Surgical and pathological outcomes after right hemicolectomy: case-matched study comparing robotic and open surgery

Transcription

1 THE INTERNATIONAL JOURNAL OF MEDICAL ROBOTICS AND COMPUTER ASSISTED SURGERY Int J Med Robotics Comput Assist Surg 2011; 7: Published online 11 May 2011 in Wiley Online Library (wileyonlinelibrary.com).398 ORIGINAL ARTICLE Surgical and pathological outcomes after right hemicolectomy: case-matched study comparing robotic and open surgery Fabrizio Luca 1 * Tiago Leal Ghezzi 1,4 Manuela Valvo 1 Sabina Cenciarelli 1 Simonetta Pozzi 1 Davide Radice 2 Cristiano Crosta 3 Roberto Biffi 1 1 Division of Abdomino-Pelvic Surgery, 2 Division of Epidemiology and Biostatistics and 3 Division of Endoscopy; European Institute of Oncology, Milano, Italy 4 Moinhos de Vento Hospital, Porto Alegre, Brazil *Correspondence to: Fabrizio Luca, European Institute of Oncology, Via Ripamonti 435, 20141, Milano, Italy. fabrizio.luca@ieo.it Abstract Objective To compare the surgical and pathological outcomes of patients with right-sided colon cancers operated on by means of open and robotic surgery with extracorporeal anastomosis. Methods Thirty-three consecutive patients who underwent robotic right hemicolectomy due to right-sided colon cancer were retrospectively well matched with 102 patients operated on by the open approach. Data were included in a prospectively maintained database. Results Mean operative time was longer in the robotic group (P < 0.001), min ( ) versus (45 240) min in the open group. Estimated intraoperative blood loss was less in the robotic group, which presented a mean of 6.1 ml versus 94.8 ml in the open group (P < 0.001). Despite the similar length of the surgical specimen and number of lymph nodes retrieved between both groups, 15 or more lymph nodes were found in the specimen in 90 out of 102 patients (88.2%) operated on by the open technique versus 33 out of 33 patients (100%) who underwent robotic hemicolectomy (P = 0.038). The median length of postoperative hospital stay was shorter in the robotic group, 5 versus 8 days (P < 0.001). No other statistically significant difference was observed in terms of pathological and postoperative results. Conclusions Robotic right hemicolectomy is an oncologically safe and effective procedure. The number of lymph nodes retrieved in the robotic group compared with the open group of our series was more homogeneous, and none of the patients operated on with this technique had a suboptimal lymphadenectomy. Further clinical trials are needed to confirm current evidence and determine whether this can influence the prognosis. Copyright 2011 John Wiley & Sons, Ltd. Keywords colon cancer; robotic surgery; colorectal surgery; da Vinci System; right hemicolectomy Introduction Accepted: 29 March 2011 The main objective of colon cancer surgery is the removal of the primary tumor, including an adequate surgical margin, together with locoregional lymph node dissection (1). Open right hemicolectomy has traditionally been considered the standard procedure for right-sided colonic tumors (2). Copyright 2011 John Wiley & Sons, Ltd.

