Epizyme – The dark side of the biotech IPO bubble

On a standalone basis, phase I results presented by Epizyme (EPZM) for its lead program, EPZ-5676, were not as negative as market reaction implies. The company was able to dose escalate through several cohorts without serious side effects, demonstrate target modulation in humans and show mild signs of efficacy. For a typical biotech company, that would be considered a reasonable phase I data package, given the possibility to explore higher potentially more efficacious doses. Unfortunately, an early stage biotech company with a market cap of ~$1B based purely on preclinical results and hype is anything but typical.


EPZ-5676 inhibits the enzyme DOT1L and is the first inhibitor of its kind to enter clinical testing.  DOT1L belongs to an emerging class of enzymes (histone methyltransferases or HMT) which are considered a promising family of targets in cancer research. On top of the general hype around epigenetic targets, excitement around EPZ-5676 was driven by an early lucrative deal with Celgene (CELG) and more importantly, a potential biomarker for patient selection.

In cells and animal experiments, Epizyme showed that cancer cells with specific mutations (MLL rearrangement or R-MLL) are highly sensitive to EPZ-5676 whereas non-mutated cells are unaffected. This, coupled with Epizyme’s strategy to focus on R-MLL cancers gave investors the impression that EPZ-5676 will be highly active in this defined subset just like Xalkori and Zelboraf are active in their respective molecularly defined target populations.

The excitement around EPZ-5676 overshadowed the lack of clinical experience with HMT inhibitors and the drug’s unfavorable PK profile, which forced Epizyme to give it via continuous IV infusion. This is in contrast to most cancer drugs that are injected every 1-3 weeks or taken orally every day. In order to justify this problematic dosing regimen, a drug must demonstrate remarkable efficacy (Amgen’s blinatumomab is a good example).

Summary of results

The phase I enrolled 16 leukemia patients in 4 dose cohorts, 8 of whom had MLL rearrangements, which should make their disease sensitive to EPZ-5676. Of the 8 R-MLL patients, 4 demonstrated what the company defines as “treatment effects” including reduction in circulating leukemia cells in the blood, differentiation or reduction of leukemia cells in the bone marrow and resolution of leukemia related symptoms. Nevertheless, in all cases the “effects” were not clinically meaningful, did not come close to a true response and did not prevent rapid disease progression on the drug. In addition, there was no dose dependent response as “treatment effects” occurred equally across the different cohorts.


On a positive note, it appears EPZ-5676 was able to modulate its target based on a pharmacodynamics marker. Using the methylation status of a known DOT1L target, there was a dose and time dependent inhibition of the methyl mark. According to the methyl mark, Epizyme reached only 60% target inhibition, which implies that the doses are not optimal.


In summary, EPZ-5676 successfully inhibits DOT1L in humans but activity to date is limited. When given in higher doses, the drug might be the breakthrough drug Epizyme hopes it to be but the current data has very little to support this hypothesis.

When hype and reality collide

Epizyme’s case reflects the problematic side of the current biotech IPO bubble, where companies are given extremely rich valuations based on little to no clinical validation (discussed here). This is the case with Epizyme, Agios (AGIO), Intrexon (XON), Verastem (VSTM) and Oncomed (OMED). In all those cases, the underlying science is exciting and innovative but valuations factor-in a high likelihood of success in the absence of any clinical proof of concept – A problematic phenomenon in an industry where most drugs fail. As a result, there are very few attractive biotech IPOs, 2 of which are discussed here.

What I find most puzzling is the company’s decision to publish results so early in the trial’s lifetime when they know expectations are so high. Did they really think such mediocre efficacy would be received positively??? The 37% drop in Epizyme’s stock is warranted not only because EPZ-5676 failed to impress, it is a warning sign for companies who take their overblown valuations for granted and assume investors will put up with underwhelming data just because a company is in fashion. Even after today’s drop, Epizyme is still too expensive in my opinion.

I will post my ASH preview and a portfolio update on Sunday

9 thoughts on “Epizyme – The dark side of the biotech IPO bubble

  1. Ohad
    ESPR: There is a new US heart guidelines, which are negative for any drugs that are not statins, including Zetia and even PCSK9. Yesterday REGN was sold-out on the news. Strange, but ESPR as well?!?
    Don’t you think the news should be positive for ESPR, since they will cover the pts. not tolerant to statins.


  2. I expect ESPR to receive the same treatment PCSK9 companies are given. The current consensus is that outcome studies are not required for initial approval and this bodes well to ESPR as well.



  3. Same for STML, in spite of the results on SL-401, which will be presented at ASH. The stock is down to $21 from the $47 high not long ago.
    Your opinion, Ohad? Most analysts have a price target near $50.


  4. Their BPDCN abstract at ASH looks very impressive (but not a lot of new data) in terms of response. Durability may become an important factor going forward as their drug cannot be given chronically (comprises an immunogenic bacterial toxin). Market size is also an open question, the company claims incidence is 2000 per year in US+EU so even if you assume 1000 patients on drug per year and a $100k price per patient, SL-401 is a $100M drug. STML’s market cap of $266M looks attractive in that sense but it’s not dirt cheap.



  5. Hi Ohad, I’m doing some desk research into the price of small molecule orphan drugs and came across your thread here. Do you know what the price of EPZ-5676 is per patient, per year please? Thank you


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s