Technique

Although the steps of the radical prostatectomy itself are anatomically similar, there are many differences between the robotic and open techniques from start to finish. Beginning with patient positioning, open RRP is traditionally performed with the table flat or slightly flexed at the midline. By contrast, RALRP typically uses steep Trendelenburg (head-down) positioning to shift the abdominal contents for optimal exposure. Prolonged Trendelenburg positioning has been associated with head edema, increased intraocular pressure, and cardiopulmonary alterations, particularly in obese patients. In addition, insufflation with carbon dioxide during RALRP may lead to problematic hypercarbia and acidosis, particularly when the intra-abdominal insufflation pressure is increased in the event of troublesome bleeding. Overall, judicious monitoring of end-tidal carbon dioxide levels and controlled ventilation are important to avoid respiratory compromise during RALRP.

Other anesthetic issues also may differ between the procedures. For example, open RRP may be performed under general, epidural, or spinal anesthesia. Indeed, the senior author exclusively uses spinal anesthesia in the majority of patients. By contrast, general anesthesia is necessary for RALRP, potentially increasing the overall “invasiveness” of the procedure.

Another readily apparent difference is the skin incision. Open RRP can be performed through a midline or Pfannenstiel 8-cm incision, as has been used by the senior author since August 2007 with improved cosmetic results. By contrast, RALRP involves 5 to 6 smaller incisions to accommodate trocars ranging from 5 to 12 mm in diameter. Because prostatic morcellation is not performed, one of these incisions must be extended sufficiently at the end of the prostatectomy to permit intact specimen removal. The relative cosmetic appeal of multiple smaller incisions higher on the abdomen, compared with a single somewhat larger incision lower on the abdomen, is a matter of personal preference.

The extent and yield of pelvic lymphadenectomy (PLND) may also differ between open RRP and RALRP. Among 1278 consecutive patients treated at University of California—San Francisco, Cooperberg and colleagues reported that PLND was less frequently performed during RALRP compared with open RRP (31.8% vs 47.8%, respectively, P <.01). Moreover, after controlling for year of surgery and CAPRA risk score (Cancer for Prostate Risk Assessment), RALRP remained significantly associated with a lower odds of PLND (odds ratio [OR] 0.18, 95% confidence interval [CI] 0.11–0.32). In addition, the mean lymph node yield was 9.3 in RALRP versus 14.4 in open RRP ( P <.01). Due to stage migration and the declining rate of lymph node metastases at diagnosis, the impact of these differences in nodal yield may be of lesser magnitude than in the past. Nevertheless, the risk of occult nodal metastases remains substantial for patients in the high-risk group, necessitating adherence to the same high surgical standards. Of note, a technique for more extended PLND during RALRP was recently described in 99 patients, involving more cephalad trocar placement. Using this method, Feicke and colleagues reported a mean lymph node yield of 19, although it required an additional 51 minutes of operative time to complete.

Another important difference between RRP and RALRP is the timing of PLND during the operation. In the classic open RRP, PLND is done at the beginning of the case, providing the opportunity to examine the lymph nodes (by palpation or, less commonly, with frozen sections) prior to prostatectomy. By contrast, PLND is typically reserved until after prostatic removal during RALRP. At that point, the excised lymph nodes may be removed either through the assistant port, in the same Endocatch bag with the prostate, or by using separate specimen bags. This action precludes the ability to examine the lymph nodes grossly or histologically before prostatectomy.

Another difference between RRP and RALRP is the technique used for the vesicourethral anastomosis. During open RRP, the anastomosis uses approximately 4 to 8 interrupted sutures. By contrast, the vesicourethral anastomosis during RALRP is typically performed in a running fashion with a single knot, as described by Van Velthoven and colleagues. The long-term impact of this technical modification on continence requires further study.

A final major difference is that open RRP is entirely an extraperitoneal procedure, whereas RALRP can be performed through a transperitoneal or extraperitoneal approach. However, the limited working space makes extraperitoneal RALRP more technically challenging. In consequence, in contrast to open RRP, most RALRP are currently performed transperitoneally. Despite its designation as “minimally invasive surgery,” one could argue that violating the peritoneal cavity makes transperitoneal RALRP considerably more invasive than open RRP.

The issue of relative “invasiveness” was studied from a unique perspective by Jurczok and colleagues by comparing levels of acute phase reactants (as a marker for the systemic response) before, during, and after laparoscopic versus open RRP. Overall, they found no difference in C-reactive protein (CRP), serum amyloid A, interleukin (IL)-6, and IL-10 between the 2 groups at any time point. The investigators concluded that “the so-far assumed less invasiveness of laparoscopic radical prostatectomy is not objectively supported by the data from this study.” By contrast, Fracalanza and colleagues reported significantly higher blood levels of CRP, IL-6, and lactate after open RRP than RALRP, although still within the normal range. Based on these findings, the investigators concluded that “RALRP induces lower tissue trauma than RRP.” Although the clinical significance of these minor variations in acute phase reactants is unclear, additional study is warranted into the physiologic response to open radical prostatectomy and RALRP at both the systemic and tissue level.

There are several major intraoperative advantages of RALRP, including 3-dimensional magnified visualization and the computer-assisted elimination of intention tremor. RALRP is also less physically and technically demanding than pure laparoscopic radical prostatectomy. Another well-documented advantage of RALRP is lower average intraoperative blood loss than open RRP, which has been verified in numerous studies. In one series, the average estimated blood loss (EBL) was 200 mL in RALRP compared with 550 mL in open RRP. In a matched analysis by Rocco and colleagues, EBL was 200 mL in RALRP versus 800 mL in RRP ( P <.0001). Farnham and colleagues similarly reported an average EBL of 191 and 664 mL in RALRP and RRP, respectively. Nevertheless, there was no difference in the transfusion rate in their series. Thus, the long-term clinical significance of such differences in intraoperative blood loss is unclear.

