BB&T and BB&T Staff Reports

Implantable cardiac defibrillators (ICDs) have had their place in the cardiac space for years, but experts think there may be an additional benefit of the devices that would reduce the number of people who die from a second heart attack. A study co-sponsored by Medtronic (Minneapolis) and GE Healthcare (Chalfont, UK), is the first randomized study testing whether two particular risk markers, when used to prescribe ICDs, prevent sudden cardiac death (SCD).

Jeff Connors, research manager of Patient Care Solutions at GE Healthcare, told BB&T that this study is designed to determine whether people who currently do not qualify to receive an ICD would benefit from one.

“What makes this different from other studies is that currently to qualify for an ICD, we look at a thing called ejection fraction, or how well the heart is pumping,“ Connors said. Only heart attack patients whose ejection fraction is below 35% currently qualify for the device, he explained. But a lot of patients who have higher ejection fraction still end up dying from a second heart attack later on.

Advances in heart attack treatments over the last few years have led to more people surviving their initial heart attack, Connors said, but the rate of death for secondary heart attacks is the same. The idea of the study is to find out who would benefit from an implantable defibrillator among those who are currently not eligible for ICD therapy, he said.

Rather than looking at ejection fraction, the study's risk markers correlate to electric properties of the heart (t-wave alternans), and the nervous system's impact on the heart (heart rate turbulence). These markers showed promise in an article appearing in the Journal of the American College of Cardiology in 2007. Based on its results, this new REFINE-ICD study is being undertaken.

Medtronic's ICD, along with GE's latest Holter system will be used in the study. The new Holter system incorporates new signal processing to enable t-wave alternans and heart rate turbulence movements via a single, non-invasive test, GE said. Earlier versions of the Holter system did not have this capability, so invasive and/or expensive tests were required. The study targets multiple global sites and more than 1,000 patients.

“If you combine these two measurements (t-wave alternans and heart rate turbulence) you can double the diagnostic ability to find out who these people are more accurately,“ Connors said. He said the study will only enroll patients who test positive for both risk factors.

Derek Exner, MD, of the Libin Cardiovascular Institute of Alberta at the University of Calgary, leads the study, which is based on his past research (the REFINE study).

The Alberta government, Western Economic Diversification Canada through the Western Economic Partnership Agreement, Medtronic, GE Healthcare, the University of Calgary, and other partners are providing $16.8 million to support the pilot phase of the project. The study may eventually grow to include 75 sites in September 2012 worldwide with a total investment of more than $40 million.

“Medtronic is committed to principled collaboration among members of academia, government and industry to better address the unmet needs of heart patients,“ said Marshall Stanton, MD, VP of clinical research for the Cardiac and Vascular Group at Medtronic. “This kind of clinical research is critical to the development of innovative technology and treatment approaches that can help transform the lives of the growing heart disease population worldwide.“

Sudden cardiac arrest is a sudden, abrupt loss of heart function – it is not a heart attack – which is caused by a blocked vessel leading to loss of blood supply to a portion of the heart muscle. Instead, most sudden cardiac arrest episodes are caused by the rapid and/or chaotic activity of the heart known as ventricular tachycardia or ventricular fibrillation.

Exner and his colleagues have developed a method to identify patients at high risk of serious heart rhythm problems after a heart attack. The institute will be the clinical coordinating center for the study and will gather patient data from 16 sites in Canada, the U.S. and Europe during an initial, two-year pilot phase.

“Sudden death is a major cause of death in Canada and around the world. Research to identify those at risk and methods to prevent death from heart rhythm problems are vital,“ Exner said. “This study may change how we manage patients after a heart attack and has the potential to save thousands of lives each and every year. It is an honor to be leading such an important study.“

Connors said there are roughly 8 million heart attacks a year in the U.S. and about 350,000 to 500,000 people a year die from sudden cardiac arrest after an initial heart attack, but only a very small portion of those patients have an ejection fraction that qualifies them for an ICD. “So this is really going to expand [ICD therapy] into a much wider population and benefit a lot of people that currently are not eligible for the therapy,“ Exner said.

Edwards in second pricey study for Sapien

Anyone responsible for delivering news to those who invest in shares of Edwards Lifesciences (Irvine, California) might have resorted to the cliché, “I have some good news and I have some bad news.“ The good news is that an FDA advisory committee voted 9-0 (with one abstention) in July that the benefits outweigh the risks associated with the firm's Sapien aortic valve, an outcome that was anticipated by the markets after FDA posted documents associated with the application in July.

