Intratumoral Electroporation of pIL-12 Plus Pembrolizumab Elicits Durable Responses in PD-1–Refractory Advanced Melanoma

The addition of plasmid IL-12 (TAVO or tavokinogene telseplasmid) to pembrolizumab (Keytruda) resulted in durable responses, as well as several complete responses (CRs), in patients with advanced melanoma who were refractory to PD-1–directed therapy, according to interim data from the phase 2 KEYNOTE-695 (NCT03132675) presented during the 2020 SITC Annual Meeting.¹

Results showed that the combination elicited an objective response rate (ORR) of 30% (95% CI, 18.0%-43.6%; n = 16/54) per investigator assessment in accordance with RECIST v1.1 criteria. Of these responders, 6% (n = 3) achieved a CR. In patients with M1c/M1d disease, the ORR was slightly higher, at 35.3% (n = 6/17). Patients who had previously received treatment with ipilimumab (Yervoy) experienced the highest ORR with the treatment, at 40% (n = 6/15). The median duration of response (DOR) is 12.2 months (95% CI, 5.6–not evaluable).

“We explored the combination of pIL-12 plus a PD-1 antibody in patients who were stringently proven to have progressed on PD-1 antibody and we saw an ORR of 30% with a 6% CR,” Adil Daud, MD, professor of medicine at the University of California San Francisco, said during in a presentation. “Extensive responses were seen.”

IL-12 is an immunoregulatory cytokine that encourages innate and adaptive immunity. By working with interferon (IFN)-γ, IL-12 expression is able to modify the tumor microenvironment, which is needed to elicit antitumor immunity, explained Daud. Although the IL-12/IFN-γ axis is not usually present in advanced melanoma, investigators have deduced that intratumoral electroporation of pIL-12 can restore this axis, thus promoting anti–PD-1 immunotherapy activity in patients who are predicted to be unresponsive to treatment.^2,3

In light of these findings, investigators set out to extend the intratumoral IL-12 platform by examining the efficacy and safety of pIL-12 and pembrolizumab in the treatment of patients with advanced melanoma who were refractory to PD-1 inhibitors.

“Gene electroporation is a technique that delivers IL-12 into the tumor microenvironment in an immunologically relevant fashion with minimal systemic spillage,” said Daud. “Previous trials have shown that this technique is safe and effective.”

The phase 2 trial enrolled patients with unresectable or metastatic melanoma who had received anti–PD-1 therapy, either as monotherapy or in combination, for 12 weeks before entry. To be eligible for enrollment, patients had to have confirmed disease progression per RECIST v1.1 criteria with no intervening treatment before entry. Moreover, they need to have measurable disease and at least 1 anatomically distinct lesion that was determined to be accessible for electroporation. The planned enrollment for the trial is 100 participants.

Participants were administered intravenous pembrolizumab at a dose of 200 mg every 3 weeks plus pIL-12, which was given to at least 1 accessible lesion on days 1, 5, and 8 every 6 weeks until all lesions were treated.

The primary end point of the trial is ORR per blinded independent review, while key secondary end points are comprised of investigator-assessed ORR, DOR, progression-free survival (PFS), intracranial PFS (iPFS), iORR, and overall survival.

Data for 56 patients were reported during the meeting. The median age was 66 years, and 55.4% (n = 31) of patients were male. More patients had an ECOG performance status of 0 versus 1, at 62.5% (n = 35) versus 37.5% (n = 21), respectively. The majority of patients had BRAF wild-type or BRAF non-V600 mutation status (78.6%; n = 44), while 21.4% had BRAF positivity (n = 12).

Most of the patients, or 75% (n = 42), had normal lactate dehydrogenase (LDH), while 19.6% (n = 11) had LDH above the upper limit of normal, and 5.4% (n = 3) had unknown status. More than half, or 53.6% (n = 30), of patients had stage IVa or IVb disease, while 30.4% (n = 17) had stage IVc or IVd disease, and 16% (n = 9) had stage IIIb, stage IIIc, or stage IIId disease. Additionally, 76.8% (n = 43) of patients had over 3 target and non-target lesions at baseline, with a mean of 8.9 lesions.

Half of the patients (n = 28) received 1 prior therapy, while 23.2% (n = 13) had received 2 to 3 lines, and 26.8% (n = 15) had received 4 or more lines. The median number of previous treatments was 1.5, while the mean was 2.9 (range, 1-17). Twenty-seven percent (n = 15) of patients had previously received ipilimumab. The median time that patients received prior treatment with a PD-1 inhibitor before KEYNOTE-695 was 5.3 months, and the median time from the last dose of that therapy before study treatment initiation was 1.2 months.

