Skip to main content

Advertisement

SpringerLink
Log in

Risk prediction for delayed clearance of high-dose methotrexate in pediatric hematological malignancies by machine learning

  • Original Article
  • Published:
International Journal of Hematology Aims and scope Submit manuscript

Abstract

This study aimed to establish a predictive model to identify children with hematologic malignancy at high risk for delayed clearance of high-dose methotrexate (HD–MTX) based on machine learning. A total of 205 patients were recruited. Five variables (hematocrit, risk classification, dose, SLC19A1 rs2838958, sex) and three variables (SLC19A1 rs2838958, sex, dose) were statistically significant in univariable analysis and, separately, multivariate logistic regression. The data was randomly split into a “training cohort” and a “validation cohort”. A nomogram for prediction of delayed HD–MTX clearance was constructed using the three variables in the training dataset and validated in the validation dataset. Five machine learning algorithms (cart classification and regression trees, naïve Bayes, support vector machine, random forest, C5.0 decision tree) combined with different resampling methods were used for model building with five or three variables. When developed machine learning models were evaluated in the validation dataset, the C5.0 decision tree combined with the synthetic minority oversampling technique (SMOTE) using five variables had the highest area under the receiver operating characteristic curve (AUC 0.807 [95% CI 0.724–0.889]), a better performance than the nomogram (AUC 0.69 [95% CI 0.594–0.787]). The results support potential clinical application of machine learning for patient risk classification.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1
Fig. 2
Fig. 3

Similar content being viewed by others

Data availability

The datasets generated and analyzed during the present study are available from the corresponding author on reasonable request.

Abbreviations

MTX:

Methotrexate

HD–MTX:

High-dose methotrexate

H:

Hour

SNPs:

Single-nucleotide polymorphisms

ALL:

Acute lymphoblastic leukemia

B-ALL:

B cell acute lymphoblastic leukemia

T-ALL:

T cell acute lymphoblastic leukemia

ANC:

Absolute neutrophil count

PLT:

Platelet

RBC:

Red blood cell

HCT:

Hematocrit

ALT:

Alanine aminotransferase

TBIL:

Total bilirubin

DBIL:

Direct bilirubin

IBIL:

Indirect bilirubin

TP:

Total protein

ALB:

Albumin

GLB:

Globulin

Cr:

Creatinine

HWE:

Hardy–Weinberg equilibrium

RF:

Random forest

SVM:

Support vector machine

CART:

Classification and regression trees

NB:

Naïve Bayes classification

SMOTE:

Synthetic minority oversampling technique

BLSMOTE:

Borderline-SMOTE

ADASYN:

Adaptive synthetic

OSS:

One-sided selection

ROC:

Receiver operating characteristic curve

AUC:

Area under the ROC curve

ACC:

Accuracy

Spec:

Specificity

Sens:

Sensitivity

PPV:

Positive predictive value

NPV:

Negative predictive value

n:

Number

References

  1. Stanulla M, Schrappe M. Treatment of childhood acute lymphoblastic leukemia. Semin Hematol. 2009;46:52–63.

    Article  PubMed  Google Scholar 

  2. Pui CH, Carroll WL, Meshinchi S, Arceci RJ. Biology, risk stratification, and therapy of pediatric acute leukemias: an update. J Clin Oncol. 2011;29:551–65.

    Article  PubMed  Google Scholar 

  3. Levêque D, Santucci R, Gourieux B, Herbrecht R. Pharmacokinetic drug-drug interactions with methotrexate in oncology. Expert Rev Clin Pharmacol. 2011;4:743–50.

    Article  PubMed  CAS  Google Scholar 

  4. Yang SL, Zhao FY, Song H, Shen DY, Xu XJ. Methotrexate associated renal impairment is related to delayed elimination of high-dose methotrexate. Sci World J. 2015;2015:751703.

    Google Scholar 

  5. Ramsey LB, Balis FM, O’Brien MM, Schmiegelow K, Pauley JL, Bleyer A, et al. Consensus guideline for use of glucarpidase in patients with high-dose methotrexate induced acute kidney injury and delayed methotrexate clearance. Oncologist. 2018;23:52–61.

