Background and objective: Leptomeningeal metastasis (LM) can occur as a late manifestation of breast cancer and has traditionally been diagnosed by CSF cytology; however, cytology suffers from low sensitivity and it is believed that many cases of LM go undiagnosed. Some studies have suggested the use of CA 15-3 in CSF (CA 15-3 CSF) to aid in the detection of LM. The purpose of this study was to compare the performance of CA 15-3 CSF to cytology for the detection and treatment monitoring of breast cancer LM.
Methods: CA 15-3 CSF requests between 2014 and 2016 were retrospectively reviewed. Fifty-two measurements from nine patients were from our health system and had corresponding CSF cytology measurements. Concordance between CA 15-3 CSF and CSF cytology was calculated. For patients with quantifiable CA 15-3 CSF, sequential determinations of CA 15-3 and cytology were compared over time to assess correlation of CA 15-3 CSF concentration and cytology with disease status.
Results: At the time of initial testing, seven of the nine patients (78%) had positive cytology. Two samples (22%) had quantifiable CA 15-3, both of which were also positive by cytology. The positive concordance between all cytology and CA 15-3 measurements was 9% (2/22), while negative concordance was 100% (30/30). Sequential CA 15-3 and cytology measurements showed a decrease in CA 15-3 that paralleled changes observed with cytology.
Conclusions: In this cohort of patients, CA 15-3 CSF performance was neither superior nor complementary to cytology for the detection of LM, nor did the combination of CA 15-3 CSF and cytology improve performance over cytology alone.
Keywords: Breast cancer; CA 15–3; CSF; Leptomeningeal metastatic breast cancer; Tumor marker.
Copyright © 2018 The Canadian Society of Clinical Chemists. Published by Elsevier Inc. All rights reserved.