Tinib versus Tinib

Competition between approved and investigational drugs is undoubtedly good for patients. In myelofibrosis (MF), Incyte’s drug ruxolitinib has been the sole JAK inhibitor approved to date, primarily on the basis of the registrational trials COMFORT-I and II. Cell Therapeutics and their partner Baxter have also set sights on the MF space, having initiated PERSIST-1 and 2 to further the cause for pacritinib, their JAK inhibitor.

Incyte’s registration trials, COMFORT-I and II, enrolled patients with baseline platelet counts >100,000/uL, and had dose reduction strategies in their protocols for emergent exacerbation of thrombocytopenia. Subsequent to these trials, Incyte has initiated additional studies that have accrued results from ruxolitinib dosing strategies to accommodate patients with platelet counts below 100,000/uL. To wit, the prescribing information for ruxolitinib includes dosing strategies for patients with platelet counts between 50,000 - 100,000/uL. Nonetheless, ruxolitinib was approved on the strength of its treatment effect in patients with >100,000/uL platelet count. The development strategy for pacritinib is focused on this point, and pays special attention to MF patients with baseline platelet counts below 100,000/uL. The first large study for pacritinib is PERSIST-1, with topline data released in March and more details released at ASCO.

Table 1 provides a comparison of spleen responses for the various pacritinib and ruxolitinib trials to date. When possible, the data have been split up into platelet subgroups to facilitate comparisons. No such comparison is perfect, but absent head-to-head data, it’s a reasonable exercise.

View fullsize

Based on the results to date, some trends are visible. First and foremost, for patients with platelets >100,000/uL, ruxolitinib provides better spleen response. This population is, arguably, two thirds of the MF population. Conversely, for patients with platelet counts below 50,000/uL, ruxolitinib yields to pacritinib. The package insert directs caregivers to hold ruxolitinib administration if platelets fall below 50,000/uL, providing pacritinib with an open opportunity. As for the remaining patients with starting platelet counts between 50,000 and 100,000/uL, it is difficult to suggest superiority of one drug over the other. Cumulatively, 38 of 134 patients (~28%) treated with ruxolitinib have had responses as compared to ~11% for pacritinib. Nonetheless, these ruxolitinib data are from open label or expanded access trials, and would likely see a step down in a controlled, randomized setting.

Next up for pacritinib is PERSIST-2, a trial that is enrolling ~300 MF patients with platelet counts less than 100,000/uL. These patients are split into 3 arms: pacritinib 400 mg QD, pacritinib 200 mg BID, and a best available therapy arm that, unlike PERSIST-1, may include ruxolitinib. If the best available therapy arm primarily features ruxolitinib use, and the spleen responses remain similar to those outlined in Table 1, there is little chance of the pooled pacritinib arms besting ruxolitinib. To be more precise, the primary endpoint of PERSIST-2 takes into count both spleen responses and patient reported Total Symptoms Scores (TSS). In COMFORT-I and PERSIST-1, approximately 50% and 25% of ruxolitinib and pacritinib treated patients, respectively, noted a 50% or greater reduction in TSS. In effect, a head-to-head trial favors ruxolitinib for the majority of the MF population. Unless PERSIST-2 enrolls a large percentage of patients with starting platelets <50,000/uL, pacritinib is unlikely compare favourably if the best available therapy arm has extensive ruxolitinib use.

The following consideration for this setting is safety. Ruxolitinib has demonstrable impacts on both red blood cell and platelet counts, requiring dose reductions or transfusions. Pacritinib’s ability to be dosed for patients with platelets <50,000/uL likely confirms its advantage in eliciting less thrombocytopenia. At this time, it is difficult to compare anemia data across the trials, in part because the transfusion dependence criteria were divergent:

• COMFORT-I: IWG criteria wherein “Baseline transfusion dependence was defined as the receipt of ≥2 units of RBC products in the 4 weeks prior to randomization, and on-study transfusion independence was defined as the absence of transfusions for any period of ≥8 weeks.”

• PERSIST-1: As per the press release, defined as “patients treated with pacritinib who were severely anemic and transfusion dependent – requiring at least six units of blood in the 90 days prior to study entry.”

By the above criteria, 41.2% vs 46.9% in the active and comparator arms in COMFORT-1 and 25% vs 0% in the active and comparator arms in PERSIST-1 became transfusion independent. 

In all, the results from PERSIST-1 suggest that pacritinib is a less potent member of the JAK inhibitor class, signaled not only by its modest efficacy, but also by its more lenient adverse event profile. Going forward, pacritinib’s dataset in PERSIST-2 will be pivotal in determining its competitiveness in the MF space. If the best available therapy arm in PERSIST-2 is able to accommodate a high percentage of patients receiving ruxolitinib, the primary efficacy endpoint is unlikely to be met.

Incyte 4Q 2014 Notes

Just a few notes from the latest quarterly conference call. For the company, it seems 2015 will be a year to set up an eventful 2016 rather than being an exciting year in and of itself.