    Article  CAS  PubMed  Google Scholar 

  6. Howard SC, McCormick J, Pui CH, Buddington RK, Harvey RD. Preventing and managing toxicities of high-dose methotrexate. Oncologist. 2016;21:1471–82.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  7. Hu YH, Zhou L, Wang SS, Jing X, Guo HL, Sun F, et al. Methotrexate disposition in pediatric patients with acute lymphoblastic leukemia: what have we learnt from the genetic variants of drug transporters. Curr Pharm Des. 2019;25:627–34.

    Article  CAS  PubMed  Google Scholar 

  8. Cao M, Guo M, Wu DQ, Meng L. Pharmacogenomics of methotrexate: current status and future outlook. Curr Drug Metab. 2018;19:1182–7.

    Article  CAS  PubMed  Google Scholar 

  9. Gervasini G, Mota-Zamorano S. Clinical Implications of methotrexate pharmacogenetics in childhood acute lymphoblastic leukaemia. Curr Drug Metab. 2019;20:313–30.

    Article  CAS  PubMed  Google Scholar 

  10. Ranchon F, Vantard N, Henin E, Bachy E, Sarkozy C, Karlin L, et al. Delayed methotrexate elimination: incidence, interaction with antacid drugs, and clinical consequences? Hematol Oncol. 2018;36:399–406.

    Article  CAS  PubMed  Google Scholar 

  11. Schmidt D, Kristensen K, Schroeder H, Wehner PS, Rosthøj S, Heldrup J, et al. Plasma creatinine as predictor of delayed elimination of high-dose methotrexate in childhood acute lymphoblastic leukemia: A Danish population-based study. Pediatr Blood Cancer. 2019;66:e27637.

    Article  CAS  PubMed  Google Scholar 

  12. Kawase A, Yamamoto T, Egashira S, Iwaki M. Stereoselective inhibition of methotrexate excretion by glucuronides of nonsteroidal anti-inflammatory drugs via multidrug resistance proteins 2 and 4. J Pharmacol Exp Ther. 2016;356:366–74.

    Article  CAS  PubMed  Google Scholar 

  13. Suzuki K, Doki K, Homma M, Tamaki H, Hori S, Ohtani H, et al. Co-administration of proton pump inhibitors delays elimination of plasma methotrexate in high-dose methotrexate therapy. Br J Clin Pharmacol. 2009;67:44–9.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  14. Inose R, Takahashi K, Nanno S, Hino M, Nagayama K. Calcium channel blockers possibly delay the elimination of plasma methotrexate in patients receiving high-dose methotrexate therapy. J Chemother. 2019;31:30–4.

    Article  CAS  PubMed  Google Scholar 

  15. Ramsey LB, Mizuno T, Vinks AA, O’Brien MM. Delayed methotrexate clearance in patients with acute lymphoblastic leukemia concurrently receiving dasatinib. Pediatr Blood Cancer. 2019;66:e27618.

    Article  PubMed  CAS  Google Scholar 

  16. Liu G, Xu Y, Wang X, Zhuang X, Liang H, Xi Y, et al. Developing a machine learning system for identification of severe hand, foot, and mouth disease from electronic medical record data. Sci Rep. 2017;7:16341.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  17. Nunnelee JD. Review of an article: the international warfarin pharmacogenetics consortium (2009) Estimation of the warfarin dose with clinical and pharmacogenetic data. NEJM 360(8): 753–764. J Vasc Nurs. 2009;27:109.

    Article  PubMed  Google Scholar 

  18. Tang J, Liu R, Zhang YL, Liu MZ, Hu YF, Shao MJ, et al. Application of machine-learning models to predict tacrolimus stable dose in renal transplant recipients. Sci Rep. 2017;7:42192.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  19. Csordas K, Lautner-Csorba O, Semsei AF, Harnos A, Hegyi M, Erdelyi DJ, et al. Associations of novel genetic variations in the folate-related and ARID5B genes with the pharmacokinetics and toxicity of high-dose methotrexate in paediatric acute lymphoblastic leukaemia. Br J Haematol. 2014;166:410–20.

    Article  CAS  PubMed  Google Scholar 

  20. Que LP, Huang K, Fang JP, Xu HG, Zhou DH, Li Y, et al. Reassessment of the risk-stratified gd-2008 all protocol. J Pediatr Hematol Oncol. 2018;40:472–7.

