Difference of opinion - Radical prostatectomy in metastatic prostate cancer: is there enough evidence? | Opinion: No.

Abstract

Despite an absence of level I data suggesting a survival benefit, interest in radical prostatectomy (RP) for patients with metastatic prostate cancer (PC) is rising (1). Tradi-tionally, RP has been reserved for clinically localized PC, and good outcomes have been demonstrated in this population (2). While both retrospective and observational studies have reported improved survival outcomes for patients with metastatic (M1) disease who undergo primary tumor treatment relative to androgen deprivation therapy alone (1, 3), prospective data – particularly for surgery – is sparse. It would be unwise, then, to pre-maturely extrapolate these results to patients with metastatic disease until the merits of such an approach are carefully considered.For the purposes of this discussion, we will consider metastatic disease to en-compass both clinically node positive (cN1) and traditional metastatic (cM1) disease (4). There are no randomized trials exploring the role of RP in the cN1 setting. Thus, we are limited to retrospective analyses to inform treatment decisions. Though the relevance of the distinction has been recently questioned (5), the majority of these studies are compri-sed of patients with occult nodal disease (i.e. clinically unapparent and discovered at the time of radical prostatectomy) rather than clinical node positive disease (i.e. > 1cm nodes identified on pre-operative imaging studies). A German group looked at patients with cT1-3, N1-2, M0 prostate cancer who underwent pelvic lymph node dissection (PLND) with androgen deprivation therapy (ADT, group 1) versus RP+PLND+ADT (group 2) (6). Biochemical recurrence (BCR)-free survival, overall survival (OS), and prostate cancer specific mortality (PCSM) favored performance of RP. Unfortunately, marked differences in stage and grade also favored the RP group, thereby limiting the generalizability of these results. Similarly, Grimm and colleagues evaluated patients with node-positive PC who either underwent RP+ADT or ADT alone and demonstrated improved BCR-free sur-vival and PCSM in the RP group (7). However, the group that was not selected for surgery had greater number of positive lymph nodes relative to the RP group, again highlighting selection biases inherent to this study design. In an attempt to account for these diffe-rences between groups, Ghavamian et al. matched 79 pN+ patients who underwent PLND and early adjuvant orchiectomy to 79 pN+ patients who underwent RP with PLND on the following parameters: number of positive nodes, clinical grade, clinical stage, patient