paediatrics Brussels 17

293

Volume 88 Number 2 2014

Quality of survival in the PNET4 study

Methods and Materials: Participants in the PNET4 trial and their parents/caregivers in 7 participating anonymized countries completed standardized questionnaires in their own language on executive func- tion, health status, behavior, health-related quality of life, and medical, educational, employment, and social information. Pre- and postoperative neurologic status and serial heights and weights were also recorded. Results: Data were provided by 151 of 244 eligible survivors (62%) at a median age at assessment of 15.2 years and median interval from diagnosis of 5.8 years. Compared with standard radiation therapy, hyperfractionated radiation therapy was associated with lower (ie, better) z -scores for executive func- tion in all participants (mean intergroup difference 0.48 SDs, 95% confidence interval 0.16-0.81, P Z .004), but health status, behavioral difficulties, and health-related quality of life z -scores were similar in the 2 treatment arms. Data on hearing impairment were equivocal. Hyperfractionated radi- ation therapywas also associatedwith greater decrement in height z -scores (mean intergroup difference 0.43 SDs, 95% confidence interval 0.10-0.76, P Z .011). Conclusions: Hyperfractionated radiation therapy was associated with better executive function and worse growth but without accompanying change in health status, behavior, or quality of life. 2014 Elsevier Inc.

comparing hyperfractionated radiation therapy with stan- dard radiation therapy, com- bined with a chemotherapy regimen common to both treatment arms, in children and young adults with me- dulloblastoma. We observed a possible benefit of hyper- fractionated radiation ther- apy to executive function but greater restriction of spinal growth and no significant benefit to health status, behavior, or quality of life.

In-depth assessment of CNS outcomes in multicenter clinical trials has typically been associated with ascertainment rates below 30% (31) and attrition bias (1) . Booklets of standardized ques- tionnaires for completion by participants and parents, however, achieved ascertainment rates of 73% in United Kingdom survivors enrolled in the PNET3 study of medulloblastoma at 7 years from diagnosis (32) . PNET4 provided an opportunity to establish whether cross-cultural adaptations of the same questionnaires could be applied across Europe. We report here on a cross- sectional follow-up study of the CNS and endocrine outcomes in surviving PNET4 participants.

Introduction

Survivors of childhood central nervous system (CNS) tumors have shown impairments of quality of survival (QoS) in multiple do- mains of function, persisting or worsening many years after completion of treatment (1-5) . Medulloblastoma survivors, particularly if treated before age 8 years, have impaired health- related quality of life (HRQoL) d a multidimensional concept that includes physical, social, cognitive, and emotional func- tioning (6-8) , with underlying difficulties in neurocognition, attention, executive function (9-13) , growth, fertility, thyroid function (14-18) , educational attainment, employment, and for- mation of long-term relationships. Rates of stroke, second tumors, and premature aging are increased (14, 19-25) . The HIT-SIOP PNET4 phase 3 European randomized controlled treatment trial (RCT) for medulloblastoma compared hyper- fractionated radiation therapy (HFRT) comprising 2 smaller frac- tional doses of 1 Gy per day, separated by an interval of at least 8 hours, with conventionally fractionated standard treatment (STRT) of 1 1.8-Gy fraction per day. Smaller fractional radiation therapy doses produce a redistribution of proliferating tumor cells, with some cells entering a radiosensitive stage and thus providing a better tumor cell kill. Nonproliferating tissues, such as normal brain, should be spared this effect of redistribution and thereby potentially be protected by HFRT according to a linear-quadratic formula relating fraction size to biological effect (26) . This sparing effect of HFRT may account for the higher full-scale intelligence quotient after treatment with twice-daily HFRT without chemotherapy than that observed in historical controls treated with once-daily STRT previously reported by Carrie et al (27) and Gupta et al (28) in medulloblastoma survivors (29) . These studies were not, however, experimental in design and relied on historical controls, rather than random allocation of treatment, to provide comparison groups. The HFRT regimen used in PNET4 was predicted to be bio- logically more effective in its effects on tumor tissue than STRT, but this was not supported by a 5-year event-free survival rate (0.79 0.02) in both treatment arms (30) . It was also predicted that the HFRT in PNET4 would produce long-term CNS effects that were either similar or possibly, by reduction of the posterior fossa HFRT boost outside the tumor bed, less deleterious than those observed after STRT.

Methods and Materials

Patients

Between 2001 and 2006, 338 participants aged 4-21 years in 10 countries had been randomly allocated to receive either HFRT or STRT treatment trial for M0 medulloblastoma in PNET4. Standard radiation therapy comprised 23.4Gy to the craniospinal axis and 54Gy to the posterior fossa and was given over 42 days in 30 daily fractions of 1.8Gy for 5 days perweek. Hyperfractionated radiation therapywas given in 68 fractions: 1.0 Gy twice per day with an 8-hour interval between fractions, given over 48 days. The total craniospinal dose was 36 Gy, and the whole posterior fossa dose was 60 Gy, with a further boost to 68 Gy to the tumor bed. In both treatment arms a maximum of 8 doses of vincristine 1.5 mg/m 2 (maximum 2 mg) was given once per week during radiation therapy and adjuvant chemotherapy. Eight cycles of cisplatin 70 mg/m 2 intravenously, lomustine 75 mg/m 2 on day 1, and vincristine 1.5 mg/m 2 intravenously on days 1, 8, and 15, began 6weeks after the end of RT, with a 6-week interval between each cycle (30) . The 244 survivors from France, Germany, Italy, The Netherlands, Spain, Sweden, and the United Kingdom who remained in remission during the 2010-2011 9-month cross-sectional follow-up study period in the QoS study were eligible for the present QoS study.

Procedure

Written consent for QoS data collection had been obtained as part of the PNET4 treatment trial by the treating clinician, who provided