NF1 Mutated Metastatic Melanoma and Response to Immune Checkpoint Inhibitor Therapy: A Retrospective Analysis

1. Abstract 1.1. Background: Based on molecular profiling, malignant melanoma is classified in four different groups. NF1-mutated tumors are a small subgroup occurring with a frequency of 13% of all malignant melanomas, usually harboring a high tumor mutational burden (TMB). Considering TMB as being a prerequisite for the effectiveness of immune checkpoint inhibitor therapy, we were wondering if this rare subtype is associated with a higher response rate to immunotherapy than it is known for the general melanoma population.

1.2. Methods: We analyzed a small cohort of 14 NF1 mutated metastatic melanoma patients and retrospectively assessed the response rate (RR) according to RECIST 1.1, Progression Free Survival (PFS) and Overall Survival (OS). We compared our results with outcome data reported in several clinical trials with immune checkpoint inhibitors.

1.3. Results: For our cohort, we noticed an objective response rate of 64%, which is higher than the response rate generally reported for anti-PD1 based therapy in a population of melanoma patients. The PFS rate at twelve months was 62%, next to an OS rate at twelve months of approximately 84%. Additional mutations co-occur in NF1-mutated patients. Although we did not find an association between the number of additional mutations and response in general, we did notice a significant correlation between mutations in TERT promotor region and tumor response (p-value 0.027).

1.4. Conclusion: Despite the small patient group, we observed a higher response rate for NF1 mutated metastatic melanoma patients treated with immunotherapy. In addition, a significant correlation between response rate and the presence of hTERT promotor mutations was observed.

Keywords: Metastatic melanoma; NF1; TERT promotor; Immunotherapy

2. Introduction Melanoma is a disease where driver mutations are known to be responsible for tumor proliferation for the vast majority of patients. Molecular profiling classifies melanoma in different subtypes. Generally, four different molecular subtypes of melanoma are identified: BRAF-mutated, NRAS-mutated, NF1-mutated and triple wild-type tumors. BRAF (V600E) mutations (present in 38,5% of cutaneous melanomas) and NRAS mutations (28,6% of melanomas) are the most prevalent alterations and are both mutually exclusive [1]. The group which is considered triple wildtype, is thought to have different mechanisms driving tumor growth, such as dependence on ERK pathway activation. The least prevalent group are melanomas with NF1 alterations. The frequency of somatic NF1 mutations overall is 12,2% (1): while this alteration has a low frequency in cutaneous melanoma (12-30%), it is vastly more present (45-90%) in a very rare subtype of melanoma, called desmoplastic melanomas which is known to respond with a high frequency to immunotherapy [2]. In a study published in Nature in 2018 by Eroglu et al, 60 patients with desmoplastic melanoma were included, and an objective response rate of 70% was observed. 45 % of these patients had a complete response, 55% a partial response [3]. Screening with NGS (next generation sequencing) to identify these mutations is standard of care since the identification of driver mutations has direct implications for the treatment of patients at least when BRAF mutations are concerned. In 2011, FDA and EMA approved vemurafenib for metastatic melanoma carrying BRAFV600 mutations. Currently, for BRAF mutated patients, we have three approved combinations of BRAF/MEK inhibition as therapeutic options [4]. However other molecular alterations are not good candidates for targeted therapy thus far as it has been shown for NRAS and KIT mutated melanomas. NF1 (neurofibromin 1) is a protein known as a tumor suppressor gene, as it downregulates RAS proteins [5], by facilitating the hydrolysis of GTP to GDP [6]. If NF1 is lost, RAS becomes refractory to negative feedback hence leading to consecutive RAS activation [6]. NF1 mutations can occur in germline, resulting in a hereditary disease known as neurofibromatosis type 1. This is an autosomal dominant disorder with an incidence of 1 in 3000 live births, characterized by café-au-lait-macules, benign neurofibromas and other tumors, mostly from the neural crest [2, 5]. As a somatic mutation, NF1 mutations are the third most frequent cause of melanoma [7]. These melanomas originate most typically on chronically sun-exposed skin in older male patients, and show in general a high mutational burden [5, 7]. The mechanism leading to a high tumor mutational burden as a possible result of NF1 mutation, still remains elusive. Very rarely, NF1 mutations are also present in patients harboring a BRAF V600 mutation, and response to BRAF/MEK inhibition has also been studied in these patients. Presumably, NF1 mutations could induce resistance to MAPK inhibition because of a sustained MAPK pathway activation [5]. Due to NF1 loss, the negative feedback on RAS activation is lacking, resulting in resistance to RAF inhibition. As a result, the pathway stays MEK dependent, and presumably susceptible to allosteric MEK inhibitors [6]. One explanation of the high mutational burden in these melanomas could be the chronically stimulated RAS pathway where the DNA damage repair machinery cannot keep up with the proliferative drive in these cells. The mechanism how mutational burden translates into a higher response to immunotherapy is not yet totally clear but likely mutations give rise to expression of more neo-antigens, and as a result the likelihood for immune recognition increases [8]. In line with this hypothesis, the rare subtype of desmoplastic melanomas (less than 4%), which significantly harbor NF1 mutations more frequently, are known to derive substantial benefit from anti PD-1 / anti PD-L1 therapy according to anecdotal reports, although this has not been systematically investigated so far [3]. However, a multitude of different markers exist which are presumably associated with response to anti-PD1 based therapy. In melanoma, hTERT promotor mutations are reported to be the most frequently occurring mutation, and the prevalence is associated with increasing age, tumor site, and histological subtype. [10] Some authors describe that the presence of a mutation in the hTERT promotor corresponds with a worse prognosis and shorter survival, because of adverse characteristics like increased thickness, ulceration and mitotic rate, being more prevalent in hTERT mutated melanomas [11]. Until now, this worse prognosis is mostly described for hTERT promotor mutations and the simultaneous presence of BRAF and NRAS mutations. It is still unclear whether the same is true for NF1 mutated tumors [12, 13]. In this retrospective analysis, we were looking at a cohort of NF1 mutated metastatic melanoma patients, who were treated at our center with anti PD1 based therapy. In particular, we were interested in response, PFS and OS in patients treated with ICI, who suffer from this rare subtype of melanoma.

