Introduction
Brain metastases are a frequent and potentially devastating complication of metastatic melanoma. In the past, up to one third of metastatic melanoma patients presented with brain metastases at diagnosis and nearly 75% of patients developed brain metastases during the course of illness [1]. Melanoma patients with brain metastases historically have had a very poor outcome [2], with previous studies reporting a median overall survival of only 3–13 months following the onset of brain metastases [3, 4, 5, 6]. Patients with concurrent systemic metastatic disease have inferior survival versus patients with an isolated central nervous system (CNS) recurrence [7].
The development of immune checkpoint inhibitors (ICI), such as anti-CTLA-4 and anti-PD-1 antibodies, significantly improved treatment outcome in patients with metastatic melanoma. For example, 6.5-year follow-up of the Checkmate-067 trial comparing the combination of ipilimumab plus nivolumab, to monotherapy with either agent alone demonstrated a 31% median progression free survival at 10 years. A melanoma-specific survival of 52% was achieved in the combination arm, which was superior to either monotherapy [8]. While the Checkmate 067 study allowed inclusion of patients with brain metastases, the small number of enrolled patients with brain metastases were not analyzed separately.
Anzeige
Previous case reports identified the potential for intracranial response following ICI monotherapy [9, 10, 11]. Subsequently, several clinical trials have confirmed the therapeutic benefit of combination ipilimumab plus nivolumab based immunotherapy in patients with asymptomatic brain metastases [1, 2, 3, 12]. Treatment of patients with clinically symptomatic brain metastases has generally produced an inferior outcome [1, 13].
The outcome of brain metastases in patients with metastatic melanoma treated with ipilimumab plus nivolumab therapy is evolving. The aim of our study was (1) to determine the current frequency and timing of onset of brain metastases and (2) to characterize the outcome of delayed brain metastases following initial ipilimumab-nivolumab therapy.
Materials and methods
Patient identification
Potential subjects for this retrospective analysis were identified by a search of a secure, password protected Health Information Portability and Accessibility Act (HIPAA)-compliant iKnowMed database (IKM-G2, McKesson, Houston, TX). This database was searched for patients who received treatment with concurrent ipilimumab and nivolumab between 1 January 2013 to 31 March 2022. Patients were eligible for inclusion in this retrospective study, if they were previously untreated and had metastatic cutaneous melanoma. Patients who had brain metastases at initial presentation of metastatic disease or following initial ipilimumab plus nivolumab therapy were included. Patients with non-cutaneous melanoma (uveal, mucosal, or acral-lentiginous) were excluded from this analysis. Patients who received combination ipilimumab and nivolumab therapy as 2nd or later lines of therapy were also excluded.
Each patient record was individually accessed, and a unique patient number was assigned. Other demographic data included race, gender, site of primary melanoma and number of organs involved with metastases. Data was recorded in an Excel spreadsheet (Microsoft, Redmond, WA version 16.82). Characterization of non-overlapping somatic mutations in BRAF, NRAS, NF1, CKIT was recorded. Following completion of data extraction, patient identifying information was deleted. This study design was reviewed by the Western IRB chair and was deemed exempt from full IRB review.
Anzeige
Treatment characteristics
Treatment consisted of either the original regimen of ipilimumab plus nivolumab (3 mg/kg ipilimumab plus 1 mg/kg nivolumab IV every 3 weeks for 4 doses) [14], or an alternate dosing regimen (1 mg/kg ipilimumab plus 3 mg/kg nivolumab IV every 3 weeks for 4 doses) [15]. Both regimens were followed by nivolumab maintenance (either 3 mg/kg or 240 mg fixed dose every 2 weeks, or a fixed dose of 480 mg monthly). The specific ipilimumab plus nivolumab regimen utilized, treatment start date, cumulative number of doses, and treatment end date were noted.
ICI doses were generally rounded to the nearest higher vial size, due to the wide effective dose range with minimal differences in toxicity. Treatment was continued until disease progression. If patients achieved a radiologically or pathologically confirmed complete remission, elective treatment discontinuation was considered [16].
