Minimally invasive vs open radical prostatectomy in high-risk prostate cancer: comparing apples and pears?
Article first published online: 13 SEP 2013
© 2013 The Authors. BJU International © 2013 BJU International
Volume 112, Issue 6, pages 711–712, October 2013
How to Cite
Joniau, S., Tosco, L., Van Poppel, H. and Spahn, M. (2013), Minimally invasive vs open radical prostatectomy in high-risk prostate cancer: comparing apples and pears?. BJU International, 112: 711–712. doi: 10.1111/bju.12302
- Issue published online: 13 SEP 2013
- Article first published online: 13 SEP 2013
We have read with interest the paper of Pierorazio et al. . The authors aimed at retrospectively analysing pathological and short-term oncological outcomes (biochemical recurrence-free survival, BFS) of patients treated surgically for high-risk prostate cancer, defined as a PSA concentration of ≥20 ng/mL, clinical T-stage of ≥2c or Gleason score at biopsy of ≥8 . The paper presents the single-institution results of a highly respected tertiary referral centre, including a high number of cases treated with open retropubic radical prostatectomy (ORRP, 743 cases, 81.4%) or minimally invasive RP (MIRP) comprising robot-assisted laparoscopic RP (RALRP, 105 cases, 11.5%) and LRP (65 cases, 7.1%).
Surgery for high-risk prostate cancer should aim at achieving either oncological radicality or local debulking. This also entails an extended pelvic lymph node dissection (ePLND), not only for staging purposes, but also with a possible therapeutic intent . In the past decade, LRP and RALRP have been proposed as less invasive alternatives to ORRP. The main purpose of the introduction of these minimally invasive techniques is to achieve oncological and functional results comparable to (or better than) ORRP with minor operative impact .
The authors  conclude that MIRP delivers oncological results comparable to ORRP. They observed BFS rates at 3 years of 56.3%, 67.8% and 41.1% for ORRP, RALRP and LRP, respectively (P = 0.6). However, these results should be interpreted with caution, as the mean follow up in the groups was 3.2, 1.95 and 1.97 years, respectively. Further to this, although the groups seem to be matched, it appears that clear differences between them still exist. This disparity in preoperative criteria is apparent in biopsy Gleason score: 40.3% of patients in the ORRP group vs 34.6% and 30.7% in the RALRP and LRP, respectively (MIRP, 33%) had biopsy Gleason score ≥8. Such imbalances may be important at longer follow-up, as high-grade cancer at biopsy has consistently been shown to majorly influence cancer-related outcomes . Moreover, the definition of high-risk prostate cancer in the present series covers a large number of patients with clinical stage <cT2c (n = 812, 88.9% of the total population), which may translate into a possible overestimation of BFS in high-risk prostate cancer. Finally, Walz et al.  evidenced that the accumulation of high-risk risk factors directly correlate with worse BFS. In the present paper, 9% of ORRP vs 5.1% of MIRP had two to three high-risk factors, again introducing an imbalance between groups. The above issues reflect the observed pathological results with organ-confined prostate cancer in 33.3% and 40.6% after ORRP and MIRP, respectively.
In present series, the authors  state that positive surgical margins rates are equivalent at 29.4%, 34.3% and 27.7% respectively for ORRP, RALRP and LRP. Taking into consideration that surgeons performed nerve-sparing surgery whenever possible according to intraoperative findings, details about surgical technique and nerve-sparing approach should be provided, otherwise a real comparison it is not feasible. Importantly, an ePLND is recommended in high-risk prostate cancer  and it has been shown that the removal of <10 lymph nodes is associated with a very low staging accuracy . In the present series, a higher number of lymph nodes has been excised in ORRP vs MIRP with median numbers of 8 (0–37) vs 6 (0–39). A comparative evaluation between different techniques is difficult without an accurate description of the templates. The difference between the median number of resected lymph nodes and the respective ranges show that not all patients underwent the same ePLND. Moreover, the percentage of positive nodes in the ORRP and MIRP groups are 10.8% and 3.5%, respectively. This can probably be explained either by a different surgical template or by different aggressiveness of disease. In future series, a more accurate definition of ePLND templates and the use of validated nomograms in selecting candidates for ePLND  should be included in order to better understand the impact of surgical technique on pathological outcomes.
Recently, Joniau et al.  proposed a super-extended PLND template, as the risk of missing positive nodes by the standard ePLND is 13%. The higher percentage of positive nodes in the ORRP group justifies the more intense use of androgen-deprivation therapy compared with the MIRP groups.
In summary, also considering that treatment periods differed between robot-assisted surgery (2005–2011) and the other treatments (2002–2011), we agree with authors that selection bias may have influenced the results. For this reason, a propensity score matching  is required to control for all covariates that could have influenced treatment choice.
Nevertheless, the paper of Pierorazio et al.  represents an important study, which may prove to be decisive in focusing the debate on the importance of patient selection by validated prognostic instruments and the need for a standardisation of surgical technique to accurately compare surgical and oncological results.