Shifting Paradigms for High-grade Prostatic Intraepithelial Neoplasia

Abstract

High-grade prostatic intraepithelial neoplasia (HGPIN) has long been considered a histologic precursor to invasive prostate adenocarcinoma (PCA), with the canonical view stating that HGPIN progresses unidirectionally to PCA. This notion is based on the observations that HGPIN lesions are often found in close proximity to PCA and share a large numberofdefiningmolecular andmorphologic featureswith invasive carcinoma. This is demonstrated in previous studies showing that numerous genomic lesions (eg, mutations in SPOP, ERG rearrangement, and deletions inNKX3.1 and PTEN) as well as epigenetic alterations are shared between HGPIN and synchronous PCA [1–6]. More recent insights into the biology of HGPIN have challenged both the clinical importance of HGPIN and the molecular relationship between HGPIN and PCA and suggest that HGPIN, like PCA, could be a heterogeneous entity with variable clinical significance. In this issue of European Urology, Shin and colleagues [7] use laser-capture microdissection and next-generation sequencing toaddressa criticalquestion:Whatgenomic lesions are already present in HGPIN, considered the histologic precursor to PCA, and what lesions herald the transition to true invasive adenocarcinoma? They defined genetic abnormalities in paired areas of HGPIN and associated prostate cancer in six patients. The authors reported only a small number of genetic abnormalities in HGPIN and, even more important, few similarities between the HGPIN and PCA areas; thephylogenetic trees generated (shown inFig. 5 of the study [7]) suggest minimal clonal relationship between HGPIN and PCA. The authors concluded that the majority of known genomic changes in PCA are cancer-specific events that occur after the transition from HGPIN. This study is innovative because little information on the mutational spectrum of HGPIN is currently available, and no other study has evaluated genomic alterations in HGPIN at such resolution [7]. The presented results, however, contrast with prior reports suggesting a higher level of overlapping alterations between HGPIN and PCA. In general, major technical challenges are associated with the analysis of HGPIN lesions. Themajor limitation of studyingHGPIN is the complexity of tissue architecture. HGPIN is usually found as dispersedglandswithahigh level of ‘‘contaminating’’ normal cells intermixed. Given this cellular heterogeneity and the limitations of current sequencing technology, it remains very challenging to conclusively prove that the genetic alterations of interest are truly absent in HGPIN. The authors need to showthat the sensitivity todetect themisequal inHGPINand associated PCA. Given the challenges of tumor purity inherent to prostate neoplasia, this is no easy feat, and it is unclear that the authors clear this admittedlyhighbar,which makes these data and the validity of the conclusions drawn from them difficult to interpret. The conclusion thatHGPIN shows fewof the critical driver genomic alterations of PCA seems timely. It comes about as the clinical relevanceofHGPIN, regarding its associationwith clinically significant PCA, has also shown a dramatic shift in recent years. Previous generations of urologists were trained that detection of HGPIN on prostate biopsy necessitated immediate rebiopsy because the high risk of development of future PCA was up 80% [8]. It now appears this was due to undersampling of invasive PCA in the sextant biopsy era [9]; the shift to standard 12to 14-core templates has resulted in a decrease in the predictive value of HGPIN on subsequent detection of PCA, and isolated HGPIN is no longer an EU RO P E AN URO LOGY 6 9 ( 2 0 1 6 ) 8 3 1 – 8 3 3