Is alpha-fetoprotein decline a prognostic factor of childhood non-seminomatous germ cell tumours? Results of the French TGM95 study. (May 2018)

• Record Type: Journal Article
• Authors: Fresneau, B.; Orbach, D.; Faure-Conter, C.; Sudour-Bonnange, H.; Vérité, C.; Gandemer, V.; Pasquet, M.; Fasola, S.; Rome, A.; Raimbault, S.; Martelli, H.; Frappaz, D.; Le Teuff, G.; Patte, C.
• Is Part Of: European journal of cancer. Volume 95(2018)
• Page Start: 11
• Page End: 19
• Publication Date: 2018-05
• Subjects: Germ cell tumours -- Children -- Alpha-fetoprotein decline -- Prognostic analysis; Cancer -- Periodicals; Neoplasms -- Periodicals; Cancer -- Périodiques; Cancer; Tumors; Electronic journals; Periodicals; Electronic journals; 616.994
• DOI: 10.1016/j.ejca.2018.02.029
• Languages: English
• ISSNs: 0959-8049

Abstract

Abstract: Purpose: In adults' non-seminomatous germ cell tumours (NS-GCT), alpha-fetoprotein (AFP) decline was identified as an important prognostic factor. We investigated its prognostic value in the French TGM95 study for childhood NS-GCT. Patients and methods: Three risk groups were defined: low risk (LR: localised and completely resected pS1, AFP<15000 ng/ml), with a 'wait-and-see' strategy; intermediate-risk (IR: localised incompletely resected, AFP<15000 ng/ml) with 3–5 vinblastine-bleomycine-cisplatin courses; high risk (HiR: AFP≥15000 ng/ml and/or metastatic) with 4–6 etoposide-ifosfamide-cisplatin courses. The multivariable prognostic analysis for progression-free survival (PFS) included age (±10 years), primary tumour site (1-testis, 2-ovary, 3-extragonadal), extent of disease (1-pS1, 2-loco-regional dissemination, 3-metastasis) and AFP (±10, 000 ng/ml). AFP decline prognostic value was investigated in IR + HiR groups using predicted time to normalisation (TTN), AFP change, and difference between observed and expected (based on AFP half-life) area under the curve (O-E AUC). Results: From January 1995 to December 2005, 239 patients (median age = 3years, 60 LR, 65 IR, 114 HiR) were included. Main sites were testis (n = 66), ovary (n = 77) and sacrococcygeal (n = 57). Five-year PFS and OS were 85% (95% confidence interval [CI] = 80–89%) and 93% (89–95%), respectively. Age ≥ 10 years (hazard ratio [HR] = 4.6, 95% CI = 2.1–10.1, p = 0.0001) and extragonadal primary (HR = 6.3, 95% CI = 2.0–19.9, p = 0.005) were significant prognostic factors. In AFP decline analysis (n = 151, 17 events), TTN (p = 0.61) and AFP change (p = 0.10) were not prognostic, whereas we showed a significant effect of O-E AUC (HR = 2.1, 95% CI = 1.0–4.2, p = 0.05). Conclusion: Age ≥ 10 years and extragonadal tumours remain as poor prognostic factors. Contrary to adults, TTN is not reliable in paediatric NS-GCT. The prognostic value of O-E AUC should be investigated in larger studies. Highlights: In TGM95 study, age ≥10 years and extragonadal tumours remain poor prognostic factors. Contrary to adults, predicted time to AFP normalisation (TTN) is not prognostic in paediatric NS-GCT. AFP decline was prognostic when modelling by the difference between observed and expected (based on AFP half-life) area under the curve (AUC).