Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2026 Sep 3.
Published in final edited form as: Melanoma Res. 2026 Jul 28;36(5):424–435. doi: 10.1097/CMR.0000000000001127

Spatial T Lymphocyte Infiltration into the Primary Melanoma Tumor Associates with Survival

Sharon N Edmiston 1,2,3, Yi-hsuan S Tsai 4, David L Corcoran 2,3, Honglin Hao 1, Eloise A Parrish 1, Paul B Googe 1,2,5, Ronglai Shen 6, Pei Fen Kuan 7, David W Ollila 2,8, Kathleen Conway 1,2,9, Nancy E Thomas 1,2
PMCID: PMC13537756  NIHMSID: NIHMS2194497  PMID: 42515829

Abstract

Tumor infiltrating lymphocyte grade is a prognostic indicator for patients with cutaneous invasive primary melanoma and assigned one of 3-levels (brisk, nonbrisk, or absent). We sought to develop objective, quantitative measures of T lymphocyte subtypes spatially within the tumor microenvironment. We applied multiplex whole-slide immunohistochemistry/ immunofluorescence to 80 melanomas, using S100 to identify tumor cells (S100+) and CD3 and CD8 to identify non-cytotoxic T lymphocytes (CD3+CD8−), cytotoxic CD8+ T lymphocytes (CD3+CD8+), and total T lymphocytes (CD3+CD8− plus CD3+CD8+). We used image analysis to spatially locate the T lymphocyte subtypes within the whole tumor, tumor center (175μm into the tumor), and margin (within 100μm outside the tumor). T lymphocyte subtypes were quantified for density as CD3+CD8− and/or CD3+CD8+ cells co-localized with intact nuclei /μm2. Melanoma-specific survival, adjusted for American Joint Committee on Cancer 8th edition stage, improved when a higher density of each T lymphocyte subtype (CD3+CD8−; CD3+CD8+) infiltrated 175μm into the tumor center or total T lymphocytes infiltrated into the whole tumor (hazard ratios 0.72–0.77; all p<0.05). Improved mlanoma-specific survival significantly associated with high density of CD3+CD8− T lymphocytes in the margin after stage adjustment. Our results indicate spatial location of T lymphocyte subset infiltration within the tumor microenvironment is important for predicting melanoma-specific survival.

Keywords: tumor infiltrating lymphocytes, pathology, skin cancer, tumor microenvironment, melanoma

Introduction

The immune system is our first line of defense against cancer and the ability to mount a robust anti-tumor immune response in early melanomas is essential for controlling disease progression. [1, 2] Tumor infiltrating lymphocytes (TIL) are an indicator of the host immune response in the tumor microenvironment (TME). In primary invasive melanomas, TIL grade is scored visually by pathologists as brisk, nonbrisk, or absent based on the density and distribution of lymphocyte infiltrate in and around the tumor [3, 4]. As demonstrated by our population-based Genes, Environment, and Melanoma (GEM) study and others [3, 5–7], brisk TIL grade is positively associated with better melanoma-specific survival (MSS), independent of American Joint Commission (AJCC) stage, suggesting the patient’s immune response to the primary tumor can improve survival from melanoma. AJCC 8th edition (AJCC8) staging recommends the routine documentation of TILs in primary melanomas at diagnosis; however, TIL grade is not incorporated into AJCC8 stage [8]. Recent meta-analyses support the association of brisk TIL grade with improved overall survival, but nonbrisk TIL grade or moderate immune infiltration was not associated with MSS in several studies [9–12].

While TIL presence associates with improved patient survival in multiple cancers including melanoma, breast, and colorectal cancer [13–16], it is unclear if localized specific T lymphocyte subtype quantifications in primary melanoma can provide a more precise or accurate prognostic indicator of survival than TIL grade alone [17, 18]. TILs are a heterogeneous group of immune cells that produce a coordinated response to disease. CD3+ and cytotoxic CD8+ T lymphocytes are markers of better prognosis in many tumor types [19], including in metastatic melanoma [20], colon cancer [21], and breast cancer [22]. Since cytotoxic CD8+ T lymphocytes are abundant in early immune response and suggested to mechanistically drive the adaptive immune response, most studies focus on CD8+ T lymphocytes [23]. However, recent studies note the importance of CD4+ T lymphocytes in an alternative role for destroying tumor cells [24, 25]. Further, Kruse et al. show differences in T lymphocyte subtype infiltration patterns with CD4+ T effector lymphocytes accumulating at the tumor periphery [24].

It is also unclear if spatial information regarding the T lymphocyte subtypes densities within the primary melanoma TME can add prognostic information to melanoma staging. The activated T lymphocyte response can be attenuated or evaded due to the changing TME [2]. The TME consists of a complex environment of tumor, stromal, epithelial, and immune cells that can be altered by growth factors and cytokines. For example, melanomas with BRAF V600E mutation produce increased cytokines, IL-1β, IL-6, and IL-8 leading to activation of stromal fibroblasts [26]. Increased cancer-associated fibroblasts lead to CD8+ T lymphocyte exclusion in tumors [27]. Also, tumors can exhibit an adaptive immune resistance through production of immunosuppressive proteins evidenced by programmed cell death protein ligand 1 (PD-L1) expression on tumor cells to inhibit T lymphocytes [28]. These mechanisms can alter the T lymphocyte subtype densities and locations, producing different tumor ecosystems [29].

In this study, we examine the immune response in primary melanoma by measuring T lymphocyte subtype densities, their spatial locations within the TME, and their relationships with patient survival. Our hypothesis is that non-cytotoxic T lymphocyte (NCTL; CD3+CD8−), cytotoxic CD8+ T lymphocyte (CTL; CD3+CD8+), and total T lymphocyte (TTL; CD3+CD8− and CD3+CD8+) densities and their TME spatial locations will inform survival outcomes for primary melanoma better than AJCC8 stage and/or TIL grade alone. In this study, CD3+CD8− T lymphocytes are a surrogate for double-negative or CD4+ T lymphocytes and CD3+CD8+ represent cytotoxic CD8+ T lymphocytes. We expect a higher proportion of CD3+CD8− T lymphocytes will be CD4+ lymphocytes since CD3+ T lymphocytes double-negative for CD4 and CD8 are estimated to be only 1–3% of the population[30]. Thus, we measured CD3+ and CD8+ T lymphocytes using whole-slide multiplex immunohistochemistry/immunofluorescence (mIHC/IF) staining and determined their distribution in the whole tumor (WhT), tumor center (TC), and tumor margin (M) using S100+ staining to define the tumor in primary melanomas and image analysis to determine spatially defined TME locations.

Methods

Patients and tissues

Formalin-fixed, paraffin-embedded (FFPE) cutaneous primary invasive melanomas (n=80) diagnosed between 2001 and 2012 were assembled from the University of North Carolina (UNC) Hospitals’ pathology archives based on original diagnoses abstracted from pathology reports. The Institutional Review Board at UNC approved the study. They determined that the research met the criteria for waiver of informed consent for research (45 CFR 46.116d) and waiver of Health Insurance Portability and Accountability Act authorization (45 CFR 164.512[i][2][ii]). Melanomas were chosen to span AJCC8 tumor stages and included common and less common (e.g., Spitz, nevoid, and desmoplastic melanomas) subtypes. These melanomas are a subgroup of the 89 melanomas utilized for methylation profiling by Conway et al. [31, 32]. Sex, age, race, anatomic site, nodal status at diagnosis, and survival data were abstracted from the UNC medical record. Patient samples were collected retrospectively prior to any treatment with targeted or immune modulating therapies. Tissues used in this study represent the original diagnostic biopsies. De-identified tissues were sectioned at 4-micron thickness and stored in a nitrogen chamber before mIHC/IF staining.

