Given how many patients who completed the double-blind portion of the study opted into the extension segment and the quality of pharmacokinetic (PK) data, phase III testing of Intec Pharma Ltd.'s Accordion Pill (AP, carbidopa/levodopa [CD/LD]) against immediate-release CD/LD in Parkinson's disease (PD) seemed likely to ring the statistical significance bell.
But it didn't happen. Shares of Jerusalem-based Intec (NASDAQ:NTEC) plunged 81.6%, or $2.44, to close Monday at 55 cents on the outcome of the study called Accordance. Principal investigator Peter LeWitt from Detroit's Henry Ford Hospital said during a conference call with investors that the gastric-retentive combo pill, formulated with technology that uses biodegradable polymeric films, "still seems the best way" to get the right amounts of PD drugs into patients at the correct rate. Even continuous delivery with a tube planted in the stomach wall doesn't work as well as AP-CD/LD, in his view.
Specifically, the Intec oral approach combines and loads active ingredients onto films, folds them into an undulated shape, and places them inside a capsule. Intec is developing AP-cannabinoids to deliver either or both of the primary cannabinoids contained in Cannabis sativa, cannabidiol and tetrahydrocannabinol, for various pain indications. Partners seem sold on the method; Intec has a feasibility agreement for the development of a custom-designed AP with Basel, Switzerland-based Novartis AG, as well as a research collaboration with Merck & Co. Inc., of Kenilworth, N.J.
But Intec's PD effort took center stage Monday, and Accordance did not achieve statistical superiority to Merck's oral CD/LD Sinemet on the primary endpoint of reduction in daily "off" time. Data became available over the weekend. Right away, the firm conducted post-hoc analyses in an effort to figure out what might have caused the blow-up. The most credible idea so far is that too many subjects in the study were underdosed, because Intec agreed with the FDA that – until the (now-finished) gastroscopy safety study was completed – the experiment would keep dosing within low parameters.
"No matter how you look at it, we did not give as much as we could have or should have, based upon what's in the literature," CEO Jeffrey Meckler said. "There's a reason this was the first analysis we did once we saw the results," he added, noting that the company was "concerned throughout the trial" about the ceiling on dosage. "I could not have told you ahead of time that this would be a problem but it was something that was on the radar," and it appears to have become an issue.
Accordance was a multicenter, global, randomized, double-blind, double-dummy, active-controlled, parallel-group study in adult subjects with advanced PD at more than 90 sites throughout the U.S., Europe and Israel. Before the 13-week randomized and double-blinded portion, the trial had two open-label periods of six weeks each during which all patients in those open-label periods were first stabilized and then optimized on Sinemet and then on AP-CD/LD. The study enrolled 462 patients in the Sinemet titration period to provide for the 320 patients that were randomized into the 13-week portion. All patients completing that stretch were eligible to continue in an open-label extension (OLE) study: treatment with AP-CD/LD for 12 more months. More than 90% of eligible patients chose to enter the OLE study. The "off" time primary efficacy endpoint was measured by Hauser home diaries, and the study was 90% powered to detect a one-hour difference between Sinemet and AP-CD/LD.
LD is typically given for PD to boost brain dopamine, adding CD to mitigate side effects. But advanced patients on LD and Sinemet can have motor-function problems because of their "off" periods between doses. Accordance's secondary endpoints included change from baseline to endpoint in "on" time without troublesome dyskinesia during waking hours, clinical global impression of improvement at endpoint (as recorded by physician and patient), and change from baseline through endpoint in the Unified Parkinson's Disease Rating Scale score parts 2 and 3. AP-CD/LD also did not achieve statistical superiority on those endpoints. Treatment-emergent adverse effects proved consistent with the known safety profile of CD/LD formulations, with no new safety issues turning up in Accordance, during the gastroscopy safety substudy, or in the OLE study, Intec said.
Execution 'suboptimal,' CEO says
Oppenheimer analyst Jay Olson last month maintained his "outperform" rating on Intec with a price target of $17. His firm consulted "two of the world's leading authorities on PD," Warren Olanow, from Mount Sinai School of Medicine, and Karl Kieburtz, from the University of Rochester. They were encouraged by, among other things, PK data unveiled at the World Congress on Parkinson's Disease and Related Disorders in Montreal. AP-CD/LD 50/500 mg three times per day met the study's primary endpoint of reducing plasma LD variability vs. standard immediate-release (IR) CD/LD when dosed five times per day (p=0.0048). Less variability also was measured for the co-efficient of variation of plasma LD levels (a key secondary endpoint; p=0.047). Significant outcomes were achieved on each of the prespecified sensitivity analyses, Intec said. Olanow said the "smoother" profile improves efficacy and safety. He regarded the phase III study as "relatively de-risked."
After the Accordance results rolled out, Olson downgraded his rating to "perform," saying in a report that "while these are top-line results and we await details, we believe this failure is a major setback for Intec" and that he was "uncertain about the read-across from AP-CD/LD to other AP applications in diseases beyond PD." Intec pointed to an analysis showing that a subset of patients who didn't require the maximum dose of 1,500 mg LD did achieve a numerical separation in "off" time on AP vs. IR, though the difference was not statistically significant, a finding that supports Intec's underdosing theory. "We asked about the secondary endpoint of change from baseline in 'on'-time without troublesome dyskinesia," Olson said. "Management indicates that the post-hoc analysis showed numerical separation between AP and IR but it was not statistically significant, again, because the number of patients was too small since only 40% of patients were at a dose that qualified."
CEO Meckler said that "certainly early on, some of the communications and I would say execution of the study were suboptimal." The protocol did not specify, for example, when patients should fill out their diaries. "There was an expectation that the diaries should be completed near the timing of the clinic visits, yet some patients didn't do it, and it seemed to be site by site. Some sites knew how to do these studies and, unfortunately, some apparently didn't." Intec will have updates in the coming weeks and months as the results are sifted further, he said. "We want to analyze these data and figure out what is the right trial to do next in this program."