Surgical outcomes of 264 patients undergoing laparoscopy prior to planned partial hepatic resection.
Findings that precluded celiotomy in patients undergoing diagnostic laparoscopy prior to planned partial hepatectomy for metastatic colorectal cancer stratified by clinical risk score.
Grobmyer SR, Fong Y, D’Angelica M, DeMatteo RP, Blumgart LH, Jarnagin WR. Diagnostic Laparoscopy Prior to Planned Hepatic Resection for Colorectal Metastases. Arch Surg. 2004;139(12):1326–1330. doi:10.1001/archsurg.139.12.1326
(1) A clinical risk score (CRS) is useful in selecting patients for diagnostic laparoscopy prior to planned resection of colorectal metastasis. (2) Preventing unnecessary celiotomy in these patients undergoing laparoscopy is associated with shorter hospital stays and earlier administration of systemic chemotherapy.
Retrospective analysis of data collected in a prospective database.
Tertiary cancer hospital.
Two hundred seventy-six patients who underwent laparoscopy prior to planned partial hepatic resection (n = 264) for colorectal metastases or prior to hepatic artery infusion pump placement for colorectal metastases (n = 12).
Main Outcome Measures
The yield of laparoscopy for patients with potentially resectable tumors was analyzed in the context of a CRS, calculated by assigning 1 point for each of the following adverse, disease-related factors: lymph node–positive primary tumor, disease-free interval less than 12 months, more than 1 hepatic tumor, hepatic tumor greater than 5 cm, and cardio embryonic antigen level less than 200 ng/mL. The CRS represents the sum for each patient. Length of hospital stay and time to initiation of chemotherapy were compared in those patients determined to be unresectable.
Staging laparoscopy prevented nontherapeutic celiotomy in 10% of patients submitted to operation for a potentially curative partial hepatectomy (26 of 264) and in 33% of patients scheduled for pump placement only (4/12). The CRS correlated closely with the likelihood of identifying radiographically occult unresectable disease: 0 or 1, 4%; 2 or 3, 21%; and 4 or 5, 38%. Likewise, the percentage of patients avoiding an unnecessary celiotomy increased progressively with increasing CRS: 0 or 1, 0%; 2 or 3, 11%; and 4 or 5, 24%. Preventing an unnecessary celiotomy with laparoscopy was associated with a decreased length of hospital stay (P<.01) and earlier initiation of chemotherapy (P = .045).
Diagnostic laparoscopy does not need to be performed routinely in all patients prior to hepatic resection for colorectal cancer metastasis. Laparoscopy has a very low yield in patients with a CRS of 1 or less and is unnecessary. The yield of laparoscopy increases with increasing CRS. Preventing celiotomy with laparoscopy is associated with a decreased length of hospital stay and earlier initiation of postoperative chemotherapy.
Hepatic resection remains the only potentially curative therapy for metastatic colorectal cancer.1- 3 With the emergence of additional effective chemotherapeutic agents4,5 (irinotecan; oxaliplatin) and strategies (chronomodulated chemotherapy6 and adjuvant pump chemotherapy7) during the past several years, the indications for liver resection have been extended to include patients who would have previously been considered unresectable. Additionally, aggressive treatment approaches that combine resection and tumor ablation, often with systemic and/or regional chemotherapy, are being used with greater frequency. Thus, as the treatment algorithms for patients with hepatic colorectal metastases continue to evolve, so does the definition of “unresectable” disease.
Despite this trend toward more aggressive operative approaches in patients with advanced disease, the presence of extrahepatic metastases and the inability to achieve complete tumor clearance remain relative contraindications to resection. Ensuring resectable disease before celiotomy has been a difficult problem in this patient population,8- 10 as it has been for patients with other types of gastrointestinal malignancies.11- 16 These difficulties have fostered efforts to reduce the incidence of unnecessary exploratory celiotomies. Diagnostic laparoscopy in particular has been used with greater frequency and has been shown to be useful in patients with hepatobiliary malignancy.10,16,17Patients with unresectable disease identified laparoscopically have a reduced hospital length of stay, reduced hospital charges, and lower morbidity compared with those undergoing a nontherapeutic celiotomy.10,12 In the subset of patients with hepatic metastases from colorectal cancer, the value of diagnostic laparoscopy has been ill defined.8- 10,18 Prior studies have suggested that laparoscopy spares up to 33% of patients an unnecessary celiotomy.18 By contrast, we previously published a yield of 10%,10,15,17 likely a reflection of improvements in imaging technology and better preoperative identification of patients with unresectable disease.
