwise comparison was intended to answer different questions (B versus A, superiority of fosinopril over hydrochlorothiazide in subjects not receiving pravastatin; C versus A, superiority of pravastatin over placebo in subjects receiving hydrochlorothiazide; D versus A, superiority of fosinopril plus pravastatin over hydrochlorothiazide plus placebo), no Bonferroni adjustment of P values was used to control experimental type I error, as suggested by Cook and Farewell.13 Differences in IMT changes between each pair of the 4 treatment groups were expressed as means±SE and 95% CI and tested by ANCOVA with treatment and baseline IMT as covariates. No logarithmic transformation of data was found to be necessary. The analysis plan also included a test of interaction to provide indication on possible additive effects of the 2 treatments on the primary outcome.
On-treatment clinic and ambulatory blood pressures and heart rates and metabolic variables were compared with baseline by paired t test, and changes between randomized groups compared by unpaired t test or multiple-measurement ANCOVA. All significance tests were 2-sided.
The present article reports intention-to-treat analyses.
A total of 950 patients were screened by ultrasound and 508 patients randomly allocated to double-blind treatment (average treatment duration 2.6 years). Baseline characteristics are listed in Table 1. There was no major between-group difference except for smokers (fewer in group D).
A total of 79.8% of patients in group A, 80.9% in B, 84.2% in C, and 84.2% in D remained on monotherapy with hydrochlorothiazide or fosinopril until the last study visit. The remainder received additionally open-label nifedipine GITS, 30 to 60 mg daily. A total of 98% of patients in groups C and D remained on pravastatin. No patients in groups A and B were on pravastatin or another lipid-lowering agent.
Treatment-Related Changes in Carotid Wall
The primary variable CBMmax significantly progressed by 0.010±0.004 mm per year (P=0.01), with a final difference from baseline of 0.025±0.011 mm (P=0.03) in group A (hydrochlorothiazide alone). There was no significant progression in the other treatment groups. Slopes of CBMmax changes in groups B (fosinopril alone; −0.002±0.004 mm; NS), C (hydrochlorothiazide+pravastatin; −0.002±0.004 mm; NS), and D (fosinopril+pravastatin; −0.002±0.004 mm; NS) were all significantly different (P=0.02 to 0.03) from slope in group A (hydrochlorothiazide alone); B versus A −0.012 mm per year, P=0.03; C versus A −0.012 mm per year, P=0.03; D versus A −0.012 mm per year, P=0.02), but not between themselves (Figure 1). Identical differences between treatment groups with identical levels of significance were found when using original measurements uncorrected by final controls for possible drift. Considering CBMmax differences between final and baseline scans (Figure 1), the same trends were observed, although only the difference between groups D and A formally achieved statistical significance (P=0.04). As indicated by the interaction test (P=0.08), the assumption of additive effects appeared unlikely; indeed, the effect of fosinopril plus pravastatin on CBMmax slope was identical to that of fosinopril alone.
Changes in yearly progression (Figure 2, left) or final difference from baseline in CC-IMT were small and nonsignificant in group A, and no significant progression in CC-IMT occurred in any of the other groups. However, a marked and significant yearly progression in Bif-IMT was found in group A with no significant progression in the other groups (Figure 2, right). Significant differences in yearly Bif-IMT progression were seen by comparing groups B (−0.023 mm per year; P=0.012), C (−0.019 mm per year; P=0.037), or D (−0.22 mm per year; P=0.007) with group A.
Carotid internal diameter changed very slightly and nonsignificantly in the 4 groups, and Spearman correlation coefficients with CBMmax changes were extremely small (A −0.0404; B 0.0117; C 0.0942; D 0.0091) and statistically nonsignificant. Consequently, adjustments of CBMmax progressions using diameter changes as covariate influenced results in an irrelevant way.
Treatment-Related Changes in Blood Pressure, Heart Rate, and Metabolic Variables
Clinic SBP, DBP, and pulse pressure (PP) decreased markedly and significantly (P<0.001) in all groups, with no significant between-group differences. Changes in ambulatory SBP, DBP, and PP were smaller but all significant versus baseline (SBP and DBP, P<0.001; PP P<0.05) and nonsignificantly different between groups (Figure 3). Heart rate did not significantly change in any group.
