Prostate cancer (PC) is the most common cancer in men in the United States, and is estimated to account for 21% of newly diagnosed cancer cases in men worldwidein 2016^[1]. Radical prostatectomy (RP) is the standard treatment for clinically localized PC. Since 1991 when laparoscopic radical prostatectomy (LRP) was first performed ^[2], LRP has undergone rapid development for its minimal invasiveness and favorable prognosis. Nevertheless, LRP remains a complex procedure and is associated with a long learning curve for the novices. In addition, when guided by a twodimensional (2D) imaging system, surgeons may lose depth perception and spatial orientation^[3-4] to makes such procedures as urethrovesical anastomosis more difficult.

To overcome the limitations of 2D visualization and space constraint, three-dimensional (3D) imaging systems was introduced in the early 1990s^[5]. In the stereotype of a 3D imaging system, however, the surgeons had to wear a heavy helmet with a headmounted display, which easily cause fatigue and dizziness and provided a poor image resolution ^[6-7]. Hence, 3D imaging systems were not further used in laparoscopic surgery. During the past 20 years, the development of 3D techniques fueled the rapid evolution of the 3D imaging system into its 4th generation, which brought drastic improvements compared with the first-generation 3D imaging systems ^[8]. Surgeons can now wear more ergonomic glasses to receive clear 3D images and safely perform laparoscopic surgeries ^[8-10], and compared with robotic-assisted laparoscopic radical prostatectomy (RALP), 3D LRP can be performed at a much lower cost ^[11].

Although 3D imaging systems are currently widely used in LRP, studies that evaluate the effectiveness of 3D imaging systems still remain scarce. In this study, we aimed to compare the perioperative data and functional and oncologic outcomes of patients with clinically localized PC receiving LPR with 3D and 2D imaging systems.

From February 2014 to January 2016, 72 consecutive patients with clinically localized PC received LRP performed by a single surgeon. The patients' perioperative data were collected and retrospectively analyzed, and their follow-up data were obtained by reviewing the medical records or call visits. Due to the retrospective nature of the study design, written informed consents from the patients were waived, but care was taken that the patients remained anonymous and their personal information was concealed in this study. This study was conducted with approval by the Ethical Committee of Nanfang Hospital, Southern Medical University.

LRP with a 2D imaging system was performed using a 10 mm 30° laparoscope with a 2D-HD Olympus^® camera, and LRP with a 3D imaging system was performed using a 10 mm and 30° lens doublechannel stereo laparoscope with 3D-HD Viking^® camera. During the operation, the surgeon wore polarized glasses to obtain 3D visualization on the 3D-HD Viking^® screen.

All the operations of LPR were performed by the same surgeon (Tan WL) following the same surgical procedure of laparoscopic extraperitoneal radical prostatectomy (LERP). Briefly, the surgical procedures were carried out first by inserting a self-made balloon device to expand the extraperitoneal space, and the bilateral pelvic lymph node were dissected in cases that had a preoperative serum prostate-specific antigen level of≥10 ng/mL and a Gleason score≥7, and/or where enlarged lymph nodes were found. The Retzius space was then expanded and an incision was made on the endopelvic fascia to dissociate the puboprostatic ligament followed by ligation of the dorsal venous complex (DVC). The bladder neck was transected, and the deferent ducts and seminal vesicles were isolated and dissected. After the incision of the Denonvilliers fascia, the prostate vessels were transected, and the nerve vessel bundles (NVB) were either transected or spared, followed by transection of the apical urethra and the DVC. Finally urethrovesical anastomosis was performed and potential anastomotic leakage was examined.

The perioperative data and functional and oncologic outcomes of the patients were collected and evaluated retrospectively. The baseline characteristics of the patients included age, body mass index (BMI), preoperative prostatic serum antigen (PSA) level and biopsy Gleason score. The perioperative data collected included operative time (from skin incision to skin closure), blood loss, days of drainage, hospital stay, days of catheterization and complications reported according to the updated Clavien classification ^[12]. Functional outcomes included urinary continence and potency data. According to the Incontinence Quality of Life (I-QoL)^[13], continence was defined as the use of no pad or one safety pad per day. We used the International Index of Erectile Function (IIEF-5) to assess the potency rate. The oncologic outcomes included pathological tumor stage, positive surgical margin (PSM), lymph node dissection (LND), positive lymph nodes (PLN) and biochemical recurrence (BCR, defined as PSA levels > 0.2 ng/mL on two consecutive occasions ^[14]). The patients were followed up for at least 6 months, and their functional and oncologic outcomes were assessed at 3 months and 6 months during the follow-up.

