The correlative data show an immune-directed mechanism that is differentiated between responders and nonresponders, suggesting the combination can effectively alter the tumor microenvironment to benefit patients otherwise unlikely to respond to anti–PD-1 monotherapy.— Alain Algazi, MD
In patients with stage III/IV melanoma, immunologically “cold” tumors were rendered immunologically active through intratumoral injections of plasmid interleukin-12 (IL-12) combined with pembrolizumab (Keytruda).1 Describing the approach at the 2017 ASCO-SITC Clinical Immuno-Oncology Symposium, Alain Algazi, MD, of the University of California, San Francisco, said, “We think we are inducing a systemic immunologic effect with this treatment.” The combined strategy produced responses in almost half the patients treated on this phase II study, he reported.
The expression of programmed cell death ligand 1 (PD-L1) in the tumor microenvironment and the proportion of programmed cell death protein 1 (PD-1) high expressors plus CTLA-4 high expressors on CD8-positive tumor-infiltrating lymphocytes are presumed measures of immune activity. A low frequency of these markers seems to indicate a “cold” tumor that will not respond well to immune checkpoint inhibition. One way to boost immunologic response may be through the delivery of cytokines, which has been used for years in melanoma, but toxicity limits its usefulness, Dr. Algazi said.
Dr. Algazi and his team developed a platform that involves injection of the plasmid for IL-12 into the tumor and the application of electric current (ie, electroporation) to facilitate its uptake and sustained expression of IL-12. Intratumoral plasmid IL-12 delivered via electroporation (IT-pIL12-EP) has been shown to increase tumor-infiltrating lymphocytes in both treated and untreated lesions. The investigators hypothesized that nonresponse could be “rescued” with the combination of IT-pIL12-EP and anti–PD-1 therapy.
“IL-12 is a potent, well-characterized proinflammatory cytokine. If we just inject protein, it disappears quickly, but if we use plasmid and electroporation, the cytokine stays in the tumor for a week at a time. We thought this might potentiate response to anti–PD-1 antibodies, so we conducted a combination study,” he said.
Phase II Study Details
The phase II study enrolled 22 patients with stage III/IV melanoma who were predicted to respond poorly to anti–PD-1 therapy based on evidence suggesting that their tumors were immunologically inactive. By flow cytometry, patients had a low proportion (< 22%) of PD-1–high expressing/CTLA-4–positive cells, a value associated with nonresponse to anti–PD-1 agents. Some were naive to anti–PD-1 agents, and others had been previously treated with immunotherapy.
Patients received pembrolizumab (200 mg every 3 weeks) and IT-pIL12-EP 3 times every 6 weeks (ie, every other cycle). Pre- and posttreatment blood and tumor specimens were collected and analyzed for immune phenotyping, gene expression, T-cell receptor diversity, and changes in the tumor microenvironment by immunohistochemistry.
Clinical and Immunologic Data
“Intratumoral delivery of IL-12 stimulated a safe but powerful systemic immune response,” Dr. Algazi reported. “Remarkably, although predicted to be an unresponsive population, the objective response rate was 40% in this population deselected for response (4 complete and 2 partial responses). If we include a patient who had disease progression but ultimately had a dramatic response, the response rate was 48%.”
Responses to IT-pIL12-EP were durable, similar to what is observed in other patients who respond to immune therapies, he added.
In responders, analysis of the tumor microenvironment revealed many significant immunologic changes not seen in nonresponders, including the number and ratios of CD8-positive to PD-L1–positive cells by immunohistochemistry, increased expression of natural killer CD8 and adaptive resistance markers (eg, interferon gamma) by NanoString technology, increased clonality of T cells, and increased numbers of tumor-infiltrating lymphocytes. Log-fold changes in markers of activated T cells were significantly higher in responders vs nonresponders (P < .05). Changes were seen in lesions that were directly injected and in nontreated lesions as well.
“So what’s going on? Responders have increased immune cell infiltration into the tumor, typically associated with increases in the CD8-positive to PD-L1–positive cell ratio. We think we are increasing the number of effector cells relative to the number of regulatory cells. In a sense, you can bring in immune cells in many patients, but the difference in responders vs nonresponders is the nature of those cells,” he said.
“The excellent safety profile and striking 40% clinical response rate are encouraging, as treated patients were predicted to be nonresponsive to pembrolizumab,” Dr. Algazi concluded. “The correlative data show an immune-directed mechanism that is differentiated between responders and nonresponders, suggesting the combination can effectively alter the tumor microenvironment to benefit patients otherwise unlikely to respond to anti–PD-1 monotherapy.”
Ann W. Silk, MD
Ann W. Silk, MD, of Rutgers Cancer Institute of New Jersey, New Brunswick, commented that the study adds to the growing body of data showing that “activating immunologically inactive tumors is a major goal, and we should be enrolling patients on these trials. Short of this, we should consider available methods of priming tumors, such as with oncolytic virus therapy or radiotherapy.” ■
Disclosure: Dr. Algazi has received institutional research funding from Astra-Zeneca, Bristol-Myers Squibb, GlaxoSmithKline, Merck, Novartis, and OncoSec. Dr. Silk reported no conflicts of interest.