Biomarkers for Outcome in Metastatic Melanoma in First Line Treatment with Immune Checkpoint Inhibitors

Introduction: A high proportion of metastatic melanoma patients do not respond to immune checkpoint inhibitors (ICI), and until now, no validated biomarkers for response and survival have been known. Methods: We performed a retrospective analysis of outcomes in patients with metastatic melanoma treated with first-line ICI at the Institute of Oncology Ljubljana from January 2018 to December 2020. The immune-related adverse events (irAEs) and serum immune-inflammation parameters (neutrophil-to-lymphocyte ratio (NLR), platelet-to-lymphocyte ratio (LR), systemic immune-inflammation index (SII) and pan-immune-inflammation value (PIV)) were analyzed as potential biomarkers for response and survival. Survival rates were calculated using the Kaplan–Meier method and then compared with the log-rank test. Multivariate regression Cox analysis was used to determine independent prognostic factors for progression-free survival (PFS) and overall survival (OS). Results: Median follow-up was 22.5 months. The estimated median progression-free survival (PFS) was 15 months (95% CI 3.3–26.2). The two-year survival rate (OS) was 66.6%. Among 129 treated patients, 24 (18.6%) achieved complete response, 28 (21.7%) achieved partial response, 26 (20.2%) had stable disease and 51 (39.5%) patients experienced a progressive disease. There was a higher response rate in patients with irAEs (p < 0.001) and high NLR before the second cycle of ICI (p = 0.052). Independent prognostic factors for PFS were irAE (HR 0.41 (95% CI 0.23–0.71)), SII before the first cycle of ICI (HR 1.94 (95% CI 1.09–3.45)) and PLR before the second cycle of ICI (HR 1.71 (95% CI 1.03–2.83)). The only independent prognostic factor for OS was SII before the first cycle of ICI (HR 2.60 (95% CI 0.91–7.50)). Conclusions: Patients with high pre-treatment levels of SII had a higher risk of progression and death; however, patients with irAEs in the high-SII group might respond well to ICI. Patients who develop irAEs and have high NLRs before the second ICI application have higher rates of CR and PR, which implicates their use as early biomarkers for responsiveness to ICI.


