Just as the clamor about Biogen’s seeking FDA approval for aducanumab with one positive Phase 3 trial started to die down, a Chinese company garnered a regulatory thumbs-up for its drug, also with one Phase 3 trial. On November 2, Shanghai Green Valley Pharmaceuticals announced that China’s National Medical Product Administration (NMPA) had conditionally approved GV-971, aka oligomannate, for the treatment of mild to moderate Alzheimer’s disease. Full approval, expected to come in spring 2020, hinges on the company submitting animal safety studies that have already been completed, according to a spokesperson for the company.

  • GV-971 conditionally approved in China to treat AD.
  • Drug met its one primary, but not secondary endpoints, in a single Phase 3 trial.
  • Global Phase 3 trial slated to begin in 2020, in United States, Europe, China.

A single Phase 3 study apparently found that GV-971, reported to alter the gut microbiome, slowed cognitive decline in people with AD. The trajectory of cognitive change in the nine-month trial raised some eyebrows—for one, the placebo group did unusually well for weeks before their cognitive scores plummeted. A global Phase 3 trial, slated to begin in 2020, will test if GV-971 passes muster in the United States, Europe, and China.

Green Valley has developed carbohydrate-based drugs for a number of chronic diseases since 1997. It touts GV-971, a mixture of oligosaccharides derived from brown algae, as restoring balance to the gut microbiota. A recent study led by the drug’s discoverer, Meiyu Geng at the Chinese Academy of Sciences in Shanghai, reported that the bacterial flora inhabiting the guts of 5xFAD and other mouse models of AD differed from the flora of wild-type animals. This triggered infiltration of T cells into the brain, where they stoked microglial activation and led to damaging neuroinflammation. GV-971 reportedly altered the microbiome such that it no longer triggered neuroinflammation. Furthermore, the saccharide reportedly reduced Aβ burden, tau hyperphosphorylation, and cognitive deficits in the mice (Wang et al., 2019). 

Since news of GV-971’s approval burst onto the scene, some commentators have scrutinized this preclinical data. They noted instances of image duplication in previous studies published by Geng’s lab, and recalled regulatory troubles with Green Valley from the past (see Science Translational Medicine blog). 

Liping Zhao, a microbiome researcher at Rutgers University, New Jersey, noted that in the preclinical study, treatment of 5xFAD mice with GV-971 appeared to boost the growth of certain taxa that contain opportunistic pathogens that could aggravate inflammation, and suppressed others known to pump out inflammation-soothing short-chain fatty acids. “Taken together, the microbiome data … did not support the hypothesis that part of the mechanism for GV-971 to alleviate AD might be reducing inflammation by way of modulating the gut microbiota,” he wrote in a comment to Alzforum. “On the contrary, the microbiome data provided in the Cell Research paper indicates a possibility that GV-971 may aggravate the dysbiosis of the gut microbiota and thus potentially increase inflammation in AD mice.” (See full comment below.)

Green Valley has been studying this oligosaccharide in people for some years. In 2008, the company completed a Phase 1 study in healthy men, followed by a Phase 2 study in 255 people with mild to moderate AD that began in 2011. Results of that Phase 2 study were presented at the Clinical Trials in Alzheimer’s Disease (CTAD) meeting in 2014. While a 900 mg daily dose of GV-971 had not significantly altered the trajectory of ADAS-Cog12 scores over 24 weeks, i.e., had failed the primary endpoint, it did appear to have slowed decline on Clinician's Interview-Based Impression of Change Plus Caregiver Input (CIBIC-Plus), a secondary endpoint (see Dec 2014 Medscape news story). As yet, no clinical trial data on GV-971 are published in the peer-reviewed literature.

Geng presented findings from the Phase 3 trial, which started in 2014, at the last CTAD conference in 2018, and Green Valley filed for approval shortly thereafter, according to the company spokesperson (Nov 2018 news). Conducted at 34 sites in China, this trial enrolled 818 people clinically diagnosed with mild to moderate AD. Roughly half were randomized to twice-daily doses of 450 mg of GV-971, the others to placebo. Participants could not take cholinesterase inhibitors, the standard of care in many countries, during the trial. Change in ADAS-Cog12 performance at 36 weeks served as the primary endpoint. Secondary endpoints—none of which were met—included changes on the CIBIC, activities of daily living (ADL), neuropsychiatric inventory (NPI), and FDG-PET. Neither amyloid PET scans nor fluid biomarkers were used anywhere in the trial.

Curious Curves. ADAS-Cog12 scores similarly improved in both treatment and placebo groups, with a group difference at four weeks sharply increasing at 36 weeks. [Courtesy of Shanghai Green Valley Pharmaceuticals.]

