Downloaded from http://journals.lww.com/jpho-online by BhDMf5ePHKbH4TTImqenVMLLdHdrcqOQVP8EIbQkc53cVSOKdCHxUTTfmNz2uNiAy40QM2F/4ac= on 10/10/2020 Establishing Cost-effective Strategies for Predicting Outcomes of Pediatric Leukemia Akshat Jain, MD, MPH E ditorial to—Flow cytometry-based absolute blast count on day 8: reliable, fast, and inexpensive method. JPHO Manuscript number: JPHO-20-122. Each year, over 2900 children (0 to 19 y of age) are diagnosed with acute lymphoblastic leukemia (ALL), making ALL the most common type of childhood cancer. 1 As survival rates improve and more children with leukemia are entering remission with standardized protocols and utilization of novel treatment regimens, there has been a concerted effort to study cost- effective strategies for better outcomes of pediatric leukemia. A recent study predicted that utilization of adherence promotion strategies in leukemia care could lead to superior health outcomes based on quality-adjusted life-years and an average of 3000 USD cost-saving per patient with a pediatric leukemia diagnosis. 1 Similar cost-benefit analyses have been performed in accurate diagnostic and prognostic testing. Most recently the Canadian model predicted 1-year cost expenditure for minimal residual disease (MRD) testing by flow cytometry in newly diagnosed patients with precursor B-cell ALL was estimated at $340,760. 2 MRD while hailed as a benchmark for prognostication in outcomes, yet still remains too expensive for developing countries, that bear a big chunk of the global morbidity and mortality from pediatric cancer. A study in 2012, looking at absolute lymphocyte count (ALC) demonstrated that high ALC on day 29 of induction day, to be an independent, clinically significant predictor of improved relapse-free survival and overall survival. 3 Newer strategies such as utilization of day 8 blast count (“Flow cytometry-based absolute blast count on day 8: reliable, fast, and inex- pensive method. JPHO Manuscript number: JPHO-20-122”) provide critical metrics needed by novel artificial intelligence predictive models, that may not only prove cost-effective and efficient but also accurate in diagnostic screening and predicting relapse rates. Machine learning models utilizing existing hematologic metrics and indices that are commonly performed in all patients with a hematologic malignancy are already under study. 4,5 A recent in-depth analysis of using cell population data generated from next-generation hematologic analyzers proved its usefulness in the screening of hematologic and non- hematologic diseases such as sepsis. 6 What limits their accuracy and validity is, finding vari- ables that are relevant, affordable, and can be obtained globally to improve the sample size of these models to be able to give a representative date. Day 8 blast count, ALC, MRD in combination could prove to be useful in refining the accuracy of predicting outcomes for pediatric leukemia. Thus investing in clinical and bench research to identify cost-effective hematologic matrices, beyond MRD, could prove vital in using technologic advances to better outcomes in pediatric cancer. REFERENCES 1. McGrady ME, Eckman MH, O’Brien MM, et al. Cost-effectiveness analysis of an adherence-promotion intervention for children with leukemia: a Markov model-based simulation. J Pediatr Psychol. 2018;43: 758–768. 2. Health Quality Ontario. Minimal residual disease evaluation in childhood acute lymphoblastic leukemia: an economic analysis. Ont Health Technol Assess Ser. 2016;16:1–83. 3. Rabin KR, Gramatges MM, Borowitz MJ, et al. Absolute lymphocyte counts refine minimal residual disease-based risk stratification in childhood acute lymphoblastic leukemia. Pediatr Blood Cancer. 2012; 59:468–474. 4. Pan L, Liu G, Lin F, et al. Machine learning applications for prediction of relapse in childhood acute lymphoblastic leukemia. Sci Rep. 2017;7:7402. 5. Zini G. Artificial intelligence in hematology. Hematology. 2005;10:393–400. 6. Syed-Abdul S, Firdani RP, Chung HJ, et al. Artificial intelligence based models for screening of hematologic malignancies using cell population data. Sci Rep. 2020;10:4583. Received for publication April 17, 2020; accepted July 6, 2020. From the Loma Linda University School of Medicine, Loma Linda, CA. The authors declare no conflict of interest. Reprints: Akshat Jain, MD, MPH, Loma Linda University School of Medicine, 1175 Campus Street, Loma Linda, CA 92350 (e-mail: akshatjainusa@gmail.com). Copyright © 2020 Wolters Kluwer Health, Inc. All rights reserved. INVITED COMMENTARY J Pediatr Hematol Oncol  Volume 42, Number 7, October 2020 www.jpho-online.com | 451 Copyright r 2020 Wolters Kluwer Health, Inc. All rights reserved.