Abstract BACKGROUND: ETV and tumor biopsy are central to the surgical management of children with primary CNS GCT. An ETV creates a communication between the intraventricular compartment and the subarachnoid spaces and decompartmentalizes the ventricular system. “Tumor spill” or shedding may result from surgical interventions, such as biopsy. ETV with simultaneous biopsy may impart a greater tendency for dissemination and possible distant relapse. This is a concern in CNS GCT given the attempts of irradiation field and dose reduction following chemotherapy. METHODS: We performed a retrospective review of the prospective database for the Children’s Oncology Group (COG) ACNS1123 study. Possible associations were explored among ETV, endoscopic biopsy, and combined ETV+biopsy with relapse, distant relapse, progression free survival (PFS), and time to distant relapse. RESULTS: Among 244 eligible patients, 97 ETV+/-biopsies were performed, and 30 relapses occurred. There were no associations among ETV and/or biopsy with relapse (Cochran-Mantel-Haenszel [CMH] test, with histology (germinoma vs. nongerminomatous germ cell tumor (NGGCT)) as stratification variable: ETV: p=0.3167, biopsy: p=0.3375, combined: p=0.3066), distant relapse (CMH test, ETV: p=0.4631, ETV+biopsy: p=0.6795), PFS (log-rank test, ETV: NGGCT p=0.1632, germinoma p=0.9288; biopsy: NGGCT p=0.1682, germinoma p=0.9701; ETV+Biopsy: NGGCT p=0.1306, germinoma p=0.7758), or time to distant relapse with death/local relapse as competing risk (Gray’s test, ETV: NGGCT p=0.5694, germinoma p=0.2327; biopsy: NGGCT p=0.3505, germinoma p=0.5747; ETV+Biopsy: NGGCT p=0.3988, germinoma p=0.6839). CONCLUSIONS: Based on a secondary analysis of prospective data from the ACNS1123 trial, ETV and biopsy did not impart a greater likelihood of relapse in children with primary CNS GCT treated with combined chemotherapy and irradiation. However, three tract recurrences did occur (all germinoma), suggesting that they may affect pattern of relapse. Current and future prospective trials should continue to explore associations among these variables and relapse, including patterns of relapse.
Read full abstract