GV-971 was safe and well-tolerated, with comparable side effects to placebo, according to the company press release. Compared with baseline scores, both the treatment and placebo groups improved on the ADAS-Cog12 by four weeks, though those on GV-971 improved slightly more than those on placebo. At 12 and 24 weeks, both groups essentially maintained their scores. At 36 weeks, the gap widened as placebo-group scores plummeted to just above baseline, while those in the treatment group reportedly bettered their baseline scores by 2.70 points, making for a 2.54-point difference between the groups.

This is slightly larger than differences reported for cholinesterase inhibitors, said Jeffrey Cummings of the Cleveland Clinic, Lou Ruvo Center for Brain Health in Las Vegas, who is a scientific adviser to Green Valley. Separation between groups was largest—with a 4.55 ADAS-Cog point difference—in participants with the lowest MMSE scores at baseline, i.e., those whose dementia was most advanced.

Lon Schneider of the University of Southern California, Los Angeles, said the trajectories were unusual, noting that the placebo group would be expected to worsen by around 1.5 points on the ADAS-Cog over six months. He noted that a similar placebo response was observed over the 24-week Phase 2 trial. “In the Phase 3 trial, after closely tracking the oligomannate group for 24 weeks, the placebo group took a nosedive, returning to its baseline by week 36. This unexpected and unexplained inflection point for the placebo group accounts for the significance on the ADAS-Cog12 at week 36,” he wrote.

Cummings also found the performance of the placebo groups in both trials odd. One possible explanation is the relatively low standard of healthcare in China. Regular doctor visits are not the norm there, and might by themselves have a salubrious effect on trial participants. Cummings said he was convinced that the trial was well-conducted, with standards typical of those in the United States and European Union.

The mechanism of this compound remains unclear, Cummings said. While the gut microbiome, which is reportedly altered in people with AD, could be involved, Cummings thinks it will be important to consider other possibilities. According to previous in vitro studies from Green Valley scientists, seaweed-derived oligosaccharides also bind to and de-aggregate various forms of Aβ, and quell inflammatory responses to them (see Hu et al., 2004; and Wang et al., 2007).

Eric Siemers, Siemers Integration LLC, Zionsville, Indiana, stressed the need for another, larger study. Siemers agreed that the trajectory in the placebo group was unusual and that the drug’s mechanism of action remains unclear. He added that the group separation at the first few time points was small. Siemers asked how much of this rather large oligosaccharide crosses into the brain, and whether it has been shown to affect the gut microbiome in people. Positive findings in a second trial would indicate the drug most likely offers symptomatic relief, he said.

“While the drug in question may soon be available in Chinese markets and there are efforts to initiate global level trials, extreme caution is warranted before the scientific and medical community can embrace this new drug for treating AD patients,” wrote Malú Tansey and Paramita Chakrabarty of the University of Florida in Gainesville, in a joint comment to Alzforum. They stressed the importance of nailing down the drug’s mechanism of action, and of evaluating it in diverse populations, especially since the microbiome varies across ethnicities.

The drug will be available to people in China by the end of 2019. Green Valley plans to seek marketing authorization in other countries as well. It farms its own seaweed, from which it extracts the oligosaccharides that make up GV-971, according to the company.

In the United States, the FDA sometimes approves drugs pending further data linking a biomarker response to a meaningful clinical outcome. In this case, by contrast, the Chinese NMPA approved GV-971 on the condition of reviewing further animal safety reports. It is unclear how the drug’s approval in China could be affected by a potential failure in the global trial, Cummings said.

That global trial is slated to begin in 2020, and will include more than 1,000 participants from the United States, Western Europe, and China, Cummings said. In the meantime, will people outside of China be able to obtain the drug? Cummings said the company is considering asking the FDA to create a compassionate-use indication, which could theoretically give people with AD access to the drug. Without a legitimate pathway for expanded use outside of China, counterfeiting is likely to take place, he said.

“Approval in China for a new treatment for Alzheimer’s disease is encouraging,” wrote Stephen Salloway of Brown University in Providence, Rhode Island. “We await publication of the study that led to approval.” Salloway noted that people who took cholinesterase inhibitors were excluded from the trial. “Further study is needed in the United States and elsewhere that includes the background use of cholinesterase inhibitors, the current standard of care,” he added.

David Holtzman of Washington University in St. Louis agreed. “It will also be important to sort out whether the effects observed in humans are through neurotransmitter modulation accounting for a symptomatic benefit, or whether they are due to altering the microbiome, with secondary effects on the brain as has been recently seen in mouse models with GV-971,” he added. Holtzman co-authored an editorial about the recently published preclinical studies (Seo et al., 2019).

