By Neil Canavan
There was much to the good at this year’s ASCO – which is not unusual – but there was some news that was notably not-so-good.
For an example to the (very) good was data presented by Loxo Oncology for their tropomyosin receptor kinase inhibitor (TRK), larotrectinib, with the touted outcomes for this novel agent joining the data march towards near-certain approval as the first ever targeted tumor agnostic therapeutic.
That said, this particular journey has been long underway, so the clinical improvements reported at ASCO were more or less expected.
What was widely unexpected were fails from the field of cancer immunotherapy. For those examples, read on.
Whack, Clang, Crash
Whack was the sound of JTX-2011, an immuno-oncology (IO) drug in development at Jounce Therapeutics, hitting a statistical wall. This was particularly surprising for two reasons. Number one, the list of Jounce’s scientific founders reads like the top acts at the legendary concert, Woodstock: James Allison, Drew Pardoll, Pam Sharma, Bob Schreiber, Tom Gajewski… These are the Hendrix, Santana, Joplin, et al, of the IO world.
Second, the science behind the drug is particularly compelling, as were the preclinical evidence of the drug’s activity.
Yet – WHACK.
To elaborate: JTX-2011 is an agonist of ICOS (Inducible CO-Stimulator of T cells). As reported by Dr. Timothy Yap, University of Texas MD Anderson Cancer Center, JTX-2011 was tested in the ICONIC study in combination with the checkpoint inhibitor, nivolumab, in patients with advanced cancers.
The rationale for pursuing this regimen is that previous testing showed the ICOS receptor being upregulated on T cells – not only after target/antigen engagement – but also by the administration of checkpoint inhibitors, such as anti- CTLA-4 (discovered by Jounce founder, Jim Allison). Further, ICOS downregulates the immune-suppressing activity of T-regulatory cells.
“Different groups have shown that ICOS is upregulated by a variety of different agents,” said Dr. Yap. Given its various mechanisms of action, “this makes (ICOS) an ideal target for combination therapy.”
In ICONIC, an adaptive phase I/II study, relapsed/refractory patients with a variety of tumor types were treated with
JTX-2011 +/- nivolumab (N=164: phase I, n=71; phase II, n=93).
Regarding patient selection, “There was no enrichment for (expression of) ICOS in phase I, but in phase II, patients needed a (ICOS) score of at least two.”
The results for ICONIC – politely put – were mixed.
For the phase I cohort, a 20% tumor control rate was observed with JTX-2011 monotherapy, which is meh, at best. For the phase II cohort, patients taking the combination achieved disease control rate of 29%. Again, modest. That said, for both cohorts the discontinuation rate was huge – 75%, and 69%, respectively.
Dr. Yap attributed the discontinuations to the advanced disease state of the patients, saying, “this may potentially confound the accurate assessment of clinical activity.”
Famed founders, solid science – what went wrong? Specifically, it is hard to say, but in general, one can assume there is some science here as yet unknown, and so for now, it’s back to the white board.
Another IO agent with some very elegant science behind it is epacadostat, manufactured by Incyte, an inhibitor of indoleamine 2,3 diosygenase (IDO1), a signaling moiety thought to be involved in several types of immuno-suppression.
“Because IDO1 is an enzyme that catabolizes tryptohan to kynurenine, increased kynurenin in the tumor micro-environment decreases T-effector cells or CD8 cytotoxic cells, and natural killer cells,” explained Dr. Georgina Long, MD, PhD, of the Melanoma Institute of Australia.
“It also increases proliferation of (suppressive) T-regulatory cells, myeloid-derived suppressor cells, and tumor associated macrophages. Taken all together… this inhibits immunogenic cancer cell death.” Therefore, targeting IDO should have great potential, and several companies were attempting to explore it (note the past tense).
In the ASCO-reported phase III, ECHO-301/KEYNOTE-252 study, epacadostat +/- pembrolizumab (anti-PD-1) was tested in patients with unresectable, or metastatic melanoma (N=706).
The proposed combination was justified by preclinical synergy with PD-L1 blocking agents, and further, phase I/II results in treatment-naïve melanoma patients showed a response rate of 55%, and a median progression-free survival of 22.8 months.
However, results for this phase III study – after a median follow-up of 12.4 months – showed exactly ZERO added benefit with the addition of epacadostat to pembrolizumab. No, really. CLANG – like getting your bell rung. Lights out. Zero.
“Additional analyses are underway to better understand these findings,” said Dr. Long, but in the meantime, the ECHO trial was stopped, and, sensing a drug-class related disaster, other IDO investigating companies pulled the plug on any number of programs planned, or already underway.
