New combinations and immunotherapies for melanoma: latest evidence and clinical utility

Abstract

Until recently there was no effective systemic therapy for metastatic melanoma. Increased understanding of tumor biology and immune regulation has led to the development of drugs targeting the mitogen-activated protein kinase (MAPK) pathway (BRAF inhibitors and MEK inhibitors) and T-cell regulation (CTLA4 antibodies). These drugs are the new standard of care, however barriers to better patient outcomes include limited responses and significant toxicities (CTLA4 antibodies) and lack of durability in the majority of cases (BRAF and MEK inhibitors). This review discusses the next stages of development of treatments in melanoma, including immune checkpoint blocking drugs targeting the PD-1/PD-L1 axis, and the use of BRAF and MEK inhibitors in combination. Both approaches lead to a higher proportion of durable responses coupled with less toxicity. In an effort to improve outcomes even further, clinical trials of combinations of MAPK inhibitors, immunotherapies and other signal pathway inhibitors are underway. Adjuvant studies of many of these drugs have commenced, with the hope of also improving outcomes in patients with early-stage melanoma.

Introduction

Melanoma is curable in the majority of early-stage cases; however, metastatic melanoma carries a poor prognosis [Balch et al. 2009]. Prior to the development of immune and molecular targeted therapies, systemic treatments were ineffective. A rapidly evolving understanding of tumor biology and immunity has provided the basis for a revolution in systemic therapies for melanoma, in particular, the identification of immune checkpoints and prevalent driver oncogenes. Drugs targeting the mitogen-activated protein kinase (MAPK) pathway and CTLA4 have entered routine clinical practice, and were the subject of a recent review in this journal [Khattak et al. 2013]. Building upon the early success of these therapies, trials involving new classes of drugs and combinations of these drugs already in clinical use are underway. This review focuses on the next stage of development of drug therapies and combinations that may improve patient outcomes further.

Combination BRAF and MEK inhibitors

Several trials combining BRAF inhibitors and MEK inhibitors for patients with V600 BRAF-mutant metastatic melanoma are underway, including trials of dabrafenib combined with trametinib [Flaherty et al. 2012], vemurafenib combined with cobimetinib [Gonzalez et al. 2012] and LGX818 combined with MEK162 [ClinicalTrials.gov identifier: NCT01543698] (Figure 1). The rationale behind this approach is twofold: to prolong the progression-free survival (PFS) by delaying or preventing the development of MAPK-dependent resistance mechanisms (reviewed in this journal) [Khattak et al. 2013]; and to reduce BRAF inhibitor related toxicities as a result of paradoxical activation of the MAPK pathway in nonmelanoma BRAF wild-type cells.

Figure 1.

The mitogen-activated protein kinase (MAPK) and phosphatidylinositol 3 kinase (PI3K) signaling pathways and drugs in development. In normal cells, growth factors bind to cell surface receptor tyrosine kinases (RTKs), triggering signaling down various pathways, including the RAS-RAF-MEK-ERK (MAPK) and PI3K-AKT-mammalian target of rapamycin (mTOR) pathways. This signaling results in highly regulated cell proliferation, growth and survival. Specific aberrations in melanomas result in abnormal constitutive activation of the MAPK pathway, and include activating mutations in BRAF (40–50%), NRAS (20%) and KIT (<5%). Several drugs are in development to target these pathways, and are often used in combination. RTK, receptor tyrosine kinase.

