AUTHOR=Gao Tianqi , Chen Fengxi , Li Man 

TITLE=Development of Two Diagnostic Prediction Models for Leptomeningeal Metastasis in Patients With Solid Tumors

JOURNAL=Frontiers in Neurology

VOLUME=Volume 13 - 2022

YEAR=2022

URL=https://www.frontiersin.org/journals/neurology/articles/10.3389/fneur.2022.899153

DOI=10.3389/fneur.2022.899153

ISSN=1664-2295

ABSTRACT=Objectives: For accurate diagnosis of leptomeningeal metastasis (LM) and avoid unnecessary examinations or lumber puncture (LP), we develop two diagnostic prediction models for patients with solid tumors.
Study design, setting and participants: It is a retrospective cohort study launched in the Second Affiliated Hospital of Dalian Medical University. 206 patients who had been admitted between January 2005 and December 2021 with a solid tumor and clinical suspicion of LM were enrolled to develop model A. 152 patients of them who underwent LPs for cytology and biochemistry were enrolled to develop model B.
Model development: Diagnostic factors included skull metastasis, active brain metastasis, progressed extracranial disease, number of involved extracranial organs, number of symptoms, protein in cerebrospinal fluid (CSF) and glucose in CSF. Outcome predictor was defined as clinical diagnosis of LM. Logistic least absolute shrinkage and selection operator regression was used to identify the relevant variables and fit the prediction model. A calibration curve and the concordance index c-index were used to evaluate calibrated and discriminatory ability. The n-fold cross validation method was used to internally validate the models. The decision curve analysis and the interventions avoided analysis were used to evaluate the clinical application.
Results: The area under the curve (AUC) values of model A and model B were 0.812 (95% CI: 0.751-0.874) and 0.901 (95% CI: 0.852-0.949). Respectively compared to the first MRI and first lumber puncture, model A and model B showed higher AUC (model A vs first MRI: 0.812 vs 0.743, p= 0.087; model B vs first lumber puncture: 0.901 vs 0.800, p= 0.010). The validated c-index were 0.810 (95% CI: 0.670-0.952) and 0.899 (95% CI: 0.823-0.977). The calibration curves show a good calibrated ability. The clinical application evaluation revealed a net clinical benefit and reduction in unnecessary intervention by using the models.
Conclusions: The models can help to improve diagnostic accuracy when were used alone or combined with conventional work-up. They also exhibit net clinical benefit in medical decisions and avoiding unnecessary interventions for LM patients. Studies focused on the external validation of our models are necessary in the future.