Introduction
The annual incidence of malignant melanoma in Europe varies between 3 and 5 people per 100,000 in Mediterranean countries and between 12 and 35 people per 100,000 in Nordic countries [1]. In Slovenia, the average annual melanoma incidence rate in the period from 2014 to 2018 was 25.6 for females and 29.5 for males per 100,000. These numbers increased compared to the average annual melanoma incidence rate from 2009 to 2013 (24.3 for females and 25 for males per 100,000), and is estimated to increase further to 28 women and 34 men per 100,000 (95% prediction interval) in the year 2021. That makes Slovenia one of the European countries with the highest annual incidence of malignant melanoma. Approximately 78% of Slovenian patients with melanoma initially present with localized disease, which is mostly due to broad public education; 19% present with regional phocytes and monocytes), molecular parameters (BRAF V600 and NRAS (neuroblastoma Ras gene)), irAEs, treatment responses and events such as relapse and death were collected from medical records.
The study was approved by the National Medical Ethics Committee (approval number: 0120-342/2020/5) and was carried out according to the Declaration of Helsinki.
The ICI was applied as an intravenous infusion at the recommended dose of 200 mg every 3 weeks for pembrolizumab and either 240 mg every 2 weeks or 480 mg every 4 weeks for nivolumab. Nivolumab was used in combination with ipilimumab as follows: 1 mg/kg nivolumab in combination with 3 mg/kg ipilimumab, administered intravenously every 3 weeks for the first 4 doses, followed by a second phase with nivolumab monotherapy, applied as described above.
The immune Response Evaluation Criteria in Solid Tumors (iRECIST) criteria were used to evaluate the tumor response [15]. The irAEs were evaluated by a clinician based on the findings of laboratory tests, clinical examinations and imaging studies, and were graded according to the National Cancer Institute Common Terminology Criteria for Adverse Events, version 5.0. [16]. Treatment interruption was defined as a time interval during which an irAE was assessed by the oncologist at a grade (CTCAE v 5.0) indicating that the ICI should be temporarily discontinued until the irAE is reversed. Treatment discontinuation was defined as a time point at which irAE was assessed at a grade (CTCAE v 5.0) indicating that the ICI should be terminated.
The systemic immune-inflammation markers were assessed from the peripheral blood platelet (P), neutrophil (N), monocyte (M) and lymphocyte (L) counts in cells/L [17][18][19]. The neutrophil-to-lymphocyte ratio (NLR) was calculated as N/L; the platelet-to-lymphocyte ratio (PLR) as P/L, the pan-immune-inflammation value (PIV) as (N × P × M)/L) and the systemic immune-inflammation value (SII) as P × N/L. We evaluated NLR, PLR, PIV and SII at baseline before the start of ICI therapy and before the second cycle of ICI. Cut-off values for low and high values were set based on data from the literature or median values were set at ≥2 for NLR, ≥180 for PLR, ≥390 for PIV and ≥730 for SII [17,18].
The OS was defined as the time from the date of the first ICI administration to the date of death from any cause. The PFS was defined as the time from the date of the beginning of ICI to the date of disease progression or death from any cause.
The characteristics of patients were categorically presented as frequencies and proportions. Age was presented as median and range. Pearson's chi-square test was used for statistical comparisons for categorical data, and the unpaired Student's t-test was used for comparing ages between groups. In the case of expected parameter values of <5 in >20% of cells, Fisher's exact test was used to facilitate the analysis of smaller population sizes. A p-value ≤ 0.05 was considered statistically significant. Spearman rho was used for the calculation of the correlation between ordinal determinants.