    Article  PubMed  Google Scholar 

  21. Hospira (2011) Label for methotrexate injection. Lake Forest, il: Hospira; available at: https://www.Accessdata.Fda.Gov/drugsatfda_docs/label/2011/011719s117lbl.pdf. Accessed 17 Jan 2020.

  22. Laboratories B (2016) The R project for statistical computing, https://www.R-project.Org/.Accessed 30 Nov 2017. Accessed 30 Nov 2017.

  23. Bunkhumpornpat C, Sinapiromsaran K. CORE: core-based synthetic minority over-sampling and borderline majority under-sampling technique. Int J Data Min Bioinform. 2015;12:44–58.

    Article  PubMed  Google Scholar 

  24. Li Y, Li H, Yao H. Analysis and study of diabetes follow-up data using a data-mining-based approach in new urban area of urumqi, xinjiang, china, 2016–2017. Comput Math Methods Med. 2018;2018:7207151.

    PubMed  PubMed Central  Google Scholar 

  25. Sufriyana H, Wu YW, Su EC. Prediction of preeclampsia and intrauterine growth restriction: development of machine learning models on a prospective cohort. JMIR Med Inform. 2020;8:e15411.

    Article  PubMed  PubMed Central  Google Scholar 

  26. Siriseriwan W (2019) Smotefamily. A collection of oversampling techniques for class imbalance problem based on smote, https://cran.r-project.org/package=smotefamily. Accessed 21 June 2019.

  27. Pozzolo AD, Olivier C, Bontempi G (2015) Unbalanced: racing for unbalanced methods selection https://cran.r-project.org/package=unbalanced. Accessed 21 June 2019.

  28. Kataoka T, Sakurashita H, Kajikawa K, Saeki Y, Taogoshi T, Matsuo H. Low serum albumin level is a risk factor for delayed methotrexate elimination in high-dose methotrexate treatment. Ann Pharmacother. 2021. https://doi.org/10.1177/1060028021992767.

    Article  PubMed  Google Scholar 

  29. Nakano T, Kobayashi R, Matsushima S, Hori D, Yanagi M, Suzuki D, et al. Risk factors for delayed elimination of high-dose methotrexate in childhood acute lymphoblastic leukemia and lymphoma. Int J Hematol. 2021;113:744–50.

    Article  CAS  PubMed  Google Scholar 

  30. Skärby T, Jönsson P, Hjorth L, Behrentz M, Björk O, Forestier E, et al. High-dose methotrexate: on the relationship of methotrexate elimination time vs renal function and serum methotrexate levels in 1164 courses in 264 Swedish children with acute lymphoblastic leukaemia (ALL). Cancer Chemother Pharmacol. 2003;51:311–20.

    Article  PubMed  CAS  Google Scholar 

  31. Xu W, Tang Y, Song H, Shi S, Yang S. Retrospective study on elimination delay of methotrexate in high-dose therapy of childhood acute lymphoblastic leukemia in China. J Pediatr Hematol Oncol. 2007;29:688–93.

    Article  CAS  PubMed  Google Scholar 

  32. Xu C, Jackson SA. Machine learning and complex biological data. Genome Biol. 2019;20:76.

    Article  PubMed  PubMed Central  Google Scholar 

  33. Altman N, Krzywinski M. The curse(s) of dimensionality. Nat Methods. 2018;15:399–400.

    Article  CAS  PubMed  Google Scholar 

  34. Kang J, Schwartz R, Flickinger J, Beriwal S. Machine learning approaches for predicting radiation therapy outcomes: a clinician’s perspective. Int J Radiat Oncol Biol Phys. 2015;93:1127–35.

    Article  PubMed  Google Scholar 

  35. Lundberg SM, Nair B, Vavilala MS, Horibe M, Eisses MJ, Adams T, et al. Explainable machine-learning predictions for the prevention of hypoxaemia during surgery. Nat Biomed Eng. 2018;2:749–60.

    Article  PubMed  PubMed Central  Google Scholar 

  36. Song L, Langfelder P, Horvath S. Random generalized linear model: a highly accurate and interpretable ensemble predictor. BMC Bioinformatics. 2013;14:5.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  37. Marchese Robinson RL, Palczewska A, Palczewski J, Kidley N. Comparison of the predictive performance and interpretability of random forest and linear models on benchmark data sets. J Chem Inf Model. 2017;57:1773–92.