Issues of the basic science of IDO science aside, Chuck Drake, MD, PhD of New York Presbyterian Medical Center offered a slightly positive spin to this wipe out. “The fact that these (response) curves are so beautifully overlapping suggests that at this dose, and this schedule… that the enzymatic pathway was not inhibited in the tumor microenvironment.” Meaning, science good – but the clinical translation? Not so much.
This one is a head-scratcher. After Nektar Therapeutics wowed the crowd with a readout at the November 2017 Society for Immunotherapy of Cancer (SITC) conference – so much so that BMS bought them shortly thereafter for $1.85 billion – the really-not-at-all-bad data reported at ASCO sent the stock price crashing (down 40% in a matter of minutes).
Why? First the facts.
NKTR-214 is prodrug version of the immune-boosting, interleukin-2 (IL-2), a cytokine that preferentially activates and expands effector T cells, and NK cells over suppressive T regs, and promotes expression of PD-L1. These dynamics of IL-2 are the mechanistic rationale for combining NKTR-214 with the checkpoint inhibitor, nivolumab.
Interestingly, Steve Rosenberg at the NIH first used recombinant IL-2 (not the prodrug highlighted here) in 1984, and he observed some remarkable outcomes. In fact, a handful of these once terminally-ill patients are still alive today. However, naked IL-2 proved to be whoppingly toxic – every patient that Rosenberg treated wound up in the ICU – so the approach was never widely adopted.
Thus, the prodrug formulation. “The unique drug design of NKTR-214 provides a safety profile that avoids that rapid and hyperactive immune stimulation and inflammation that is usually associated with (severe) toxicities of giving high-dose IL-2,” said PIVOT investigator, Dr. Adi Diab, of MD Anderson Cancer Center. “The prodrug design also allows for administration every three weeks, which is very convenient for the patients.”
The trial, first reported at SITC, and then ASCO, is PIVOT, an ongoing, open-label, phase I/II study of NKTR-214, plus nivolumab, in patients with advanced solid tumors, including melanoma, renal cell cancer, lung cancer, urothelial cancer, and triple negative breast cancer (N=38, phase I; N=94 evaluable, reported here, with target accrual of N=330).
Here are the clinical results comparing the SITC data with the ASCO update: “The overall response rate (ORR) initially observed for melanoma has been maintained (64%), with patients having a deepening of response – wether they were PDL-1 positive, or PDL-1 negative,” said Dr. Diab.
For renal cell carcinoma, the ORR was 46% at SITC, and 71% at ASCO. For the lung cancer cohort (N=5) the ORR was 60% at ASCO (SITC data not reported) – and all five responding patients were deemed PD-L1 negative.
For new patients in the ongoing dose expansion cohorts, the ORR for melanoma stands at 50%; for renal cancer, 46%; and for urothelial cancer, 60%. “Again, many of these patients were PD-L1 negative, with several of these patients having 100% response in target legion,” said Dr. Diab, further noting that some of these patients converted to PD-L1 positive status while on therapy.
So, overall, a fairly decent showing for Nektar – after which the stock collapsed. So… what happened?
Some, like Heather L. McArthur, MD, MPH Cedars-Sinai Medical Center, were impressed with the approach and puzzled by the fall.
“These were some dramatic waterfall plots across three different tumor types – in melanoma, renal cell carcinoma, and urothelial cancer,” said Dr. McArthur. And, although cross-study comparisons are discouraged in clinical trials, she felt compelled to do so, given the negative vibe in the air.
“The data we just saw in metastatic melanoma in the NKTR 214/nivo combination had a response rate of 50%, (compare that) with two prior nivolumab studies showing responses of 40%-44%, which suggests that (this drug) maybe has successfully pushed the bar forward.” Especially in light of the fact that those small historical increments in overall responses translated into dramatic survival benefits. “For example with ipilimumab as compared to chemo – there was a small overall response rate benefit, but a dramatic one, two, three, four-year survival benefit. So this kind of overall response rate might actually translate into something very impactful.”
Meaning, investors should be more circumspect.
Alternatively, as concisely put by Axel Hoos, senior vice president of Oncology R&D at GlaxoSmithKline Pharmaceuticals (GSK), investors need to learn some math. “There are those who do not understand how to interpret the statistics of small numbers.” Meaning, you cannot assume trends in improvement from a tiny sampling.
Asthika Goonewardene, lead biotech analyst for Bloomberg Intelligence, concurs. “People were too stupidly optimistic that the ORRs would go back up versus what was in the abstracts – a poor understanding that an N of 7 is less predictive and has more room to get exaggerated than an N of 25.”
Finally, the negative impression – flawed or not – was made worse by the timing. Nektar’s data set was not available until the very last minute, as opposed to days or weeks before the meeting. The sudden data drop may have sent the unprepared scurrying towards the door.