The bad news? FDA wants the firm to enroll a new five-year trial of as many as 1,000 subjects across 75 centers as a condition of approval. The panel agreed with the agency's approach, and investors clearly heard the bad news, sending shares of the firm's stocks on the New York Stock Exchange (NYSE; New York) down more than $2 each on a volume of more than 1.2 million shares as the hearing came to a close.

Some may wonder whether the Sapien is worth the bother. The Sapien XT, a smaller version that seems sure to create fewer hazards for patients because of it's substantially smaller size, is scheduled to complete follow-up in six years, a convergence of time lines that surely is not lost on the firm's managers and shareholders.

On the other hand, the U.S. market for catheter-delivered heart valves is expected to hit $1 billion by 2015, leaving Edwards with a choice of taking another hit to sell in the U.S. or leaving money on the table. Making matters worse for industry's expectations of the American market is that the Sapien has been available in Europe since 2007, a device lag of four years and counting, while the XT earned a CE mark last year.

FDA and the firm agreed in broad terms to a post approval study (PAS) built around a continued enrollment protocol attached to the PARTNER trial and conducted as a registry, but the second PAS would be a hypothesis-driven trial with composite safety and efficacy endpoints. The size of this enrollment stands in sharp contrast to the less than 400 patients in the cohort of interest at the advisory committee.

FDA staff mentioned the increased incidence of stroke seen in the PARTNER patients on the study article vs. those in controls, but they indicated that their thinking about a second PAS was driven at least as much as by concerns about device durability, quality of life (QoL) measures, and differences between the patients in the control and device arms.

One of the major concerns for FDA about the trial was the incidence of stroke. FDA's Julie Swain, MD, a cardiovascular surgeon, told the panel, “when we look at stroke, the less-than 30 days level is about 4.3 times higher“ on the device “and about 2.5 times higher at one year,“ adding that anti-coagulation therapy might help ward off such events. However, Swain also noted that total neurological events were higher on the device, with 16 such events compared to seven among controls. Another difference here was the incidence of transient ischemic attacks, which affected three on the device and none among controls.

Jodi Akin, VP for global clinical affairs at Edwards, noted that the firm “has already agreed to continue the PARTNER study for another five years“ for continued access patients, but FDA's request would require an infrastructure similar to PARTNER, which she said “would be duplicative of PARTNER and PARTNER II,“ the latter of which is the U.S. trial just underway for the XT model. According to, PARTNER II commenced enrollment in February with a target enrollment of 600.

One of the sticking points for the advisory committee and for FDA was the fact that the characteristics of the study and control patients were somewhat dissimilar on several points. Martin Leon, MD, of Columbia University (New York), and one of the co-primary investigators in the trial, mentioned that mean age in both arms was roughly 83.

While each arm initially consisted of 179 patients, only 33 of the enrollees in the device arm had experienced previous infarcts compared to 47 in the control arm. Previous bypass also favored the device (58 compared to 73 among controls), as did prior balloon dilation of the affected valve (25 to 39). On the other hand, 47 of the device patients had undergone a percutaneous intervention in the coronary arteries compared to only 39 among the controls.

Regarding the post-hoc analysis in which Edwards used “major stroke“ as part of the composite endpoint rather than stroke generally, Leon said “during the trial, we realized we needed greater sensitivity“ to determine the source of strokes, so the investigators and Edwards crafted the post-hoc measure “based on a modified Rankin score of two or greater“ for the intent-to-treat population, Leon said.

FDA's presentation was based on a 30-day rate of stroke/transient ischemic attack of 7.3% for the device arm vs. 1.7% among controls, but Leon's presentation indicates 5.6% for major stroke for the device at 30 days and 1.1% for controls. At one year, those numbers become 8.4% and 3.9%, respectively.