Response assessment
Responses were described using RECIST 1.1 criteria [17]. Data collection concluded 1 August 2022 (with a minimum potential follow-up of 18 months).
The date of progression or the date of last clinic follow-up (if in remission) was recorded. If a subject died, the date of death and cause of death were recorded. Progression-free survival was calculated from the date of onset of brain metastases detected by radiologic imaging. CNS progression-free survival was defined as the time in which a patient did not develop new onset brain metastases or experience hyperprogression of existing brain metastases following completion of ICI therapy. Overall survival was also calculated from the date of onset of brain metastases.
Statistical analysis
Descriptive statistics, such as median, standard deviation, and data range were calculated. Overall survival (OS) and progression free survival (PFS) rates were analyzed using the Kaplan and Meier method [18]. Comparison between groups was performed using a log-rank test [19]. Analysis of potential predictive tests was performed using the student’s t-test [20].
Results
A total of 73 patients received ipilimumab plus nivolumab as their initial therapy for metastatic cutaneous melanoma in our clinic. Only 20 patients developed brain metastases (27.4%). These 20 patients were further analyzed. Fourteen of 73 patients (19.2%) were found to have brain metastases at the time of initial diagnosis of metastatic disease. Only 6 patients of the remaining 59 (11.2%) developed delayed onset of brain metastases following initiation of ipilimumab plus nivolumab treatment.
Baseline patient clinical characteristics are shown (Table 1). All 20 of the patients with brain metastases included in this study were Caucasian, and 55% were male. The median age of patients was 64.5 ± 15.9 years (± SD) years old (with a range from 28 to 86 years old). All patients underwent testing for potentially targetable somatic mutations. Most patients underwent a broad panel of oncogene mutation testing by Next Gen sequencing (Foundation Medicine CDx testing, Foundation Medicine, Cambridge, MA). The most common non-overlapping “driver” mutation identified in patients who developed brain metastases was a BRAF V600E mutation in 12 patients (60% of all MBM patients). An NRAS point mutation was present in 3 patients (15% of MBM patients). NF1 mutations were found in 2 patients (10% of MBM patients). A c-KIT mutation was found in 1 patient (5%). Two additional patients (10%) underwent only limited point mutation testing in specific oncogenes and thus their mutational status was not fully characterized.
Table 1
Patient demographics
UPN | Age | Sex | Stage | “Driver” Mutation Status | KPS | LDH | Primary site | Initial sites of Metastases | Brain metastases |
|---|---|---|---|---|---|---|---|---|---|
1 | 57 | F | IV | BRAF V600E | 90% | 394 | Hand | Brain, LN, SQ | at onset |
2 | 48 | M | IV | BRAF V600E | 80% | 1454 | Abdomen | Brain, liver, LN, lung, SQ | at onset |
3 | 86 | M | IV | BRAF V600E | 70% | 892 | Leg | Brain, LN | at onset |
4 | 82 | M | IV | No BRAF, NRAS, KIT¶ | 90% | 402 | Unknown | Brain, liver | at onset |
5 | 48 | M | IV | BRAF V600E | 90% | 210 | Back | Brain, bowel, LN, stomach | at onset |
6 | 77 | M | IV | No BRAF V600E¶ | 90% | 231 | Ear | Brain, lung | at onset |
7 | 46 | M | IIC | NRAS Q61K | 50% | 218 | Trunk | Brain, lung | at onset |
8 | 78 | F | IV | BRAF V600E | 80% | 206 | Trunk | Brain, stomach | at onset |
9 | 52 | F | IIIB | BRAF V600E | 90% | 164 | Arm | Brain, LN | at onset |
