A phase I dose-escalation, safety/tolerability, and preliminary efficacy study of the intratumoral administration of GEN0101 in patients with advanced melanoma

Advanced melanoma is refractory to anticancer drugs, and many new immunotherapies have been developed in response to this in recent years. Of note, while programmed cell death protein-1 (PD-1) and cytotoxic T lymphocyte antigen-4 (CTLA-4) inhibitors are superior to chemotherapy in terms of their clinical effectiveness for advanced melanoma patients, combination therapy with anti-PD-1 and anti-CTLA4 antibodies can cause severe immune-related adverse events, which are sometimes life-threatening [1]. Therefore, new drugs with few side effects that improve the efficacy of immune-checkpoint inhibitors are urgently needed. As candidates, we focused on the clinical usefulness of intratumoral oncolytic virotherapy instead of systemic therapy.

Hemagglutinating virus of Japan (HVJ), also known as Sendai virus, belongs to the Paramyxovirus family with single-strand RNA [2]. HVJ-envelope (HVJ-E), an inactivated form of HVJ, was developed as a drug delivery vector with membrane fusion activity [3]. The envelope itself has an antitumor effect through the induction of direct killing of cancer cells and activating CD8-positive cells and natural killer (NK) cells [4], 5. In addition, induced tumor-specific cytotoxic T lymphocytes (CTLs) were maintained even when dacarbazine was used in combination with HVJ-E for melanoma [6]. Fragmented RNA in HVJ-E can be taken up into cells by membrane fusion and induce type-1 interferon via the RIG-I pathway [7]. Not only apoptosis but also necroptosis was caused by HJV-E in neuroblastoma [8]. Since HVJ-E exerts an antitumor effect via the above diverse mechanism, we performed a first-in-human phase I/IIa dose-escalation clinical study for patients with advanced melanoma that could be treated intratumorally by HVJ-E at 1000 and 3000 milli neuraminidase units (mNAU) [9]. As a result, the safety and tolerability were confirmed with no dose-limiting toxicity (DLT). CD8⁺ and CD4⁺ cells were recruited to the tumor by HVJ-E. Tumor shrinkage was found at the directly treated and untreated sites. Furthermore, the appearance of vitiligo reflecting the results of the immune response was observed. This result indicated that HVJ-E can activate tumor immunity and may be effective in combination with immune-checkpoints inhibitors. Subsequent to this favorable result, we planned to use higher doses of HVJ-E (GEN0101) at 30,000 and 60,000 mNAU in patients with stage IIIC-IV advanced melanoma.

In the present phase I clinical trial, we performed intratumoral administration of GEN0101 in two dosages to confirm its safety and tolerability as the main purpose and to assess the antitumor effect as the secondary purpose.

In this clinical study, the safety and tolerability of the intratumoral injection of GEN0101 at two different doses in patients with advanced melanoma were demonstrated. Local regression of the tumor at injection sites and regression of the metastatic tumor in lung was suggested to be mediated by strong antitumor immunity. Thus, the treatment with GEN0101 has been suggested to induce systemic as well as local antitumor immunities, as described below. These data suggest that GEN0101 may represent a new treatment for patients with advanced melanoma.

Notably, there were no serious adverse events in which a causal relationship could be ruled out. DLT was not observed in this trial. Adverse events of CTCAE grade ≥ II, including injection site reactions, occurred due to GEN0101 administration; however, all of these were temporary. The adverse events observed in this study were also already reported in the HVJ-E clinical study and for other intralesional immunotherapies [9] [13]. In addition, no DLT was observed at the same dose of GEN0101 in another clinical trial for prostate cancer [14]. A local skin reaction at the injection site may reflect a substantial immune response to GEN0101. This adverse event was considered to result from the induction of recruitment of immune cells, including CD8⁺ T cells, as was observed in a previous clinical study for melanoma [9]. The decreased neutrophil count observed in this study was suggested to be due to the chemokine-mediated recruitment of neutrophils from bloodstream to tumor tissues. Intratumoral injection of HVJ-E reportedly induces the expression of CXC motif ligand 2, which is a chemokine for neutrophil, and infiltration of N1 neutrophil into tumor tissue in mouse model of melanoma [15]. N1-type neutrophils have an antitumorigenic phenotype in tumor microenvironment [16]. Erythema and heat due to local inflammation repeatedly appeared but receded over time. In addition, a transient fever was able to be suppressed by premedication, such as non-steroidal anti-inflammatory drugs (NSAIDs) and acetaminophen. The fever was considered to have been induced by cytokines, such as interferon-γ. Since a dose of 60,000 mNAU was confirmed to be safe, this dose will likely continue to be used in future clinical trials.

The antitumor effect was comprehensively evaluated based on RECIST as in the clinical study. The results showed an objective response rate (ORR) of 33.3% (2/6) in the total group and 66.7% (2/3) in the high-dose group. This was better than that of previously performed clinical studies with lower doses of GEN0101 (3000 and 10,000 mNAU) [9]. In addition, the tumor response also showed dose-dependency with intratumoral administration; the response rate of the target lesions increased from 44% in the low-dose group to 78% in the high-dose group, suggesting that the intratumoral administration of GEN0101 increased a response rate in a dose-dependent manner.

