On November 14, 2013, Epizyme (NASDAQ:EPZM) presented results of the Phase I trial data for EPZ-5676, a small molecule inhibitor of DOT1L.
EPZ-5676 was developed for the treatment of acute leukemias with MLL rearrangement (MLL-r). This rearrangement abnormally activates DOT1L, which leads to an abnormal methylation and broad expression of genes causing leukemia. Epizyme has previously published extensive preclinical work showing the effect of the inhibition on leukemic cells.
Let's briefly recapitulate what Epizyme did and why, because this will give us a better perspective on published results. The purpose of a Phase I trial is to establish a toxicity profile of a new drug. One of the measures of the toxicity is so-called Maximum Tolerated Dose (MTD), which is the last safe dose that does not cause severe toxicities in patients. Toxicities are very carefully monitored and graded. When the toxicity reaches unacceptable levels, the investigator has to stop the dose escalation or even stop the trial, if the toxicity is life threatening. Patients enrolled in a Phase I trial exhausted all therapeutic options such as chemotherapy, allogenic stem cell transplant and a higher Phase (II or III) trial. The last option for these patients is a Phase I trial. The purpose of any Phase I trial is to establish whether the drug is safe, tolerable and to establish the safe dose for future trials with healthier patients. Secondary goals include pharmacodynamic (PD) and pharmacokinetic (PK) measurements.
EPZ-5676 is a new class of a drug and the company designed the trial very conservatively. Investigators enrolled four cohorts of patients and treated them with four different doses 12, 24, 36 and 54 mg/m2 followed by a week without the drug to evaluate for toxicity. At the present point, the fifth cohort is enrolling. It is important to note, that the higher dose cohort cannot start until it is clear that the previous dose was safe. If the cohort reaches the dose that causes toxicity, the investigator stops enrolling and has to go back to a lower dose. The trial population was composed of a mix of patients with MLL-r , which is supposed to be targeted by EPZ-5676 and without the MLL-r, which is not supposed to be affected.
What were the results?
Four cohorts completed their treatment and there were no dose limiting toxicities. There were no adverse effects that would warrant stopping the trial or reducing the dose. Only one patient had a drug related neutropenia, which is a Grade 3 adverse effect. One patient had brain bleeding due to the progression of the disease, which was not related to the drug. EPZ-5676 PK data showed a nice linear increase of the systemic drug while the dose and PD data showed effects of the drug measured by methyl marks. When the drug was discontinued, the methyl mark bounced back. The methyl mark is a good marker of drug activity, because the drug is an inhibitor of a methyltransferase.
Interestingly, Epizyme's investigators also observed treatment effects in 4/8 patients with MML-r and no effect in patients without MLL-r. It is interesting to take a close look at the effect in some patients. EPZ-5676 treatment showed a marked reduction of the blast count and maturation of leukemic cells. This was exactly predicted by Epizyme's preclinical research. The company also noted a resolution of some clinical symptoms such as fevers, cachexia (body mass wasting) and leukemia cutis (leukemic skin ulcers).
To realize how remarkable these results are, it is important to realize what EPZ-5676 did treat. Any cancer in a Phase I trial is a cancer that resisted numerous previous therapies. Patients in this trial had a median number of 4 previous therapies. These previous therapies damaged the cancer, but the cancer cell mutated and became resistant to all therapies. It would not be a surprise if in these cancer cells, DOT1L would not be a major genetic driver anymore and would not respond at all to the targeted therapy. This is precisely the reason one does not expect much of a therapeutic effect in a Phase I.
However, based on these great results from the Phase I, the company is progressing with the expansion study and continues escalating the dose.
These results mean three important things. First, the drug does not cause severe toxicities and does not kill patients. Patients did not even reach the drug's MTD, so the actual therapeutic dose might be much higher. For a novel class of a drug, that's pretty great. Some of us might remember HSP inhibitors and MMP inhibitors and associated toxicities. Second, the effect is proportional to the dose and can be measured. Third, the preclinical science has been confirmed in humans. Patients who are supposed to respond (with MLL-r) show a measurable clinical response, while patients without MLL-r do not show any response.
The second and the third point are particularly important when we think what the ultimate goal is: FDA approval. The FDA does not approve drugs that show random effects, which cannot be quantified.
So why did the stock fall 40%?
When I saw the Phase I results, I was pleasantly surprised. However, I did not expect much of a price effect, maybe a slight move towards $40. Even a great Phase I result is usually not a reason for a big move up. But I certainly did not expect a 40% fall. With all honesty, I have no idea what happened. Nevertheless, I will try to speculate about possible reasons and suggest what it means for potential long-term investors.
