Hand-Assisted Laparoscopic Versus Robot-Assisted Laparoscopic Partial Nephrectomy: Comparison of Short-Term Outcomes and Cost

Introduction

T he 2009 AUA Clinical Guidelines for the management of clinical T1 renal masses cite partial nephrectomy (PN) as the preferred standard of care for select small renal masses (SRMs). ¹ Many urologists have shifted toward minimally invasive approaches to PN, such as laparoscopic partial nephrectomy (LPN), in an attempt to offer improved convalesance and lessen morbidity over open partial nephrectomy (OPN). In the largest comparison of LPN to OPN, LPN was found to offer the advantages of less operative time, decreased operative blood loss, and a shorter hospital length of stay (LOS). ² However, in the same study, LPN was associated with longer warm ischemia time (WIT), more postoperative complications (particularly urologic), and an increased number of subsequent procedures. Further, patients offered LPN were more likely to have smaller more peripheral tumors as compared with OPN. Clearly, these differences highlight the technical difficulty of LPN. In an attempt to overcome the technical challenges of LPN, two adjuncts have emerged: hand-assisted laparoscopic partial nephrectomy (HALPN) and, more recently, robot-assisted laparoscopic partial nephrectomy (RALPN).

Many publications have extolled the benefits of both of these procedures. While the results are encouraging, several questions remain in the literature. First, the literature currently lacks well-designed and standardized studies comparing RALPN to LPN. Second, few studies have performed an adequate cost analysis comparison of LPN and RALPN. Assuming the significant cost of the da Vinci surgical robot, such a comparison is warranted. Third, to our knowledge there are no studies directly comparing RALPN to HALPN. Herein, we sought to evaluate the outcomes of HALPN and RALPN performed at our institution in a standardized fashion, with subsequent cost analysis comparison.

Methods

After institutional review board approval, the records of all HALPN and RALPN over a 4-year period were reviewed (2006–2010). All cases were performed by two fellowship-trained endourologists with extensive prior experience with robot-assisted laparoscopic urologic surgery. Patient age, body mass index (BMI), anesthesia society association (American Society of Anesthesiologists [ASA]) score, estimated blood loss (EBL), radiographic tumor size, histopathology, operative and room times, hospital LOS in days, and complications were reviewed. Pre- and postoperative creatinine (Cr) and hemoglobin (Hgb) values were used to assess a change in both groups. Preoperative imaging for all patients was reviewed and a nephrometry radius, endo/exophytic, nearness to collecting system, anterior/posterior, lines of polarity (RENAL) score (www.nephrometry.com) was calculated. All patient charts were reviewed for amount and type of narcotic administration postoperatively. These values were standardized using morphine equivalents. ³ Complications were graded using the Dindo–Clavien grading system. ⁴ Student's t-test and Fischer's exact test were utilized for statistical comparison of all parameters and proportions, respectively, with p<0.05 deemed as significant.

HALPN was performed at our institution in similar fashion to the technique described by Strup et al. ⁵ Often, we employed the use of a Gel-port laparoscopic system hand port (Applied Medical, Rancho Santa Margarita, CA), two 12-mm trocars, and occasionally a third 5-mm trocar. Usually a Harmonic Ace Scalpel (Ethicon Endo-Surgery, Inc., Cincinnati, OH) was used to aid in dissection, though an electrocautery-enabled laparoscopic scissor may be used instead. Upon surgeon discretion, renal hilar vessel clamping may be obtained via laparoscopic bulldog clamps or omitted. If hilar vessel clamping is to be omitted, in-situ manual compression is used for hemostasis during extirpation and renorrhaphy.

Gettman et al described the technique for RALPN, and it is similarly performed at our institution. ⁶ We routinely use the da Vinci S Robot (Intuitive Surgical Systems, Sunnyvale, CA) with a four-arm technique, though the fourth arm may be omitted per the surgeon's discretion. Often a 12-mm and a 5-mm trocars are inserted for the bedside assistant to use. Hilar vessel clamping with bulldogs, extirpation, and hemostasis are performed in a similar manner to HALPN. At the surgeon's discretion, usually for exophytic tumors, a Simon-Pole clamp may be used instead of renal hilar vessel clamping. Renohrrhaphy, nephropexy, and closed suction drain placement are performed in similar fashion to HALPN.

Postoperatively, a complete blood cell count and chemistry are performed on POD#1. Serial hematocrits were performed at regular intervals based on hemodynamic and lab stability. Diet is advanced on POD#1 and the patient is discharged when the patient has demonstrated adequate pain control, tolerance of regular diet, and stable hematocrit and renal function.

