Baxter International Inc. announced topline data yesterday that showed its intravenous immunoglobulin (IVIG) therapy, known as Gammagard™, failed to meet primary endpoints of slowing cognitive and functional decline in a Phase 3 study of 390 patients with mild to moderate Alzheimer's disease (see company press release). People on Gammagard showed no improvement over those on placebo on the Alzheimer’s Disease Assessment Scale-Cognitive Subscale (ADAS-Cog), the Modified Mini-Mental State (3MS) Examination, or the Alzheimer's Disease Cooperative Study-Activities of Daily Living (ADCS-ADL). All current studies of the therapy in AD patients will be stopped, the company said.

Subgroup analyses provided a glimmer of hope, showing hints of cognitive benefit in patients with moderate Alzheimer's. Baxter will continue to analyze the data and present details, including imaging data, at the Alzheimer’s Association International Conference (AAIC) in July in Boston, Massachusetts. Several other companies are testing similar IVIG therapies in Phase 2 or 3 trials. Scientists involved in those studies expressed disappointment at Baxter’s news, but told Alzforum they will wait to see the full dataset before making any decisions. Other researchers noted that the finding deals a major blow to a promising AD therapeutic.

“The biomarker and imaging data to be presented in July may be helpful in interpreting the pattern of clinical effects, but overall the results were not encouraging for IVIG as a therapeutic avenue for AD,” Reisa Sperling at Brigham and Women’s Hospital, Boston, Massachusetts, wrote to Alzforum (see full comment below). She was not involved in the research.

Made from pooled human antibodies, Gammagard is approved for treatment of immune and inflammatory diseases. In an 18-month Phase 2 study on 24 AD patients, the drug appeared to slow clinical decline and brain shrinkage (see ARF related news story; ARF news story). The Phase 3 study, called the Gammaglobulin Alzheimer’s Partnership (GAP), was conducted in collaboration with the Alzheimer’s Disease Cooperative Study at 45 centers in the U.S. and Canada. Participants received either 400 mg/kg IVIG, 200 mg/kg IVIG, or placebo every two weeks for 18 months. The drug was well tolerated, with few adverse events.

Despite the overall negative results, pre-specified subgroup analyses showed that those taking 400 mg/kg who had moderate AD or carried an ApoE4 allele did better than the placebo group on the two cognitive assessments. However, the trial was not powered to detect statistical significance in subgroups. Norman Relkin at Weill Cornell Medical College, New York City, who ran the trial, wrote to Alzforum, “There are some positive findings in the pre-planned subgroup analyses that are encouraging. In my opinion, the future of the approach hinges on how robust those subgroup results prove to be and whether there are additional signals in the remainder of the clinical, imaging, and biomarker datasets.” (See full comment below.)

Finding therapeutic benefit in people with moderate AD stands in contrast to another experimental immunotherapy—Eli Lilly’s solanezumab—which improved cognition only at mild stages of the disease (see ARF related news story; ARF news story). The finding could suggest a somewhat different mechanism of action from solanezumab. On the other hand, it might simply reflect random chance in small samples, suggested Lon Schneider at the University of Southern California, Los Angeles. He noted that cognition appeared to decline slightly in patients with mild AD taking Gammagard compared to those on placebo, and that other subgroup results appeared contradictory as well. “All this speaks to play of chance and type 1 error,” Schneider said. “What you really have is just not much evidence to take [Gammagard] forward.”

Several other companies are developing IVIG-based AD therapies. CLS Behring AG in Bern, Switzerland, is evaluating its IVIG product for use in AD trials (ARF related news story). Grifols Biologicals, Inc., Los Angeles, California, is running a Phase 3 trial of IVIG in 350 AD patients. Octapharma AG, based in Lachen, Switzerland, previously reported that its IVIG missed its endpoints in a Phase 2 trial of 58 AD patients (see ARF related news story). Nonetheless, the trial did show improved glucose metabolism in treated patients compared to those taking placebo. Richard Dodel of Philipps University, Marburg, Germany, who led the study, said he is curious to see if those findings hold up in imaging data from the Gammagard Phase 3 study. “We have to wait. I think the story is not over,” he told Alzforum.

Shawn Kile at Sutter Neuroscience Institute, Sacramento, California, is currently testing Octapharma’s IVIG in a Phase 2 trial of 52 patients with mild cognitive impairment. Preliminary results suggest the treatment may slow brain atrophy in this group (see ARF related news story). “The GAP study will help us hone our approach and design of future investigations using IVIG, which still remains a potential treatment for Alzheimer's disease,” he wrote to Alzforum (see full comment below).––Madolyn Bowman Rogers.


  1. The IVIG results are very disappointing, as this means we are likely at least two to three years away from an approved disease-modifying therapy for AD dementia patients, as large-scale trials for other agents are still enrolling. The finding of a somewhat better response among ApoE4 carriers may reflect the higher likelihood of these patients harboring amyloid pathology, based on results from other Phase 3 trials, and thus potentially more likely to respond to immunotherapy. It is unclear how to interpret the trend level effects seen in a small subgroup of moderate AD dementia, but this might also reflect higher likelihood of amyloid pathology in more impaired patients or possibly effects of antibodies in IVIG targeting other aggregated proteins, such as tau, that may be more dynamic later in the disease process. The biomarker and imaging data to be presented in July may be helpful in interpreting the pattern of clinical effects, but overall the results were not encouraging for IVIG as a therapeutic avenue for AD.

