A novel machine learning-derived decision tree including uPA/PAI-1 for breast cancer care

Nathalie Reix; Massimo Lodi; Stéphane Jankowski; Sébastien Molière; Elisabeth Luporsi; Suzanne Leblanc; Louise Scheer; Issam Ibnouhsein; Julie-Charlotte Benabu; Victor Gabriele; Alberto Guggiola; Jean-Marc Lessinger; Marie-Pierre Chenard; Fabien Alpy; Jean-Pierre Bellocq; Karl Neuberger; Catherine Tomasetto; Carole Mathelin

doi:10.1515/cclm-2018-1065

Article

A novel machine learning-derived decision tree including uPA/PAI-1 for breast cancer care

Nathalie Reix , Massimo Lodi , Stéphane Jankowski , Sébastien Molière , Elisabeth Luporsi , Suzanne Leblanc , Louise Scheer , Issam Ibnouhsein , Julie-Charlotte Benabu , Victor Gabriele , Alberto Guggiola , Jean-Marc Lessinger , Marie-Pierre Chenard , Fabien Alpy , Jean-Pierre Bellocq , Karl Neuberger , Catherine Tomasetto and Carole Mathelin

Published/Copyright: December 20, 2018

Published by

Become an author with De Gruyter Brill

Submit Manuscript Author Information

From the journal Clinical Chemistry and Laboratory Medicine (CCLM) Volume 57 Issue 6

Abstract

Background

uPA and PAI-1 are breast cancer biomarkers that evaluate the benefit of chemotherapy (CT) for HER2-negative, estrogen receptor-positive, low or intermediate grade patients. Our objectives were to observe clinical routine use of uPA/PAI-1 and to build a new therapeutic decision tree integrating uPA/PAI-1.

Methods

We observed the concordance between CT indications proposed by a canonical decision tree representative of French practices (not including uPA/PAI-1) and actual CT prescriptions decided by a medical board which included uPA/PAI-1. We used a method of machine learning for the analysis of concordant and non-concordant CT prescriptions to generate a novel scheme for CT indications.

Results

We observed a concordance rate of 71% between indications proposed by the canonical decision tree and actual prescriptions. Discrepancies were due to CT contraindications, high tumor grade and uPA/PAI-1 level. Altogether, uPA/PAI-1 were a decisive factor for the final decision in 17% of cases by avoiding CT prescription in two-thirds of cases and inducing CT in other cases. Remarkably, we noted that in routine practice, elevated uPA/PAI-1 levels seem not to be considered as a sufficient indication for CT for N≤3, Ki 67≤30% tumors, but are considered in association with at least one additional marker such as Ki 67>14%, vascular invasion and ER-H score <150.

Conclusions

This study highlights that in the routine clinical practice uPA/PAI-1 are never used as the sole indication for CT. Combined with other routinely used biomarkers, uPA/PAI-1 present an added value to orientate the therapeutic choice.

Keywords: breast cancer; chemotherapy; machine learning; over- and under-treatment; survival; uPA/PAI-1

Corresponding author: Nathalie Reix, PhD, Clinical Biologist, Laboratoire de Biochimie et Biologie Moléculaire, Hôpitaux Universitaires de Strasbourg, 1 place de l’Hôpital, Strasbourg, France; and ICube UMR 7357, Université de Strasbourg/CNRS, Fédération de Médecine Translationnelle de Strasbourg (FMTS), 4 rue Kirschleger, Strasbourg, France, Phone: 00 33 3 69 55 08 27; Fax: 00 33 3 69 55 18 85

Acknowledgments

The authors thank Sandrine Kandel for her contribution.

Author contributions: All the authors have accepted responsibility for the entire content of this submitted manuscript and approved submission.
Research funding: This work was supported by a French non-profit association “SEVE, Seins et Vie”.
Employment or leadership: Some authors are signing under the Quantmetry affiliation. Quantmetry is a private society developing applications of artificial intelligence.
Honorarium: None declared.
Competing interests: The funding organization(s) played no role in the study design; in the collection, analysis, and interpretation of data; in the writing of the report; or in the decision to submit the report for publication.