10 | 78 | M | IIA | NF1 E977* | 90% | 198 | Arm | Brain, lung | at onset |
11 | 69 | F | IV | BRAF V600E | 80% | 189 | Unknown | Brain | at onset |
12 | 52 | M | IV | BRAF V600E | 100% | 235 | Trunk | Bone, brain | at onset |
13 | 68 | M | IV | BRAF V600E | 70% | 217 | Back | Bone, brain, gallbladder, LN, lung | at onset |
14 | 67 | M | IIA | NF1 T257 fs*48 | 100% | 312 | Trunk | Brain, lung | at onset |
15 | 28 | F | IIA | BRAF V600E | 100% | 138 | Trunk | Adrenal, bone, brain, LN | delayed |
16 | 62 | F | III | NRAS Q16R | 90% | 298 | Trunk | Brain, liver, LN, lung | delayed |
17 | 85 | F | IV | BRAF V600E | 90% | 362 | Unknown | Brain, liver, LN | delayed |
18 | 71 | M | IIID | c-KIT A829P | 100% | 163 | Trunk | Adrenal, brain, gallbladder, lung, SQ | delayed |
19 | 60 | F | IVC | NRAS Q61R | 90% | 407 | Trunk | LN, lung, liver | delayed |
20 | 43 | F | III | BRAF V600E | 90% | 223 | Trunk | Adrenal, brain, lung, retroperitoneal | delayed |
Anzeige
The clinical outcome of our patients with brain metastases is shown (Table 2). Eleven of fourteen patients with brain metastases at the time of diagnosis of metastatic disease underwent initial surgical resection of brain metastases followed by immunotherapy with ipilimumab plus nivolumab. Adjuvant radiosurgery to the resection cavity was then performed. Any additional brain metastases were also treated with radiosurgery. Thirteen out of fourteen patients with MBM at presentation (92.9%) were diagnosed due to the development of neurologic symptoms. Ten of these 14 patients were receiving systemic corticosteroids at the time of ICI treatment.
Table 2
Outcomes of ipilimumab and nivolumab treated patients with brain metastases
UPN | ICI Regimen | ICI Doses | Time to onset (mo) | Number of brain mets | Neurologic symptoms | Steroids prior to ICI? | Steroid Dosing | Surgical resection | PFS (mo) | CNS PFS (mo) | OS (mo) | Current Status |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
1 | I + N (s) | 4 | 0 | 2 | Y | None | N/A | Yes | 7.8 | 21.3 | 21.3 | NED |
2 | F I + N (f) | 1 | 0 | 1 | Y | DX | 4 mg, bid | No | 1.3 | 1.5 | 2.6 | DOD |
3 | I + N (f) | 2 | 0 | 2 | Y | DX | 4 mg, bid | Yes | 3.7 | 3.7 | 3.7 | DOD |
4 | I + N (s) | 3 | 0 | 2 | Y | DX | 4 mg, bid | Yes | 10.1 | 10.1 | 10.1 | DOD |
5 | I + N (f) | 3 | 0 | 1 | Y | DX | 2 mg, daily | Yes | 10.5 | 10.5 | 10.5 | NED |
6 | I + N (s) | 4 | 0 | 1 | Y | None | N/A | Yes | 4.4 | 7.7 | 7.7 | DOD |
7 | I + N (f) | 4 | 0 | 6 | Y | DX | 4 mg, qid | Yes | 3.4 | 3.4 | 3.4 | DOD |
8 | I + N (f) | 3 | 0 | 1 | Y | None | N/A | Yes | 6 | 6 | 6 | DOD |
9 | I + N (f) | 4 | 0 | 3 | N | None | N/A | No | 34 | 34 | 34.5 | NED |
10 | I + N (f) | 4 | 0 | Y | DX | 1 mg, bid | Yes | 5.2 | 5.2 | 5.2 | SD | |
11 | I + N (f) | 4 | 0 | 3 | Y | DX | 2 mg, daily | Yes | 24.6 | 24.6 | 24.6 | NED |
12 | I + N (f) | 4 | 0 | > 20 | Y | DX | 1 mg, bid | Yes | 8.9 | 8.9 | 8.9 | DOD |
13 | I + N (s) | 2 | 0 | 1 | Y | DX | 2 mg bid | Yes | 2.1 | 1.8 | 2.1 | DOD |
14 | I + N (f) | 3 | 0 | 11 | Y | DX | 2 mg, bid | Yes | 1.8 | 1.8 | 2.1 | DOD |
15 | I + N (s) | 2 | 0 | 1 | N | None | N/A | No | 1.2 | 1.2 | 9.3 | AWD |
16 | I + N (s) | 4 | 1.7 | 12 | Y | None | N/A | No | 1.7 | 1.7 | 5.4 | DOD |
17 | I + N (f) | 4 | 2.1 | 6 | Y | None | N/A | No | 2.1 | 2.1 | 6.9 | DOD |
18 | I + N (f) | 4 | 6.3 | > 10 | N | None | N/A | No | 6.2 | 6.2 | 9.7 | DOD |
19 | I + N (s) | 4 | 14.4 | 1 | N | None | N/A | No | 14.2 | 14.2 | 19.7 | DOD |
20 | I + N (s) | 1 | 7.5 | 4 | N | None | N/A | No | 1.4 | 7.4 | 10 | DOD |
