Martin Mackay, global head of R&D for Alexion Pharmaceuticals Inc., told BioWorld Today the firm is "very much in preparation mode" for talks with regulators about complement protein C5 binder Soliris (eculizumab) for refractory generalized myasthenia gravis (MG).

After its narrow failure in the phase III trial called Regain – an almost-hit that hardly quelled optimism for marketing clearance – Alexion offered new data from the same study that brighten the picture even more. "I'm always at pains to say that what we presented today were all pre-specified analyses," Mackay said, speaking from the 14th International Congress on Neuromuscular Diseases in Toronto. "It's not like a post-hoc look at the data."

The complement protein C5 binder "looks approvable," said Piper Jaffray analyst Joshua Schimmer. "Our view of Soliris for MG is that there appears to be a minority of patients who have an exceptional response to Soliris, and this dynamic may have led to a missed primary endpoint despite a very active therapy for some."

Regain's primary efficacy endpoint of change from baseline in the MG-Activities of Daily Living (MG-ADL) Profile total score, a patient-reported assessment, at week 26, did not reach statistical significance (p=0.0698) as measured by a worst-rank analysis. Those data were reported last month. New, more-encouraging results saw daylight at the Hot Topics session of the Toronto meeting and showed that 18 of 22 pre-defined endpoints and pre-specified analyses in the study, based on the primary and five secondary endpoints, achieved "p" values of <0.05. Soliris is already sold by New Haven, Conn.-based Alexion for paroxysmal nocturnal hemoglobinuria and atypical hemolytic uremic syndrome. (See BioWorld Today, June 8, 2016.)

In a research report, Schimmer opined that Soliris for MG is "not only approvable, but a drug that will garner meaningful uptake among patients, most of whom are identified and waiting for a new therapeutic option. Even if 50 percent of suitable MG patients achieve only modest response to Soliris and discontinue therapy, our $750 million sales estimate for this indication in 2021 looks conservative to us, and we are nearly twice consensus [estimates] for this indication in the outer years."

Schimmer has been a believer since the early days. "MG appears to be among a small group of autoimmune diseases caused by an antibody and is characterized by potentially profound muscle weakness," he wrote in an April report. "A subset of MG patients have IgG1 or IgG3 antibodies against the acetylcholine receptor, with disease pathology which appears to be driven primarily by complement activation," a fact on which he based his prediction at the time that "Soliris will have a transformational role for a subset of MG patients with severe/refractory disease."

That's what Alexion is shooting for. "We want to make a huge difference to their lives in a tiny number of patients," Mackay said.

It could happen. Suntrust Robinson Humphrey analyst Yatin Suneja said "there was some discussion [at the meeting] on why MG-ADL was chosen as a primary endpoint" and the presenter reminded attendees that "regulators have been pushing for more patient-reported outcome measures over the past five years. However, given the 'overwhelmingly' positive 'totality' of the Regain data, we still believe an approval for Soliris in refractory generalized MG is possible. We also believe that two (bowel perforation, prostate adenocarcinoma) of the four patients on Soliris who discontinued despite showing seven-point improvements in MG-ADL would have been noted as having a benefit in more conventional analyses (vs. worst-rank analysis [deployed by Regain], which has never been used before in clinical trials for MG)."

Mackay said "one discontinued because he was getting worse. The other three actually were getting better and had marked improvements in their MG-ADL scores."

COMPETITION'S GOOD

The details of the findings from Regain matter. Its prospectively defined secondary efficacy endpoint of change from baseline in Quantitative Myasthenia Gravis (QMG) total score, a physician-administered assessment of MG clinical severity, with Soliris compared to placebo at week 26 achieved a "p" value of 0.0129 as measured by the worst-rank method. The second and third prospectively defined secondary efficacy endpoints of responder status in MG-ADL and QMG achieved "p" values of <0.05. The proportion of patients with at least a three-point reduction in MG-ADL total score and no rescue therapy from baseline to week 26 with Soliris vs. placebo achieved a "p" value of 0.0229, and the proportion of patients with at least a five-point reduction in QMG total score and no rescue therapy from baseline to week 26 with the drug vs. placebo achieved a "p" value of 0.0018.

Mackay zeroed-in on sensitivity analyses. "The reason you do [them] isn't just to do fun things with statistics," he said. "It really tests the validity of the endpoints, particularly in areas where there is no defined pathway like refractory MG." In QMG, all four sensitivity analyses came up strong, "but MG-ADL wasn't a complete washout, because three of the four were positive," he said. Entry criteria called for patients having failed at least two prior therapies, but "the reality was that over 50 percent had failed over three or more," Mackay said. "One of our investigators described it well. He said he had exhausted all treatments for these patients, and that's very much the population that we focused on."

Even before the new results were detailed, Suntrust's Suneja reported on a conference call with a key opinion leader (KOL) in MG who gave Soliris a greater than 50 percent chance of winning the FDA nod. His report June 20 said "the majority of MG patients have a good response with first- and second-line therapies such as steroids and immunosuppressants," but available therapies "are inadequate for a small subset (5-10 percent; 1,800-3,000) of MG patients." For those, "our KOL believes that Soliris would be appropriate. He also believes that physicians would likely use Soliris in advance of chronic plasma exchange, while noting that 'intravenous immunoglobulin [therapy] doesn't work that well' in the refractory setting."

Alexion's Mackay said such a move would be the physician's choice. "We want to give treatment options," he added.

Competition for Soliris could come from the CD20 antibody Rituxan (rituximab, Roche AG), undergoing a phase II study sponsored by Yale University in generalized MG. Also in play is Argen-x BV, of Breda, the Netherlands, with antibody fragment ARGX-113. The company reported healthy volunteer phase I data at the end of last month and is expected to unveil plans for the phase II program in MG in September.

"I never make light of competition, but if you think about the [ultra-rare] areas we work in, the fact that there's competition or potential competition is really a good thing," Mackay said.