As well as defining responses to immunotherapy, PD-L1 expression can be used to define which patients with metastatic non-small cell lung cancer (NSCLC) respond better to chemotherapy, Australian research shows.
However inter-assay variability can impact outcomes, according to investigators from the University of Sydney.
In a meta-analysis of 4,378 patients in the chemotherapy control arm of 12 first-line RCTs, comparing outcomes with platinum-doublet chemo versus immune checkpoint inhibitors, people with high PD-L1 expressing tumours had longer overall survival.
The study, published in Future Oncology, found chemo-treated patients with high PD-L1 tumours according to Dako 22C3 or 28-8 assays had a higher ORR than those with low/negative PD-L1 (HR 0.77; p=0.0002).
“Low PD-L1 tumours were also associated with a reduced risk of death compared with PD-L1 negative tumors (HR: 0.87, 95% CI: 0.76–0.99, p = 0.04),” the study said.
However no significant difference in OS was found for high versus low PD-L1 tumours (HR: 0.92, 95% CI: 0.82–1.03, p = 0.14).
In studies where Ventana SP142 assays were used to quantify PD-L1, there was a significant difference in OS when comparing PD-L1 low versus negative tumours (HR: 0.78, 95% CI: 0.66–0.93, p = 0.004).
However the same was not observed when comparing high PD-L1 tumours with negative (HR: 0.87, 95% CI: 0.71–1.08, p = 0.20) or high PD-L1 tumours compared with low (HR: 1.17, 95% CI: 0.97–1.41, p = 0.10).
“These findings raise questions about how best to incorporate chemotherapy into ICIs as a treatment for mNSCLC and warrant further research,” the study concluded.
“In trials included in the present meta-analysis, a mean of only 32.3% of patients assigned to ICIs received second-line chemotherapy, and thus, the majority did not receive a potentially beneficial treatment, particularly those with high PD-L1-expressing tumours.”
However the investigators noted that their findings were not consistent with other studies and they contradict current biological understanding.
“A greater understanding of PD-L1 and its association with survival outcomes will ensure a more nuanced discussion of prognosis with patients,” they said.
Lead investigator on the study, Dr Rachel Woodford, acknowledged there wasn’t a consistent gradation of treatment effects for ORR as there was for OS.
“I do acknowledge that they are incremental differences and I think it’s always an issue when you are assessing subtleties, albeit in a big group of patients. It can be quite difficult to find a signal.”
“Our conclusion based on that, and consistent with a recent pathology-based study, was that any PD-L1 expression was better than no PD-L1 expression.”
“They were looking at resected specimens and correlating the degree of PD-L1 to benefit in the adjuvant chemotherapy setting, which was gratifyingly fairly similar to what we were demonstrating albeit in the metastatic setting.”
Dr Woodford, currently a fellow at the Melanoma Institute Australia, told the limbic that the findings were interesting given chemotherapy is meant to be immunosuppressive through leukodepletion and induction of antigen intolerance.
“However … there is this growing recognition that certain chemotherapies can be immunogenic so to speak….through depletion of regulatory T cells and other pathways, which may in fact be somewhat the biological basis for the differences we were observing.”