No Pony in There: Bapi Fails Mild to Moderate ApoE4 Carriers
Today, Pfizer conceded that results of the first of four Phase 3 trials of intravenous bapineuzumab, a monoclonal antibody against Aβ, failed to meet its co-primary endpoints in some 1,100 patients with mild to moderate Alzheimer’s disease who carry at least one ApoE4 allele. The endpoints were the ADAS-cog and the DAD.
The announcement of topline clinical results took researchers by surprise, mostly with regard to its timing. They had expected two of the four widely anticipated Phase 3 trials—the two run in the U.S. that had started earlier—to be announced together. At the recent Alzheimer’s Association International Conference held 14-19 July 2012 in Vancouver, Canada, word began spreading that topline results for those two trials would come in early August, and clinical and biomarker data would be presented this September at the European Federation of Neurological Societies Conference in Stockholm, Sweden. The latter is still true, but apparently, efficacy in the carrier study was so clearly absent that the Janssen AI and Pfizer Joint Steering Committee for the Alzheimer Immunotherapy Program decided to stop dosing patients in the ongoing open-label extension trial. They informed regulators and site investigators, and made the topline results public earlier.
Beyond timing, the finding was a setback, but not altogether a surprise, to clinician-researchers. After all, the Phase 2 study that formed the basis for the Phase 3 program missed primary endpoints and showed a potential benefit only for ApoE4 non-carriers, not for carriers, in a post-hoc exploratory subgroup analysis (Salloway et al., 2009). “This is a replication of that finding. It is no different from what we saw previously in a much smaller group. We are hoping that we will also replicate the benefit for carriers that we saw in Phase 2,” Steven Salloway told Alzforum. Salloway is at Brown University and Butler Hospital, Providence, Rhode Island, a bapineuzumab trial site. He also chairs the steering committee for the 301 (mild to moderate non-carriers) and the present 302 trials of bapineuzumab run by Janssen’s Alzheimer Immunotherapy Program.
“These results are disappointing. But we need to evaluate the full dataset from this trial and the three other trials still under way. Then we can draw conclusions regarding efficacy, safety and target engagement, and see the implications for AD therapeutic research,” said Paul Aisen of the University of California, San Diego.
Pfizer is conducting a second Phase 3 trial in ApoE4 carriers in Europe. That trial is ongoing, but Janssen AI's media spokeswoman Ellen Rose said that its site investigators would inform patients there and in the U.S. of the negative result of the U.S. trial in the coming days. Janssen AI is the J&J company that works with Pfizer on bapineuzumab. Moreover, the companies will expedite interim analysis of that trial, implying that they may be considering whether to discontinue dosing if there is no hint of efficacy. The decision to stop dosing in the U.S. study's open-label extension implies that there was no efficacy signal at all in the carriers who received treatment in the double-blind phase.
Efficacy was the primary problem, not so much safety, according to Salloway. Apparently, the data safety monitoring board that evaluated the unblinded data found no new safety concerns. “As we know, there is a rate of Aria-E, but they tended to be on the mild side and were usually asymptomatic,” Salloway said.
At this point, the dataset is incomplete. Still, it is probably a real finding that bapineuzumab treatment is not effective for mild to moderate AD in E4 carriers, Salloway said. "Beyond that, we do not know yet what effect it will have in non-carriers. It is too early to make a real evaluation of who might benefit, and at what stage, from bapineuzumab.”
What is clear, however, is that this trial has little bearing on Alzheimer’s immunotherapy in general, Salloway said, “We cannot interpret broadly. We have to deal with one compound at a time.” Other companies that have AD immunotherapies in Phase 3 and 2 include Lilly, Eisai, Novartis, Roche, and others.
To cynics who might suspect corporate communications to have timed release of the bad news for after AAIC to evade the collective glare of the news media, Rose insisted, “That is not possible.” Indeed, stories sprouted online within minutes of Pfizer’s release. Instead, Rose said, what drove the timing was communication of the results and discussion of the extension trial halt with regulators and investigators.
Reisa Sperling of Brigham and Women’s Hospital in Boston, another site trialing bapineuzumab, issued this statement: “We are very disappointed about the lack of clinical efficacy in the ApoE ε4 carriers at the stage of mild to moderate dementia. However, we are hopeful that we might see a more positive clinical result in the ApoE non-carriers. … We eagerly await the biomarker evidence, including PET amyloid imaging and spinal fluid markers, to determine whether there are any signals of disease-modifying activity that would support earlier intervention.”
Two papers reported a lowering of CSF tau (Blennow et al., 2012) and brain amyloid by PIB PET (Rinne et al., 2010) with bapineuzumab. For details, see press release.—Gabrielle Strobel.
