High stakes, thin data: Why cardiovascular research must do better for hemodialysis patients
When I prescribe a heart medication or recommend a procedure for a patient receiving hemodialysis (treatment that filters waste and excess fluid from the blood when the kidneys have failed), I often am asked how much this intervention will help them.
In this situation, I may reply, “We try to apply what we know from analyzing data involving patients who are not on hemodialysis to understand whether it will benefit you; but we can’t always be sure.” For patients like these, who spend considerable time in the hospital, the evidence gap in outcomes for patients receiving hemodialysis is a research failure.
Patients receiving hemodialysis face some of the highest cardiovascular risks of any population in medicine, with staggering rates of heart attack, stroke, heart failure and sudden cardiac death. Yet when landmark trials of statins, heart failure therapies or atrial fibrillation strategies are designed, these patients are routinely excluded up front or included only in small, underpowered subgroups.
We often have to rely on data on patients without end-stage kidney disease or observational data to make decisions. The result is a paradox: those most likely to have heart events are those least likely to be represented in the evidence that shapes guidelines, reimbursement and bedside care.
In our recent JACC state-of-the-art review on cardiovascular trials in patients receiving hemodialysis, my co‑authors from the cardiovascular clinical trialist forum and I mapped out what has been tested in this population. For one example, the most robust evidence we have relates to statins, which have been shown to help prevent heart attacks and strokes in the general population. However, statin trials such as 4D and AURORA, and subgroup analyses from the SHARP trial, largely failed to show clear cardiovascular benefit in patients receiving dialysis despite robust effects in earlier‑stages of kidney disease.
Trials focused on dialysis‑specific complications, such as calcimimetics, vitamin D or dialysis strategies, have mostly shown that patients are no better or worse after these medications, with a few notable exceptions, such as proactive IV iron in the PIVOTAL trial.
When we default to using data from non‑dialysis populations to make treatment decisions for hemodialysis patients, we ask them to accept the risks of drugs and devices without reasonably knowing whether they will benefit directly. This scenario is both clinically challenging and ethically problematic.
At the same time, drug sponsors often avoid conducting dedicated dialysis trials because of the perceived complexity in trial design, recruitment challenges and the fear that high event rates and competing risks will “muddy” primary endpoints used to determine whether the drug is safe and effective. While including only patients most likely to benefit in the trial improves the primary outcome, patients on hemodialysis are left without reliable treatment options because we rarely study them.
In our review, we highlight solutions to the challenges that have led to this issue. There are clear paths forward if we treat generation of data for patients on hemodialysis as an ethical obligation. Regulators could require or strongly incentivize participation in phase 1 studies and dedicated outcome trials in this population when cardiovascular drugs are developed for broad use. Trial sponsors and health systems can support pragmatic trials embedded in dialysis units by streamlining consent, linking registries, maintaining regular follow-up schedules and joint lab-related follow-ups. Trials also may be restructured to track outcomes that account for the added risks faced by hemodialysis patients.
An often-overlooked perspective is that, if such trials are attempted, they may, in fact, benefit from shorter follow-up periods than trials in patients not on hemodialysis. This is because of the high cardiovascular risk encountered by patients on hemodialysis; these patients may have a greater number of heart attacks, strokes or other cardiovascular disease earlier. So, the efficiency of trials in this population may be underestimated. Our goal must be investment in trials that include only patients on hemodialysis at the same time as, or directly after, demonstrating benefit in the general population. This will require investment from patients, trialists, funders and regulators.
Most importantly, we need to invite people on hemodialysis into the research agenda, not just as subjects but as partners. Many of my own patients have told me they would participate in trials if someone explained why the scientific question matters for them and the potential benefit. Clear communication, trust‑building in dialysis centers, collaboration with nephrology and cardiology along with patients receiving hemodialysis across institutions can turn a “hard‑to‑recruit” population into a highly engaged one.
At its core, this is about equity in cardiovascular innovation. Therapies that transform outcomes for the general population but leave behind those with kidney failure are only a partial success. As cardiologists, nephrologists, scientists and policymakers, we should hold ourselves to a higher standard: no cardiovascular breakthrough is truly complete until it has been fairly tested in the patients who may need it most.
By Dr. Arsalan Hamid, a cardiology fellow at Baylor College of Medicine
