Diagnostics & Imaging Week Washington Editor

GAITHERSBURG, Maryland — FDA reviewers at the Center for Devices and Radiological Health are looking at another batch of close calls this year, and its circulatory system advisory panel held two of such toughies last week on Thursday and Friday.

On Thursday, the panel examined — and by a 9-2 vote recommended against — the application for the Chronicle implantable hemodynamic monitoring system, from Medtronic (Minneapolis), apparently sunk by its failure to meet a primary efficacy endpoint — reduced trips to the hospital for care.

Virtually every member of the panel praised the study design, the company’s presentation and the device’s purpose. But their discussion indicated that a major problem was in painting the Chronicle as a therapeutic device when it would have been presented better as a diagnostic.

Implanted in the chest and about the size of a pacemaker, the Chronicle monitors the right ventricle to gauge intra-cardiac pressure in patients with heart failure by means of a lead inserted into the heart. It is one of a new wave of home monitoring systems pushing into the marketplace that use either the Internet or standard telephone line transmissions to deliver data to doctors.

The principal reason this device makes a Wall Street darling of its sponsor and others in the sector is that heart failure the total U.S. tab for heart failure is estimated at $332 billion.

In the company’s Phase III study — the Chronicle Offers Management to Patients with Advanced Signs and Symptoms of Heart Failure (COMPASS-HF) trial — all 274 subjects received the study device, with data from the control subjects withheld from clinicians for six months.

Randall Brockman, MD, a medical officer at the FDA’s Office of Device Evaluation, told the panel that “while there was a trend toward the primary endpoint of a reduction in overall heart-failure hospital equivalents in the Chronicle group, it was not statistically significant.” In the first six months, 72 of the reporting patients had to go to the hospital, compared with 99 in the subjects who did not report their data.

When combined with several other measures of reports of care episodes, the reporting group enjoyed an event rate that was only 21% lower than the controls, a P score of .33, not indicating statistical significance.

Medtronic did well on the primary safety objective, zero device-related complications at six months for 80% of subjects. The Chronicle attained this number easily, ringing in at zero events for 91.5% of the enrollees.

The secondary safety endpoint, freedom from pressure-sensor failure at six months for nine of 10 patients, likewise was met easily as none of the leads failed at the half-year interval.

Panel member Jeffrey Borer, MD, a professor of cardiovascular medicine at Weill Cornell University Hospital (Ithaca, New York) gave the panel a brief review of the application, and described the device as a “very complicated, but very ingenious tour de force.”

Borer said that in the first phase trial, the Chronicle was compared to the Swan-Ganz catheter, used only in a clinical setting, and that the Chronicle “showed relatively good equivalence” for data collection. The Phase II study evaluated durability and safety in 148 patients who were followed for periods as long as 73 months. Some adverse events surface in this phase, but “after some manufacturing flaws were resolved,” the agency deemed these fixes to be appropriate, Borer said.

Borer said that the fact that only the subjects were blinded “could have had some impact on the disposition of cases,” and that the lack of a statistically significant improvement in the primary endpoint may have been due to “a lower-than-expected event rate in the control group,” possibly engendered, in turn, by a high rate of contact between the controls and their doctors compared to more typical heart failure patients.

The enrollees all fell into either Class III or IV in the New York Heart Association (NYHA) scale, and those in the IV class, who are profoundly ill and usually confined to bed, were more likely to show up in the hospital if their data sets were filed with their doctors.

However, NYHA Class IV patients made up only 40 of the enrollees, which severely limited the value of this analysis. The sub-group analysis for the NYHA Class III patients showed that only 35 of the patients whose data was sent in experienced an event (out of 112) compared with 51 of the 122 patients whose data were stored for six months. But the addition of the Class IV patients diluted the effect seen in Class III patients.

Borer noted “a number of unanswered questions” but said the Chronicle “is certainly a great research tool” because “we would learn a tremendous amount,” from its use.

During the discussion, panelist John Teerlink, MD, the director of the heart failure clinic at the San Francisco Veterans Affairs Medical Center (California), said, “as a heart clinician, the benefit [of the device] is so tantalizing, I really want this to work.”

But he noted risks that included potential harmful changes to pharmacological therapy, device-related complications, and “opportunity cost” in the form “of real estate on the chest,” a reference to space in the body potentially useful for a pacemaker or defibrillator implant. He also pointed out that MRI is contraindicated with the device.

Sharon-Lise Normand, PhD, a professor of public healthcare policy and a biostatistician at the Harvard School of Public Health (Boston), said, “the unit of analysis here is really the physician” because the data provided by the Chronicle is likely to influence a doctor’s recommendation. “You’re really testing what the physician will do,” making the current analysis “a nested design.”

Eugene Blackstone, MD, a consultant to the panel and a cardiovascular surgeon at the Cleveland Clinic (Cleveland) said he had “a vague feeling that the fundamental problem is the endpoints that are chosen.”

Jeffrey Brinker, MD, a professor of cardiology at Johns Hopkins School of Medicine (Baltimore), said he did not want “to throw the baby out with the bathwater” and insisted that FDA should come up with a narrowed indication because the data “should be available for select patients.”

Though voting against approval, the panel offered recommendations for approaching approach another submission.

John Somberg, MD, professor of medicine at Rush University Medical College (Lake Bluff, Illinois), suggested providing more data. “Don’t be cheap with the input, because you get killed with the output,” he said, and that he would be “very supportive” of more data from the same trial but with more enrollees.

David Steinhaus, MD, the Medtronic’s chief medical officer, in a statement said that while the company recognized that the Compass trial had not met its primary endpoint, it “believe[s] in the benefit of this technology” and will develop “the appropriate path for approval.”

Shares of Medtronic closed at $50.15 on March 1, down 21 cents on a day when the New York Stock Exchange grappled with the continuing fall-out from the Asian stock market slide and the Feb. 28 computer glitch that backed up trade orders for the Dow Jones index. The Dow had shed about 200 points as the hearing wound out, but recovered all but 36 of those points by day’s end.