#DreamRCT candidate: Prevent DeaDD

Go ahead I dare you to tell me that is not a killer name

This DreamRCT entry was written by Swapnil Hiremath

This is another candidate for the #DreamRCT in Nephrology. Thanks to Joel, for hosting this on his Medium feed and helping with the editing. There have been some notable candidates already, including uric acid causing CKD; the Phosphate trial, IMAGINE, Phatom-1 and O My.

To these, I would like to add Prevent DeaDD: Prevent DEAth in Dialysis patients with Defibrillators.

Addressing traditional cardiovascular risk factors in dialysis patients have not been an effective strategy to reduce mortality. Notably, statins which are so effective at secondary prevention in non-dialysis patient have failed to reduce mortality in 4D and AURORA. ACEi inhibitors failed to reduce mortality in FOSIDIAL. One of the explanations is that, though cardiovascular mortality is the primary cause of death in dialysis patients, it is not due acute plaque rupture and myocardial infarction. Alternative causes of cardiovascular demise could be cardiomyopathy and decreased left ventricular dysfunction, which are both common in incident dialysis patients, and increase in severity with time on dialysis. Sudden cardiac death due to arrhythmia is another attractive candidate for cardiovascular mortality in dialysis patients. This is particularly appealing due to a few observations:

• The frequency of electrolyte disorders that could predispose to arrhythmia (hyperkalemia, I’ve got my eyes on you)
• A clear association between long interdialytic interval and mortality, suggesting hyperkalemia may be a culprit
• Increased mortality associated with lower dialysate potassium concentrations
• Decreased mortality with beta-blockers in HDPAL

To test this possibility I suggest a RCT of implanted cardiac defibrillators. Prevent DeaDD is a prospective, randomized, controlled, open label trial. The intervention group would have an ICD and the control group would receive usual care. The trial would enroll both incident and prevalent dialysis patients. In addition to this, they would require to have an indication for an implantable defibrillator for primary prevention (patients with a prior MI and LVEF < 30% and/or patients with a cardiomyopathy NYHA class II or III and LVEF < 35%). Though these are proven indications in non-dialysis patients, history shows that we should be cautious about extrapolating these benefits to dialysis patients. And there is evidence that this doubt is resulting in few dialysis patients recieving ICDs despite established (at least in renally intact cohorts) indications.

The primary outcome is all cause mortality. Since this is an open label trial, measurement of the outcome needs to be done carefully to avoid information bias, all-cause mortality is perfect because of the lack of subtltlety. Secondary outcomes are cardiovascular mortality and hospitalization, adjudicated by a blinded endpoint adjudication committee.

In addition to these, are important safety outcomes, especially given previous problems:

• acute complications with device insertion (bleeding, hematoma, local infection)
• long term complications (thrombosis, infections, central vein occlusions)
• Vascular access complications

Another secondary endpoint is patient quality of life (using the standardized KDQoL and SF36 scales). This is assessed at enrollment, one year, and at the study conclusion. This is a pragmatic trial, with all other factors ( access decisions, dialysis adequacy, medication use, coronary revascularization, etc.) left to the discretion of the treating nephrologist. It is possible that some patients in the control group will end up with defibrillators for other indications, as happens in other trials. However, the primary analysis will be strictly intention-to-treat.

Interim analyses will be carried out when 1/3rd and 2/3rd of the total predicted events occur to assess for futility or efficacy with predefined stopping rules. The power calculation will take these interim analyses into account.

Now, I am not the first one to be interested in this area. There have been cohort studies that suggest that ICDs are not as beneficial in dialysis patients (or CKD patients, for that matter) as they are in the general population. The obvious problem with these studies, is that they compare survival in dialysis patients with survival in the general population. There are a few published studies (ahem, like mine) which do choose the correct control group, namely matched dialysis patients with no defibrillators. There is also a recent meta-analysis with patient-level data from ICD trials, which suggests that there may be a non-significant benefit in CKD patients.

In fact, there is an ongoing study, ICD 2 that is attempting to answer this vary question. However, it is a pilot RCT, with only 200 patients, and I suspect it is having problems with recruiting since the completion date has been pushed back from 2012 to 2017!

An interesting question to examine regarding this study, is what will happen if it is a positive trial. Though consideration of cost is not yet commonly done south of the 49th parallel, it is of huge significance to the parts of the world that still use the metric system. Once an intervention is shown to be both more costly and more effective than the standard of care, it needs to show itself to be cost-effective. Whether it is ‘cost-effective’ depends on the societal ‘willingness to pay’, and this has been traditionally tied to the cost of dialysis. This was made for historical reasons, since we, as a society, had already decided to fund dialysis. It turns out that this works out to $50,000 per QALY (quality of life year) gained. Interestingly, the intervention of interest here, defibrillators, also turn out to have a remarkably similar cost-per-QALY. Now the strange paradox of these facts means that an effective intervention in dialysis patients will be inherently not cost-effective at this willingness to pay (read this fine paper to understand the fine analysis behind this statement).

For decades, nephrologists have been flailing about trying to improve dialysis survival. Repeated attempts to correct traditional CV risk factors have been futile. With this straightforward approach to the heart of the matter, I would like you to consider the Prevent DeaDD trial to accomplish this goal.