The combination of BRAF and MEK inhibitors was first tested in the phase I/II trial of dabrafenib and trametinib (CombiDT) in patients with V600E or V600K BRAF-mutant metastatic melanoma. Initial data showed that response rates were higher with CombiDT than those previously reported for dabrafenib monotherapy [Infante et al. 2011], but only 19% of patients who failed prior BRAFi therapy achieved a response [Flaherty et al., 2011]. The randomized phase II section of this trial showed a higher response rate (76% versus 54%, p = 0.03), a longer median PFS [9.4 months versus 5.8 months, hazard ratio (HR) 0.39, p < 0.001], and fewer oncogenic toxicities in MAPK inhibitor naïve patients on ‘full-dose’ CombiDT 150/2 (dabrafenib at 150 mg twice daily and trametinib at 2 mg daily) compared with dabrafenib monotherapy [Flaherty et al. 2012]. All BRAF inhibitor class toxicities, including hyperkeratosis, alopecia, arthralgia and rash, were less frequent. The rate of cutaneous squamous cell carcinoma (SCC) with CombiDT was one-third that of dabrafenib monotherapy (7% versus 19% respectively). Fever was the most common toxicity and occurred in approximately 70% of patients (5% grade 3/4) treated with CombiDT 150/2, and only 26% of patients (0% grade 3/4) treated with dabrafenib alone [Flaherty et al. 2012; Hauschild et al. 2012]. The mechanism is not understood, but clinically, fever occurs early, is often repetitive and can be managed with brief dose interruption or, in the case of recurrent fever, with corticosteroid prophylaxis. A single institution substudy showed that dose reduction was not effective prophylaxis [Menzies et al. 2012]. Two phase III trials comparing CombiDT 150/2 with dabrafenib (COMBI-D) [ClinicalTrials.gov identifier: NCT01584648] or vemurafenib (COMBI-V) [ClinicalTrials.gov identifier: NCT01597908] monotherapy in patients with V600E or V600K metastatic melanoma are ongoing.

The combination of vemurafenib and the MEK inhibitor cobimetinib (GDC-0973) was tested in a phase I trial of 70 patients with Cobas-positive metastatic melanoma, of which 38 (54%) had disease that failed to respond to prior vemurafenib treatment [Gonzalez et al. 2012]. The Cobas 4800 BRAF V600 mutation test (Roche Molecular Systems, Pleasanton, CA, USA) is a polymerase chain reaction based test that is very sensitive and specific for the more common V600E BRAF mutation, however it only detects 40–70% of V600K, and no other V600 mutations in melanoma tumor samples [Aakre et al. 2012; Anderson et al. 2012; Halait et al. 2012; Kwok et al. 2012; Wenceslao et al. 2012]. Of 25 BRAF inhibitor naïve evaluable patients, all had a reduction in tumor size. In 32 patients previously treated with BRAF inhibitor the response rate was less impressive (19%). As with CombiDT, cutaneous SCCs were much less frequent (1.4%) than reported with single agent vemurafenib (20–25%), while other grade 3 toxicities such as nonacneiform rash (7.1%), arthralgia (4.3%), and fatigue (1.4%) were similar [Chapman et al. 2011; Sosman et al. 2012a]. Reported MEK inhibitor induced adverse events included creatine kinase elevation, diarrhea (4% and 6% grade 3, respectively) and chorioretinopathy (4.3% all grades). Rates of other known MEK inhibitor class effects have not yet been specifically reported, such as cardiac dysfunction and acneiform rash. The recommended dose for phase III testing was vemurafenib 960 mg twice daily, and cobimetinib 60 mg once daily (21 days on, 7 days off) as the maximum tolerated dose (MTD) was reached when cobimetinib was dosed daily. The phase III trial of this combination versus vemurafenib monotherapy is underway [ClinicalTrials.gov identifier: NCT01689519].

Initial results from a phase I trial of the BRAF inhibitor LGX818 [Dummer et al. 2013], as well as the phase I/II trial of the combination of LGX818 and the MEK inhibitor MEK162 [Kefford et al. 2013] in patients with BRAF V600-mutant metastatic melanoma were recently reported. These trials specifically allowed any V600 BRAF mutation, including rare variants such as V600R [Klein et al. 2013]. The LGX818 monotherapy trial enrolled 26 BRAF inhibitor naïve, and 28 BRAF inhibitor pretreated patients. The MTD for LGX818 was 600 mg daily; however, the recommended phase II dose (RP2D) was 450 mg once daily. Initial results show a confirmed response rate of 58% in BRAF inhibitor naïve patients, and 11% in BRAF inhibitor pretreated patients. In contrast to vemurafenib and dabrafenib, photosensitivity, liver transaminase elevations, and pyrexia were rare toxicities. The combination LGX818 and MEK162 trial in BRAF V600 mutant solid tumors included nine BRAF inhibitor naïve and 14 BRAF inhibitor pretreated patients with melanoma treated across a range of doses. MTD was not reached, and two RP2Ds of 450 mg and 600 mg once daily of LGX818 in combination with 45 mg twice daily of MEK162 were declared. The 600/45 dose will be taken forward into the phase II part of trial. Initial results show a confirmed response rate of 88% in BRAF inhibitor naïve patients, and 18% in BRAF inhibitor pretreated patients. No pyrexia, photosensitivity, or SCC have been reported to date. A phase III trial of combined LGX818 and MEK162 versus LGX818 versus vemurafenib is planned.

