Cancers (Jun 2023)

Prediction of Prostate Cancer Biochemical and Clinical Recurrence Is Improved by IHC-Assisted Grading Using Appl1, Sortilin and Syndecan-1

  • Jessica M. Logan,
  • Ashley M. Hopkins,
  • Carmela Martini,
  • Alexandra Sorvina,
  • Prerna Tewari,
  • Sarita Prabhakaran,
  • Chelsea Huzzell,
  • Ian R. D. Johnson,
  • Shane M. Hickey,
  • Ben S.-Y. Ung,
  • Joanna Lazniewska,
  • Robert D. Brooks,
  • Courtney R. Moore,
  • Maria C. Caruso,
  • Litsa Karageorgos,
  • Cara M. Martin,
  • Sharon O’Toole,
  • Laura Bogue Edgerton,
  • Mark P. Ward,
  • Mark Bates,
  • Stavros Selemidis,
  • Adrian Esterman,
  • Sheena Heffernan,
  • Helen Keegan,
  • Sarah Ní Mhaolcatha,
  • Roisin O’Connor,
  • Victoria Malone,
  • Marguerite Carter,
  • Katie Ryan,
  • Andres Clarke,
  • Nathan Brady,
  • Sonja Klebe,
  • Hemamali Samaratunga,
  • Brett Delahunt,
  • Michael J. Sorich,
  • Kim Moretti,
  • Lisa M. Butler,
  • John J. O’Leary,
  • Douglas A. Brooks

DOI
https://doi.org/10.3390/cancers15123215
Journal volume & issue
Vol. 15, no. 12
p. 3215

Abstract

Read online

Gleason scoring is used within a five-tier risk stratification system to guide therapeutic decisions for patients with prostate cancer. This study aimed to compare the predictive performance of routine H&E or biomarker-assisted ISUP (International Society of Urological Pathology) grade grouping for assessing the risk of biochemical recurrence (BCR) and clinical recurrence (CR) in patients with prostate cancer. This retrospective study was an assessment of 114 men with prostate cancer who provided radical prostatectomy samples to the Australian Prostate Cancer Bioresource between 2006 and 2014. The prediction of CR was the primary outcome (median time to CR 79.8 months), and BCR was assessed as a secondary outcome (median time to BCR 41.7 months). The associations of (1) H&E ISUP grade groups and (2) modified ISUP grade groups informed by the Appl1, Sortilin and Syndecan-1 immunohistochemistry (IHC) labelling were modelled with BCR and CR using Cox proportional hazard approaches. IHC-assisted grading was more predictive than H&E for BCR (C-statistic 0.63 vs. 0.59) and CR (C-statistic 0.71 vs. 0.66). On adjusted analysis, IHC-assisted ISUP grading was independently associated with both outcome measures. IHC-assisted ISUP grading using the biomarker panel was an independent predictor of individual BCR and CR. Prospective studies are needed to further validate this biomarker technology and to define BCR and CR associations in real-world cohorts.

Keywords