Metabolic changes (Table 2) consisted in marked and highly significant reductions in total and LDL cholesterol in the groups receiving pravastatin, but a significant although smaller decrease also occurred in group B. HDL cholesterol significantly increased in group D, and triglycerides in group A. Serum urate significantly increased and serum potassium significantly decreased in groups receiving hydrochlorothiazide.
There were too few cardiovascular events among PHYLLIS patients to allow statistical evaluation: 4 myocardial infarctions (3 in A, 1 in D), 1 stroke (D), 1 cardiovascular death (D), and 2 cases of cancer (1 each in B and C). Patients with creatine-phosphokinase increases to >600 U/L were 1, 5, 5, and 2 in each group.
PHYLLIS conclusions rest on careful recording of duplicate carotid scans by trained sonographers using the same high-precision instruments and probes, with centralized reading, study personnel blinded to treatment, and correction for possible drift or bias. Fosinopril and pravastatin exerted their effects predominantly or exclusively at the bifurcation, a preferred location for the hemodynamic effects of hypertension and for atherosclerosis, and this strongly supports the conclusion that fosinopril and pravastatin blunt progression of carotid atherosclerosis.
PHYLLIS shows ACE inhibitors may exert an antiatherosclerotic action, a matter until now undecided, confirming for fosinopril versus hydrochlorothiazide the finding of the Study to Evaluate Carotid Ultrasound changes in patients treated with Ramipril and Vitamin E (SECURE)4 for ramipril versus placebo, and explaining the negative results of PART-2,5 because this study only investigated common carotid IMT, which was uninfluenced by fosinopril in PHYLLIS. The IMT effect of fosinopril versus hydrochlorothiazide (−0.012 mm per year) had similar order of magnitude as ramipril versus placebo in SECURE4 (−0.008 mm per year), calcium antagonists versus diuretics in VHAS1 (−0.010 mm per year) and INSIGHT2 (−0.0081 mm per year), and versus atenolol in European Lacidipine Study on Atherosclerosis (ELSA)3 (−0.006 mm per year).
PHYLLIS also shows that lipid-lowering drugs reduce progression of carotid IMT in hypertensive patients, not only in those with vascular or coronary disease as in previous studies.6–12 The extent of the CBMmax effect of pravastatin in PHYLLIS (−0.012 mm per year) was of the same magnitude as in previous placebo-controlled studies measuring carotid CBMmax or Mmax (−0.009 to −0.019 mm per year)6–8,10,12 In PHYLLIS, the effect of pravastatin was predominantly found at the bifurcation, as in a previous study,10 although other studies found an effect on CC-IMT.7–9,11,12 In PHYLLIS, CC-IMT did not significantly progress under hydrochlorothiazide, and this might have prevented any further pravastatin (as well as fosinopril) action at that level.
Several tentative explanations can be offered for the absence of additive effect of fosinopril plus pravastatin: (1) because either fosinopril or pravastatin fully blocked progression of carotid IMT, no further effect could be seen by using the 2 drugs together; (2) a higher statin dose with greater cholesterol reduction was necessary; (3) additive effects may have required a time span >2.6 years; and (4) associating an ACE inhibitor and a statin may influence composition rather than thickness of carotid wall or plaques.