First, let me say what I think would be some very good reasons to sell after a Phase I and what could lead to a massive sellout and drop of the price: 1) the trial stopped because some patients died, 2) the trial did not complete the dose escalation because patients reached dose limiting toxicities very early, 3) the drug caused acceleration of the disease, 4) although we usually do not expect much of an effect in Phase I, having a zero effect of the drug against the genetically defined target would be concerning.
But none of that happened, so why the massive drop in price?
OK, here are the reasons I could think of, leaving out a paranoid or conspiracy possibility. If anybody else knows about something I missed either from the data or the conference call, please let me know in the discussion.
1) Technical reasons. Someone made enough money buying up shares between $20-30 and decided that waiting for a big payout is too long or risky. Maybe even thought that these good Phase I data are going to serve as a buffer. However one sell led to another and that caused a bunch of stop loss orders to kick in and the stock spiraled down. A 40% drop causes a bit of a nervous itch even in the face of the calmest person so not surprisingly people are freaked out and afraid to get back in. However, I do not believe that the correction happened because the stock was overpriced. In my opinion, there is no such thing as a fair stock price in a starting biotech company without any approved drug. If the company makes a successful drug, the price will multiply; if not there is no value.
2) MTD not reached. I would not come up with this one unless I saw it published by the MT Newswires here: EPZM MT Newswires. Whoever sold shares because the MTD was not reached, should never invest in a biotech stock again, because he/she does not understand the most basic rules of clinical trials. The MTD is the highest dose you can give to patients without causing toxicity. A high MTD means that you can give enough of the drug to hit the cancer without harming the patient. A low MTD is bad news, because the patient will get sick from the drug before a cancer cell will notice it. Again, the trial did not reach its MTD, because there was NO TOXICITY that would establish the MTD. So investigators can keep increasing the dose. Eventually, the drug's MTD will be reached. Remember, the difference between a drug and a poison is in the dose. Nowhere else is this more true than in cancer therapeutics.
3) Only 4/8 MLL-r patients responded. Again, the fact that anyone in the trial cohort responded is remarkable. And the response was not minimal. Although the Phase I numbers are small, my bet is that patients without a measurable response might have developed other driver mutations in addition to DOT1L. Higher phase trials are necessary to establish the therapeutic benefit, but one can feel more confident that healthier patients will respond at least at this level.
4) None of the MLL-r negative patients responded. This reason was suggested in online discussions as one of the possibilities for the panic. That maybe someone expected the drug to work in non-MLL-r leukemias as well. I have no idea whether someone had this thought or not. However, it made me think what would I have done if, for example, the drug had an effect also in 4/8 non-MLL-r leukemias. The stock price might have even moved higher. But I know that I would pull myself out of the company. You might think that I am crazy. If the drug works in more patients, even in patients who are not expected to respond that should be great, right? Wrong. Remember, that each Phase has its defined goals. The goal of a Phase I is mostly the toxicity but also serves as a proof of the science. A drug can go all the way to a Phase III and then fails. And there are usually two reasons: a flawed science and a poorly defined patient population. It does not matter how hard the company tries. If the science is flawed, sooner or later, it will catch up with the company and the drug will fail. Epizyme presented the drug as a DOT1L inhibitor. If patients without the mutation responded, the company has to find the scientific reason why it happened.
Otherwise it is possible, with such a small number, that it is just a random nonspecific effect. If some patients with MLL-r respond and some patients without MLL-r respond, that basically equals a chance if there is no scientific reason. Maybe it was a placebo effect or maybe a disease just took a little break in some patients. With small numbers you just don't know. Many drugs show random effects in early stages, but if there is not a strong science that explains the response, the drug will eventually fail in later clinical trials. The second reason is similar. The FDA hates drugs that work randomly and does not approve them. You might truly save the lives of 3 out of a hundred patients with a disease. However, if you cannot identify potential responders, the drug will not get approved.
In summary, Epizyme showed that EPZ-5676 works as predicted by preclinical work. The study showed that the drug has minimal toxicities and the company is moving to the next stage in clinical development.
I think it will be exciting to see the work presented at ASH and I will try to write an analysis of the work presented after the meeting.
Anyway, the recent events show again that stocks are risky, and this is particularly true for early stage biotech stocks. This article is not meant as a recommendation; you should do your own research and consult with professionals before investing in stocks.