  2. The topline results were announced today to satisfy SEC requirements and Baxter's obligation to its investors to report material findings in a timely fashion. However, analysis of data from the GAP study is very much ongoing. It is too soon to talk about the overall findings or the long-term implications. All we can really say right now is that the primary endpoints of the study were not met, and there are some positive findings in the preplanned subgroup analyses that are encouraging. In my opinion, the future of the approach hinges on how robust those subgroup results prove to be and whether there are additional signals in the remainder of the clinical, imaging, and biomarker datasets. I think we stand to learn a lot from this study, and I look forward to presenting a more complete picture at the AAIC meeting on 16 July 2013.

  3. Prior to commenting, I need to disclose that I am the principal investigator of an ongoing IVIG study for mild cognitive impairment due to Alzheimer's disease, which is supported by Sutter Institute for Medical Research and Octapharma.

    While the GAP trial did not demonstrate statistical significance for its primary endpoints, it does provide us with informative data about IVIG for Alzheimer's disease. First, there was a trend of improved cognitive function in the IVIG 0.4 g/kg every two weeks dosage arm, which has also been the optimal and most effective IVIG dosing in prior investigations. Second, this positive trend occurred in those subjects with more evidence of having dementia truly caused by Alzheimer's disease, i.e., genetic marker (ApoE4) or more clinically evident symptoms (moderate disease). Third, this larger trial supports the safety data of prior studies confirming that IVIG appears to be a relatively safe immunotherapeutic strategy for AD without any new safety concerns.

    I believe the GAP study will help us to hone our approach and design of future investigations using IVIG, which still remains a potential treatment for Alzheimer's disease.

  4. We clearly need to wait until the release of the complete data before drawing conclusions. However, what clearly strikes me here is the difference between the interpretations made from clinical versus basic science data. On the one hand, randomized clinical trial results are considered negative when primary endpoints are not reached. We all understand the purpose of this rule, which may, however, be a little stringent for multifactorial heterogeneous diseases. How can we expect to find one single drug that will benefit all subtypes of AD patients? This “one size fits all” approach cannot apply to a complex, elusive disease such as AD.

    On the other hand, I can’t help thinking that similar results from a preclinical study performed in animal models would have been twisted into a big “positive” publication. Indeed, the difference in response between ApoE4 carriers and non-carriers is so striking that it clearly deserves the attention of the AD research community. When in most clinical studies, mildly symptomatic ApoE4 non-carriers tend to respond better to treatments, the preliminary data from the GAP study suggest the opposite: Cognitive benefits after IVIg treatment were detected in patients with moderate ApoE4-positive AD. Again, we’ll have to wait until the full disclosure of the data because chance may have played a role here. Nevertheless, this looks like a very promising pharmacogenetic finding, which could have important mechanistic implications on how we should treat AD relative to ApoE status.

    Disclosure: I have been involved in preclinical research in IVIg supported by private and public funds.

  5. With every AD clinical trial having missed primary endpoints and therefore being deemed as "failed," one wonders whether the fundamental design of AD trials is such that they are destined to fail?

    Most human AD trials run for approximately an 18-month period, which is roughly equivalent to a two-week treatment in mice (based on their average lifespans). In AD mouse models, not many two-week treatments have resulted in statistically significant improvement in memory performance. Most successful treatments need two to three months to show memory improvements, which, one could argue, is equivalent to seven to 10 years in humans. So, are preclinical strategies failing in humans because we do not test them long enough?

    Second, the Assessment Score numbers (ADAS-COG, ADCS-ADL, and 3M) in the GAP study show a large variation—the standard deviation numbers are bigger than the mean values, which could be concealing positive effects. The use of biomarkers in place of assessment scores is not a good alternative since changes in biomarkers and cognitive performance are not always correlated. So, unless we can refine testing protocols to drastically reduce the variability, future trials will be similarly handicapped.

    One final point—as expected the mild AD cohort in the GAP study deteriorated much more slowly than the moderate AD patients during the 18-month period. However, the deterioration in the mild cohort was so small that IVIg-treated patients’ numbers were worse than the placebo group. We might face the same problem in prevention trials, where the treatment starts at presymptomatic stages of the disease. These are formidable but soluble challenges that will require considerable debate.

Make a Comment

To make a comment you must login or register.


News Citations

  1. Chicago: More Phase 2 News—PBT2 and IVIg
  2. Toronto: In Small Trial, IVIg Slows Brain Shrinkage
  3. Phase 3 Solanezumab Trials "Fail"—Is There a Silver Lining?
  4. The Solanezumab Benefit: Oh, So Small, But Probably Real
  5. Zuers—Meeting Mixes Translational News and Debate
  6. Research Brief: Octapharma IVIg Iffy in Phase 2 Trial
  7. Quick-and-Early IVIG Therapy: Hints of Promise?

External Citations

  1. Phase 3 study
  2. company press release
  3. Phase 2 study
  4. Alzheimer’s Disease Cooperative Study
  5. Phase 3 trial
  6. Phase 2 trial
  7. Phase 2 trial

Further Reading