Intralesional Bacillus Calmette-Guerin (BCG) treatment has been established as immunotherapy for malignant melanoma. The long-term survival of stage IV melanoma patients in the extended MMAIT-IV trial showed a median OS and 5- and 10-year survival of 39.1 months and 43.3 and 33.3%, respectively, with BCG plus placebo [17]. Morton et al. reported that 90% of injected lesions achieved CR, whereas 17% of non-injected lesions showed regression [18]. Furthermore, 30% of patients remained disease free for 6–74 months. BCG demonstrated significant potential as intratumor immunotherapy for malignant melanoma. However, severe complications, such as ulceration, skin necrosis, and abscess, have been reported following repeated injection with BCG [19]. The severe skin complications, in particular, have been a major problem for melanoma patients, hampering their ability to safely receive intralesional BCG therapy. T-VEC, which is also intratumorally administered, like BCG and GEN0101, exerts an antitumor effect by expressing GM-CSF and has been approved by the FDA for malignant melanoma. A phase III clinical trial of T-VEC alone for melanoma patients demonstrated 31.5% of ORR (CR; 16.9%, PR; 14.6%) and 19.3% of durable response rate [20]. An overall survival and a median time to CR were 23.3 (19.5–29.6) and 8.6 months, respectively. The ORRs of lesions injected by T-VEC and GEN0101 in this study were equivalent in 67.2 and 61%, respectively [21]. The effect of high-dose GEN0101 may be comparable to that of T-VEC, as the local site response rate in the high-dose group was 78%. Another oncolytic virus therapy, coxsackie A21, was intratumorally administered to patients in a phase II trial of 57 stage IIIC-IVM1c melanoma patients [22]. A total of 38.6% of patients achieved an immune-related progression free survival of 6 months. The ORR was 28.1% in that trial. High-dose GEN0101 appears to exert a similar antitumor efficacy to these oncolytic viruses.

In addition to direct killing, antitumor immunity induced by GEN0101 may be exerted as therapeutic effect. HVJ-E recruits CD8⁺ T cells at the administration site as well as at the non-administration site to evoke antitumor immunity [9]. Therefore, it is considered that antitumor immunity was also induced by GEN0101 in the present trial; an increase in the NK cell activity and IFN-γ was confirmed (Fig. 5). The NK cell activity increased immediately after the administration of GEN0101, peaking at 6 weeks and followed by a gradual decrease in the low-dose group. This tendency was also seen during clinical studies, suggesting that the NK cell response was transient. There was little change in the NK cell activity in the high-dose group, possibly because of the high level of NK cell activity (about 80%) at baseline. IFN-γ levels increased transiently after six administration sessions at two weeks, which was considered to contribute to the activation of CTL. According to an in vitro study, HVJ-E can induce dendritic cell maturation [4]. Tumor-specific T cells were trained by dendritic cells infiltrated from local tumor [4]. CTL against local tumor antigen was also detected in a mouse malignant melanoma model [6]. Based on these findings, we suspect that activated NK cells and CTLs stimulated by the production of IFN-γ spread to distant metastasis and might have thus contributed to the reduced size of lung metastasis (Figs. 4 and 5). Systemic effects of HVJ-E may be mediated by both innate and adaptive immunity in humans. Therefore, the more potent efficacy may be obtained via combination therapy with an immune checkpoint inhibitor.

In terms of immune induction, T-VEC induces CD8⁺ CTL activity [23]. In addition, CD8⁺ T cells existing around the tumor have been reported to be involved in a favorable response to the treatment with anti-PD-1 antibodies in patients with malignant melanoma [24]. Therefore, combination therapies of T-VEC and immune checkpoint inhibitors have been attempted in order to improve the therapeutic effect. In phase II clinical trial of T-VEC in combination with ipilimumab, an anti-CTLA4 antibody, the response rate was 39% in the combination group and 18% in the ipilimumab-alone group, with a significant difference. In that clinical trial, ipilimumab was injected after the intratumoral administration of T-VEC [25]. Intralesional BCG followed by ipilimumab was not tolerated in advanced melanoma and showed no evidence of clinical benefit [26]. In another clinical trial, the combination of T-VEC and ipilimumab showed an ORR of 50% in patients with melanoma [27]. A phase Ib clinical trial was also conducted in combination with another immune checkpoint inhibitor (the anti-PD-1 antibody Pembrolizumab), targeting untreated patients with stage IIIB to IV melanoma [28]. In that clinical trial, an ORR was 61.9%, with 33.3% showing CR. An over 50% reduction in tumor volume was observed in 82% of injected lesions. Interestingly, the response was correlated not with the baseline number of CD8⁺ cells around the tumor but instead with the intratumoral infiltration of CD8⁺ cells after the administration of T-VEC. In Phase II, single-arm, biomarker study of T-VEC monotherapy, T-VEC increased the ratio of CD3⁺/CD8⁺ T cells expressing granzyme B, CTLA-4, and PD-1 [29]. Such mechanism may be one reason for the high response rate obtained in combination therapy. Similar to T-VEC, GEN0101 was predicted to induce the dense intratumoral infiltration of CD4⁺ and CD8⁺ T cells, which may have contributed to tumor cell death, according to previous clinical study [9]. Therefore, GEN0101 combined with immune checkpoint inhibitors may enhance the clinical effectiveness against malignant melanoma. In particular, since GEN0101 was originally developed as a vector, entities such as cytokine genes, anticancer agents, and siRNA are expected to be able to be enclosed in GEN0101, which will induce the more efficient antitumor effect.

In summary, the results of this clinical trial suggest that GEN0101 at 30,000 and 60,000 mNAU is tolerable and effective for the treatment of melanoma patients. The response rates of target lesions in the low- and high-dose groups were 44 and 78%, respectively. This result implies that the antitumor effects on cutaneous target lesions increase in a dose-dependent manner. Furthermore, antitumor immunity in lung metastatic lesion was induced by the local injection of GEN0101. To investigate the antitumor immunity induced by GEN0101 in combination with the immune checkpoint inhibitors, we started a phase Ib/II clinical trial of the combination of GEN0101 and pembrolizumab in patients with advanced melanoma. We expect this phase Ib/II clinical trial to show an improvement in the antitumor effect of GEN0101.