Machine Learning Predicts Oxaliplatin Benefit in Early Colon Cancer

Author(s): Lujia Chen, PhD¹; Ying Wang, PhD²; Chunhui Cai, PhD¹; Ying Ding, PhD^3,4; Rim S. Kim, MD^2,5; Corey Lipchik, BS²; Patrick G. Gavin, PhD^2,6; Greg Yothers, PhD^3,4; 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, MD^2,12,14; Peter C. Lucas, MD, PhD^2,12,14; Soonmyung Paik, MD^2,15; Min Sun, MD, PhD^12,14,16; Katherine L. Pogue-Geile, PhD²; Xinghua Lu, MD, PhD^1,16

Source: https://doi.org/10.1200/JCO.23.01080

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.

PURPOSE

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.

METHODS

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.

RESULTS

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).

CONCLUSION

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

Oxaliplatin-Based Adjuvant Chemotherapy in Older Patients With Stage III Colon Cancer: An ACCENT/IDEA Pooled Analysis of 12 Trials

There is more movement towards using time to relapse as an endpoint for adjuvant studies. This report highlights that there is still a role for oxaliplatin in older patients with Stage III CRC. The time to relapse was not significantly different in patients who are younger vs older. TTR is suggested as a more informative endpoint as older patients uniformly have a lower DFS and OS in most trials.

Neoadjuvant Immunotherapy in Locally Advanced Mismatch Repair–Deficient Colon Cancer

Epi and nivo for mismatch repair-deficient (dMMR) colon cancer had 98% response rate with 68% complete response, non-recurred with a median of follow up of 26 months. Does this mean bye-bye surgery??? Probably. We will need to assess those who have complete response and have a close follow up. CtDNA may play big role here, along with a need for randomized trials.

Adjuvant Hyperthermic Intraperitoneal Chemotherapy in Patients With Locally Advanced Colon Cancer (COLOPEC): 5-Year Results of a Randomized Multicenter Trial

Negative study for the use of HIPEC in the adjuvant setting of high-risk or locally advanced colorectal cancers, including those with perforation and peritoneal invasion. HIPEC doesn’t seem to reduce the risk of peritoneal recurrence after potentially curative resection.

Prognostic Impact of Early Treatment and Oxaliplatin Discontinuation in Patients With Stage III Colon Cancer: An ACCENT/IDEA Pooled Analysis of 11 Adjuvant Trials

Stopping oxaliplatinum and cont. of 5FU is better than stopping all chemotherapy. An option for patients with worsening neuropathy or other oxaliplatinum toxicity.

Physical Activity in Stage III Colon Cancer: CALGB/SWOG 80702 (Alliance)

Get your patients moving, even in the adjuvant setting! Of course, a lot of bias is possible here, but there was nearly a 10% difference in cure at three years.