Additional results showed that among the 16 responders to the combination, 24% (n = 13) experienced a partial response, and 19% (n = 10) achieved stable disease. Approximately half of patients (52%; n = 28) experienced disease progression. Two patients were not evaluable for best ORR.

When looking at the treated (n = 22) and untreated (n = 20) lesions from 16 responders to treatment, investigators reported that the ORR for patients with less than 50% of target lesions injected was 40% (n = 6/15), while the ORR for patients with 100% of target lesions injected was 32% (n = 7/22).

Investigators also noted that patients had immunologically quiescent lesions at the time of screening that quickly inflamed with a productive immune infiltrate. Moreover, an increase in CD8+ T cells in the tumor microenvironment while patients were receiving treatment was found to be linked with an active transcriptomic profile with more intratumor clonality following just 1 cycle of therapy.

Lastly, peripheral blood mononuclear cells were collected at the time of screening. Following 2 cycles of study treatment, they were analyzed via flow cytometry. Results showed that responders had significant on-treatment increases in short-lived effector cells and substantially less polymorphonucler myeloid-derived suppressor cells versus those who were not responsive to treatment. Moreover, genomic data was collected from 2 g of fecal material from the patient, and whole-genome sequencing revealed that responders had a significant on-treatment increase in the incidence of Clostridiales versus nonresponders.

Daud shared the case of 1 patient with stage IVb disease who experienced pseudoprogression in a target lesion following treatment with nivolumab (Opdivo) ended up achieving a response to treatment with the combination. The patient received nivolumab for approximately 2.5 months, from March 2018 to May 2018, and around 1 month later, had confirmed progressive disease. Later, the patient received pIL-12 for approximately 6 months until no more lesions were left to treat. They then received maintenance pembrolizumab. The patient was on treatment for approximately 2 years. As of August 2020, the patient has been determined to be disease free with no new cancer treatment administered.

In another case, the patient with stage IVc disease had progressed on treatment with ipilimumab plus nivolumab and ended up achieving a response to the study combination. The patient had received pembrolizumab for about 2.5 months, from May 2018 to July 2018. Approximately 1 month later, the patient had disease progression. They then received ipilimumab/nivolumab from about 2.5 months, from August 2018 to October 2018. Again, the patient experienced progressive disease about 1 month later. The patient then received pIL-12 for about 6 months until no more lesions were left to treat; this was followed by maintenance pembrolizumab. The patient received the study treatment for approximately 21 months so far, and they continue to benefit from a response.

Another patient had with stage IVc disease received prior treatment with dabrafenib (Tafinlar) plus trametinib (Mekinist) for approximately 10 months, ranging from July 2018 to May 2019. They achieved stable disease for about 2 months. They then went on to receive ipilimumab/nivolumab for about 1.5 months, from August 2019 to October 2019 and subsequently progressed about 6 weeks later. Then, for about 1 month, December 2019 to January 2020, the patient received nivolumab. Again, about 6 weeks later, they experienced disease progression. At this time, they received pIL-12 for about 3 months until no more lesions were left to treat; this was followed by maintenance pembrolizumab. The patient has been receiving treatment for about 7 months and continues to respond.

With regard to safety, the most commonly reported all-grade treatment-related toxicities included fatigue (26.8%, grade 1/2), procedural pain (23.2% grade 1/2), diarrhea (19.6%, grade 1/2), nausea (10.7%, grade 1/2), and rash (10.7%, grade 1/2). Three patients experienced grade 3 toxicities: cellulitis, enteritis, and Lichen planus. The median time for pIL-12 electroporation treatment was 11 minutes.

References

1. Fernandez-Penas P, Carlino MS, Tsai KK, et al. Durable responses and immune activation with intratumoral electroporation of pIL-12 plus pembrolizumab in actively progressing anti-PD-1 refractory advanced melanoma: KEYNOTE 695 interim data. Presented at: 2020 SITC Annual Meeting; November 10-15, 2020; Virtual. Abstract 799.
2. Algazi A, Bhatia S, Agarwala S, et al. Intratumoral delivery of tavokinogene telseplasmid yields systemic immune responses in metastatic melanoma patients. Ann Oncol. 2020;31(4):532-540. doi:10.1016/j.annonc.2019.12.008
3. Algazi AP, Twitty CG, Tsai KK, et al. Phase II trial of IL-12 plasmid transfection and PD-1 blockade in immunologically quiescent melanoma. Clin Cancer Res. 2020;26(12):2827-2837. doi:10.1158/1078-0432.CCR-19-2217