Introduction

Patients diagnosed with advanced cutaneous melanoma are at significant risk of developing melanoma brain metastases (MBM). Melanoma results in the third highest number of cancer patients who develop brain metastases, trailing only more common cancers, such as lung and breast cancer [1]. However, the actual percentage of metastatic melanoma patients who eventually develop brain metastases, termed incidence proportion percentage, is currently the highest of all cancers [2]. In the past, approximately 28-40% of melanoma patients presented with brain metastases at the time their metastatic disease was diagnosed [2,3]. In a more recent literature review, 33% of patients still were found to have MBM at the time of initial diagnosis of metastatic cutaneous melanoma [4]. Historically, another 25%-44% of metastatic melanoma patients developed the delayed onset of MBM, even if brain metastases were not present at the time treatment was initiated [4-7]. Development of either initial or delayed MBM was associated with increased morbidity and mortality [8-10]. Approximately 50% of metastatic melanoma-related deaths resulted from central nervous system (CNS) progression [11].

Recent advances in cancer therapy, especially the development of targeted therapy (TT) and immune checkpoint inhibitor (ICI) therapy have markedly improved the survival of patients with metastatic melanoma [12-15]. Even in patients with existing MBM, both ICI and TT have demonstrated intracranial responses [16,17]. However, the combination of CTLA-4 and PD-1 blockade (ipilimumab plus nivolumab) appears to provide more durable activity in the CNS when compared to TT [18,19]. Due to these treatment advances, the frequency and timing of brain metastases appears to be evolving. Our own institutional data has shown that the incidence of delayed brain metastases following any form of ICI-based cancer immunotherapy appeared to be low (~7%) [20].

As the incidence of melanoma brain metastases appears to be decreasing following immunotherapy, we identified small number of patients with onset of brain metastases after initial ipilimumab plus nivolumab treatment to further evaluate their characteristics and clinical outcomes.

Materials and Methods Patient Enrolment

Study candidates were identified by a retrospective chart review of a Health Insurance Portability and Accountability Act (HIPAA)compliant IKnowMed medical record program (McKesson, Houston, TX). A search of this data base was conducted to identify patients treated by a single physician (WS) with ipilimumab and concurrent nivolumab. These records were individually accessed to verify that each patient had metastatic cutaneous melanoma. Patients who were treated with ipilimumab plus nivolumab for other cancers were excluded. Additional exclusion criteria included a diagnosis of non-cutaneous melanoma (e.g. acrallentiginous, ocular, or mucosal melanoma). Patients who already had melanoma brain metastases (MBM) at the time of their initial diagnosis of metastatic melanoma were excluded from this analysis. Patients who received ipilimumab and nivolumab as 2nd or later lines therapy for metastatic disease were also deemed ineligible. Patient records were individually accessed and relevant data extracted into a spreadsheet for analysis (Microsoft Excel version 16.83, Redmond, WA). After data entry, the spreadsheet was de-identified. Since this was a retrospective review of existing data, there was no plan to modify treatment based on the results or contact patients directly. This study design was reviewed by the Western IRB chair and deemed exempt from full Institutional Review Board (IRB) review.

Data Collection

Details recorded including age at the start of ICI treatment, gender, and comorbidities. Description of the primary site and initial sites of metastatic disease were used to define the melanoma stage at the time of diagnosis. Pertinent pre-treatment variables, such as lactate dehydrogenase levels, PD-1 ligand (PDL1) expression, tumor mutational burden (TMB) was recorded. The “driver” mutation status identified by Next-Gen molecular sequencing (including BRAF, NRAS, NF-1, KIT mutations, or “quadruple negative”) was also recorded.

Treatment Regimens

All patients in this series were treated with first line ipilimumab plus nivolumab therapy. Patients were treated either with standard regimen (ipilimumab 3 mg/kg with nivolumab 1 mg/kg every 3 weeks) [21], or the alternate dosing regimen (ipilimumab 1 mg/ kg with nivolumab 3 mg/kg every 3 weeks) [22]. Doses were generally rounded up to the nearest vial size, as the toxicity and effectiveness of these agents is similar across a broad dose range. The number of induction and maintenance doses administered were recorded. Detailed information regarding the treatment course was collected, including the treatment start and end dates and subsequent treatments administered following the initial ICI therapy.