The quantitative data were tested using Student's t-test or Mann-Whitney test as appropriate. The qualitative data were analyzed using Chi-square test or Fisher's exact test. A two-tailed P value less than 0.05 was considered to indicate a statistically significant difference. All the data were analyzed using SPSS 13.0 software (IBM Corp., Armonk, NY, USA).

The characteristics of the patients were summarized in Tab. 1. There were no significant differences between the two groups (3D LRP vs 2D LRP) with regard to age, BMI, preoperative serum PSA level or biopsy Gleason score.

表 1(Table 1)

Table 1 Characteristic of patients at baseline Characteristics 3D LRP (n=36) 2D LRP (n=36) P Age (year, Mean±SD) 65.14±9.15 66.03±5.74 0.623 BMI (kg/m2, Mean±SD) 23.91±3.10 24±2.30 0.881 PSA (ng/mL, Mean±SD) 12.09±10.22 12.02±10.58 0.977 Biopsy Gleason score (n [%]) 6 14 (38.89%) 16 (44.44%) 0.111 7 20 (55.56%) 13 (36.11%) ≥8 2 (5.55%) 7 (19.45%) BMI: Body mass index; PSA: Prostatic serum antigen.

Table 1 Characteristic of patients at baseline

Tab. 2 showed the perioperative data of the two groups. Only the operative time and volume of blood loss differed significantly between the two groups (P < 0.001). The differences between the groups in the mean day of drainage, mean postoperative hospital stay, mean days of catheterization, overall rate and severity of complications, or incidence of minor complications (Clavien grade≤2) did not reach statistically significant levels. The total costs during hospitalization were similar between the two groups.

表 2(Table 2)

Table 2 Perioperative data in the two groups Variable 3D LRP (n=36) 2D LRP (n=36) P Operative time (min, Mean±SD) 167.72±26.42 218.11±35.96 < 0.001 Blood loss (mL, Mean±SD) 86.11±57.78 177.78±102.43 < 0.001 Days of drainage (day, Mean±SD) 6.33±2.29 7.33±3.20 0.132 Postoperative hospital stay (day, Mean±SD) 11.47±4.93 12±5.23 0.661 Days of catheterization (Mean±SD) Complication (Clavein grade) 25±6.78 25.39±6.88 0.81   Grade 1 Anastomosis leakage (n=2) Anastomosis leakage (n=3)   Grade 2 Transfusion (n=1), incision infection (n=1) Transfusion (n=1), pneumonia (n=1)   Grade 3a 0 0   Grade 3b 0 0   Grade 4a-b/5 0 0   Total (%) 11.11% (4) 13.89% (5) 0.722 Total cost (RMB Yuan, Mean±SD) 44155±11438 47046±17396 0.408

Table 2 Perioperative data in the two groups

The overall continence rates differed significantly between 3D and 2D groups within 3 months after the operation, but were comparable over a follow-up period of 6 months (Tab. 3). Twelve patients in 3D LRP group and 12 in 2D LRP group had an either unilateral or bilateral nerve-sparing procedure, and potency in these patients were examined at 3 and 6 months postoperatively. As shown in Tab. 3, the overall potency rates were similar between the two groups at both 3 and 6 months after the operation.

表 3(Table 3)

Table 3 Functional outcomes of between 2D and 3D LRP groups Outcome 3D LRP (n=36) 2D LRP (n=36) P Continence rate   Pre-operative 100% 100%   3 month follow-up 88.89% (32/36) 63.89% (23/36) 0.026   6 month follow-up 94.44% (34/36) 91.67% (33/36) 0.643 Nerve-sparing procedure   No 24 (66.67%) 24 (66.67%)   Unilateral 2 (5.56%) 3 (8.33%)   Bilateral 10 (27.77%) 9(25%) Potency rate (IIEF 5≥17)   Pre-operative 69.44% (25/36) 63.89% (23/36) 0.617   3 month follow-up 41.67% (5/12) 25% (3/12) 0.667   6 month follow-up 58.33% (7/12) 41.67% (5/12) 0.684

Table 3 Functional outcomes of between 2D and 3D LRP groups

The oncologic outcomes were listed in Tab. 4. Pathological examination of the surgical specimens did not reveal significant differences between the two groups in terms of pathological tumor stage or Gleason score. The overall PSM rates and PLN rates were both similar between the two groups. No significant difference was found in the BCR-free rate at 6 months between the two groups.

