High TMB Proposed As Pembrolizumab Biomarker For Advanced Solid Cancers

Author: By Lynda Williams, Senior medwireNews Reporter 

medwireNews: Patients with a high tumour mutational burden (TMB) may be more likely to achieve an objective response to pembrolizumab therapy than other individuals with an advanced solid tumour, indicates research published in The Lancet Oncology. 

“This analysis of the KEYNOTE-158 study showed that [tissue] TMB-high status identified a subgroup of patients who could have a robust and durable tumour response to pembrolizumab monotherapy”, say Aurélien Marabelle, from Gustave Roussy in Villejuif, France, and co-authors. 

“Therefore, [tissue] TMB could be a useful biomarker to predict the efficacy of pembrolizumab monotherapy in patients with previously treated recurrent or metastatic advanced solid tumours.” 

The international phase II trial recruited adults who had received at least one prior line of treatment for unresectable or metastatic mesothelioma or anal, biliary, cervical, endometrial, neuroendocrine, salivary, small-cell lung, thyroid or vulvar cancer. 

The 1066 patients were treated with up to 35 cycles of pembrolizumab 200 mg given at 3-week intervals and followed-up for a median 37.1 months. 

Overall, 790 patients enrolled into the trial at least 26 weeks before data cutoff were evaluable for TMB, 13% of whom had a TMB-high status, defined as 10 or more mutations/Mb. 

A RECIST objective response occurred in 29% of the TMB-high patients versus 6% of those with a lower TMB, including a complete response in 4% and 2%, respectively. Median duration of response was unreached at time of analysis for the TMB-high group versus 33.1 months for those with a lower TMB. 

Median progression-free survival (PFS) was 2.1 months regardless of TMB status, and median overall survival (OS) was also comparable between TMB-high patients and those with a lower TMB, at 11.7 versus 12.8 months. 

However, the investigators note that the latter parts of the survival curves had separated, favouring the TMB-high versus non-TMB-high patients for PFS at 2 years (22 vs 7%) and OS at 3 years (32 vs 22%). 

“With median duration of response not reached and a possible separation in the Kaplan-Meier overall survival curves at 3 years, it is possible that a further overall survival benefit in the [tissue] TMB-high group might become apparent after longer follow-up, particularly considering that more than 100 patients remained at risk at 3 years in the overall efficacy population”, write Aurélien Marabelle et al. 

The author of a linked comment highlights that the large numbers of participants in the study with cervical or small-cell lung cancer have a poor prognosis and high treatment resistance, as well as high prevalence of TMB-high status. 

“The short time-to-progression and the early deaths of such patients might have affected the first part of the progression-free survival and overall survival curves, rapidly decreasing irrespective of [tissue] TMB status”, postulates Melissa Bersanelli, from the University Hospital of Parma in Italy. 

“After 18–24 months, the patient population might have been positively selected, consequently leading to a plateau in overall survival and progression-free survival”, she writes. 

Of note, TMB score was not significantly associated with PD-L1 combined positive score in the population, even when separately assessing patients with or without an objective response. 

“These elements suggest the potential of combining these factors, possibly together with the tumour microsatellite status, into a biomarker score to improve the predictive ability of each individual biomarker”, hypothesises Melissa Bersanelli. 

Noting that the study indicated the utility of TMB status in tumour types that are usually poorly immunogenic, the commentator suggests that such a “combined biomarker score could be useful to improve the selection of appropriate treatment options in these and other tumour types, for which treatment choices are currently driven only by PD-L1 expression, as in the case of non-small-cell lung cancer.” 

References 

Marabelle A, Fakih M, Lopez J, et al. Association of tumour mutational burden with outcomes in patients with advanced solid tumours treated with pembrolizumab: prospective biomarker analysis of the multicohort, open-label, phase 2 KEYNOTE-158 study. Lancet Oncol; Advance online publication 10 September 2020. https://doi.org/10.1016/S1470-2045(20)30445-9

Bersanelli M. Tumour mutational burden as a driver for treatment choice in resistant tumours (and beyond). Lancet Oncol; Advance online publication 10 September 2020. https://doi.org/10.1016/S1470-2045(20)30433-2