PATH-07. QUALITY ASSURANCE IN CEREBROSPINAL FLUID CYTOLOGY ASSESSMENT FOR MEDULLOBLASTOMA STAGING LEADS TO POTENTIAL IMPROVED RISK-GROUP ASSESSMENT IN THE PROSPECTIVE MULTICENTER HIT-2000 TRIAL. (4th December 2020)

• Record Type: Journal Article
• Authors: Hagel, Christian; Sloman, Veronika; Mynarek, Martin; Petrasch, Katharina; Obrecht, Denise; Deinlein, Frank; Schmid, Renate; von Bueren, André O; Friedrich, Carsten; Juhnke, B Ole; Gerber, Nicolas U; Kwiecien, Robert; Girschick, Hermann; Höller, Alexandra; Zapf, Antonia; von Hoff, Katja; Rutkowski, Stefan
• Is Part Of: Neuro-oncology. Volume 22(2020)Supplement 3
• Page Start: iii425
• Page End: iii426
• Publication Date: 2020-12-04
• Subjects: Brain Neoplasms -- Periodicals; Brain -- Tumors -- Periodicals; Brain -- Cancer -- Periodicals; Nervous system -- Cancer -- Periodicals; 616.99481
• DOI: 10.1093/neuonc/noaa222.643
• Languages: English
• ISSNs: 1522-8517

Abstract

Abstract: BACKGROUND: Cerebrospinal fluid (CSF) dissemination of medulloblastoma (M1 stage) is a high-risk prognostic factor. However, because diagnostic criteria for M1 staging are missing we specified process-related and cytomorphological parameters influencing the predictive value of the CSF status. PATIENTS AND METHODS: CSF samples and cytology reports from 405 medulloblastoma patients of the prospective multicenter trial HIT-2000 were reviewed and related to 5-year progression free survival (5y-PFS). RESULTS: Tumor cells were detected in 237/1073 CSF cytospins. M1-patients and M2/3 patients with radiologically detected metastases showed a worse 5y-PFS than M0 patients (54% and 52% vs. 76%; p=0.01 and p<0.001). Lumbar sampling was more sensitive than ventricular sampling. M0 diagnosed specimens containing >50% lytic cells and/or less than 10 nucleated cells showed a decreased 5y-PFS (61%). Further investigation of cytological parameters revealed a poor outcome for cases harboring > 3 tumor cell clusters and individual tumor cells (5y-PFS 33%) vs. cases with ≥ 2 individual tumor cells but no clusters (5y-PFS 61%). In bi-variable Cox-regression, ≥ 2 vs. 0 or 1 tumor cells were associated with a Hazard Ratio (HR) of 0.52 (95%-Confidence Interval (CI): 0.12, 2.30; p=0.39), whereas > 3 vs. no tumor cell clusters were associated with a HR of 8.94 (95%-CI: 1.66, 48.22; p=0.01). CONCLUSIONS: CSF staging in medulloblastoma should comprise lumbar specimens with <50% lytic cells and a minimum of 10 nucleated cells. The predictive value of CSF cytology in M1 cases may predominantly depend on tumor cell clusters. The latter finding needs to be confirmed in prospective trials.