表 4(Table 4)

Table 4 Oncologic results of the patients in the two groups Outcome 3D LRP (n=36) 2D LRP (n=36) P Pathological tumor stage (n [%]) 1   T2 32 (88.89%) 33 (91.67%)   T3 4 (11.11%) 3 (8.33%)   T4 0 0 Pathologic Gleason score (n [%]) 0.487   6 9(25%) 12 (33.33%)   7 23 (63.89%) 18(50%)   ≥8 4 (11.11%) 6(16.67%) PSM (n [%]) 3 (8.33%) 5(13.89%) 0.708 LND (n) 29 27 PLN (n [%]) 3(10.34%) 1(3.7%) 0.612 BCR-free rate at 6 months (n [%]) 32 (88.89%) 31 (86.11%) 0.722 PSM: Positive surgical margin; LND: Lymph node dissection; PLN: Positive lymph nodes; BCR: Biochemical recurrence.

Table 4 Oncologic results of the patients in the two groups

In this study, we found significant differences in the outcomes of LRP for clinically localized PC using different imaging systems. 3D LRP was associated with a shorter operative time, a smaller volume of blood loss and a better early urinary continence outcome as compared with 2D LRP. The incidence of surgical complications, potency outcome and BCR-free rate, however, were not significantly different between the two groups.

The early 3D laparoscopic technique, with its poor image resolution and discomfort resulting from the heavy helmet ^{[6, 7, 15]}, failed to show much difference from the 2D systems ^{[16, 17]}and was therefore not used as a routine procedure^[18]. The recent advancement in 3D laparoscopic technology substantially improves the performance precision and hand-eye coordination in laparoscopy and offers a greater depth perception during operation while causing minimal dizziness of the surgeon^{[8, 19]}, which benefits not only the surgical novices to reduce their learning curves ^[20] but also experienced surgeons to facilitate the surgical performance^{[7, 21]}. Compared with the robot-assisted laparoscopic surgery systems with multi-angular device tips and 3D visualization^[22], which is especially useful in urologic operations^{[23, 24]}, laparoscopy assisted by a 3D vision system is associated with a much lower cost and offers a more economic option for the patients.

3D LRP significantly facilitates the performance of challenging procedure of urethrovesical anastomosis and has shown better outcomes than conventional 2D LRP ^{[10, 25, 26]}. Aykan et al^[10]] conducted a retrospective study of 2D LRP and 3D LRP in 95 consecutive patients and concluded that 3D LRP had a shorter operative time, less blood loss and a higher early continence rate than 2D LRP, which is consistent with our results. In a more recent study, Bove et al ^[25]compared 3D LRP against 2D LRP (43 patients vs 43 patients) and found similar outcomes in terms of the operative times, intraoperative blood loss and postoperative recovery of continence. However, another multicenter, open-label, randomized trial indicated that 3D LRP may have limited advantages over 2D LRP only in terms of shortened operative time^[26]; but considering the number of surgeons participated in this study (as many as 9) and their varying levels of experience with the procedures (the most experienced surgeon performed only 24 LRP operations), the validity of this conclusion needs to be carefully weighed against the compounding factors associated with the surgeons' experience.

In this study, all the LRP operations were performed by a single surgeon, which makes the results more reliable and less biased. Considering the rich experience of the surgeon with the two procedures long before this study had even begun, we believe that the difference in the outcomes between the two procedures were much more a result of the advancement of the surgical procedures in 3D LRP almost independent of the surgeon's surgical skills. In addition, the similar baseline characteristics between the two groups reduced the selection bias to the minimal and made the results more reliable.

This study also has some limitations. We included only 72 consecutive patients, and due to the retrospective nature of the study, the patients were not randomly assigned. Some of the follow-up data were obtained through call visits, which could cause recall bias. Furthermore, the duration of the follow-up was relatively short for evaluating the functional outcomes and BCR-free rates. Future randomized trial involving a larger sample size is needed to further confirm the results of this study.