We can exclude that the greater antiatherogenic action of fosinopril versus hydrochlorothiazide was related to greater BP reduction because clinic and ambulatory values were never reduced to a greater extent by fosinopril. A calcium antagonist with documented action on carotid IMT2 was used in a minority of patients and exactly in the same proportion in groups A and B, and in a slightly lower proportion in groups C and D. Animal experiments with various ACE inhibitors14 have shown efficacy on various models of atherosclerosis and suggested actions on endothelium or other vascular components. The antiatherosclerotic effect of pravastatin in PHYLLIS was not attributable to greater BP reduction but was associated with marked decrease in total and LDL cholesterol, some increase in HDL cholesterol, and possibly, antiproliferative and antinflammatory actions of statins.15
There were too few events in PHYLLIS patients to be correlated with carotid changes or treatment regimens. Therefore, it cannot be concluded from our observations that the smaller carotid IMT progression occurring with fosinopril (versus hydrochlorothiazide) or pravastatin added to hydrochlorothiazide is necessarily associated with a lower cardiovascular risk. However, the IMT changes observed with fosinopril (versus hydrochlorothiazide) and associating pravastatin with hydrochlorothiazide were of similar magnitude to those induced by lipid-lowering drugs in patients with coronary heart disease accompanied by reduced incidence of cardiovascular events.6,7 Cholesterol changes were also of the same magnitude as in a recent study in hypertensives in which lipid lowering significantly decreased cardiovascular event incidence.16
The finding that ACE inhibitors and diuretics have a comparable effect on cardiovascular morbidity and mortality17 (nonetheless, a nonunanimous finding18) does not necessarily suggest that differences in carotid atherosclerosis progression described in PHYLLIS between fosinopril and hydrochlorothiazide have no clinical relevance. On the contrary, it may support the opinion of recent European hypertension guidelines19 that event-based trials may be too short and commonly performed in too complicated hypertensives to detect differences in morbidity/mortality between different active antihypertensive treatments. Trials showing treatment-related differences in organ damage (ie, on alterations that in the history of cardiovascular disease are intermediate in time) may provide useful indications on a longer-term perspective.
A. Zanchetti, G. Crepaldi, M.G. Bond.
E. Ambrosioni, G. Baggio, C. Dal Palù, G.V. Gallus, B. Magnani, G. Mancia, M. Mancini, R. Paoletti, A. Ventura, G.B. Leproux, A. Magni.
Ultrasound Coordinating Center
Ambulatory Blood Pressure Monitoring Coordinating Center
G. Mancia, G. Parati.
Statistical Analysis Center
G.V. Gallus, F. Veglia.
Data Monitoring Unit
M. Ricci, E. Serrotti.
A. Rappelli, P. Dessi-Fulgheri, E. Espinosa, R. Catalini, O. Zingaretti (Ancona); A. Capurso, N. Barile, C. Macchiarulo, A. Venezia (Bari); A. Gaddi, S. Rimondi, L. Finazzo, C. Mussoni (Bologna); E. Agabiti-Rosei, M.L. Muiesan, C. Monteduro, M. Salvetti (Brescia); R. Balestreri, G. Lotti, S. Bertolini, N.R. Musso, C. Barone, O. Franza (Genova); G. Leonetti, L. Sampieri, C. Cuspidi, R. Chianca, M. Bombelli, G. Foglia, S. Pellizzoli, G. Pontiggia (Milano); C. Sirtori, F. Pazzucconi, D. Baldassarre (Milano); F. Pasanisi, A. Iannuzzi, F. Faccenda (Napoli); A. Semplicini, A. Pontebasso, A. Calabro, G. Brisotto (Padova); G. Ciuffetti, M. Politano, G. Brunetti, B. Fiorucci (Perugia); A. Salvetti, A.H. Basen, S. Buralli (Pisa); L. Campanacci, G. Guarnieri, G. Bellini, L. Cattin, M. Fisicaro, M. Tonizzo (Trieste); A. Lechi, M. Muggeo, M. Ribul, D. Travia (Verona).
PHYLLIS was an investigator-generated trial sponsored by Bristol-Myers Squibb Italy, Rome, and Menarini, Florence. All authors have received research grants or lecture honoraria from the sponsors.
Dr Sperti is an employee of Bristol-Myers Squibb, and Dr Magni is an employee of Menarini.
- Received June 21, 2004.
- Revision received August 24, 2004.
- Accepted September 16, 2004.
Zanchetti A, Agabiti-Rosei E, Dal Palù C, Leonetti G, Magnani B, Pessina A. The Verapamil in Hypertension and Atherosclerosis Study (VHAS): results of long-term randomized treatment with either verapamil or chlorthalidone on carotid intima-media thickness. J Hypertens. 1998; 16: 1667–1676.
Simon A, Gariépy J, Moyse D, Levenson J. Differential effects of nifedipine and co-amilozide on the progression of early carotid wall changes. Circulation. 2001; 103: 2949–2954.