3. Methods The patient population included in our analysis was from our melanoma database in UZ Leuven. All patients were diagnosed and treated at our site. Next generation sequencing is a standard of care diagnostic procedure and includes a panel of 96 genes, including the NF1 gene, relevant for diagnostic and therapeutic purposes in solid tumors. Only patients with available sequencing data were included. The melanoma database was extended to October 2018, because only since then NF1 was part of the Next Generation Sequencing (NGS) panel. A search for all NGS executed tests, irrespective of the type of malignant disease was done in September 2020, and 4077 NGS tests were identified. Melanoma was found to be the tumor of origin in 4.7 % of the cases, resulting in 194 NGS tests performed in malignant melanomas. 13 of these 194 NGS analyses did not yield a result because of numerous reasons, mostly due to poor DNA quality. 181 NGS tests were available for analysis, with 24 (13,3%) showing a NF1 alteration. 11 of these were classified as VUS, the other 13 were thought to be pathogenic or presumably pathogenic (Figure 1). When comparing our NF1 mutation positive population with other reports in the literature, a comparable mutation frequency of around 13-14% was found [14]. A database of 24 NF1 mutant tumors was established by extracting clinical and demographic data from the patients hospital charts. Of those 24 individual patients, 6 had undergone a curative resection of their melanoma and had no documented relapse until the last date of follow-up in January 2021. 4 out of 6 patients were treated in the adjuvant setting, one patient in this group relapsed during adjuvant treatment. 14 patients had upfront metastatic disease at the time of diagnosis. For the purpose of this analysis, we only focused on those patients with metastatic disease, hence 14 metastatic patients were then further analyzed. Characteristics as gender and age were extracted, but also stage of disease, first line therapy, number of metastasis sites, lactate dehydrogenase (LDH) level at time of diagnosis, type of melanoma (cutaneous, mucosal or unknown) and presence or absence of brain metastasis. Response was assessed in all patients with CT scan or PET-CT scan according to RECIST v1.1, and response assessment was done every twelve weeks. In the event of progressive disease, response assessment was complemented with criteria used in iRE CIST unless patients showed unequivocal disease progression. Progression Free Survival (PFS) was defined as time from initiation of anti-PD1 based therapy until disease progression per RECIST or death from any cause. Overall Survival (OS) was defined as time from treatment initiation until death from any cause. Statistical analysis used the Kaplan-Meier method for estimating PFS and OS. The cumulative incidence function was used for estimating time to best response; death without response was treated as a competing event. The Mann-Whitney U test was used to compare groups on ordinal variables. The Fisher exact test was used for group comparisons on binary variables. A two-sided 5% significance level was adopted for all tests. Analyses have been performed using SAS software (version 9.4 of the SAS system for Windows).