Hematoxylin and eosin (H&E) staining was performed on the first and last section of each tissue to identify presence of primary melanoma [33]. Histologic review of all melanomas was conducted by a dermatopathologist (PG) to determine histologic subtype, Breslow thickness, ulceration, mitoses, and TIL grade. Melanomas were classified according to published histopathology criteria [16, 17]. Breslow thickness corresponds to the thickest Breslow depth noted in the original biopsy prior to wide local excision. Ulceration was recorded as present or absent. Mitoses were recorded as ≤1 or >1per mm2 [18]. TIL grade was scored as brisk, nonbrisk, or absent, by using a previously defined grading system [2–4]. Pathologists define TIL grade as brisk when a moderate to marked TIL infiltrate exists along the vertical component and/or as a diffuse lymphocytic infiltrate within the tumor; nonbrisk when a mild to moderate TIL infiltrate exists along the vertical component or as scattered foci within the tumor; and absent when few or no lymphocytes infiltrate along the vertical component or in the tumor [3, 4, 6]. Melanomas were restaged according to AJCC8 specifications [34] using medical record and pathology review data.

Multiplex Immunohistochemistry / Immunofluorescence Stains

All mIHC/IF stains and quantification of T lymphocytes in melanoma were performed in the UNC Pathology Services Core from whole mount slides. T lymphocytes were localized using monoclonal antibodies against CD8 (SP16) (Cell Marque, Cat# 108R, RRID: AB_2892088) and CD3 (LN10) (Leica Biosystems, Cat # PA0553, RRID: AB_10554601) and melanoma cells were localized using a polyclonal antibody against S100 (Agilent, Cat# Z0311, RRID: AB_10013383). The source, concentrations, pretreatment protocols and incubation times of all primary antibodies used in this study are listed in Table S1, as well as all reagents, detection kits, fluorescently labeled secondary antibodies, and Tyramide Signal Amplification (TSA) reporter systems used for staining [35].

mIHC/IF stains were performed in the Leica Bond-III fully automated staining platform (Leica Biosystems Inc., Norwell, MA). Slides were dewaxed in Bond™ Dewax solution (AR9222) and hydrated in Bond Wash solution (AR9590). Epitope retrieval for all targets was performed for 20 minutes in Bond-epitope retrieval solution 1 pH6.0 (AR9661) or solution 2 pH9.0 (AR9640). The epitope retrieval was followed with 5 minutes endogenous peroxidase blocking using Bond peroxide blocking solution (DS9800). The antibodies in Table S1 were applied and detected serially. Stained slides were counterstained with Hoechst 33258 and mounted with ProLong® Diamond Antifade Mountant (#H3569, Life Technologies, Carlsbad, CA). Positive controls (melanoma; tonsil) and negative controls (no primary antibody) were included during each run. Single stain controls were done for mIHC/IF in which one primary antibody was omitted to make sure that cross reactivity between the antibodies did not occur. Assay reproducibility was reliable and reproducible across repeats within multiple tissues having <10% coefficients of variation below standard recommendations (coefficients of variation ≤20%) [36] (Table S2).

Imaging and image analysis

H&E slides were digitally imaged in the Aperio ScanScope XT (Leica Biosystems Inc., Buffalo Grove, IL) using a 20x objective. High resolution acquisition of whole-slide mIHC/IF slides in the DAPI, AF488, Cy3, and Cy5 channels was performed in the Aperio ScanScope FL (Leica) using the 20x objective or the 10x objective in the Aperio-VERSA digital pathology scanner. Nuclei stained with Hoescht 33258 were visualized in the DAPI channel (blue).

Definiens Architect XD version 64 2.3 software was used for image segmentation and analysis of slides stained with S100_Alexa555; CD3_ Cy5; and CD8_AF488 using tissue detection, TME, and co-localization solutions. The TME algorithm was trained to define tumor mask based on the S100+ stain and then create the regions of interest (ROI): whole tumor (WhT) and tumor margin (M). The WhT and M ROIs are mutually exclusive. The width and position of a margin ROI was guided via a thickness threshold. The thickness threshold was set at 100μm for the M surrounding the S100 defined WhT. Cells expressing S100, CD3, and/or CD8, and Hoechst 33258 nuclear stain were measured for number of stained cells and intensity across the whole slide and within each ROI. Densities of CD3+, CD8+, and CD3+CD8+ co-expressing T lymphocytes in the ROI were calculated as the number of cells per area (μm2). For the following analyses, (CD3+CD8−) were noted as NCTL (non-cytotoxic T lymphocytes); (CD3+CD8+) as CTL (cytotoxic CD8+ T lymphocytes); and (CD3+CD8+ and CD3+CD8−) as TTL (Total T lymphocytes). Given that the lymphocyte percentages and cell densities were highly correlated for NCTL (r=0.968, p<10e-6 and CTL (r=0.970, p<10e-6), mean T lymphocyte subtype (NCTL, CTL, and TTL) densities were used for the analyses. Considering that CD3+ T lymphocytes double-negative for CD4 and CD8 are only 1–3% of the T lymphocyte population [30], CD3+CD8− T lymphocytes were used as a surrogate for CD4+ T lymphocytes.

The TME was analyzed using the Definiens Architect XD software to segment tumor tissues into ROIs (Figure 1). The S100+-stained melanoma areas selected by the Definiens Architect XD software were visually compared to tumor area identified by the dermatopathologist, and we confirmed that the software accurately identified S100+ stained tumor. ROIs were defined in relation to the S100 positively stained melanoma, with WhT being the entire S100+-stained melanoma areas and M defined as 100 μm outside and surrounding the S100+-stained WhT areas. Using the Definiens Architect XD software, we also created an RO1 of the tumor center (TC), defined as the central tumor area 175 μm internal to the edge of the S100+-stained tumor, for a secondary analysis. Subsequent analyses focused on the WhT, TC, and M ROIs.

Figure 1: Definiens defined TME regions of interest (ROIs) in primary melanoma.

Figure 1:

Open in a new tab

(A) A schematic is shown defining whole tumor (WhT) (orange) and margin (M) (beige) ROIs. (B) A schematic is shown defining whole tumor (WhT) (orange), tumor center (TC) (apricot), and margin (M) (beige) ROIs. The TC was defined as 175um into the tumor. The M was defined as within 100um outside and surrounding the WhT.

Statistical Methods

Standard descriptive statistics were calculated to define the clinicopathological characteristics of the primary melanomas and T lymphocyte subtype densities within ROIs. Kruskul-Wallis tests and Wilcoxon pairwise analysis tests were used to compare differences in clinicopathological characteristics for categorical variables between the mean T lymphocyte subtype densities within the ROIs.

MSS is reported with patients censored if they died due to any cause other than melanoma. For progression-free survival (PFS), the progression event is the incidence of any disease recurrence defined as in transit (over 2 cm from the primary tumor but not reaching the regional lymph node station), regional (within the regional lymph node station), and distant (occurring beyond the regional lymph node station) and censored for patients who develop a second or third primary melanoma distant to the index lesion. Survival curves by high, medium, and low tertiles of T lymphocyte subtype densities within the WhT, TC, or M were constructed using the Kaplan-Meier (KM) method to examine time to event (i.e., MSS or PFS) and compared using log-rank tests [37]. KM survival curves are unadjusted for clinicopathologic variables.