References

1. Harris LN, Ismaila N, McShane LM, Andre F, Collyar DE, Gonzalez-Angulo AM, et al. Use of biomarkers to guide decisions on adjuvant systemic therapy for women with early-stage invasive breast cancer: American Society of Clinical Oncology Clinical Practice Guideline. J Clin Oncol 2016;34:1134–50.10.1200/JCO.2015.65.2289Search in Google Scholar PubMed PubMed Central

2. Harris L, Fritsche H, Mennel R, Norton L, Ravdin P, Taube S, et al. American Society of Clinical Oncology 2007 update of recommendations for the use of tumor markers in breast cancer. J Clin Oncol 2007;25:5287–312.10.1200/JCO.2007.14.2364Search in Google Scholar PubMed

3. Sturgeon CM, Duffy MJ, Stenman UH, Lilja H, Brünner N, Chan DW, et al. National Academy of Clinical Biochemistry laboratory medicine practice guidelines for use of tumor markers in testicular, prostate, colorectal, breast, and ovarian cancers. Clin Chem 2008;54:e11–79.10.1373/clinchem.2008.105601Search in Google Scholar PubMed

4. Molina R, Barak V, van Dalen A, Duffy MJ, Einarsson R, Gion M, et al. Tumor markers in breast cancer – European Group on Tumor Markers recommendations. Tumour Biol 2005;26:281–93.10.1159/000089260Search in Google Scholar PubMed

5. Duffy MJ, Harbeck N, Nap M, Molina R, Nicolini A, Senkus E, et al. Clinical use of biomarkers in breast cancer: Updated guidelines from the European Group on Tumor Markers (EGTM). Eur J Cancer 2017;75:284–98.10.1016/j.ejca.2017.01.017Search in Google Scholar PubMed

6. Senkus E, Kyriakides S, Ohno S, Penault-Llorca F, Poortmans P, Rutgers E, et al. Primary breast cancer: ESMO Clinical Practice Guidelines for diagnosis, treatment and follow-up. Ann Oncol 2015;26(Suppl 5):v8–30.10.1093/annonc/mdv298Search in Google Scholar PubMed

7. Institut national du cancer (INCa). Rapport intégral – UPA/PAI-1, ONCOTYPE DXtm, MAMMAPRINT® – Valeurs pronostique et prédictive pour une utilité clinique dans la prise en charge du cancer du sein. Collection Etat des lieux et des connaissances/Pratique clinique; 2014, http://www.e-cancer.fr/Expertises-et-publications/Catalogue-des-publications/Rapport-integral-UPA-PAI-1-ONCOTYPE-DXtm-MAMMAPRINT-R-Valeurs-pronostique-et-predictive-pour-une-utilite-clinique-dans-la-prise-en-charge-du-cancer-du-sein.Search in Google Scholar

8. Hanf V, Schütz F, Liedtke C, Thill M, AGO Breast Committee. AGO Recommendations for the Diagnosis and Treatment of Patients with Early Breast Cancer: Update 2015. Breast Care (Basel) 2015;10:189–97.10.1159/000431346Search in Google Scholar PubMed PubMed Central

9. Saadoun H, Lamy PJ, Thezenas S, Pouderoux S, Bibeau F, Montels F, et al. Prognostic impact of the inclusion of uPA/PAI-1 tumor levels in the current adjuvant treatment decision-making for early breast cancer. Future Oncol 2014;10:195–209.10.2217/fon.13.177Search in Google Scholar PubMed

10. Look MP, van Putten WL, Duffy MJ, Harbeck N, Christensen IJ, Thomssen C, et al. Pooled analysis of prognostic impact of urokinase-type plasminogen activator and its inhibitorPAI-1 in 8377 breast cancer patients. J Natl Cancer Inst 2002;94:116–28.10.1093/jnci/94.2.116Search in Google Scholar PubMed

11. OncoLogiK, http://www.oncologik.fr/index.php/Interregion:Sein_(principes_de_prise_en_charge).Search in Google Scholar

12. OncoLogiK, http://www.oncologik.fr/index.php/Interregion:Sein_(principes_de_prise_en_charge)#Arbres_de_d.C3.A9cisions_selon_RH_et_HER2.Search in Google Scholar

13. Reix N, Malina C, Chenard MP, Bellocq JP, Delpous S, Molière S, et al. A prospective study to assess the clinical utility of serum HER2 extracellular domain in breast cancer with HER2 overexpression. Breast Cancer Res Treat 2016;160:249–59.10.1007/s10549-016-4000-zSearch in Google Scholar PubMed PubMed Central

14. Thike AA, Chng MJ, Fook-Chong S, Tan PH. Immunohistochemical expression of hormone receptors in invasive breast carcinoma: correlation of results of H-score with pathological parameters. Pathology 2001;33:21–5.10.1080/00313020123290Search in Google Scholar

