Comparison of Extraperitoneal and Transperitoneal Pelvic Lymph Node Dissection During Minimally Invasive Radical Prostatectomy

Introduction

P elvic lymph node dissection (PLND) at the time of radical prostatectomy (RP) has prognostic and potential therapeutic benefits in men with prostate cancer. ^{1 –3} Which men benefit from PLND and the appropriate PLND template (limited, standard, or extended) has been debated extensively in the urologic literature. ^1,4 Large open RP series, however, have demonstrated that more extensive PLNDs result in higher lymph node yields (LNYs) as well as higher rates of lymph node invasion after controlling for clinical variables but are associated with higher complication rates. ^1,5,6

With the widespread application of minimally invasive RP (MiRP), questions have been raised regarding the technical feasibility of performing a routine PLND either laparoscopically or with robot assistance. ⁷ While data exist regarding the feasibility and safety of transperitoneal PLND, data on extraperitoneal, minimally invasive PLND are more scarce. The present study directly compares standard extraperitoneal PLND and transperitoneal PLND with regard to LNY, percent lymph node involvement (%LNI), and complication rates during MiRP.

Patients and Methods

Between 2001 and 2010, 914 patients underwent MiRP (laparoscopic or robot-assisted) by one surgeon (CPP) transperitoneally (five or six port approach) or extraperitoneally (four or five port approach). The 361 who did not receive PLND were excluded from analysis. Preoperatively, patients were classified into low-, intermediate-, and high-risk categories in accordance with D'Amico risk categories. ⁸ Low-risk patients generally received a limited (external iliac) bilateral PLND consisting of all fibrofatty and lymphatic tissue inferior to the bifurcation of the common iliac artery, bound inferiorly by the femoral canal, laterally by the pelvic side wall, superiorly by the femoral vein, and inferomedially by the obturator nerve. Intermediate- and high-risk patients generally received a standard bilateral PLND consisting of the above external iliac template, plus all fibrofatty and lymphatic tissue in the obturator fossa, both deep and proximal to the obturator nerve up to the hypogastric artery. Pelvic lymph nodes were submitted to the pathology department for permanent section as right and left pelvic lymph nodes.

Patients were divided into two groups based on operative approach (transperitoneal [2001–2010] vs extraperitoneal [2005-2010] MiRP). Extraperitoneal MiRP became the predominant form of surgery by 2006, although transperitoneal procedures continued to be performed throughout the study period in select cases, particularly in men with previous bilateral hernia repairs and/or umbilical hernia repairs, and men with narrow pelvises. Patient risk stratification did not play a conscious role in selection of the trans- or extraperitoneal approach.

The primary surgeon's prospectively established database approved by the Institutional Review Board was retrospectively queried for clinical, pathologic, and operative data. LNY, %LNI, operative time, estimated blood loss (EBL), and perioperative complications (Clavien grade II–V) were compared between groups. LNY and %LNI were further compared between groups in men in similar risk categories. In an attempt to minimize experience bias, the last 100 extraperitoneal cases were compared with the last 100 transperitoneal cases with respect to LNY and %LNI among men with intermediate- and high-risk disease. Finally, oncologic outcomes were studied in men who had positive lymph nodes at lymph node dissection.

Statistical analysis was performed via STATA version 11 using the two-sample t test and chi-square test where appropriate. Log transformation was performed where appropriate. Statistical significance was defined as P<0.05.

Results

Overall, 175 patients had a Montsouris-style transperitoneal MiRP+PLND, 8.6% of these using a robot-assisted approach. ⁹ Three hundred and seventy-eight patients had extraperitoneal MiRP+PLND as per Rassweiler, 3.2% by a robot-assisted approach. ¹⁰ Demographic data including age and race were similar between transperitoneal and extraperitoneal groups. Body mass index was significantly higher in the extraperitoneal group. Comparison of preoperative data demonstrated a significantly higher percentage of patients with low-risk disease in the extraperitoneal group (51.2% vs 37.5%, P=0.003). In addition, the extraperitoneal group had a significantly lower percentage of patients with intermediate-risk disease (39.8% vs 52.3%, P=0.006) (Tables 1, 2).

