AUA 2024: Development of a Prediction Model of Survival Amongst Patients with Metastatic Hormone Sensitive Prostate Cancer Using Big Data: Interim Results from the European Network of Excellence for Big Data in Prostate Cancer (PIONEER)

( The 2024 American Urological Association (AUA) annual meeting held in San Antonio, TX between May 3 and May 6, 2024, was host to the advanced prostate cancer moderated poster session. Dr. Rossella Nicoletti aimed to develop a prediction model of survival amongst patients with metastatic hormone-sensitive prostate cancer (mHSPC) using big data from the PIONEER study, the European network of excellence for big data in prostate cancer evaluating real-world trends and outcomes for European patients with mHSPC.

Dr. Nicoletti began by highlighting that androgen deprivation therapy (ADT) alone used to be the standard of care (SOC) for mHSPC. However, with the introduction of androgen receptor pathway inhibitors (ARPIs), taxane-based chemotherapy and its combinations with ARPIs (triplets) along with ADT have proven to be more effective than ADT monotherapy and are now the SOC. Real-world data have been acknowledged for their capacity to generate real-world evidence, filling gaps in knowledge that traditional research methods may overlook. Their aim was to create a predictive model for survival in patients with mHSPC undergoing treatment with one of the approved regimens for mHSPC. This prediction model was based on patient clinical characteristics using real-world world data.

For developing the prediction model, they used data from patients with mHSPC across a distributed network of observational databases and divided the patients into two large cohorts:

  • Cohort 1: Male patients without prior ADT with mHSPC
    • 1.1 Metachronous mHSPC
    • 1.2 Synchronous mHSPC
  • Cohort 2 The start of ADT as a proxy definition of mHSPC disease.
    • 2.1 Metachronous mHSPC
    • 2.2 Synchronous mHSPC

This analysis included 94,261 mHSPC patients in cohort 1 and 77,123 in cohort 2. The clinical and demographic variables increasing the odds of mortality were: 

  • Age >90
  • Non-adenocarcinoma variants
  • M1c disease (visceral metastasis)
  • Liver metastasis

The variables decreasing the Odds of death were:

  • Age <80
  • Adenocarcinoma of the prostate
  • M1a-M1b disease
  • ISUP Grade Group 3 or lower
  • cT1 Disease. 

With this prediction model, they were only able to predict survival in the first year since diagnosis with an accuracy of 75%, mainly because there are multiple confounding factors for mortality, however, the major factor is that mHSPC patients are heterogenous, and there is a need to better risk stratification of these patients. The graphic below shows the performance of the prediction model for each year after index diagnosis.PIONEER study prediction accuracy
They found that 60% of deaths tend to occur in the first two years after the date of mHSPC diagnosis, and the number of events (deaths) decreased between years 2-5.PIONEER study time at risk
Dr Nicoletti finished her presentation by concluding:

  • This is the largest study in Europe using real-world data from the PIONEER project in mHSPC.
  • There is vast heterogeneity in the demographic and clinical characteristics of patients with mHSPC, as well as in treatment approaches.
  • Their prediction model shows age, non-adenocarcinoma variants, and visceral metastasis, increased the odds of death in mHSPC. However, further validation is needed to confirm the reliability of their findings.

Presented by: Rossella Nicoletti, MD, Urology Resident, Careggi University Hospital in Florence, Italy.

Written by: Julian Chavarriaga, MD - Society of Urologic Oncology (SUO) Clinical Fellow at The University of Toronto, @chavarriagaj on Twitter during the 2024 American Urological Association (AUA) annual meeting held in San Antonio, TX between May 3rd and May 6th, 2024