Abstract
Background
Microsatellites are an adverse prognostic feature in melanoma that upstages node-negative disease to N1c in the AJCC 8th edition staging system. While Breslow thickness and microsatellites are established prognostic markers, their relationship in primary cutaneous melanoma has not been well examined.
Materials and Methods
We conducted a single-institution retrospective review of an IRB-approved prospective database of patients with primary cutaneous melanoma (January 2016-January 2026), excluding those with metastatic disease at diagnosis. Microsatellites are microscopic foci of melanoma cells discontinuous from the primary tumor and located in the dermis, subcutis, separated from the main tumor by normal tissue with no fibrosis or inflammation. Breslow thickness was measured from the granular layer (or the base of ulceration) to the deepest invasive melanoma cell. Patients were grouped by Breslow-based T-stage, and microsatellite frequency was calculated for each group. Fisher’s exact test evaluated the association between T-stage and microsatellite presence.
Results
Among 1,904 patients, 87 (4.6%) had microsatellites, with one patient having two separate microsatellite-positive tumors. Microsatellite frequency increased progressively with T-stage: 0.1% of T1 tumors (1/924), 2.4% of T2 tumors (14/579), 9.1% of T3 tumors (28/309), and 17.8% of T4 tumors (45/253) (Fisher’s exact p<0.0001). Within the microsatellite-positive cohort, T-stage distribution was 1.1% for T1, 15.9% for T2, 31.8% for T3, and 51.1% for T4.
Discussion
We found that microsatellite frequency increases with Breslow thickness. Since microsatellites upstage tumors to N1c with implications for immunotherapy decisions, surgeons and dermatopathologists must be diligent in performing and assessing oncology wide excision specimens.
Introduction
The American Joint Committee on Cancer (AJCC) staging system relies heavily on Breslow thickness and ulceration for prognostic risk stratification of primary cutaneous melanoma. 1 However, this does not capture the full spectrum of the biology surrounding a primary melanoma. Microsatellites, defined as microscopic foci of melanoma cells separate from the primary tumor, represent an adverse prognostic feature.2,3 Their presence automatically upstages a patient to N1c, or stage III disease.1,4
Back in the early eighties, 2 studies2,5 demonstrated that microsatellites are uncommon in thin melanoma primaries but occur more often in thicker tumors, suggesting that microsatellite formation is closely linked to increasing Breslow thickness. Subsequent studies have shown that microsatellites are strongly associated with other high-risk histopathologic features, including ulceration, higher Clark level, and elevated mitotic rate.3,5 Together, these histopathologic findings indicate that microsatellites reflect aggressive tumor biology and stratify risk within stage III melanoma patients. 3 However, to our knowledge, no study has systematically examined the relationship between microsatellite frequency and T-stage using the AJCC staging criteria.
Accordingly, our objective is to systematically examine the relationship between Breslow thickness-based T-stage using the 8th edition AJCC staging criteria and microsatellite frequency in order to define microsatellite distribution across tumor thickness categories and inform surgical risk stratification.
Materials and Methods
We conducted a single-institution retrospective cohort study using a prospectively maintained, IRB-approved melanoma registry managed using REDCap electronic data capture tools,6,7 at the University of North Carolina (IRB # 25-0846, approved 04/13/2025). Patients diagnosed with primary invasive cutaneous melanoma between January 1, 2016 and January 1, 2026 were eligible for inclusion. Patients with metastatic disease at initial presentation, mucosal or uveal melanoma, or recurrent melanoma at previously treated sites were excluded. Patients with multiple synchronous or metachronous primary melanomas were included, with each primary tumor analyzed as a separate lesion. Cases with incomplete pathology data, specifically missing Breslow thickness or microsatellite status, were excluded.
The primary outcome of interest was the association between AJCC T stage and microsatellite presence. Demographic and clinical data, including age at diagnosis, sex, race, primary tumor location, biopsy type, and melanoma histologic type, were extracted from the registry and cross-checked against electronic medical records (Epic Systems Corporation, Verona, WI) for accuracy. 8 Tumor-infiltrating lymphocytes (TILs), ulceration, regression, and lymph vascular invasion (LVI) status were also recorded from original pathology reports using standard dermatopathological criteria and AJCC 8th edition definitions, where applicable.
