Initial Use in Melanoma

As with all antineoplastics the first testing of IFNs in melanoma was in the metastatic disease setting. In 1978, the American Cancer Society initiated a multicenter trial testing IFN-α in patients with metastatic melanoma. Forty-five patients were enrolled among whom there was one partial responder and minimal responses in two others. A similar study was performed by Retsas and colleagues in which 17 patients with melanoma were treated with IFN-α and one partial response was seen.

Studies continued as more dosing information became available. Several phase I/II studies were performed with very similar results. In these trials there were 2 responses out of 15 patients, 4 responses out of 23 patients, and 3 responses out of 20 patients. Tumor response rates around 16% were observed with some late responders. It is unknown if there is a survival benefit for IFN in the metastatic setting because no randomized trials comparing it with cytotoxic therapy or supportive care have been performed. Most of the responders had a low tumor volume. This led to the hypothesis that the greatest benefit of IFN-α would be in patients with microscopic residual disease.

Adjuvant Testing

Randomized phase III trials have been performed testing both high-dose IFN-α2b and pegylated (PEG) IFN-α2b in the adjuvant setting in high-risk melanoma patients ( Table 1 ). The first trial was Eastern Cooperative Group E1684, a randomized controlled study of 287 patients with T4 or N1 melanoma. Patients were given IFN-α2b, 20 MU/m ² /d intravenously for 1 month followed by 10 MU/m ² /d three times per week subcutaneously for 48 weeks versus observation. In this study a significant prolongation of relapse-free survival (RFS) ( P = .0023) was observed with an increase from 1.0 to 1.7 years disease-free survival. In addition a significant prolongation in overall survival (OS) ( P = .0237) was also observed with an increase from 2.8 to 3.8 years. The benefit of therapy was greatest among patients with lymph node involvement.

Based on the positive results from E1684 the decision was made to move forward with a comparison of high- and low-dose IFN in an Intergroup trial. E1690 was a three-arm prospective, randomized trial of patients with IIB and III melanoma. A total of 642 patients were enrolled and randomized to high-dose IFN at the same doses as E1984, low-dose IFN (3 MU/m ² /d three times per week subcutaneously for 2 years), or observation. This trial showed a RFS benefit for the high-dose IFN arm by Cox multivariable analysis ( P = .03). However, there was no significant RFS benefit seen for the low-dose arm. In addition there was no OS benefit comparing either the high-dose or low-dose arm with observation. The investigators concluded that these results may have been confounded by the high proportion of patients who received IFN-α2b at progression in the observation arm.

Vaccination is another adjuvant therapy being explored for use in melanoma treatment. The ganglioside GM2 is a serologically well-defined melanoma antigen that has shown efficacy when combined with various adjuvants. Intergroup E1694 compared high-dose IFN alfa-2b versus vaccination with GM2 conjugate. A total of 880 patients with stage IIB/III melanoma were randomized to each treatment arm. The trial was closed early after an interim analysis indicated inferiority of GM2 compared with IFN. IFN had a significant RFS ( P = .0015) and OS benefit ( P = .009) compared with the GM2 conjugate vaccine ( Table 2 ).

Three large meta-analyses of the published results of IFN-based trials in the adjuvant setting were performed with similarly mixed results. The first analysis by Wheatley and colleagues included 12 randomized controlled trials and showed a significant RFS benefit ( P = .000003) and unclear OS benefit ( P = .1). This meta-analysis did not include E1694 because of the control arm being vaccine and suggested an increased benefit for IFN at increased doses.

The second two meta-analyses showed a risk reduction in RFS and OS for IFN therapy. The first by Wheatley and colleagues included 13 randomized controlled trials and showed a 13% risk reduction in RFS and a 10% risk reduction in OS. The final meta-analysis by Mocellin and colleagues from 2010 showed an 18% risk reduction in RFS and an 11% risk reduction in OS. Overall these meta-analyses showed a consistent benefit to the use of IFN in the adjuvant setting for high-risk melanoma.

