Association of ERG/PTEN Status with PSA Recurrence after Radical Prostatectomy
Rohit Mehra, Arul M Chinnaiyan, Lakshmi P Kunju, Angela Wu, Rajal B Shah, Nallasivam Palanisamy, Robert Lonigro, Scott A Tomlins. University of Michigan, Ann Arbor, MI; Caris Pathology, Irving, TX
Background: Previous studies have demonstrated that ERG rearrangement is present in more than 45% of clinically localized prostate cancers and more than 40% of lethal castration resistant metastatic prostate cancers. Our group has previously demonstrated a significant association between PTEN deletion and ERG rearrangement both in localized cancers and metastases (Modern Pathology Aug;22(8):1083-93, 2009). Here, we attempt to delineate the cooperative role of ERG and PTEN, if any, in prostate cancer outcome.
Design: We evaluated ERG and PTEN status on 235 clinically localized prostate cancer cases represented on tissue micro arrays (TMAs) using a previously validated FISH strategy. We also employed array comparative genomic hybridization (aCGH) to asses ERG and PTEN on an additional 59 clinically localized prostate cancer cases. Excluding the overlapping cases on TMAs and those used for aCGH, a total of 261 clinically localized prostate cancer cases with long term follow-up were evaluable for both ERG and PTEN.
Results: 3 cases of clinically localized prostate cancer demonstrated SPINK1 overexpression, 1 case demonstrated RAF1 rearrangement and 5 cases demonstrated ETV1/ETV5 rearrangements and hence were excluded from this study. Out of the remaining 252 evaluable cases, 63 cases exhibited clinical recurrence over a median follow-up of 5.9 years. Overall, 123/252 cases (49%) demonstrated ERG rearrangement and 34/208 (16%) cases demonstrated PTEN loss. Hemizygous and homozygous deletion of PTEN was seen in 10/129 (8%) and 2/129 (1.5%) of ERG negative cases, respectively. In contrast, hemizygous and homozygous deletion of PTEN was seen in 14/123 (11%) and 18/123 (15%) of ERG positive cases, respectively. PTEN loss was significantly associated with time to PSA recurrence compared to no PTEN loss (Logrank test, p<0.001), however, ERG rearrangement vs. no rearrangement was not associated with time to PSA recurrence (p=0.26). Patients who exhibited ERG rearrangement and loss of PTEN had no significant difference in time to recurrence compared to patient with wild type ERG and loss of PTEN (p=0.97).
Conclusions: Our findings using FISH and aCGH technologies confirm a mutual co-operative role of ERG and PTEN in the pathogenesis of prostate cancer, particularly for homozygous PTEN deletion. ERG did not stratify outcome either alone or in combination with PTEN in this cohort.
Category: Genitourinary (including renal tumors)
Tuesday, March 5, 2013 9:30 AM
Poster Session III # 68, Tuesday Morning