Severe Toxicity Free Survival Following Childhood Acute Lymphoblastic Leukemia

Background:

Childhood acute lymphoblastic leukemia (ALL) is the most common childhood cancer comprising around 25% of all childhood cancers. Stepwise modifications of antileukemic therapy have led to a rise in 5-year survival probability from less than 30% in the 1970s to above 90% today. Cure, however, comes at a price; survivors are burdened by acute as well as long-term toxicities, that is, adverse effects of treatment. Among survivors, the cumulative incidence of severe, disabling, life-threatening, or fatal chronic health conditions was recently reported to be 21.1% at 20 years from ALL diagnosis, which is substantial, not least when considering the peak incidence of childhood ALL around the age of 3 years, meaning that many 20-year survivors are young adults. Hence, focusing on long-term health sequelae is more relevant than ever. Traditional cancer outcome measures consist of overall survival and cancer-related-event-free survival, but for cancers with high survival probability, such as for childhood ALL, these traditional outcome measures become insufficient since they ignore the burden of therapy. Severe, long-term toxicities should be considered in treatment outcome evaluation, but the lack of internationally standardized capturing and reporting of late effects is a barrier. To address this, an international collaboration of ALL consortia initiated a project aiming to select physician-defined severe toxicities following ALL treatment. This work resulted in consensus definitions of 21 severe health conditions named Severe Toxicities, which could each be considered an unacceptable price for cure. With examples as heart failure, brain damage, and chronic lung disease, the 21 conditions are of such severity that ALL treatment possibly would have been modified if the toxicity had been predictable at time of diagnosis. These Severe Toxicities should be integrated in treatment outcome evaluation alongside the overall survival and the cancer-related event-free survival for a more comprehensive evaluation of treatment protocols.

The occurrence of these 21 Severe Toxicities has not been evaluated before. This international study aims to quantify the occurrence of the Severe Toxicities in 5 large cohorts of childhood ALL patients from Europe, USA, and Australia. Even though the overall survival as well as the cancer-related-event-free survival for different cohorts may be similar, differences in the use and dose-intensity of steroids, chemotherapy, radiotherapy, and hematopoietic stem cell transplantation may lead to different toxicity patterns. The original Severe Toxicity definitions were recently modified to meet statistical requirements for valid analyses and to ensure that the Severe Toxicities can be classified uniformly and prospectively across different cohorts.

Comparing the prevalence and patterns of Severe Toxicities across different treatment protocols will potentially reveal modifiable treatment-related factors associated with risk of individual and/or multiple Severe Toxicities. A global decision to routinely report Severe Toxicity is essential for a more comprehensive knowledge of late effects, which may guide future research towards improved treatment strategies aiming to reduce toxicities, and hereby improving quality of life, without compromising cure.

The 21 Severe Toxicities include the following conditions:

• Hearing loss
• Blindness
• Heart failure
• Arrhythmia
• Coronary artery disease
• Heart valve disease
• Gastrointestinal failure
• Hepatic failure
• Insulin dependent diabetes
• Renal failure
• Pulmonary failure
• Osteonecrosis
• Amputation and physical deformation
• Cognitive dysfunction
• Seizures
• Psychiatric disease
• Paralytic, neuropathic, myopathic, or movement disorders
• Vocal cord paralysis
• Cytopenia
• Immunodeficiency
• Second malignant neoplasm and benign CNS tumor The outcomes considered in the statistical analyses are time-to-event outcomes for each of the 21 Severe Toxicities and the two derived outcomes: the time to first event of death or any of the Severe Toxicities, and the cumulative number of different Severe Toxicities events by time.

Aims:

1. To investigate Severe-Toxicity-Free Survival in childhood ALL survivors
2. To investigate the occurrence of Severe Toxicities in childhood ALL survivors treated on contemporary treatment protocols, investigating the cumulative number of different Severe Toxicities as well as each Severe Toxicity separately Furthermore, the study will explore associations with potential risk factors, e.g., demographic variables such as sex and age at diagnosis, and treatment-related variables such as type of chemotherapy and use of hematopoietic stem cell transplantation with and without total body irradiation.

Data registration:

Data will be collected from medical charts. Medical chart review can be combined with data extraction from existing databases. The method of data capture will vary between centers depending on the degree of already existing toxicity registration and the current use of databases.

Data analysis:

All analyses will use time since diagnosis as the time scale. Severe-Toxicity-Free Survival will be presented graphically by a Kaplan-Meier curve, and exploratory risk factor analyses will be based on Cox regression analyses. Number of different Severe Toxicities by time will be presented graphically as mean cumulative count and exploratory risk factor analyses will be based on Poisson regression models assuming piecewise constant underlying rates and allowing for overdispersion. The cumulative probability of observing each of the Severe Toxicities will be presented graphically using the Aalen-Johansen estimator with death as the competing event, while exploratory risk factor analyses will be based on Cox regression model for the toxicity-specific hazard treating death as a censoring.

In the exploratory risk factor analyses, the linearity assumption for quantitative covariates will be evaluated using linear splines. There are no prespecified hypotheses, all analyses are exploratory, and statistical significance will not be claimed. Estimated associations will be presented with nominal 95% confidence limits, that is, confidence limits that are not adjusted for multiple testing.

Ethics:

All requested data concerns toxicity registration in the protocol period (within 5 years after ALL diagnosis) or potental risk factors such as gender and age at time of diagnosis. The study will be conducted in accordance with Law nr. 502 of 23/05/2018 "Lov om supplerende bestemmelser til forordning om beskyttelse af fysiske personer i forbindelse med behandling af personoplysninger og om fri udveksling af sådanne oplysninger" (The Data Protection Act), Regulation (EU) 2016/679 (GDPR) and other relevant regulation. Each study group is responsible for obtaining relevant local approvals and follow relevant regulation, hereunder GDPR.