Tumor-Infiltrating Lymphocytes in Patients with Melanoma: Which Is Its Prognostic Value?

Abstract

Oncological impact of tumor-infiltrating lymphocytes (TILs) in melanoma remains controversial. We aimed to determine the significance of TILs on melanoma-specific survival (MSS), recurrence-free survival (RFS), and sentinel lymph node status (SLN). A retrospective analysis of patients undergoing melanoma resection during the period 2009–2019 was performed. Using the Melanoma Institute Australia grading system for TILs, the cohort was divided into two groups: group 1 (G1), patients with TILs grades 1, 2, or 3 and Group 2 (G2), patients with TILs grade 0. From a total of 386 melanoma resections, 151 (39%) were included in G1 and 39 (10%) in G2. Among the 151 patients who underwent SLN biopsy, the positivity rate according to the TILs grades 0, 1, 2, and 3 was 32%, 18%, 14%, and 0%, respectively, p = 0.02. With an average follow-up of 48 months, the 5-year MSS (G1: 86% vs G2: 75%, p = 0.002) and the 5-year RFS (G1: 81% vs G2: 60%, p = 0.004) were significantly higher in G1 than G2. Tumor-infiltrating lymphocytes in melanoma are associated with the SLN status and with a better MSS and RFS.

Introduction

Fifty years have passed since Clark et al. first described the lymphocytic infiltration of primary cutaneous melanoma [1]. However, it was not until recently that immune host response against melanoma has gained much attention with the advent of immunomodulating therapies for advanced stages disease.

Twenty years later, Clark classification system for TILs (tumor-infiltrating lymphocytes) remained relevant, but with the advent of targeted immunotherapy against TILs, refinement of the Clark classification system was proposed [2]. Melanoma Institute Australia (MIA) proposed a new grading system for TILs consisting of grade 0 through grade 3. This was based on the density (mild, moderate, or marked) and distribution (focal, multifocal, or diffuse) of TILs into the dermis [3] (Table 1) (Fig. 1).

Table 1.

Grading system for TILs of The Melanoma Institute Australia (MIA)

TILs distribution and density	TILs grade
Absent TILs	Grade 0
Mild or moderate focal TILs infiltrate Mild multifocal TILs infiltrate	Grade 1
Marked focal TILs infiltrate	Grade 2
Moderate or marked multifocal TILs infiltrate
Mild diffuse TILs infiltrate
Moderate or marked diffuse TILs infiltrate	Grade 3

TILs tumor-infiltrating lymphocytes

Fig. 1.

Hematoxylin and eosin (H&E)-stained melanoma sections at 40 × magnification demonstrating grades 0–3 of Melanoma Institute Australia grading system for tumor infiltrating lymphocytes. (A) Grade 0, (B) grade 1, (C) grade 2, and (D) grade 3

In the current scenario, the prognostic impact of TILs on survival outcomes and sentinel lymph node status remains controversial. The aim of this study was to analyze the association of TILs in melanoma with the sentinel lymph node status, melanoma-specific survival (MSS), and recurrence-free survival (RFS).

Material and Methods

Study Design and Population

Data was collected prospectively from all patients who underwent melanoma resection from January 2009 to December 2019. Only the resections with information on TILs according to Melanoma Institute Australia (MIA) grading system in the histopathology reports and patients with a minimum follow-up of 12 months were included in the study. Patients with more than one primary melanoma, melanoma in situ, and histopathological reports without TILs description were excluded. The sample was divided into two groups: the first group (G1) involved patients who had TILs grade 1, 2, or 3, and the second group (G2) consisted of those who had grade 0.

The diagnosis of melanoma was established by corroboration of clinical findings and histopathology in all cases. TILs were evaluated on H&E stained sections and scored according to the MIA grading system [3]. Current sentinel lymph node biopsy (SLNB) indications are Breslow’s tumor depth of more than 0.8 cm, ulceration, mitotic rate ≥ 2/mm², and lymphovascular invasion. However, indications for SLNB were not uniform throughout the study period. Adverse prognostic features were defined as Breslow ≥ 4 mm, mitotic rate ≥ 2 mitoses/mm², and ulceration. Clinical follow-up was performed postoperative on day 7 and day 30, thereafter every 3 months for the first 2 years, and every 6 months for the following 3 years. After 5 years from surgery, patients were checked annually.

Routine abdominal and lymph node ultrasound, thoracic radiography, and clinical exam were performed annually during the follow-up. Magnetic resonance imaging of the brain, PET-CT, and biopsies were performed only if recurrence were suspected. The institutional review board (IRB) approved this study. Written informed consent was exempted by the IRB owing to the study’s retrospective nature.

In all patients, the implemented surgical technique was according to NCCN melanoma guideline recommendations. In patients with SLNB indication, a lymphoscintigraphy was performed within 1 to 6 h before the surgery. Intra-operative patent blue dye injection around melanoma skin and gamma probe was utilized to confirm the localization of the sentinel lymph node.