Hazard ratios (HRs) and 95% CIs for MSS or PFS by mean T lymphocyte subtype densities within the WhT, TC, and M were estimated in univariable Cox-proportional hazard regression models. T lymphocyte densities were treated as continuous variables and linear trend was tested using Wald tests. Multivariable Cox-models also included pathologic AJCC8 stage (IA/1B, IIA/IIB/IIC, IIIA/IIIB/IIIC, IV) at diagnosis and TIL grade (brisk, nonbrisk, absent, undetermined). R version 4.0.2 was used for analyses.

Comparisons between groups were performed with student’s t-tests and ANOVA where appropriate, and unadjusted and adjusted p-values are reported.

Results

Clinicopathologic characteristics

The clinicopathologic characteristics of the 80 cutaneous primary invasive melanomas in this study are displayed in Table S3. Of the 80 patients, 61.3% are male, 76.3% are 50 years or older, 93.8% are white of European origin, and 50% had superficial spreading melanoma (SSM). The melanomas have a mean Breslow thickness of 1.6 millimeters (mm) (range of 0.4–17.0 mm), 36.3% demonstrate ulceration, 65% have presence of mitoses equal to or greater than 1per mm2, and 34% have nodal involvement. TIL grades for melanomas are 22.5% brisk, 53.8% nonbrisk, 22.5% absent, and 1.3% undetermined. The tumor stages span all AJCC8 pathologic stage classifications with the majority being stage I (27.5%) or II (40%).

Spatial profiles of T lymphocyte density

Primary melanoma (n=80) sections were mIHC/IF stained for S100-CD3-CD8 and the nuclear stain and segmented into WhT, TC, and M ROIs (Figure 1). The M was defined as 100 μm outside and surrounding the S100+ stained tumor (Figure 1A). The TC was defined as 175 μm into the S100 stained tumor (Figure 1B). Density distributions of NCTL, CTL, and TTL subtypes in the WhT, TC, and M ROIs are displayed in Table S4. Overall, NCTLs had a higher mean density than CTLs in the WhT and TC (p=0.0001). However, patients with a robust CTL response (>850 cells/μm2) have a greater mean density of CTLs than NCTLs in the WhT and TC.

The spatial locations of T lymphocyte infiltrates also vary which can be seen in tumor images and bar graphs of T lymphocyte densities by ROIs (Figure 2). Figure 2A shows a primary melanoma scored as brisk TIL grade, with T lymphocyte infiltration into the WhT and TC but few T lymphocytes in the M. Figure 2B shows a melanoma with a nonbrisk TIL grade; the T lymphocytes mostly accumulated in the tumor periphery and the M. Figure 2C shows a melanoma scored with absent TIL grade that had only a few detectable T lymphocytes evident in the M. Figure 2D displays the density for each T lymphocyte subtype within the ROIs for each TIL grade. Brisk TIL grade shows a high infiltration of CTL in the WhT and TC. However, the melanoma with nonbrisk TIL grade has a significant reduction in CTL with an increase in NCTL density in the WhT/TC, and an increased density of T lymphocytes in the margin; while a melanoma with absent TIL grade has fewer T lymphocytes and these were mainly in the margin.

Figure 2: Multiplexed Immunohistochemistry / Immunofluorescence (mIHC/IF) in primary melanoma with brisk, nonbrisk, or absent TIL grade in the ROIs.

Figure 2:

Open in a new tab

Shown are three melanomas (A) a superficial spreading melanoma with 1.3 mm Breslow thickness and brisk TIL grade; (B) a lentigo maligna melanoma with 4.05 mm Breslow thickness and nonbrisk TIL grade; and (C) a nodular melanoma with 4.1 mm Breslow thickness and absent TIL grade. For each melanoma, shown are the (a) H&E; (b) Definiens defined WhT (orange) and M (tan) ROIs; (c) whole tissue mIHC/IF [scale bars, 100μm (A; C) and 2mm (B)]; (d) mIHC/IF in the white box region [scale bars, 50μm (A, d-g; B, d-g; C, d-g)]; (e) S100 melanoma cells in cyan (Alexa555) in the region; (f) CD3+ T lymphocytes in pink (Cy5) in the region; and (g) CD8+ T lymphocytes in green (Alexa488) in the region. The Hoechst 33258 nuclear counterstain is in dark blue. (D) For each melanoma, a bar graph displays the median densities defined by counts of non-cytotoxic T lymphocytes (NCTL; CD3+CD8−), cytotoxic T lymphocytes (CTL; CD3+CD8+), or Total T lymphocytes (TTL; CD3+CD8− and CD3+CD8+) co-localized with intact nuclei per μm2 (determined by the Hoechst 33258 nuclear counterstain) in the WhT, TC, and M.

Relationship of T lymphocyte densities to clinicopathologic characteristics

Boxplots show the relationships of the clinicopathological characteristics to mean total T lymphocyte (TTL) densities in the WhT, TC, and M (Figure S1). No statistically significant associations were found for sex, age, or race categories. TTL density is associated with AJCC pathologic stage in the WhT (p=0.02) and TC (p<0.01) but not the M. Melanomas with Breslow thickness < 2mm or that were non-ulcerated had higher mean TTL density in the WhT (p=0.05; p=0.02, respectively) and TC (p<0.01, p<0.01, respectively) but not the M.

SSM, lentigo maligna melanoma (LMM) and unclassified (UNCL) tumors had higher mean TTL density in the WhT and TC, while LMM had lower mean TTL density in the M compared to other histologic subtypes. Box plots show UNCL and SSM display the highest T lymphocyte density scores while nodular melanoma (NM) and acral lentiginous melanoma (ALM) display the lowest T lymphocyte density scores (Figure S1). Wilcoxon pairwise analysis show SSM and UNCL histological subtypes had significantly increased T lymphocyte density when compared to ALM in the WhT (p<0.05) (Figure S1A). Pairwise analysis within the TC, shows SSM has increased T lymphocyte density when compared to ALM (p<0.01) and LMM (P<0.05) (Figure S1B). Melanomas with brisk TIL grade had higher mean NCTL, CTL, and TTL densities in the WhT and TC than nonbrisk or absent TIL grade tumors (Figure 3). Notably, nonbrisk TIL grade melanomas had increased NCTL density in the M. Overall, increased T lymphocyte densities in the tumor regions associated with better clinical prognostic features.

Figure 3: Associations of T lymphocyte subtype densities with TIL grade in the ROIs.

Figure 3:

Open in a new tab

Box plots of T lymphocyte subtype densities by TIL grade for NCTL (CD3+CD8−), CTL (CD3+CD8+) and Total T lymphocytes (CD3+CD8− and CD3+CD8+) in the (A) whole tumor (WhT), (B) tumor center (TC), and (C) margin (M). ANOVA p values are displayed on each graph.

Survival and T lymphocyte density

MSS and PFS were examined using the KM method according to NCTL, CTL, or TTL density divided into high, medium, and low tertiles measured within the WhT, TC, or M (Figure 4). Log-rank tests comparing survival outcomes between groups demonstrated improved MSS when a high NCTL density was in the WhT or a high NCTL, CTL, or TTL density was in the TC (all p<0.05) (Figure 4A). PFS was improved when a high NCTL, CTL, or TTL density was in the WhT or TC (all p<0.05) (Figure 4B). For densities in the M, only TTL was associated with improved PFS (p=0.02).