Cost

The cost of a surgical intervention is a complex issue to study. Not only is it affected by the cost of the surgical equipment but also by surgeon-related factors, operative time, length of hospital stay, hospital volume and charges, lost work time, rates of hospital readmission and complications, marketing costs, and many other factors.

To start, the initiation of a robotic prostatectomy program involves considerable expense, including the cost to purchase the robotic system (approximately $1.5 million), as well as annual maintenance fees of $100,000 to $200,000, and at least $1000 for disposable instruments per case. Furthermore, the initial learning curve has been associated with substantial expense. Using a theoretical model with different rates of surgeon improvement, Steinberg and colleagues estimated this cost to range from $95,000 with a learning curve of 24 cases, to $1,365,000 for a learning curve of 360 cases. A major contributory factor to this increased cost clearly is the significantly longer operative times during the early learning curve, which have been well documented in numerous studies.

In a recent cost-benefit analysis, Steinberg and colleagues estimated that at least 78 cases are needed per year to maintain profits following the purchase of a robotic system. The investigators concluded accordingly that surgical volume at a particular center and patient demand may represent important determinants for the profitability of implementing a RALRP program. To minimize the costs associated with RALRP would involve a high-volume center with an experienced surgeon, short length of stay, and the absence of complications, readmissions, or secondary therapy.

A few studies have demonstrated significantly shorter length of hospital stay with RALRP than RRP, which would help to balance the aforementioned costs. Nevertheless, many other studies reporting similar clinical care pathways and length of hospitalization for RALRP versus RRP would negate this advantage.

In the words of Emanuel, “novelty cannot be equated with benefit.” Particularly in light of the current economic crisis and discussions regarding health care reform, a critical assessment of new technology is warranted, including RALRP.

Cost

The cost of a surgical intervention is a complex issue to study. Not only is it affected by the cost of the surgical equipment but also by surgeon-related factors, operative time, length of hospital stay, hospital volume and charges, lost work time, rates of hospital readmission and complications, marketing costs, and many other factors.

To start, the initiation of a robotic prostatectomy program involves considerable expense, including the cost to purchase the robotic system (approximately $1.5 million), as well as annual maintenance fees of $100,000 to $200,000, and at least $1000 for disposable instruments per case. Furthermore, the initial learning curve has been associated with substantial expense. Using a theoretical model with different rates of surgeon improvement, Steinberg and colleagues estimated this cost to range from $95,000 with a learning curve of 24 cases, to $1,365,000 for a learning curve of 360 cases. A major contributory factor to this increased cost clearly is the significantly longer operative times during the early learning curve, which have been well documented in numerous studies.

In a recent cost-benefit analysis, Steinberg and colleagues estimated that at least 78 cases are needed per year to maintain profits following the purchase of a robotic system. The investigators concluded accordingly that surgical volume at a particular center and patient demand may represent important determinants for the profitability of implementing a RALRP program. To minimize the costs associated with RALRP would involve a high-volume center with an experienced surgeon, short length of stay, and the absence of complications, readmissions, or secondary therapy.

A few studies have demonstrated significantly shorter length of hospital stay with RALRP than RRP, which would help to balance the aforementioned costs. Nevertheless, many other studies reporting similar clinical care pathways and length of hospitalization for RALRP versus RRP would negate this advantage.

In the words of Emanuel, “novelty cannot be equated with benefit.” Particularly in light of the current economic crisis and discussions regarding health care reform, a critical assessment of new technology is warranted, including RALRP.

Perioperative outcomes

With respect to the perioperative period, there is conflicting evidence regarding differences in length of hospital stay (LOS) and convalescence between RRP and RALRP. Among Medicare beneficiaries treated from 2003 to 2005, Hu and colleagues reported a significantly longer LOS following open RRP than laparoscopic approaches (4.4 vs 1.4 days, P <.0001). However, the LOS in this study may not reflect other contemporary RRP series.

At the authors’ institution, the same clinical care pathway is used for open RRP and RALRP. Following an uncomplicated open or robotic procedure, patients are given a clear liquid diet and patient-controlled anesthesia (PCA) on the night of surgery. On the first postoperative day, the diet is gradually advanced, the PCA is discontinued, and frequent ambulation is encouraged. The timing of discharge (ie, postoperative day 1 or 2) is then determined based on patient-specific factors (eg, tolerating a regular diet, pain control, ambulating) and physician preference, but not by surgical approach. Nelson and colleagues similarly reported a common clinical care pathway and LOS for RRP and RALRP patients treated at Vanderbilt. Prior studies have also shown similar postoperative pain after RRP and RALRP. For example, Jurczok and colleagues reported no difference in morphine equivalents between patients treated by the 2 approaches.

After discharge, patients at the authors’ institution treated by RRP and RALRP are provided with the same instructions regarding the return to work and resumption of physical activity. The duration of catheterization is determined based on patient-specific factors and physician discretion, rather than surgical approach.

With regard to complication rates, published studies have varying results when comparing RRP to RALRP. In the study by Hu and colleagues, RALRP was associated with a significantly lower odds of short-term perioperative complications (OR 0.7, 95% CI 0.6–0.9), but a significantly higher frequency of strictures (OR 1.4, 95% CI 1.04–1.9). As expected, in both open RRP and RALRP, the risk of perioperative complications and strictures has been shown to decrease with surgeon experience.