For the current analysis, we focused on Breslow thickness and the presence of microsatellites. Microsatellites were defined according to American Joint Committee on Cancer (AJCC) 8th edition criteria: microscopic nests of melanoma cells discontinuous from the primary tumor, located in the dermis or subcutis, separated from the main tumor by uninvolved dermis, without intervening fibrosis or inflammation. 4 All assessments were made by board-certified dermatopathologists as part of UNC’s routine clinical pathology reporting. Breslow thickness was defined as the distance from the top of the granular layer of the epidermis (or the base of ulceration, if present) to the deepest invasive melanoma cell, measured perpendicularly to the skin surface. Thickness measurements were extracted directly from original pathology reports. Tumors were grouped by Breslow thickness into T1 (≤1.0 mm), T2 (>1.0-2.0 mm), T3 (>2.0-4.0 mm), and T4 (>4.0 mm) according to AJCC 8th edition staging criteria.
All analyses were performed using R version 4.3.2 (R Foundation for Statistical Computing, Vienna, Austria). 9 Categorical variables were summarized as counts and percentages; continuous variables were summarized as mean ± standard deviation (SD) or median with interquartile range (IQR), depending on distribution. Baseline demographic characteristics were compared between microsatellite-positive and microsatellite-negative cohorts using chi-square or Fisher’s exact tests for categorical variables and the Wilcoxon rank-sum test for continuous variables. Age at diagnosis, anatomic location, ulceration, LVI, regression, and TILs were analyzed at the tumor level, as some patients had more than 1 melanoma. Exact binomial (Clopper-Pearson) methods were used to calculate 95% confidence intervals for proportions. The association between T stage and microsatellite presence was evaluated using Fisher’s exact test. Pairwise comparisons between adjacent T stages (T1 vs T2, T2 vs T3, T3 vs T4) were also performed using Fisher’s exact test. Multivariable logistic regression was performed to evaluate the independent associations of ulceration, LVI, and regression with microsatellite presence. A two-sided P value of <0.05 was considered statistically significant.
Results
Patient and Tumor Characteristics Stratified by Microsatellite Status
Of the 2065 tumors, 924 (44.7%) were T1, 579 (28.0%) T2, 309 (15.0%) T3, and 253 (12.3%) T4. Microsatellite frequency increased progressively with advancing T stage (Figure 1). Microsatellites were identified in 0.1% of T1 tumors (1/924; 95% CI, 0.0-0.6%), 2.4% in T2 (14/579; 95% CI, 1.3-4.0%), 9.1% in T3 (28/309; 95% CI, 6.1-12.8%), and 17.8% in T4 (45/253; 95% CI, 13.3-23.1%) (Figure 1). The association between T stage and microsatellite presence was statistically significant (Fisher’s exact test, P < 0.0001). Microsatellite frequency by T stage and Breslow thickness
Pairwise comparisons between adjacent T-stages demonstrated significant increases in microsatellite frequency at each step: T1 vs T2 (P = 1.4 × 10-5), T2 vs T3 (P = 1.9 × 10-5), and T3 vs T4 (P = 0.0025). Among the 88 microsatellite-positive lesions, 1 (1.1%) was T1, 14 (15.9%) T2, 28 (31.8%) T3, and 45 (51.1%) T4, indicating that 83% of all microsatellite-positive cases occurred in T3 or T4 tumors (Figure 2). Distribution of microsatellite-positive tumors across T-stage distribution (n = 88)
Among microsatellite-positive tumors, primary tumor locations included head/neck (n = 27, 30.7%), trunk (n = 22, 25.0%), lower extremity (n = 21, 23.9%), upper extremity (n = 17, 19.3%), and cutaneous vulvar melanoma (n = 1, 1.1%) (Figure 3). When compared with microsatellite-negative tumors, there was no statistically significant difference in primary tumor location (P = NS). Ulceration was present in 37 microsatellite-positive tumors (42.0%), LVI in 21 (23.9%), and regression in 15 (17.0%). Compared to microsatellite-negative tumors, microsatellite-positive tumors demonstrated significantly higher rates of ulceration (42.0% vs 18.3%, P < 0.001), LVI (23.9% vs 1.7%, P < 0.001), and regression (17.0% vs 8.9%, P = 0.014). TILs were brisk in 25 (28.4%), non-brisk in 38 (43.2%), and absent in 25 (28.4%), with no significant difference observed in TILs distribution (P = NS) (Table 1). Distribution of microsatellite-positive tumors across anatomic site (n = 88)
Multivariable Analysis of Factors Associated with Microsatellite Presence
Discussion
Our study provides the first systematic characterization of microsatellite frequency across the current AJCC T-stage categories, demonstrating a progressive increase from 0.1% in T1 melanomas to 2.4% in T2, 9.1% in T3, and 17.8% in T4 lesions (P < 0.0001). This stepwise increase in microsatellite frequency with higher T-stage suggests that microsatellite development may be closely linked to increasing tumor depth rather than being a random chance occurrence. This provides quantitative evidence that microsatellite development parallels tumor thickness as defined by the current 8th edition AJCC staging system.