    Article  CAS  PubMed  Google Scholar 

  38. Gregers J, Christensen IJ, Dalhoff K, Lausen B, Schroeder H, Rosthoej S, et al. The association of reduced folate carrier 80G>A polymorphism to outcome in childhood acute lymphoblastic leukemia interacts with chromosome 21 copy number. Blood. 2010;115:4671–7.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  39. Luteijn RD, Zaver SA, Gowen BG, Wyman SK, Garelis NE, Onia L, et al. SLC19A1 transports immunoreactive cyclic dinucleotides. Nature. 2019;573:434–8.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  40. Laverdière C, Chiasson S, Costea I, Moghrabi A, Krajinovic M. Polymorphism G80A in the reduced folate carrier gene and its relationship to methotrexate plasma levels and outcome of childhood acute lymphoblastic leukemia. Blood. 2002;100:3832–4.

    Article  PubMed  Google Scholar 

  41. Wang SM, Sun LL, Zeng WX, Wu WS, Zhang GL. Effects of a microRNA binding site polymorphism in SLC19A1 on methotrexate concentrations in Chinese children with acute lymphoblastic leukemia. Med Oncol. 2014;31:62.

    Article  PubMed  CAS  Google Scholar 

  42. Lopez-Lopez E, Ballesteros J, Piñan MA, de Toledo JS, Garcia de Andoin N, Garcia-Miguel P, et al. Polymorphisms in the methotrexate transport pathway: a new tool for MTX plasma level prediction in pediatric acute lymphoblastic leukemia. Pharmacogenet Genomics. 2013;23:53–61.

    Article  CAS  PubMed  Google Scholar 

  43. Radtke S, Zolk O, Renner B, Paulides M, Zimmermann M, Möricke A, et al. Germline genetic variations in methotrexate candidate genes are associated with pharmacokinetics, toxicity, and outcome in childhood acute lymphoblastic leukemia. Blood. 2013;121:5145–53.

    Article  CAS  PubMed  Google Scholar 

  44. Liu SG, Gao C, Zhang RD, Zhao XX, Cui L, Li WJ, et al. Polymorphisms in methotrexate transporters and their relationship to plasma methotrexate levels, toxicity of high-dose methotrexate, and outcome of pediatric acute lymphoblastic leukemia. Oncotarget. 2017;8:37761–72.

    Article  PubMed  PubMed Central  Google Scholar 

  45. Kotnik BF, Jazbec J, Grabar PB, Rodriguez-Antona C, Dolzan V. Association between SLC19A1 gene polymorphism and high dose methotrexate toxicity in childhood acute lymphoblastic leukaemia and non Hodgkin malignant lymphoma: introducing a haplotype based approach. Radiol Oncol. 2017;51:455–62.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  46. Xu WQ, Zhang LY, Chen XY, Pan BH, Mao JQ, Song H, et al. Serum creatinine and creatinine clearance for predicting plasma methotrexate concentrations after high-dose methotrexate chemotherapy for the treatment for childhood lymphoblastic malignancies. Cancer Chemother Pharmacol. 2014;73:79–86.

    Article  CAS  PubMed  Google Scholar 

  47. Nader A, Zahran N, Alshammaa A, Altaweel H, Kassem N, Wilby KJ. Population pharmacokinetics of intravenous methotrexate in patients with hematological malignancies: utilization of routine clinical monitoring parameters. Eur J Drug Metab Pharmacokinet. 2017;42:221–8.

    Article  CAS  PubMed  Google Scholar 

  48. Hurkmans EGE, Klumpers MJ, Vermeulen SH, Hagleitner MM, Flucke U, Schreuder HWB, et al. Analysis of drug metabolizing gene panel in osteosarcoma patients identifies association between variants in SULT1E1, CYP2B6 and CYP4F8 and methotrexate levels and toxicities. Front Pharmacol. 2020;11:1241.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  49. Minematsu T, Sugiyama E, Kusama M, Hori S, Yamada Y, Ohtani H, et al. Effect of hematocrit on pharmacokinetics of tacrolimus in adult living donor liver transplant recipients. Transplant Proc. 2004;36:1506–11.

