4:35 PM
 | 
Jan 31, 2019
 |  BC Innovations  |  Tools & Techniques

TMB needs company

Why predicting responses to checkpoint blockade will take more than just TMB

Tumor mutation burden is looking less like a magic bullet and more like just one of at least three variables that could jointly predict responses to checkpoint inhibitors. Combining TMB measurements with markers of T cell inflammation and antigen presentation could pinpoint responders in a wide range of cancers.

However, to avoid the pitfalls that came with measuring PD-L1, the TMB field will need to converge on standards for both methodology and validation that will allow results to be compared across trials.

By counting mutations in a patient tumor sample, TMB measures how different a cancer has become from its tissue of origin. That serves as a proxy for the likelihood the cancer cell will be targeted as non-self by the immune system.

Cancers with high TMB counts are therefore expected to be more susceptible to drugs that boost immune responses.

Numerous studies in the last five years have bolstered that hypothesis, showing high TMB count is associated with greater response to checkpoint inhibitors. The largest of these, published by Memorial Sloan Kettering Cancer Center researchers in Nature Genetics this month, showed the correlation held up in a study of 1,662 patients representing a dozen different cancer types, treated with one or more of seven different checkpoint inhibitors.

The mounting evidence for TMB led Bristol-Myers Squibb Co. to change a primary endpoint in its Phase III CheckMate -227 trial of Opdivo nivolumab plus Yervoy ipilimumab. Last February, BMS showed the anti-PD1 and anti-CTLA-4 combo increased progression-free survival (PFS) vs. chemotherapy in first-line non-small cell lung cancer (NSCLC) patients with high TMB, regardless of PD-L1 status -- a subpopulation not targeted by the original study design (see “Necessary Adjustment”).

At least 37 checkpoint inhibitor trials are using TMB as a stratification factor or endpoint across a range of solid and hematological cancers.

“TMB was one of these special biomarkers that had the opportunity to impact all patients with cancer, not just patients with specific disease types,” said David Fabrizio, cancer immunotherapy leader and VP of product development at the Foundation Medicine Inc. subsidiary of Roche.

“TMB is doing a good job on its own, but it's not the whole story.”

David Fabrizio, Foundation Medicine

Yet despite the pile of positive signals, regulator scrutiny has raised doubts that TMB can define a drug-responsive population on its own.

In October, BMS reported the EU’s Committee for Medicinal Products for Human Use (CHMP) requested an overall survival (OS) analysis of patients with low TMB, defined in the study as <10 mutations per megabase (mut/Mb). This analysis showed the hazard ratio (HR) for OS with Opdivo/Yervoy vs. chemotherapy was similar whether patients had high or low TMB -- 0.77 and 0.78, respectively.

And on Jan. 24, BMS withdrew its supplemental Biologics License Application (sBLA) to FDA for Opdivo/Yervoy in patients with TMB ≥10 muts/Mb, citing the need for more data on the relationship between TMB and PD-L1, which will be available in 1H19.

“TMB is doing a good job on its own, but it’s not the whole...

Read the full 2524 word article

User Sign in

Trial Subscription

Get a 4-week free trial subscription to BioCentury Innovations

Article Purchase

$100 USD
More Info >