Cummings told Alzforum that Green Valley has not yet completed a trial to assess interactions between GV-971 and cholinesterase inhibitors, but that one is planned. Therefore, the upcoming global Phase 3 trial will likely prohibit use of cholinesterase inhibitors among its participants, a factor he acknowledged could hinder enrollment.—Jessica Shugart

Comments

  1. Approval in China for a new treatment for Alzheimer’s disease is encouraging. We await publication of the study that led to approval. From the data presented at CTAD, cholinesterase inhibitors were not allowed and the drug worked best for those with a Mini Mental State Exam score of 11-15. There was improvement relative to placebo on the ADAS-Cog but not on the Activities of Daily Living scale and with a trend on the Clinician’s Global Impression of Change. The mechanism of action is novel, but it is hard to know the exact mechanism that might provide clinical benefits. Further study is needed in the United States and elsewhere that includes the background use of cholinesterase inhibitors, the current standard of care.

  2. The compound GV-971, Sodium oligomannate, was recently tested in a people who, based on clinical criteria, had mild to moderate dementia believed to be due to Alzheimer’s disease. Over a 36-week period, participants on GV-971 performed better than placebo-treated individuals by ~2.5 points on the ADAS-Cog test. This is a similar effect on this test to studies using cholinesterase inhibitors such as donepezil. Based on this and other data, GV-971 was approved to treat patients with mild to moderate dementia of the Alzheimer’s type in China.

    It is exciting that a new treatment is being approved for mild to moderate dementia in China. To my knowledge, the patients who participated in the trials in China were not on cholinesterase inhibitors or memantine. It will be important in future studies in the United States, Europe, and other places to compare the effects of GV-971 to the currently utilized treatments in the field, or when added to these treatments. It will also be important to sort out whether the effects observed in humans are through neurotransmitter modulation accounting for a symptomatic benefit, or whether they are due to altering the microbiome with secondary effects on the brain as has been recently seen in mouse models with GV-971.

  3. The conditional marketing approval by China’s FDA, the NMPA, for GV-971 also limited its price to twice that of donepezil. If donepezil is as cheap in China as it is at Costco (i.e., $30.30 per year with a GoodRx coupon), then $60.60 is a steal for this plant-derived product, whether or not it’s effective and as long as it’s safe. The prior Phase 2 study and this Phase 3 trial report adverse events generally lower than placebo, suggesting that oligomannate is remarkably safe and perhaps without side effects.

    However, the company provides curious and tenuous results for its efficacy, even if we accept the outcomes at face value. The Phase 2, 24-week trial of 255 patients showed significance on the CIBIC+, but not on any other outcome (presented at CTAD in 2014, results at clinicaltrials.gov). The Phase 3, 36-week P3 trial with 788 patients reported a significant 2.54 ADAS-cog12 point advantage over placebo at 36 weeks, but also was not supported by any other outcome (presented at CTAD 2018). Both trials looked rather similar up to 24 weeks, showing substantial improvement for both drug and placebo on the ADAS-cog12 within four weeks and stabilization of this improvement in both treatment groups over 24 weeks, with very small and questionably clinically meaningful drug-placebo differences of 1.1 and 0.7 ADAS-cog12 points. In the P3 trial, after closely tracking the oligomannate group for 24 weeks, the placebo group took a nosedive, returning to its baseline by week 36. This unexpected and unexplained inflection point for the placebo group accounts for the significance on the ADAS-cog12 at week 36. 

    Some features of the study may have a bearing on its external validity. They include an apparent marked practice effect on the ADAS-cog12 when patients with dementia often don’t show much, and the lack of cognitive decline over 24 to 36 weeks, as seen in the vast majority of mild to moderate AD trials (Schneider and Sano, 2009). 

    There also may be issues with internal validity that would become evident once we know more about the statistical analyses. They might include the rather small standard errors suggesting smaller-than-expected standard deviations and less variance in the ADAS-cog outcomes than in other trials.

    Patients for this study were not allowed to take cholinesterase inhibitors, and as has been demonstrated in U.S. samples, the differences in clinical characteristics and course between those taking AChEIs and those not taking them may affect outcomes. Any trial in the United States likely would have to enroll patients on AChEIs. The lack of clarity and details inherent in conference presentations and press releases limits inferences we can make.

    References:

    . Current Alzheimer's disease clinical trials: methods and placebo outcomes. Alzheimers Dement. 2009 Sep;5(5):388-97. PubMed.

  4. The research team reports that the initial results from the Phase 3 trial look very promising. The team claims that the drug was extremely well-tolerated, with literally no side effects; that it stabilized the decrease in ADAS-Cog within four weeks of initiation. Yet without data on other measures, such as CSF biomarker levels, FDG-PET and Aβ PET signals, plasma levels of cytokines and some indication as to how the microbiome is progressively changed in these patients, the outcome of this trial should be treated as promising but preliminary. Other critical information, such as whether these patients were on any other cognitive enhancers or undergoing neurocognitive therapies, was not included in their reports and is needed to evaluate whether the positive outcomes can be attributed directly to GV-971. Although preclinical studies have hinted that GV-971 has multifaceted roles in AD mouse models, such as altering Aβ metabolism, controlling neuroinflammation, and normalizing gut microbiome, we do not yet know to what extent these parameters changed in these patients.

