Outcomes have significantly improved for patients with melanoma in the wake of newer targeted treatments. Yet, brain metastases from melanoma remain a significant unmet need, with up to 60% of patients with advanced melanoma developing brain metastases.
Historically, all brain metastases were treated with surgery and/or radiotherapy.
Now, investigators are looking to separate the treatment approach by the primary tumor type, with many targeted therapies, immunotherapies, and combinations under investigation for the treatment of patients with melanoma-associated brain metastases (MBM).
THE HISTORICAL APPROACH
For primary melanoma, treatment with radiation in the frontline is not suggested as results from studies have shown varying benefits from initial treatment with radiation and it has been suggested that primary melanoma may be radioresistant.2However, for brain metastases, radiation is recommended.
Surgical resection is usually performed on patients with a single, symptomatic brain metastasis.1When the tumors are resectable, surgery has a significant benefit over radiotherapy. The results of a retrospective study of 686 patients with MBM showed a median overall survival (OS) of 8.7 months for patients undergoing surgery compared with 3.4 months with radiotherapy alone.3Among patients who received both, the median OS was 8.9 months (P <.001). Of those who received surgery, most only had 1 brain metastasis.
Beyond a single metastasis or an inaccessible tumor, radiation therapy is the go-to treatment modality with either whole-brain radiotherapy (WBRT) or stereotactic radiosurgery (SRS).1While WBRT does tend to show a benefit in local control, its role is being questioned due to the associated deficits in memory and learning.
The phase III Alliance trial of 213 patients with 1 to 3 brain metastases, most commonly from a lung primary cancer, treated with SRS with or without additional WBRT showed that intracranial tumor control with SRS plus WBRT was 84.9% at 1 year compared with 50.5% with SRS alone (P <.001).4However, patients treated with the combination showed more frequent neurocognitive deterioration than the SRS-alone arm in terms of immediate recall (31% vs 8%, respectively; P = .007), delayed recall (51% vs 20%; P = .002), and verbal fluency (19% vs 2%, P = .02). An OS benefit was not noted from the addition of WBRT. Efforts are ongoing to mitigate these effects by avoiding the hippocampus during radiation, as few metastases occur near the hippocampus, or by adding memantine (Namenda), an NMDA receptor antagonist.5
SRS is considered the best approach for achieving local control of 1 to 3 small brain lesions, and is increasingly being recommended due to the negative effects of WBRT.4,5
Existing targeted agents are increasingly being investigated to see if they can improve the prognosis for patients with MBM as they have for patients without metastases.BRAF-targeted therapies vemurafenib (Zelboraf) and dabrafenib (Tafinlar) have both been tested asBRAFmutations are common in patients with melanoma and are frequently tested for.5
In a pilot study of 24 patients withBRAFV600-mutant melanoma metastatic to the brain, whose metastases were unresectable and previously treated, the median progression-free survival (PFS) was 3.9 months (95% CI, 3.0-5.5) and the median OS was 5.3 months (95% CI, 3.9-6.6) with vemurafenib therapy.6Partial responses (PRs) were noted at both intracranial and extracra- nial sites in 10 patients (24%), and stable disease was seen in 9 (38%). Seven of 19 patients experienced more than 30% intra- cranial tumor regression. The benefit was similar to that seen in patients with melanoma without brain metastases. Patients who responded poorly tended to have activation mutations inPIK3CAor inactivation ofPTEN.5
A phase II study of vemurafenib looked at patients with melanoma with previously treated or untreated brain metastases.7Of the 90 patients in the untreated cohort, 2 achieved an intracranial complete response and 14 achieved PRs, for a best overall response rate of 18%. The median PFS in the brain was 3.7 months in the untreated cohort and 4 months in the previously treated cohort. The median OS was 8.9 months in cohort 1 and 9.6 months in cohort 2. No significant central nervous system toxicities were noted.
In the phase II BREAK-MB trial, 172 patients withBRAFV600-mutated metastatic melanoma were treated with dabrafenib, separated into a cohort for patients whose brain metastases were untreated and a cohort for patients with recurrent disease following treatment for their brain metastases.8In the first cohort, 39.2% (95% CI, 28.0%-51.2%) of patients with V600E mutations and 6.7% (95% CI, 0%-31.9%) of patients with V600K mutations experienced an overall intracranial response. In the second cohort, 30.8% (95% CI, 19.9%-43.4%) of patients with V600E mutations and 22.2% (95% CI, 6.4%-47.6%) of patients with V600K mutations experienced a response. Frequent serious adverse events (AEs) included pyrexia; intracranial hemorrhage, 1 of which was treatment-related; and squamous-cell carcinoma (6% each).
