What Health Execs Need to Know about Tissue-Agnostic Cancer Drug Trials

August 17, 2018

Basket trials are designed to test therapies that target a specific genetic mutation found in the tumor, regardless of tumor histology. This emerging trend raises a number of key questions for the industry.

Cancer treatment has made significant strides toward the promise of personalized medicine in recent years. Keytruda’s microsatellite instability-high (MSI-H) or mismatch repair deficient (dMMR) approval last year serves as a proof of concept for tissue-agnostic development via basket trials. Basket trials are designed to test therapies that target a specific genetic mutation found in the tumor, regardless of the location of the tumor within the body (i.e. tumor type). If successful, basket trials can result in the approval of drugs that treat patients with a specific genetic mutation rather than a specific type of cancer (e.g. breast cancer, lung cancer, etc.)

Today, at least 31 unique therapies are being studied in basket trials that are phase 1/2 or later. For manufacturers, this development approach could provide access to a number of patients with different types of cancers and a fast route to market. While there are clear benefits, histological variance creates significant heterogeneity in tissue-agnostic, biomarker-defined populations. Not only does this pose developmental risks, but it may also impact how key stakeholders react to and incorporate such tissue-agnostic therapies into clinical practice.

Potential barriers to adoption

First, the FDA still requires tissue-agnostic trial data to be separated by tumor site in the product’s label, allowing physicians, insurers, and other stakeholders to identify in which tumors the therapy works best. From an access and willingness to prescribe perspective, tissue-agnostic therapies may need tremendous efficacy and/or unmet need to encourage truly broad use and overcome small sample sizes compared to conventional histology defined trials. Patient identification may also be a challenge as next-generation genomic sequencing (NGS) used to identify rare biomarkers hasn’t been widely adopted to date and could receive significant pushback from insurers due to cost considerations. Furthermore, many molecular targets may not be included in commonly used NGS panels.

It will likely be a challenge for a basket trial to produce data that is as strong as some conventional therapies that are already incorporated into the National Comprehensive Cancer Network (NCCN) guidelines and/or clinical practice for a specific tumor type. For example, Herceptin’s registrational trials in HER2-positive metastatic breast cancer enrolled 220 patients while Keytruda was studied in two patients with MSI-H positive metastatic breast cancer. As such, Keytruda’s use in MSI-H positive tumors has only been recommended by the NCCN for tumors with large sample sizes in its registrational trial, and tumors associated with Lynch syndrome, which has long been linked to the MSI-H biomarker. Specifically, the NCCN has added Keytruda as a category 2A option to the list of “other regimens” recommended for the treatment of these tumor types. For all other cancers, as specified in its label, Keytruda is indicated for MSI-H positive “solid tumors that have progressed following prior treatment and who have no satisfactory alternative treatment options.” With minimal efficacy data and low biomarker prevalence, use in these other tumor types may be limited.

For a novel oncolytic that launches with a first approval for a tissue-agnostic indication, these clinical hurdles will likely be accentuated. At the time of its tissue-agnostic label approval, Keytruda’s efficacy in many tumor types was well established. Without this type of supplemental tumor specific data, clinicians may be more hesitant to prescribe a new therapy.

Potential genetic/biomarker testing challenges

Second, testing for a novel biomarker may not be as easily incorporated into clinical practice as it was for MSI-H which was already being tested in routinely used immunohistochemistry assays. For a novel biomarker requiring the use of NGS, cost will be a barrier to patient identification. Private insurers may not be receptive to universal testing of patients for a rare biomarker. For indications like NSCLC, in which there are many actionable biomarkers and associated therapies available today, this may not be a significant barrier; however, for many indications, NGS may not be readily supported.

Furthermore, there is no guarantee a novel biomarker will be included on commonly used NGS panels. MSI-H/dMMR is now being reported on molecular profiling panels, but the level of efficacy and degree of unmet need required to drive widespread incorporation of a novel biomarker into NGS panels remains to be elucidated.

Potential access barriers

Beyond clinical barriers, the current tumor-site-oriented paradigm may also present many challenges from an access perspective. How payers will incorporate tissue-agnostic therapies into their formulary is still an unanswered question; however, it appears likely that in the near term, payers will continue to evaluate therapies based on tumor-site-specific data within the tissue-agnostic clinical data package. This may place restrictions on the use of tissue-agnostic therapies in the broadest possible eligible patient population. Thus, we are left with a number of unanswered questions today, and even more questions in the future, if tissue-agnostic approvals gain significant traction:

Key questions

  • From the clinical perspective. How will tumor-specific guidelines change in response to additional approval of tissue-agnostic therapies? Will they remain tumor-specific? Will clinical practice change in any fundamental way (e.g., genetic testing, treatment algorithms/pathways, etc.)? How will clinicians manage competing treatment algorithms in the practice setting? For example, for instances where there is overlap in tumor biomarkers (e.g., MSI-high and BRCA in ovarian), how will clinicians use combination therapy, appropriate sequencing, or strict tradeoffs between targeted therapies?
  • From an access perspective. How will health technology assessments and agencies compare benefits across tumor types (e.g., a two-month improvement in progression free survival is meaningfully different in hepatocellular carcinoma vs. prostate cancer)? How will payers manage a growing pool of tissue-agnostic therapies?
  • From an industry perspective. How will industry change forecasting and performance tracking of products approved across a broad range of tumor types? Will this trend result in a need for additional services (e.g., personalized genetic marker maps, machine-assisted treatment algorithms, real-world evidence, etc.)? What are the advantages and disadvantages of this development pathway relative to other clinical development options? Will these shift as tissue-agnostic therapies become more prevalent?

The future of registrational basket trials

Thus far, the emergence of registrational basket trials has been limited (only two currently ongoing), perhaps due to the many short-term challenges associated with the development and commercialization of tissue-agnostic therapies enumerated above. However, we’ve identified at least 34 mid-phase and/or exploratory tissue-agnostic trials currently in the clinic.

In the near term, pending trial results, interest in tissue-agnostic basket trial development will likely grow, especially as NGS continues to become more widespread in oncology. Whether the number of tissue-agnostic indication approvals will ever be substantial enough to drive a significant shift in treatment approaches remains an open question. The growing pool of mid-phase/exploratory tissue-agnostic trials suggests we could be headed in that direction, leading to many long-term implications on the clinical application, commercialization and development of targeted oncolytics in the next several years.


Jillian Godfrey Scaife is a principal at Trinity Partners. During her tenure at Trinity, she has developed a diverse skill set, which enables her to lead cross-functional projects involving quantitative data analytics, primary market research, and statistical analyses-to generate strategic insights for Trinity’s clients.