2. UroToday Home
  3. Recent Abstracts
  4. Urologic Oncology
  5. Bladder Cancer

AI-Powered Biomarker to Predict Response to High-Grade NMIBC Treatments - Expert Commentary

High-grade non-muscle-invasive bladder cancer (NMIBC) is usually treated with intravesical bacillus Calmette-Guerin (BCG) as first-line adjuvant therapy, with sequential intravesical gemcitabine and docetaxel (Gem/Doce) emerging as an alternative. A recent study evaluated whether an artificial intelligence-augmented histologic assay (CHAI biomarker) could predict differential treatment responses to guide clinical decision-making.

In this retrospective analysis, the researchers assessed 253 patients with treatment-naive high-grade NMIBC who received either BCG (63%) or Gem/Doce (37%). The CHAI biomarker was present in 30% of patients. Among BCG-treated patients, those with the biomarker present had significantly worse high-grade recurrence-free survival (HG-RFS) compared to biomarker-absent patients (hazard ratio [HR] 2.0, 95% confidence interval [CI] 1.1-3.6, p = 0.023). In contrast, among Gem/Doce-treated patients, biomarker status did not significantly affect HG-RFS (HR 0.47, 95% CI 0.13-1.7, p = 0.2). For biomarker-positive patients, 24-month HG-RFS rates were 56% (95% CI 43-73%) with BCG versus 90% (95% CI 79-100%) with Gem/Doce (HR 5.4, 95% CI 1.6-18.3, p = 0.007). The biomarker-treatment interaction term was statistically significant (p = 0.029), indicating the biomarker's predictive rather than merely prognostic value.

This interesting study adds to the nascent field of AI-based predictive biomarkers in bladder cancer. Limitations include its retrospective nature, data from only two centers, potential unknown confounders in treatment allocation, and wide confidence intervals due to smaller sample sizes in the biomarker-present subgroups. Despite these limitations, this innovative study represents a novel use of an AI-based biomarker to predict which high-grade NMIBC patients are less likely to benefit from BCG and may instead benefit from alternative treatments like Gem/Doce. These findings could help advance precision medicine in NMIBC management. Understanding the mechanistic basis and model interpretability will be important for future efforts in this field.

Written by: Bishoy M. Faltas, MD, Chief Research Officer, Englander Institute for Precision Medicine, Gellert Family - John P. Leonard, MD, Research Scholar, Associate Professor of Medicine, Cell and Developmental Biology, Weill Cornell Medicine, New York- Presbyterian Hospital, NY

References:

  1. Packiam VT, McElree IM, Ghodoussipour S, Nimgaonkar V, Krishna V, Kim JK, Allison DB, Richards JR, Rajan KDA, Chen SJ, Lotan Y, Williams SB, Zhang H, Watson D, Vrabac D, Abuzeid WM, Joshi A, Kamat AM, O'Donnell MA, Hensley PJ. Presence of an Artificial Intelligence–powered Predictive Biomarker Is Associated with a Poor Response to Intravesical Bacillus Calmette-Guerin but Not to Intravesical Sequential Gemcitabine/Docetaxel in Patients with High-grade Non–muscle-invasive Bladder Cancer. Eur Urol Oncol. 2025 Apr 9. doi: 10.1016/j.euo.2025.04.006.
Read the Abstract