Checkpoint inhibitors have been incorporated into treatment for patients with melanoma and could be beneficial for patients with MBM. In a nonrandomized phase II trial, patients with untreated or progressive brain metastases from either melanoma (n = 18) or nonsmall cell lung cancer were treated with the PD-1 inhibitor pembrolizumab (Keytruda).9Responses were achieved in 4 patients with melanoma (22%; 95% CI, 7%-48%). Grade ≥3 treatment-related or serious AEs included colitis, pneumonitis, fatigue, and grade 4 hyperkalemia (6% each). Significant neurological AEs from the melanoma cohort included transient grade 3 cognitive dysfunction and grade 1-2 seizures.
In a phase II study of ipilimumab (Yervoy), a CTLA-4 checkpoint inhibitor, 14 of 72 patients achieved disease control in the brain, including 2 patients whose brain tumors were symptomatic and who were receiving corticosteroids.10Activity was more frequent in patients whose metastases were small and asymptomatic.
BecauseBRAFinhibition and radiotherapy have separately shown benefit among patients with MBM, the combination is also being tested in these patients, asBRAFinhibitors may have the ability to increase radiosensitivity in melanoma.2
A study of 52 patients with MBM treated with SRS with or withoutBRAFinhibitor therapy showed increased local control for patients receiving combination therapy over radiotherapy alone.11Patients withBRAFmutations who were treated with the combination showed a local control rate of 85.0% at 1 year compared with 67.1% inBRAFwild-type patients treated with the combination.BRAF-mutant patients treated with SRS alone had a 1-year local control rate of 51.5% (P = .0077). However, the OS rate and rate of distant brain failure did not vary significantly.
In another study of patients treated with SRS with or without aBRAFinhibitor, the median survival times were greater among patients withBRAFmutations treated with the combination compared with wild-type patients (P <.01) and patients withBRAFmutations treated with SRS alone.12
Increased potential neurotoxicity from this combination is of concern. Individual cases of radiation necrosis and intracranial hemorrhage following therapy have been noted, although overall rates are unknown.2To mitigate these effects, the ECOG group recommends holdingBRAFinhibitor therapy for ≥1 day before and after SRS and for ≥3 days before and after fractionated radiosurgery.13
Patients with melanoma have demonstrated strong responses toBRAFinhibition combined with a MEK inhibitor to reduce resistance toBRAF-targeted therapy. A trial of patients with metastatic melanoma, although not focused on patients with MBM, demonstrated an objective response rate of 64% from dabrafenib and trametinib (Mekinist) compared with 51% from vemurafenib monotherapy (P <.001) in patients with aBRAFV600 mutation.14
The phase II NIBIT-M1 study of ipilimumab with fotemustine showed clinical activity in patients with metastatic melanoma, including in those with MBM.15Ten patients with asymptomatic brain metastases (50%) achieved disease control, including 5 whose metastases became undetectable by imaging.
Radiotherapy combined with immunotherapy is also considered as radiotherapy could increase the permeability of the blood-brain barrier for checkpoint inhibitors. The results of retrospective studies of patients treated with ipilimumab and radiotherapy have shown that the combination is more effective than radiotherapy alone in patients with brain metastases.1
A retrospective analysis of 2 studies examined 26 patients with MBM who were also treated with SRS within 6 months of receiving nivolumab therapy.16Local brain metastasis control at 1 year following SRS treatment was 85% overall. Patients with unresected metastases showed a median OS of 11.8 months from the date of SRS therapy and 12 months from the date of nivolumab initiation. Neurotoxicity was experienced in 1 case of grade 2 headaches following SRS. No other treatment-related neurotoxicities were reported.
COMBINATION OUTCOME COMPARISON
A retrospective study of patients with MBM compared outcomes of 96 patients treated with 1 of 5 combination approaches with SRS: antiPD-1 therapy (n = 21), anti–CTLA-4 therapy (n = 25),BRAFinhibitor therapy (n = 18),BRAFand MEK inhibitors (n = 12), and SRS with chemotherapy (n = 20).17Patients were included in the analysis if their metastases were unresected and had received single-fraction SRS within 3 months of systemic or cytotoxic therapy.
Distant MBM control rates were 38%, 21%, 8%, 20%, and 5%, respectively (P = .008) at 1 year. No significant differences in local control were noted between the treatment arms (P = .25).
Following treatment with radiotherapy, systemic PFS rates at 1 year were 41%, 27%, 12%, 39%, and 5%, for combination therapy with antiPD-1 therapy, anti–CTLA-4 therapy,BRAFinhibitor therapy,BRAFand MEK inhibitors, and chemotherapy, respectively (P = .04). At 1 year, the OS rates following SRS were 48%, 41%, 24%, 65%, and 10%, respectively (P = .01).
The comparison demonstrates that combination therapy withBRAFand MEK inhibitors and SRS, a PD-1 inhibitor and SRS, or with a CTLA-4 inhibitor and SRS all provide increased benefits in patients with MBM. Each of these combination approaches significantly improved OS rates on both univariate and multivariate analyses compared with chemotherapy.
The authors noted that systemic treatment combined with radiation was well tolerated with no significant neurotoxicites reported.
As many of these trials were retrospective, further study is required to determine the true value of these approaches and which is best for patients with MBM.