2 Robotic right hemicolectomy 299 Although the first laparoscopic colorectal resection was performed almost 20 years ago and several publications have supported the evidence of significant improvements in short-term outcomes, laparoscopic colorectal surgery is still limited to relatively few centers with significant expertise (2 5). After the FDA approval for clinical use in abdominal surgery in 2000, the da Vinci robotic surgical system has been safely and successfully used in a variety of colorectal procedures (6,7). Since the first robotic colorectal surgery performed in 2002, the da Vinci robot has been used mainly for the treatment of left-sided colonic tumors and rectal cancer (7,9,10). As a consequence, the outcomes of robotic right hemicolectomy have not been widely examined (3,7,8). The aim of this study was to compare surgical and pathological outcomes of patients with right-sided colon cancer operated on either by the open procedure or the robotic technique. Materials and Methods Thirty-three consecutive patients who underwent robotic right hemicolectomy with extracorporeal anastomosis between February 2008 and July 2010, in the Division of Abdomino-Pelvic Surgery of the European Institute of Oncology, Milan, Italy, were compared with 102 patients who underwent open surgery from July 1994 to December The two groups were matched with respect to sex, age, body mass index (BMI), American Society of Anesthesiologists (ASA) score, history of previous abdominal surgery, AJCC/UICC stage and tumor location ( Right-sided colon cancers were defined as cancer located at the cecum, ascending colon, hepatic flexure, or proximal transverse colon. Exclusion criteria were T4 tumor, AJCC/UICC stage IV, multivisceral resection, and segmental resection. Data were extracted from a prospectively maintained database. The following outcomes were analyzed in each group: length of postoperative hospital stay, operative time, estimated operative blood loss, intra and postoperative blood transfusion, postoperative hemoglobin drop, length of postoperative hospital stay, postoperative complications, reoperation, hospital readmission, mortality, pathological TNM AJCC/UICC stage (categories T and N), surgical specimen length (colon, ileal, and total), surgical margin and number of harvested lymph nodes. Previous abdominal surgery was defined as any surgical procedure, laparoscopic or open, that could lead to the formation of peritoneal adhesions. Inguinal herniorrhaphy was not considered as previous abdominal surgery. Postoperative hemoglobin drop was defined as the difference between the preoperative hemoglobin value and the lowest value during the postoperative hospital stay. Robotic right hemicolectomy All patientsunderwentmechanical bowel preparation and were administered prophylactic antibiotics intravenously Figure 1. Position of the robotic cart at induction of anesthesia. All the operations were performed by two senior surgeons with considerable experience in robotic colorectal surgery. Pneumoperitoneum is induced through a Veress needle using carbon dioxide insufflated to a pressure of 12 mmhg. The procedure starts with the introduction of a 12 mm optical trocar 2 cm left lateral to the umbilicus in order to enable laparoscopic exploration of the abdomen, determination of the feasibility of a minimally invasive procedure, performance of laparoscopic liver ultrasonography, and insertion of the remaining ports under direct vision. The table is then turned into a moderate left tilt. The small bowel is displaced to the left abdomen with the help of laparoscopic bowel graspers. The robotic cart is then positioned on the right side of the patient at an angle of degrees and docked to the trocars (Figure 1). The disposition of the trocars is illustrated in Figure 2. A bipolar fenestrated grasper and harmonic shears are used through robotic trocars 1 and 2, respectively. A 12 mm trocar inserted in the left lateral flank is used by the assistant surgeon. The procedure begins with isolation of the ileocolic trunk at its origin. The mesocolic medial-to-lateral dissection continues with the identification of duodenum, right ureter, and gonadal vessels. The ileocolic trunk, right colic, and right branch of the middle colic vessels are isolated and divided between haemostatic clips. A linear laparoscopic stapler is used for

