Validation of an Artificial Intelligence-Based Prognostic Biomarker in Patients with Oligometastatic Castration-Sensitive Prostate Cancer - Beyond the Abstract

The optimal management of patients with metastatic castration-sensitive prostate cancer (mCSPC) remains an area of active investigation, with strategies ranging from metastasis-directed therapy (MDT) to combination systemic therapy. Many clinical and biological factors impact prognosis and patterns of progression for metastatic disease and can be used to guide treatment decisions. These factors include metastatic disease burden (e.g. low versus high volume), timing of metastasis (synchronous versus metachronous), and presence of specific mutations.

For oligometastatic castration-sensitive prostate cancer (omCSPC), an intermediate state between localized disease and widespread metastasis usually defined as having 5 or fewer metastases, phase 2 randomized clinical trials have demonstrated a role for MDT in metachronous disease.1–4 MDT can delay hormone therapy use and represents an important standard of care option for patients with omCSPC. However, omCSPC is a heterogeneous disease state, which requires more accurate and accessible prognostic and predictive biomarkers to determine which patients do well with MDT alone versus those that require treatment intensification.

The multimodal artificial intelligence biomarker (MMAI) biomarker developed by ArteraAI is calculated from digitized histopathology slides and clinical variables. The MMAI biomarker has been validated using prospectively collected data across large, randomized trials and was shown to outperform traditional risk group stratification.5–7 Based on these studies, the MMAI biomarker has been incorporated into the NCCN Guidelines (4.2024) as a risk stratification and predictive tool for localized prostate cancer.

In this study, we investigated the MMAI biomarker in patients with omCSPC.8 The cohort included patients (n=222) from Johns Hopkins Hospital and Ghent Hospital who were treated with standard of care (SOC) or randomized in the STOMP and ORIOLE trials to observation versus MDT. SOC treatments included MDT, androgen deprivation therapy (ADT), and ADT plus androgen receptor signaling inhibitor (ARSI) or docetaxel.

The primary objective was to evaluate overall survival (OS). Secondary objectives included evaluating time to castration-resistant prostate cancer (ttCRPC) and whether MMAI score could predict response to MDT. Analysis of response to MDT was focused on patients within the randomized STOMP and ORIOLE trials, with an endpoint of metastasis-free survival (MFS, defined as development of new metastases from time of randomization or death by any cause). Patients were stratified by median MMAI score to “high” versus “low” score. We found that high score was associated with worse OS (HR=6.46, 95% CI=1.44-28.9; p=0.01) and ttCRPC (HR=2.07, 95% CI=1.15-3.72; p=0.015). In the STOMP/ORIOLE cohort (n=51), MMAI score also appeared to predict for benefit from MDT. Patients with high score (p=0.039) but not low score (p=0.69) showed improved MFS with MDT (p-interaction=0.04).

Altogether, these data are a promising step toward developing a prognostic and predictive biomarker for omCSPC. The results are largely hypothesis generating, providing a foundation for further investigation. Future studies include validation of the MMAI biomarker in a larger cohort of metastatic patients, and ideally within a prospective trial.

Written by: Jarey H Wang,1 Matthew P Deek,2 Adrianna A Mendes,1 Yang Song,1 Amol Shetty,3 Soha Bazyar,3 Kim Van der Eecken,4 Emmalyn Chen,5 Timothy N Showalter,6 Trevor J Royce,5 Tamara Todorovic,5 Huei-Chung Huang,5 Scott A Houck,5 Rikiya Yamashita,5 Ana P Kiess,1 Daniel Y Song,1 Tamara Lotan,1 Theodore DeWeese,1 Luigi Marchionni,7 Lei Ren,3 Amit Sawant,3 Nicole Simone,8 Alejandro Berlin,9 Cem Onal,10 Andre Esteva,5 Felix Y Feng,11 Phuoc T Tran,12 Philip Sutera,13 Piet Ost14

  1. Johns Hopkins University, Baltimore, MD, USA.
  2. Rutgers Robert Wood Johnson Medical School, New Brunswick, NJ, USA.
  3. University of Maryland Medical Center, Baltimore, MD, USA.
  4. Ghent University Hospital, Ghent, Belgium.
  5. Artera Inc., Los Altos, CA, USA.
  6. Artera Inc., Los Altos, CA, USA; University of Virginia, Charlottesville, VA, USA.
  7. Weill Cornell Medical College, New York, NY, USA.
  8. Thomas Jefferson University, Philadelphia, PA, USA.
  9. Princess Margaret Cancer Centre, Toronto, Canada.
  10. Baskent University, Ankara, Turkey.
  11. Artera Inc., Los Altos, CA, USA; University of California San Francisco, San Francisco, CA, USA.
  12. University of Maryland Medical Center, Baltimore, MD, USA.
  13. University of Rochester Medical Center, Rochester, NY, USA.
  14. Department of Human Structure and Repair, Ghent University, Ghent, Belgium; Department of Radiation Oncology, Iridium Network, Antwerp, Belgium.
References:

  1. Ost P, Reynders D, Decaestecker K, et al. Surveillance or metastasis-directed therapy for oligometastatic prostate cancer recurrence (STOMP): Five-year results of a randomized phase II trial. Journal of Clinical Oncology. 2020;38(6_suppl):10-10. doi:10.1200/jco.2020.38.6_suppl.10
  2. Phillips R, Shi WY, Deek M, et al. Outcomes of Observation vs Stereotactic Ablative Radiation for Oligometastatic Prostate Cancer: The ORIOLE Phase 2 Randomized Clinical Trial. JAMA Oncology. 2020;6(5):650-659. doi:10.1001/jamaoncol.2020.0147
  3. Palma DA, Olson R, Harrow S, et al. Stereotactic Ablative Radiotherapy for the Comprehensive Treatment of Oligometastatic Cancers: Long-Term Results of the SABR-COMET Phase II Randomized Trial. JCO. 2020;38(25):2830-2838. doi:10.1200/JCO.20.00818
  4. Tang C, Sherry AD, Haymaker C, et al. Addition of Metastasis-Directed Therapy to Intermittent Hormone Therapy for Oligometastatic Prostate Cancer: The EXTEND Phase 2 Randomized Clinical Trial. JAMA Oncology. 2023;9(6):825-834. doi:10.1001/jamaoncol.2023.0161
  5. Esteva A, Feng J, van der Wal D, et al. Prostate cancer therapy personalization via multi-modal deep learning on randomized phase III clinical trials. npj Digit Med. 2022;5(1):1-8. doi:10.1038/s41746-022-00613-w
  6. Spratt DE, Tang S, Sun Y, et al. Artificial Intelligence Predictive Model for Hormone Therapy Use in Prostate Cancer. NEJM Evidence. 2023;2(8):EVIDoa2300023. doi:10.1056/EVIDoa2300023
  7. Ross AE, Zhang J, Huang HC, et al. External Validation of a Digital Pathology-based Multimodal Artificial Intelligence Architecture in the NRG/RTOG 9902 Phase 3 Trial. European Urology Oncology. Published online February 1, 2024. doi:10.1016/j.euo.2024.01.004
  8. Wang JH, Deek MP, Mendes AA, et al. Validation of an artificial intelligence-based prognostic biomarker in patients with oligometastatic Castration-Sensitive prostate cancer. Radiotherapy and Oncology. 2025;202. doi:10.1016/j.radonc.2024.110618
Read the Abstract