In an era of rapid advancements in imaging techniques and more aggressive surgical intervention, the routine use of laparoscopy in patients with hepatic metastases of colorectal cancer is probably unnecessary since most do not benefit from the procedure. On the other hand, some patients are probably at greater risk than others for harboring radiographically occult unresectable disease and the potential benefit of laparoscopy is greater.10 The present study seeks to clarify the role of laparoscopy prior to planned celiotomy in patients with hepatic metastases from colorectal cancer and specifically seeks to validate the utility of a preoperative clinical risk scoring system (using a clinical risk score [CRS])3 for stratifying patients most and least likely to benefit from laparoscopy in this setting.
Data on patients who underwent diagnostic laparoscopy between December 1997 and December 2002 prior to planned hepatic resection for colorectal cancer metastases, as well as patients who underwent diagnostic laparoscopy prior to hepatic artery infusion pump placement, were collected from a prospectively maintained database and analyzed retrospectively. Patients who underwent laparoscopy to confirm the presence of unresectable disease were not included. Additional data were obtained through a review of hospital records and included demographics, extent of preoperative imaging, extent of laparoscopic examination, complications related to laparoscopic examination, surgical findings, and length of hospital stay. The results of diagnostic laparoscopy in the first 103 of these patients undergoing planned hepatic resection (December 1997 to December 1999) were the subject of an earlier report.10 The current study represents our cumulative experience with laparoscopy for metastatic colorectal cancer.
Our general approach to patients with potentially resectable hepatic colorectal metastases has been described previously.10 All patients had histologically confirmed metastatic colorectal cancer to the liver, whether or not resection was performed. Cases were reviewed at a weekly Hepatobiliary Disease Management conference attended by surgeons, oncologists, radiologists, and gastroenterologists. Patients typically had a portion of the preoperative evaluation performed by the referring physician; additional imaging tests were obtained as indicated. For the purposes of the current study, only imaging tests and carcinoembryonic antigen levels obtained within 6 weeks of surgery were considered preoperative and used for analysis.
Data pertaining to the primary tumor and the metastatic liver disease were reviewed and used to calculate a CRS.3 The CRS comprises 5 preoperative clinical variables previously shown to be independent predictors of survival after resection.3 One point was given for each of the following and the CRS represented the sum: lymph node–positive primary tumor, disease-free interval (time of colorectal primary tumor to diagnosis of liver metastasis) less than 12 months, more than 1 hepatic tumor, largest diameter of hepatic tumor greater than 5 cm (based on preoperative imaging), and carcinoembryonic antigen level less than 200 ng/mL. Complete CRS data were not available on all patients. The operative findings and yield of diagnostic laparoscopy were analyzed in total and after stratification of patients according to CRS.
Our technique of diagnostic laparoscopy has been previously described.15 Laparoscopy was performed through upper abdominal port sites immediately prior to planned celiotomy. Laparoscopic ultrasonography was performed at the discretion of the operating surgeon. Laparoscopic examination was complete if anterior and posterior surfaces of the right and left hepatic lobes, the gastrohepatic omentum, porta hepatis, pelvis, and peritoneal cavity were well visualized. Biopsies of suspicious extrahepatic lesions noted at the time of laparoscopy were performed and the results sent for frozen section. Many patients submitted to operation were enrolled in a protocol of postresectional regional plus systemic chemotherapy and underwent simultaneous placement of a hepatic artery infusion pump. Additionally, some of these patients were simultaneously given informed consent information for possible resection plus ablation with hepatic artery pump placement or pump placement alone (both on protocol) in the event that unresectable disease confined to the liver was identified. All hepatic artery infusion pumps were placed through a celiotomy. The finding of extrahepatic disease or additional hepatic disease not amenable to complete extirpation constituted irresectability. Similarly, patients with extrahepatic disease did not undergo hepatic artery pump placement. A small number of patients with synchronous hepatic metastases were considered for simultaneous resection of the colorectal primary tumor and liver disease. Nontherapeutic celiotomy refers to performance of a celiotomy after which the patient did not undergo any treatment directed at the metastatic liver disease. No patients in this series underwent laparoscopic hepatic resection.
