Childhood medulloblastoma—a single institution’s historical perspective on survival and functional morbidity

Purpose To compare results from a third (1995–2010) cohort of children with medulloblastoma with two previous series (J Neurosurg 86:13–21, 1997; Arch Dis Child 54:200–203, 1979) to analyse the effects of management changes aimed at improving both overall and event-free survivals (OS and EFS) and functional outcomes. Methods Review of neuro-oncology and imaging databases and previously published results. Results There was no statistically significant improvement in the 5-year OS for 104 children diagnosed 1995–2010, 61.5% (95% CI, 52.9, 71.6), compared with 50% of the 80 children presenting 1980–1990 (J Neurosurg 86:13–21, 1997) (difference 11.5%; 95% CI, 2.8, 25.4). Five-year OS for 96 children suitable for risk-stratification was overall 66% (95% CI, 57.9, 75.8); standard risk 77.8% (95% CI, 67.4, 89.7); high risk < 3 years 50.0% (95% CI, 32.3, 77.5); high risk ≥ 3 years 54.5% (95% CI, 37.2, 79.9); 5-year EFS were standard risk 68.5% (95% CI, 57.2, 82.1); high risk < 3 years 40.0% (95% CI, 23.4, 68.4); and high risk ≥ 3 years 36.4% (95% CI, 20.9, 63.2); overall 55.2% (95% CI, 46.1, 66.1). Of 62/63 ≥ 5-year survivor, 9 died later from tumour relapse and 4 from second malignancy. Functional outcomes of 62 of the 63 ≥ 5-year survivors: 67.7% had educational issues requiring remedial input; 18% restricted mobility indoors and outdoors; 59.7% hearing impairment (42% prescribed aids). Conclusions Comparison of this single-institution series with its predecessor found that revised chemotherapy and RT protocols and greater accuracy of risk stratification did not result in statistically significant improvements in either survival or treatment-related functional disability. Extended (> 5-year) follow-up is essential if 20% of late deaths from relapse and second malignancies are not to be overlooked.