Figure 4: Kaplan–Meier curves and log rank p-values of T lymphocyte subtype densities in the whole tumor (WhT), tumor center (TC), and margin (M) in primary melanoma.

Figure 4:

Open in a new tab

(A) Kaplan–Meier curves of melanoma-specific survival by tertiles of NCTL, CTL, and TTL densities in the WhT, TC, and M. (B) Kaplan–Meier curves of progression-free survival by tertiles of NTCL, CTL and TTL densities in the WhT, TC, and M.

In Cox proportional hazard models (Table 1) with T lymphocyte subtype densities (as continuous variables), higher mean NCTL, CTL, or TTL density in the WhT or TC was associated with better MSS in unadjusted or the base model adjusted for age and sex (all p<0.05). With the addition of AJCC8 pathologic stage, increased T lymphocyte subtype densities remained significantly associated for improved MSS with infiltration of each T lymphocyte subtype into the TC (all p<0.05), while only NCTL infiltration was significant in the WhT (p=0.02). After adjusting for AJCC8 stage and TIL grade only higher mean NCTL in the TC (p=0.01) showed a significant association for MSS while the association was attenuated for TTL and CTL. Similar results were seen in the WhT analysis with the association being attenuated after controlling for AJCC8 stage and TIL grade together. For T lymphocytes located in the margin, only high NCTL were associated with improved MSS after controlling for AJCC8 stage (p=0.04) or AJCC8 stage and TIL grade together (p=0.05).

Table 1:

Hazard ratios for melanoma specific survival and progression free survival

A. Melanoma-specific survival Whole Tumor (WhT)1 Tumor Center (TC)1 Margin (M)1
NCTL CTL TTL NCTL CTL TTL NCTL CTL TTL
Hazard ratio unadjusted (95% CI) 0.62 (0.47–0.80) 0.82 (0.68–0.98) 0.69 (0.54–0.89) 0.67 (0.57–0.79) 0.77 (0.65–0.91) 0.69 (0.58–0.82) 0.85 (0.71–1.02) 0.93 (0.78–1.11) 0.89 (0.75–1.05)
MSS Cox p-value <0.01 0.03 <0.01 <0.001 <0.01 <0.001 0.08 0.40 0.16
Hazard ratio adjusted for age and sex2 (95% CI) 0.59 (0.43–0.80) 0.83 (0.68–1.00) 0.69 (0.53–0.90) 0.65 (0.53–0.78) 0.78 (0.66–0.92) 0.68 (0.56–0.82) 0.86 (0.72–1.03) 0.93 (0.78–1.12) 0.90 (0.76–1.06)
MSS Cox p-value <0.001 0.05 <0.01 <0.001 <0.01 <0.001 0.10 0.45 0.21
Hazard ratio adjusted for AJCC8 stage3 (95% CI) 0.67 (0.47–0.94) 0.88 (0.72–1.07) 0.77 (0.58–1.02) 0.68 (0.54–0.86) 0.83 (0.69–1.00) 0.73 (0.58–0.91) 0.80 (0.65–0.99) 0.91 (0.75–1.10) 0.85 (0.69–1.03)
MSS Cox p-value 0.02 0.21 0.07 <0.01 0.05 <0.01 0.04 0.33 0.1
Hazard ratio adjusted for AJCC8 stage and TIL grade4 (95% CI) 0.77 (0.51–1.14) 1.00 (0.80–1.25) 0.91 (0.66–1.25) 0.70 (0.52–0.93) 0.92 (0.75–1.13) 0.78 (0.60–1.03) 0.78 (0.61–1.00) 0.95 (0.77–1.18) 0.85 (0.67–1.07)
MSS Cox p-value 0.19 0.99 0.55 0.01 0.43 0.08 0.05 0.66 0.17
B. Progression-free survival Whole Tumor (WhT)1 Tumor Center (TC)1 Margin (M)1
NCTL CTL TTL NCTL CTL TTL NCTL CTL TTL
Hazard ratio unadjusted (95% CI) 0.57 (0.43–0.75) 0.78 (0.66–0.92) 0.66 (0.52–0.83) 0.55 (0.45–0.68) 0.70 (0.59–0.82) 0.58 (0.48–0.71) 0.83 (0.71–0.97) 0.88 (0.76–1.03) 0.85 (0.74–0.99)
PFS Cox p-value <0.001 <0.01 <0.001 <0.001 <0.001 <0.001 0.02 0.10 0.04
Hazard ratio adjusted for age and sex2 (95% CI) 0.55 (0.40–0.75) 0.78 (0.66–0.92) 0.65 (0.50–0.84) 0.54 (0.44–0.67) 0.70 (0.59–0.83) 0.0.58 (0.47–0.71) 0.83 (0.71–0.98) 0.88 (0.75–1.03) 0.86 (0.74–0.99)
PFS Cox p-value <0.001 <0.01 <0.001 <0.001 <0.001 <0.001 0.03 0.11 0.04
Hazard ratio adjusted for AJCC8 stage3 (95% CI) 0.62 (0.45–0.87) 0.82 (0.69–0.98) 0.71 (0.55–0.92) 0.58 (0.45–0.74) 0.74 (0.62–0.88) 0.62 (0.49–0.78) 0.77 (0.63–0.93) 0.84 (0.71–1.00) 0.79 (0.67–0.95)
PFS Cox p-value <0.01 0.02 <0.01 <0.001 <0.001 <0.001 0.01 0.05 0.01
Hazard ratio adjusted for AJCC8 stage and TIL grade4 (95% CI) 0.76 (0.52–1.11) 0.92 (0.75–1.13) 0.85 (0.63–1.14) 0.61 (0.46–0.81) 0.81 (0.67–0.98) 0.67 (0.51–0.88) 0.77 (0.62–0.96) 0.90 (0.74–1.09) 0.81 (0.66–1.00)
PFS Cox p-value 0.15 0.44 0.28 <0.001 0.03 <0.01 0.02 0.26 0.05
Open in a new tab

Abbreviations: AJCC8, American Joint Committee on Cancer AJCC 8th edition; CI, confidence index;TIL, tumor infiltrating lymphocyte; MSS, melanoma-specific survival; PFS, progression-free survival; NCTL, CD3+CD8- non-cytotoxic T lymphocytes; CTL, CD3+CD8+ cytotoxic T lymphocytes; TTL, CD3+CD8- and CD3+CD8+ lymphocytes.

1

T cell density (#cells/mm2) in the whole tumor and tumor center.

2

Base model adjusted for age (continuous) and sex.

3

Base model plus pathologic AJCC8 stage (IA/1B, IIA/IIB/IIC, IIIA/IIIB/IIIC, IV).

4

Base model plus pathologic AJCC8 stage and TIL grade (brisk, nonbrisk, absent).

All high T lymphocyte subtype densities in the WhT or TC were each associated with improved PFS and remained significantly associated adjusting for AJCC8 stage in the models (all p<0.05). Improved PFS associated with higher mean NCTL, CTL and TTL remained significant for infiltration into the TC after adjusting for AJCC8 stage (all p<0.01) or AJCC8 stage and TIL grade together (all p<0.05). However, the association with PFS with a high mean density of any T lymphocyte subtype infiltration in the WhT was attenuated after adjusting for AJCC8 stage and TIL grade together. A high density of NCTL and TTL in the margin was significant for improved PFS and remained significant after adjusting for AJCC8 stage (p=0.01) or AJCC8 stage and TIL grade together (p≤0.05). However, a high density of CTL in the margin was significant only after adjusting for AJCC8 stage (p=0.05).

