Substantial variation in SBP response to common antihypertensive drugs

Introduction and methods

Background

While multiple classes of highly effective blood pressure (BP)–lowering drugs are widely available [1], it is unknown whether the optimal choice of BP-lowering therapy varies between patients and whether personalized hypertension treatment can maximize clinical benefit.

Aim of the study

The study aim was to investigate and quantify the potential for personalized drug therapy in hypertension to maximize BP-lowering effects.

Methods

The PHYSIC (Precision Hypertension Care) trial was a randomized, double-blind, repeated crossover trial conducted at an outpatient research clinic of the Uppsala University Hospital in Sweden. In this trial, 280 patients aged 40–75 years with grade 1 hypertension (i.e., systolic BP (SBP): 140–159 mmHg) who were pharmacologically untreated or used BP-lowering monotherapy at inclusion (indicating a low-risk primary prevention sample) were included.

After completion of a 2-week run-in period with placebo, participants were assigned to a sequence of 6 treatment periods administered in random order. A treatment period (7–9 weeks’ duration) consisted of treatment with candesartan 16 mg (ARB), lisinopril 20 mg (ACEi), amlodipine 10 mg (calcium channel blocker), or hydrochlorothiazide 25 mg (thiazide). In addition, every patient repeated 2 of the treatment periods selected at random. Between each treatment period, there was a 1-week washout period with placebo. The primary analysis of the trial comprised 1486 completed treatment periods (median length: 56 days) in 270 patients.

Outcome

The primary endpoint was daytime (10:00–20:00 h) ambulatory SBP, measured at the end of each treatment period.

Main results

Variability in drug treatment effects

• SBP response to different treatments varied substantially (overall P<0.001). A higher mean BP was seen for patients treated with hydrochlorothiazide compared with any of the other 3 drugs (all P<0.001), for amlodipine compared with lisinopril (P=0.02), and for candesartan compared with lisinopril (P<0.001).
• Large variations in mean SBP were also observed between patients, within patients taking the same treatment, and between treatments in the same patient.

Potential for personalized treatment

• Using a fitted model, personalized treatment using single-drug therapy would on average lower SBP by 4.4 mmHg in the trial population compared with a fixed drug choice.
• As lisinopril was on average the most efficacious of the 4 drugs, personalized treatment would still lead to an SBP reduction of 3.1 mmHg compared with lisinopril.
• Major differences in treatment response were ruled out for the choices of lisinopril versus candesartan and hydrochlorothiazide versus amlodipine.
• The other 4 comparisons showed marked differences in treatment response, with particularly large benefits to be made by personalizing the choice between candesartan and amlodipine and between lisinopril and amlodipine.

Conclusion

References

Find this article online at JAMA.