Ipilimumab and nivolumab therapy of patients with MBM at diagnosis was associated with a modest overall benefit in progression-free survival (PFS)(5.2 months; (95% confidence interval [CI], 0.5 to 9.9 months) (Fig. 1A). Neurologic progression free survival (CNS PFS) was 6 months (95% CI, 1.1 to 10.9 months)(Fig. 11B). The overall survival (OS) of patients with brain metastases at diagnosis was 7.7 months (95% CI, 2.9 to 12.5)(Fig. 2). It should be noted that 4/14 patients (28.6%) with initial brain metastases and concomitant systemic disease achieved a durable complete remission with ipilimumab plus nivolumab therapy.
Fig. 1
(Panel A) Progression-free survival of patients with brain metastases at diagnosis versus patients who developed delayed onset of brain metastases. (Panel B): CNS progression-free survival of patients with brain metastases at diagnosis versus patients who developed delayed onset of brain metastases
×
![]()
Anzeige
Fig. 2
Overall survival of patients with brain metastases at diagnosis versus patients who developed delayed onset of brain metastases
×
![]()
Three of four complete remission patients received the alternate ipilimumab plus nivolumab regimen. One received the standard dosing regimen. All four of the patients with complete response (CR) have been able to electively discontinue therapy after radiologic confirmation of their complete remission. All 4 patients had an underlying BRAF V600E mutation. Three of 4 initially presented with neurologic symptoms associated with their brain metastases and underwent neurosurgical resection. These 3 subsequently received radiosurgery to the resection cavity and were receiving significant doses of systemic steroids at the time of immunotherapy treatment. receiving significant doses of systemic steroids at the time of immunotherapy treatment. It should be noted that the remaining 9 of 10 remaining patients all died due to CNS progression. One patient was best characterized as stable disease and was lost to follow-up.
The outcome of patients with brain metastases at presentation stands in contrast to the 6 of 59 patients who developed delayed onset brain metastases following initiation of ipilimumab plus nivolumab therapy. Onset of delayed brain metastases in all these patients occurred within 15 months of starting immunotherapy. Delayed onset of brain metastases following immunotherapy was uncommon (only 10.2% of patients receiving ipilimumab plus nivolumab therapy). Five of these 6 patients had died at the time of this analysis. Their median PFS was only 2.1 months after diagnosis of brain metastases (95% CI, 2.6 to 6.8 months). CNS PFS was 6.2 months (95% CI, 1.8 to 10.6 months) and median OS of 9.7 months (95% CI, 4.8 to 14.6 months). The remaining patient is currently remains on treatment at 24 months after multiple surgical resections of brain metastases and ongoing therapy.
Discussion
Metastatic melanoma has historically been associated with the frequent development of brain metastases [1]. Some studies have estimated the incidence of brain metastases to be 40–50% [21]. However, autopsy reports have suggested a much higher incidence (70–80%) [21]. Metastatic melanoma currently ranks 3rd in terms of the number of patients who develop brain metastases across different types of cancer. Only more common cancers, such as lung and breast cancer, result in a greater number of patients with brain metastases [22]. Another measure of the frequency of brain metastases is the percentage of patients with a given cancer who develop brain metastases (incidence proportion percentage). Patients with metastatic melanoma currently have the highest overall incidence proportion percentage [23].
