Tumor location in laparoscopic group. Numbers of conversions to laparotomy are given in parentheses.
Postoperative change in prothrombin rate, total bilirubin level, and aspartate aminotransferase level. To convert bilirubin to micromoles per liter, multiply by 17.1.
Rate of overall survival according to treatment group (A), according to treatment group excluding postoperative mortality (B), and disease free according to treatment group (C).
Laurent A, Cherqui D, Lesurtel M, Brunetti F, Tayar C, Fagniez P. Laparoscopic Liver Resection for Subcapsular Hepatocellular Carcinoma Complicating Chronic Liver Disease. Arch Surg. 2003;138(7):763-769. doi:10.1001/archsurg.138.7.763
Laparoscopic liver resection for subcapsular hepatocellular carcinoma in patients with chronic liver disease is associated with lower morbidity than open resections.
A case-comparison study.
A tertiary referral center.
Patients and Intervention
From December 1, 1998, to November 30, 2000, 13 patients with chronic liver disease who underwent laparoscopic resection of hepatocellular carcinoma formed the laparoscopic group (LG). Tumors were 5 cm or smaller, subcapsular, and located in anterolateral segments (segments II-VI). A control group was created by matching each laparoscopic case with patients identical for liver disease, tumor size, and location and type of hepatectomy who underwent open liver resection. Fourteen patients fulfilled the criteria and formed the open group (OG).
Main Outcome Measures
Postoperative mortality and morbidity.
One segment or less was resected in 21 patients and 2 in 6 patients. Operative duration and cumulative portal triad clamping times were longer in the LG (267 ± 79 minutes vs 182 ± 57 minutes, P = .006; 68 ± 24 minutes vs 25 ± 19 minutes, P = .006, respectively). Mortality rates were 0% in the LG and 14% (2/14) in the OG (P = .2). Postoperative liver failure and ascites occurred in 8% (1/13) in the LG and 36% (5/14) in the OG (P = .15). Surgical margin was not different in the 2 groups. Three-year survival was significantly higher in the LG (89% vs 55%; P = .04), but 3-year recurrence rates were similar (46% vs 44%).
Our study suggests that, despite longer operative and clamping times without clinical consequences, the rate of decompensation of liver disease could be lower after laparoscopy.
HEPATOCELLULAR carcinoma (HCC) is the most common primary cancer of the liver. The disease is associated with a 5-year survival rate of less than 5% without treatment. Recently, its incidence has slightly increased in the United Kingdom,1 France,2 and the United States.3 Screening for HCC has become part of the treatment of patients with chronic liver disease (CLD), and small (<3-cm) asymptomatic HCCs have been increasingly recognized.4 Therapeutic options for small HCCs in patients with CLD include hepatic resection, percutaneous ablative techniques (ethanol injection and radiofrequency), and liver transplantation. Liver transplantation is the best theoretical treatment for these patients by removing both the tumor and underlying CLD. However, it cannot be applied on a large scale because HCC is a common condition; there is a donor shortage, and tumor progresses during the waiting time. Therefore, liver resection and percutaneous treatments, although associated with high recurrence rates, remain an important part of the treatment of these patients.
Our approach in patients with small HCCs is to propose, as a first step, percutaneous ablative techniques for lesions located deep in the liver parenchyma and liver resection with limited functional parenchymal sacrifice for small superficial lesions. Liver resection in cirrhotic patients has been limited by high morbidity rates caused by postoperative decompensation of underlying CLD. The advent of laparoscopic techniques may improve the treatment of such patients. In 1996, we began a prospective study of laparoscopic liver resections in selected patients,5 including HCC in patients with CLD since 1998. The aim of this study was to review our results of laparoscopic limited resection of subcapsular HCC in patients with CLD and, in an attempt to assess the relevance of this approach, to compare its results with those of a matched retrospective group of open liver resections.
In 1996, a prospective evaluation of laparoscopic liver resection was initiated. This was made possible by the coexistence in our department of a program of advanced laparoscopic surgery and a hepatobiliary and liver transplant unit.5 The first patients had liver tumors that developed on normal liver parenchyma, and after 2 years the indications were extended to patients with HCC and CLD. From December 1, 1998, to November 30, 2000, 13 patients with HCC complicating CLD, selected for liver resection, agreed to a laparoscopic approach. There were 3 women and 10 men, with a mean age of 63.0 ± 9.5 years (range, 47-74 years). The lesion was solitary in all patients. Lesions were subcapsular, localized in the anterior or lateral segment of the liver according to Couinaud classification5- 7 (ie, segments II-VI) (Figure 1). Patients, tumors, and liver characteristics are summarized in Table 1. These 13 patients formed the laparoscopic group (LG). Conversion to laparotomy occurred in 2 patients.