    While the drug in question may soon be available in Chinese markets, and there are efforts to initiate global level trials, extreme caution is warranted before the scientific and medical community can embrace this new drug for treating AD patients. Specifically, there should be concerted efforts to nail down the mechanism of action as well as establishing reproducibility of the beneficial effects in a larger cohort of patients of different ethnicities since ethnic differences influence gut microbiome composition.

    That said, if these results can be replicated in additional, larger cohorts of diverse ethnicities, this drug may hold promise in other neurological disorders where disruption of the gut-brain axis and dysbiosis has been implicated in etiology, including autism spectrum disorders and Parkinson’s disease.  

  5. Out of curiosity, I read the preclinical Cell Research paper on how GV-971 modulated the gut microbiota of transgenic (Tg) mice (Wang et al., 2019). 

    I was surprised to find out that the microbiome data did not support the hypothesis that GV-971 may alleviate inflammation via modulation of the gut microbiota.

    This relevant data is in Fig. 4.

    Fig. 4a is a principal coordinate analysis of the gut microbiome composition at the species level (operational taxonomic units) based on the Bray-Curtis distance for Tg and GV-971-treated Tg mice. This shows that these two groups of mice have significantly different gut microbiota. This could mean that GV-971 changed the gut microbiota of Tg mice.

    Fig. 4b is a heat map of significantly changed gut bacteria represented at the genus level between Tg and GV-971-treated Tg mice. Among the taxa that were significantly promoted by GV-971 were those that contained opportunistic pathogens that may aggravate inflammation. A typical example is the family Desulfovibrionaceae. Members of this family, when found in human or animal gut, are endotoxin and hydrogen sulfide producers, which can induce or aggravate inflammation. This means that potentially proinflammatory gut bacteria were promoted by GV-971. Among the taxa that were significantly diminished by GV-971 were those that can produce short-chain fatty acids such as Roseburia spp. Members of genus Roseburia can produce butyrate that can mitigate inflammation. Increased abundance of this genus has been associated with weight loss and reduced insulin resistance in mice. Astonishingly, this potentially beneficial, anti-inflammatory group of gut bacteria was reduced by GV-971.

    Taken together, the microbiome data provided in the Cell Research paper, being associative in nature, do not support the hypothesis that part of the mechanism for GV-971 alleviating AD might be by reducing inflammation by way of modulating the gut microbiota.

    On the contrary, the microbiome data provided in the Cell Research paper indicates a possibility that GV-971 may aggravate the dysbiosis of the gut microbiota and thus potentially increase inflammation in AD mice.

    References:

    . Sodium oligomannate therapeutically remodels gut microbiota and suppresses gut bacterial amino acids-shaped neuroinflammation to inhibit Alzheimer's disease progression. Cell Res. 2019 Oct;29(10):787-803. Epub 2019 Sep 6 PubMed.

  6. The authors should publish the full study as soon as possible. The observed effect size is relatively small, and is only slightly greater than the minimal clinically important difference of ADAS Cog. So, I don't view this as a breakthrough. If they sell this drug at the same price as Aricept, I think it is not really good news for Chinese patients.

Make a Comment

To make a comment you must login or register.

References

Research Models Citations

  1. 5xFAD (B6SJL)

News Citations

  1. Fits and Starts: Trial Results from the CTAD Conference

Paper Citations

  1. . Sodium oligomannate therapeutically remodels gut microbiota and suppresses gut bacterial amino acids-shaped neuroinflammation to inhibit Alzheimer's disease progression. Cell Res. 2019 Oct;29(10):787-803. Epub 2019 Sep 6 PubMed.
  2. . Acidic oligosaccharide sugar chain, a marine-derived acidic oligosaccharide, inhibits the cytotoxicity and aggregation of amyloid beta protein. J Pharmacol Sci. 2004 Jun;95(2):248-55. PubMed.
  3. . A marine-derived acidic oligosaccharide sugar chain specifically inhibits neuronal cell injury mediated by beta-amyloid-induced astrocyte activation in vitro. Neurol Res. 2007 Jan;29(1):96-102. PubMed.
  4. . The microbiome: A target for Alzheimer disease?. Cell Res. 2019 Oct;29(10):779-780. PubMed.

External Citations

  1. announced 
  2. Phase 3
  3. Meiyu Geng
  4. Science Translational Medicine blog
  5. Phase 2 study
  6. Dec 2014 Medscape news story

Further Reading

No Available Further Reading