BioCentury
ARTICLE | Product Development
Biorender

Product Development

Basket trials carry tumor-agnostic drug development forward

Tumor-agnostic drug development is expanding into new biomarkers and drug classes, using a basket of basket trial designs

Tumor-agnostic drug development is expanding into biomarkers and drug classes, using a basket of basket trial designs to test them.

By lauren Martz, Senior Editor
August 21, 2020 10:34 PM UTC

Tumor-agnostic drug development has been slow to catch on, but with wider adoption of gene sequencing and a steady flow of new targeted therapy approvals, companies are starting to re-draw the lines that divide tumor types.

BioCentury’s analysis of the clinical trial landscape for tumor-agnostic therapies shows the field is branching out into a broader set of therapeutic classes, biomarker types and trial structures that should help pick up the pace of approvals in the coming years.

The idea behind tumor-agnostic drug development is to define a patient’s cancer by its genetic or molecular makeup rather than its location in the body. Because disease biology can vary widely across tumors originating from the same tissues,  matching the biology to the therapeutic mechanism should increase the odds of treatment success. 

The hold up is that redefining tumors goes against decades of practice in clinical trials, regulatory review and patient care that have all centered on tissue-specific treatments. 

While companies have increasingly incorporated biomarkers into clinical research, they have generally stayed within the confines of historical tumor definitions. For example, c-MET inhibitor Tabrecta capmatinib from Novartis AG (NYSE:NCS; SIX:NOVN) was approved this year for a genetically defined subset of non-small cell lung cancer (NSCLC) patients.

Such approvals are important for pairing patient subsets with targeted therapies, but restricting the label to specific tissues of origin leaves behind patients with other classically defined tumor types harboring the same mutations. 

BioCentury’s analysis of biomarker-defined basket trials suggests tumor-agnostic drug development is catching on.

Three approvals have set the stage for broader adoption of tumor-agnostic drug development.

The first was the 2017 approval of Keytruda pembrolizumab from Merck & Co. Inc. (NYSE:MRK) for microsatellite instability-high (MSI-H) and mismatch repair deficient (dMMR) solid tumors in patients who progressed on prior therapies or have no satisfactory alternative treatments. In June, the  PD-1 inhibitor was approved for the first-line tumor-agnostic setting.

Trk inhibitors Entrectinib larotrectinib from Bayer AG (NYSE:BAYN) and Rozlytrek entrectinib from the Genentech Inc. unit of Roche (SIX:ROG; OTCQX:RHHBY) were approved in 2018 and 2019 to treat tumors with NTRK gene fusions (see “Why Tissue-Agnostic Drug Development Needs NGS”).

More recent approvals also border on tumor-agnostic. The May approval of Retevmo selpercatinib from Eli Lilly and Co. (NYSE:LLY) captures multiple tumor subtypes with RET fusions, but stops short of allowing its use in all cancers with the genomic alteration (see “Approval of Lilly’s Targeted Lung Cancer Therapy”).

Now, more than three years after the first tumor-agnostic approval, trials evaluating therapies across multiple classically defined tumor types are finding their way into development plans and new companies are forming around the concept.

The clinical programs are mostly in the form of basket trials: a type of master protocol that studies the same therapy in multiple patient subgroups.

BioCentury’s analysis of biomarker-defined basket trials suggests tumor-agnostic drug development is catching on, and starting to branch out beyond the biomarkers and drug classes validated in the initial approvals.

Sponsors are now deploying different basket trial structures for different purposes: to capture the largest possible biomarker-defined patient populations for targeted therapies, to identify the best biomarkers for therapeutic candidates, and to identify the best treatments for genetically defined patient populations.

Casting a wider net

Though kinase inhibitors and immunotherapies are the most popular classes of therapies in the tumor-agnostic pipeline, other classes that haven’t yet seen tissue-independent approvals such as PARP inhibitors and cell therapies are also represented in the 26 biomarker-focused basket trials listed for cancer in ClinicalTrials.gov.

Among the 80 biomarkers under evaluation in those trials, indicators of high mutation burden or genomic instability, the markers for the original Keytruda approval, are still the most popular. However, about 70% of the top biomarkers — defined as those being used in more than one trial — are individual genes that pair with specific targeted therapies.

New company launches and IPOs also reflect the growing interest in tumor-agnostic drug development.

