Propensity-Matched Analysis of Open Versus Robotic Primary Retroperitoneal Lymph Node Dissection for Clinical Stage II Testicular Cancer

Open RPLND (O-RPLND) has been the gold standard technique for primary dissection of the retroperitoneal lymph nodes since the early 1900s. The first minimally invasive RPLND was described in 1992 by Rukstalis and Chodak. They performed a laparoscopic RPLND at the University of Chicago. However, the widespread adoption of laparoscopic RPLND has been hampered by several limiting factors including, a steep learning curve, the complexity of the operation that is hard to replicate outside of tertiary referral centres, the lack of reproducibility in the community setting without expert laparoscopic surgeons and the uncertain oncologic efficacy of the procedure.

The first robotic RPLND (R-RPLND) was performed in 2006 by Davol and Rukstalis. The robotic approach has allowed greater dissemination due to several advantages of the robotic platform including enhanced vision with 3D optics, magnification, improved dexterity with the Endowrist of the DaVinci robotic platform, and ergonomic advantages for the surgeons that previously had to stand for 8-10 straight hours to perform these operations. The advent of minimally invasive techniques has been associated in retrospective series with lower estimated blood loss (EBL), shorter length of stay (LOS), less postoperative pain, faster recovery and return to work, superior cosmetic results, and lower risk of incisional hernia. Notwithstanding, the robotic approach is associated with increased costs due to a longer operative time and instrument expenditures.

There is a lack of prospective randomized trials comparing R-RPLND to O-RPLND, controlled data is certainly needed. We performed a propensity score matched analysis to limit the impact of selection bias on survival and perioperative outcomes and allow it to approximate results that might be observed in a randomized trial.

Our study used the prospectively maintained eCancerTestis database and we conducted a retrospective chart review of all patients who underwent primary RPLND at the Princess Margaret Cancer Centre between 1990 to 2022. We identified a total of 178 patients who underwent primary RPLND. Of these 137 underwent O-RPLND and 41 R-RPLND. Before PSM, the groups were not balanced, and patients treated with O-RPLND had a larger median retroperitoneal mass size (3.0 vs. 1.7cm), and were less likely to have had elevated tumour markers before pRPLND (22% vs. 7%) We decided to use the following variables for propensity-score matching:

Age at RPLND
Obesity (classified as BMI ≥ 30 kg/m2)
Histology (Seminoma/NSGCT)
Clinical stage (de novo CSII or CSI and relapsed as CSII)
Preoperative retroperitoneal mass size (cm)
Embryonal carcinoma % (orchiectomy pathology)
Tumor marker status pre-RPLND
Location of retroperitoneal mass (Paracaval, interaortocaval, paraaortic)

Our primary endpoint was to compare the relapse-free survival (RFS) of both surgical approaches, hypothesizing there should be no difference between open and robotic RPLND. The secondary endpoints were to evaluate perioperative outcomes including operative time, LOS, EBL, and surgical complications.

We have described in detail in our article our surgical technique for R-RPLND. The Princess Margaret Cancer Centre philosophy is that while performing R-RPLND we should aim to replicate the open approach at every point in the operation, without sacrificing the oncological safety of the procedure. We perform full bilateral template R-RPLNDs with bilateral nerve sparing when feasible, furthermore, all lumbar arteries and the inferior mesenteric artery (IMA) when not directly involved with disease are preserved. Lastly, all RPLNDs were performed by two expert urologic oncologists each having performed over 500 RPLNDs (MASJ and RJH), and all robotic cases were performed by one surgeon (RJH).

After propensity score matching 26 patients were matched in the R-RPLND (26/41) to 38 in the O-RPLND (38/137) group, and the groups were well balanced.

We found no significant differences in RFS between open and R-RPLND at a median follow-up of 23.5 months (HR 0.65, 95% CI 0.07-6.31, p= 0.7097). Of note, among the 4 relapses in the cohort, none were “in field” of the dissection.

R-RPLND was associated with:

Lower blood loss (median 200 vs. 300cc, p=0.03)
Reduced length of stay (median 1 vs. 5 days, p<0.0001)
Longer operative time (median 8.8 vs. 4.3hrs, p<0.0001) *Operative time decreased as experience was gained, but this is also likely offset by an increase in complexity of cases performed robotically as more experience was gained.
No significant differences in the:
Complications (p=0.8141) or Grade III complications (p=0.7659)
Lymph node yield (median 31 vs. 34, p=0.4356)
RPLND pathology (p=0.6637)

Our study is the first to compare primary open with robotic RPLND using a propensity score matching system to adjust for potential unbalanced factors at baseline. It is not free of limitations including but not limited to small sample size, the presence of unmeasured confounding variables, selection bias, differences in surgical technique, post-operative management, and pathologic processing. However, the strengths of our study are the inclusion of only CS II patients undergoing pRPLND, the use of a nerve-sparing bilateral template, and the avoidance of adjuvant chemotherapy.

To conclude our study showed that primary robotic RPLND offers low morbidity and improved perioperative outcomes while maintaining the oncological efficacy observed with the open approach and it should be endorsed by clinical practice guidelines provided it is carried out by surgeons at high-volume centres for testis cancer with experience in RPLND.

Written by: Julian Chavarriaga MD, & Robert J. Hamilton MD, MPH, FRCSC

Division of Urology, Department of Surgical Oncology, Princess Margaret Cancer Centre, University Health Network and the University of Toronto, Toronto, Ontario, Canada

Read the Abstract

Login