Keytruda (pembrolizumab, Merck & Co. Inc.) has become the first drug to be approved based solely on the presence of a molecular biomarker without regard to tumor location.

The tissue agnostic approval "for the treatment of adult and pediatric patients with unresectable or metastatic, microsatellite instability-high (MSI-H) or mismatch repair deficient solid tumors that have progressed following prior treatment and who have no satisfactory alternative treatment options," marks yet another milestone for the drug.

Looking for actionable mutations is the foundational principle of targeted therapy. Other drugs, from Gleevec (imatinib) on, have been approved based on the presence of specific molecular targets in the tumors they are approved for.

But so far, FDA approvals of targeted therapies have also always specified the anatomical origin of the tumor. Keytruda itself has been approved tumor type by tumor type up to now.

Tuesday's approval skips such anatomical constraints, requiring only that the tumors have MSI-H/deficient mismatch repair (dMMR) status.

Such tumors have an extremely high mutational load even compared to tumors with other types of DNA repair defects, such as BRCA-mutated tumors, though Roy Baynes told BioWorld Today that such other defects are also "fertile areas for exploration," and there is anecdotal evidence that they, too, can sensitize tumors to immunotherapy.

"What actually causes [tumors] to be seen as immunogenic is quite complex," said Baynes, senior vice president and head of global clinical development at Merck Research Laboratories. But in a general sense, "it was noticed fairly early on that cancers with high mutational burdens were more likely to be immunogenic."

Based on that observation, MSI-H/dMMR tumors "seemed like a good place to look," Baynes said. "And that's exactly as it's panned out."

Dung Le, who is an oncologist at the Johns Hopkins Bloomberg Kimmel Institute and was the principal investigator of the clinical trial of Keytruda in cancers with mismatch repair defects, and her colleagues tested the drug in roughly 150 patients with solid tumors enrolled in one of five uncontrolled, single-arm clinical trials. The overall response rate in those patients was about 40 percent. In 80 percent of those patients who responded at all, the responses lasted for more than six months, once again confirming the outstanding durability of responses to immunotherapy.


Tuesday's approval of Keytruda marks the broadest success of basket trials to date. Basket trials are a relatively new type of trial design that allows patients with the same mutation in tumors across several anatomical locations to enter the same clinical trial.

Although the original rationale behind basket trials was that mutation would trump location, so far the evidence suggests that at least for targeted therapies, treatment success is often determined by both.

In 2015, the results of a basket study investigating Zelboraf (vemurafenib, Roche AG) in BRAF mutated tumors beyond the drug's original indication of melanoma, response rates of around 40 percent were seen in non-small-cell lung cancer and the rare disease Langerhans' cell histiocytosis.

But other BRAF-mutated tumor types, most notably colorectal cancer, mostly shrugged off the drug, leading the investigators to conclude that "the histologic context is an important determinant of response in BRAF V600-mutated cancers."

At the 2017 annual meeting of the American Association for Cancer Research (AACR), data from the phase II SUMMIT basket trial investigating the use of neratinib (Puma Biotechnology Inc.) in patients with HER2 or HER3-mutated tumors across more than 20 anatomical locations showed that by far the most frequent responses were seen in breast cancer. However, SUMMIT's principal investigator David Hyman argued at the AACR meeting that the higher response rate in breast cancer was partly due to more clinical experience with how to use the drug in its flagship indication. (See BioWorld Today, April 4, 2017.)

Another possibility, Le told BioWorld Today, is that with respect to targeted therapies, "different tumor types may have different molecular bypass mechanisms," so that targeted therapy response does depend on anatomical location.


Next up in basket trials is Loxo Oncology Inc., which will be presenting pivotal data on its tropomyosin receptor kinase (TRK) fusion inhibitor larotrectinib (LOXO-101) in a late breaker session at the upcoming annual meeting of the American Society of Clinical Oncology (ASCO).

Loxo chief business officer Jacob Van Naarden told BioWorld Today that "on the regulatory front, it allows people to breathe a sigh of relief that the FDA is open to this concept. . . . It's always nice to have some regulatory precedence behind you."

He said the company had "a preclinical rationale based on biology," where TRK fusions were transforming in a number of different models, for testing larotrectinib in a basket trial.

But, he added, "There's the biology story, and then there's the clinical data."

Regardless of biological rationale, "you have to run the clinical trial," he said. "The BRAF story is a perfect counterexample of how genetics can be misleading."

Even for Keytruda patients covered by Tuesday's approval, anatomy has not become fully irrelevant. Repair-deficient colorectal cancer got a broader approval than other tumor types, with tumors that have progressed following treatment with certain chemotherapies eligible for Keytruda treatment.

On the other hand, Merck noted in its release announcing the approval that "the safety and effectiveness of Keytruda in pediatric patients with MSI-H central nervous system cancers have not been established."

Baynes said, though, that neither the warning on pediatric brain tumors nor the broader indication for colorectal tumors was due to response differences.

The brain tumor warning, he said, came from a combination of lack of data in this specific indication, and a worry about pseudoprogression.

Tumors can appear to grow after treatment with checkpoint blockers, as immune cells infiltrate the tumor.

In pseudoprogression the treatment is ultimately effective. But in the cramped quarters of the skull, temporary swelling can itself be a very serious problem.

The broader approval for colorectal cancer, in turn, is due to the fact that the trial was initially focused on colorectal cancer, and so "the colorectal population was studied in somewhat greater detail," Baynes said. But, he added, "responses are quantitatively and qualitatively similar" in other tumor types.