Given the preliminary data of increased response rates, longer PFS, and less toxicity with combination BRAF and MEK inhibitors, it is anticipated that they will replace BRAF inhibitor monotherapy as the preferred first-line MAPK inhibitor treatment for BRAF-mutant metastatic melanoma in the near future. The results of the first phase III trials of combination therapy should be available in late 2013 (COMBI-V). At present, little is known regarding mechanisms of resistance to combined BRAF and MEK inhibitors, but much research is currently focused on this issue.

Anti-PD-1 and anti-PD-L1 antibodies

Unlike CTLA-4 antibodies, the PD-1/PD-L1 antibodies aim to potentiate the antitumor T-cell response at a tumor-specific level, by impairing the interaction of the inhibitory receptor PD-1 on T cells with PD-L1 expressed on tumor cells (Figure 2 and reviewed by Topalian and colleagues) [Topalian et al. 2012a].

Figure 2.

The PD-1/PD-L1 axis and antibodies in development. T cells interact with tumor cells in peripheral tissues. Tumor cells can present antigen to the T-cell receptor, resulting in a stimulatory signal to the T cell (+). Tumor cells may also express PD-L1, which interacts with PD-1 on activated T cells, and results in inhibition (–) of the antitumor T-cell response.

Nivolumab (BMS-936558, MDX-1106, ONO-4538) is a fully human Immunoglobulin G4 (IgG4) monoclonal PD-1 antibody and was the first of its class to be tested in a phase I trial of 107 patients with metastatic melanoma [Sosman et al. 2012b]. There were no exclusions based on mutation status or histopathological subtype. Although approximately 25% of patients had received three or more lines of systemic therapy, responses were seen throughout the range of doses given every 2 weeks (0.1–10 mg/kg), with an overall RECIST response rate of 31% (41% in the 3 mg/kg group). Most importantly, the median duration of response was over 2 years. Nivolumab was well tolerated. Toxicities were immune related, mild, and less frequent and less severe than those observed with ipilimumab. Common toxicities were fatigue, rash, diarrhea, and pruritis. Grade 3/4 toxicity occurred in 21% of patients, including lymphopenia, fatigue, diarrhea, nausea and anemia (1–3% each). Pneumonitis was a rare but significant adverse event, resulting in the death of three patients without melanoma in the wider trial that included a range of solid tumors (n = 296) [Topalian et al. 2012b]. There was no association between drug dose and efficacy or toxicity, and the 3 mg/kg dose was chosen for the first-line phase III trial of nivolumab versus dacarbazine [ClinicalTrials.gov identifier: NCT01721772] in metastatic melanoma, excluding patients with a history of ocular melanoma.