Time to initiation of systemic chemotherapy following operation was compared in patients who had unresectable disease identified by laparoscopy only and in patients who had unresectable disease identified at the time of celiotomy.
Statistical analyses were performed using StatView 5.0. Continuous variables were compared using the t test and categorical variables were compared using the Fisher exact test. Numeric data are expressed as mean ± standard error of the mean. A P value of <.05 was considered statistically significant for the purposes of this study. This study was approved by the institutional review board of Memorial Sloan Kettering Cancer Center, New York, NY.
Between December 1997 and December 2002, 276 patients underwent diagnostic laparoscopy prior to planned hepatic resection (n = 264) or hepatic artery pump placement only (n = 12). Patients in the former group all had disease that was considered potentially resectable based on preoperative imaging and were submitted to operation for resection as the primary objective. The demographics and clinical characteristics of these patients are outlined in Table 1. One hundred sixty-eight of 264 patients had a complete laparoscopic examination. The most common reason for incomplete examination was adhesions from prior surgery (n = 259). Complications from laparoscopy occurred in only 4 patients (1.4%): 2 serosal tears and 2 small-bowel injuries. All injuries were repaired primarily without further complication.
Forty-eight (18%) of 264 patients scheduled for partial hepatectomy had neither a resection nor any other cancer-specific procedure performed (Figure 1). Twenty-six (54%) of the 48 patients in this group were spared a celiotomy as a result of the laparoscopic findings. Overall, 26 (10%) of 264 patients were spared nontherapeutic celiotomy as a result of laparoscopy. Laparoscopy reduced the nontherapeutic celiotomy rate from 18% to 8%. Disease identified at laparoscopy that precluded celiotomy was additional, unsuspected hepatic tumors in 11 patients (42%) and extrahepatic metastases in 15 patients (58%). Of the 22 patients (8.3%) who had a nontherapeutic celiotomy, the intraoperative findings precluding resection that were missed at laparoscopy were as follows: extrahepatic metastases in 16 patients (perihepatic lymph nodes [n = 11], peritoneal disease [n = 5]) and additional hepatic tumors in 6 patients. Adhesions limited the laparoscopic examination in 18 (82%) of these 22 patients. Patients who underwent diagnostic laparoscopy only had a significantly shorter mean ± SEM hospital length of stay compared with those who had a nontherapeutic celiotomy 1.3 ± 0.6 days vs 5.5 ± 0.6 days (P<.01). For patients on whom data were available, irresectable disease identified by laparoscopy only (n = 8) was associated with a significantly shorter mean ± SEM interval to initiation of postoperative systemic chemotherapy than irresectable disease identified by celiotomy (n = 23), 14.2 ± 3.3 days vs 20.7 ± 1.4 days (P = .045).
The distribution of CRSs for all patients scheduled to undergo hepatic resection is demonstrated in Table 2. The CRS predicted the likelihood of finding radiographically occult unresectable disease, which was 4% in those patients with a CRS of 0 or 1, 21% in those with a CRS of 2 or 3, and 38% in those with a CRS of 4 or 5 (Table 2). Similarly, the CRS correlated closely with the yield of laparoscopy. Laparoscopy did not prevent an unnecessary celiotomy in patients with a CRS of 0 or 1 (Table 2), compared with 11% of patients with a CRS of 2 or 3 and 24% with a CRS of 4 or 5 (P<.01). Findings at laparoscopy that precluded celiotomy for each risk score are shown in Figure 2. The surgical outcomes of patients stratified by CRS are shown in Table 2.