The following terminology was used: left lateral resection for bisegmentectomy of segments 2 and 3; segmentectomy for resection of 1 segment, according to its theoretical borders and defined by its number; and atypical resection for resection of less than 1 segment.
The surgical technique for laparoscopic liver resection in our unit has been described previously.5 Briefly, the procedures were performed with carbon dioxide pneumoperitoneum, and abdominal pressure was electronically maintained below 14 mm Hg. A 30° laparoscope was used in all cases. The liver was examined by vision and intraoperative ultrasonography to confirm the number and size of the lesions and define their relationship with the intrahepatic vascular structures. A tape was placed around the porta hepatis and passed through a 16F rubber drain for use as a tourniquet for portal triad clamping. Intermittent clamping was applied, with 15-minute clamping and 5-minute release periods. Hepatic transection was performed with a harmonic scalpel (Ultracision; Ethicon, Endo-Surgery, Issy-les-Moulineaux, France) and ultrasonic dissector (Dissectron; Satelec, Merignac, France). Bipolar electrocoagulation was used for minor bleeding, and larger structures were secured with clips. Portal pedicles and major hepatic veins were divided by application of a linear stapler. The resected specimen was placed in a plastic bag and externalized, without fragmentation, through a separate incision, along either a previous appendectomy incision or a new suprapubic horizontal incision. This incision was immediately closed and the abdomen reinflated. The surgical field was irrigated and checked for bleeding or bile leak, and residual fluid was removed by suction. Abdominal drainage was usually omitted, and the pneumoperitoneum was vented.
Patients with open liver resection were selected from our computer database of patients operated on from 1990 to 2000 by the same surgical team. For each patient in the LG, all patients who underwent open liver resection for subcapsular HCC with CLD matched for severity of liver disease (Child-Pugh criteria,8 esophageal varices, and fibrosis histologic score9), tumor size, and type of liver resection were identified by computer analysis and formed the open control group (OG).10 During the study period, 112 patients with HCC underwent open liver resection and 14 (12%) of these fulfilled the selection criteria. During the selection process, we were unaware of morbidity and mortality in the patients considered for the OG.
The 2 groups were compared on an intention-to-treat basis without exclusion of the 2 patients whose procedures were converted to laparotomy. The studied criteria were duration of surgery and clamping, perioperative transfusions, pathological margins, operative mortality and morbidity with special attention to postoperative ascites and liver failure (defined as occurrence of encephalopathy, total bilirubin level >3.5 mg/dL [>60 µmol/L], or prothrombin rate <30% of normal level), postoperative variations of results of liver function tests, and hospital stay.
Comparison between groups used the Fisher exact test or Mann-Whitney test for qualitative and quantitative variables, as appropriate. Survival and cumulative recurrence were estimated with actuarial methods. Difference was analyzed with the log-rank test. Significance was defined as P<.05.
As shown in Table 1, the 2 groups were well matched for preoperative criteria. Postoperative care was also identical, including perioperative antibiotic prophylaxis and fluid management. Two patients in each group received a single course of preresection transarterial chemoembolization. No other adjuvant or neoadjuvant treatment was used in either group.
Operative duration and cumulative clamping time were significantly longer in the LG by an average of 80 minutes (P = .006) and 40 minutes (P = .006), respectively (Table 2). The reasons for conversion to laparotomy in 2 patients in the LG were hemorrhage from a hepatic vein branch during segmentectomy (segment 6) in one case and insufficient exposure for adequate oncologic resection in the other. Conversion for hemorrhage was not associated with shock or need for highly urgent laparotomy, and the patient did not require transfusion.
One patient in the LG (8%) had intraoperative transfusion. Four patients in the OG (27%) received transfusions, but the difference did not reach significance.
Pathological margins were of similar size in the 2 groups, and 2 patients in each group had a margin less than 5 mm. In 1 patient in the LG, tumor contact occurred during transection, leading to conversion.
There were no deaths in the LG and 2 deaths in the OG (14%) (Table 3). These 2 patients had limited subsegmental resection with 30-minute continuous portal triad clamping and they received transfusions (4 and 10 units). They developed postoperative liver failure, and 1 experienced variceal bleeding requiring emergency endoscopic treatment. Both died of sepsis and multiple-organ failure.
In the LG, 4 patients (36%) had 5 complications. Ascites and transient liver failure occurred in 1 patient (8%) each. The patient with transient liver failure also developed nosocomial pulmonary infection (Legionella pneumonia) and required temporary tracheostomy. A second case of pulmonary infection occurred in a patient who was a heavy smoker and had chronic obstructive pulmonary disease; this patient required 6 days of mechanical ventilation. These 2 patients received 2 units of packed red blood cells in the postoperative period because of mild anemia. Both recovered completely. One obese patient underwent reoperation on day 8 for an incisional hernia occurring on a 12-mm port orifice. One of the 2 patients whose procedure had been converted to laparotomy developed transient postoperative ascites, while the other had an uncomplicated outcome.