Lambrolizumab (MK-3475) is a humanized monoclonal IgG4 PD-1 antibody, which was studied in a phase I trial that included 132 patients with metastatic melanoma. Sixty-seven percent of patients had BRAF wild-type melanoma and 9% had brain metastases [Iannone et al. 2012]. Two doses were tested, 2 and 10 mg/kg, given every 2 or 3 weeks. The overall response rate was 51% in the 85 patients with melanoma dosed at 10 mg/kg, using the immune-related response criteria [Wolchok et al. 2009]. Patients responded whether they were ipilimumab naïve with a response rate of 55% (n = 58) or had previously progressed on ipilimumab (response rate 41%, n = 27). Only 15.9% of the full cohort of 132 patients with melanoma developed an immune related adverse event, and only 5.3% were grade 3/4. All grade 3/4 toxicities were observed at 10 mg/kg, and included nephritis (n = 1), pleuritic pain (n = 1), pancytopenia (n = 1), pneumonia (n = 1), abdominal pain/vomiting (n = 1) and thyroiditis (n = 2). Pneumonitis occurred in 3% of patients, all grade 1/2, and was managed with dose interruption, and in one case, steroids. A phase II trial of lambrolizumab at two dose levels versus chemotherapy is underway [ClinicalTrials.gov identifier: NCT01704287], and a phase III trial is planned.

BMS-936559, a fully human IgG4 PD-L1 antibody, was tested in 55 patients with metastatic melanoma as part of the phase I trial [Brahmer et al. 2012]. Fifty-six percent of the patients with melanoma had received prior immunotherapy and 9% had received prior BRAF inhibitor therapy. Infusions were given every 2 weeks. The overall response rate was 17% and ranged from 6% to 29% across dose levels (0.3–10 mg/kg), as assessed by RECIST criteria. Interestingly, the highest response rate was observed at the 3 mg/kg dosage. Of the nine patients who responded, five had an ongoing response for over a year, and overall, 27% of patients had stable disease for over 6 months. Toxicity was generally mild and manageable. From the total patient cohort (n = 207), 9% of patients had grade 3 toxicity and 39% had an immune adverse event of any grade, including rash, hypothyroidism, hepatitis, sarcoidosis, endophthalmitis, diabetes mellitus, and myasthenia gravis. There was no significant difference in toxicities across dose levels, apart from infusion reactions that were more common in those who received 10 mg/kg.

Although not compared head to head, current evidence suggests that the PD1/PDL1 axis inhibitors are more active and less toxic than ipilimumab due to the more tumor-specific mode of immune activation. The possibility that tumor PD-L1 expression may be a biomarker to predict response to PD-1 antibodies makes these drugs even more attractive to patients and physicians. In the nivolumab trial, no responses were seen in 17 patients with solid tumors who did not express PD-L1, while 9 of 25 patients with PD-L1 expression had a response to treatment [Topalian et al. 2012b]. Pretreatment tumor biopsies have been collected and sent centrally for all patients on the lambrolizumab (MK-3475) trial, and the results are awaited. Optimization of the immunohistochemicial assay for detection of PD-L1 is ongoing, but potentially, immunotherapy may also be tailored to patients as occurs with MAPK inhibitors.

Future directions

MAPK inhibitors and new immunotherapies are superior to previous treatment regimens, but they still have major limitations. Although MAPK inhibitors in combination result in responses in almost every patient that are more durable than single agent responses, acquired resistance remains the greatest obstacle, and most patients relapse within a year. Nevertheless, even with targeted therapies, there is a subgroup of patients that may benefit for a prolonged period of time. The 1-year survival for the phase III trial of vemurafenib was 56% [Chapman et al. 2012], the 2-year survival for the phase II trial was 32% [Kim et al. 2012b], and the 3-year survival for the phase I trial was 26% [Kim et al. 2012a]. Preliminary data suggest that PD-1/PD-L1 immunotherapies provide faster and more frequent responses compared with ipilimumab, but the durability of response is currently unknown.

Several new strategies are emerging to treat melanoma based upon lessons learnt from the success and limitations of the aforementioned treatments, and incorporating the results of recent research into melanoma biology, immune regulation, and drug resistance (Figure 1). Several trials of combinations of MAPK inhibitors, immunotherapies, and other signal pathway inhibitors are underway.