3 300 F. Luca et al. examinations were performed by means of the same technique and by the same team of pathologists. The histological diagnosis was coded according to the TNM AJCC/UICC classification system. Solutions for lymph node clearance were not employed for pathological examination. Statistical analyses were conducted with SAS software, version 9.2. Descriptive statistics were calculated for all variables. Data are presented as mean and range or number of patients and percentage unless otherwise stated. The two-sample two-sided Wilcoxon test or the unpaired t-test were employed to compare continuous variables when appropriate. The chi-square or two-sided Fisher s exact test were used as appropriate to compare categorical variables. The p values <0.05 were considered to indicate statistical significance (two-tailed test). All patients received a detailed explanation of the procedure and provided informed consent. This study was not supported by any commercial fund or sponsorship. Results Figure 2. Placement of the trocars: RT1-2, robotic ports (8 mm); OT, optical trocar (12 mm); A, assistant trocar (12 mm) the intracorporeal transection of the terminal ileum. The colon is retrieved and sectioned through a small transverse laparotomy in the right upper quadrant. Stapled endto-side or side-to-side anastomosis is then performed according to surgeon preference. The colon is then returned to the abdominal cavity, the minilaparotomy is closed and the pneumoperitoneum is re-established. A drain is positioned near the anastomosis and exteriorized through the left lower quadrant port. Open right hemicolectomy Patients of the open surgery group were treated applying the same standards of preoperative and postoperative care as those of the robotic surgery group. Surgery was performed by experienced senior colorectal surgeons. Right colon mobilization was performed in a lateralto-medial approach following the same oncological principles described above. Stapled end-to-side or sideto-side anastomosis was performed according to surgeon preference. Outcome measures and statistical analysis Operative morbidity and mortality were defined as any postoperative complication or death that occurred within 30 days after surgery. The patients were discharged when they were tolerating a normal diet, and pain was well controlled with oral analgesics. All the pathological Thirty-three patients who underwent robotic right hemicolectomy were compared with 102 patients who underwent surgery by means of the open approach. Both groups were statistically well matched in terms of clinical pathological characteristics, as shown in Table 1. The operative time was 54.5 min longer in the robotic group (P < 0.001), with a mean of ( ) min in the robotic group and (45 240) min in the open group. Estimated intraoperative blood loss was less Table 1. Clinical pathological characteristics for patients matching Open (n = 102) Robotic (n = 33) P-value Gender Male 50 (49.0) 16 (48.5) Female 52 (51.0) 17 (51.5) Age a (years) 66.1 (34 88) 65.2 (44 87) BMI a (kg/m 2 ) 24.8 (16 35) 25.4 (20 32) History of abdominal surgery b 52 (51.0) 17 (51.5) ASA score I 9 (8.8) 4 (12.1) II 59 (57.9) 25 (75.8) III 34 (33.3) 4 (12.1) AJCC/UICC stage I 17 (16.7) 9 (27.3) II 42 (41.2) 16 (48.5) III 43 (42.1) 8 (24.2) Tumor location Cecum 43 (42.1) 15 (45.4) Ascending colon 33 (32.4) 15(45.4) Hepatic flexure 11 (10.8) 2 (6.1) Transverse colon 15 (14.7) 1 (3.1) ASA American Society of Anesthesiologists. a Mean (range) value expressed. b Any surgical procedure, laparoscopic or open, that could lead to peritoneal adhesions. Inguinal herniorrhaphy was not considered as prior abdominal surgery.

4 Robotic right hemicolectomy 301 in the robotic group, which presented a mean of 6.1 ml versus 94.8 ml in the open group (P < 0.001). The mean of postoperative hemoglobin drop was 1.69 and 1.55, respectively, in the open and robotic group (P = 0.445). Intraoperative and postoperative blood transfusions were more frequent in the open group (4.9% and 12.8%) than in the robotic group (0% and 9.1%); these differences, however, were not statistically significant (0.334 and 0.760, respectively). The length of postoperative hospital stay was statistically reduced in the robotic group (P < 0.001). Patients who underwent robotic surgery had a median postoperative stay of 5 days, while those operated on through the open technique stayed 8 days. Although not statistically significant, a lower percentage of infectious, non-infectious, and overall postoperative complications (9.1% vs. 20.6%, 15.1% vs. 17.6%, and vs. 33.3%) was observed among patients who underwent robotic surgery. Three patients of the open group underwent reoperation due to postoperative surgical complications (one case of anastomosis leak and two cases of abdominal wall dehiscence). No reoperation was needed in the robotic group (P = 1.000) and no anastomotic leak occurred. One patient needed readmission for conservative treatment of a surgical site infection after a robotic right hemicolectomy. No readmission was needed in the open group (P = 0.244). No mortality was reported in bothgroups (P = 1.000). Perioperative surgical outcomes are reported in Table 2. Pathological examinations did not show a statistically significant difference between the two groups in terms of category T (TNM) (P = 0.928), while regarding category N (TNM), 57.8%, 30.4%, and 11.8% of patients in Table 3. Pathological outcomes Open (n = 102) Robotic (n = 33) P-value Surgical specimen length a (cm) Colon 22.0 (11 73) 19.1 (10 32) Ileal 7.3 (1 50) 5.5 (2 12) Total 29.3 (15 37) 24.6 (15 37) Negative