Atlanta, GA—Immune checkpoint inhibitors represent a tremendous advance in the treatment of several types of cancers. Although approximately 20% to 25% of patients will have durable responses with these agents, it has been challenging to find biomarkers to identify who these patients are.
PD-L1 expression is a suboptimal biomarker, and researchers are looking at other potential biomarkers. Results of a retrospective exploratory analysis of the MYSTIC study suggest that high tumor mutation burden (TMB) can identify patients with non–small-cell lung cancer (NSCLC) who will have a survival benefit with the checkpoint inhibitor durvalumab (Imfinzi) and the investigational CTLA-4 antibody tremelimumab.
Moreover, high TMB is independent of PD-L1 expression, said lead investigator Solange Peters, MD, PhD, Chair, Thoracic Malignancies, Oncology Department, Centre Hospitalier Universitaire Vaudois, Switzerland, who discussed the results at the 2019 American Association for Cancer Research meeting. Furthermore, the analysis showed that blood TMB is as reliable as tissue TMB, overcoming one of the main disadvantages of inadequate tissue samples from patients with lung cancer and other tumors.
“TMB was predictive of an overall survival benefit with first-line durvalumab and durvalumab plus tremelimumab in metastatic NSCLC. In tissue TMB greater than or equal to 10, overall survival was improved with durvalumab, with or without tremelimumab, versus chemotherapy, but the small data set limits interpretation,” said Dr Peters. “Low TMB identified patients who did better on chemotherapy,” she added.
MYSTIC Study Details
MYSTIC was a phase 3 clinical trial that randomized 1118 patients with stage IV NSCLC to 1 of 3 treatment arms for first-line therapy—durvalumab alone; durvalumab plus tremelimumab; and platinum-based chemotherapy. Patients were enrolled regardless of level of PD-L1 expression.
The primary end point was not met. In the overall analysis of the MYSTIC study, no significant difference was found in overall survival (OS) between the 2 treatment arms. The OS results according to tissue TMB and blood TMB were exploratory end points.
Patients with the cutoff of tissue TMB ≥10 mutations/megabase (mut/Mb) of DNA had a median OS of 18.6 months with durvalumab, 16.6 months with durvalumab plus tremelimumab, and 11.9 months with chemotherapy. The difference between the 2 durvalumab arms did not reach statistical significance. Patients with a lower TMB (ie, <10 mut/Mb) did better with chemotherapy alone.
A separate analysis showed that using a higher threshold of TMB—≥20 mut/Mb—was associated with improved survival for durvalumab plus tremelimumab versus chemotherapy.
“We observed markedly improved overall survival and greater clinical benefit with durvalumab plus tremelimumab versus chemotherapy, indicating the potential contribution of tremelimumab in this setting,” Dr Peters noted.
“The results of the exploratory analysis support prospective study of blood TMB as a predictive marker for immunotherapy,” she said, but noted that raising the threshold for a cut point restricts the therapy to fewer patients. “This should be studied in as many clinical trials as possible.”
TMB a Predictive Biomarker for Immunotherapy Benefit
“TMB is a predictive biomarker for a survival benefit from immunotherapy that is independent of PD-L1 expression. However, TMB had not previously been studied to show a survival benefit for immune checkpoint inhibitor versus chemotherapy,” Dr Peters told the audience.
Tissue TMB is measured by next-generation sequencing from Foundation Medicine. “Blood TMB was evaluated for the first time in MYSTIC, using a Guardant sequencing platform that comprises a 500-gene panel. In a subset of patients with matched tissue and blood samples, blood TMB showed good correlation with tissue TMB,” Dr Peters told the audience. “Blood TMB is easier, less invasive, and faster,” she added.
The investigators used several thresholds for blood TMB, including >16 mut/Mb. “We found better survival compared to chemotherapy in all patients defined as high TMB, irrespective of threshold chosen,” Dr Peters said.
Using blood TMB ≥20 mut/Mb as a threshold, OS was superior with high TMB and a significant benefit with durvalumab plus tremelimumab versus chemotherapy was seen using this higher threshold.
In patients with high blood TMB >20 mut/Mb, the 2-year OS was 48% with the combination of durvalumab plus tremelimumab, 33.8% with durvalumab alone, and 19% for chemotherapy.
“In patients with blood TMB <20 mut/Mb, no significant survival difference was observed for immunotherapy, and chemotherapy is still a good option,” she said.
“PD-L1 and blood TMB are independent biomarkers for survival. They do not correlate. You look at different patients when you use these biomarkers,” Dr Peters said.
No difference was observed in safety outcomes when patients with high TMB were compared with the overall study population.