University of Southern California Keck School of Medicine
The topline results of the Phase 3, bapineuzumab trial 302 are disappointing but not surprising. That is mainly because the prior Phase 2, 201 trial was a multiple ascending dose trial, not planned as an efficacy trial, and didn’t show reliable efficacy evidence. Only when data from the four doses were pooled, separated into ApoE4 carrier and non-carrier groups, and several post-hoc models and tests were applied did the sponsor report some of the many statistical tests to be nominally significant, albeit ignoring the contributions of multiplicity and play of chance (Salloway et al., 2009). This amounted to, at best, no evidence for efficacy in the ApoE4 carriers, and a wisp of efficacy in the non-carrier group. Notably, that indication of efficacy can be attributed to the 32 placebo-treated non-carriers declining nearly twice as much as expected and substantially more than the carrier group, i.e., an ADAS-cog worsening of nearly 12 points over 18 months and a CDR-box score of over 4 when expected change is about 7 and 2.5, respectively. In effect, the company’s Phase 3 wager, launching four trials with more than 4,000 patients, is informed by only the 79 non-carriers from the 229 patients in the Phase 2 trial.
It will be a surprise if the results of the soon-to-be-reported trial in ApoE4 non-carriers are positive, as there is absent prior evidence as well. Positive outcomes would be welcomed, particularly interesting, and expose a range of hypotheses.
The fact that the sponsors are discontinuing open-label treatment for study participants, not waiting for the imminent results of the non-carriers trial, and expedited an interim analysis of the mainly European carrier trial, suggests that they don’t see a pathway for regulatory approval of IV bapineuzumab for ApoE4 carriers, and are hoping for a limited approval targeted to non-carriers.
Moreover, the idea that the mild AD groups from the 301 and 302 trials would be combined to particularly test if bapineuzumab is effective in relatively milder AD appears to be dashed as well. If there is a potential for efficacy in mild AD overall, then we would have expected open-label treatment to have been continued. If there is no clinical signal with the mild AD ApoE4 carriers, then a statistically significant effect would be unlikely when the mild AD carriers and non-carriers are combined, again leaving the only possible claim for ApoE4 non-carriers.
The four bapineuzumab Phase 3 trials, the solanezumab trials, and earlier Phase 3 trials of GSIs, GSMs, and other substances intended to affect aspects of Aβ processing were each launched with virtually no Phase 2 data supporting clinical efficacy other than in subgroup and post-hoc analyses. At best, pharmacokinetic, pharmacodynamic, and biomarker effects were offered as indicators of future efficacy. Even here, the reported bapineuzumab tau biomarker effects from only a few dozen patients were rather tenuous, dependent on an outlier or pooling samples.
The strategy of going directly to large Phase 3 trials without Phase 2 efficacy evidence marks a very high-risk strategy for failure. Mark Trusheim and colleagues quantified the effects of such stratified medicine decisions through analysis and simulations. For bapineuzumab in particular, interestingly and counterintuitively, he determined that a certain ApoE-targeted, high-risk approach could make economic sense under certain business circumstances (Trusheim et al., 2011).
The potential for positive outcomes from the ongoing bapineuzumab and solanezumab trials remains; but future Phase 3 trials might revert to the confirmatory trials they were originally meant to be rather than the huge exploratory exercises they’ve become. Phase 2 should be able to prove the concept or at least demonstrate the clear potential for efficacy. Of course, that is easier said than done.
Disclosure: LSS has consulted with Johnson and Johnson, Elan, Roche, Pfizer, Wyeth, and Baxter with respect to recent anti-Aβ treatments, and USC has been a clinical site for AD trials sponsored by Johnson and Johnson, Lilly, Pfizer, Genentech, and Baxter.
Trusheim MR, Burgess B, Hu SX, Long T, Averbuch SD, Flynn AA, Lieftucht A, Mazumder A, Milloy J, Shaw PM, Swank D, Wang J, Berndt ER, Goodsaid F, Palmer MC. Quantifying factors for the success of stratified medicine. Nat Rev Drug Discov. 2011 Nov;10(11):817-33. PubMed.
Salloway S, Sperling R, Gilman S, Fox NC, Blennow K, Raskind M, Sabbagh M, Honig LS, Doody R, van Dyck CH, Mulnard R, Barakos J, Gregg KM, Liu E, Lieberburg I, Schenk D, Black R, Grundman M, . A phase 2 multiple ascending dose trial of bapineuzumab in mild to moderate Alzheimer disease. Neurology. 2009 Dec 15;73(24):2061-70. PubMed.
Universitat Autònoma de Barcelona
The finding that immunotherapy against amyloid-β may work in ApoE3 carriers but not in ApoE4 carriers supports that ApoE4 inflicts damage to the brain in a manner independent of amyloid-β. Therefore, combination therapy targeting amyloid-β AND ApoE4 appears to be in order.
In addition, more research is necessary to understand the molecular basis and the consequences of the gain of toxic function of ApoE4. I personally favor the notion that ApoE4 causes dysfunction of astrocytes—incidentally, the major producers of ApoE in the brain. This would result in alterations of neurovascular coupling, redox homeostasis, synaptic plasticity, and energy metabolism, all functions tightly controlled by astrocytes.