Machine Learning Predicts Oxaliplatin Benefit in Early Colon Cancer

Author(s): Lujia Chen, PhD1; Ying Wang, PhD2; Chunhui Cai, PhD1; Ying Ding, PhD3,4; Rim S. Kim, MD2,5; Corey Lipchik, BS2; Patrick G. Gavin, PhD2,6; Greg Yothers, PhD3,4; Carmen J. Allegra, MD7; Nicholas J. Petrelli, MD8; Jennifer Marie Suga, MD9; Judith O. Hopkins, MD10; Naoyuki G. Saito, MD, PhD11; Terry Evans, MD12; Srinivas Jujjavarapu, MD, PhD13; Norman Wolmark, MD2,12,14; Peter C. Lucas, MD, PhD2,12,14; Soonmyung Paik, MD2,15; Min Sun, MD, PhD12,14,16; Katherine L. Pogue-Geile, PhD2; Xinghua Lu, MD, PhD1,16
Anjan J Patel MD

Dr. Anjan Patel's Thoughts

COLOXIS is an interesting computer-based learning model that seems to be able to predict which patients are likely to benefit from adjuvant chemotherapy in colorectal cancer. This, along with ctDNA tests, may change whom we choose to offer adjuvant chemotherapy.


A combination of fluorouracil, leucovorin, and oxaliplatin (FOLFOX) is the standard for adjuvant therapy of resected early-stage colon cancer (CC). Oxaliplatin leads to lasting and disabling neurotoxicity. Reserving the regimen for patients who benefit from oxaliplatin would maximize efficacy and minimize unnecessary adverse side effects.


We trained a new machine learning model, referred to as the colon oxaliplatin signature (COLOXIS) model, for predicting response to oxaliplatin-containing regimens. We examined whether COLOXIS was predictive of oxaliplatin benefits in the CC adjuvant setting among 1,065 patients treated with 5-fluorouracil plus leucovorin (FULV; n = 421) or FULV + oxaliplatin (FOLFOX; n = 644) from NSABP C-07 and C-08 phase III trials. The COLOXIS model dichotomizes patients into COLOXIS+ (oxaliplatin responder) and COLOXIS– (nonresponder) groups. Eight-year recurrence-free survival was used to evaluate oxaliplatin benefits within each of the groups, and the predictive value of the COLOXIS model was assessed using the P value associated with the interaction term (int P) between the model prediction and the treatment effect.


Among 1,065 patients, 526 were predicted as COLOXIS+ and 539 as COLOXIS–. The COLOXIS+ prediction was associated with prognosis for FULV-treated patients (hazard ratio [HR], 1.52 [95% CI, 1.07 to 2.15]; P = .017). The model was predictive of oxaliplatin benefits: COLOXIS+ patients benefited from oxaliplatin (HR, 0.65 [95% CI, 0.48 to 0.89]; P = .0065; int P = .03), but COLOXIS– patients did not (COLOXIS– HR, 1.08 [95% CI, 0.77 to 1.52]; P = .65).


The COLOXIS model is predictive of oxaliplatin benefits in the CC adjuvant setting. The results provide evidence supporting a change in CC adjuvant therapy: reserve oxaliplatin only for COLOXIS+ patients, but further investigation is warranted.

Author Affiliations

1Department of Biomedical Informatics, University of Pittsburgh, Pittsburgh, PA; 2NSABP/NRG Oncology, Pittsburgh, PA; 3NRG Oncology Statistics and Data Management Center, Pittsburgh, PA; 4Department of Biostatistics, University of Pittsburgh, Pittsburgh, PA; 5AstraZeneca, Oncology Translational Medicine, Gaithersburg, MD; 6AstraZeneca Respiratory and Immunology, Gaithersburg, MD; 4Department of Biostatistics, Univesity of Pittsburgh, Pittsburgh, PA; 7Department of Medicine, University of Florida Health, Gainesville, FL; 8Helen F. Graham Cancer Center and Research Institute at Christiana Care, Newark, DE; 9Kaiser Permanente Oncology Clinical Trials, KP NCI Community Oncology Research Program (NCORP), Vallejo, CA; 10Novant Health Forsyth Medical Cancer Institute/Southeast Clinical Oncology Research NCORP, Kernersville, NC; 11Melvin and Bren Simon Cancer Center, Indiana University School of Medicine, Indianapolis, IN; 12UPMC Hillman Cancer Center, Pittsburgh, PA; 14UPMC Hillman Cancer Center, University of Pittsburgh School of Medicine, Pittsburgh, PA; 15Yonsei University College of Medicine, Yonsei Biomedical Research Institute, Seoul, Republic of South Korea; 16DeepRx Inc, Pittsburgh, PA

Leave a Comment

Your email address will not be published. Required fields are marked *

Related Articles