Anzeige
Surgical resection of brain metastases is usually considered if lesions are solitary or symptomatic, particularly if they develop in non-eloquent areas of the brain [24]. In rare patients with isolated central nervous system (CNS) lesions, this can result in protracted remissions [25]. If there are multiple brain metastases, stereotactic ablative radiosurgery (SART) or whole brain radiotherapy (WBRT) are commonly employed as treatment options. However, patients treated with either WBRT or SART alone only gain a modest survival benefit [4, 26, 27].
With the advent of ICI-based immunotherapy, the treatment outcome has significantly improved for patients with small, asymptomatic brain metastases. For example, the NIBIT M2 trial randomized melanoma patients with asymptomatic brain metastases into three treatment groups: one receiving fotemustine only, another receiving ipilimumab and fotemustine, and a third group receiving ipilimumab plus nivolumab combination therapy. The recent 7-year follow up of trial results showed a significantly improved outcome in the cohort of patients receiving standard dose ipilimumab plus nivolumab therapy [12]. This group had a 7-year overall survival of 42.8% compared to only 10.0% in the group receiving fotemustine alone and 10.3% in the group receiving ipilimumab plus fotemustine [12].
In the Australian ABC trial, immunotherapy naïve patients with asymptomatic brain metastases were randomized into two different treatment groups, either receiving combined standard ipilimumab and nivolumab therapy or nivolumab monotherapy [3]. At a median follow up of 17 months, intracranial response was achieved in 46% of patients receiving combined immunotherapy compared to only 20% of patients receiving nivolumab monotherapy [3]. A small number of patients with symptomatic brain metastases, similar to those in our study, were also evaluated. These patients had a markedly inferior outcome [3]. After a median follow up of 17 months, only one patient in the symptomatic cohort (6%) had a partial intracranial response, in contrast to our data.
In another trial (Checkmate 204), 165 patients were enrolled into two separate cohorts based on whether they had symptomatic or asymptomatic melanoma brain metastases. Patients were treated with the standard ipilimumab and nivolumab regimen. At 3-year follow up, 85% of patients from the asymptomatic cohort had achieved an intracranial response, resulting in an intracranial progression free survival of 54.1% and a 3-year overall survival of 71.9%. However, in the symptomatic cohort, the objective response rate was significantly lower (only 16.7%). These patients had a markedly decreased intracranial progression free survival (18.9%) and overall survival (36.6%) [1].
DeCOG investigators published a pooled analysis of 380 ipilimumab plus nivolumab treated patients from German cancer centers [28]. Symptomatic brain metastases were present in 31% of patients prior to treatment with ipilimumab plus nivolumab. With a median follow-up of 18 months, the estimated 2-year survival rate was 41%. Potential prognostic factors for OS included the serum lactate dehydrogenase (LDH) level, the number of MBM, and ECOG performance status. Clinical complete remission was rare. Addition of SART to ipilimumab plus nivolumab therapy appeared to be beneficial. It should be noted that this was quite a heterogenous group of patients. Not all had brain metastases at onset, some received ICI treatment as 2nd or later lines of therapy, including following CNS progression on BRAF plus MEK inhibitor treatment [28].
Taken together, the results of these trials all strongly support combination immunotherapy as first line treatment for patients with small, asymptomatic brain metastases [1, 3]. In contrast, these trials suggest much more limited activity of this regimen in symptomatic patients. Of note, the timing of onset of brain metastases and the somatic mutation spectrum was not described in these studies.