Mean LNY was significantly higher overall in the extraperitoneal group (6.5 vs 5.3, P=0.002) (Table 3). When stratified by risk category, LNY was significantly higher with extraperitoneal surgery in low-risk patients only. There was no difference in LNY in intermediate/high-risk patients by either the extraperitoneal or transperitoneal approach. There was no difference in LNY between laparoscopic or robot-assisted MiRP, P=0.41. %LNI was not significantly different between groups regardless of risk category (Tables 2, 3).

Assessment of the last 100 consecutive extraperitoneal and 100 consecutive transperitoneal cases demonstrated intermediate-high–risk disease in 48% and 65% of cases, respectively. Within these subgroups, mean standard PLND LNY in the extraperitoneal group and transperitoneal group was 7.04 (range 1–26) and 5.74 (range 1-14), respectively (P=0.156). LNI was observed in 4.2% of the extraperitoneal group and in 3.1% of the transperitoneal group (P=0.755).

Overall, eight patients had positive lymph nodes at MiRP—three from the transperitoneal MiRP group and five from the extraperitoneal MiRP group. In patients with lymph node metastases, 3/30 (10%) nodes were positive in the transperitoneal cases and 20/51 (39.2%) nodes were positive in the extraperitoneal group (P=0.37). Of these eight patients, six experienced biochemical recurrence within 12 months, one chose immediate adjuvant hormonal therapy, and one has remained without evidence of biochemical recurrence for more than 1 year. This last patient had a lymph node density of 10% (1/10) and Gleason 4+5=9 pT_3a disease with negative margins; he has elected surveillance for the time being.

EBL and complication rates were comparable between operative approaches. Average operative time was significantly longer for the transperitoneal approach (3.8 h vs 2.8 h, P=0.004) (Table 4), likely reflecting that the majority of transperitoneal operations were performed earlier in the surgeon's experience. There were no Clavien V (death) complications in either group. Notable complications in the transperitoneal group included three wound infections, three ileus, and five thromboembolic events. The extraperitoneal group had a similar distribution of complications including three ileus, four wound infections, and three thromboembolic events. No perioperative transfusions were needed in either group, and a symptomatic lymphocele did not develop in any patient. Complications did not differ statistically between groups receiving limited vs standard PLND.

Discussion

Oncologic adequacy must be demonstrated when applying new technology to cancer for which effective therapy already exists. Robotics and laparoscopy allow for less invasive access to the prostate, but there is concern that visualization and access to the pelvic lymph nodes may be compromised, particularly during extraperitoneal MiRP, because of the angles involved and the linearity of the laparoscopic instruments.

The present study demonstrates that limited and standard-template PLND can be effectively and safely performed both transperitoneally and extraperitoneally during laparoscopic or robot-assisted MiRP, particularly in the men who may most benefit from PLND. Specifically, there was no significant difference in standard-template LNY between approaches in men with intermediate- or high-risk disease, even when procedures performed later in the series were compared with each other (in an effort to account for increasing surgeon experience with each technique). Furthermore, we did not find any differential effect on EBL or complication rate depending on approach. To our knowledge, this is one of the first reports directly comparing routine transperitoneal and extraperitoneal PLND during MiRP in a consecutive series.

Both the role and extent of PLND at the time of RP have been debated in the urologic literature. Studies have demonstrated that more extensive lymph node dissections yield higher lymph node counts, and higher LNYs are associated with increased %LNI. ^1,6 These findings have led to the general conclusion that men with intermediate- and high-risk disease derive prognostic benefit from more extensive dissections.