15. Borstnar S, Sadikov A, Mozina B, Cufer T. High levels of uPA and PAI-1 predict a good response to anthracyclines. Breast Cancer Res Treat 2010;121:615–24.10.1007/s10549-009-0691-8Search in Google Scholar PubMed

16. Jacobs VR, Augustin D, Wischnik A, Kiechle M, Hoess C, Steinkohl O, et al. Concordance rates of biomarkers uPA and PAI-1 results in primary breast cancer vs. consecutive tumor board decision and therapy performed in clinical hospital routine: Results of a prospective multi-center study at certified breast centers. Breast 2016;29:208–12.10.1016/j.breast.2016.06.014Search in Google Scholar PubMed

17. Vénat-Bouvet L, Fermeaux V, Leobon S, Saidi N, Monteil J, Mollard J, et al. Adjuvant chemotherapy in node-negative breast cancer: UPA/PAI-1 determinations for 163 cases. Anticancer Res 2014;34:1213–7.Search in Google Scholar

18. Breiman L, Friedman JH, Olshen RA, Stone CJ. Classification and regression trees. Belmont, CA: Wadsworth, 1984.Search in Google Scholar

19. Therneau T, Atkinson B, Ripley B. rpart: Recursive Partitioning and Regression Trees. R package version 4.1-10. https://cran.r-project.org/package=rpart; 2015.Search in Google Scholar

20. Kolben T, Augustin D, Armbrust R, Kolben TM, Degenhardt T, Burgmann M, et al. Impact of guideline-based use of uPA/PAI-1 on patient outcome in intermediate-risk early breast cancer. Breast Cancer Res Treat 2016;155:109–15.10.1007/s10549-015-3653-3Search in Google Scholar PubMed

21. Mazouni C, Spyratos F, Romain S, Fina F, Bonnier P, Ouafik LH, et al. A nomogram to predict individual prognosis in node-negative breast carcinoma. Eur J Cancer 2012;48:2954–61.10.1016/j.ejca.2012.04.018Search in Google Scholar PubMed

22. Goldhirsch A, Ingle JN, Gelber RD, Coates AS, Thürlimann B, Senn HJ, et al. Thresholds for therapies: highlights of the St Gallen International Expert Consensus on the primary therapy of early breast cancer 2009. Ann Oncol 2009;20:1319–29.10.1093/annonc/mdp322Search in Google Scholar PubMed PubMed Central

23. Gulshan V, Peng L, Coram M, Stumpe MC, Wu D, Narayanaswamy A, et al. Development and validation of a deep learning algorithm for detection of diabetic retinopathy in retinal fundus photographs. J Am Med Assoc 2016;316:2402–10.10.1001/jama.2016.17216Search in Google Scholar PubMed

24. Gonçalves-Ribeiro S, Sanz-Pamplona R, Vidal A, Sanjuan X, Guillen Díaz-Maroto N, Soriano A, et al. Prediction of pathological response to neoadjuvant treatment in rectal cancer with a two-protein immunohistochemical score derived from stromal gene-profiling. Ann Oncol 2017;28:2160–8.10.1093/annonc/mdx293Search in Google Scholar PubMed

25. Somashekhar SP, Sepúlveda MJ, Puglielli S, Norden AD, Shortliffe EH, Rohit Kumar C, et al. Watson for oncology and breast cancer treatment recommendations: agreement with an expert multidisciplinary tumor board. Ann Oncol 2018;29:418–23.10.1093/annonc/mdx781Search in Google Scholar PubMed

26. Coates AS, Winer EP, Goldhirsch A, Gelber RD, Gnant M, Piccart-Gebhart M, et al. Tailoring therapies--improving the management of early breast cancer: St Gallen International Expert Consensus on the Primary Therapy of Early Breast Cancer 2015. Ann Oncol 2015;26:1533–46.10.1093/annonc/mdv221Search in Google Scholar PubMed PubMed Central

27. Viala M, Alexandre M, Thezenas S, Lamy PJ, Maran-Gonzalez A, Gutowski M, et al. Prognostic impact of the inclusion of uPA/PAI-1 for adjuvant treatment decision-making in ER+/Her2- pN0 early breast cancers. Breast Cancer Res Treat 2017;165:611–21.10.1007/s10549-017-4373-7Search in Google Scholar PubMed

Received: 2018-09-28

Accepted: 2018-11-06

Published Online: 2018-12-20

Published in Print: 2019-05-27

You are currently not able to access this content.

Articles in the same Issue

https://doi.org/10.1515/cclm-2018-1065

Keywords for this article

breast cancer; chemotherapy; machine learning; over- and under-treatment; survival; uPA/PAI-1