In the OG, 7 patients (50%) had 11 complications. Ascites occurred in 5 cases (36%) and 1 case required repeated paracenteses. Liver failure occurred in 5 cases (36%), including the 2 deaths. In 3 cases, ascites and liver failure were associated. Differences from the LG were not significant. As shown in Figure 2, there were no differences in the laboratory test results between the 2 groups except for a higher aspartate aminotransferase peak in the LG, returning to normal value within the same time range. Mean hospital stay in the LG and the OG was 15.3 ± 8.6 days (median, 8 days; range, 5-69 days) and 17.3 ± 18.9 days (median, 12.5 days; range, 6-36 days), respectively.
Three patients in the OG underwent liver transplantation for recurrent HCC 2.3, 2.5, and 6.3 years after resection. They were not excluded from survival and recurrence analysis after LT. No patient in the LG had undergone transplantation at last follow-up. Overall 3-year survival was significantly higher in the LG (89% vs 55%; P = .04) (Figure 3A). Survival analysis with exclusion of the 2 postoperative deaths suppressed this difference (Figure 3B). Recurrence-free survival was similar between the 2 groups (Figure 3C). Liver recurrence occurred in 5 patients in the LG and in 7 in the OG, for a recurrence rate at 3 years of 44% and 46%, respectively. No patient, in either group, developed recurrence at the site of resection, and all intrahepatic recurrences occurred at a remote site. No port site recurrences were observed in the LG.
This study demonstrates the technical feasibility and safety of laparoscopic resection of small subcapsular HCCs. Furthermore, laparoscopy, despite a longer operative time, showed a trend toward reduced morbidity in comparison with open resection. This study was limited by the relatively small number of patients and the retrospective nature of the control group. Laparoscopic liver resection is an emerging technique that requires prospective evaluation, which we undertook in 1996.5 Since the beginning of our experience, we considered resection of HCC in patients with CLD a good case for laparoscopic approach, and our initial promising results in this indication prompted us to perform the present study. A prospective randomized study of open vs laparoscopic liver resections would be ideal, but the number of potential candidates with HCC is too limited to undertake a realistic study. This is why we designed this case-control study.
The procedure can be considered safe because there were no deaths and no unusual complications. It should be emphasized that laparoscopic liver resections require specific training and dedicated technology.5 A significant increase in operative time has been observed with laparoscopic procedures; this is, at least in part, due to a learning curve effect and should improve in the future.11 Laparoscopy is frequently subject to the potential adverse effects of pneumoperitoneum. The influence of pneumoperitoneum on intraoperative hemodynamics is still controversial.12 A recent prospective study from our unit showed the same hemodynamic tolerance of Pringle maneuver during laparoscopy with pneumoperitoneum and laparotomy.13 In addition, we have not observed clinically significant gas embolus in our experience with laparoscopic liver resection, despite monitoring with transesophageal echocardiography. This may be due to the high solubility of carbon dioxide. However, caution remains warranted.
Open hepatectomies, even limited ones, in cirrhotic patients are associated with a mortality rate of 6% to 12% and a morbidity rate of 10% to 40%.14,15 In this study, ascites occurred in 1 patient in the LG and in 5 in the OG; liver failure occurred in 1 patient in the LG and in 5 in the OG (8% vs 36%; P = .15). The only case of postoperative ascites in the LG occurred in a patient whose procedure was converted to laparotomy. Furthermore, 2 deaths occurred in the OG and none in the LG (14% vs 0%; P = .2). These results might be hampered by the retrospective nature of the control group, whereas laparoscopic cases were more recently performed. However, these were our first laparoscopic cases and therefore corresponded to our learning curve. In addition, this potential historical bias is reduced by the design of the study, leading to a control group similar to the LG at the time of hepatectomy for Child-Pugh classification and fibrosis score, size and subcapsular location of tumor, type of hepatectomy, and degree of portal hypertension. The 2 patients who died in the OG had portal hypertension, shown by the presence of esophageal varices, but a similar number of patients in the LG had esophageal varices. The 2 deaths in the OG occurred after 30-minute continuous clamping, but this duration remains moderate. In recent years, we have used intermittent clamping routinely for both open and laparoscopic procedures in patients with CLD. This is particularly important in laparoscopically treated patients because of longer and less predictable transection and, therefore, clamping times.
