The procedure outcome of laparoscopic resection for 'small' hepatocellular carcinoma is comparable to vlaparoscopic radiofrequency ablation.

BACKGROUND: The aim of this study was to compare the effectiveness of laparoscopic liver resection (LLR) and laparoscopic radiofrequency ablation (LRFA) in the treatment of small nodular hepatocellular carcinoma (HCC). PATIENTS AND METHODS: We enrolled 50 cirrhotic patients with similar baseline characteristics that underwent LLR (n = 26) or LRFA (n = 24), in both cases with intraoperative ultrasonography. Operative and peri-operative data were retrospectively evaluated.
RESULTS: LLR included anatomic resection in eight cases and non-anatomic resection in 18. In LRFA patients, a thermoablation of 62 nodules was achieved. Between LLR and LRFA groups, a significant difference was found both for median diameters of treated HCC nodules (30 vs. 17.1 mm; P < 0.001) and the number of treated nodules/patient (1.29 ± 0.62 vs. 2.65 ± 1.55; P < 0.001). A conversion to laparotomy occurred in two LLR patient (7.7%) for bleeding. No deaths occurred in both groups. Morbidity rates were 26.9% in the LLR group versus 16.6% in the LRFA group (P = 0.501). Hospital stay in the LLR and LRFA group was 8.30 ± 6.52 and 6.52 ± 2.69 days, respectively (P = 0.022). The surgical margin was free of tumour cells in all LLR patients, with a margin <5 mm in only one case. In the LRFA group, a complete response was achieved in 90.3% of thermoablated HCC nodules at the 1-month post-treatment computed tomography evaluation.
CONCLUSIONS: LLR for small peripheral HCC in patients with chronic liver disease represents a valid alternative to LRFA in terms of patient toleration, surgical outcome of the procedure, and short-term morbidity.

INTRODUCTION

Both hepatic resection and interstitial therapies influence the natural history of hepatocellular carcinoma (HCC) by increasing the survival of patients with small-size nodules, but disease recurrence after either treatment remains an issue. Several studies have compared the outcomes of radiofrequency ablation (RFA) and hepatic resection, showing a trend towards better survival and significantly better disease-free survival compared with the RFA group.[123] Laparoscopic RFA (LRFA) has proved to be superior to the percutaneous approach in lesions that are difficult or impossible to be treated in such a way or severe liver disease.[4] Until date, thermoablation of HCC has always been compared to hepatic resection performed through an open approach, carrying the bias of comparing a non-invasive technique to a quite invasive one. In fact, open resection has been associated with longer operative times, higher blood loss and post-operative morbidity.[56] It is now widely accepted that laparoscopy may be used in liver surgery, by including patients with HCC and cirrhosis. Laparoscopic liver resection (LLR) of HCC has been proposed as a less invasive technique alternative to open resection and provided better post-operative results and similar recurrence rate when compared with the open counterparts.[78]

The aim of our study was to evaluate retrospectively the peri/post-operative outcomes of small peripheral HCC treated with LLR or LRFA, being these surgical approaches similar in terms of invasiveness.

PATIENTS AND METHODS

Patient Enrolment

Between June 2005 and November 2010, we evaluated 133 consecutive patients with an established diagnosis of HCC. After completing staging, 50 patients were enrolled in the study. The overall assessment of the severity of liver disease was made at the time of patient enrolment according the Child-Pugh classification.[9] Plasma levels of α-fetoprotein were measured. Patients with severe impairment in the pre-operatively assessed coagulation tests (platelets <40 × 109/L) were excluded from the procedure. A pre-operative assessment was performed in all patients with a spiral computed tomography (CT) and/or gadolinium-enhanced magnetic resonance imaging (MRI) to document the number, size, and the segmental location of all liver lesions. Only patients with ‘small’ HCC were taken into account. According to Yao et al., ‘small’ HCC was defined for a single HCC nodule ≤6.5 cm, or with ≤3 lesions when the largest of which was ≤4.5 cm.[10]

Patient Grouping

The choice of LLR or LRFA was based on the site of the tumour: A LLR was accomplished when the site of the tumour was located in a resectable segment, whereas an LRFA was performed when the tumour was deeply located, requiring major hepatic resection. Furthermore, LRFA permitted treatment of deep-sited lesions with very difficult or impossible percutaneous approach.