The survival analyses were performed using the Kaplan-Meier method and compared by log-rank test. The prognostic significance of the variable of interest (age, ECOG performance status, melanoma type, metastatic site, LDH, comorbidities, irAE, response rate and immune-inflammatory parameters (NLR, PLR, PIV and SII) was calculated using the Cox proportional hazards regression model, and expressed as hazard ratio (HR) and 95% confidence interval (CI). All the variables showing a p ≤ 0.05 in the univariate models were included in the multivariable model. The variables showing a p ≤ 0.05 in the multivariable models were considered to be independent prognostic factors. All statistical analyses were performed using SPSS v.28.0 (IBM Corp., Armonk, NY, USA).

Patients and Treatment
From January 2018 to December 2020, 311 patients were treated with ICI for malignant melanoma at the Institute of Oncology Ljubljana. Of these, 129 patients fulfilled the inclusion criteria for the study (Figure 1).

Patients and Treatment
From January 2018 to December 2020, 311 patients were treated with ICI for malignant melanoma at the Institute of Oncology Ljubljana. Of these, 129 patients fulfilled the inclusion criteria for the study (Figure 1). Patients (all Caucasian) were treated with ICI as the first-line treatment for metastatic melanoma. The median age of patients was 66.2 (30.1-84.5) years, and 61.3% were males. In total, 75% of patients had primary skin melanoma; 47.3% were metastatic to skin or soft tissue, including muscle and/or nonregional lymph nodes (stage M1a); and 17% had central nervous metastases. More than half (56.6%) of patients had concomitant diseases, the most prevalent of which was arterial hypertension (39.5%), and 7.8% had a history of an autoimmune disease (one patient had sarcoidosis; others had thyroid autoimmune disease). BRAF mutations were present in 23.3% and N-RAS in 6.2% of patients. Over threequarters of patients (76.7%) were treated with pembrolizumab, and 37.2% of patients developed irAE. Patients' baseline clinical and pathological characteristics, as well as the type of ICI treatment, are summarized in Table 1.  Patients (all Caucasian) were treated with ICI as the first-line treatment for metastatic melanoma. The median age of patients was 66.2 (30.1-84.5) years, and 61.3% were males. In total, 75% of patients had primary skin melanoma; 47.3% were metastatic to skin or soft tissue, including muscle and/or nonregional lymph nodes (stage M1a); and 17% had central nervous metastases. More than half (56.6%) of patients had concomitant diseases, the most prevalent of which was arterial hypertension (39.5%), and 7.8% had a history of an autoimmune disease (one patient had sarcoidosis; others had thyroid autoimmune disease). BRAF mutations were present in 23.3% and N-RAS in 6.2% of patients. Over three-quarters of patients (76.7%) were treated with pembrolizumab, and 37.2% of patients developed irAE. Patients' baseline clinical and pathological characteristics, as well as the type of ICI treatment, are summarized in Table 1. We compared clinical, pathological and systemic immune-inflammatory markers between the group of patients with occurrence irAE and the group without irAE ( Table 2). Patients in the group with irAE were more often treated with ipilimumab + nivolumab. They had also higher systemic inflammation indices (SII) before the first cycle of ICI and a higher frequency of BRAF mutations. Table 2. Comparison of clinicopathological and systemic inflammatory characteristics between groups of patients who either developed immune-related adverse effects (irAE) or did not (statistically significant marked bold). M1a-distant metastasis to skin or soft tissue, including muscle and/or nonregional lymph nodes; M1b-lung; M1c-other visceral sites; M1d-the central nervous system. LDH-lactate dehydrogenase, ECOG-Eastern Cooperative Oncology Group, PS-performance status, PIV-pan-inflammation value, SII-systemic immune-inflammation index, PLR-platelet-tolymphocyte ratio, NLR-neutrophil-to-lymphocyte ratio, ICI-immune checkpoint inhibitor.

