Amicus (FOLD) is a biotech company that focuses on orphan diseases. Its research area is mainly Fabry Disease along with other lysosomal enzyme defects. The company has a big-pharma partner, GlaxoSmithKline (GSK). In this article I will take a closer look at Amicus' Migalastat HCI, a small-molecule drug candidate for Fabry Disease.
Fabry Disease is characterized with a defect in the enzyme alpha-galactosidase-A. This leads to accumulation of globotriaosylceramide (an intermediate metabolite in lipid metabolism, also known as GL-3) in many tissues and therefore a multisystemic disease develops. The prevalence of Fabry disease is estimated to be 1 per 40,000 people. Fortunately, there are treatment options: Two enzymes, agalsidase-alpha (Replagal) and agalsidase-beta (Fabrazyme), reportedly help in normalizing renal function, cardiac function and cerebrovascular flow. Unfortunately, these enzyme replacement therapies cost around $200,000 per year. (Source: Medscape Emedicine)
A mutation in alpha-galactosidase-A enzyme causes misfolding of this protein. In the cell, misfolded proteins/enzymes in the endoplasmic reticulum (ER) need to be degraded, so is the misfolded mutant alpha-galactosidase-A. These misfolded enzymes still retain some catalytic activity which might save the patient, but nonetheless ER degrades these enzymes as they cannot escape ER's quality control.
Amicus' approach is to stabilize the misfolded mutant alpha-galactosidase-A enzyme with small chemical compounds and therefore allow the stabilized enzymes to escape from ER's quality control. Even 10-30 percent of normal enzymatic activity prevents accumulation of the globotriaosylceramide in the body and prevents any damage caused by accumulated globotriaosylceramide.
Amicus is using a "stabilizing compound" (migalastat HCI) which binds to the catalytic site of the alpha-galactosidase-A enzyme. This compound is mimicking the substrate, binding to the catalytic site and producing a "transitional state" for the enzyme. This transitional state is normally seen during catalysis. "Transitional state" as you can predict is a transitional form of the enzyme, which drives its function. But this stabilizing compound is not substrate for alpha-galactosidase-A. Moreover migalastat is a competititive inhibitor given its binding ability to bind to the same site as its actual substrate. (If you are interested you can read more on induced-fit model of enzymatic catalysis.)
Yes, Amicus has stabilized the enzyme by a competitive inhibitor (migalastat) of the alpha-galactosidase-A enzyme. But cutting-edge technology plays a role after this point: The stabilized migalastat - alpha-galactosidase-A enzyme complex escapes from ER quality control and enters the lysosome, where the pH is low (around 5). Low pH causes migalastat to dissociate from the catalytic site of alpha-galactosidase-A, and now the partially active alpha-galactosidase-A is free to do its job in the lysosome like it normally would.
Great chemistry, great enzymology and great biology. In this article I am going to analyze whether those 3 greats are likely to turn into great medicine.
There are lots of different mutations which cause Fabry Disease. These include point mutations, deletions etc (PMID: 7504405). My first concern is whether migalastat is likely to work with all these mutant alpha-galactosidase-A enzymes. An enzyme with a specific mutation might be stabilized by migalastat and that enzyme might be able to escape from ER quality control. But other mutations might prevent migalastat-enzyme binding. It is also possible that some mutations render the enzyme completely dysfunctional. So, my opinion is that migalastat must have been studied in subpopulations of Fabry disease patients who carry the same mutation.
Amicus is now continuing to its phase 3 study for migalastat (NCT00925301); we do not know whether the company is studying a subgroup of patients with same mutation. It announced disappointing results from the study on 19th of December 2012 which caused %47 decline in stock prize.
The company stated regarding Study 011 (migalastat HCl vs. Placebo): This difference did not achieve statistical significance (p=0.3) according to the pre-specified primary endpoint analysis.
My guess is that differences in mutations of alpha-galactosidase-A caused varying responses to the drug. Therefore, a statistical difference between placebo and migalastat arms were not found in that study. Migalastat might have worked in one mutation type, but we cannot know for certain.
Amicus also stated in that press release: Study 011 randomized 67 patients (24 males and 43 females) diagnosed with Fabry disease who had genetic mutations amenable to chaperone monotherapy in a cell-based assay.
This means that it has not grouped patients according to mutation types. Rather, it worked on patients with different mutations. I believe the researched group consisted of patients who had some remnant alpha-galactosidase-A activity. And cell-based assay probably indicates that they have measured alpha-galactosidase-A activity in blood or tissue samples taken from patients. Because if you stabilize enzymes with no catalytic activity in order to prevent their degradation by ER quality control system, then you will get more enzymes in lysosome with no catalytic activity. Therefore, it is more logical to use migalastat in patients with some alpha-galactosidase-A activity.
Marc Dunoyer, Global Head of GSK Rare Diseases, said in that press release: "GSK and Amicus are committed to advancing migalastat HCl as a monotherapy in Fabry patients with amenable mutations. While these 6-month data are encouraging, there is additional work to be done. We continue to analyze the 6-month results and look forward to receiving the 12-month results from this study. In addition the results of Study 012, our second Phase 3 Fabry monotherapy study, will add to the totality of our data and give us a more complete picture of the clinical effect of migalastat HCl. This study, an 18-month comparison of migalastat to ERT, with iohexol GFR as the primary endpoint, is fully recruited and due to report in 2014."