Our findings both confirm and substantially extend prior observations regarding the relationship between Breslow thickness and microsatellites. An earlier report, 5 which represented secondary observations within a surgical margin study, demonstrated that microsatellites were rare in melanomas <3 mm thick but present in approximately 37% of thicker lesions. However, this analysis relied on broad, Breslow-based thickness groupings that did not use the finite cut points as currently defined by the AJCC 8th edition. 4 Our current results refine this understanding by demonstrating a graded, stepwise increase in microsatellite frequency across all current AJCC T-stage categories, consistent with earlier work establishing microsatellites as an adverse pathologic feature with prognostic impact in addition to simply Breslow thickness and ulceration.3,5 Our multivariable analysis identified LVI (OR 14.6), ulceration (OR 2.6), and regression (OR 1.9) as independently associated with microsatellite presence. The high OR associated with LVI is not surprising since both LVI and microsatellites represent micrometastatic tumor, the former in lymphatic or vascular lumens and the latter in stroma. 10
Large contemporary outcome studies11,12,13 have looked at the association of microsatellites and sentinel node positivity rates and have demonstrated a wide range (43-71%) of patients with metastatic disease in the sentinel nodes. Collectively, this suggests that microsatellites, a lymphatic based metastasis, do not guarantee that the patient also has nodal disease. Furthermore, these studies and the current one, all support that microsatellite development reflects biologically advanced disease associated with increasing tumor thickness. This has important clinical implications for both surgeons and pathologists. In our study, 83% of all microsatellite-positive tumors occurred in T3 or T4 primaries, strongly suggesting that heightened vigilance during pathologic evaluation of diagnostic biopsy specimens and wide local excision specimens is particularly warranted in these melanomas. Identification of microsatellites automatically upstages patients to N1c (stage III) disease.11,13 Given these clinical implications, accurate identification of microsatellites is critical. However, prior studies14,15 have demonstrated variation in histopathological interpretation of melanoma prognostic and staging features, which may contribute to inconsistent routine reporting of microsatellites. Together with our findings, these observations support the need for systematic, standardized pathologic examination protocols, particularly for T3/T4 melanomas, to ensure accurate staging and optimal clinical management.
Several limitations of this study should be acknowledged. First, this is a single-institution analysis, which could limit the generalizability of our findings to other clinical settings. Second, the retrospective design relies on routine clinical pathology reporting by a limited number of institutional dermatopathologists. While this will improve internal consistency in microsatellite identification, it does not eliminate the potential for reporting variability. Finally, we did not analyze microsatellite presence in relation to BRAF mutation status, sentinel lymph node biopsy findings, or results of staging studies, which may provide additional insight into the biological and clinical significance of microsatellites.
In conclusion, this study demonstrates a strong, progressive relationship between increasing Breslow thickness and microsatellite frequency across the current AJCC T-stage categories. From a surgical perspective, these findings highlight the critical importance of meticulous pathologic evaluation of diagnostic biopsy specimens and wide local excision specimens, particularly in patients with T3 and T4 melanomas, where the likelihood of microsatellite detection is highest. Because microsatellites upstage patients to stage III (N1c) and may influence therapeutic decisions, accurate detection by the dermatopathologist is essential. Future work should focus on the standardization of histopathologic examination of diagnostic biopsies and wide excision specimens, especially in T3 and T4 melanoma specimens. We acknowledge that this may be more time consuming for surgeons and dermatopathologists, but our patients deserve the same diligent, potentially stage-altering examination of their specimens that is afforded to sentinel nodes.16,17
Footnotes
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.