4. Results and Discussion 14 patients with metastatic, irresectable malignant melanoma were included in this retrospective analysis, all of them harboring a NF1 mutation, in addition to a multitude of secondary mutations associated with this rare melanoma subtype. All these patients had upfront metastatic disease at the time of their diagnosis. Median age was 73 years, 9 patients were male, 5 patients were female. hTERT promoter mutation was the second most frequent mutation in our cohort. At the time of initiation of therapy, 29% had elevated LDH. 7 out of 14 (50%) included patients had a primary cutaneous melanoma, while 3 (21%) were diagnosed with a mucosal melanoma. For the remaining 4, the primary tumor was not known. 5 patients (36%) had brain metastasis at the time of diagnosis. Almost half of the included patients (43%) had more than three metastatic sites (Table 1). More than half of the NF1 mutations were classified as VUS (57%). 11 out of 14 patients harbored an additional TERT promotor mutation, next to the NF1 mutation. Three patients (21%) didn’t harbor a mutation in the TERT promotor gene. Most of the included patients harbored more mutations than only one NF1 mutation. Besides the most prevalent hTERT promotor mutations (79%), additional BRAF mutations (3 patients (21%): two mutations in exon 15, one in exon 1), and additional NRAS mutation (4 patients (29%); three mutations in exon 3, one in exon 2) were detected. A variety of other mutations were also found although less frequently (CDKN2A (36%), TP53 (21%), KIT (21%), MET (14%), SMARCA4 (14%), GNAQ (7%), PDGFRA (7%), BAP1 (7%), MAP2K1 (7%), GNA11 (7%), PIK3CA (7%)) (Table 7). 13 out of 14 patients received immunotherapy in first line. Only one patient received pembrolizumab in second line. 4 out of 13 patients started their treatment with combination immunotherapy ipilimumab – nivolumab, whereas 3 patients received nivolumab in monotherapy. 7 patients received pembrolizumab mono. With regard to response, we observed an objective response rate of 64% in our total patient population. 36 % (n=5) of the patients showed either stable disease (n=3; 21%) or progressive disease (n=2; 14%) as their best response. A total of 28% of the patients achieved a complete response, whereas 36% had a partial response (Table 2). When subdividing by type of treatment, patients treated with combination immunotherapy showed a response in 50% of the cases (2/4), whereas patients treated with anti-PD1 monotherapy, responded in 70% of the cases (7/10). As shown in the swimmer plot analysis, responses usually occurred within the first 4 months of treatment (Figure 2). The median time of follow up in our study was 11.7 months. Considering progression free survival, in our cohort 6/14 patients had a progression event during the time of follow-up compared to 8/14 who had no progression (Table 3). The PFS rate was 86% (CI 54-96%) at three months, and the estimated PFS at six, twelve and twenty-four months was approximately of 62% (CI 32-82%) (Table 4). The Kaplan Meier curve for progression free survival shows a plateau which is expected and typical for ICI treated patients (Figure 3). Overall survival estimates at three months are 100%, 92% (CI 57- 99%) at six months, 84 % (CI 49-96%) at twelve months and 72 % (CI 34-90%) at twenty-four months (Table 6). The number of deaths in this small cohort was 29% (n=4) (Table 5). Figure 4 shows the overall survival curve as estimated by the Kaplan-Meier method.

5. Conclusion We want to emphasize that this small descriptive study had a very limited patient number and interpretation of the results should therefore be done with caution. However, we saw a high response rate in this small number of NF1 mutated patients treated with immunotherapy, which presumably is higher compared to reported efficacy data generated in large phase III studies. We didn’t find a clear correlation between the total number of additional mutations and response rate. This is likely due to the fact we did not measure TMB, and the number of additional mutations might not be a perfect surrogate for this purpose. We also found a significant correlation between response rate and the presence of a mutation in the promotor of TERT. Reports in the literature show a negative prognostic impact of the presence of this mutation especially if simultaneously present with NF1 mutation. In summary, our data suggest that the outcome of anti PD1 based ICI therapy in NF1 mutated melanoma is at least comparable with the outcome seen in other melanoma subtypes and should therefore also be considered in patients with this rare genomic alteration.

