The Evidence: Rigel Pharmaceuticals’ (RIGL) Fight for Fostamatinib’s FDA Approval

Read the article disclaimer first. This article may be updated with important information in the near future. Subscribe to Healthonym’s free newsletter to receive an alert if and when that happens.


Fostamatinib disodium (known as R935788 and R788 during development) is the generic name for Tavalisse, an oral medication sponsored by Rigel Pharmaceuticals, Inc. (Ticker: RIGL) with an expected PDUFA action date of 4/17/2018.

Fostamatinib is a prodrug. That is to say, it only becomes an active compound once it’s processed by intestinal enzymes into the active form of the drug, known as R406.

This medication is indicated for the treatment of primary persistent/chronic immune thrombocytopenia, sometimes called immune thrombocytopenic purpura and idiopathic thrombocytopenic purpura, (ITP).

In other words, fostamatinib is used to treat a condition that results in the long-term (> 3 month) depletion of platelets, sometimes despite numerous treatments.

Moreover, while the initial cause of primary ITP is unknown, we do know that the immune system ends up depleting the platelets in this condition anyways.


A macrophage deforming itself in order to capture and destroy a platelet.

A macrophage deforming itself in order to capture and destroy a platelet.

And so, fostamatinib works by counteracting an aberrant immune response by inhibiting a biological catalyst called spleen tyrosine kinase (Syk).

Syk is necessary for white blood cells, called macrophages, to properly function. When triggered, Syk signals these cells to re-arrange their shape in order to capture and destroy anything the immune system has marked for elimination.

In ITP, the immune system erroneously marks platelets for eradication. So if fostamatinib stops Syk from working properly, we can minimize the destruction of platelets as a result.


In ITP, a person loses platelets through a combination of possible mechanisms, including:


1. The elimination of platelets by macrophages. This is what fostamatinib tries to stop.

2. The destruction of platelets by other white blood cells, such as cytotoxic (poisonous) T-cells.

3. The eradication of platelets by immune system proteins called complements.

4. The inhibition of the production of new platelets.

5. The destruction of cells that produce new platelets.

Based on fostamatinib’s mechanism of action, it should be clear that this drug focuses on #1.

And even though #1 is a major reason for platelet loss, this nonetheless means there are plenty of other ways by which a person with ITP can lose platelets for which fostamatinib is expected to have no significant therapeutic effect.

This, combined with our discussion of numerous issues below, may force fostamatinib to resign itself to becoming a third or even fourth-line treatment option with a small patient base and thus limited revenue potential compared to alternatives.


In this drug’s case, data from basic science tells us that at least 40% of patients with ITP are expected to lose platelets in ways that fostamatinib cannot counter.

That is to say, over 40% of people with chronic/persistent ITP are not expected to respond to fostamatinib no matter what.

This percentage is confirmed by actual clinical data with trials on this drug, which actually paint a bleaker picture because a greater percentage of patients than this fail to effectively respond to the drug (especially if we ignore the overly optimistic post-study analyses conducted by RIGL – which we largely will).

That being said, the limited efficacy might be explained by things other than the drug’s target and mechanism of action, including interactions with other medications and the influence of other conditions a patient may have. But there is not enough quality phase II and III data to be sure if this is truly the case.

Regardless, there is at least one very good thing about fostamatinib’s mechanism of action.

Because fostamatinib works specifically to stop macrophages from destroying a person’s platelets, this should minimize the total number of side-effects it should be expected to have when compared to more general treatment approaches, such as steroids.


Authors associated with RIGL like to term fostamatinib as a “selective”, or sometimes a bit more appropriately, as a “relatively selective” Syk inhibitor.

The choice of words here is interesting as it would superficially imply that such selectivity improves the safety profile of this drug from the basic science perspective we just went over. After all, drugs that are more biochemically selective are less likely to lead to unwanted effects.

With respect to this, there is definitely some preclinical evidence to suggest that fostamatinib/R406 is somewhat preferentially selective for Syk when compared to other similar biological targets. So fostamatinib is, in a sense, relatively selective (in select cases, strongly so).

However, there is also plenty of evidence that shows fostamatinib may exert enough selectivity, even clinically relevant selectivity, for other targets.

Moreover, fostamatinib/R406 might exert even stronger preferences for some biological targets than Syk itself.

This, of course, is worrisome from a safety profile perspective and it may help explain why more recent patent filings in this arena have been for drug molecules that are more selective than fostamatinib appears to be.

Consequently, fostamatinib’s interactions with other targets may lead to (most notably):

1. Numerous gastrointestinal side effects.

2. Increased blood pressure.

3. An increased risk of infection and/or abnormalities associated with some types of white blood cells, the cells that protect us from infection.

These basic science concerns were confirmed in numerous clinical studies.


We’ll get to those studies after going over the preclinical data.

On that note, upon review of the numerous animal studies, data, and respective results on the drug, the following could be said with a reasonable degree of certainty:

1. Fostamatinib appears to protect animals from ITP. Meaning, animal studies confirm this drug’s potential efficacy for this condition.