Response Assessment

Information about the best objective response rate (BORR) at one year after the initiation of therapy was recorded, as well as data on ICI toxicity and steroid usage. The study evaluated the response to treatment assessed one year from the start of ipilimumab plus nivolumab therapy using the RECIST 1.1 criteria [23]. This assessment categorized patients as having a complete response (CR) if there was no detectable disease, a partial response (PR) if there was more than a 30% reduction in disease, stable disease (SD) if there was less than a 30% reduction or less than a 20% increase in disease, and progressive disease (PD) if there was more than a 20% increase in the index lesion or the presence of new lesions. This assessment considered both primary and metastatic lesions. The date of progression, MBM response status and date and cause of death (if applicable) were noted. If patients were alive, the date of the last follow-up was recorded. Data analysis concluded on September 11, 2023.

Statistical Analysis

Progression-free survival (PFS) and overall survival (OS), from the date of treatment start to the date of progression or date of death, was analyzed via the technique of Kaplan and Meier [24]. Descriptive statistical measures, including the median and standard deviation were calculated via the Excel spreadsheet.

Results

From 2017 to 2022, 171 patients were treated at our institution with ipilimumab plus nivolumab combination therapy. This included 75 patients who received ipilimumab plus nivolumab therapy as firstline therapy for metastatic cutaneous melanoma.

Of these 75 patients, 29 received the standard ipilimumab plus nivolumab regimen (38.7%) while 45 patients received the alternate regimen (60%). An additional patient received the standard regimen and then was switched to the alternate regimen (1.3%), due to insurance requirements. Overall, 32 patients achieved a CR (42.6%) following ipilimumab plus nivolumab treatment. This included 13 patients who received standard dosing regimen (40.6%), and 19 patients received the alternate dosing regimen (59.4%). Four of the 29 patients who received the standard regimen developed delayed MBM (13.8%). Only 3 of the 45 patients who received the alternate dosing regimen developed delayed MBM (7.1%)(p>0.05). The demographics of these 7 patients that developed delayed onset MBM following the start of ICI treatment are shown (Table 1).

All seven patients were Caucasian. Their median age was 63 (± 17.5 SD) years old (range 35-86). Four patients were female (57.1%), and 3 patients were male (42.9%). Of the patients who developed delayed onset of brain metastases, 4 patients had BRAF V600 mutation (57.1%), 1 patient had an NRAS Q61R mutation, 1 patient had a NF-1 R1362* truncation mutation, and 1 patient had a KIT A829P mutation. At the initial start of ipilimumab plus nivolumab treatment, 6 patients had normal lactate dehydrogenase (LDH) levels (85.7%), and only 1 patient had an elevated LDH levels. Two patients were receiving steroids during initial CKI therapy (28.6%). All 7 patients experienced some degree of immunotherapy related toxicity, such as fatigue, rash, fever, headache, nausea, diarrhea, and colitis. Five of these 7 patients also received other forms of treatment after diagnosis of MBM, such as BRAF/MEK inhibitors (3), whole brain radiation therapy (WBRT) (3), surgical resection (2), radiosurgery (1), and nilotinib (1) (Table 2). The remaining two patients deteriorated so quickly that they did not receive further therapy.

Patient Outcomes:

Individual patient treatment outcomes for the seven patients are shown (Table 2). The median length of time from treatment initiation to diagnosis of MBM was 4.8 months (± 7.7 months) (Figure 1). Only one patient developed CNS progression more than 1 year following the start of treatment.

This BRAF mutant patient had an unusually complicated clinical course. Initial ipilimumab plus nivolumab therapy resulted in the development of fevers and hepatic toxicity, leading to early cessation of treatment. The patient subsequently received multiple cycles of treatment with TT after multiple episodes of systemic disease progression. Following eventual CNS progression, the patient had surgical resection of a CNS lesion with stereotactic radiotherapy to the resection cavity. Subsequently, there was development of additional multifocal brain metastases. At this time, the patient was retreated with ipilimumab plus nivolumab with addition of whole-brain radiotherapy, prior to eventual demise due to brain metastases.

We evaluated treatment response and survival in patients with delayed CNS progression. Median progression-free survival (PFS) was only 4.8 ± 4.4 months from the date of CNS progression (Figure 2A). There were no clinical responders to salvage therapies (Table 2). The development of delayed-onset brain metastases had a very adverse effect on patient outcomes. The median overall survival (OS) from the diagnosis of MBM in these 7 patients was 8.4 ± 10.6 months (Figure 2B). All patients with delayed-onset MBM died due to their CNS disease, rather than from systemic progression.

Table 1: Patient Demographics.

Table 2: Patient treatment outcome.

Figure 1: The median length of time from treatment initiation to diagnosis of MBM.

Figure 2A: Median progression-free survival (PFS) was only 4.8 ± 4.4 months from the date of CNS progression.

Figure 2B: The median overall survival (OS) from the diagnosis of MBM.