PEG-IFN

To further reduce the toxicity associated with high-dose IFN there was an effort to see if PEG-IFN alfa-2b would facilitate prolonged exposure with reduced toxicity. In EORTC 18991, a total of 1256 patients with resected stage III melanoma were randomly assigned to observation versus PEG-IFN-α2b at 6 μg/kg per week for 8 weeks followed by 3 μg/kg per week for an intended duration of 5 years. There was a statistical difference in RFS between the two groups ( P = .01). However, there was no difference in OS between the two groups. Based on these results PEG-IFN alfa-2b was approved by the Food and Drug Administration in 2011.

A post hoc meta-analysis of EORTC trials showed that tumor stage and ulceration were predictive factors for the efficacy of adjuvant IFN/PEG-IFN therapy. The efficacy was lower in stage III-N2 patients with ulceration and uniformly absent in patients without ulceration. EORTC 18081 is currently enrolling patients with ulcerated primaries and no nodal disease to 2 years of PEG-IFN versus placebo to confirm this effect.

Toxicity

IFN is a difficult treatment to tolerate and up to 50% of the initial patients in E1684 required a dose reduction or treatment discontinuation. Efforts have been made to improve compliance with therapy and some centers have been able to get 90% of their patients, who do not have early relapse, through 1 year of therapy. However, the toxicity is clearly a limiting factor to adoption and continuation of IFN.

The predominant side effects seen in trials include myelotoxicity, elevation of liver enzymes, nausea and vomiting, flulike symptoms, and neuropsychiatric symptoms. Within the spectrum of flulike symptoms patients often report fevers, chills, anorexia, weight loss, diarrhea, rash, and fatigue. These symptoms can often be managed with supportive care or a dose reduction. However, they are often unacceptable to melanoma patients who tend to be younger and need to continue working.

Patient selection for treatment with IFN is important. It should only be considered in patients with a risk of relapse greater than 30%. In addition it should be avoided in patients with a history of significant heart disease, liver toxicity, or depression.

Adjuvant Dosing and Schedule

Based on the low number of patients in the original trials that were able to complete a full year of therapy the possibility of a shorter course has been explored. The Hellenic Cooperative Oncology group conducted a noninferiority phase III trial to evaluate the efficacy and safety of 4 weeks versus 42 weeks of high-dose IFN in patients with stage IIB, IIC, and III melanoma. This trial showed no significant difference in OS and RFS between the two different regimens.

E1697 assessed the benefit of 4 weeks high-dose IFN compared with observation in patients with intermediate- and high-risk melanomas. The trial was terminated early after 1150 patients were enrolled because an interim analysis showed no improvement in RFS or OS for 4 weeks high-dose IFN versus observation. Based on these results 1 year of high-dose adjuvant IFN remains the standard adjuvant therapy for appropriate patients.

Predicting Response

Unfortunately, adjuvant therapy for melanoma continues to have a limited benefit with substantial toxicity. Attempts to identify patients who would have the highest benefit from adjuvant IFN have been ongoing. It has been observed that patients who exhibit symptoms of autoimmunity are more likely to respond to IFN. Such immune toxicities include hypothyroidism, hyperthyroidism, the antiphospholipid-antibody syndrome, and vitiligo.

Studies to look at the role of autoimmunity in these patients have been ongoing. In particular Gogas and colleagues evaluated 200 patients who were part of a larger randomized trial to look for markers of response. Serum was tested for antithyroid, antinuclear, anti-DNA, and anticardiolipin antibodies, and patients were examined for vitiligo. The development of autoimmunity was an independent prognostic marker for improved RFS and OS ( P <.001).

Further study to uncover why certain patients are more likely to develop autoimmunity than others is ongoing. To date no genetic marker or molecular profile that consistently correlates with IFN response has been found.

Future Directions

As newer immunomodulatory agents and targeted therapy emerge in the metastatic melanoma setting these agents are being transitioned into the adjuvant setting. Novel treatments are being compared with high-dose IFN or placebo. ECOG 1609 is a large trial currently accruing patients to a comparison of high-dose IFN versus two doses of ipilimumab in patients at high risk for recurrence. Enrollment to this trial is ongoing.

Despite controversy and limited adoption, IFN is the standard of care for adjuvant treatment of patients with high-risk melanoma. It remains to be seen if IFN will have a continued role in the future of melanoma treatment.