2. Fostamatinib may increase the chances of birth defects. Proper labeling would help take care of this issue.

3. Fostamatinib might increase a patient’s blood pressure. This high blood pressure can be managed with other medications or the discontinuation of the drug. It appears that high blood pressure meds will not significantly influence the therapeutic levels of fostamatinib.

4. Fostamatinib has the potential to suppress part of the immune system, especially when tested in longer term animal studies.

5. The available animal data pointed to the possibility that it will be difficult to find a dose for this medication that is simultaneously safe, effective, and a better alternative to existing treatment options (including first line therapies). This notion is partly confirmed by clinical data as well.

#5 is actually the most important thing here. While #4 seems pretty bad we should remember that many treatments for ITP have this and many other safety issues but are still on the market.

So it’s not so much the adverse effects that matter on their own, it’s rather how safe and effective fostamatinib is compared to alternatives that really matters here (and whether or not clinical trial data strongly supports its touted safety/efficacy profile).


Let’s see what the clinical data has to say about all of this. We’ll start with the phase I trials.

Numerous generally well-conducted phase I studies using healthy subjects assessed the pharmacology and safety of fostamatinib.

Overall, most side effects were mild to moderate in nature and mainly limited to gastrointestinal issues and dizziness. Some issues related to immune system suppression were made apparent but the data provided didn’t allow for a full appreciation of exactly what was going on.

There are no overt issues with the phase I trials but they don’t really tell us much with respect to safety or efficacy either.


The phase II study was far more revealing and quite interesting from numerous standpoints.

There were two major endpoints in the trial and the relevant platelet counts the study authors had selected for these endpoints were appropriate according to known standards for this condition, its treatment, and study intent.

Depending on how you look at the statistics here, seemingly 50%-75% of patients in this small (16 person) trial showed some form of positive response to fostamatinib as seen with appropriately increased platelet counts.

On the surface of things, the results of this phase II study seemed quite encouraging. This is the good news.

But there were some important caveats we found and we’ll go over one of them.

One major caveat was that one of the study’s endpoints, with positive findings no less, was irrelevant. Irrelevant to the FDA, not the sponsor though, for this endpoint’s findings can be used to guide the design of phase III studies (and perhaps this actually happened).

But for us, that’s largely beyond the point. That’s because the methodology used to come up with this endpoint (and its positive findings) is fraught with numerous potentially serious statistical perils we won’t discuss herein for brevity’s sake.

The FDA knows this and largely ignores the results of such findings when it comes to approving a medication as a result.

There is a potential workaround for problems like this, one that can actually give this apparently problematic endpoint’s positive findings more credibility in the eyes of the FDA.

However, there was no mention that any of the techniques used “to fix” or allay some (but not all) of this endpoint’s statistical concerns were employed by the study’s authors (or they were and the results weren’t so great and thus went unmentioned).

The takeaway for us is that we must assume that a nicely sized chunk of the good news coming from this phase II study, with respect to this drug’s effectiveness, is actually of no serious consequence to the chances of this drug’s approval and might actually disguise a weak hand.


The other side of the phase II coin is safety. Safety-wise, gastrointestinal side effects were noted as per the phase I studies with a not insignificant number of withdrawals due to these side-effects.

This brings into question how many people, even as a third or fourth tier therapy option, will truly benefit from the drug if they are unable to tolerate it.

Furthermore, elevated blood pressure was confirmed as a possible side effect as was a decrease in white blood cell counts among other adverse events. This simply reiterates all of the preclinical and clinical concerns we went over above.

That being said, given the very small sample size of this study, it’s hard to make heads or tails of what any of this efficacy or safety data really means. The data here can be twisted in a positive and negative manner quite easily as a result.

The conclusion here is that we should rightfully be uncertain about the implications of the study’s results with respect to safety and even the appropriate dose for the medication.

Finally, this phase II study was unable to adequately answer why there was so much variability in patient response to the medication as a result of a technicality that the FDA may want RIGL to revisit.

It’s best to look towards the quantitatively and qualitatively better phase III studies for more weighted clues as to what’s really going on here.


RIGL conducted a phase III study, a parallel phase III study, and an extension study at significant risk of bias. We go over the more methodologically rigorous phase III studies in this article as a result.

These phase III studies were, by and large, qualitatively well-designed and conducted based on what we know. Quantitatively, though, there are numerous concerns.

One of the phase III studies met its primary endpoint. Critically, the parallel one did not. Pooled data from both trials suggests that fostamatinib is effective.

This kind of analysis may make it seem like the parallel trial’s failure isn’t a big deal. Yet there are two issues with this type of thinking.


First, pooling data can be hazardous if not done correctly. We don’t know enough about the parallel phase III studies to be sure either way if the pooled data is appropriate or not.

If it is, it can indeed provide a stronger case for safety/efficacy and we should give RIGL the benefit of the doubt.

Even so, we must be aware that one of the major reasons for NDA rejection is inconsistent trial results exactly such as these. So just because the pooled data looks good, that doesn’t mean the FDA will care as it further reconciles it with numerous other concerns.

