Relapsed/refractory acute lymphoblastic leukemia (ALL) remains a leading cause of cancer-related mortality in children despite highly effective frontline therapies.1 Modern risk stratification based on ALL-associated cytomolecular genetic alterations and end-of-induction minimal/measurable residual disease (MRD) has informed therapeutic allocation in current clinical trials,2, 3 some of which are testing precision medicine approaches with kinase inhibitors4, 5 or CD19- or CD22-directed immunotherapies.6 However, definitions of complete remission (CR), primary treatment failure (TF), and relapse in ALL have not been uniform, making outcomes comparisons difficult across pediatric oncology consortia and their associated trials.
A recent collaborative effort to harmonize response definitions for pediatric ALL was undertaken by Dr. Swantje Buchmann and colleagues via the Ponte di Legno (PdL) Consortium with a goal of creating international consensus for use across future clinical trials. The investigators unsurprisingly first noted appreciable differences amongst 17 pediatric leukemia consortia in their disease detection methods, timing of CR assessment, definitions of CR and TF, and MRD threshold levels to determine CR and relapse. These definitions influence critical treatment decisions, such as the need for post-induction therapy intensification and when to declare TF or relapse and move quickly to alternative salvage approaches.
The first task was to define uniform methodology for ALL response assessment. Prior morphologic blast estimation by light microscopy has evolved in the past 20 years to MRD detection by flow cytometry (FC) immunophenotyping or polymerase chain reaction (PCR) assessment of aberrant TCR or IGH rearrangements. MRD positivity is the most important independent predictor of clinical outcomes in childhood B-ALL.7, 8 FC- and PCR-based MRD techniques are highly specific and sensitive, with a detection capability of 10-4 or 10-5.9 New next-generation sequencing (NGS) MRD approaches are further capable of detecting and tracking specific TCR or BCR rearrangement-based subclones at a sensitivity of at least 10-6, and the prognostic significance of NGS MRD versus FC or PCR-based approaches is under active study in current pediatric ALL trials.10 The PdL group concluded that MRD measurement via FC or PCR is the “gold standard” for assessment of CR achievement by prior clinical trial--determined thresholds and that newer NGS methodologies are also acceptable if quality standards are established. CR is to be defined as MRD less than 1 percent and/or M1 cytomorphology (< 5% marrow blasts by microscopy) without evidence of extramedullary disease and to be determined at end of induction or end of consolidation for cooperative group trials.
A second task was to define TF with uniform residual disease thresholds and timings of assessment. With this definition, the PdL consensus was that TF is a lack of CR attainment and is to be determined no sooner than end-of-consolidation to allow for appropriate risk-adapted consolidation therapy delivery with agents not given in induction.
A third task was to determine a uniform definition of ALL relapses, which can only occur after the achievement of prior CR. The investigators endorsed classifying relapse by anatomic sites (medullary and/or extramedullary) and quantification of marrow involvement by MRD techniques. Compared to the historic morphologic relapse definition of 25 percent or greater marrow involvement (M3), the PdL group consensus was that medullary relapse has occurred if MRD is 5 percent or higher, with confirmation by one additional sensitive genetic assay (cytogenetics, fluorescence in situ hybridization, PCR or real-time PCR detection of an ALL-specific marker) or if MRD is 1 percent or higher with confirmation by two additional ALL-specific diagnostic assays. The investigators also defined extramedullary central nervous system (CNS) relapse as a new occurrence of CNS3 status (≥5 white blood cells in cerebrospinal fluid and blasts detected in cytospin) or CNS2 status (< 5 white blood cells in CSF with positive cytospin) on two occasions at least one week apart after initial CNS1 status or CNS remission achievement. Other extramedullary sites of relapse (if isolated) must be confirmed by imaging and/or biopsy.
In Brief
The PdL Consortium comprehensively addressed prior discrepancies in definitions of CR, TF, and relapse in childhood ALL across major international pediatric oncology consortia with particular emphasis on the necessity of modern MRD technologies in response assessment. An important caveat is potential disparities in access to high-sensitivity MRD methodologies in resource-poor settings. Nonetheless, it is hoped that this major international harmonization will facilitate more uniform data comparison across clinical trials and, at a pragmatic level, allow children with confirmed lower-level ALL relapses to be treated in potentially better health with novel investigational agents on clinical trials with a goal of long-term cure.
Competing Interests
Dr. Ding and Dr. Tasian indicated no relevant conflicts of interest.
