Agios Pharmaceuticals, Inc (AGIOS) – Mitapivat

Forecasting Dashboard

< Return to POC2 drug list



Please add insights, questions and comments here and respond to others’ questions / comments.

You can also see the comments your peers have submitted in their initial assessments in the “Forecasting and relevant links” tab.

Please do not disclose information about your identity or reference years/types of expertise. Those actions contribute to authority bias and decrease the predictive power of collective intelligence.

If you press the subscribe button, next to the email icon below, you can receive an email alert when people post in this drug’s discussion.

4 3 votes
How valuable was the briefing?
Notify of
Oldest Most Voted
Inline Feedbacks
View all comments

Here are the current EPO and Aranesp labels.



I was mistaken about whether transfusion frequency was used as a primary approval endpoint (the hazards of trying to read these quickly during a discussion) — some studies looked at mortality, but others did look at transfusion frequency and hemoglobin level.

It looks like most (if not all) of the studies in CKD patients described in these labels were not placebo controlled, and compared lower to higher doses of drug (or more vs less aggressive treatment of hemoglobin level).

Some of the studies in cancer patients were placebo controlled: e.g. Studies C1 and C2 (Epogen label), and Study C1 (Aranesp label). Placebo controlled studies are also described in the Epogen label for HIV and surgical patients. In cancer and HIV patients, the studies show significant reduction in transfusions required; in surgical patients there was significant reduction in transfusion for patients with a pretreatment Hb of 10-13 g/dL, but not in the 13-15 g/dL stratum or in the <10 g/dL stratum (too few patients in the latter case to make a determination).

Mark asked me to summarize a conclusion from these labels. The question had been whether transfusion frequency has been used as an approvable endpoint before. The answer is “yes”, but since this is a completely different patient population, we can’t necessarily extrapolate how feasible this endpoint will be in PK deficiency patients.

If the baseline transfusion frequency in the trials for these drugs were similar to that for the PK deficiency population that Agios has recruited, then these trials might be of some use for extrapolating feasibility of the endpoint in PK deficiency patients. But the TREAT trial, for example, did not report the transfusion frequency of the study subjects prior to the trial, and in CHOIR they excluded patients that required “frequent” transfusion — which 6x/yr would probably count as (this is the minimum frequency required for inclusion in ACTIVATE-T).

Here is the data from the Forma SAD/MAD presented last month –

The primary endpoint of the study is a bit concerning, as the company ( has stated previously that no standard for transfusions exists amongst the community and that unless you have at least 6 per year, it is unlikely that you’ll see consequences. Also, due to the small number of patients likely to be enrolled in the study, the statistical power will only be high enough to detect a small difference (~30%) and while there aren’t other options on the market, it makes me wonder whether that is clinically meaningful; that is, whether that translates into QOL outcomes for the patients suffering with disease.

One more comment on MOA: Is this drug only affecting mutant PK or does it also impact wild-type? Let’s assume for the sake of argument that its primary benefit is on the mutant PK. Then in the current circumstance, individuals with regular transfusions have two populations of rbc’s (wild-type transfused and mutant), and need to have regular transfusions as the wild-type ones die off. The schedule would basically be based on the lifespan of WILD-TYPE rbc’s. In order to lower the frequency of transfusions meaningfully, there would likely need to be a substantial increase in the mutant cell lifespan/activity to the point where they can provide a meaningful percentage of oxygen-carrying by themselves. If the baseline was low, this make be a very substantial increase to make any difference.

If you read the Phase 2 results,, it is mentioned that selection was based on a balance amongst the most common mutations witnessed in patients with PKD. I see your point, though. I don’t know the baseline for RBC life in people with PKD, but wild-type is 60 days. I believe that is true.

In the ACTIVATE trial (,do%20not%20receive%20regular%20transfusions), this product achieved its primary endpoint of a significantly significant difference between drug and placebo arm in raising blood hemoglobin by at least 1.5 g/dL. Specifically, 40% of patients on drug had an increase and 0% of placebo. Pretty good evidence that the drug is active.

The differences between that trial and the current trial are as follows:

1) Population in the original trial was in those infrequently transfused (less severe), while this population is regularly infused (more severe)
2) Original trial was placebo controlled, this is single arm
3) Original endpoint was increase in hemoglobin, this trial is decrease in transfusions after optimizing dose

As I understand the MOA (,enrolled%20nine%20patients%20to%20date), this product works by activating existing PK activity. If the baseline activity is much lower in more severe patients, there may not be sufficient baseline activity for an activator to be useful. That only 40% of less severe patients responded is a concern that this product may only be useful in a subclass of patients, potentially those with higher baseline activity.

The endpoint for this trial is new and untested. Its variability is potentially high, and it is unclear that it would be acceptable to the FDA in a single-arm trial, particularly a small one (about 27 participants). The team does not list a regulatory scientist on their leadership page and their prior experience is mostly in cancer.

Particularly since this is a single-arm trial, the success level is critical. It must be a very substantial reduction to be believable and given that there are reasonable alternatives (transfusions), there needs to be a good case if this reaches the market for payers to reimburse.

What is “success” in this trial that we will be voting on?

Excellent points about baseline activity…..need enough substrate for mechanism to be useful. Also good insight about no known regulatory person, however, not unusual to use consultants for same. I would hope this company would have had EOP2 visit with FDA and discussed registration study approach?

from a report: The ACTIVATE-T study of regularly transfused PK deficiency patients has a primary endpoint requiring at least a 33% reduction in transfusion burden from baseline at 24 weeks. Bowden said the study is open-label, rather than placebo-controlled, because Agios worried it would struggle to enroll even 20 patients for the study, although it ended up with enrollment of 27. Randomizing some patients to placebo in such a small trial might not have yielded useful efficacy or safety data, the exec said.

Given the small patient pop/high unmet need, FDA would likely accept that endpoint. Assume for our purposes, we need to predict meeting, or not, that 33% reduction in transfusions.

Ive read the synopsis for the phase II study ( My first real learning about this potential product. Definitely warranted progressing to phase III, seems plausible proof of hypothesis. Things I want to know more about is off target effects (aromatase inhibition) and reliability of genotype predictability. Things I’m very encouraged about is large proportion of participants in the extension phase as well as the seeming transience of adverse effects.

Forecasting and relevant links

Forecaster Actions

Here are what your peers have submitted so far:    Summary of Initial Asssessments

Please submit your initial assessment by Dec 28th by clicking “Initial Assessment” below:

Initial Assessment

Supporting Materials

Agios Website