Sign in Subscribe
Cancer

Accurate Prediction of Pathological Response to Neoadjuvant Chemo-Immunotherapy in NSCLC: A Novel AI-Assisted Approach

Strategic Insights

— 2 min read

Researchers from Liaoning Cancer Hospital and Institute have made a groundbreaking discovery in the field of immunotherapy, developing a novel artificial intelligence (AI) model that accurately predicts pathological response to neoadjuvant chemo-immunotherapy in non-small-cell lung cancer (NSCLC). The study, published in the Journal for ImmunoTherapy of Cancer, utilized a retrospective cohort of 509 consecutive NSCLC cases from four Chinese thoracic-oncology centers and prospectively enrolled 50 additional patients. The AI model, named NeoPred, was trained on pre-treatment and pre-surgical computed tomography (CT) scans and achieved an impressive accuracy of 0.627 in an external validation set, surpassing the performance of board-certified radiologists.

Key Takeaways:

  • Researchers from Liaoning Cancer Hospital and Institute developed an AI model, NeoPred, that accurately predicts pathological response to neoadjuvant chemo-immunotherapy in NSCLC.
  • The model, trained on pre-treatment and pre-surgical CT scans, achieved an accuracy of 0.627 in an external validation set, outperforming expert radiologists.
  • NeoPred was developed using a retrospective cohort of 509 consecutive NSCLC cases from four Chinese thoracic-oncology centers and prospectively enrolled 50 additional patients.
  • Incorporating clinical variables into the model increased the accuracy to 0.787 in the external validation set.
  • The study concluded that NeoPred "reliably and non-invasively predicts pathological response to neoadjuvant chemo-immunotherapy in NSCLC, outperforms unaided expert assessment, and significantly enhances radiologist performance."
  • Additional multinational trials are needed to confirm the generalizability and support surgical decision-making.
  • Ongoing research may lead to improved surgical outcomes and a better understanding of chemo-immunotherapy in NSCLC.

Statistics:

  • The AI model, NeoPred, achieved an accuracy of 0.627 in an external validation set.
  • NeoPred's performance outpaced board-certified radiologists, with a mean accuracy of 0.720 in a prospective cohort.
  • Incorporating clinical variables increased the accuracy to 0.787 in the external validation set.
  • The model's performance persisted in radiologically stable-disease subgroups, achieving an external AUC of 0.742 (95% CI: 0.468 to 1.000) and a prospective AUC of 0.833 (95% CI: 0.497 to 1.000).
  • A total of 509 consecutive NSCLC cases were used in the retrospective cohort, and 50 additional patients were prospectively enrolled.

Sources:

  • NewsRx. Liaoning Cancer Hospital and Institute Researchers Focus on Immunotherapy (NeoPred: dual-phase CT AI forecasts pathologic response to neoadjuvant chemo-immunotherapy in NSCLC). Immunotherapy Weekly. June 18, 2025; p 451.
  • Journal for ImmunoTherapy of Cancer. NeoPred: dual-phase CT AI forecasts pathologic response to neoadjuvant chemo-immunotherapy in NSCLC. 2025,13(5). (http://www.immunotherapyofcancer.org/).

Read more