Variables and Outcomes

Data collection included age, gender, American Society of Anesthesiologists (ASA) classification, and disease stage. Histopathological features of the primary tumor were recorded for each case: tumor thickness, mitotic rate (per mm2), presence or absence of ulceration, and TIL grade. Survival status of patients during the follow-up, disease recurrence (local or distant), and sentinel lymph node status were recorded.

Statistical Analyses

For statistical analysis, the program R statistical software (http://www.rproject.org) was used. Cox regression models were used for multivariate and univariate analyses. For the purpose of melanoma-specific survival analysis, date of surgery was considered as the starting point and melanoma-specific death as the end point. Deaths resulting from different causes other than melanoma were censored from the analyses. Recurrence-free survival time was defined as the time from surgery to first locoregional recurrence or distant metastasis.

Survival rates were calculated using the Kaplan–Meier method, and survival comparisons were performed using log-rank tests. The Student’s t test was used to compare continuous variables, whereas the χ2 test was used for categorical variables. A p value < 0.05 was considered statistically significant for all tests.

Results

During the study period, a total of 386 melanoma resections were performed, out of which 190 (49%) met the inclusion criteria. One hundred and fifty-one (79%) patients had TILs of grade 1, 2, or 3 (G1), and 39 (21%) patients had grade 0 (G2) according to MIA grading system. Demographic variables were homogeneous, and there were no significant differences in TNM stages or histopathological prognostic indicators of the primary tumor between the two groups (Table 2).

Table 2.

Demographic variables and oncological status

Demographic	G1 n: 151	G2 n: 39	p
Female, n (%)	69 (46%)	15 (38%)	0.35
Male, n (%)	82 (54%)	25 (62%)	0.35
Mean age, (range) years	61 (11–91)	63 (30–90)	0.38
TNM stage
I, n (%)	70 (46%)	17 (44%)	0.76
II, n (%)	60 (40%)	13 (33%)	0.46
III, n (%)	21 (14%)	9 (23%)	0.16
IV, n (%)	0 (0%)	0 (0%)	1.00
Prognostic factors
Thin melanoma, Breslow ≤ 0.75 mm, n (%)	33 (22%)	6 (15%)	0.37
High mitotic rate (≥ 2 mitoses/mm2), n (%)	68 (45%)	21 (51%)	0.33
Ulceration, n (%)	36 (22%)	9 (22%)	0.92

From the 151 (73%) patients with SLNB indication (G1: 120 and G2: 28 patients), positive sentinel lymph node frequency was 32%, 18%, 14%, and 0% for primary melanoma with TILs grades 0, 1, 2, and 3, respectively (p = 0.02). The recurrence rate was 18% (27 patients) in G1 and 33% (13 patients) in G2, p = 0.04. We compared the recurrence rate according to whether it was local or systemic between both groups, and we found no statistically significant difference between them. There were 11 (41%) local recurrences and 16 (59%) distant metastases in G1 versus 7 (54%) and 6 (46%), respectively, in G2, p = 0.44.

Univariate and multivariate Cox proportional hazard model are shown in Table 3 and Table 4. Multivariate Cox regression analysis demonstrated that the absence of TILs was associated with worse MSS in the univariate (HR: 3.26, 95% CI 1.47–7.21, p = 0.003) and in the multivariate analysis (HR: 2.65, 95% CI 1.14–6.20, p = 0.02). In addition, in the univariate analysis, absence of TILs was associated with statically significant worse RFS (HR: 2.61, 95% CI 1.33–5.10, p = 0.005); however, this effect was absent in the multivariate analysis (HR: 1.88, 95% CI 0.89–3.95, p = 0.09).

Table 3.

Univariate Cox regression analysis of clinical and pathologic factors predicting melanoma-specific survival and recurrence-free survival

		MSS			RFS
Variable	HR	95% CI	p	HR	95% CI	p
Female sex	0.64	0.28–1.48	0.30	0.66	0.34–1.28	0.66
Age at diagnosis, years	1.03	1–1.05	0.05	1.03	1.01–1.05	0.01
Absence of TILs	3.26	1.47–7.21	0.003	2.61	1.33–5.10	0.005
Breslow, mm	1.43	1.19–1.71	0.0001	1.45	1.23–1.72	< 0.0001
Ulceration	2.59	1.18–5.66	0.02	2.43	1.26–4.68	0.008
High mitotic rate (≥ 2 mitoses/mm2)	2.11	0.94–4.73	0.07	2.09	1.06–4.10	0.03

MSS melanoma-specific survival; RFS recurrence-free survival; TILs tumor-infiltrating lymphocytes

p < 0.05 are denoted in bold

Table 4.

