May 2018 Edition Vol.11, Issue 5

AACR 2018 Roundup: The Practice-Changing, and the Potentials

By Neil Canavan

For proof that the annual meeting of the American Association for Cancer Research (AACR) is now a mini-version of that oncology colossus, ASCO, consider that two big pharma entities presented pivotal readouts – back to back plenary talks, no less – of two, phase III, practice-changing, non-small cell lung cancer (NSCLC) clinical trials with combination immunotherapy – and this at a meeting once known for smears on gels, cells on heat maps, myriad mouse data, and only a mere spritz of phase I data points.

Welcome to the main meal.

Practice-changing lung data

Keynote-189: In this phase III trial, presented by Leena Gandhi, MD, PhD, of the Isaac Perlmutter Cancer Center, NYU Langone Health, the chemotherapy doublet of pemetrexed and a platinum agent, plus the PD-1 inhibitor, pembrolizumab, was pitted against the doublet alone, in a setting of first-line use in patients with advanced NSCLC (N=616).

Randomized in a 2:1 fashion, Keynote-189 demonstrated superior efficacy in overall survival for the immunotherapy combination compared with chemotherapy alone. “There is a clear difference,” said Gandhi. The hazard ratio for death was 0.49 (p<0.00001), meaning, the risk of death within the first year after an advanced NSCLC diagnosis was reduced by nearly 50%. “The median overall survival was 11.3 months in the control arm but has not yet been reached in the active arm,” said Gahdhi, adding that, “A landmark analysis at 12 months suggests that 69.2% would be alive at 12 months versus 49.4% for the control arm.”

So, a huge win. Of particular note, this vaunted IO-spurred efficacy was also observed – though to a much lesser degree – in patients that express only very low levels of the IO drug’s molecular target, that being the membrane receptor, PD-1. What this means at the mechanistic level is anyone’s guess, but clinically, it means that advanced NSCLC patients will be eligible to receive pembrolizumab regardless of what biomarker testing for PD-1 indicates.

Checkmate-227: This trial – also a phase III investigation – compared the combination of two IO drugs, ipilimumab and nivolumab, with standard chemotherapy (Initial report presented at AACR: N=299).

Results showed that the IO/IO combination in patients with a high tumor mutation burden (TMB) had a hazard ratio of 0.58 for progression-free survival with the immunotherapy versus chemotherapy (p<0.0002). Further, these responses were durable: 68% response at one year for IO/IO versus 25% for the chemo doublet (Figure 1).


Given prior phase II reporting these results were not a surprise – the big reveal here had more to do with TMB testing, which seems able to predict who will respond to this very expensive drug combination. (Note: High TMB is defined as ≥10 mutations/megabase.)

Question: Is TMB now king? “TMB not a wholly new or distinct test,” said Matt Hellman, MD, of Memorial Sloan Kettering Cancer Center, New York, “Rather, TMB is like data that we’re already getting routinely from next-gen sequencing, such as the IMPACT panel (a 468-gene readout) done at my institution.”

The problem is standardization. The panel used in Checkmate 227 is from Foundation Medicine, which also looks at hundreds of genes but is not the same panel used in IMPACT. Therefore, if TMB is included in the label for the IO/IO combination in NSCLC (which seems likely) what is an institution – let alone a community provider – to do?

There is also the issue of timing. The Foundation test has a turnaround time of several weeks – a long time for a metastatic cancer patient to wait to begin treatment. But Hellman isn’t worried: “We’ve come so far already with next-gen sequencing in a really sort time period, so I’m optimistic that this field will continue to improve to allow this data to be available in real time.

IO before surgery

As is common practice with chemotherapy, cancer patients who are eligible to have their tumors removed surgically can benefit greatly from drug treatment prior to being wheeled off to the OR. This type of “neoadjuvant” treatment is being investigated with IO.

Results for neoadjuvant treatment with the PD-1 inhibitor, nivolumab, in the setting of early-stage NSCLC were presented by Drew Pardoll, MD, PhD, of Johns Hopkins’ Bloomberg-Kimmel Institute for Cancer Immunotherapy. “Every day thousands of patients around the world will get operated on for resectable cancer, and will be told by their surgeons, ‘We got it all’,” said Pardoll, but that is frequently not the case – the five-year survival rate for such patients is less than 50%. “So, there is a great need for additional therapies to be applied in the context of the surgical, resectable setting.”

Historically, the use of chemotherapy as neoadjuvant was to reduce the tumor burden, thereby enabling cleaner margins upon removal, but that’s not the goal with neoadjuvant IO. “The notion here is that these tumors already have tumor-specific T cells sitting in the tumor,” explained Pardoll, “But they’re restrained from acting by the engagement of PD-1 (the receptor) by PDL-1 (the ligand) in the tumor.” Thus, using anti-PD-1 before surgery results in an increase of activity within the tumor. “Hopefully, a portion of these (T cells) will spill out of the tumor into the circulating blood and find those residual few tumor cells that have escaped the tumor mass, intercept them, and prevent them from creating the relapse that ultimately kills patients who have had their primary tumor resected.”

That was the hypothesis tested (N=20).

The outcome: Using nivolumab monotherapy prior to surgery resulted in 45% of resected tumors having a major pathologic response, with a “robust” infiltration of PD-1+ T cells in the excised tissue. “In fact, much to our surprise, a much larger number of tumors showed significant pathologic regressions as defined by replacement of tumor cells by non-tumor cells (like lymphocytes)”. Further, tissue removed from three patients fell under the definition of pathologic complete response – with no viable tumor cells at all. “And this is just four weeks after they were given nivo,” said Pardoll.

