Recurrent glioblastoma patients treated with pembrolizumab (Keytruda) before and after surgery had a substantial survival benefit compared to those who only received the anti-PD-1 immunotherapy prior to resection, according to findings from a randomized phase I study.
Among 32 treated patients, overall survival was 417 days in those who received neoadjuvant and then adjuvant pembrolizumab compared to 228.5 days for those who received adjuvant therapy alone (HR 0.39, 95% CI 0.17-0.94, P=0.04), Robert Prins, PhD, of UCLA Jonsson Comprehensive Cancer Center, and colleagues reported in Nature Medicine.
“We actually showed that the PD-1 blockade really does induce an anti-tumor immune response, and that this is enhanced if you give it before surgery,” Prins told MedPage Today.
Glioblastoma is the most common primary malignant central nervous system tumor, with an incidence of 3.2 per 100,000 in the general population. And it carries a dire prognosis, with a median overall survival of just 14.6 months with a standard regimen of surgery, radiation therapy, and temozolomide. The prognosis for patients with recurrent disease is even grimmer.
The researchers noted that treatment with anti-PD-1 immunotherapies has been shown to be effective as monotherapy in treating some cancers, but mostly in the adjuvant setting.
However, they pointed to previous preclinical work that demonstrated a significantly greater efficacy of neoadjuvant compared to adjuvant immunotherapy in treating metastatic breast cancer, a concept validated by clinical studies in resectable lung cancer and melanoma.
But, “past experience with checkpoint blockade in malignant gliomas has not seen a lot of success,” Prins said. “We were really interested in knowing what was going on at the tumor site.”
As to what this means in treating glioblastoma going forward, Prins said that “timing matters.”
“Neoadjuvant immunotherapy is a very different situation than neoadjuvant chemotherapy,” he explained. “That’s designed to shrink the tumor to a margin where it can be taken out safely. Neoadjuvant immunotherapy actually serves as a primer to initiate the immune response, and then you can take out the bulk of this immunosuppressive tumor, and give more of continued immunotherapy that can sustain and enhance what was initially begun in the neoadjuvant setting.”
In the current study, 35 patients with recurrent and surgically resectable glioblastoma were enrolled and randomized across seven institutions. Of those, 16 patients were randomized into the neoadjuvant pembrolizumab group and 19 into the adjuvant-only group (3 of these patients withdrew from the study before surgery). Patients all had recurrent glioblastoma, were candidates for surgery, and were at least age 18 years. Eligibility criteria included Karnofsky performance status ≥70, and radiotherapy in first-line treatment.
Patients in both arms received 200 mg of pembrolizumab every 3 weeks post-surgery. However, the patients in the neoadjuvant group also received a 200-mg infusion of pembrolizumab 9 to 19 days before surgery.
“We collected tissue from the surgical resection from every patient,” Prins said, adding that he and his colleagues looked at cell populations in the blood, genetic signatures, sequenced T-cell receptors, and performed mRNA expression profiling.
“We think what is happening is that when there is a progressive tumor, there actually is an endogenous immune response from the patients — it’s just dysfunctional,” said Prins. “When you give PD-1 blockade — in this case pembrolizumab — before surgery in this setting, it can enhance the function of these T cells. When these T cells are activated with this blocking antibody they actually secrete these cytokines like interferon gamma that actually bind to the tumor cells, are internalized, and send a signal to them to stop dividing.”
Going forward, Prins and colleagues plan on expanding the current study.
“We are also going to design new combinations that will take advantage of what we’ve learned,” he said. “For example, we’re going to combine this neoadjuvant timing of checkpoint blockade with a personalized vaccine.”
This study was funded in part by Merck, Adaptive Biotechnologies, the National Institutes of Health SPORE in Brain Cancer, the Cancer Research Institute, and others.
Prins disclosed no competing interests. Co-authors reported relationships with Merck, Bristol-Myers Squibb, General Electric, Amgen, Lilly, Agios, Puma Biotechnology, Debiopharm Group, and Roche; and two work for Adaptive Biotechnologies.