And so, in no particular order:

1. On the PV front, their estimates for the addressable population appears to be roughly equal to the myelofibrosis population of around 25000. Nonetheless, the expectation is that take-up will be slow as this patient population is ripe for a watchful-waiting approach rather than a quick-get-them-on-this-drug approach. Therefore, strong sales in PV would be an upside to their 525-565 million 2015 net product revenue. I don’t expect strong sales in that indication to materialize.
2. The FGFr inhibitor, INCB54828, is an FGFr1, 2 and 3 inhibitor.
3. Of the two PI3Kd inhibitors, they’re calling INCB40093 highly selective whereas INCB50465 is being termed a Delta inhibitor. They note that in the preclinical models, the profile of INCB50465 is clean enough to dose at high levels, but the biological relevance of dosing to 99%+++ inhibition of the target remains unclear to me.
4. It was noted that there will be no milestone payments received from Eli Lilly during 2015. With the remaining phase 3 trials of baricitinib in RA due to read out this year, the comment suggests that there will be no NDA filing in 2015 (provided results merit one). Perhaps not too surprising, but nonetheless suggests no baricitinib related revenue until late 2016, if at all.
5. They will initiate an INCB39110 + Gemzar + Abraxane trial in first line pancreatic cancer later this year. The only reason this was interesting is for the lack of mention of any CRP related enrollment or stratification. Others will note that there are currently two Janus trials testing ruxolitinib in pancreatic cancer using CRP levels as part of the enrollment criteria. Perhaps CRP wasn’t mentioned for this INCB39110 trial because it is still at an exploratory stage. Nonetheless, it merits attention.

Notes From The Incyte Presentation At The 2015 JP Morgan Conference

Some highlights from the recent update by Herve Hoppenot at the JP Morgan Healthcare Conference:

- The company anticipates results from the Janus 1 and 2 trials to be available in 2016. As a recap, these trials are focusing on the use of ruxolitinib with capecitabine in patients at different stages of pancreatic cancer. These trails are pursuing a prespecified patient population based on the exploratory phase 2 data suggesting a relationship between CRP and ruxolitinib treatment. Whether that relationship is causal or correlative yet unrelated remains to be seen. Typically, markers such as CRP are reliably indicative of disease but have not been great prognostic companions for treatment subgroups.

- The controlled phase 2 trials of combination ruxolitinib therapy in colorectal, nonsmall cell lung and breast cancers continue. Data are expected in 2016.

- In an attempt to protect / expand their lead in the JAK space, the company is pursuing development of its selective JAK1 inhibitors, INCB39110 and 52793. The latter is the newer addition to the clinic and, compared to 39110, has ~5x the increased selectivity for JAK1 vs JAK2. The rationale behind this is to minimize the JAK2 inhibition of a mixed JAK1/2 inhibitor like ruxolitinib, thereby attempting to minimize myelosuppression. INCB39110 is already in the clinic in combination studies for nonsmall cell lung and pancreatic cancers. The more interesting INCB39110 trial is in B-cell malignancies where it is being tested in combination with Incyte’s internal PI3Kd inhibitor INCB40093. The newer molecule, INCB52793, is just entering dose finding phase I studies.

- Addressing the recent Agenus deal, the message was as I suspected: this is mostly a tools deal. Incyte believe that their internal engine is strong on the small molecule front, but they see a relative weakness in biologics. The Agenus platform will allow them to screen for fully humanized antibodies against targets of their choosing. Not too surprising, since biologics are a very relevant and meaningful platform and Incyte had not yet shown much acumen in this regard.

- On the PI3Kd front, Incyte has two candidates: INCB40093 and a more recent compound, INCB50465. The latter compound is roughly an order of magnitude more potent in preclinical assays. However, the benefit of this increased potency is unclear. Interestingly, although their slide deck showed preclinical data rationalizing the combination of INCB40093 and INCB39110, no such data were shown for INCB50465. Therefore, it remains uncertain if the increased potency has a material benefit in either combination or monotherapy use.

- Two new molecules, INCB54828 and 54329 are FGFR and bromodomain inhibitors (see this link for a general background), respectively. For INCB54828, no data were broken down for its relative selectivity among FGFR1, 2 or 3. For the bromodomain inhibitor, the requisite mouse model data were shown… but as usual, such data are necessary for advancement but not overly useful in predicting clinical efficacy. These compounds do appear to be moving ahead, however, as the bromodomain inhibitor was noted as entering the clinic imminently.

- A wild card for Incyte this year will be the rheumatoid arthritis program. The first phase 3 readout occurred late last year and was positive. Readouts for the remaining 3 trials are expected this year, and of particular interest to me is the RA-BEAM trial that will  include structural endpoints and include Humira as a comparator. The biologic market for RA is very large, but to date the small molecule competitor, Xeljanz (tofacitinib), has not made much of a dent. I think Incyte’s baricitinib has advantages over Pfizer’s tofacitinib, with the latter being a JAK3 inhibitor and the former a mixed JAK1/2. Cross-talk between JAK1 and 3 is possible, and I think a slight safety advantage may emerge for baricitinib versus tofacitinib. Whether this combination of efficacy and safety makes a dent on biologics remains to be seen. My suspicion is that it does not in the near term (1-5 years), but for a company the size of Incyte, obtaining revenue from the RA market would be significant.