Susan Jeffrey

July 29, 2011

July 29, 2011 (Paris, France) — Two new reports provide additional information on the safety of bapineuzumab, a humanized monoclonal antibody being developed for the treatment of Alzheimer's disease (AD).

Presented at the Alzheimer's Association International Conference 2011 here, the reports provide a more in-depth look at vasogenic edema (VE), now referred to as amyloid-related imaging abnormalities (ARIA), seen during the previously reported phase 2 studies of bapineuzumab in light of a new review of these magnetic resonance imaging (MRI) abnormalities by the Alzheimer's Association Research Roundtable published in the July 2011 issue of Alzheimer's & Dementia.

Dr. Reisa Sperling

Reisa Sperling, MD, associate professor of neurology at Harvard Medical School, Cambridge, Massachusetts, and director of the Alzheimer's Disease Neuroimaging Program at the Massachusetts Alzheimer's Disease Research Center at Massachusetts General Hospital, Boston, reported that reexamination of MRI scans taken during phase 2 studies including 210 patients treated with bapineuzumab turned up 36 cases of ARIA, for an incidence of about 17%.

Interestingly, 15 of these cases were asymptomatic and not detected during the trial, so the patients received study drug through the presence of ARIA without becoming symptomatic, a development Dr. Sperling sees as reassuring. With powerful new biologic agents come the potential for more severe adverse effects, she told Medscape Medical News.

"The question is can we control them, and I feel more comfortable now that we're going to be able to get a handle on these that hopefully the majority of people don't get symptoms associated, and maybe we'll one day think of this as evidence of proof of mechanism, which would be great."

A separate report provided longer-term follow-up to 78 weeks from one of those phase 2 studies, study 251.

Dr. Stephen Salloway

Stephen Salloway, MD, Department of Neurology and the Memory and Aging Program at Butler Hospital and professor of clinical neurosciences and psychiatry at Brown Medical School in Providence, Rhode Island, reported safety data on bapineuzumab out to 78 weeks.

"Overall, bapineuzumab was well tolerated," Dr. Salloway told a press conference here. Adverse effects were classified as mild or moderate, and no new safety signals were reported after 5 years of exposure to treatment, he said.

"This is extremely important because we want to know how the drug performs over an extended period, and this gives us additional experience and encouragement about moving forward," Dr. Salloway said.

The phase 2 program is supported by Pfizer Inc and Janssen Alzheimer Immunotherapy.

Moving Amyloid

Cerebral microhemorrhages and VE have been reported in a number of trials of amyloid-lowering therapies and are speculated to be related to the movement of amyloid out of plaques, a kind of good news/bad news scenario where the drug may actually be doing what it's designed to do but causing adverse effects of edema or microhemorrhage.

The reports prompted the US Food and Drug Administration (FDA) to recently advise companies running these trials to exclude patients who had 2 or more microhemorrhages at study entry and to discontinue therapy for any patient developing incident microhemorrhages during the study.

Concerned that these new rules would significantly limit enrollment in trials, the Alzheimer's Association Research Roundtable, with member scientists drawn from the pharmaceutical, biotechnology, diagnostics imaging, and cognitive testing industries, along with senior staff and advisers from the association, struck a working group to look further into the occurrence of VE and microhemorrhages and make recommendations on inclusion and exclusion criteria and suggested monitoring for patients in ongoing trials of amyloid-lowering therapies.

The report of the working group, now published in Alzheimer's & Dementia, recommended changes in the terminology, now favoring ARIA as the general term to indicate the relationship with the presumed mechanism — removal of amyloid — since VE and microhemorrhages occur also in the natural course of AD.

ARIA-E refers to abnormalities attributed to edema and ARIA-H relates to those attributed to hemorrhage.

The working group report was sent to FDA for review, and in an editorial published with it, Maria Carrillo from the Alzheimer's Association and Rachel Schindler from Pfizer Inc report that the FDA has since revised their advice to reflect their recommendations.

Phase 2 Studies Revisited

In the new report presented here, Dr. Salloway and colleagues presented extension data from study 251, a previously reported phase 2, open-label study of 4 doses of bapineuzumab. Study drug was given intravenously, with 13 weeks between infusions.

A total of 194 patients were enrolled; 94 (48.5%) received treatment for at least 3 years, and 46 (23.7%) received it for at least 4 years. The average exposure in the group was 2.6 years.

Overall, the treatment was well tolerated, he said. "Basically the adverse events and the serious adverse events in this open-label phase were similar to what we saw in the double-blind phase," Dr. Salloway said. Adverse events related to bapineuzumab were generally mild to moderate, and there were no new safety signals.

ARIA-E were the most common drug-related adverse event, occurring in 9.3%. "This is very similar to what we saw in the phase 2 and actually the rate of ARIA-E goes down with time," he noted. The occurrence was 6.7% in the first 3 infusions, decreasing to 2.7% during infusions 4 through 10, Dr. Salloway noted.

"In terms of clinical outcomes, the numerical treatment differences that we saw and reported in Neurology in the phase 2 trial persisted into the open-label phase," differences that were not statistically significant," he said. He emphasized that no efficacy conclusions can be drawn from these findings.