One notable combination is that of MAPK inhibitors and immunotherapy. This is based upon individual drug activity, a lack of overlapping toxicity, and translational evidence that BRAF inhibition (and combined BRAF and MEK inhibition) leads to increased expression of melanoma differentiation antigens, and an influx of tumor infiltrating lymphocytes [Boni et al. 2010; Frederick et al. 2013; Wilmott et al. 2012]. The first combination, vemurafenib and ipilimumab [ClinicalTrials.gov identifier: NCT01400451] was terminated due to hepatotoxicity, a side effect of both drugs [Ribas et al. 2013], however other combinations, such as dabrafenib and ipilimumab with/without trametinib [ClinicalTrials.gov identifier: NCT01767454], may not have this issue.

Combination immunotherapy trials are also underway, including ipilimumab and nivolumab in combination [ClinicalTrials.gov identifier: NCT01024231], and in sequence [ClinicalTrials.gov identifier: NCT01783938]. Other immunotherapy combination approaches are planned, such as the combination of ipilimumab and radiotherapy [ClinicalTrials.gov identifier: NCT01689974].

Several other signal pathway inhibitors are being tested in metastatic melanoma either as monotherapy or in combination with MAPK inhibitors. These include inhibitors of receptor tyrosine kinases (platelet-derived growth factor receptor, insulin-like growth factor receptor, MET, vascular endothelial growth factor receptor), the phosphatidylinositol 3 kinase (PI3K) pathway [PI3K, AKT, mammalian target of rapamycin (mTOR)], downstream in the MAPK pathway, and at the cell cycle level (CDK4/6). Examples of trials in progress or planning include vemurafenib with BKM120, a PI3K inhibitor [ClinicalTrials.gov identifier: NCT01512251], vemurafenib with everolimus or temsirolmus, mTOR inhibitors [ClinicalTrials.gov identifier: NCT01596140], LGX818 with LEE011, a CDK4/6 inhibitor [ClinicalTrials.gov identifier: NCT01777776], and MEK162 with AMG479, an insulin-like growth factor 1 receptor antibody [ClinicalTrials.gov identifier: NCT01562899].

Second-generation BRAF inhibitors are also in development attempting to specifically inhibit mutant BRAF, and not activate MAPK signaling in wild-type BRAF cells [Le et al. 2013; Lo et al. 2013]. These ‘paradox breakers’ are expected to have less toxicity and enable greater BRAF inhibition in mutant cells, with the hope of an increased clinical benefit over current BRAF inhibitors, particularly as they may also be active against BRAF splice variants and cells with NRAS mutations, both known BRAF inhibitor resistance mechanisms.

Adjuvant therapy

The greatest role for new systemic treatments may be in the adjuvant setting. The risk of distant relapse and death in patients with high-risk early stage melanoma (IIC/III) is approximately 50% [Balch et al. 2009]. At present, the only approved treatment, interferon, is toxic, and has little impact on overall survival [Mocellin et al. 2010; Petrella et al. 2012]. The impressive activity of MAPK inhibitors and immune checkpoint blockade antibodies in the metastatic setting provides hope that they may reduce the risk of recurrence and improve overall survival when given adjuvantly. A trial of ipilimumab versus placebo [ClinicalTrials.gov identifier: NCT00636168] and a trial of a melanoma antigen (Mage-A3) vaccine versus placebo [ClinicalTrials.gov identifier: NCT00796445] are in follow up. Trials of vemurafenib versus placebo in resected stage IIC/III melanoma (BRIM8) [ClinicalTrials.gov identifier: NCT01667419] and CombiDT versus placebo in resected stage III melanoma (COMBI-AD) [ClinicalTrials.gov identifier: NC-T01682083] have commenced.

Conclusion

MAPK and immune checkpoint inhibitors have revolutionized the management of patients with metastatic melanoma, but further improvements are required to build upon the early success of these therapies. Critical areas requiring rapid progress include efficiently combining the ‘best in class’ drugs already shown to have single-agent activity rather than multiplication of trials using drugs of the same classes; the development of new drugs with new targets for combination therapy; biomarkers to better predict individual patient response to therapy; and adjuvant trials of the most effective drugs and combinations to improve cure rates for early melanoma. In the near future genomic profiling of patient tumors will extend beyond a limited number of genes, with the anticipated effect of unveiling increased targets and eventually increased treatments to improve patient outcomes further.