Response Rates and Survival Outcomes
The median follow-up time was 22.5 months. The median PFS was 15 months (95% CI 3.3-26.6). In the group of patients without irAE, the median PFS was 9.3 months, compared with 32.8 months in the group with irAE (p = 0.01). Numbers in bold are statistically significant.
Univariate and multivariate analyses of possible factors affecting PFS are presented in Table 3. Table 3. Univariate and multivariate Cox regression analyses of prognostic factors associated with prognosis. PFS-progression free survival, OS-overall survival, irAE-patients with metastatic melanoma who developed immune-related side effects due to immunotherapy, NirAE-patients with metastatic melanoma who did not develop immune-related side effects due to immunotherapy, CRcomplete response, PR-partial response, SD-stabile disease, PD-progressive disease, PIV-paninflammation value, SII-systemic immune-inflammation index, PLR-platelet-to-lymphocyte ratio, NLR-neutrophil-to-lymphocyte ratio, LDH-lactate dehydogenase, >4.31 microkat/L. Numbers in bold are statistically significant.  In our cohort, it was revealed that patients with irAE had more than halvedthe hazards of progression (HR 0.41) that their counterparts without irAE did (Figure 2A). However, patients with high SII indices before the first cycle of ICI, and those with high PLR before the second cycle of ICI, had a risk of progression almost two times higher than the corresponding patients with low levels of SII and PLR (HR 1.94 and 1.71, respectively) ( Figure 2B,C).

Factors
In our cohort, it was revealed that patients with irAE had more than halvedthe hazards of progression (HR 0.41) that their counterparts without irAE did (Figure 2A). However, patients with high SII indices before the first cycle of ICI, and those with high PLR before the second cycle of ICI, had a risk of progression almost two times higher than the corresponding patients with low levels of SII and PLR (HR 1.94 and 1.71, respectively) ( Figure 2B,C).  In the whole cohort, 1-year OS was 80% and 2-year OS was 66.6%. Median OS was not reached. An analysis of possible prognostic factors and their effects on OS (univariate and multivariate Cox analysis) is presented in Table 3. The only independent prognostic factor for OS was revealed to be SII before the first cycle of ICI (p = 0.003; Figure 3). Patients with high SIIs had hazards of death 2.6 times higher than patients with low SII. In patients with low SII, 2-year OS was 77%, compared to 55% in patients with high SIIs (p < 0.003; Figure 3).
In the whole cohort, 1-year OS was 80% and 2-year OS was 66.6%. Median OS was not reached. An analysis of possible prognostic factors and their effects on OS (univariate and multivariate Cox analysis) is presented in Table 3. The only independent prognostic factor for OS was revealed to be SII before the first cycle of ICI (p = 0.003; Figure 3). Patients with high SIIs had hazards of death 2.6 times higher than patients with low SII. In patients with low SII, 2-year OS was 77%, compared to 55% in patients with high SIIs (p < 0.003; Figure 3).
In total, 24 (18.6%) patients achieved a complete response, 28 (21.7%) achieved a partial response, 26 (20.2%) had stable disease and 51 (39.5%) experienced progressive disease. The response to ICI was significantly associated with the occurrence of irAEs (p < 0.001) (suppl. Figure S3) and high NLR before the second cycle of ICI (p = 0.052). Patients with irAEs and high NLR before the second ICI had higher rates of complete and partial responses.
The subgroup of patients with high SIIs before the ICI, but who developed irAEs due the ICI treatment, had better PFS (Figure 4). In total, 24 (18.6%) patients achieved a complete response, 28 (21.7%) achieved a partial response, 26 (20.2%) had stable disease and 51 (39.5%) experienced progressive disease. The response to ICI was significantly associated with the occurrence of irAEs (p < 0.001) (suppl. Figure S3) and high NLR before the second cycle of ICI (p = 0.052). Patients with irAEs and high NLR before the second ICI had higher rates of complete and partial responses.
The subgroup of patients with high SIIs before the ICI, but who developed irAEs due the ICI treatment, had better PFS (Figure 4).