Multivariate cox regression analysis of clinical and pathologic factors predicting melanoma-specific survival and recurrence-free survival

		MSS			RFS
Variable	HR	95% CI	p	HR	95% CI	p
Female sex	1.16	0.47–2.89	0.75	1.27	0.60–2.69	0.53
Age at diagnosis, years	1.04	1–1.07	0.02	1.04	1.01–1.06	0.003
Absence of TILs	2.65	1.14–6.20	0.02	1.88	0.89–3.95	0.09
Breslow, mm	1.34	1.04–1.72	0.02	1.34	1.07–1.66	0.009
Ulceration	2.20	0.85–5.73	0.10	1.90	0.86–4.23	0.11
High mitotic rate (≥ 2 mitoses/mm2)	1.21	0.43–3.43	0.72	1.32	0.57–3.05	0.52

MSS melanoma-specific survival; RFS recurrence-free survival; TILs tumor-infiltrating lymphocytes

p < 0.05 are denoted in bold

The median patient follow-up was 48 months (12–120 months). Five-year overall survival for the entire study cohort was 84%, and the 5-year RFS was 77%. The 5-year RFS was 81% in G1 and 60% in G2 (p = 0.004, log-rank test; Fig. 2). There was a statistically significant difference in MSS between groups (5-year MSS: 86% in G1 vs 75% in G2, p = 0.002, log-rank test; Fig. 3).

Fig. 2.

Recurrence-free survival in 190 patients with melanoma according to tumor-infiltrating lymphocyte grade (G1: grades 1, 2, and 3. G2: grade 0)

Fig. 3.

Melanoma-specific survival in 190 patients with melanoma according to tumor-infiltrating lymphocyte grade (G1: grades 1, 2, and 3. G2: grade 0)

Twenty-two patients received chemotherapy or immunotherapy (G1:16 and G2:6 patients). Excluding these patients from the analysis, there was a significant increase in the difference between survival curves of MSS (5-year MSS: 94% in G1 vs 75% in G2, p < 0.0001, log-rank test) and RFS (5-year RFS: 91% in G1 vs 67% in G2, p < 0.0005, log-rank test). In fact, after excluding these patients, TILs were an independent risk factor for MSS (HR: 5.29, 95% CI 1.68–16.67, p = 0.004) and RFS (HR: 3.50, 95% CI 1.28–9.57, p = 0.01). There were 73 (48%) patients with adverse prognostic features in G1 and 22 (56%) in G2. In this subgroup of individuals, we found that TILs also have a significant impact on MSS (5-year MSS: 80% in G1 vs 68% in G2, p = 0.03, log-rank test) and RFS (5-year RFS: 76% in G1 vs 52% in G2, p = 0.02, log-rank test).

Discussion

The objective of this study was to analyze oncological impact of TILs in melanoma. We found that (a) TILs grade is a predictor of SLN status and (b) patients with grade 0 TILs according to MIA grading system had a worse 5-year MSS and RFS than those with grades 1, 2 and 3. Robert Virchow was a pioneer in describing the connection between inflammation and cancer. In 1863, he suggested that leukocytes found within tumors were the cells of origin of cancers arising at sites of chronic inflammation [4]. Subsequent years were debated if their presence favored tumor growth or whether it represented the host immune response against the tumor [5]. After Clark et al. described and classified the lymphocytic infiltrate into absent, non-brisk, and brisk categories according to their distribution and intensity in primary cutaneous melanoma, Day et al. and Tuthill et al. found, in accordance with our results, that patients with TILs infiltration had a significantly better 5-year survival rate than patients without lymphocytic infiltration [1, 2, 6, 7]. However, in contrast with our findings, other authors have demonstrated no correlation between TILs and oncological outcomes [2, 8–10].

Azimi et al. who investigated the prognostic impact of TILs in melanoma, in the largest study to our knowledge, found that TILs grade was an independent predictor of survival, with high 5-year melanoma-specific survival and recurrence-free survival rates (83% and 76%, respectively, with a median follow-up of 43 months) [3].

Furthermore, a recent meta-analysis concluded that TILs were associated with a better MSS and RFS [11]. Contrary to these results, many trials have not observed any correlation between TILs and oncological prognosis [8, 9]. In our series, we found that the absence of TILs in melanoma was an independent risk factor for worse MSS, and although it was not statically significant, there was a trend towards better RFS in patients with lymphocytic infiltration.

In a study published in 2007, Taylor et al. reported that TILs predicted SLN positivity, similarly to the findings of Azimi et al. [3, 9]. However, other studies have shown no correlation between TILs and SLN status [12–14]. In 2017, Tas et al. showed that the absence of TILs in primary melanoma was a predictive factor for sentinel node positivity but not for disease recurrence [15]. In line with this, in our series, we found an inverse correlation between TILs and positive SLN. Interestingly, we also found that TILs were a protective factor for disease recurrence.

The main limitation of this study is its retrospective nature. In addition, a relatively low number of patients included in the analysis may explain the lack of statistical significance in some of the comparisons.

In conclusion, tumor-infiltrating lymphocytes in patients with melanoma have an inverse relationship with the positivity of sentinel lymph node and direct correlation with statistically significant improvement on both melanoma-specific survival and recurrence-free survival.

Author Contribution

All authors contributed to the study conception and design. Material preparation, data collection, and analysis were performed by Angeramo C.A., Laxague F., and Armella E.D. The first draft of the manuscript was written by Angeramo C.A., Laxague F., and Armella E.D., and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.

Declarations

Ethics Approval

The institutional review board (IRB) approved this study. The written informed consent was waived by the IRB owing to the study’s retrospective nature.

Conflict of Interest

The authors declare no competing interests.