# Defining the high-risk category of patients with cutaneous melanoma: a practical tool based on prognostic modeling

## TL;DR

This study creates a practical tool to identify high-risk melanoma patients using a prognostic model based on histological features.

## Contribution

A novel 5-factor logistic regression model is developed to predict melanoma recurrence using routine histological data.

## Key findings

- Five independent variables were identified as most significant for predicting recurrence.
- The model achieved 86.1% sensitivity and 72.7% specificity in predicting recurrence.
- The model uses routine histological features to identify high-risk patients.

## Abstract

Although most cutaneous melanoma (CM) in its early stages is treatable, the risk of recurrence remains high and there is a particular ambiguity on patients prognosis. This drives to identification of prognostic biomarkers for predicting CM recurrence to guide appropriate treatment in patients with localized melanoma.

This study aimed to develop a prognostic model for assessing the risk of recurrence in patients with CM, enabling prompt prognosis-driven further clinical decision-making for high-risk patients.

This case-control study included 172 patients with CM recurrence (high-risk group) and 30 patients with stable remission (low-risk group) 3 years after primary diagnosis. The impact of sex, age at diagnosis, anatomical site, histological characteristics (the histological type, pathological stage, ulceration; the depth of invasion, mitotic rate, lymphovascular invasion, neurotropism, association with a nevus, tumor-infiltrating lymphocyte density, tumor regression and BRAF codon 600 mutation status) on CM recurrence was evaluated.

Five independent variables, including nodal status, a high mitotic rate, Breslow thickness, lymphovascular invasion, perineural invasion and regression features were identified as the most significant. A 5-factor logistic regression model was developed to assess the risk of melanoma recurrence. The sensitivity and specificity of the model were 86.1% and 72.7%, respectively.

The developed model, which relies on routine histological features, allows the identification of individuals at high risk of CM recurrence to tailor their further management.