Discussion
We present data regarding possible new clinical biomarkers for predicting ICI efficacy in metastatic melanoma treatment-naïve patients. According to our data analysis, high pre-treatment levels of SII were connected with a high risk of progression and death. Subgroup analysis of patients with high SII according to irAEs revealed that patients with irAEs had higher PFS in comparison with those who did not. Patients who developed irAEs and had high NLR before the second ICI application were the best responders to ICI. The median PFS was 15 months (95% CI 3.3-26.6), and median OS was not reached. This corresponds with data from already-published studies. In Keynote 006 patients with metastatic melanoma treated with pembrolizumab, the median PFS was 9.7 months (95% CI, 5.8-12.0), and the median OS was not reached after a follow-up period of 48 months [20]. In CheckMate 067, the median PFS was 11.5 months (95% CI, 8.7 to 19.3) in the nivolumab-plus-ipilimumab group, 6.9 months (95% CI, 5.1 to 10.2) in the nivolumab group and 2.9 months (95% CI, 2.8 to 3.2) in the ipilimumab group, and the median OS

Discussion
We present data regarding possible new clinical biomarkers for predicting ICI efficacy in metastatic melanoma treatment-naïve patients. According to our data analysis, high pre-treatment levels of SII were connected with a high risk of progression and death. Subgroup analysis of patients with high SII according to irAEs revealed that patients with irAEs had higher PFS in comparison with those who did not. Patients who developed irAEs and had high NLR before the second ICI application were the best responders to ICI. The median PFS was 15 months (95% CI 3.3-26.6), and median OS was not reached. This corresponds with data from already-published studies. In Keynote 006 patients with metastatic melanoma treated with pembrolizumab, the median PFS was 9.7 months (95% CI, 5.8-12.0), and the median OS was not reached after a follow-up period of 48 months [20]. In CheckMate 067, the median PFS was 11.5 months (95% CI, 8.7 to 19.3) in the nivolumab-plus-ipilimumab group, 6.9 months (95% CI, 5.1 to 10.2) in the nivolumab group and 2.9 months (95% CI, 2.8 to 3.2) in the ipilimumab group, and the median OS was more than 60.0 months (median not reached) in the nivolumab and ipilimumab groups and 36.9 months in the nivolumab group, as compared with 19.9 months in the ipilimumab group (HR with nivolumab and ipilimumab vs. ipilimumab, 0.52; HR with nivolumab vs. ipilimumab, 0.63) [7]. Grade 3-4 irAEs occurred in 17% of the patients, and 1 patient died from treatment-related sepsis in Keynote 006 [20]. The correlation of irAEs and pan-immune-inflammation values with survival was not initially assessed for the metastatic patients in these two studies, but in 2020, Eggermont et al. published the results of a prospective study's sub-analysis of pembrolizumab vs. placebo in high-risk stage III melanoma patients treated with pembrolizumab. This revealed a positive correlation of irAEs with longer recurrence-free survival [21]. In the same year, we published our retrospective analysis results regarding Slovenian metastatic melanoma patients treated with ICI, showing a higher survival probability of more than 80% in patients with irAEs vs. less than 60% in patients without irAEs [14]. The meta-analysis of irAEs in patients with different cancers and their correlation with the treatment efficacy, which included 52 papers comprising 9156 patients and pooled data analysis, demonstrated a greater and statistically significant probability of achieving an objective tumor response in patients with irAEs compared to those without (OR 3.91). Patients who developed irAEs presented prolonged PFS (HR 0.54) and OS (HR 0.51) rates. This refers mostly to NSCLC and melanoma patients treated with ICI, regardless of the grade of the irAE or discontinuation of the treatment [22]. Our study's research data suggest the same notion, that irAE could be used as a potential biomarker for higher response rates and lower hazards of progression and death in patients treated with ICI.
In a metastatic colorectal setting, in patients with high microsatellite instability cancers treated with checkpoint inhibitors, both PIV and SII correlate with the PFS and OS, according to pooled data from the Valentino and Tribe trials. Patients with high PIV had 1.66-times-higher hazards of progression and two-times-higher hazards of death compared to patients with low PIV [13]. In metastatic melanoma patients treated with ICI, the data are not consistent enough to make reliable conclusions. Namely, based on Susok et al.'s prospective study, which included an analysis of 62 patients treated with immunotherapy for unresectable stage III and IV melanoma, PIV and SII did not seem to be significant predictors for clinical outcome. However, according to Fuca and al.'s retrospective study analysis of 228 metastatic melanoma patients treated with checkpoint inhibitors, a high baseline PIV was independently associated with lower OS (adjusted HR: 2.06) and PFS (ad-justed HR 1.56). High PIV was also associated with primary resistance to immunotherapy (odds ratio (OR): 3.98) [12,35].
According to our data analysis, high pre-treatment SII had a 1.94-times-higher risk of progression and 2.6-times-higher risk of death compared with patients with low SII, which was similar to the results of the colorectal cancer study [13]. A subgroup analysis of patients with high SII according to irAE revealed that patients with irAEs had higher PFS in comparison with those who did not (p = 0.002) (Figure 4). Patients with high PLR before the 2nd ICI application had 1.71-times-higher risks of progression. Regarding the response rate to ICI, high NLR before the 2nd ICI application significantly correlated with better response, and those patients had higher CR and PR. This is the main contribution of our study, made in a time where there are no conclusive data regarding the potential role of irAEs and immune-inflammation parameters as biomarkers for ICI efficacy; thus, our research adds valuable data to this topic. First of all, the development of irAEs in patients with metastatic melanoma treated with ICI as a first-line treatment can be used as a potential biomarker to determine the subgroup which will likely achieve a good response and which has a better PFS rate. The other important message of our study is the potential use of initially high SII at presentation, as well as high PLR before the second ICI application, as early biomarkers for unresponsiveness to ICI. However, even if patients had an initially high SII but developed irAEs, they showed longer PFS rates than those without irAEs. Patients that developed irAEs and had high NLR before the second ICI application had higher rates of CR and PR.
The strength of our study is that we provide the results for the response and survival rates of patients treated at a single, wide, national and comprehensive cancer center. We strictly followed the inclusion criteria, regularly evaluated treatment efficacy according to iRECIST and used accurate vitality data. The limitations of our study include its retrospective study design.

Conclusions
In a time in which there are no conclusive data on the potential role of irAEs and immune-inflammation parameters as biomarkers for ICI efficacy in the treatment of metastatic melanoma patients, our research study analysis carries important messages. Patients with high pre-treatment levels of SII have higher risk of progression and death; however, patients with irAEs in the high-SII group might be good responders to ICIs. High PLR before the second cycle suggests a higher hazard of progression. Furthermore, patients who develop irAEs and have high NLR before the second ICI application have higher rates of CR and PR, which implicates their usefulness as early biomarkers for responsiveness to ICI.  Institutional Review Board Statement: The study was approved by the National Medical Ethics Committee (Approval number: 0120-342/2020/5) and was carried out according to the Declaration of Helsinki.

Informed Consent Statement:
For this retrospective study, written informed consent was not required.

Data Availability Statement:
The research data for this study are not publicly available on legal and ethical grounds. Regulation (EU) 2016/679-General Data Protection Regulation (GDPR) protection of natural persons with regard to the processing of personal data and the free movement of such data. Further enquiries can be directed to the corresponding author.