Innate reasoning

Checkpoint inhibitors (e.g. nivo, pembro) augment the activity of the adaptive arm of the immune system – but that’s just half the story; whereas adaptive immunity is specific, targets of the innate arm of the immune system are far more general. A common component of the adaptive system are toll-like receptors (TLRs), which come in at least 10 molecular flavors, all with slightly different functions.

The drug, CMP-001, a virus-like particle loaded with a 30-mer oligonucleotide, leverages the activity of TLR9 (Figure 2).

“Toll-like receptors of the innate immune system are displayed on macrophages and dendritic cells, explained Suzanne Topalian, MD, of Johns Hopkins, “And they help protect us against viruses and bacteria, so they recognize components of microbes,” There are currently two approved agents that take advantage of this system. BCG, a microbe used to treat bladder cancer, and imiquimod, a topical agent that works through TLR7 used for melanoma.

For CMP-001, a phase I trial was conducted with 85 metastatic melanoma patients who were resistant to anti-PD-1 treatment. The drug was administered by direct injection into tumors, with concurrent administration of pembrolizumab.

Of 85 patients treated, 18 patients had a notable response, two of these being complete responses. Three patients have maintained response at one year. Not a homerun, to be sure, but remember that these patients have run out of other treatment options.

Of greater interest was reduction observed in tumor masses that were not injected – the so called “abscopal” effect – one of the holy grails of immunotherapy.

Mohammed Milhem, MBBS, University of Iowa, reported the MCP-001 results.

“We did see systemic, as well as local effects,” said Milhem. “In fact, there was a very powerful abscopal effect in many visceral lesions (nodal, hepatic, splenic) that were not injected.”

Investigation in other tumor types is underway.

New Blu

Reporting on a phase I trial with the small molecule, RET-targeted inhibitor BLU-667 was Vivek Subbiah, MD, of MD Anderson Cancer Center (Figure 3).

“RET is a driver in multiple cancer types,” said Subbiah. In healthy tissue RET helps to control organ development and tissue homeostasis, whereas oncogenic RET drives tumorogeneis through activating point mutations, or genetic fusions. “As you know, there are patients with EGFR, ALK, or ROS1 alterations (which have available treatment options) while, unfortunately, patients with RET altered cancers have not benefitted from the development of oncology targets to date.”

Enter BLU-667, a drug of sub-nanomolar potency, which targets include the predicted RET gate-keeper mutation, V804M. In most cancers this genetic variant is rare, but it is present in >60% of medullary thyroid cancers (MTC), and roughly 2% of NSCLCs.


Though reporting is still very early, in a phase I trial of 51 RET-altered patients several remarkable responses have already been observed. Subbiah related two of them:

“Really, this is what makes phase I investigators wake up in the morning and go to work.” One patient, a 27-year-old male with a history of highly invasive MTC, “…so aggressive that he couldn’t breathe…” was started on a very low level of drug and responded dramatically. Following initial response, the patient was titrated up to 400 mg, “and right now, almost 11 months later, he is going back to full-time work” said Subbiah. “This is phenomenal for a phase I study.”

A second patient, a male, 74 years of age with RET-altered NSCLC, had progressed on several lines of chemo and was in a wheelchair when he enrolled on the trial. “Within two months he was a different person,” said Subbiah, “Up and walking. Fantastic.”

The treatment was well tolerated, with no grade 4 events, and few grade 3 – mostly for hypertension.

The study is ongoing, and results to date were published in Cancer Research the day of the AACR presentation.

A matter of FATE

Not much data in this last presentation, except – stay tuned to the technology. As described by Bob Valamehr, PhD, of FATE therapeutics, the business future of CAR T-cells may lie in the tricked-out T and NK (natural killer) cells being manufactured at FATE. (Much of the technology described here is coming out of Michel Sadelain’s lab – he being one of the co-founders of the CAR-T giant, JUNO.)

To date, CAR T-cells are genetically altered T cells derived from and returned to a single patient (which avoids issues of transplant rejection). Whereas Sadelain, via FATE, is developing “off the shelf” cells that can be used to treat anyone (Figure 4).

The core of the technology is induced pluripotent stem cells (iPSCs), cells that, theoretically, can be differentiated into any mature cell type, like T cells. This has two distinct advantages over current CAR T-cell design. The first being the already mentioned universal use, the second being batch control – ensuring a consistent product.

Using the iPSC approach, a single cell (rather than thousands) is transformed into a CAR-T in a way that eliminates the potential for host rejection. From there, that single cell can multiplied into billions of identical daughter cells that can be used to treat hundreds of patients.

“Using this approach,” said Valamehr, “We can effectively place multiple attributes into a cell at the single cell level and create a massive cell bank.” Thus, a homogenous, cost-effective product.

The proof-of-principle construct, FT819, reported at AACR, is a CAR-T product which has not one, but two targeting moieties – the poster child of CARs, CD19, as well as the CD16 Fc receptor, which augments the CARs activation signaling capacity.

FT819, reported at AACR, has performed well in cell killing assays, and FATE hopes to file an IND within the next 12 months.

(Neil Canavan is the author of, “A Cure Within: Scientists Unleashing the Immune System to Cure Cancer” published by Cold Spring Harbor Laboratory Press.)

Post a Comment