Prognostic value of insulin‐like growth factor II mRNA binding protein 3 in patients treated with radical prostatectomy

Study Type – Prognosis (case series) Level of Evidence 4 What's known on the subject? and What does the study add? Insulin‐like growth factor II mRNA binding protein 3 (IMP3) is associated with poor outcomes in a variety of malignancies. The role of IMP3 in protate cancer remains poorly understood. IMP3 expression was associated with features of aggressive biology and aggressive prostate cancer recurrence after surgery. Although IMP3 is differentially expressed in patients with features of biologically aggressive prostate cancer, it does not have independent prognostic value in patients treated with RP. OBJECTIVE To evaluate the association of insulin‐like growth factor II mRNA binding protein 3 (IMP3) with pathological features and outcomes in patients treated with radical prostatectomy (RP). PATIENTS AND METHODS Immunohistochemical staining for IMP3 was performed on archival tissue microarray specimens from 232 consecutive patients treated with RP for clinically localized disease. None of the patients received neoadjuvant or adjuvant radiation or hormone therapy. IMP3 expression was histologically categorized as normal or abnormal. Disease recurrence was classified as aggressive if metastases were present, post‐recurrence prostate‐specific antigen (PSA) doubling time was less than 10 months, or if the patients failed to respond to salvage local radiation therapy. RESULTS The median follow‐up was 69.8 months (interquartile range [IQR]: 40.1–99.5). IMP3 expression was abnormal in 42 (18.1%) of 232 patients. IMP3 expression was associated with extracapsular extension ( P = 0.020), seminal vesicle invasion ( P = 0.024), lymphovascular invasion ( P = 0.036) and a high pathological Gleason score ( P = 0.009). The 5‐year PSA recurrence‐free survival for IMP3‐negative patients was 83% (standard error [SE]= 3) vs 67% (SE = 8) in IMP3‐positive patients (log‐rank test, P = 0.015). In a multivariable analysis that adjusted for the effects of surgical margins, extracapsular extension and seminal vesicle invasion, PSA (hazard ratio [HR]: 1.04, P = 0.013), lymph node metastasis (HR: 16.7, P < 0.001) and a high pathological Gleason score (HR 4.3, P = 0.008) were significantly associated with PSA recurrence‐free survival, whereas IMP3 expression was not ( P = 0.11). Similarly, IMP3 expression was only associated with aggressive recurrence (HR 3.2, P = 0.006). CONCLUSION IMP3 expression is abnormal in approximately one‐fifth of prostate cancers. Although IMP3 is differentially expressed in patients with features of biologically aggressive prostate cancer, it does not have an independent prognostic value in patients treated with RP.