The estimated total cost of each procedure was calculated based on operative room (time cost of room, anesthesia time cost, and supply) and hospital stay cost. Operative room time cost, anesthesia time cost, and supply costs were obtained from hospital administration. Robotic cost and maintenance contract costs were obtained from hospital administration and represent a value that may vary based on institution or location. Hospital room costs (for in-patient hospital stay) and pharmacy costs were tabulated based on standardized institutional data. Pharmacy costs were obtained from our institution and applied to narcotics and anti-emetics administered for each patient. Mean cost per cohort is provided.

Surgical pick lists were obtained for both procedures. Cost of common items (bioadhesive, lapra-ty, etc.) and unique items was tabulated: laparoscopic cholecystectomy drape set, Gel-port laparoscopic system (Applied Medical, Rancho Santa Margarita, CA), and Harmonic Ace Scalpel (Ethicon Endo-Surgery, Inc., Cincinnati, OH) for HALPN; robotic/cardiovascular drape set and robotic instrument arms (endo-shears, bipolar forceps, prograsp forceps, and two-needle drivers) for RALPN. A robotic premium cost, the financial opportunity cost to use the robot, was estimated by dividing the initial purchase price by the duration in which materials are amortized at our institution, typically 7 years. We then added to that the yearly contract cost to obtain a yearly cost of owning a robot. This yearly cost was divided by the average case volume over the past 3 years to yield a robotic premium per case. Surgeon fees were not applied to total cost as they were equivalent, given the same CPT code utilized for both procedures.

Results

Discussion

Nephron sparing surgery (NSS), when technically feasible, has become the preferred paradigm for the treatment of SRMs. As chronic kidney disease is known to be associated with increased hospitalization, cardiovascular morbidity, and even mortality, PN has become the standard for the treatment of SRMs. ⁷ The 2009 AUA guidelines on the management of SRMs state that for both healthy patients and those with major comorbidities, PN is the standard of care, while RN reserved as an alternate standard of care for tumors not technically feasible for resection or as a co-standard for those patients in whom it would pose too high a surgical risk. ¹

While OPN requires a large incision with attendant pain, increased time to convalescence, and increased LOS, LPN has become an attractive alternative with similar oncologic and nephrologic outcomes, improvements in pain, convalescence time, and LOS. In a comparison of 1800 patients who underwent PN at three high-volume tertiary care centers, 771 patients who underwent LPN enjoyed shorter operative times (3.3 hours or 198 minutes), decreased intraoperative blood loss (300 mL), and shorter hospital stay (3.3 days) with equivalent nephrologic outcomes at 3 months. Additionally, the cancer-specific survival was equivalent at 3 years. The limitation, however, was that the patients who underwent LPN were more likely to have smaller and more peripheral tumors, required longer WIT, and sustained higher postoperative complications with a resultant requirement for further procedures. ² Certainly, this represents the outcome of a difficult laparoscopic technique, even in highly trained hands.

HALPN was first reported in 2000 by Wolf et al with favorable outcomes compared with OPN (shorter LOS and time to convalescence) and LPN (better hemostasis). ⁸ This was soon followed by a report from Stifelman et al who described their initial experience in 11 patients. ⁹ These authors concluded that the in-situ hand facilitated dissection, vascular control, hemostasis, and suturing. This early experience revealed an average operative time of 273 minutes, EBL of 319 mL, and LOS of 3.3 days with no transfusions and one conversion to open. In 2004, Brown et al described their experience with 30 consecutive patients who underwent HALPN. ¹⁰ Unique to this cohort was that none of their patients underwent hilar vessel dissection or clamping. Hemostasis was achieved via the use of manual compression and a fibrin-soaked gelatin sponge sewn into the resection bed and subsequently activated with thrombin. Their patients had an average EBL of 415 mL, operative time of 218 minutes, and LOS of 4.0 days. On subanalysis, the authors noted that for tumors <5 mm from the collecting system or hilar vessels, defined as central in their report, there was a 3–4-fold increase in EBL, transfusion rates, and urinary leakage. Hilar vessel clamping was recommended for all central lesions.

