The changing burden of long-term health outcomes in survivors of childhood acute lymphoblastic leukaemia: a retrospective analysis of the St Jude Lifetime Cohort Study

Abstract

Background: Treatment for childhood acute lymphoblastic leukemia has evolved over the past five decades, with moderation of traditional chemotherapy and radiotherapy and the introduction of targeted immune-based and cellular-based therapies. The affect of these changes on late health outcomes has not been assessed. Using data from the The St Jude Lifetime (SJLIFE) Cohort, we aimed to characterise the magnitude of morbidity and patterns of late health outcomes among survivors of childhood acute lymphoblastic leukaemia treated over time.

Methods: The St Jude Lifetime (SJLIFE) Cohort is a retrospective cohort study with prospective follow-up and ongoing data accrual designed to facilitate longitudinal, clinically-based assessment of health outcomes among survivors of paediatric malignancies. 980 survivors included in this analysis were diagnosed with paediatric acute lymphoblastic leukaemia at St Jude Children's Research Hospital (SJCRH) between Aug 28, 1963, and July 19, 2003, were aged 18 years old and older at enrolment, had a minimum follow-up of 10 years after diagnosis, and completed an initial on-campus SJLIFE assessment as of data cutoff (June 30, 2015). 272 community control participants, matched to survivors on 5-year age blocks in each sex, were recruited for comparison. Cumulative chemotherapy and radiation dose exposures and major medical events during and after therapy were retrieved from the medical records of the survivors. History or physical examination, laboratory analysis, physical fitness, and neurocognitive testing were done. Health conditions were graded according to a modified version of the Common Terminology Criteria for Adverse Events. Neurocognitive domains of attention (Trial Making Test Part A and Conner's Continuous Performance Test-II) and executive function (Trail Making Test Part B, Controlled Oral Word Association Test, and Wechsler Adult Intelligence Scale-III Digit Span Test Backward) were measured and age-adjusted Z scores were calculated. Mean cumulative count was used to calculate the age-standardised cumulative burden of health conditions over time. This cohort study is registered at ClinicalTrials.gov, number NCT00760656.

Findings: 980 survivors of acute lymphoblastic leukaemia (50% women, median age at diagnosis 5 years [IQR 3·1-9·1 years], and median time from diagnosis of 30·0 years [22·7-36·3]) had a median age of 35·8 years (29·4-42·9) at assessment compared with 35·1 years (28·7-42·6) for 272 controls. Survivors had significantly more growth hormone deficiency, hypogonadism, and neuropathy than controls. By age 30 years, survivors of acute lymphoblastic leukaemia had, on average, 5·4 (95% CI 5·1-5·8) grade 1-4 health conditions, including 3·2 (2·9-3·4) grade 2-4 health conditions, compared with 2·0 (CI 1·7-2·2) grade 1-4 and 1·2 (1·03-1·4) grade 2-4 health conditions among controls. The cumulative burden of grade 2-4 health conditions involved multiple organ systems for survivors treated on protocols between 1962-91, but after elimination of cranial radiotherapy for children with acute lymphoblastic leukaemia, conditions now predominately include musculoskeletal and endocrine disorders for survivors on protocols between 1991-2007.

Interpretation: Although changes in paediatric acute lymphoblastic leukaemia treatment protocols have improved overall survival, the burden of late morbidity remains high for these patients. We show that the pattern of late toxic effects has markedly changed over time, with survivors having a reduction in health conditions that are immediately life-threatening, however, maintaining health status and quality of life for survivors of paediatric acute lymphoblastic leukaemia requires continued medical surveillance, counselling, and lifestyle modifications.

Funding: US National Cancer Institute and the American Lebanese Syrian Associated Charities.

Figure 1:

Study diagram of SJLIFE population and participation in ALL analysis

Figure 2:

Cumulative burden of health conditions (A) Grade 1–4 events. (B) Grade 2–4 events. Among 980 survivors and 272 controls, but two survivors and 20 controls were censored before attaining age 20 years.

Figure 3:

Cumulative burden of grade 2–4 health conditions in survivors by therapy protocol Among 980 survivors, 72 survivors who were treated according to a Total protocol, but who did not consent to enrolment, were excluded and one survivor in Total therapy XIII–XIV was censored before age 20 years. Thus, 907 survivors are shown here.

Figure 4.

Distribution of the cumulative burden of grade 2-4 health conditions among ALL survivors by Total Therapy protocol^*┼ * Age standardized to 30 years old ^┼ Excludes 72 participants who were treated according to but did not consent to enrollment on a Total protocol.

Figure 5.

Frequency of impaired aerobic dysfunction and mobility among ALL survivors by Total Therapy protocol* * Adjusted for age at evaluation, gender, BMI, and neuropathy ^┼ Excludes 72 participants who were treated according to but did not consent to enrollment on a Total protocol.

Figure 6:

Age-adjusted z-scores for attention and executive function by radiation dose and exposure in ALL survivors compared to population norms and controls