There are potential explanations for the apparent benefit of laparoscopy. Postoperative ascites is the most frequent complication of open hepatectomy in cirrhotic patients, even for limited resections.16 The main predictor of the development of postoperative ascites is the degree of preoperative portal hypertension.16 In the present study, severity of portal hypertension was comparable in both groups by the design of the study. The reduced incidence of postoperative ascites can therefore be attributed to the surgical approach. Potential mechanisms include preservation of the abdominal wall collateral circulation by avoiding long abdominal incisions and muscle division, and preservation of the round ligament, which may contain significant collateral veins. Other possibilities include less mobilization and manipulation of the liver during the laparoscopic approach, which might reduce liver trauma,17,18 less division of lymphatic channels,19 and less intraoperative fluid requirements. The reduced liver failure rate in the LG could also result from the same mechanisms. It should be emphasized that no major changes in postoperative care, particularly concerning fluid management, occurred during the study period. A significantly better 3-year survival rate was observed after laparoscopic resection, but survival was identical when the 2 postoperative deaths were excluded. Therefore, survival after laparoscopic resection was at least equivalent to that observed after open resection. Recurrence-free survival was the same, suggesting that tumor recurrence rate was not increased by the use of the laparoscopic approach.
Tumor clearance was macroscopically satisfactory in all but 1 case. Intraoperative exposure of tumor led to conversion to establish an adequate margin in 1 early laparoscopic case. Except for this case, the anticipated resection could be achieved in laparoscopic cases. Mean resection margins were identical in both groups (9 mm), and 2 patients in each group had a margin less than 5 mm. These margins reflect the subcapsular location of the lesions and the type of resection aiming at preserving functional parenchyma. It should be emphasized that no patient, in either group, developed recurrence at the site of resection, and that all intrahepatic recurrences occurred at a remote site. Although resection margin must remain a permanent concern during hepatectomy for cancer, its value as a prognostic factor has probably been overestimated. Indeed, several studies including large numbers of patients with HCC,20- 22 but also with cholangiocellular carcinoma23 or colorectal metastases,24,25 showed no negative effect on survival of a margin less than 10 mm. Clearance margin does not seem to affect oncologic results provided the resection is complete and tumor is not exposed.
Cirrhotic patients with single small HCCs and CLD with preserved liver function create a therapeutic dilemma for the clinician because of the various possible modalities. The reference treatment remains surgical resection, although it is seriously challenged by percutaneous ablative techniques. Both of these methods treat the tumor but leave the liver disease in place, which accounts for high recurrence rates due to multicentric carcinogenesis.26 Orthotopic liver transplantation is the only treatment that offers a prospect of cure, since it removes both the tumor and the underlying liver disease. In the case of small HCCs, it offers the best 5-year survival rate.27,28 However, because of the severe organ shortage, liver transplantation can be offered only to a limited number of patients. Recently, Majno et al,29 using a Markov-based decision analytic model, reported that even if liver transplantation is the best treatment, liver resection remains an effective means of treatment in view of growing transplant waiting lists. Moreover, a proper selection of candidates for resection promotes better results than transplantation, which is significantly hampered by the growing incidence of dropouts due to tumor progression resulting from the increasing waiting time.30 For patients with small HCCs (<3 cm) complicating CLD, an alternative strategy is salvage transplantation,29,31 which includes resection or ablation techniques first and transplantation if the tumor recurs or if the liver function deteriorates. Indeed, resection may offer a primary treatment of the tumor, leaving the possibility of a timely transplantation.
Percutaneous radiofrequency ablation is an efficient first-line procedure, but its use is limited in the case of subcapsular lesions by the risk of iatrogenic tumor seeding, recently confirmed by the Barcelona group.32 Although avoiding direct puncture of tumors with larger radiofrequency needles is recommended, this may be very difficult to achieve in cases of surface tumor. Therefore, our first-line approach in patients with small HCCs has been to perform liver resection with limited functional parenchymal sacrifice in cases of subcapsular lesions and to treat deeper lesions with percutaneous ablative techniques, avoiding direct tumor puncture.
In conclusion, laparoscopic resection of subcapsular HCCs in cirrhotic patients is safe but requires specific training, and it may have a higher burden on operating room ressources. It seems to offer advantages over open resection in terms of morbidity, the same or better 3-year survival, and a similar recurrence rate. Equivalent or better results observed with laparoscopy promote this approach as a promising avenue of research for resection of subcapsular HCCs, and further evaluation is warranted.
Corresponding author: Daniel Cherqui, MD, Department of Digestive Surgery, Hôpital Henri Mondor, 51, Avenue De Lattre de Tassigny, 94010 Creteil, France (e-mail: firstname.lastname@example.org).
Accepted for publication December 22, 2002.