PulsePredic Technology: Interview With Samuel Dudley, MD, PhD

PulsePredic is a new blood test that helps determine who benefits most from an ICD. Dr. Dudley is the Director of Cardiac Electrophysiology and Professor of Medicine, Section of Cardiology, at the University of Illinois at Chicago (UIC). He is also President and CEO of ROS Technologies, the innovator in the development of this blood test. 

Discuss the background of the PulsePredic and how it was developed. 

The PulsePredic blood test predicts sudden death risk in people with heart failure and ejection fractions less than 35 percent. Development started in 2005 from an observation that the cardiac sodium channel message (mRNA) was abnormally processed during heart failure. The intellectual property is held variably by the Veterans Administration, Emory University, and the University of Illinois in Chicago. The test is being developed by ROS Technologies, in collaboration with UIC, and in part is funded by an STTR grant from the NIH. 

Tell us about the purpose of the PulsePredic blood test and how it works. In which group of patients is the PulsePredic must useful for? 

The purpose of the test is to refine utilization of implanted cardiac defibrillators (ICD). The test has a high degree of sensitivity and specificity for prediction of appropriate ICD events, at least in the initial numbers that we’ve done in a clinical trial. The test can identify those people that would not need the device within the next year who would otherwise have an indication (i.e., an ejection fraction [EF] of less than 35 percent) and who may not need the device. It will also identify those people who have high abnormal mRNA splicing variants, and therefore would be advantaged by having an ICD. We know that somewhere between 60–70 percent of the patients with implanted defibrillators do not have appropriate ICD interventions, and between 15–40 percent of people who have the indication for an ICD never get one implanted. PulsePredic could help encourage utilization in that group that is now either declining and is not being offered an ICD and decrease utilization in the group that would not benefit from ICD implantation in the next year even though their ejection fraction is less than 35 percent. 

Describe the clinical trial in which it was studied. When was the study done? 

The trial results originally debuted at the ESC Congress 2012 in Munich. The trial consisted of 134 people in four groups: a control group, a group with congestive heart failure but no implanted device, a group with heart failure and an EF of less than 35 percent with an ICD but no appropriate ICD shocks, and a group with an EF less than 35 percent with an ICD in place with an appropriate ICD shock for ventricular arrhythmia. Each group had a higher number of cardiac sodium channel mRNA variants in their white blood cells as compared to the control group; however, the group that received an ICD shock had the highest variant levels, and they were much different than those people who had an ICD in place who were not shocked. The area under the curve for prediction of sudden death risk using either one of the variants was approximately 0.96. Thus, the test was highly sensitive and specific in discriminating subjects who were going to get shocked in a one-year time range. 

I should emphasize this is retrospective trial — we drew blood and looked back at the ICD records. The patients were studied over the last year and a half at the University of Illinois at Chicago, and the population in the trial is predominantly African American (60%) and Hispanic (20%), although in regression analysis we haven’t found any differences in race or gender. The test works in both ischemic and nonischemic patients. We’re currently planning the confirmatory, prospective trial for validation.

Tell us about further research being done with PulsePredic. When could the PulsePredic 

test be commercially available and approved by the FDA? 

Single-site validation for the trial has started, and we hope to initiate a larger 700- to 800-person trial soon. We hope to be up and running with this prospective validation trial by early spring, with between four and six sites. During the trial, we will draw blood and follow up with patients after their ICD implant for appropriate ICD shocks at one year. After that point, we hope to apply for a CE mark and potential FDA approval. 

How could this test change the way we treat patients at risk for sudden death?

Most tests available today aren’t very good in determining sudden death risk. The only criterion we currently and commonly use is EF, and in turns out that EF is not a very good predictor either — that is why 70 percent of people who get a defibrillator don’t have an appropriate event. In fact, there are few tests that predict sudden death, only: 1) T wave alternans; 2) signal-averaged ECG, which has a good negative predictive value, but not a good positive predictive value; and 3) EP studies, which don’t have a particularly good positive predictive value. There are no other blood tests available, so this could revolutionize how we decide on implanting defibrillators and really improve targeting the highest risk group. In refined ICD use, you could get the defibrillator to the patients who actually need it, and you would reduce expenditure, time, and complication rates in the people who didn’t need the device. 

There are different possible implications for this test depending on where you are in the world; for example, in the U.S. we implant defibrillators prophylactically, and this test will help target the correct group for implantation. In Japan, they implant mostly for secondary prevention, and this test encourages physicians to implant devices prophylactically. Therefore, the effect of the test worldwide is going to be different depending on where and how people use implanted devices. 

ROS Technologies has follow-up therapeutics that address reduced sodium channel levels that this test measures, which represent potential future therapies that could be linked with the diagnostic. If PulsePredic performs the way it has in the early clinical trials, it should dramatically improve our selection of the patients who actually need the device and those who don’t, so the use of ICDs will be much more refined, and the lost costs of the implanted devices in the people who didn’t need them will be reduced considerably. 

What kind of feedback did you receive about PulsePredic at the ESC Congress 2012?

It was very positive. With the performance of the test being so good and the positive initial trial results, the common response was that it was too good to be true! If the test continues to do well in the high-risk group, we may be able to start screening the intermediate risk group for ICD implantation, which would be completely new and totally revolutionize the way we manage malignant arrhythmic risk. 

Any limitations of the PulsePredic so far? 

There are the usual barriers to product development. So far we have not seen medications affect or alter the predictive power. We’ve looked mostly at class II, III, and IV heart failure patients, so it’s difficult to know if it will be predictive in completely asymptomatic heart failure subjects yet. We have about 30–40 percent women in the trials, and it looks like the test performs similarly in women. We have not seen a statistical difference between races, but that needs to be confirmed. The test is an RNA-based essay, so it is very similar to hepatitis C or HIV testing; it’s a little more cumbersome to do than protein-based testing, but certainly it is an established technology that is readily available in the clinical world now.

Is there anything else you’d like to add?

It is a very exciting technology, but it is a small trial. If it is valid and it comes to pass, it will completely change the way we risk stratify people for sudden death. I should emphasize to you that PulsePredic is not just a biomarker or some random protein that may or may not be pathologically related to arrhythmias. The way that this test works is that there is a splicing abnormality in the cardiac sodium channel (mRNA), so the message for making the protein gets truncated. This makes channels that are too short, and these short channels do not work. We already know that the cardiac sodium channel levels are critical for proper electrical activity. For example, Brugada syndrome is an inherited sudden death condition in which there is an abnormality in just one of the two cardiac sodium channel genes. This abnormal splicing measured by PulsePredic reduces cardiac sodium channels to the same level you see with Brugada syndrome. It turns out that the channel mRNA processing defect is identical in white cells as it is in the heart, and that is why we can use white cells as a surrogate for the heart. The test grew out of our studies in heart tissue. While looking for surrogates for hard to obtain human cardiac tssue, we investigated more easily obtained white cells. It turned out that white cells transcribe cardiac sodium mRNA, so it occurred to us that maybe the body works the same way. That is how we developed the blood test. It was really all by informed chance!