Laparoscopic Liver Resection Group

Twenty-six LLR for HCC was retrospectively reviewed. Selection criteria for the laparoscopic approach in HCC patients included the following criteria:

Well compensated Child's class A/B cirrhosis;

Oesophageal varices ≤grade 2;

Platelet count ≥40 × 109/L and/or international normalized ratio >1.5;

'small’ lesion accessible to the laparoscopic approach (i.e., located in the peripheral segments of the liver) and treatable by limited resection (<3 segments).

Only one patient was operated on in an emergency situation for a haemoperitoneum from a ruptured HCC located in segment 3. Most resections were intended to be anatomic (i.e., resection of ≥1 anatomic segments) in order to resect the portal territory of the tumour. Non-anatomic resections included the tumour and an intended 1-cm tumour-free margin.

Laparoscopic Radiofrequency Ablation Group

Twenty-four LRFA were retrospectively reviewed. Patients enrolled had at least one of the following criteria:

Large tumours but with a diameter <5 cm, or multiple lesions requiring repeated punctures;

Superficial lesions adjacent to visceral structures that could be displaced by laparoscopic manoeuvres;

Deep-sited lesions with a very difficult or impossible percutaneous approach.

Exclusion criteria were the same as for the LLR group. However, main portal branches thrombosis or severe liver disease (Child-Pugh class C) did not contraindicate the procedure.

Surgical Treatment and Follow-up

The surgical technique adopted for LLR and LRFA was described elsewhere.[1112] The liver was examined by vision and intraoperative laparoscopic ultrasonography (LUS) to confirm the number and size of the lesions and define their relationship with the intra hepatic vascular structures. A definitive histological diagnosis of both HCC and liver cirrhosis has been obtained by an intraoperative biopsy of all patients undergoing LRFA or by resected liver specimen in LLR group. In this case, the surgical margins, together with histological characteristics of the lesion were evaluated as well. Liver ultrasound and CT (and/or MRI) were performed within 1-month after treatment to assess the response to LRFA or LLR. A complete response to LRFA was achieved when no enhancement or a thin peripheral enhancement rim (representative of an inflammatory response) was observed, while an incomplete response to LRFA was defined as persistent nodular enhancement.

Statistical Analysis

Descriptive statistics for continuous variables were expressed as mean values ± standard deviations. Continuous variables were compared by Wilcoxon rank sum test. Categorical variables were entered in contingency tables and then evaluated by Fisher exact test. Univariate logistic regression was performed by assuming morbidity as dependent variable. Statistical significance was assumed for P < 0.05 with a two-tailed null hypothesis. Statistical analysis was performed using the R software/environment (R Foundation for Statistical Computing, Vienna, Austria). R is an open source project that is distributed under the GNU General Public License (Free Software Foundation, Inc., Cambridge, Massachusetts, US). At the time of this writing, R version, 3.0.2 was available.

RESULTS

Table 1 summarizes the demographics of the LRH and LRFA groups. The two groups were similar regarding age, sex, and the aetiology of cirrhosis. Size of dominant lesion expressed as median diameter was significantly larger (P < 0.001) in the LLR (30 mm) versus. The LRFA group (17.1 mm), whereas the number of multiple lesions was significantly higher in the LRFA group [P < 0.001; Table 1]. As a result, also the mean number of treated nodules per patient was significantly greater (P < 0.001) in the LRFA group (2.65 ± 1.55) than in the LLR group (1.29 ± 0.62). Child class and oncologic staging were more advanced in the RFA group. Eighty-eight percent of patients in the LLR group were child class A compared with 54.1% in the LRFA group. Anatomical resection based on the segmental division of the liver was performed in eight cases (30.7%), whereas a US-guided wedge resection was chosen in 18 cases. The type of liver resections is given in Table 2.