"However, 4 large, randomized, phase 3, double-blind trials are under way, making good progress to evaluate safety and efficacy, so we hope later next year to have more information about bapineuzumab, including the effect on biomarkers, such as imaging and [cerebrospinal fluid]," Dr. Salloway concluded.

ARIA Re-Read Project

The study reported by Dr. Sperling involved a systematic review of all phase 3 studies, including study 251 that Dr. Salloway reported extension data on, as well as studies 201 and 202, to estimate the incidence and risk factors for ARIA-E.

More than 2500 fluid attenuated inversion recovery images from 262 subjects were blindly reviewed by 2 independent neuroradiologists for the presence of ARIA-E and risk analyses performed to characterize risk factors.

A total of 36 incident cases of ARIA-E were identified among those with at least 2 posttreatment MRIs. "That's 17%, which might be a little higher than you've heard in some other estimates, but that's partly because this includes the higher doses that were originally used in the 201 study," Dr. Sperling said.

Among the 36 patients were 21 who had been identified during the actual trials, but 15 patients that had remained asymptomatic despite ongoing treatment. Eight patients became symptomatic, however, including headache, confusion, visual disturbance, and some walking abnormalities, she noted.

Microhemorrhages, or ARIA-H, occurred in 17 of the 36 ARIA-E patients (47%) vs 7 of the 177 patients without ARIA-E (4%).

The risk analysis showed an increase in risk for ARIA-E in APOEε4 carriers; those homozygous for the risk allele had about 5 times the risk of developing ARIA-E, she noted. They also confirmed the previously reported relationship with increasing dose, but no other factors were significantly associated.

Dr. Sperling presented a separate paper here reporting 3 cases of apparent ARIA-E in a phase 2 trial of the gamma secretase inhibitor BMS-708163 (Bristol-Myers Squibb). "This was actually somewhat of a surprise because we previously thought that this might be something that might be specific to immunotherapy, and I can say these 3 cases look very, very similar to what's been seen in multiple immunotherapy programs," she told Medscape Medical News.

Fortunately, from cerebrospinal fluid studies in these patients, it doesn't seem the process is inflammatory like progressive multifocal leukoencephalopathy with natalizumab in multiple sclerosis, she noted. "It doesn't look like it's a meningocephalitis or something else that had been previously seen with the vaccines." However, like chemotherapy and multiple sclerosis agents now being developed, adverse effects may be more severe with more powerful agents.

"I hope that that won't be the case in Alzheimer's disease, but I think we will have to be prepared for biologically active side effects," she said.

Although the precise mechanism of ARIA is still unknown, some early data from animal models suggest they may relate to amyloid clearance. She showed 1 slide of positron emission tomographic scans from a patient in whom ARIA-E developed, showing that the areas of high amyloid burden before treatment were the same areas affected by ARIA-E and the same areas that showed focal clearance of amyloid 3 months later.

"So we think that what it says right now is that it probably is related to how much amyloid you have and how much amyloid you move," she said. The increased risk in APOEε4 carriers may relate, for example, to their known increased amyloid levels in blood vessels, "so we may need to dose them slightly differently to be able to get efficacy but avoid the risk of having significant imaging abnormalities."

The current guidance is that patients in all therapeutic trials of treatments that affect amyloid should be monitored every 3 months for evidence of ARIA-E and ARIA-H that frequently occur together, she noted.

Tolerance for Adverse Effects Depends on Efficacy

Mary Sano, MD, PhD, professor of psychiatry at Mount Sinai School of Medicine in New York City, moderated the press conference here. Asked to comment on these reported results, she pointed out that the imaging abnormalities are apparent in many clinical trials investigating agents that clear amyloid, "so I think it's important that we take time to understand what they mean."

Dr. Mary Sano

In this discussion, she said, "we've been offered a theory of how that abnormality may be demonstrating that an agent is actually working, so I think that's very important."

"Of course, the willingness to tolerate any of this will depend ultimately on whether the drug is efficacious, and we need phase 3 studies to understand that," Dr. Sano concluded.

Ron Petersen, MD, from Mayo Clinic Foundation in Rochester, Minnesota, said he believes that VE is a real phenomenon but still may not prove to be a limiting factor. "It doesn't appear that this line of research will be halted because of side effects," he said. "The question is, is it going to work?"

If the therapy does work, then monitoring, perhaps using MRI, would become important, he said.

But patients can develop VE and still be asymptomatic, although the potential is there for bleeds and other symptoms, Dr. Petersen added.

Dr. Petersen sits on the Data and Safety Monitoring Committee for 2 trials of active immunization being conducted by Janssen. "It may be interesting as to whether there's a difference between active and passive immunization on this [effect]," he said. "I don't know the answer to that yet."

The trial was funded by Janssen Alzheimer's Immunotherapy Program and Pfizer Inc. Dr. Petersen sits on the Data and Safety Monitoring Committee for 2 trials of active immunization being conducted by Janssen.

Alzheimer's Association International Conference (AAIC) 2011: Presentation O4-08-07, P4-392, P2-200. Presented July 18 and 20, 2011.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.