Furthermore, a thorough PLND may also have therapeutic benefit. Pagliarulo and associates ¹¹ and Masterson and colleagues ³ demonstrated improved outcomes in patients without nodal metastases who received a more extended PLND. This is in contrast to DiMarco and coworkers ¹² who found no association between prostate cancer outcomes and LNY in patients without nodal metastases. ¹² Similarly, Allaf and associates ¹ demonstrated improved biochemical recurrence-free survival in patients with positive lymph nodes and lymph node density less than 15%. ¹

Although there are no definitive recommendations and no evidence of improved overall survival regarding PLND for prostate cancer, some benefit appears to exist, particularly for patients with higher risk disease states. Nevertheless, the increase in complications attributable to extended PLND suggests that proper risk stratification makes sense before considering extended template dissection. ⁵

Several studies have compared extraperitoneal and transperitoneal laparoscopic and robot-assisted MiRP. Initial studies have demonstrated similar oncologic and functional outcomes between transperitoneal and extraperitoneal laparoscopic RP. ^{11,13 –17} For example, retrospective studies by Porpiglia and coworkers ¹⁸ and Eden and colleagues ¹⁹ demonstrated faster operative time and increased recovery of continence in patients undergoing an extraperitoneal approach. ^18,19 Ruiz and associates ¹⁷ demonstrated equivalent immediate oncologic outcomes including pathologic stage and rates of positive surgical margins in men undergoing either transperitoneal or extraperitoneal laparoscopic RP. Similarly, Joseph and colleagues ²⁰ described successful application of extraperitoneal, robot-assisted laparoscopic prostatectomy. These studies, however, did not specifically assess the feasibility or adequacy of concomitant minimally invasive PLND.

Extended and standard PLNDs have been described using a minimally invasive, transperitoneal approach. Katz and coworkers ¹⁶ found that both standard and extended PLND could be performed via a laparoscopic transperitoneal approach with an acceptable side effect profile. Similarly, Eden and associates ²¹ demonstrated an advantage of transperitoneal extended PLND over extraperitoneal standard PLND with regard to LNY and %LNI, which is not surprising because it is the templates rather than the approaches that were compared ²¹ ; no direct comparison was indeed made between similar nodal dissection templates attempted transperitoneally or extraperitoneally.

Finally, Wyler and colleagues ²² did demonstrate that extended PLND could be performed either transperitoneally or extraperitoneally without a difference in nodal yield, although their series contained only 15 extraperitoneal cases. These reports and some expert opinion suggest that an extraperitoneal MiRP allows for acceptable standard-template PLND, but we know of no sizeable studies that have directly compared standard PLND template yields between transperitoneal and extraperitoneal MiRP. This is an important question, given the prevailing opinion that extraperitoneal MiRP is more difficult and potentially limits access to the pelvic lymph nodes.

There are several limitations to our study. We acknowledge its retrospective, nonrandomized nature, with inherent weaknesses typical to such a study design. The decision of the operative approach was based on surgeon preference, which likely resulted in some selection bias. Most measured comparisons between the transperitoneal and extraperitoneal groups were, in fact, similar, however. Robot-assisted MiRP cases represent a smaller proportion, and with shorter follow-up, because access to a robot came later in the chronology of the study. The lower percentage of cases performed by a Da Vinci-assisted approach could limit the applicability of the current study given the case penetration of robotics in United States. However, we noted no differences in nodal yield by template whether or not a robot was used.

The data presented are from a highly specialized minimally invasive surgeon whose initial cases were transperitoneal and whose later cases were increasingly extraperitoneal. Therefore there may be experience bias: Increasing experience may have resulted in greater nodal yields even within a limited PLND template, possibly explaining why nodal yields were higher in low-risk patients with the extraperitoneal approach. We tried to mitigate such a bias by comparing cases that were done after the surgeon had performed many with either approach; ie, by comparing the most recent transperitoneal and extraperitoneal cases. These analyses, performed on men who received PLND and who had intermediate and high-risk disease, demonstrated no significant difference in LNY based on surgical approach.

Finally, mean LNY were lower than in some comparable published series, which may in part be explained by differences in counting lymph nodes between institutions, in part by the fact that lymph nodes were sent as one specimen from each side only, and in part by the fact that no patients received extended PLND, which would have boosted mean series yields.

Conclusion

The extraperitoneal MiRP approach does not appear to compromise the oncologic efficacy or safety of a standard PLND compared with a similar template attempted during transperitoneal MiRP. A limited or standard template extraperitoneal PLND can be safely and effectively performed either laparoscopically or with robot assistance. Further studies are needed to determine the feasibility of performing an extraperitoneal minimally invasive PLND in an extended template fashion.