ALL-RIC trial protocol: a comparison of reduced dose total body irradiation (TBI) and cyclophosphamide with fludarabine and melphalan reduced intensity conditioning in adults with acute lymphoblastic leukaemia (ALL) in complete remission

INTRODUCTION

The usage of a T-cell depleted, reduced intensity conditioning (RIC) approach to haematopoietic cell transplantation (HCT) in adult patients with acute lymphoblastic leukaemia (ALL) over 40 years of age and in first complete remission (CR) has resulted in encouraging rates of event-free and overall survival in a population of adults with high risk disease. However, relapse rates remain high-with disease progression being the major cause of treatment failure. Using different, more powerful conditioning approaches is the logical next step in examining the role of RIC allogeneic HCT in adult ALL.

METHODS AND ANALYSIS

The ALL-RIC trial is a two-arm, phase II, multicentre, randomised clinical trial in adult patients with ALL in first or second CR, who are undergoing allogeneic HCT. Comparison of a novel RIC transplant conditioning regimen using reduced-dose total body irradiation (TBI), cyclophosphamide and alemtuzumab, is made against a standardised RIC approach using fludarabine, melphalan and alemtuzumab. The primary outcome of the study is disease-free survival at 3 years, defined as time from randomisation to the first of either relapse or death from any cause. Patients who are still alive and progression-free at the end of the trial will be censored at their last date known to be alive. Secondary outcomes include overall survival and non-relapse mortality.

ETHICS AND DISSEMINATION

The protocol was approved by the East Midlands-Leicester Central Research Ethics committee (18/EM/0112). Initial approval was received on 12 June 2018. Current protocol version (V.6.0) approval obtained on 18 November 2019. The Medicines and Healthcare products Regulatory Agency (MHRA) also approved all protocol versions. The results of this trial will be disseminated through national and international presentations and peer-reviewed publications.

TRIAL REGISTRATION NUMBER

EudraCT Number: 2017-004800-23.ISRCTN99927695.

Does Anatomical Contextual Information Improve 3D U-Net-Based Brain Tumor Segmentation?

Effective, robust, and automatic tools for brain tumor segmentation are needed for the extraction of information useful in treatment planning. Recently, convolutional neural networks have shown remarkable performance in the identification of tumor regions in magnetic resonance (MR) images. Context-aware artificial intelligence is an emerging concept for the development of deep learning applications for computer-aided medical image analysis. A large portion of the current research is devoted to the development of new network architectures to improve segmentation accuracy by using context-aware mechanisms. In this work, it is investigated whether or not the addition of contextual information from the brain anatomy in the form of white matter (WM), gray matter (GM), and cerebrospinal fluid (CSF) masks and probability maps improves U-Net-based brain tumor segmentation. The BraTS2020 dataset was used to train and test two standard 3D U-Net (nnU-Net) models that, in addition to the conventional MR image modalities, used the anatomical contextual information as extra channels in the form of binary masks (CIM) or probability maps (CIP). For comparison, a baseline model (BLM) that only used the conventional MR image modalities was also trained. The impact of adding contextual information was investigated in terms of overall segmentation accuracy, model training time, domain generalization, and compensation for fewer MR modalities available for each subject. Median (mean) Dice scores of 90.2 (81.9), 90.2 (81.9), and 90.0 (82.1) were obtained on the official BraTS2020 validation dataset (125 subjects) for BLM, CIM, and CIP, respectively. Results show that there is no statistically significant difference when comparing Dice scores between the baseline model and the contextual information models (p > 0.05), even when comparing performances for high and low grade tumors independently. In a few low grade cases where improvement was seen, the number of false positives was reduced. Moreover, no improvements were found when considering model training time or domain generalization. Only in the case of compensation for fewer MR modalities available for each subject did the addition of anatomical contextual information significantly improve (p < 0.05) the segmentation of the whole tumor. In conclusion, there is no overall significant improvement in segmentation performance when using anatomical contextual information in the form of either binary WM, GM, and CSF masks or probability maps as extra channels.