Zanchetti A, Bond M G, Hennig M, Neiss A, Mancia G, Dal Palù C, Hansson L, Magnani B, Rahn K-H, Reid JL, Rodicio J, Safar M, Eckes L, Rizzini P. Calcium antagonist lacidipine slows down progression of asymptomatic carotid atherosclerosis. Principal results of the European Lacidipine Study on Atherosclerosis (ELSA), a randomized, double-blind, long-term trial. Circulation. 2002; 106: 2422–2427.
Lonn EM, Yusuf S, Dzavik V, Doris Cl, Yi Q, Smith S, Moore-Cox A, Bosch J, Riley WA, Teo KK. Effects of ramipril and vitamin E on atherosclerosis: the Study to Evaluate Carotid Ultrasound changes in patients treated with Ramipril and Vitamin E (SECURE). Circulation. 2001; 103: 919–925.
MacMahon S, Sharpe N, Gamble G, Clague A. Mhurchu C N, Clark T, Hart H, Scott J, White H. Randomized, placebo-controlled trial of the angiotensin-converting enzyme inhibitor, ramipril, in patients with coronary or other occlusive arterial disease. J Am Coll Cardiol. 2000; 36: 438–443.
Furberg CS, Adams HP, Applegate WB, Byington RP, Espeland MA, Hartwell T, Hunninghake DB, Lefkowitz DS, Probstfield J, Riley WA, Young B. Effect of lovastatin on early carotid atherosclerosis and cardiovascular events. Circulation. 1994; 90: 1679–1687.
Salonen R, Nyyssönen K, Porkkala E, Rummukainen J, Belder R, Park J-S, Salonen JT. Kuopio Atherosclerosis Prevention Study (KAPS). A population-based primary preventive trial of the effect of LDL lowering on atherosclerosis progression in carotid and femoral arteries. Circulation. 1995; 92: 1758–1764.
Mercuri M, Bond MG, Sirtori CR, Veglia F, Crepaldi G, Feruglio FS, Deskovic G, Ricci G, Rubba P, Mancini M, Gallus G, Bianchi G, D’Alò G, Ventura A. Pravastatin reduces carotid-intima-media thickness progression in an asymptomatic hypercholesterolemic Mediterranean population: the Carotid Atherosclerosis Italian Ultrasound Study. Am J Med. 1996; 101: 627–634.
MacMahon S, Sharpe N, Gamble G, Hart H, Scott J, Simes J White H. Effects of lowering average or below-average cholesterol levels on the progression of carotid atherosclerosis. Results of the LIPID Atherosclerosis Substudy. Circulation. 1998; 97: 1784–1790.
Hedblad B, Wikstrand J, Janzon L, Wedel H, Berglund G. Low-dose metoprolol CR/XL and fluvastatin slow progression of carotid intima-media thickness. Main results from the β-Blocker Cholesterol-lowering Asymptomatic Plaque Study (BCAPS). Circulation. 2001; 103: 1721–1726.
Ridker PM, Rifai N, Pfeffer MA, Sacks FM, Moye LA, Goldman S, Flaker GC, Braunwald E. Inflammation, pravastatin, and the risk of coronary events after myocardial infarction in patients with average cholesterol levels. Cholesterol and Recurrent Events (CARE) Investigators. Circulation. 1998; 98: 839–844.
Sever PS, Dahlöf B, Poulter NR, Wedel H, Beevers G, Caulfield M, Collins R, Kjeldsen S, Kristinsson A, McInnes GT, Mehlsen J, Nieminen M, O’Brien E, Ostergren J; ASCOT investigators. Prevention of coronary and stroke events with atorvastatin in hypertensive patients who have average or lower-than-average cholesterol concentrations, in the Anglo-Scandinavian Cardiac Outcomes Trial—Lipid Lowering Arm (ASCOT-LLA): a multicentre randomised controlled trial. Lancet. 2003; 361: 1149–1158.
Wing LMH, Reid CM, Ryan P, Beilin LJ, Brown MA, Jennings GLR, Johnston CI, McNeil JJ, Macdonald GJ, Marley JE, Morgan TO, West M. A comparison of outcomes with angiotensin-converting-enzyme inhibitors and diuretics for hypertension in the elderly. N Engl J Med. 2003; 348: 583–592.