The yield of laparoscopy was not significantly related to the number of preoperative studies performed. The yield of laparoscopy stratified by the number of preoperative studies was as follows: 1 study, 4 of 42 (number of patients avoiding celiotomy of total number of patients with score) (10%); 2 studies, 13 of 117 (11%); 3 studies, 8 of 78 (10%); 4 studies, 1 of 21 (5%); and 5 studies, 0 of 6 (0%) (P = .80).
Twelve patients in the series had diagnostic laparoscopy prior to planned hepatic artery infusion pump placement. These patients had extensive hepatic disease that precluded resection based on preoperative imaging studies. Four patients in this group (33%) were spared unnecessary celiotomy as a result of the laparoscopic finding of extrahepatic metastatic disease.
There have been numerous reports describing the value of diagnostic laparoscopy in upper gastrointestinal malignancies,8- 13,15,16,18- 26 although relatively few have focused on patients with hepatic colorectal metastases. To our knowledge, the present report represents the largest published series of diagnostic laparoscopy in the surgical management of patients with hepatic metastasis from colorectal cancer and updates our previously reported experience. This study confirms the benefit of laparoscopic identification of irresectable disease in decreasing the length of hospital stay and demonstrates, for the first time, an associated shortening of the interval to initiation of system chemotherapy in these patients. The study also shows that the vast majority of patients do not benefit from the procedure, given the overall yield of 10%.10 Efforts to target laparoscopy to patients at high risk for harboring radiographically occult unresectable disease are therefore necessary.
The present study stratifies patient risk based on clinical, disease-specific variables, each previously shown to be an independent predictor of outcome after resection.3 Although perhaps crude, the CRS does provide some insight into disease biology and disease extent beyond that obtainable through imaging studies. The value of this scoring system is that it is based on clinical data readily obtainable prior to operation.
The results show that the CRS is useful for identifying patients at greatest risk for radiographically occult unresectable disease and therefore most likely to benefit from laparoscopy. Diagnostic laparoscopy prevented a nontherapeutic celiotomy in none of the patients with a CRS of 0 or 1 and is thus of little benefit in this subset unless being used to evaluate a specific radiological finding that might alter management. On the other hand, a CRS of 4 or 5 was associated with the highest yield of laparoscopy, with 24% of patients avoiding a nontherapeutic celiotomy. Given the high yield of the procedure in this group, staging laparoscopy should be used liberally, if not routinely. Patients with a CRS of 2 or 3 had a more modest 11% yield. Decisions on the use of staging laparoscopy in these patients should perhaps be based on factors specific to the case, such as findings on preoperative imaging studies and the possible performance of alternative procedures in the event that unresectable disease confined to the liver is identified. In patients with advanced disease who are being considered for hepatic artery pump placement only, the yield of laparoscopy is sufficiently high to warrant its frequent use.
One might predict that increasing the number of noninvasive preoperative imaging studies would decrease the yield of laparoscopy. However, the results of the present study do not support this concept. The yield of laparoscopy was not significantly related to the number of preoperative imaging studies performed. The quality of the studies, on the other hand, almost certainly affects the yield of laparoscopy, although this is difficult to assess and was not evaluated in the present study. Clearly, however, advances in imaging technology, in addition to the greater number of surgical treatment options and the increasingly aggressive treatment approach to patients with hepatic colorectal metastases, have contributed to the declining yield of laparoscopy during the past several years.
In conclusion, diagnostic laparoscopy does not need to be performed in all patients prior to partial hepatectomy for metastatic colorectal cancer. Patients being considered for resection who have a high CRS and those with advanced liver disease who are being considered for hepatic artery pump placement only are at greatest risk for radiographically occult unresectable disease. The CRS should be used in the context of the overall operative plan and in conjunction with information obtained on preoperative imaging studies to identify patients most likely to benefit from diagnostic laparoscopy.
Correspondence: William R. Jarnagin, MD, Department of Surgery, Memorial Sloan Kettering Cancer Center, 1275 York Ave, New York, NY 10021 (firstname.lastname@example.org).
Accepted for Publication: June 18, 2004.
Previous Presentation: Portions of this study were presented at the American Hepato-Pancreato-Biliary Association annual meeting; February 28, 2003; Miami, Fla.