Discussion

Our image analyses illustrate the importance of both the density and location of T lymphocyte subtypes in primary cutaneous melanoma. MSS is improved when a higher density of each T lymphocyte subtype (NCTL, CTL, and TTL) infiltrates 175μm into the TC or a high density of TTL infiltrates into the WhT, even after adjusting for AJCC8 stage. With the added adjustment for TIL grade, improved MSS maintained significant associations with high density of NTCL infiltrates into the TC. For T lymphocytes accumulating in the margin surrounding the tumor, only a high density of NCTL were associated with improved MSS after adjusting for AJCC8 stage or AJCC8 stage and TIL grade together. These results suggest that NCTL (double-negative or CD4+ T lymphocytes) are active throughout the tumor and retain important immune function at the edge of the tumor while effector CD8+ CTL infiltrate into the center of the tumor. Our results indicate that both density and spatial location of the T lymphocyte subtype infiltration within the tumor is important for predicting MSS and warrants additional study.

A strong immune response is widely accepted as being linked to better survival outcomes in melanoma and other cancers, such as breast, lung, gastric and colorectal [14, 15, 38–40]. T lymphocytes, a key component in TIL grade, have been shown in multiple studies to correlate with survival based on CD3 and CD8 expression levels [41–43]. Meta-analyses by Fu et al. [9] support the positive association of CD3+, CD4+, and CD8+ T lymphocytes with improved overall survival in melanoma. Consistent with these findings, our study demonstrates that higher densities of any T lymphocyte subtypes in the WhT, TC, or M, are a favorable prognostic indicator, being significantly associated with longer PFS, even after controlling for stage.

Our results indicate that the spatial localization of T lymphocyte concentration within the tumor microenvironment (TME) is important for predicting MSS. Although an abundant immune response has consistently been associated with PFS and overall survival, associations with MSS have shown inconsistencies in previous studies [9–12]. Our data suggest that this may be due to spatial location of T lymphocytes within the TME. MSS is specifically associated with a high T lymphocyte density infiltrating 175μm into the center of the tumor. This is especially true for CTL where the effects are attenuated in the whole tumor after adjusting for AJCC8 stage and TIL grade suggesting that effective cytotoxic CD8+ T lymphocytes infiltrate to the center of the tumor. Although, an exception may be seen in early-stage tumors representing an initial immune response showing active recruitment of CTL into the tumor as evidenced by PFS significantly association with CTL in the margin only after controlling for stage. Research focusing on T lymphocyte location within the TME indicates that CD8+ T lymphocytes positioned near or within tumors are particularly crucial for survival. Toki et al. [44] analyzed melanoma tissue microarrays and found that high tumoral expression of CD3 or CD8 predicts PFS, but only high CD8 expression is a significant predictor of overall survival (OS). Similarly, De Logu et al. [41] reported that a high density of both CD3+ and CD8+ T lymphocytes correlates with better MSS, especially when a CD8+ T cells are in close proximity to melanoma cells [45]. Attrill et al. [46] also found that CD8+CD103+ T lymphocytes near melanoma cells are associated with improved recurrence-free survival. Our findings show that a high density of cytotoxic CD8+ T lymphocytes infiltrating 175μm into the tumor center is linked to better MSS, emphasizing the importance of T cell penetration into the tumor.

Recent studies illustrate that a diverse immune response consisting of dendritic cells along with CD4+ and CD8+ T lymphocytes is needed to eliminate cancer cells. Espinosa-Carrasco et al. found that intratumoral cellular triads of CD4+, CD8+, and dendritic cells was important for effective killing of cancer cells [47]. A study examining response to immune checkpoint inhibitors in hepatocellular carcinoma found CD8+ T cells and CXCL13+ CD4+ T helper cells interact with dendritic cells to stimulate an effective response [48]. Consistent with these findings, our data shows the importance of high density of NCTL (double-negative or CD4+ T lymphocytes) along with CD8+ CTL present in the tumor significantly improves survival. NCTL infiltration consistently demonstrated strong associations with survival. Recent studies suggest that CD4+ T lymphocytes accumulated in the tumor periphery may interact with major histocompatibility complex (MHC)-II and be effective in destroying tumor cells that express MHC-II markers[24]. These data suggest that CD4+ T lymphocyte density within the tumor is an important survival marker. Our data also shows an increase of NCTL in the margin of nonbrisk melanomas. Thus, T lymphocytes limited to the tumor periphery may indicate an enrichment of CD4+ T lymphocytes with reduced CD8+ CTL infiltration. This suggests that specific T lymphocyte subtypes may be identified by spatial locations within the TME. Our study indicates that along with density and spatial localization of T lymphocytes, it may be important to identify that both CD4+ and CD8+ T lymphocytes have infiltrated into the tumor center to effectively inform prognosis.

Brisk TIL grade in primary melanoma significantly associates with better prognosis. However, moderate immune infiltration or nonbrisk TIL grades did not consistently correlate with MSS across several studies. While TIL grade uses the location of the lymphocytic infiltrate in the definitions, it does not clearly delineate the location of the densities in the scoring. Thus, joint consideration of T lymphocyte density and more precise location in the TME may be more prognostic (and potentially predictive) than TIL grade alone. Our method provides a more quantitative measure of T lymphocytes and the spatial location within the TME. Our future studies aim to use artificial intelligence to create a digital pathology algorithm to simulate this analysis on H&E tissues.

Strengths of our study include centralized pathology review allowing collection of TIL grade and MSS; treatment naïve melanomas collected before current treatment regimens. Also, unlike most studies where margin contains the tumor periphery, our ROIs for margin and tumor are exclusive of each other, and, thus, the T lymphocytes densities do not overlap. Our study examined primary melanoma across all pathologic stages with a majority of early stage I and II melanoma with mean Breslow thickness of 1.6mm. The increased number of lower stage melanomas allows us to better examine TILs in early melanomas compared to other studies where the majority of primary melanomas have a Breslow >2mm.

Limitations of our study are the technology, modest sample size, and lack of an independent test set. Our study was limited by the number of immune markers utilized since the visualization techniques applied limit the number of antibodies that can be examined on a single slide. Although we did not measure CD4, separating CD3+CD8− and CD3+CD8+ cells created a proxy for CD4 lymphocytes noted as NCTL. However, our use CD3+CD8− lymphocytes as a surrogate for CD4+ T lymphocytes (both helper and Tregs) may exhibit some cellular misclassification since these cells may represent a mixture of CD3+ T lymphocytes double-negative for both CD4 and CD8 including αβ T cells, gδd T cells, and NKT-lineage subsets [30, 49]. Our algorithm could be applied to an expanded multiplex immunofluorescence assay using new technologies such as EVOS S1000 Spatial Proteomics (Thermo Fisher Scientific) technology that allows visualization of 9 markers simultaneously. Other new multiplex technologies[50], such as Ultivue InSitu-plex, PhenoCycler (CODEX), multiplexed ion-beam imaging (MIBI), and the NanoString GeoMx protein assays, will enable future studies to investigate a greater expanse of immune regulatory cell subtypes within the TME to better identify targets for melanoma treatments. With 53.8% of our samples designated nonbrisk, our ability to accurately measure differences by TIL grade was limited. Our study is limited by the absence of data on biopsy-related scarring given that scarring in the skin may affect T lymphocyte motility. Finally, the ability to visualize melanoma cells to create our ROIs may be problematic in melanomas with diffuse cellularity intermixed with stroma such as desmoplastic melanomas. Our study uses S100 staining to identify melanoma cells which is noted to be the best marker for desmoplastic melanomas [51]. We will expand melanocytic markers (i.e. SOX10, MLANA) in our future studies to improve ROI detection. Our study has one desmoplastic melanoma, and our algorithm was able to create ROIs for tissues from this melanoma.

