Hypertension Drug Effects Differ From Person to Person, PHYSIC Trial Shows
In the absence of simple biomarker tests to predict BP responses to different agents, the findings remain theoretical, not practical.
Substantial variation in the effect of antihypertensive medications from one individual to the next argues for the need to target the right therapy to the right person, according to data from the Swedish PHYSIC trial.
The variations were seen between treatments, between participants, within participants on the same therapy, and between treatments in the same participant on standard first-line agents from the four drug classes recommended for initial therapy in the American and European hypertension guidelines.
Johan Sundström, MD, PhD, (Uppsala University Hospital, Sweden), the study’s lead author said his group was surprised by the magnitude of the heterogeneity they observed.
“The additional reduction in blood pressure by receiving the ‘optimal’ blood-pressure medication was about twice as large as the additional reduction would be by doubling the dose of a blood pressure medication,” he told TCTMD in an email. “This is a clinically important effect size.”
The investigators say a personalized approach could be expected to reduce systolic BP an additional 4.4 mm Hg compared with starting patients off with a random guideline-recommended first-line therapy.
Speaking with TCTMD, Robert M. Carey, MD (University of Virginia Health System, Charlottesville), who was the second author on the US hypertension guidelines, agreed that the magnitude of the heterogeneity in response is surprising.
“I don't think there is any hard literature except for this article highlighting the heterogeneity of blood pressure responses in a good comparison trial with good statistical methods,” he said. “It's significant enough that it does suggest that a personalized approach to treatment may be possible in the context of primary hypertension.”
In an accompanying editorial, Carey notes that the findings are of considerable interest given that only about one-quarter of adults with hypertension are well controlled on medication to < 130 mm Hg/80 mm Hg, with evidence that rates of poor control on therapy are increasing.
Carey adds that while the small study demonstrates the feasibility of a personalized approach to hypertension and the reality of substantial heterogeneity in drug response, the findings “are more theoretical than immediately practical for the implementation of personalized antihypertensive drug therapy.”
While such an approach may eventually supplement or even supplant current methods of antihypertensive therapy selection, larger trials are needed to clarify the practical ramifications of targeted therapy, Carey adds Specifically, it would be necessary to clarify what would be required of physicians to accomplish this, since the type of short-term testing of individual drugs like in the study would be cumbersome in clinical practice.
In response to that, Sundstrom told TCTMD that his group isn’t advocating for the year-long process used in their study to test out the different regimens on each patient.
“We could rule out any personalization potential for some treatment contrasts, so those need not be tested in clinical practice. Beyond that, identifying mechanisms to enable this personalization in routine clinical practice should now be a priority, given the size of the likely benefits,” Sundstrom said.
PHYSIC Trial Surprises
The PHYSIC trial randomized 280 hypertensive participants between the ages of 40 and 75 years (54% men; mean age 64) who were either not taking pharmacologic therapy for hypertension or were on prescribed monotherapy and being seen at an outpatient research clinic. The mean length of hypertension was 3 years, with approximately 60% of patients having used BP-lowering monotherapy. The mean in-office BP prior to the intervention 154/89 mm Hg.
Participants were randomized to drugs from each of the four guideline-recommended drug classes (ACE inhibitors, ARBs, calcium-channel blockers, and diuretics), with repeat treatment periods for two randomly selected agents (for a total of six treatment periods per patient). This crossover format with repeat testing of the same drug aided in establishing both within-patient and between-patient differences in BP response.
All patients were treated for periods of 7 to 9 weeks with 16 mg candesartan, 20 mg lisinopril, 10 mg amlodipine, and 25 mg hydrochlorothiazide, with 1-week washout periods with placebo between each treatment period and 24-hour ambulatory BP monitoring at the end of each treatment period.
The variations in BP measurements were widespread and significant (P < .001). On average, BP at the end of each study period was higher on hydrochlorothiazide than other medications; higher on amlodipine than lisinopril; and higher on candesartan than lisinopril.
“It was clear that some patients had lower blood pressure from one drug than from another,” Sundström said. “This challenges the strategy recommended in current treatment guidelines, where four drug groups are equally recommended for everyone with high blood pressure.”
While lisinopril emerged as the most efficacious of the drugs at the selected doses, the investigators suggest that a personalized treatment approach versus simply selecting lisinopril “would still lead to a 3.1 mm Hg systolic BP improvement.”
In the case of some drugs, large differences in response were not seen. Such was the case for candesartan versus lisinopril, for example, as well as amlodipine versus hydrochlorothiazide.
“The absence of any potential benefit from choosing between the two agents inhibiting the renin angiotensin-aldosterone system provides some reassurance about the validity of the study—these two agents share multiple aspects of their mechanisms of action,” Sundström and colleagues write. “In the same way, the benefits of personalization observed for four of the five other comparisons between drugs with quite different mechanisms of action aligns with expectations, though the reason for the absence of a potential for benefit from personalizing hydrochlorothiazide vs amlodipine therapy is unclear.”
According to Carey, the PHYSIC results should encourage further exploration of this work, “with longer follow-up data to determine whether there is improvement in long-term clinical outcomes compared with current strategies.”
Carey says the study also supports a role for biomarkers in drug-response prediction, although that avenue is limited at the moment due to the availability of very few phenotypic markers.
“In the past, for example, plasma renin activity was recommended for this purpose, but its use was later abandoned due to lack of specificity,” Carey writes. “Nevertheless, this trial supports the discovery of new biomarkers that can accurately predict BP responses to single antihypertensive agents.”
To TCTMD, he said the importance of the PHYSIC study is that “it leaves no doubt of the heterogeneity in response,” and keeps the door open to further efforts to match patients to the most effective antihypertensive for them.
Sundström J, Lind L, Nowrouzi S, et al. Heterogeneity in blood pressure response to four antihypertensive drugs: a randomized clinical trial. JAMA. 2023;329:1160-1169.
Carey RM. Is personalized antihypertensive drug selection feasible? JAMA. 2023;329:1153-1154.
- Sundström reports owning stock in Symptoms Europe AB and Anagram Kommunikation AB.
- Carey reports no relevant conflicts of interest.