References 1. Krauthammer M, Kong Y, et al. Exome sequencing identifies recurrent mutations in NF1 and RASopathy genes in sun-exposed melanomas. Nat Genet. 2015; 47(9): 996-1002.

2. Philpott C, Tovell H, et al. The NF1 somatic mutational landscape in sporadic human cancers. Human Genomics. 2017; 11(13).

3. Eroglu Z, Zaretsky JM, et al. High response rate to PD-1 blockade in desmoplastic melanomas. Nature. 2018; 553(7688): 347-50.

4. Sun J, Carr MJ, Khushalani NI. Principles of Targeted Therapy for Melanoma. Surg Clin North Am. 2020; 100(1): 175-88.

5. Kiuru M, Busam K. The NF1 gene in tumor syndromes and melanoma. Lab Invest. 2017; 97(2): 146-57.

6. Nissan M, Pratilas C, et al. Loss of NF1 in cutaneous melanoma is associated with RAS activation and MEK dependence. Cancer Res. 2014; 74 (8): 2340-50.

7. Cirenajwis H, Lauss M, et al. NF1-mutated melanoma tumors harbor distinct clinical and biological characteristics. Molecular oncology. 2017; 438-51.

8. Johnson DB, Frampton GM, et al. Targeted next generation sequencing identifies markers of response to PD-1 blockade. Cancer Immunol. Res. 2016; 4 (11): 959-67.

9. Saman MV, Garrigós C, Duran I. Biomarkers of response to PD-1/ PD-L1 inhibition. Critical Reviews in Oncology/Hematology. 2017; 116: 116-124.

10. Sheen YS, Chu CY et al. Co-occurrence of TERT promotor mutations with BRAF or NRAS alterations correlates with worse prognosis in melanoma. British Journal of Dermatology. 2021; 184: 384-392.

11. Manrique-Silva E, Rachakonda S, Millan-Esteban D et al. Clinical, environmental and histological distribution of BRAF, NRAS and TERT promoter mutations among patients with cutaneous melanoma: a retrospective study of 563 patients. Br J Dermatol. 2021; 184:504-13.

12. Gao K, et al. TERT promoter mutations and long telomere length predict poor survival and radiotherapy resistance in gliomas. Oncotarget. 2016; 7: 8712-25.

13. Melo M, et al. TERT promoter mutations are a major indicator of poor outcome in differentiated thyroid carcinomas. J Clin Endocrinol Metab. 2014; 99: 754-65.

14. Zhang T, Dutton-Regester K, et al. The genomic landscape of cutaneous melanoma. Pigment Cell melanoma Res. 2016; 29 (3): 26683.

15. Park C, Kim M, et al. Clinical application of Next-Generation Sequencing-Based Panel to BRAF Wild-Type Advanced Melanoma identifies Key Oncogenic Alterations and Therapeutic Strategies. Molecular Cancer Ther. 2020; 19(3): 937-44.

16. Scherrer E, Rau Reina et al. Systematic literature review for the association of biomarkers with efficacy of anti-PD-1 inhibitors in advanced melanoma. Future Oncol. 2021; 17 (20): 2683-92.

17. Schachter J, Ribas A, et al. Pembrolizumab versus ipilimumab for advanced melanoma: final overall survival results of a multicentre, randomised, open-label phase 3 study (KEYNOTE-006). Lancet. 2017; 390: 1853-62.

18. Robert C, Long GV, et al. Nivolumab in previously untreated melanoma without BRAF mutation. N Engl J Med. 372: 320-330.

19. Larkin J, Charion-Sileni V, et al. Five-year survival with combined nivolumab and ipilimumab in advanced melanoma. N Engl J Med. 2019; 381: 1535-46.

20. Teixido C, Castillo P. et al. Molecular markers and Targets in Melanoma. Cells. 2021; 10: 2320.

De Backer Cleo. NF1 Mutated Metastatic Melanoma and Response to Immune Checkpoint Inhibitor Therapy: A Retrospective Analysis. Annals of Clinical and Medical Case Reports 2022