So they know. But I think, given the small number of Fabry disease patients, it is really hard to group patients with the same mutation. Lots of different mutations cause Fabry disease and there is no one dominating mutation. That's why it is not possible to segregate patients according to their mutations. Moreover, even if migalastat works in one or several types of mutations, the market capital will be really small..
Amicus is also interested in improving the half-life of synthetic alpha-galactosidase-A (Fabrazyme etc.) by using migalastat. This is reasonable because the cost of the enzyme replacement therapy (ERT) is huge, and such an approach can make ERT more affordable. In this scenario, synthetic alpha-galactosidase-A/migalastat is taken into the cell by mannose-6-phosphate receptors which are located on the cell surface (Fabrazyme is taken into the cell by these receptors). Then, this complex is transported to the lysosomes. Due to the lysosomes acidic environment, migalastat would dissociate from enzyme and it would start to do its job there. This seems more rational and more likely to succeed, because enzymatic structure variability is not an issue, it is an enzyme that works and it is all synthetic (company knows the structure, protein sequence etc).
I could not find any information about the metabolism of the Fabrazyme (and other synthetic alpha-galactosidase-A enzymes). If Fabrazyme is degraded (or denatured) due to its instability in blood (pH, ionic environment etc.), then migalastat is very likely to work (increasing the half-life of the enzyme). If Fabrazyme is degraded by a protease, we cannot predict whether the alpha-galactosidase-A/migalastat complex will be more resistant to degradation or not. If Fabrazyme is eliminated through urine, we cannot predict whether alpha-galactosidase-A/migalastat complex will be more resistant to tubular secretion or filtration in the kidney.
The only limitation for this use of migalastat would be that alpha-galactosidase-A/migalastat complex may not be captured by mannose-6-phosphate receptors due to changes in enzymatic structure. This is, however, not very likely as these receptors target mannose-6-phosphate groups on the surface alpha-galactosidase-A and a compound which binds to the catalytic site of the enzyme is not likely to impact mannose-6-phosphate groups on the alpha-galactosidase-A enzyme.
A Phase 2a Study to Investigate the Effect of a Single Dose of Migalastat HCl on Active Agalsidase Activity in Fabry Patients Receiving Enzyme Replacement Therapy
As you can see in the poster, migalastat increased plasma active alpha-galactosidase-A. This might be due to:
1- Increased stability of the enzyme made it more resistant to denaturation and/or more resistant to a possible degradation by proteases and/or more resistant to a possible elimination-secretion by kidney tubules.
2- Alpha-galactosidase-A/migalastat complex cannot be captured by mannose-6-phosphate receptors due to alterations in enzymatic structure. This is bad.
To rule out the 2nd scenario, my guess is that they have performed skin biopsies, measured alpha-galactosidase-A activity in skin tissue samples to show that alpha-galactosidase-A/migalastat complex is captured by mannose-6-phosphate receptors and is sent to the lysosomes in the cells. This way they showed that the alpha-galactosidase-A activity is increased in skin tissue samples.
My main concern is that skin also has blood vessels and an increased blood level of alpha-galactosidase-A may lead to increased alpha-galactosidase-A activity in the skin tissue samples. In order to get clear data (to more certainly rule out the second scenario), I would just wash skin tissue samples to get rid of blood (I do not know if they have done this). Ideally, I would also perform immunoflourescence microscopy to show that alpha-galactosidase-A is in fact inside the lysosomes. Either they have not posted such results or I couldn't find them, so we do not know.
Regardless, for such use of migalastat, the FDA will definitely want to see that migalastat is decreasing the need for Fabrazyme (or any other synthetic drugs) or at least that migalastat is decreasing the dosing frequency of the Fabrazyme without affecting the total needed Fabrazyme dose.
In summary, migalastat may work as a monotherapy in cases with some types of mutations, but for now we do not know what mutations they are. Even if it seems to work in several types of mutations, market capital of migalastat will be really small for that use - Fabry disease itself is a rare disease. Migalastat may also work with ERT, and it is more likely, but I would be much more confident to say that migalastat is going to work with ERT after I see clear data showing that migalastat is increasing the amount of active Fabrazyme in the lysosomes of liver, kidney or skin tissue.
Given the reasons I described above, migalastat HCI as a monotherapy trial is not likely to succeed and the market already responded to data from its migalastat HCI as a monotherapy trial (study 011). Moreover, I claim that a success will be a surprise to the market, because a p value like 0.3 is far away from 0.05. This means that p value is away from success.
Migalastat is more likely to work with ERT, but there are several unanswered questions above. If the things will continue in the right direction (some more data supporting migalastat that I mentioned above), I am pretty sure migalastat can be used in combination with ERT.
Reported in March 2013, Amicus has approximately $93MM and a cash burning rate around $17MM per quarter.
Take home message: Keep away from Amicus before the announcement of migalastat data. After the likely small crashes (small, because the market already responded to Study 011 data on 19th December 2012), you can buy Amicus to bet for migalastat-ERT combination.