Last month, Elevation Oncology Inc. emerged from stealth to develop therapies for cancers driven by rare mutations (see “Elevation Debuts with Tumor-Agnostic mAb”). In April, University of Geneva spinout FoRx Therapeutics AG launched with a focus on synthetic lethality and tumor-agnostic development (see “A Broader Look at DNA Damage Repair”).

A pair of tumor-agnostic therapeutics developers -- Black Diamond Therapeutics Inc. (NASDAQ:BDTX) and Arcutis Biotherapeutics Inc.  (NASDAQ:ARQT)  -- made their NASDAQ debut this year. A third, PMV Pharmaceuticals Inc., is poised to go public following a crossover-heavy $70 million series D round earlier this month (see “PMV Likely One of the Next Preclinical IPOs”).

Baskets of baskets

A growing interest in novel clinical trial structures such as master protocols and the rise of tumor-agnostic development have combined to entice sponsors to explore multiple trial structures that fit within the basket trial framework.

Each trial type addresses a different need in the development of tumor-agnostic drugs.

Basket trials are already commonplace in Phase I trials, for standard development based on tissue of origin as well as tumor-agnostic development, because testing a therapy across multiple classically defined tumor types can help sponsors identify indications to evaluate in subsequent trials.

At least 15,027 interventional clinical trials listed in ClinicalTrials.gov are testing biologics or other therapies in multiple cancer indications, 3,269 of which are Phase I studies. In most cases, those studies aren’t based on biomarkers and aren’t intended to support tumor-agnostic label claims. 

Companies advancing tumor-agnostic drug development are designing later stage basket trials in addition to Phase I studies, and they are taking three main approaches to biomarker-driven designs. 

Biorender

Almost three-quarters of the 26 biomarker-based basket trials (18) are in the first category. These test a single therapy in patients with a common biomarker, regardless of where their tumors are located. The purpose of the design, which is usually run by an industry sponsor, is to prove a drug’s efficacy in a genetically or molecularly defined subset of patients. It can support a tumor-agnostic approval, and can help companies capture the largest market segments possible.

Some of those trials use the therapy’s target as the biomarker, while others use broader markers that can be applied to various therapies such as genomic instability.

The studies are also split between individual biomarkers and families of biomarkers. For example, inhibitors of the EGFR family of targets, including HER2, are being studied in patients with mutations across the EGFR genes. The goals are to capture a more complete set of patients likely to respond, and to determine if certain mutations better correlate with positive outcomes.

The second type of basket trial is biomarker-finding. In some cases, it isn’t clear which markers will correlate with responses, so patient subgroups with multiple biomarkers are designed into the trials.

An example is the Phase II Merck KGaA (Xetra:MRK) study of PD-L1 inhibitor Bavencio avelumab, which is being tested in tumors with high levels of PD-L1, EBV-positive tumors, MSI-H tumors and cancers with POLE and POLD1 mutations. 

The third bucket, multi-basket trials, line up multiple therapies with different biomarkers to find the best pairings. Unlike the previous two buckets, multi-basket trials are often run by patient organizations or academic institutions.

BioCentury’s analysis found four trials that use the multi-basket structure, including a trial run by the Chinese Academy of Medical Sciences that aims to identify effective therapies for rare cancers based on their biomarkers. The trial is studying 12 therapies across 10 biomarkers.

Another example is Europe’s Basket of Baskets study run by Cancer Core Europe (CCE). The first arm of that study is testing Genentech’s Tecentriq atezolizumab across 11 biomarker-defined subpopulations.

TARGETS
c-MET (MET; HGFR) – c-Met receptor tyrosine kinase 

EGFR (ErbB1; HER1) – Epidermal growth factor receptor

HER2 (EGFR2; ErbB2; neu) – Epidermal growth factor receptor 2

NTRK – Neurotrophic tyrosine kinase receptor

PARP – Poly(ADP-ribose) polymerase

PD-1 (PDCD1; CD279) – Programmed cell death 1

PD-L1 (B7-H1; CD274) – Programmed cell death 1 ligand 1

POLE – DNA-directed DNA polymerase ε

POLD1 – DNA polymerase δ1 catalytic subunit

RET – Ret proto-oncogene

Biorender

Tumor-agnostic drug development has been slow to catch on, but with wider adoption of gene sequencing and a steady flow of new targeted therapy approvals, companies are starting to re-draw the lines that divide tumor types.

BioCentury’s analysis of the clinical trial landscape for tumor-agnostic therapies shows the field is branching out into a broader set of therapeutic classes, biomarker types and trial structures that should help pick up the pace of approvals in the coming years...