NHLBI Workshop on Future Directions for Hypertension Treatment Trials

September 17, 2003

Executive Summary

The Workshop on Future Directions for Hypertension Treatment Trials was held on September 17, 2003, at the Natcher Conference Center, chaired by Dr. Robert Califf with 55 participants--25 outside scientists plus NIH and FDA staff. (For a list of the non-NIH participants, see separate document.) The final agenda and statement of the objectives is also provided as a separate document. The agenda was compressed into one day only, because of the impending arrival of Hurricane Isabel on 9/18. Therefore, the refinement and prioritization of recommendations were completed by e-mail and conference calls.

The following recommendations emerged from the Breakout Groups and whole group discussions:

There is considerable evidence that clinicians have not fully accepted the guideline goal of treating systolic blood pressure (BP) to <140 mm Hg, while persons with systolic BP around this level and even below account for a large proportion of the BP-related morbidity/mortality. A trial was recommended to test the effect on cardiovascular (CV) events of lowering systolic BP from untreated levels of 130-149 mm Hg, or partially treated levels >140 mm Hg. The design might involve treatment to stated goals using a wide variety of antihypertensive (AHT) drugs open label (as in ACCORD), or use limited titration using a masked design (as in SHEP), or use allocation to a fixed 2- or 3-drug combination versus placebo. Participants would be selected to have increased risk of events by virtue of ALLHAT-like criteria, excluding diabetes (being tested in ACCORD), possibly focusing or over-sampling based on markers of chronic kidney disease (CKD), i.e. micro-albuminuria or reduced glomerular filtration rate (class 3 GFR).

After ALLHAT, a confirmatory "network meta-analysis", and JNC7, the emergent question of drug selection seems to be what drug to add to the majority of patients who should be treated with a thiazide-like diuretic. A trial was recommended to compare up to 4 classes of drugs as second-step therapy for effects on major CV events. Classes suggested for consideration were: an angiotensin converting enzyme (ACE) inhibitor, an angiotensin-1 receptor blocker (ARB), a calcium channel blocker, an aldosterone antagonist, and perhaps a beta-blocker. While some views of the state-of-the science would indicate that studying either an ACE inhibitor or an ARB would be suitable, other evidence from mechanistic studies, animal experimentation, and clinical trials suggest that these two drugs would be important to compare for stroke prevention in a high stroke risk population. Most viewed a focus on stroke prevention as a secondary goal of such a trial, as would assessing effects on kidney outcomes. Enrichment of the study population with CKD patients (as defined above) would be desirable.

Both (1) and (2) were highest priorities for trials focused on AHT treatment and clinical event outcomes. Various preferences for one or the other were expressed. Designs addressing both kinds of questions together were considered.

The rationale and extensive ongoing work in pharmacogenetics of hypertension were discussed, including GenHAT, an ancillary study of ALLHAT addressing the utility of candidate genes in predicting both the BP response and end-point results associated with the four randomized drug treatments. It was agreed that improvement of treatment may be possible with a gene-directed approach. Nevertheless, it would be premature to pursue a large trial of gene-guided therapy at this time. However, there was a strong recommendation that genetic considerations (particularly with regard to participant consent) be built into any large trial from the onset.

The idea of selecting an AHT regimen based on particular effects on subclinical CV disease above and beyond its BP-lowering effects was discussed, with a detailed focus on left ventricular mass (LVM) and chronic kidney disease (particularly subclinical albuminuria, an independent CV risk factor). (Other manifestations of subclinical disease, such as conduit arterial stiffness and carotid intimal-medial thickness, were not considered in detail.) Participants were skeptical that regimens could be identified that differentially lowered LVM or urinary protein excretion, and did not differ in BP effects. They were also doubtful that there would be widespread applicability of such an approach to LVM, due to the technical requirements and cost of evaluating LVM with precision. Detection of urinary protein excretion in a spot urine specimen is easy and inexpensive, and therefore should be considered as a secondary outcome of future large trials.

Physiological phenotyping as a guide to selection of therapy was not discussed in any detail.

Some participants expressed the view that it is not very realistic to target BP in isolation, especially if it is only slightly elevated. More applicable in practice and potentially conferring more benefit for the individual and the population would be intervention on multiple risk factors (quantified by a risk score) with multiple interventions. The extreme example of this approach is the "polypill" about which British authors have recently written—a combination of several AHT drugs in low dose, a statin, aspirin, and folate. Another version of the global risk approach would be to tailor a regimen to those with the metabolic syndrome, restricted to those with no individual risk factor that would in itself mandate any of the treatment components. There was no consensus on a specific approach to a clinical events trial of multiple interventions.

A few participants raised issues related to lifestyle interventions. It was pointed out that combined lifestyle intervention can lower BP as much as single-drug therapy, although it is probably much more difficult to implement in community practice. Indeed, major questions pertain to translation of known efficacious interventions to the community for treatment and prevention. (See 6 below.) This area was not a major focus for this Workshop, since it was organized to address clinical end-point trials. Some but by no means all participants were of the opinion that some lifestyle interventions should be tested for reduction of morbidity and mortality. One example of such a trial is the large Look AHEAD trial of weight loss and exercise in obese/overweight type 2 diabetics. However, the feasibility of obtaining clear answers in this arena is as yet unknown.

There was wide agreement that the greatest public health need in hypertension is to improve the reach and effectiveness of treatment, since only about 1/3 of hypertensive persons are treated and controlled to <140/<90 mm Hg. Several proposals dealt explicitly with alternate ways of delivering hypertension management. Although they were proposed as events trials, it was agreed that BP control would be a reasonable outcome measure for a program of various "health services trials". It was pointed out that 2 recent NHLBI Requests For Applications have encompassed such hypertension-related objectives, but that the number and scope of funded projects are limited.

The possibility was raised of establishing a Hypertension Clinical Research Network, involving both academic and community sites, to pursue not only a clinical events trial but several of the other research goals as well. Because of its potential efficiencies, this approach received considerable support.

The Workshop recommendations will be considered by the NHLBI in developing future initiatives in hypertension based on available funding.

Abbreviations: ACCORD=Action to Control Cardiovascular Risk in Diabetes trial; ALLHAT=Antihypertensive and Lipid Lowering Treatment to Prevent Heart Attack Trial; JNC7=Report of the Seventh Joint National Committee on Prevention, Detection, Evaluation and Treatment of High Blood Pressure; SHEP=Systolic Hypertension in the Elderly Program

August 2004