We performed a retrospective analysis to evaluate the incidence and timing of the brain metastases in a uniformly treated cohort of metastatic melanoma patients receiving ipilimumab plus nivolumab. We found that 14 of our 20 MBM patients presented when they were initially diagnosed with systemic metastases. Most of these patients (11/14) had a BRAF V600E mutation. A small number had other RASopathy mutations (NF1 2/14, NRAS 1/14 All but two of the patients presenting with brain metastases underwent surgical resection, with routine addition of radiosurgery to the resection cavity. The clinical response rate following immunotherapy was reduced in these symptomatic patients [1, 3]. However, 4/14 patients were still able to achieve a durable complete remission and were able to eventually discontinue ICI therapy [16]. One additional patient has had protracted survival, despite persistent disease It is also notable that all four of these durable remission patients had a BRAF V600E mutation. Patients with NRAS or NF-1 mutations had inferior treatment outcomes. Of the four long term CR patients, three of them were treated with the alternate or “flipped” ipilimumab plus nivolumab regimen, suggesting that this regimen has significant CNS activity, which should be confirmed in further studies.
Our current study provides further support for previous observations that the risk of delayed brain metastases appears markedly reduced following ICI-based immunotherapy [25, 29]. Furthermore, patients who present with brain metastases prior to the start of combination immunotherapy had a better potential outcome than those who developed delayed onset of brain metastases. It is likely that patients with delayed CNS recurrence following initiation of ipilimumab plus nivolumab treatment had already been selected for immunotherapy resistance. The actual resistance mechanisms will require further evaluation. It should be noted that in our patient series, patients with delayed brain metastasis generally failed to respond to subsequent salvage therapy with further ICI treatment, targeted therapy, radiotherapy or surgery. Novel and more effective treatment approaches are badly needed for these patients.
Study limitations
A strength of our study is that it represents a population of metastatic cutaneous melanoma patients treated in a consistent fashion with ipilimumab plus nivolumab by a single investigator. A potential limitation is the relatively small number of patients, which could lead to unintended biases. Thus, this study should be considered hypothesis generating. This study represents a retrospective analysis of patient treatment spanning a lengthy period (8 years). The effect of changes in patient management have evolved over this time. For example, the use of the alternate ipilimumab plus nivolumab dosing regimen has become more prevalent over this interval [15]. In the Checkmate 511 study, the clinical outcomes of the standard and the alternate ipilimumab plus nivolumab regimens on progression-free and overall survival appear be to virtually superimposable in patients without brain metastases [15]. In contrast, we have reported an increased frequency of complete responses of brain metastases favoring the alternate treatment regimen that will need to be further evaluated in prospective studies [30].
Conclusions
Historically, patients with symptomatic melanoma brain metastases at diagnosis had a very poor outcome. Our data shows that combined immunotherapy with ipilimumab and nivolumab represent a useful first line treatment option for these patients. We observed a potential for durable clinical complete remission (28.6%), particularly if patients had a BRAF V600E mutation. Complete remission of both brain metastases and systemic disease in these 4 patients allowed eventual elective treatment discontinuation, with prolonged survival. It should be noted that complete responses were possible, even when symptomatic patients were receiving glucorticosteroid therapy at the initiation of ICI treatment (2 of these 4 patients). Three of these 4 patients also had surgical resection of large symptomatic metastases, and radiosurgery. Thus, we were not able to separately analyze which treatment components were useful in producing these long-term remissions.
The frequency of development of new brain metastases following initiation of ipilimumab plus nivolumab treatment appeared to be markedly reduced by ICI treatment (10.2%). Delayed-onset brain metastases occurred only within the first 15 months following initiation of combination immunotherapy. This observation indicates a need to monitor the brain periodically with scans, particularly during the first two years of therapy. Furthermore, patients with delayed onset brain metastases appeared to have been selected for highly resistant disease. Five of these 6 patients rapidly progressed (median 2.1 months) and died. There was a response to salvage immunotherapy with addition of targeted therapy in only one of these patients (who is receiving ongoing treatment, despite repeated disease relapses). Thus, improved treatment options are badly needed to more successfully treat patients with delayed onset brain metastases.
Acknowledgements
We would like to express our appreciation to the patients and their families, and the clinical staff of the Comprehensive Cancer Centers of Nevada. Careful review of the manuscript by Suzanne Samlowski M.Arch. is also appreciated.
Declarations
Ethical approval
The design of this retrospective anonymized data analysis was reviewed by the Western IRB chair and was considered exempt from full IRB review.
Informed consent
Not applicable.
Competing interests
The authors declare no competing interests.
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