Our results found a high density of T lymphocyte subtypes (NCTL, CTL, TTL) in the WhT and TC significantly associated with prognostic factors (AJCC8 stage, Breslow, ulceration) and improved survival. While MSS was significantly associated with T lymphocytes entering 175μm into the tumor center, high T lymphocyte density in the margin only shows significant MSS after controlling for stage and TIL grade when a high density of NCTL is present in the margin. The significance of NCTL (double-negative or CD4+ T lymphocyte) infiltration into the center of the tumor highlights the importance of a diverse T lymphocyte response. Specific T lymphocyte subtypes may be identified by spatial locations within the TME. Our data suggests that T lymphocytes in the center of the tumor are likely to be enriched for effector CD8+ T lymphocytes, while T lymphocytes in tumor margin likely indicate an enrichment of double-negative or CD4+ T lymphocytes with reduced CD8+ CTL infiltration. Future studies will be necessary to determine the best methodology for TIL spatial quantitation. It will be important to examine an expanded set of immune and tumor biomarkers to measure cellular heterogeneity within the evolving TME and determine how other cells, such as macrophages and fibroblasts, may alter tumor immunity. In our upcoming studies, we are performing immunophenotyping to better identify T lymphocyte phenotypes localized within the ROIs. If T lymphocyte subsets are primarily enriched in specific ROIs (e.g., activated, effector T cells in the TC), this could provide clinically useful information. Future studies will examine if TMB, mutation status, or CpG-methylation phenotype is associated with T lymphocyte immunophenotypes and their localization within the ROIs examined in this study. Incorporation of spatial T lymphocyte quantification in TIL grading may shed light on tumor biology and potentially help identify patients best able to respond to immune modulating therapies. Our data indicates that measures of density, subtype, and spatial localization of T lymphocytes in the tumor microenvironment are important for prognosis.

Supplementary Material

Supplement_Tables
Supplement_Figure_S1
Supplement_Figure_S1_Legend

Acknowledgments

Supported by the National Cancer Institute grants R21CA134368, R33CA160138, and R03CA199487 to K Conway and NE Thomas, P01CA206980 to M Berwick and NE Thomas, R01CA233524 to NE Thomas, M Berwick, CB Begg and H Anton-Culver. 1R01CA310060 to R. Shen and NE Thomas and P30CA016086 to HS Earp. SNE was supported in part by a grant from the National Institute of General Medical Sciences under award 1T32GM135128. Research reported in this publication was also supported by the National Institute of Environmental Health Sciences of the National Institutes of Health under award number P30ES010126, the University of North Carolina Center for Environmental Health and Susceptibility. This work was also supported by the Lineberger Comprehensive Cancer Center Translational Pathology Laboratory Core Facility and Bioinformatics Core. We would like to acknowledge Pamela A. Groben MD, Professor Emeritus, Departments of Dermatology and Pathology and Laboratory Medicine, for participating in the pathology review and Nana Feinberg and Bentley Midkiff in the UNC Pathology Services Core for processing multiplex immunohistochemistry/immunofluorescence.

Abbreviations:

TIL

tumor infiltrating lymphocyte (TIL)

TME

tumor microenvironment (TME)

ROI

regions of interest (ROI)

NCTL

non-cytotoxic T lymphocytes (CD3+CD8−), (double-negative or CD4+ T lymphocyte)

CTL

cytotoxic CD8+ T lymphocytes (CD3+CD8+)

TTL

total T lymphocytes (CD3+CD8− and CD3+CD8+)

WhT

whole tumor

TC

tumor center (175μm into the tumor)

M

tumor margin (within 100μm outside the tumor)

MSS

melanoma-specific survival (MSS)

mIHC/IF

immunohistochemistry/immunofluorescence (mIHC/IF)

Footnotes

Disclosure: The authors declare no conflicts of interest.