Should Edwards agree to the second post-approval study, the company may face regulatory double jeopardy. Responding to a question from one of the panelists about the consequences if the hypothesis-driven post-approval study came up short, FDA's Bram Zuckerman, MD, chief of the division of cardiovascular devices at the Office of Device Evaluation, said, “FDA can bring the results back to this advisory panel or make an internal decision.“ Zuckerman said this “will be one component of an eventual strategic plan“ from which “everyone will benefit.“

Analysts had a lot to say about how the outcome might affect the company's prospects. A July 21 statement by Canaccord Genuity (Boston), affirmed an earlier “hold“ recommendation with a target price of $90 per share. Analyst Jason Mills said the risk profile associated with an estimated aggressive U.S. Sapien ramp “relative to the current valuation is too high.“

A July 20 review by Wells Fargo (New York) analyst Larry Biegelsen said one question for investors is whether Edwards will scale back an estimate of sales of $150-250 million the first year. The document pegs the valuation range of Edwards' transcatheter valve business at $80 per share, citing risks of delays in time to market, possible delays due to slow uptake, and “new competition.“ Still, Biegelsen believes that transcatheter aortic valves represent “the most compelling opportunity in the cardio space.“

Healthcare VC investing shows some signs of revival

Venture Capital firms have not been shy about spilling their concerns in recent years about VC funding in the medical device industry. Global economic recession, the virtual absence of a medical device IPO market, steep decline in VC returns, healthcare reform, uncertainties and approval delays at the FDA, reimbursement challenges, syndicate investor risks, and even concerns about physician and industry alliances have all impacted the number of VC dollars backing device companies as of late.

But all hope is not lost.

Medical devices continued to lead venture deals in healthcare with the overall sector seeing deals up 7% and funding climbing by 16% on a year-over-year basis, according to the latest quarterly venture capital report from CB Insights (New York). In fact, healthcare VC investments hit a five-quarter high with nearly $1.9 billion of venture investing, even though healthcare deal flow continues to drop, a trend that has been seen since Q3 of last year, according to the CB Insights report. There were a total of 147 healthcare VC deals during the second quarter, down just a smidgen from 150 last quarter, 173 in Q4 2010 and 184 in 3Q10.

CB Insights emphasizes that its report includes only VC funding, fundings of only private companies (public companies are excluded from the numbers even if they received VC funding), U.S.-based companies, and only deals that actually closed during the second quarter.

Geographically, it's no surprise that California and Massachusetts dominated healthcare venture investment in the second quarter, seeing 56% of all deals and 75% of venture funding in healthcare. Massachusetts in particular had a good quarter as the state took 31% of healthcare VC dollars compared to 19% last quarter, 17% the previous quarter, and only 13% during 3Q10.

Looking at the numbers by state a different way, the report shows that 34% of the 89 VC deals in Massachusetts during the second quarter were made in the healthcare sector. In California, only 15% of the 336 VC deals during 2Q11 were in healthcare.

Cameron Health (San Clemente, California), a defibrillator maker, raked in the highest funding amount in the sector during the quarter when it completed a $107 million equity financing round led by Alloy Ventures (Palo Alto, California) and Delphi Ventures (Menlo Park, California). Cameron recently completed enrollment in a 330-patient pivotal FDA study involving 33 centers around the world and expects to file for FDA approval early next year. Its S-ICD System has been implanted in roughly 800 patients worldwide. According to Cameron, the S-ICD system is the first totally subcutaneous implantable defibrillator designed to provide life-sustaining therapy to patients at risk of sudden cardiac arrest. Because it is subcutaneous, it eliminates the need for electrical wires (leads) to be placed in the heart. The quarter's other largest healthcare VC deals involved drug companies like Tesaro (Waltham, Massachusetts; $101 million) and Radius (Cambridge, Massachusetts; $91 million).

Early stage healthcare investments bounced back from a five-quarter low in 1Q11. However, that trend appears to be driven primarily by a handful of larger late-stage VC deals.

By sector, 43% of the quarter's healthcare VC deals by were in medical devices compared to 19% for drug development, 6% for pharmaceuticals, and 13% for biotech. By dollar amount, 44% of VC investment in healthcare went to the device sector followed by 23% for drug development, 16% for pharma, and only 5% for biotech.

Another interesting trend highlighted in the CB Investments report for the second quarter is seen in the size of VC healthcare deals. According to the report, the quarter saw numerous healthcare mega-deals which pulled the aggregate investment numbers higher, but these were offset by smaller deals across the board. Ultimately, this drove down median deals for all stages of healthcare investment in 2Q11.

