A small randomized trial from Dana-Farber Cancer Institute researchers has found that pembrolizumab immunotherapy prior to surgery is more effective for improving overall survival (OS) than adjuvant immunotherapy administered after tumor removal in patients with recurrent glioblastoma.
The study, published in Nature Medicine (2019;25(3):477-486), represents one of the numerous research initiatives from Dana-Farber researchers focused on identifying treatment strategies aimed at prolonging survival in a cancer that continues to feature a relatively low 3-year survival rate (Neuro Oncol 2017;19(suppl_5):v1-v88).
Neoadjuvant checkpoint blockade has been of interest for many different types of cancers, Patrick Y. Wen, MD, of the Dana-Farber Cancer Institute, told Oncology Times. Yet findings from the phase III Checkmate 143 with nivolumab were negative for recurrent glioblastoma (Neuro Oncol 2017; https://doi.org/10.1093/neuonc/nox036.071).
“The general feeling for PD-1 antibodies is that they don't work well in glioblastomas because they generally code tumors with very few infiltrating lymphocytes,” explained Wen. “The rationale for this study was to try and understand how well the anti-PD-1 monoclonal antibody pembrolizumab did in these tumors.
“We randomized patients with recurrent glioblastomas who needed surgery to receive the drug before surgery primarily to study the tumor to see if the PD-1 antibody had any immune effect,” he continued. “In collaboration with researchers at UCLA, we found that even after 2 weeks following the injection of pembrolizumab, the treatment featured some immune effects.”
In their study, researchers at Dana-Farber randomized a total of 35 patients with recurrent glioblastoma to either neoadjuvant pembrolizumab (n=16) or adjuvant-only pembrolizumab (n=16).
Patients were balanced in terms of age, sex, Karnofsky Performance Status, isocitrate dehydrogenase mutation status, pre- or post-surgery tumor volume, O6-methylguanine DNA methyltransferase methylation status, and steroid administration. A total of 24 patients with progressive disease received further therapy with bevacizumab, and some patients also received dexamethasone at different times during the study and median 476-day follow-up (interquartile range, 414-522 days).
At the cut-off date, the number of deaths in the neoadjuvant and adjuvant-only groups was nine and 12, respectively. The intention-to-treat analysis demonstrated a greater increase in OS in patients who were treated with pembrolizumab prior to surgery versus after surgery (95% CI, 0.17-0.94; p=0.04, log-rank test). Likewise, the median OS was shorter in the adjuvant- versus neoadjuvant-only groups (228 days vs. 417 days, respectively). The median progression-free survival was also longer in the neoadjuvant group (99.5 days vs. 72.5 days; HR 0.43; 95% CI, 0.20-0.90; p=0.03).
Treatment with pembrolizumab was generally well-tolerated in the overall cohort, with no new or unreported toxicities in the neoadjuvant group. Approximately 67 percent (n=10) of patients who received neoadjuvant immunotherapy experienced grade 3-4 adverse events.
Discontinuation of pembrolizumab due to toxicities was recorded in only two patients of the neoadjuvant group. Treatment-related toxicities included muscle weakness (50%), headache (47%), and hyperglycemia (37%).
Pembrolizumab was also associated with an increase in the IFN-γ-related gene signature, an increase in cell-cycle genes, and patchy increases in tumor-infiltrating lymphocytes in PD-L1 expression.
“What was unexpected was that the patients who got the drug before surgery lived almost twice as long compared with those who did not receive the drug before surgery,” according to Wen. “At first, we thought this was an accident, so we looked really hard at the prognostic factors between the two groups, but there was basically no difference. What we think happened was that there probably are a small number of T cells in the tumor that are sensitive against the tumor, and the PD-1 antibody activates them.”
Other studies involving neoadjuvant administration of PD-1 antibodies in melanoma, lung cancer, and head and neck cancer have compelled the researchers to believe their findings are not a fluke.
One clear limitation of the study was its small sample size, necessitating further trials that can mirror the findings of the Dana-Farber researchers. Additionally, the study investigators suggest that the immune effects observed during the trial may have been affected by an imbalance in the number of pembrolizumab doses administered to patients in each treatment group.
“Right now, I think that the treatment for recurrent glioblastomas is terrible,” noted Wen. “We really haven't progressed in treatment for these patients in the last few decades, but I think the hope is that this approach suggests that maybe PD-1 antibodies, when given in the right situation, might have some efficacy and perhaps we can combine it with other agents.”
Until then, he added, Dana-Farber will continue expanding its research efforts by adding an additional 25 patients to their study cohort. Hopefully, this as well as further expanded investigations will be able to replicate their findings.
Brandon May is a contributing writer.
Top 10 Best Hospitals for Cancer