Annual Meeting of the NCI Cohort Consortium (Abstract Submission): Submission #4

Presenter Information
---------------------
First Name: Hannah
Middle Initial: E.
Last Name: Guard
Degree(s): M.S.
Position/Title/Career Status: Research Assistant IV
Organization: Department of Epidemiology, Harvard T.H. Chan School of Public Health
Email: hguard@hsph.harvard.edu

Additional Authors
------------------
List of Additional Authors:
- First Name: Lia
  Last Name: De Paula Oliveria
  Degrees: M.D.
  Affiliation: Department of Pathology, Johns Hopkins University School of Medicine
- First Name: Aidan
  Last Name: Daluiski
  Affiliation: Department of Epidemiology, Harvard T.H. Chan School of Public Health
- First Name: Adrianna
  Last Name: Amaral de Aragao
  Degrees: M.D.
  Affiliation: Department of Pathology, Johns Hopkins University School of Medicine
- First Name: Jane B.
  Last Name: Vaselkiv
  Degrees: M.P.H.
  Affiliation: Department of Epidemiology, Harvard T.H. Chan School of Public Health
- First Name: Sinead
  Last Name: Flanagan
  Degrees: M.B. B.Ch B.A.O., M.P.H.
  Affiliation: Department of Histopathology and Morbid Anatomy, Trinity Translational Medicine Institute, Trinity College Dublin
- First Name: Colleen
  Last Name: McGrath
  Degrees: M.S.
  Affiliation: Department of Epidemiology, Harvard T.H. Chan School of Public Health
- First Name: Magdalena
  Last Name: Brinker
  Degrees: M.S.
  Affiliation: Department of Epidemiological Methods and Etiological Research, Leibniz Institute for Prevention Research and Epidemiology – BIPS and Faculty of Human and Health Sciences, University of Bremen
- First Name: Michelangelo
  Last Name: Fiorentino
  Degrees: M.D., Ph.D.
  Affiliation: Pathology Unit, Maggiore Hospital-AUSL Bologna and Department of Experimental, Diagnostic and Specialty Medicine (DIMES), University of Bologna
- First Name: Massimo
  Last Name: Loda
  Degrees: M.D.
  Affiliation: Department of Pathology and Laboratory Medicine, New York Presbyterian Hospital, Weill Cornell Medicine
- First Name: Corinne
  Last Name: Joshu
  Degrees: Ph.D., M.P.H., M.A.
  Affiliation: Department of Epidemiology, Johns Hopkins Bloomberg School of Public Health; Sidney Kimmel Comprehensive Cancer Center at Johns Hopkins
- First Name: Lorelei
  Last Name: Mucci
  Degrees: Sc.D., M.P.H.
  Affiliation: Department of Epidemiology, Harvard T.H. Chan School of Public Health
- First Name: Konrad
  Last Name: Stopsack
  Degrees: M.D., M.P.H.
  Affiliation: Department of Epidemiological Methods and Etiological Research, Leibniz Institute for Prevention Research and Epidemiology – BIPS and Faculty of Human and Health Sciences, University of Bremen
- First Name: Tamara
  Last Name: Lotan
  Degrees: M.D.
  Affiliation: Department of Pathology, Johns Hopkins University School of Medicine


Abstract Information
--------------------
Abstract Title: Tumor expression of ERG, PTEN, and p53 and lethal prostate cancer in two prospective cohort studies
Abstract:
Background: Tumor biomarkers, including PTEN loss and TP53 mutations, can identify prognostic subtypes of prostate cancer. We evaluated p53 and PTEN loss with or without ERG tumor protein expression and lethal prostate cancer in two prospective cohort studies. 

Methods: 1,085 participants of the Health Professionals Follow-up Study and Physicians’ Health Study diagnosed with prostate cancer (1986-2017) and treated surgically were followed for lethality (metastasis or prostate cancer death) through 2022. PTEN loss, ERG, and p53-mutations in prostate tumors were assessed using validated immunohistochemistry assays on tissue microarrays. We estimated hazard ratios (HR) and 95% confidence intervals (CI) for associations between each biomarker and lethality using Cox proportional hazards models adjusted for Gleason score, pathologic TNM stage, and the other biomarkers. We also assessed the joint association of PTEN loss and ERG with lethality. We evaluated prognostic discrimination of the biomarkers in addition to clinical factors across time using time-varying area-under-the-curve (tAUC).

Results: Over median 19 years follow-up, 113 lethal events occurred. The prevalence of each subtype was: 53% ERG-positive, 4% p53-positive, 14% complete PTEN loss (overall), 9% PTEN loss/ERG-positive, 5% PTEN loss/ERG-negative. PTEN loss (HR: 2.3, 95% CI: 1.5, 3.6) and p53 (HR: 2.9, 95% CI: 1.7, 5.1) were associated with lethal prostate cancer. PTEN loss/ERG-negative tumors had 4.3-fold higher risk of lethality (95% CI: 2.4, 7.8) compared to PTEN intact/ERG-negative tumors. The risk was not elevated in PTEN loss/ERG-positive tumors (HR: 1.3, 95% CI: 0.7, 2.3). tAUCs for p53 (0.55-0.57) and PTEN (0.59-0.65) were low. tAUCs for clinical factors ranged from 0.80 to 0.86 over time. The addition of the biomarkers did not add information for prognosis.

Conclusion/Discussion: Complete PTEN loss in ERG-negative tumors and p53 were associated with higher risk of lethal prostate cancer, but they did not meaningfully improve prediction of lethality beyond Gleason grade and stage.