Strup et al outlined their experience in 2004 and 2005 with significant improvements in EBL, transfusion rates, and urinary leak since the adaptation of hilar vessel clamping and deep bolstering sutures. ¹¹ In their later report, ⁴ 76 patients with a mean 2.7-cm-diameter renal mass underwent HALPN with mean operative time of 214 minutes, EBL of 234 mL, 13.2% transfusion rate (prehilar vessel clamping era), and WIT of 26.3 minutes. Postoperative complications were seen in 18% (14 of 76), most frequent of which (5 of 14, 36%) was persistent Jackson–Pratt drainage. Most recently, Bylund et al described their experience with HALPN in 104 patients. ¹² These authors focused on their technique to obtain hemostasis and collecting system closure without formal suturing of the pelvicalyceal system. In their cohort, mean lesion size was 2.8 cm, WIT was 24.6 minutes (13 of 104, 12.5% were performed without hilar vessel clamping), EBL was 220 mL with 4.8% requiring transfusions, operative time was 189 minutes, and median LOS was 3 days. Only 2 of 47 patients who had documented collecting system entry demonstrated persistent urinary leakage; both resolved with conservative management.

Our series is in line with previous reports on HALPN outcomes. Our mean operative time of 149 minutes and EBL of 136 mL compare favorably with the literature while our LOS is slightly greater at 4.2 days. Only one patient required hilar vessel clamping for 23 minutes. An additional patient required transfusion and our overall complication rate was 18%. Clavien grade 3 or greater complications were seen in only one patient. Four patients suffered from incisional hernias; however, the severity of the hernias was unknown, representing a limitation of the retrospective nature of the study.

RALPN was first described in 2004 by Gettman and associates. ⁵ Clinical outcomes for RALPN are promising with recent series showing improvements in WIT and operative time with similar EBL, positive margin rates, and complication rates. Benway et al compared the outcomes of 118 patients who underwent LPN with 129 patients undergoing RALPN. ¹³ There was no difference in operative time and EBL was less in the RALPN cohort. There was a nonstatistically significantly shorter LOS in the RALPN group as well. An update on this RALPN series revealed operative time of 210 minutes, WIT of 23.9 minutes, and EBL of 131 mL with two conversions to open and a major complication rate of 8% (15 of 183). ¹⁴ At the AUA 2011 Annual Meeting, Munver et al reported on their multi-institutional experience with over 1000 RALPNs with promising results. ¹⁵ Mean tumor size was 2.9 cm, operative time was 198 minutes, WIT was 24.7 minutes, and mean EBL was 216 mL. There were 56 intraoperative complications and 118 major postoperative complications.

Our early series of 21 reveals favorable outcomes as compared with these results. For our mean tumor size of 2.5 cm, our mean operative time was 212 minutes, EBL was 178 mL, WIT (in 17 patients who underwent hilar clamping) of 29 minutes, and LOS was 3.5 days. There were 2 conversions to open, one for intractable bleeding and one for deep tumor infiltration. We experienced no major complications.

One limitation that we attempted to overcome was the use of a renal metric. It is difficult to compare outcomes from various studies without the use of some standard by which the difficulty of the extirpation may be assessed. In fact, none of the prior mentioned studies use a renal metric and few stratified these tumors by complexity. One study described complex tumors as those with collecting system entry on extirpation, but made no mention of diameter, exophytic versus endophytic nature, anterior versus posterior, proximity to the hilum, or how polar the tumor was. ¹⁶ All five of these parameters are components of the RENAL nephrometry score. While there may be unique features of a renal tumor that preclude categorization, the RENAL nephrometry score has been shown to be reproducible, reliable, and even predictive. ¹⁷ In fact, at the AUA 2011 Annual Meeting, Stroup et al reviewed all patients who underwent PN over a 6-year period and found that RENAL nephrometry score was predictive of which modality was chosen (OPN, LPN, or RALPN). ¹⁸ Another abstract at the same meeting revealed that increasing RENAL nephrometry scores correlated with adverse outcomes (urinary leak, EBL greater than one standard deviation above the mean, positive surgical margin, Clavien grade 3 or greater complication, or chronic renal insufficiency). ¹⁹ Clearly, size alone is not a reliable evaluation. We recommend the implementation of a renal metric, such as RENAL nephrometry, to allow for consistent and reliable comparisons between studies.

In our series all imaging was reviewed and RENAL nephrometry scores were assigned. Patients in the HALPN group and RALPN group had mean nephrometry scores of 5.24 and 6.4, respectively. These were not statistically different. Based on this data the outcomes of HALPN and RALPN are not confounded by differences in difficulty of extirpation.