    Article  CAS  PubMed  Google Scholar 

  50. Tharwat A, Moemen YS, Hassanien AE. Classification of toxicity effects of biotransformed hepatic drugs using whale optimized support vector machines. J Biomed Inform. 2017;68:132–49.

    Article  PubMed  Google Scholar 

Download references

Acknowledgements

This work was supported by Grants from the National Natural Scientific Foundation of China (No. 81503166) and the Natural Scientific Foundation of Hunan province in China (2018JJ3846).

Author information

Author notes

  1. Min Zhan and Zebin Chen contributed equally to this work.

Authors and Affiliations

  1. Department of Pharmacy, Shenzhen Children’s Hospital, Shenzhen, 518036, People’s Republic of China

    Min Zhan, Zebin Chen & Guoqiang Wang

  2. Department of Research and Development, Shenzhen Advanced Precision Medical CO., LTD, Shenzhen, 518000, People’s Republic of China

    Changcai Ding

  3. Department of Pharmacy, Xiangya Hospital Central South University, Changsha, 410008, People’s Republic of China

    Qiang Qu

  4. Department of Hematology/Oncology, Shenzhen Children’s Hospital, Shenzhen, 518036, People’s Republic of China

    Sixi Liu & Feiqiu Wen

Authors
  1. Min Zhan
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Zebin Chen
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Changcai Ding
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Qiang Qu
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Guoqiang Wang
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. Sixi Liu
    View author publications

    You can also search for this author in PubMed Google Scholar

  7. Feiqiu Wen
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

MZ: reviewed the medical records of the patients, participated in the statistical analysis, evaluated the results, and drafted the manuscript. ZC: contributed to the study design, processed the data, and revised the manuscript. CD: conducted machine learning model building. QQ: revised the manuscript. GW: processed the data. SL: reviewed the medical reports of the patients. FW: designed and organized the study, and evaluated the results. All authors discussed and approved the final manuscript.

Corresponding author

Correspondence to Feiqiu Wen.

Ethics declarations

Conflict of interests

The authors declare no potential conflicts of interest.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Below is the link to the electronic supplementary material.

Supplementary file1 (DOCX 21 KB)

12185_2021_3184_MOESM2_ESM.tiff

Supplementary Fig. 1. The performance of models developed by machine learning algorithms without resampling in the training dataset: ROC (a); Sensitivity (b); Specificity (c); Accuracy (d). Data presented as box plot showing min, first quartile, median, third quartile, and maximum (TIFF 549 KB)

12185_2021_3184_MOESM3_ESM.tiff

Supplementary Fig. 2. The performance of models developed by machine learning algorithms combination with SMOTE in the training dataset: ROC (a); Sensitivity (b); Specificity (c); Accuracy (d). Data presented as box plot showing min, first quartile, median, third quartile, and maximum (TIFF 552 KB)

12185_2021_3184_MOESM4_ESM.tiff

Supplementary Fig. 3. The performance of predicted models developed by machine learning algorithms combination with ADASYN in the training dataset: ROC (a); Sensitivity (b); Specificity (c); Accuracy (d). Data presented as box plot showing min, first quartile, median, third quartile, and maximum (TIFF 557 KB)

12185_2021_3184_MOESM5_ESM.tiff

Supplementary Fig. 4. The performance of predicted models developed by machine learning algorithms combination with BLSMOTE in the training dataset: ROC (a); Sensitivity (b); Specificity (c); Accuracy (d). Data presented as box plot showing min, first quartile, median, third quartile, and maximum (TIFF 549 KB)

12185_2021_3184_MOESM6_ESM.tiff

Supplementary Fig. 5. The performance of predicted models developed by machine learning algorithms combination with OSS in the training dataset: ROC (a); Sensitivity (b); Specificity (c); Accuracy (d). Data presented as box plot showing min, first quartile, median, third quartile, and maximum (TIFF 548 KB)

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Zhan, M., Chen, Z., Ding, C. et al. Risk prediction for delayed clearance of high-dose methotrexate in pediatric hematological malignancies by machine learning. Int J Hematol 114, 483–493 (2021). https://doi.org/10.1007/s12185-021-03184-w

Download citation

  • Received:

  • Revised:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s12185-021-03184-w

Keywords

Search

Navigation