SAN FRANCISCO — The immune checkpoint inhibitor pembrolizumab (Keytruda) had minimal activity in previously treated neuroendocrine tumors (NETs), consistent with other data on immunotherapy for NETs, according to a study reported here.
Only four of 107 patients had objective responses, none of which was a complete response. In the few patients who did respond to pembrolizumab, the activity was durable. The safety profile of pembrolizumab was consistent with previously reported clinical experience with the drug, reported Jonathan Strosberg, MD, of Moffitt Cancer Center in Tampa, Florida, and colleagues.
The latest in a procession of negative data does not necessarily represent a death knell for immunotherapy in NETs, but investigators have few solid leads to a way forward, acknowledged Strosberg in a presentation at the Gastrointestinal Cancers Symposium.
“Perhaps this should not be surprising,” Strosberg said of the results. “We’ve become increasingly aware of the fact that tumor mutational burden corresponds with response to immune checkpoint inhibitors, and neuroendocrine tumors — at least well-differentiated ones — are known to be on the relatively low mutational burden side of the spectrum.”
“That being said, among the responders there were a small number of durable responses,” he noted. “It would be great if we could figure out a way to predict who will respond, but I can say thus far, based on the cumulative data based on single-arm checkpoint inhibitor studies, we really don’t have an idea… . It seems clear from the study that PD-L1 expression is not a predictive marker, so we really don’t know who is more likely to respond at this point.”
In an early trial of pembrolizumab in patients with advanced solid tumors, the PD-1 inhibitor in a subset of patients with carcinoid and pancreatic NETs proved to be tolerable. Strosberg reported findings from an analysis of pembrolizumab in patients with previously treated NETs enrolled in another trial of patients with various types of advanced solid tumors.
The ongoing international, multicohort phase II trial, known as KEYNOTE-158, included patients with selected advanced solid tumors that progressed on standard-of-care therapy. The trial had a primary endpoint of objective response rate, and secondary endpoints included duration of response, progression-free survival (PFS), and overall survival (OS).
Patients received pembrolizumab for a maximum of 2 years, and follow-up to assess OS is ongoing. The subgroup with previously treated NETs had a median treatment duration with pembrolizumab of 4.5 months and a median follow-up of 18.6 months.
The NET subgroup had a median age of 59, and all of the patients in the subgroup had stage M1 disease. The subgroup included 17 (15.9%) patients with PD-L1-positive tumors, and the primary tumor site for a majority of patients was the pancreas (38.3%) or small intestine (21.5%). Treatment history was one prior therapy in 29% of patients, two in 23.4%, three in 15.0%, and four or more in 28.9%.
The overall response rate was 3.7%, and all responses occurred in patients with PD-L1-negative tumors. Additionally, no responding patients had tumors with microsatellite instability, another potential marker of responsiveness to immunotherapy.
The treatment led to stable disease in 57% of patients. However, Strosberg noted that NETs are generally slow growing, which complicated the evaluation of stable disease. A relatively modest median PFS of 4.1 months suggested the treatment had limited impact on stable disease — “we don’t have a strong indication of disease stabilization, at least not in a majority of patients.”
The type, frequency, and severity of treatment-related and immune-mediated adverse events were consistent with previously reported data.
During the discussion that followed the presentation, Philip A. Philip, MD, PhD, of Karmanos Cancer Institute in Detroit, asked what the future might hold for checkpoint inhibitors in NETs. Strosberg replied that the data have been “relatively negative across the board.” However, other possibilities exist.
“I think perhaps combination therapy, perhaps new sensitizers — and all sorts of new ideas have been thrown out, including angiogenesis inhibitors — and perhaps completely different types of immunotherapy, such as adoptive immunotherapy,” said Strosberg. “CAR T-cells targeting the somatostatin receptor are in early stages of development. Perhaps the answer lies there.”
Strosberg added, “Right now, I would consider using this drug sporadically in patients who have exhausted all standards of care…My standard approach, with respect to immunotherapy, is to go with a double — CTLA-4, PD-1 inhibitor — simply because we don’t yet have data to suggest that approach is ineffective.”
The KEYNOTE-158 trial was supported by Merck.
Strosberg disclosed relevant relationships with Novartis, Ipsen, Lexicon, Pharmacyclics, and Advanced Accelerator Applications.