surgical margin 102 (100) 33 (100) No of harvested lymph nodes a 25.4 (8 74) 26.6 (15 46) Category T (TNM) (3.0) 2 (6.2) 2 17 (17.2) 6 (19.4) (70.7) 21 (67.7) 4 10 (10.1) 3 (9.7) Category N (TNM) 0 59 (57.8) 25 (75.8) 1 31 (30.4) 6 (18.2) (11.8) 2 (6.1) a Mean (range) value expressed. the open group were, respectively, classified as pn0, pn1, and pn2, versus 75.8%, 18.2%, and 6.1% in the robotic group (P = 0.200). The mean length of colon and ileum surgical specimens was, respectively, 22.0 cm. and 7.3 cm. in the open group, and 19.1 cm. and 5.5 cm. in the robotic group (P = and 0.172). Microscopic examination confirmed negative surgical margins in all patients. At least 15 or more lymph nodes were retrieved in 90 patients (88.2%) treated by means of the open technique versus 33 patients (100%) in the robotic group (P = 0.038). The mean number of lymph nodes harvested was 25.4 and 26.6 respectively in the open and robotic arms (P = 0.355). Pathological outcomes are presented in Table 3. Table 2. Perioperative surgical outcomes Open (n = 102) Robotic (n = 33) P-value Operative time (min) a (45 240) ( ) <0.001 Estimated blood loss 94.8 (0 400) 6.1 (0 100) <0.001 (ml) a Hemoglobin drop 1.69 ( ) 1.55 ( ) (g/dl) a Length of 8.0 (7 59) 5.0 (4 32) <0.001 postoperative stay (days) b Blood transfusions (patients) Intraoperative 5 (4.9) 0 (0) transfusion Postoperative 13 (12.8) 3 (9.1) transfusion Complications c Non-Infectious 18 (17.6) 5 (15.1) Infectious 21 (20.6) 3 (9.1) Overall 34 (33.3) 8 (24.2) Reoperation c 3 (2.9) 0 (0) Hospital readmission c 0 (0) 1 (3.0) Mortality c 0 (0) 0 (0) a Mean (range) value expressed. b Median (range) value expressed. c During the first 30 postoperative days. Discussion Intraoperative blood loss was significantly greater in patients treated with open surgery. Robotic intervention was associated with a longer operative time; by contrast, hospital stay was significantly shorter in the robotic group with a median of 5 vs. 8 days. These findings have also been described in studies comparing open surgery with traditional laparoscopy (10). There is a general agreement that the da Vinci System provides a three-dimensional magnified view, stable camera platform, wristed instruments, tremor elimination, and an ergonomically correct position, and thereby helps in reducing the mental and physical stress of the surgeon. This was directly measured by van der Schatte Olivier et al., who demonstrated an increase in work efficiency and a lower stress load in a robotic compared with a laparoscopic study group (11). This issue has different implications and is not only a question of reducing surgeon fatigue. A link between stress and surgical performance was discussed in the interesting review of Arora et al. and they concluded that these aspects may have substantive implications for patient safety (12).

5 302 F. Luca et al. Different authors from centers with significant expertise demonstrated that robotic surgery for the treatment of right colon cancer is feasible and oncologically adequate. The present study confirms that all the pathological parameters analyzed, such as the length of the specimen and the number of lymph nodes retrieved, were comparable between the two groups. However, when we considered the cut-off number of 15 lymph nodes there was a significant difference in favor of the robotic group. In fact, although the average number was similar (25.4 vs. 26.6) the number of lymph nodes retrieved in the robotic group of our series was more homogeneous when compared with the open group. The reason for this remains unclear, but a potential explanation was suggested by D Annibale et al. They observed that the da Vinci System allows better standardization of the surgical technique of right hemicolectomy, positively increasing the percentage of correct lymphatic resections (13). It is now generally accepted that the assessment of lymph node status in colon cancer patients represents one of the most important quality indicators by which to measure and compare the different surgical treatment that patients receive, since lymph node retrieval is related to local recurrence and mortality (14 16). Chen et al. after analyses of 82,896 patients included in the Surveillance, Epidemiology, and End Results (SEER) cancer registry maintained by the National Cancer Institute, showed that adequate lymphadenectomy, as measured by the analysis of at least 15 lymph nodes, correlates with improved survival, independent of stage, patient demographics, and tumor characteristics (17). This observation was recently supported by the conclusions of the study by Downing et al. Using a population-based dataset of all patients affected by colorectal cancer included in the SEER database they found that, across all T-stages, using successive lymph node cut-offs (6 26), the highest optimal number of LN resected was 15 (18). Lowering the percentage of suboptimal lymphadenectomies could thus improve the global quality of our surgery and, maybe, positively