Another limitation is various experiences with RALPN. It is evident that outcomes have improved since the first description of RALPN in 2004. ^5,12 Operative times, WIT, and complication rates have all decreased substantially. At our institution, we are yet early in our experience with RALPN and this may skew the outcomes in favor of our HALPN cohort, particularly, as a large part of the cost difference was due to operative time. Despite this, both surgeons at our institution are endourology fellowship trained and have substantial robotic experience, including RALPN. It is not clear whether experience with the robot in one procedure translates to another. Certainly a learning curve is appreciated. Pierorazio et al evaluated the outcomes of all patients who underwent LPN and RALPN by a single surgeon. ²⁰ They examined outcomes in groups of 25 and found that outcomes plateaued after the first 25 cases. This data suggests that the learning curve was approximately 25. Lavery et al compared a single-surgeon experience with RALPN and LPN outcomes and found that RALPN operative time and WIT were equivalent with LPN outcomes after 5 cases. ²¹ Deane et al suggested that with open experience a laparoscopic naive surgeon can achieve similar outcomes to LPN with RALPN after 10 cases. ²² While our RALPN volume is still somewhat low (21 cases), it was performed by a surgeon with RALPN experience during his fellowship.

Not only is surgeon's experience important but operating room staff familiarity is critical to room times. In fact, our experience revealed that operating room staff proficiency in RAL radical prostatectomy was achieved after 30 cases (unpublished data from our institution).

Financial impact should be included in any comparison of surgical techniques. The increasing incidence of detection of SRMs and an ever growing elderly demographic further illustrated the need for financial consideration. Even modest cost savings for a single patient may translate to significant reductions in healthcare costs. The Center for Medicare and Medicaid Services reports that the average payee receives more than $15,000 per year in benefits. Further, by 2030, 80 million baby boomers will have retired and average annual payee benefits will exceed $25,000 per year. The result of these costs may translate into a $70 trillion fiscal gap between projected federal expenditures and receipts. ²³ It is imperative that new technologies and techniques are delivered in a cost-effective manner. An emerging alternative modality for treatment of SRMs is ablative therapy. While this method is generally reserved for more elderly patients with increasing co-morbidities, the effectiveness and cost efficiency of this alternative treatment have been studied. Castle et al reported that ablative therapy was more effective than open, laparoscopic, or robotic PN for cT1a renal masses. ²⁴ Long-term follow-up was very limited for this study at only 6 months and increasing costs of more surveillance imaging were not considered. Currently we are evaluating the effectiveness and cost efficiency of ablation to hand-assisted and robotic partial nephrectomies with long-term follow-up at our institution.

Our financial analysis revealed that HALPN was considerably ($1165) less costly than RALPN in a routine and uncomplicated patient. Major contributors to cost were the increased operating room and anesthetic cost due to the increased operative time of the RALPN cohort, the robotic instruments, and the robotic premium. The robotic premium represents a large portion of the increased cost compared with hand assisted. Decreasing or eliminating the cost of the robot, for instance, if the robot was donated, or if the case volume increased such that it would decrease the robotic premium significantly, would certainly change our current analysis and make robotic partial more cost efficient. In this analysis, longer operative times for the robotic cohort contributed to increased cost. If operative times were decreased by 30 minutes or 1 hour, then cost savings would amount to a savings of $275 or $550, respectively. To render the cost comparable, operative time would need to decrease by more than 2 hours. Given the timing required to dock and undock the robotic, this is not technically feasible.

These costs, and hence the cost difference, may decrease with increased robotic volume and decreased room time. However, it is these costs that generate the majority of the cost difference in a program early in its RALPN experience. Our findings are similar to a recently published meta-analysis citing decreased equipment cost associated with LPN resulting in overall savings of nearly $1500 despite the shorter LOS of RALPN. ²⁵ This analysis only included open PN, LPN, and RALPN, without mention of HALPN.

Our study has several limitations. First, the retrospective nature of our report is not ideal. A randomized study would provide better evidence to support our data and we encourage such endeavors in the future. Second, we are early in our experience with RALPN. This may have skewed the data in favor of HALPN. Additionally, the lack of an LPN cohort limits our ability to compare both of these adjunctive approaches.

Conclusions

HALPN is an accepted approach to PN for SRMs. RALPN has similarly emerged as a viable option in this patient population. At our institution, HALPN and RALPN offer similar clinical outcomes. RALPN, however, is associated with longer operating room times with a slightly shorter hospital stay. The increased operative times and equipment cost of RALPN yielded an added cost of approximately $1150 per case. Despite its surgical viability, one must weigh the potential benefits of RALPN, such as increased dexterity and shorter LOS, against the increased cost of the procedure. It may be necessary to re-evaluate RALPN and perhaps optimize its use to those cases that would otherwise not be easily done with LPN or HALPN.