I've written in the past about a class of proteins known as heat shock protein inhibitors as a targeted anticancer therapy, and there are a few that have been in clinical trials, including IPI-504 from Infinity Pharmaceuticals and STA-9090 from Synta Pharmaceuticals, with others also in development. The only trial that has actually been the subject of a completed clinical trial in NSCLC is IPI-504. Dr. Lecia Sequist from Massachusetts General Hospital led this multicenter trial (10 participating sites, including Moffitt Cancer Center: Dr. Gray is one of the study authors) of IPI-504 administered at 400 mg/m2 IV on days 1, 4, 8, and 11 every 21 days to 76 patients who had previously received an epidermal growth factor receptor (EGFR) tyrosine kinase inhibitor (TKI). In fact, these patients had generally received many prior treatments, with a median of 4 prior lines of therapy among the participants, with nearly all having already received an EGFR inhibitor. In addition, 45% of patients were never-smokers and 63% were women, so this is not a typical cross-section of the NSCLC population.

Nearly all patients were tested for the presence of an EGFR mutation, but they were eligible whether they had evidence of an EGFR mutation or were EGFR wild type (no mutation). One hypothesis of the investigators leading the trial was that IPI-504 could be an active agent in patients with an EGFR mutation, based on some preclinical lab-based research.

But the results didn't turn out quite as the investigators had anticipated. The response rate (RR) in the entire trial population was 7% (5/76 patients), with a breakdown of 10% (4/40) for the patients with EGFR wild type, and 4% (1 of 28) with an EGFR mutation. While these results were below the threshold RR of 20% the investigators were hoping to see, they also did testing for an ALK rearrangement and KRAS mutation, along with a battery of other molecular markers when tissue was available. They found that while there was no clear association of better or worse outcomes in patients with a KRAS mutation, but they found the intriguing result that two of the three patients they identified with an ALK rearrangement had significant tumor shrinkage consistent with our definition of an objective response. Here are the "waterfall plots" of response vs. progression, with bars on the right extending down from the horizontal line reflecting tumor shrinkage (and longer downward bars corresponding to more shrinkage), while the upward bars on the left correspond to tumor progression (with the length upward corresponding to more growth):

(click on image to enlarge)

In the figure above, the top panel represents the breakdown in patients by EGFR mutation status, the middle panel represents the breakdown by KRAS mutation status, and the bottom panel is by ALK rearrangement presence or absence (fewer patients had tissue available for this test, which hadn't been identified as a high priority test when this trial was developed).

This agent was generally well tolerated, with most side effects in the mild range, and the most common side effects being nausea, vomiting, diminished appetite, fatigue, diarrhea, cough, and muscle/joint aches. One interesting side effect is that a third of patients had transiently purple urine due to excretion of a metabolite of the drug. While about half of patients had an abnormal elevation in their liver function tests, this was moderate to severe in less than 10% of the study population. However, my understanding is that liver function test elevations have been more frequently abnormal in other trials and different study populations, and that this is limiting the enthusiasm and pace at which the company develops the drug further.

To my knowledge, there are no subsequent plans in the works to follow this up and develop IPI-504 specifically for the relatively small population with an ALK rearrangement, but if I happened to determine that a patient has an ALK rearrangement and developed progression on an ALK inhibitor like crizotinib, or if they couldn't pursue a trial with crizotinib, I'd be very interested in seeing whether they might be able to enroll on a trial with STA-9090 or another heat shock protein inhibitor.

This work also provides an illustration that there can be a gulf between what we might expect based on the lab work and what actually happens in the clinic -- IPI-504 definitely has some activity, but it was actually greater in the patients who didn't have an EGFR mutation. Even more importantly, it illustrates the benefit of looking at molecular subgroups, because this is a provocative lead that we would have definitely missed if the investigators hadn't looked for the presence of an ALK rearrangement and detected the pattern.