reduce the risk of recurrence. We agree with Prasad et al. that among all robotic colorectal resections, right hemicolectomy posesses the characteristics that make it the ideal procedure for the surgeon at the beginning of the learning curve (7). However, if our data are confirmed in clinical trials with a higher number of patients and a longer follow-up, the role of robotic surgery in the treatment of right colon cancer couldextendbeyonditsuseasasimplelearningtool and also have important oncologic implications. Robotics could also improve the as yet poor adoption of minimally invasive colon cancer surgery, estimated to be between 6 and 12% in western countries (19,20). In this series we decided to perform an extracorporeal anastomosis in order not to increase the complexity of the robotic operation, with a laparoscopic procedure such as intracorporeal ileocolic anastomosis which is generally considered challenging (21). We believe that the development of dedicated robotic instruments will also help overcome the technical difficulties of this procedure. In conclusion, this study confirms that robotic right hemicolectomy is a safe and effective procedure for the treatment of colon cancer. The number of lymph nodes retrieved in the robotic group of our series was more homogeneous. None of the patients operated on with this technique had a suboptimal lymphadenectomy while the number of lymph nodes found in the specimen were less than 15 in nearly 12% of the patients in the open group. More data are needed in order to determine whether this can influence the oncological outcome in patients treated for right colon cancer. Acknowledgements The authors thank William Russell-Edu, Librarian of the European Institute of Oncology for his contribution to the bibliographic research of this study. References 1. Park IJ, Choi GS, Kang BM, et al. Lymph node metastasis patterns in right-sided colon cancers: is segmental resection of these tumors oncologically safe? Ann Surg Oncol 2009; 16: Chung CC, Ng DC, Tsang WW, et al. Hand-assisted laparoscopic versus open right colectomy: a randomized controlled trial. Ann Surg 2007; 246: Tan WS, Chew MH, Ooi BS, et al. Laparoscopic versus open right hemicolectomy: a comparison of short-term outcomes. Int J Colorectal Dis 2009; 24: Hazebroek EJ. COLOR: a randomized clinical trial comparing laparoscopic and open resection for colon cancer. Surg Endosc 2002; 16: Bucher P, Pugin F, Morel P. Single port access laparoscopic right hemicolectomy. Int J Colorectal Dis 2008; 23: Ballantyne G, Moll F. The da Vinci telerobotic surgical system: the virtual operative field and telepresence surgery. Surg Clin North Am 2003; 83: de Souza AL, Prasad LM, Park JJ, et al. Robotic assistance in right hemicolectomy: is there a role? Dis Colon Rectum 2010; 53: Ostrowitz MB, Eschete D, Zemon H, DeNoto G. Robotic-assisted single-incision right colectomy: early experience. Int J Med Robot 2009; 5: Weber PA, Merola S, Wasielewski A, Ballantyne GH. Teleroboticassisted laparoscopic right and sigmoid colectomies for benign disease. Dis Colon Rectum 2002; 45: Nakamura T, Onozato W, Mitomi H, et al. Retrospective, matched case-control study comparing the oncologic outcomes between laparoscopic surgery and open surgery in patients with right-sided colon cancer. Surg Today 2009; 39: van der Schatte Olivier RH, Van t Hullenaar CD, Ruurda JP, Broeders IA. Ergonomics, user comfort, and performance in standard and robot-assisted laparoscopic surgery. Surg Endosc 2009; 23: Arora S, Sevdalis N, Nestel D, et al. The impact of stress on surgical performance: a systematic review of the literature. Surgery 2010; Mar; 147: D Annibale A, Pernazza G, Morpurgo E, et al. Robotic right colon resection: evaluation of first 50 consecutive cases for malignant disease. Ann Surg Oncol 2010; 17: Urbach DR, Baxter NN. Reducing variation in surgical care. BMJ 2005; 330: Wright FC, Law CH, Berry S, Smith AJ. Clinically important aspects of lymph node assessment in colon cancer. JSurgOncol 2009; 99: Kim YW, Kim NK, Min BS, et al. The influence of the number of retrieved lymph nodes on staging and survival in patients

6 Robotic right hemicolectomy 303 with stage II and III rectal cancer undergoing tumor-specific mesorectal excision. Ann Surg 2009; 249: Chen SL, Bilchik AJ. More extensive nodal dissection improves survival for stages I to III of colon cancer: a population-based study. Ann Surg 2006; 244: Downing SR, Cadogan KA, Ortega G, et al. The number of lymph nodes examined debate in colon cancer: how much is enough? JSurgRes2010; 163: NICE implementation uptake report: laparoscopic surgery for colorectal cancer. NICE technology appraisal 105. Available from: Reichenbach DJ, Tackett AD, Harris J, et al. Laparoscopic colon resection early in the learning curve: what is the appropriate setting? Ann Surg 2006; Jun; 243: Jamali FR, Soweid AM, Dimassi H, et al. Evaluatingthedegree of difficulty of laparoscopic colorectal surgery. Arch Surg 2008; Aug; 143: