Intralesional therapy for cancer—that is, direct injection of an anti-cancer drug into the lesion—was first taken seriously back 1975. A story in Cancer reported on a 77-year-old patient with 64 intracutaneous metastases and a pulmonary metastatic deposit. Over an eight-month period, he received inoculations with Bacille Calmette-Guérin (BCG) in 17 of his lesions, and each resolved. At the same time, the pulmonary metastasis regressed more than 50 percent. The treatment was locally ablative and, at the same time, it induced host immune anti-tumor activity in regional and distant uninjected metastases through a systemic adjuvant response.
[See Mastrangelo MJ, Bellet RE, Berkelhammer J, et al. “Regression of pulmonary metastatic disease associated with intralesional BCG therapy of intracutaneous melanoma metastases.” Cancer 1975; 36:1305–1308.]
Passion for the approach cooled after a subsequent trial, in which researchers found anaphylactic reactions and death occurred due to disseminated BCG. When randomized trials of BCG also failed to confirm a significant clinical benefit, intralesional therapy went out of fashion.
Directly attacking the lesions, however, remains an obvious way to treat cancers, and in the treatment of melanoma, there are three candidate drugs currently under investigation as intralesional agents: OncoVEXGM-CSF, now called T-VEC (Talimogene laherparepvec), Allovectin-7 and Rose Bengal (RB) disodium.
OncoVEXGM-CSF, now called T-VEC (Talimogene laherparepvec), is Amgen’s entry in the intralesional sweepstakes and is an investigational oncolytic immunotherapy designed to selectively replicate in tumor tissue and to initiate a systemic anti-tumor immune response.
In early 2014, Amgen announced findings from a pre-specified retrospective analysis of patients with metastatic melanoma that showed T-VEC reduced the size of injected tumors and also non-injected tumors that had metastasized to other parts of the body.
The analysis recorded tumor-level responses from a Phase 3 study evaluating T-VCE in patients with injectable unresected stage IIIB, IIIC or IV melanoma compared to granulocyte-macrophage colony-stimulating factor (GM-CSF).
Of the 295 patients treated with T-VCE, almost 4,000 tumor lesions were tracked for this analysis. Half of these lesions were injected with T-VEC at least once, while the rest were not injected, including visceral tumor lesions (tumors involving solid organs such as the lungs and liver). The results showed a 50-percent or greater reduction in tumor size in 64 percent of injected tumors. In addition, one-third of uninjected non-visceral tumors and 15 percent of visceral tumors were also reduced by at least 50 percent. There were 35 melanoma-related surgeries performed during this trial, of which 30 percent successfully removed all residual disease.
The most frequently observed adverse events in the Phase 3 study were fatigue, chills and pyrexia. The most common serious adverse events include disease progression in both groups, and cellulitis and pyrexia in the talimogene laherparepvec group. Serious adverse events occurred in 26 percent of talimogene laherparepvec patients and 13 percent of GM-CSF patients. Immune-mediated events were reported infrequently.
T-VEC did stumble slightly by missing on overall survival, a secondary endpoint in the Phase 3 study. However, Amgen has filed with the FDA for approval and with the European Marketing Agency. In addition, it has started at trial in collaboration with Merck to test T-VEC in combination with Merck’s newly approved checkpoint inhibitor Keytruda.
A less happy outcome was in store for Vical’s Allovectin-7 (velimogene aliplasmid). For nearly two decades, velimogene aliplasmid was under investigation. It is an HLA-B7/beta2-microglobulin plasmid formulated with cationic lipids. It increases the ability of the immune system to recognize cancer cells and kill them. In 1999, the U.S. FDA granted Allovectin-7 orphan drug designation for intralesional treatment of invasive and metastatic melanoma. What was particularly promising was the fact that its mechanism of action is not melanoma-specific; it has the potential to be used in other types of solid tumors.
In 2010, Vical reported encouraging results in Phase 2 trial, a single-arm, open-labeled study in which 127 chemo-refractory or chemo-intolerant patient subjects were treated with high-dose Allovectin-7. The median age of patients enrolled in the study was 60, and patients as old as 98 were treated with Allovectin-7, yet there were no treatment-related Grade 3 or Grade 4 adverse events, and no withdrawals from the trial for tolerability. The overall response rate for the 127 patients receiving the high-dose treatment was 11.8 percent, with four complete responders and 11 partial responders. The median duration of response was 13.8 months, ranging from a minimum of six months to a maximum of 66 months and still ongoing. Median survival was 18.8 months. These data compared favorably against historical controls from other studies in metastatic melanoma. Findings from the Phase 2 trial were incorporated into the design of a Phase 3 pivotal trial through a Special Protocol Assessment agreement with the U.S. FDA.
The Phase 3 trial, initiated in January 2007, evaluated Allovectin-7 as first-line therapy in patients with Stage III or IV recurrent metastatic melanoma. Vical completed enrollment in February 2010 of approximately 390 chemo-naive patients randomized on a 2:1 basis: approximately 260 for treatment with Allovectin-7 and approximately 130 for treatment with either dacarbazine or temozolomide.
Unfortunately for patients and the company, Vical announced on August 12, 2013, that the trial did not pan out as hoped. It failed to demonstrate a statistically significant improvement compared to first-line chemotherapy for either the primary endpoint of objective response rate at 24 weeks or more after randomization or the secondary endpoint of overall survival.
Rose Bengal (RB) disodium (PV-10)
Provectus Biopharmaceuticals' PV-10 is a preparation of rose bengal (a dye used by eye doctors, and before that, in clothing) currently being investigated at both St. Luke's Cancer Center and Moffitt Cancer Center. In a Phase 2 PV-10 trial, when all existing lesions were injected with PV-10, tumors were no longer detectable (complete response) in 50 percent of the patients (Confidence Interval: 31-69 percent). This subgroup analysis supports the potential of PV-10 as a single agent and provides a rationale for a PV-10 Phase 3 randomized controlled trial in locally advanced melanoma patients.
In the peripheral blood of patients after PV-10 injection, researchers saw a significant increase in circulating T-cells, including CD3+ and cytotoxic CD8+ cells. This suggests an immunologic-mediated antitumor response is engendered by PV-10.
The study showed PV-10 elicits a high rate of response in injected tumors through its ablative effect, and, additionally, that the durability of response as well as the bystander response in uninjected tumors implicate an additional immunologic mechanism secondary to ablation. This appears to be similar to the experience noted in Cancer back in 1975.
Tumors were no longer detectable (complete response or CR) in 26 percent of the study population. This response was particularly evident in patients who had all existing lesions injected with PV-10 (i.e., All Lesions Treated subgroup, 50 percent CR; Confidence Interval: 31-69 percent). These 28 patients had as many as 20 lesions confined to the skin and experienced a mean progression-free survival of 9.8 months. For an additional 26 patients who had all their disease treated, with the exception of 1-2 designated, untreated bystander lesions, mean progression-free survival was 8.9 months.
A separate study of PV-10 assessed response of injected and uninjected B16 melanoma tumors in mice receiving PV-10 alone or in combination with one of three agents designed for co-inhibitory blockade. The tested agents targeted either CLTA-4, PD-1 or PD-L1, the three most common clinical targets for co-inhibitory blockade. In each case, combination of PV-10 with co-inhibitory blockade led to improved tumor response and enhanced anti-tumor immunity of T-cells. Further testing with the anti-PD-L1 agent showed that these improvements could apply to both injected and uninjected tumors.
Based on the findings from St. Luke's and from Moffitt, Provectus has submitted a Phase 3 protocol to the FDA for a Phase 3 randomized controlled trial of PV-10 in patients with unresectable locally advanced cutaneous melanoma that will assess response to PV-10 versus that of systemic chemotherapy in patients who have disease limited to cutaneous and subcutaneous sites and who have failed or are ineligible for systemic immunotherapy. In addition, Provectus is investigating the efficacy of PV-10 for other indications including liver, breast and colorectal cancers.
Injecting a tumor with a cancer-killing drug seems to make perfect sense, but as research has shown, it matters a great deal what the drug is. A generation ago, intralesional BCG appeared to work, but when serious study of it showed problems, researchers walked away, not only from that use of the drug but also from the approach itself. Now, intralesional treatments are making a comeback in clinical trials because the drugs being used appear to be better suited for it than BCG was.
Peter R. Culpepper is chief financial officer and chief operating officer at Provectus Biopharmaceuticals. He has spent 20 years in the financial field working for a wide range of companies and industries in the United States and abroad, especially high-growth startups. This commentary originally ran on our Cancer Research News website at www.ddncancer.com.