paediatrics Brussels 17

296

Kennedy et al.

International Journal of Radiation Oncology Biology Physics

Executive Function

Health Status

100

HFRT:

=0.430

=0.900

P

P

1.0

80

=0.098

P =0.005

P

STRT:

0.8

60

0.6

40

0.4

5% 15%

12% 25%

20

BRIEF GEC score

0.2

0

0.0

n=19 n=20

parent report n=49 n=51

n=50 n=54

HUI3 global health status utility score parent report n=46 n=48 (7-17 yrs)

self report

self report

(7-17 yrs)

(18-29 yrs)

(11-29 yrs)

Behavioural Difficulties

Quality of life (<18 years)

Quality of life ( 18years)

=0.060

P

100

20

100

=0.570

=0.753

P

P

=0.250

=0.420

P

P

80

80

16

60

12

60

40

8

40

53% 48%

41% 48%

20

4

20

PedsQL Total score

15% 12%

14% 24%

SDQ Total Difficulties score

QLQ C-30 Global health status score 0 n=22

0

0

n=22

n=40 n=44

n=40 n=43

parent report n=51 n=52

parent report n=50 n=50

self report

self report

self-report (18-29 yrs)

(7-17 yrs)

(7-17 yrs)

(11-29 yrs)

(11-17 yrs)

Fig. 1. Scores on age-appropriate measures of quality of survival by treatment allocation. Error bars indicate SDs. Executive Function: Higher scores indicate worse function. Percentages within bars refer to abnormally elevated scores. BRIEF GEC Z Behavior Rating Inventory of Executive Function Global Executive Composite. Health Status: Higher scores indicate better health. HUI Z Health Utilities Index. Behavioral Difficulties: Higher scores indicate worse function. Percentages within bars refer to borderline or abnormal scores. SDQ Z Strengths and Difficulties Questionnaire. Quality of life: Higher scores indicate better quality. Percentages within bars refer to “at risk” scores. PedsQL Z Quality of Life Inventory; QLQ-C30 Z core 30-question version of the European Organization for Research and Treatment of Cancer Quality of Life Questionnaire. HFRT Z hyperfractionated radiation therapy; STRT Z standard radiation therapy.

Because treatment allocation was random, differences in outcomes are inherently unlikely to be attributable to known or unknown differences in premorbid characteristics. Exclusion from the analysis of participants reported to have premorbid developmental impairment did not alter the findings. The simi- larity of nonparticipant and participant baseline characteristics in both treatment arms makes attrition bias unlikely, and the retention rate for QoS follow-up information at 4-9 years from diagnosis of 62% is high relative to a rate of < 30% (31) in other pediatric neuro-oncology studies and a mean rate of 68% (range, 41-100%) at 1 year follow-up in RCTs for pediatric chronic conditions (42) . A system is now in place to include in future trials the option of direct entry of patients’ responses to elec- tronic versions of these questionnaires on personal computers or other devices with Internet access, but its effect on ascertainment in this context remains to be determined. Radiation therapy fields, as distinct from fractionation schedules, were the same in both treatment arms except for the (small) boost to the tumor bed in HFRT, and there is no reason to suppose that field alignment differed between the 2 arms. The BRIEF questionnaire, which relates to everyday function, has been widely used in patients with acquired brain injury with

Discussion) suggested a higher incidence of ototoxicity after HFRT. Self-reported social and employment outcomes, only applicable to participants whose age at assessment was 18 years (all of whom were also aged 8 years at diagnosis), seemed to be similar in the 2 treatment arms excepting an excess, in the group allocated to HFRT, of those driving a motor vehicle ( Supplementary Table e3 ).

Discussion

The PNET4 trial is the first clinical RCT of craniospinal HFRT versus STRT for medulloblastoma and the first pediatric brain tumor treatment trial to ascertain QoS information internationally across Europe. Compared with survivors who had received STRT, BRIEF scores for executive function in everyday life suggested a possible benefit to those who had received HFRT 6 years after enrollment in the PNET4 trial, but the fact that this group did not show associated benefits on measures of health status or quality of life is equally important. The HFRT group also suffered a greater decrement from height z -score at diagnosis despite GH treatment, and use of a hearing aid was more commonly reported after HFRT.