Overall, the VC industry fared well during the second quarter. The report notes that deals climbed 25% and funding climbed 29% as the quarter registered $7.6 billion of VC funding invested in 768 deals representing a nine-quarter high for both. Even better, this performance follows first quarter which saw deals and funding at pre-recession levels. “We know that all you bubble prognosticators are probably licking your lips reading this,“ the authors of the CB Investments report wrote. “But before you make these predictions, take a look at the quarter in a bit more detail as the quarter's stats are not good news for everyone.“

FDA's 'when-to-file' 510(k) guidance limits predicates

FDA's Center for Devices and Radiological Health is steadily moving forward on its “to do“ list, and released a draft guidance in July on when to file a new 510(k) for changes to an existing device. Much of the guidance seems to replicate its predecessor from 1997, but at least one passage hints at a substantive change.

This new guidance includes the provision that when determining whether a change to a device requires a new 510(k) filing, “the modified device should not be compared to multiple devices, only to the most recently cleared version of that device“ that is already listed in the filing for the altered device. The agency reinforces the point with the passage that a modified device “should not be compared to any other device produced by the same manufacturer or another manufacturer, even if the other device could serve as a predicate to the modified device.“

This language represents a substantial change from the 1997 guidance, which states that a device or diagnostic maker “should compare the change or changes to their device as previously found to be substantially equivalent“ with other devices. The language of the 1996 guidance seems prohibitive of nothing on this score.

As for the question of the use of multiple predicates to determine whether a new 510(k) is called for, the 1997 guidance seems to be silent, making the proposal to ban the use of multiple predicates in this context another regulatory innovation.

The 1997 guidance addresses the elimination of an indication from the label in brief terms, stating that if a device “was cleared for use with three specific indications and the firm decides to market the device for only two of those indications, [this] would not require submission of a new 510(k).“ However, the new guidance states that if a firm decides to remove one indication due to a market withdrawal or correction, this would trigger a filing. However, a lack of demand for the device for a specific indication would not trigger a filing requirement should the manufacturer decide to delete the indication from the label.

The 1997 guidance uses somewhat different language from the latest iteration to deal with indications as related to patient populations. The prior guidance states, “if the expansion [of indication] is to a population with similar demographics, diagnosis, prognosis, co-morbidity and potential for complications as the original, then a new 510(k) is not ordinarily expected.“ The new guidance states, however, that because changes to the indications for use are generally “major“ changes to the intended use,“ those changes “generally will require submission of a new 510(k).“ The new guidance states further that a change from general populations to specific populations “usually“ will necessitate a new filing.

With regard to a change of design undertaken to address an indication for expanded populations, the 2011 guidance states that design changes “that allow use in a new, expanded, or more specific patient population also carry a high potential to significantly affect safety or effectiveness, and therefore, these changes should result in a new 510(k) submission.“

FDA tackles the topic of the impact of multiple changes to a device in both guidances, but takes a somewhat different approach in its discussion in the later version. The 2011 guidance recommends that makers of devices and diagnostics “have a mechanism or standard operating procedures in place for evaluating whether a proposed change meets the regulatory threshold for a new 510(k).“ The new guidance includes discussions of how manufacturing and sterilization might create a need for a new 510(k) as well.
The new guidance states further that when more than one change has been made to device design and only one of those changes triggers a filing requirement, a manufacturer should include all those changes in a filing “as well as a comparison of the changed device to the device as it was described in the most recently cleared 510(k).“ This applies “even [to] those that did not trigger the need for a new 510(k),“ although the sponsor should make the distinction between the two.

Furthermore, manufacturers “should scientifically justify their conclusions that modifications, individually and collectively, could not affect safety or effectiveness“ should the manufacturer determine no new filing is needed, the draft guidance states. As always, “a copy of this documentation should be maintained“ in company records.

In the 1997 guidance, FDA allowed firms to “document the application of the model along with the necessary records of the validation of all changes to the device“ when the sponsor decides not to file, but this passage from the 1997 guidance also explicitly permits a manufacturer to “compare their device to a competitor's legally marketed device“ in making such a determination.

One feature of the new draft guidance of interest to makers of diagnostics is a passage indicating a concern regarding interpretation of data generated by a diagnostic. The 2011 draft guidance states that labeling for a device “that provides diagnostic information [that] is modified to include additional or new instructions on how to interpret data from the device“ constitutes “a major change in intended use that could significantly affect the treatment of the patient and . . . requires submission of a 510(k).“ The 1997 edition does not appear to address this question.