Literature Citations

  • 1.Gajewski TF, Louahed J, Brichard VG. Gene signature in melanoma associated with clinical activity: a potential clue to unlock cancer immunotherapy. Cancer J. 2010;16(4):399–403. doi: 10.1097/PPO.0b013e3181eacbd8. [DOI] [PubMed] [Google Scholar]
  • 2.Chen DS, Mellman I. Oncology meets immunology: the cancer-immunity cycle. Immunity. 2013;39(1):1–10. doi: 10.1016/j.immuni.2013.07.012. [DOI] [PubMed] [Google Scholar]
  • 3.Clark WH Jr., Elder DE, Guerry D IV, Braitman LE, Trock, Schultz D, et al. Model Predicting Survival in Stage I Melanoma Based on Tumor Progression. JNCI: Journal of the National Cancer Institute. 1989;81(24):1893–904. doi: 10.1093/jnci/81.24.1893. [DOI] [PubMed] [Google Scholar]
  • 4.Busam KJ, Antonescu CR, Marghoob AA, Nehal KS, Sachs DL, Shia J, et al. Histologic classification of tumor-infiltrating lymphocytes in primary cutaneous malignant melanoma. A study of interobserver agreement. Am J Clin Pathol. 2001;115(6):856–60. doi: 10.1309/G6EK-Y6EH-0LGY-6D6P. [DOI] [PubMed] [Google Scholar]
  • 5.Thomas NE, Busam KJ, From L, Kricker A, Armstrong BK, Anton-Culver H, et al. Tumor-Infiltrating Lymphocyte Grade in Primary Melanomas Is Independently Associated With Melanoma-Specific Survival in the Population-Based Genes, Environment and Melanoma Study. Journal of Clinical Oncology. 2013;31(33):4252–9. doi: 10.1200/jco.2013.51.3002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Tuthill RJ, Unger JM, Liu PY, Flaherty LE, Sondak VK. Risk Assessment in Localized Primary Cutaneous Melanoma: A Southwest Oncology Group Study Evaluating Nine Factors and a Test of the Clark Logistic Regression Prediction Model. American Journal of Clinical Pathology. 2002;118(4):504–11. doi: 10.1309/wbf7-n8kh-71kt-rvq9. [DOI] [PubMed] [Google Scholar]
  • 7.Azimi F, Scolyer RA, Rumcheva P, Moncrieff M, Murali R, McCarthy SW, et al. Tumor-Infiltrating Lymphocyte Grade Is an Independent Predictor of Sentinel Lymph Node Status and Survival in Patients With Cutaneous Melanoma. Journal of Clinical Oncology. 2012;30(21):2678–83. doi: 10.1200/jco.2011.37.8539. [DOI] [PubMed] [Google Scholar]
  • 8.Gershenwald JE, Scolyer RA, Hess KR, Sondak VK, Long GV, Ross MI, et al. Melanoma staging: Evidence-based changes in the American Joint Committee on Cancer eighth edition cancer staging manual. CA Cancer J Clin. 2017;67(6):472–92. Epub 2017/10/14. doi: 10.3322/caac.21409. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Fu Q, Chen N, Ge C, Li R, Li Z, Zeng B, et al. Prognostic value of tumor-infiltrating lymphocytes in melanoma: a systematic review and meta-analysis. Oncoimmunology. 2019;8(7):1593806. Epub 2019/05/31. doi: 10.1080/2162402x.2019.1593806. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Maibach F, Sadozai H, Seyed Jafari SM, Hunger RE, Schenk M. Tumor-Infiltrating Lymphocytes and Their Prognostic Value in Cutaneous Melanoma. Front Immunol. 2020;11:2105. Epub 20200910. doi: 10.3389/fimmu.2020.02105. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Morrison SL, Han G, Elenwa F, Vetto JT, Fowler G, Leong SP, et al. Is the presence of tumor-infiltrating lymphocytes predictive of outcomes in patients with melanoma? Cancer. 2022;128(7):1418–28. Epub 2022/02/02. doi: 10.1002/cncr.34013. [DOI] [PubMed] [Google Scholar]
  • 12.Yang J, Lian JW, Chin YH, Wang L, Lian A, Murphy GF, et al. Assessing the Prognostic Significance of Tumor-Infiltrating Lymphocytes in Patients With Melanoma Using Pathologic Features Identified by Natural Language Processing. JAMA Netw Open. 2021;4(9):e2126337. Epub 20210901. doi: 10.1001/jamanetworkopen.2021.26337. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Galon J, Costes A, Sanchez-Cabo F, Kirilovsky A, Mlecnik B, Lagorce-Pages C, et al. Type, density, and location of immune cells within human colorectal tumors predict clinical outcome. Science. 2006;313(5795):1960–4. Epub 2006/09/30. doi: 10.1126/science.1129139. [DOI] [PubMed] [Google Scholar]
  • 14.Galon J, Mlecnik B, Bindea G, Angell HK, Berger A, Lagorce C, et al. Towards the introduction of the 'Immunoscore' in the classification of malignant tumours. J Pathol. 2014;232(2):199–209. Epub 2013/10/15. doi: 10.1002/path.4287. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Pages F, Mlecnik B, Marliot F, Bindea G, Ou FS, Bifulco C, et al. International validation of the consensus Immunoscore for the classification of colon cancer: a prognostic and accuracy study. Lancet. 2018;391(10135):2128–39. Epub 2018/05/15. doi: 10.1016/s0140-6736(18)30789-x. [DOI] [PubMed] [Google Scholar]
  • 16.Angell HK, Bruni D, Barrett JC, Herbst R, Galon J. The Immunoscore: Colon Cancer and Beyond. Clinical Cancer Research. 2020;26(2):332–9. doi: 10.1158/1078-0432.Ccr-18-1851. [DOI] [PubMed] [Google Scholar]
  • 17.van Houdt IS, Sluijter BJ, Moesbergen LM, Vos WM, de Gruijl TD, Molenkamp BG, et al. Favorable outcome in clinically stage II melanoma patients is associated with the presence of activated tumor infiltrating T-lymphocytes and preserved MHC class I antigen expression. Int J Cancer. 2008;123(3):609–15. Epub 2008/05/24. doi: 10.1002/ijc.23543. [DOI] [PubMed] [Google Scholar]
  • 18.Weiss SA, Han SW, Lui K, Tchack J, Shapiro R, Berman R, et al. Immunologic heterogeneity of tumor-infiltrating lymphocyte composition in primary melanoma. Hum Pathol. 2016;57:116–25. Epub 20160726. doi: 10.1016/j.humpath.2016.07.008. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Quigley DA, Kristensen V. Predicting prognosis and therapeutic response from interactions between lymphocytes and tumor cells. Molecular Oncology. 2015;9(10):2054–62. doi: 10.1016/j.molonc.2015.10.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Erdag G, Schaefer JT, Smolkin ME, Deacon DH, Shea SM, Dengel LT, et al. Immunotype and immunohistologic characteristics of tumor-infiltrating immune cells are associated with clinical outcome in metastatic melanoma. Cancer Res. 2012;72(5):1070–80. Epub 2012/01/24. doi: 10.1158/0008-5472.Can-11-3218. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Pagès F, Kirilovsky A, Mlecnik B, Asslaber M, Tosolini M, Bindea G, et al. In Situ Cytotoxic and Memory T Cells Predict Outcome in Patients With Early-Stage Colorectal Cancer. Journal of Clinical Oncology. 2009;27(35):5944–51. doi: 10.1200/jco.2008.19.6147. [DOI] [PubMed] [Google Scholar]
  • 22.Ali HR, Provenzano E, Dawson SJ, Blows FM, Liu B, Shah M, et al. Association between CD8+ T-cell infiltration and breast cancer survival in 12,439 patients. Ann Oncol. 2014;25(8):1536–43. Epub 2014/06/12. doi: 10.1093/annonc/mdu191. [DOI] [PubMed] [Google Scholar]
  • 23.Kumar BV, Connors TJ, Farber DL. Human T Cell Development, Localization, and Function throughout Life. Immunity. 2018;48(2):202–13. Epub 2018/02/22. doi: 10.1016/j.immuni.2018.01.007. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Kruse B, Buzzai AC, Shridhar N, Braun AD, Gellert S, Knauth K, et al. CD4(+) T cell-induced inflammatory cell death controls immune-evasive tumours. Nature. 2023;618(7967):1033–40. Epub 20230614. doi: 10.1038/s41586-023-06199-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Wu K, Fong L. CD4(+) T cells help myeloid-mediated killing of immune-evasive tumors. Trends Cancer. 2023;9(10):777–9. Epub 2023/08/07. doi: 10.1016/j.trecan.2023.07.013. [DOI] [PubMed] [Google Scholar]
