38IN - Do we need a new taxonomy for prostate cancer?

Date 27 September 2014
Event ESMO 2014
Session Precision medicine in prostate cancer
Topics Prostate Cancer
Presenter Mark Rubin
Citation Annals of Oncology (2014) 25 (suppl_4): iv15-iv16. 10.1093/annonc/mdu298
Authors M. Rubin
  • Institute For Precision Medicine, Department Of Pathology & Laboratory Medicine, Weill Cornell Medical College, 10065 - New York/US




Our current ability to diagnose and grade clinically localized, hormone naïve prostate cancer has served us well. The Gleason grading system, developed empirically, is one of the most robust pathology tumor grading schemes. Hormone naïve prostate cancer, unlike some other common cancers (e.g., renal, breast, and ovarian), presents with little histologic variability, presenting most commonly as an adenocarcinoma and rarely with variant histology (e.g., small cell or pure ductal). However, we are now in a new era where a variety of hormonal treatments has led to longer patient survival and the emergence of less common prostate cancer variants characterized best by both histologic and molecular alterations. Metastatic prostate cancer biopsies following treatment now create a need for a new taxonomy for prostate cancer to capture the range of disease not previously appreciated by pathologists, oncologists, or oncologic surgeons. Defining castration resistant prostate cancer (CRPC) may require including information regarding androgen receptor signaling (e.g., AR status and/or AR signaling signatures), ETS fusions (e.g., TMRPSS2-ERG, BRAF), PTEN/PI3K/AKT, P53, and RB status. Defining neuroendocrine transdifferentiation is also emerging as an important subclass of advanced prostate cancer with distinct histologic and molecular features including amplification of AUKRA/MYCN and loss of AR. Finally, efforts in precision medicine with next generation sequencing are identifying subclasses of prostate cancer that demonstrate significant alterations in DNA damage repair. For example, tumors harboring PTEN/p53 or SPOP mutations may benefit from the use of PARP inhibitors in co-targeting therapeutic strategies creating a synthetic lethal phenotype similar to BRCA1 mutations. Taken together, a future prostate cancer taxonomy must include reproducible histologic definitions useful in the setting of treated disease as well as robust molecular testing. Future clinical trials such as those conducted by PCF/Stand-Up-2-Cancer teams should pave the way for the empirical development of a new prostate cancer molecular-based taxonomy.


The author has declared no conflicts of interest.