Concerns such as the following: the available data (and equivocal results) suggest that the trials were numerically too small to draw strong conclusions about safety and efficacy.

There appears to be a strong probability that chance played an outsized role in the positive and negative results, making any conclusion about safety or efficacy that much more difficult.

This means the results may be skewed too much towards RIGL’s favor or they might make the results look worse than they truly are. We can’t be sure either way and must be careful in being overly optimistic about the available data.

After all, in an ultimately dismissed lawsuit, RIGL was accused of potentially delaying the distribution of negative trial data in the past.

Regardless, it’s also hard to scientifically (albeit not financially) reason why RIGL used such small sample sizes in their phase III studies in light of the fact that plenty of FDA approved orphan drugs with far smaller patient populations used much larger sample sizes than RIGL to prove safety and efficacy.

Another concern is that the trials used a highly controversial method of assigning patients to the treatment arm or placebo group. While there may be legitimate reasons for using this method of patient assignment in some cases, the FDA generally appears to frown upon this methodology.

If RIGL had justifiable reasons for using this controversial method, they nonetheless might’ve been able to use a more rigorous study design alongside it in order to allay the method’s many concerns and, at the same time, provide for stronger evidence for fostamatinib’s efficacy and safety.

It would’ve been win-win.


What is abundantly clear, based on all of the above, is that there will be a need to look towards other data for, at minimum, a better interpretation of possible safety concerns.

The FDA knows to look for this additional information and RIGL probably volunteered it anyways.

The best data for all of this actually comes from phase II and III clinical trials on fostamatinib for rheumatoid arthritis and lymphoma.

Yes, the indications and endpoints are different but there are enough similarities, such as the doses used, to draw some appropriate inter-study comparisons in some areas.

Moreover, the generally well-conducted arthritis and lymphoma clinical trials are, on average, much larger and provide us with more robust information than the ITP clinical trials.

Unfortunately for RIGL, the individual and pooled results of these far larger studies imply that, overall, fostamatinib has numerous safety issues that were likely simply missed or not emphasized in the smaller ITP clinical trials.

This notion is tempered by the fact that comorbid conditions and medications may have confounded these results.

Notably, but less crucially given inter-trial differences, the studies tended to show that the more rigorous the trial design, the less efficacious fostamatinib appeared to be across all studies and indications.

The FDA will be well aware of all of this when it reconciles the safety data from the arthritis and lymphoma trials with those of the smaller and equivocal ITP trials.


The other important point revealed via all the available preclinical and clinical trial data, including ITP-based data, was the fact that fostamatinib may have a very narrow range at which it is both adequately effective yet appropriately safe.

That is, it may be quite difficult to establish a safe and simultaneously effective dose for this drug. As we can all understand, this is a potentially serious pitfall.


Putting the clinical data aside, let’s turn our attention to a small regulatory-based note.

The FDA designated fostamatinib as an orphan drug. In other words, the FDA recognizes that this drug is intended to treat what is a rare disease or subset thereof.

Incentives for drug development apply as a result but they are not relevant to our discussion.

One of the pluses of having this designation is the FDA’s tolerance of clinical trials with far smaller sample sizes than for standard NDAs/BLAs.

To be clear: the FDA has approved orphan drugs with studies conducted on literally a handful of patients and with what were, frankly, anything but rigorous study designs.

So we must emphasize that the small sample sizes in fostamatinib’s ITP case, on their own, are not the concern here.

The actual problem is that this is combined with inconsistent or irrelevant phase II/III results and, among other pitfalls discussed before, known safety issues found in larger non-ITP trials but not emphasized in the less rigorous ITP trials.

If we’re going to highlight a good thing with respect to fostamatinib’s orphan drug designation it’s that orphan drugs, all else equal, are more likely to be approved than standard NDAs so this definitely plays in RIGL’s favor.


So far, getting Syk inhibitors to FDA approval has proven troublesome. This isn’t for a lack of effort or want either.

Many Syk inhibitors have simply failed or stalled along the way for one reason or another. Even studies on fostamatinib for lymphoma and rheumatoid arthritis have unfortunately ultimately come up short.

While saying “this time it’s different” is nice, we shouldn’t close our eyes to everything outlined herein either and the issues (unexplored in this article) that point to numerous problems with Syk inhibitors and thus explanations for why they haven’t been approved.

Again it should be noted that, individually, it’s unlikely that any of the safety, efficacy, quantitative, or methodological issues we outlined are grounds for failure.

Many of the same safety issues exist in standard therapies. Even if this drug is effective for a few people, it may be better than nothing given the issues surrounding chronic ITP. Even the question marks surrounding this drug’s comparable safety/efficacy to standard treatments can be countered.

But combined together, all of these issues would likely hamper any standard NDA’s chances of approval especially in light of better alternatives.

The major saving grace here is that this drug has orphan status and that may allow for some (potentially significant) leeway on some issues in the eyes of the FDA that otherwise wouldn’t be afforded to another drug candidate.


Read the article disclaimer.

Published: 3/14/2018

Last Updated: 3/14/2018

No one associated with this article has any financial stake in, or ties to, RIGL.