  • 26.Whipple CA, Brinckerhoff CE. BRAF(V600E) melanoma cells secrete factors that activate stromal fibroblasts and enhance tumourigenicity. Br J Cancer. 2014;111(8):1625–33. Epub 2014/08/15. doi: 10.1038/bjc.2014.452. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Jenkins L, Jungwirth U, Avgustinova A, Iravani M, Mills A, Haider S, et al. Cancer-Associated Fibroblasts Suppress CD8+ T-cell Infiltration and Confer Resistance to Immune-Checkpoint Blockade. Cancer Res. 2022;82(16):2904–17. Epub 2022/06/25. doi: 10.1158/0008-5472.Can-21-4141. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Ribas A Adaptive Immune Resistance: How Cancer Protects from Immune Attack. Cancer Discov. 2015;5(9):915–9. Epub 2015/08/15. doi: 10.1158/2159-8290.Cd-15-0563. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Pérez-Romero K, Rodríguez RM, Amedei A, Barceló-Coblijn G, Lopez DH. Immune Landscape in Tumor Microenvironment: Implications for Biomarker Development and Immunotherapy. Int J Mol Sci. 2020;21(15). Epub 2020/08/06. doi: 10.3390/ijms21155521. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Wu Z, Zheng Y, Sheng J, Han Y, Yang Y, Pan H, et al. CD3(+)CD4(−)CD8(−) (Double-Negative) T Cells in Inflammation, Immune Disorders and Cancer. Front Immunol. 2022;13:816005. Epub 20220210. doi: 10.3389/fimmu.2022.816005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Conway K, Edmiston SN, Parker JS, Kuan PF, Tsai YH, Groben PA, et al. Identification of a Robust Methylation Classifier for Cutaneous Melanoma Diagnosis. J Invest Dermatol. 2019;139(6):1349–61. Epub 20181206. doi: 10.1016/j.jid.2018.11.024. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Conway K, Tsai YS, Edmiston SN, Parker JS, Parrish EA, Hao H, et al. Characterization of the CpG Island Hypermethylated Phenotype Subclass in Primary Melanomas. J Invest Dermatol. 2022;142(7):1869–81.e10. Epub 2021/11/30. doi: 10.1016/j.jid.2021.11.017. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Fischer AH, Jacobson KA, Rose J, Zeller R. Hematoxylin and eosin staining of tissue and cell sections. CSH Protoc. 2008;2008:pdb.prot4986. Epub 2008/01/01. doi: 10.1101/pdb.prot4986. [DOI] [PubMed] [Google Scholar]
  • 34.Gershenwald JE, Scolyer RA, Hess KR, Sondak VK, Long GV, Ross MI, et al. Melanoma staging: Evidence-based changes in the American Joint Committee on Cancer eighth edition cancer staging manual. CA Cancer J Clin. 2017;67(6):472–92. Epub 20171013. doi: 10.3322/caac.21409. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Warford A, Akbar H, Riberio D. Antigen retrieval, blocking, detection and visualisation systems in immunohistochemistry: a review and practical evaluation of tyramide and rolling circle amplification systems. Methods. 2014;70(1):28–33. Epub 2014/03/19. doi: 10.1016/j.ymeth.2014.03.001. [DOI] [PubMed] [Google Scholar]
  • 36.Baker GM, Murphy T, Block T, Nguyen D, Lynch FJ. Development and validation of an immunohistochemistry assay to assess glucocorticoid receptor expression for clinical trials of mifepristone in breast cancer. Cancer Manag Res. 2015;7:361–8. Epub 2015/12/18. doi: 10.2147/cmar.S91546. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Lee ET, Go OT. Survival analysis in public health research. Annu Rev Public Health. 1997;18:105–34. Epub 1997/01/01. doi: 10.1146/annurev.publhealth.18.1.105. [DOI] [PubMed] [Google Scholar]
  • 38.Matsumoto H, Thike AA, Li H, Yeong J, Koo SL, Dent RA, et al. Increased CD4 and CD8-positive T cell infiltrate signifies good prognosis in a subset of triple-negative breast cancer. Breast Cancer Res Treat. 2016;156(2):237–47. Epub 2016/03/11. doi: 10.1007/s10549-016-3743-x. [DOI] [PubMed] [Google Scholar]
  • 39.Wang Y, Zhu C, Song W, Li J, Zhao G, Cao H. PD-L1 Expression and CD8(+) T Cell Infiltration Predict a Favorable Prognosis in Advanced Gastric Cancer. J Immunol Res. 2018;2018:4180517. Epub 2018/07/14. doi: 10.1155/2018/4180517. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Chen L, Chen L, Ni H, Shen L, Wei J, Xia Y, et al. Prediction of CD3 T cells and CD8 T cells expression levels in non-small cell lung cancer based on radiomic features of CT images. Front Oncol. 2023;13:1104316. Epub 2023/03/03. doi: 10.3389/fonc.2023.1104316. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.De Logu F, Galli F, Nassini R, Ugolini F, Simi S, Cossa M, et al. Digital Immunophenotyping Predicts Disease Free and Overall Survival in Early Stage Melanoma Patients. Cells. 2021;10(2). Epub 2021/03/07. doi: 10.3390/cells10020422. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Vargas GM, Shafique N, Xu X, Karakousis G. Tumor-infiltrating lymphocytes as a prognostic and predictive factor for Melanoma. Expert Rev Mol Diagn. 2024;24(4):299–310. Epub 20240205. doi: 10.1080/14737159.2024.2312102. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Salgüero I, Roustán G, Requena L, Suárez D, García-Fresnadillo D, Redondo JI, et al. Immunophenotypic Differences in Tumor-Infiltrating Lymphocytes and Neovascularization Between Primary Cutaneous Melanoma With and Without Metastasis: An Immunohistochemical Study of 80 Cases. Am J Dermatopathol. 2021;43(11):811–8. Epub 2021/02/04. doi: 10.1097/dad.0000000000001907. [DOI] [PubMed] [Google Scholar]
  • 44.Toki MI, Merritt CR, Wong PF, Smithy JW, Kluger HM, Syrigos KN, et al. High-Plex Predictive Marker Discovery for Melanoma Immunotherapy-Treated Patients Using Digital Spatial Profiling. Clin Cancer Res. 2019;25(18):5503–12. Epub 2019/06/14. doi: 10.1158/1078-0432.Ccr-19-0104. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45.De Logu F, Ugolini F, Iannone LF, Simi S, Maio V, de Giorgi V, et al. Spatial Proximity and Relative Distribution of Tumor-Infiltrating Lymphocytes and Macrophages Predict Survival in Melanoma. Lab Invest. 2023;103(12):100259. Epub 2023/10/16. doi: 10.1016/j.labinv.2023.100259. [DOI] [PubMed] [Google Scholar]
  • 46.Attrill GH, Lee H, Tasker AT, Adegoke NA, Ferguson AL, da Silva IP, et al. Detailed spatial immunophenotyping of primary melanomas reveals immune cell subpopulations associated with patient outcome. Frontiers in Immunology. 2022;13. doi: 10.3389/fimmu.2022.979993. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47.Espinosa-Carrasco G, Chiu E, Scrivo A, Zumbo P, Dave A, Betel D, et al. Intratumoral immune triads are required for immunotherapy-mediated elimination of solid tumors. Cancer Cell. 2024;42(7):1202–16.e8. Epub 20240620. doi: 10.1016/j.ccell.2024.05.025. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 48.Magen A, Hamon P, Fiaschi N, Soong BY, Park MD, Mattiuz R, et al. Intratumoral dendritic cell-CD4(+) T helper cell niches enable CD8(+) T cell differentiation following PD-1 blockade in hepatocellular carcinoma. Nat Med. 2023;29(6):1389–99. Epub 2023/06/16. doi: 10.1038/s41591-02302345-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 49.Velikkakam T, Gollob KJ, Dutra WO. Double-negative T cells: Setting the stage for disease control or progression. Immunology. 2022;165(4):371–85. Epub 20220105. doi: 10.1111/imm.13441. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 50.Sheng W, Zhang C, Mohiuddin TM, Al-Rawe M, Zeppernick F, Falcone FH, et al. Multiplex Immunofluorescence: A Powerful Tool in Cancer Immunotherapy. Int J Mol Sci. 2023;24(4). Epub 2023/02/26. doi: 10.3390/ijms24043086. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 51.Chen LL, Jaimes N, Barker CA, Busam KJ, Marghoob AA. Desmoplastic melanoma: a review. J Am Acad Dermatol. 2013;68(5):825–33. Epub 20121223. doi: 10.1016/j.jaad.2012.10.041. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplement_Tables
NIHMS2194497-supplement-Supplement_Tables.pdf (191.5KB, pdf)
Supplement_Figure_S1
NIHMS2194497-supplement-Supplement_Figure_S1.pdf (2.3MB, pdf)
Supplement_Figure_S1_Legend
NIHMS2194497-supplement-Supplement_Figure_S1_Legend.pdf (78.7KB, pdf)

RESOURCES