Table 1

Characteristics of patients grouped for laparoscopic procedure

Table 2

Liver resection procedures

In the resection group, the lesions were mostly located in the anterolateral segments of the liver (73%), whereas the majority of LRFA patients had multiple lesions located in both anterolateral and posterosuperior segments of the liver (54.1%) [Table 3]. No significant difference in HCC diameter (P = 0.175) was found between LLR and LRFA groups. Intraoperative LUS identified a larger number of new malignant lesions missed at pre-operative imaging in the LRFA group when compared with the resection group (13 vs. 3, P < 0.001). Sixteen additional procedures (mainly cholecystectomies) were associated in 13 (50%) LLR patients, whereas six other procedures were associated in 5 (20.8%) LRFA patients (P = 0.042).

Table 3

Surgical-related variables grouped for laparoscopic procedure

Operative time significantly differed between LLR and LRFA procedures (297 ± 120 vs. 147 ± 55 min; P < 0.001), as well as blood loss [Table 3]. Twelve patients required blood transfusion only in the LLR group. A conversion to laparotomy was required in two LLR patients (7.69%) due to important continuous bleeding during parenchymal transection while no conversion occurred in the LRFA group. Length of hospital stay was significantly higher in LLR group than in LRFA group (8.30 ± 6.52 vs. 6.52 ± 2.69 days; P = 0.022). Although a major morbidity occurred in the LLR group versus LRFA group (26.9% vs. 16.6%), no statistical significance was reached (P = 0.501). By entering the laparoscopic procedure as independent variable in a univariate logistic regression model for morbidity, no statistical significance was found (P = 0.385). No deaths occurred in both LLR and LRFA groups. Complications in the LRFA group included transient liver failure in two patients, transient fever with abdominal pain in two patients and the formation of liver abscess on a thermoablated nodule requiring percutaneous drainage in one. Eight complications occurred in 7 (26.9%) LLR patients. They included transient liver failure (n = 3), intraabdominal collections including biliary collection drained percutaneously (n = 1), or seroma (n = 3) and ascites well controlled by diuretics (n = 1). In the LLR group, the median size of the resected tumour was 31.5 mm, median surgical margin was 13.5 mm and exceeded 5 mm in 24 patients (92%). In the LRFA group, a complete response with a total tumour necrosis was achieved in 56 of the 62 thermoablated HCC nodules (90.3%) at the 1-month post-treatment CT evaluation.

DISCUSSION

In recent years, the progress of laparoscopic procedures and the development of dedicated technologies have made endoscopic hepatic surgery feasible and safe.[56] The laparoscopic approach to RFA has proved to be superior to the percutaneous approach in lesions that are difficult or impossible to be treated in such a way or severe liver disease.[13] Analogously, the laparoscopic approach to resection of HCC may be considered in selected cases. In fact, the use of this technique has been limited to easily accessible lesions, namely, tumours in the peripheral portion of the liver's anterolateral segments (segments 2, 3, 5 and 6, and the inferior part of 4 according to the classification of Couinaud).[14] Accordingly, many surgeons consider that lesions located in the posterior or superior part of the liver (segments 1, 7, and 8, and the superior part of 4) are not appropriate because of the limited visualization and the difficulty of controlling bleeding.[1516]

Some reports appeared showing similar operative outcomes and oncological results when comparing laparoscopic resection of posterosupeior and anterolateral HCC nodules.[17] In this series, we have laparoscopically also approached lesions located in difficult liver portions in 27% of the cases. In LRFA patients, lesions located in the AL segments were pre-dominant, but in case of multiple nodules a posterosuperior lesion was treated in more than half of patients.

Different indications to LRFA and to LLR are responsible for the slightly different patients characteristics of the two groups. In fact, patients with normal liver function and larger tumours were resected, whereas those with liver dysfunction, multiple tumours, and portal hypertension were ablated. As expected, operative time and blood loss significantly differed between LLR and LRFA procedures, although additional procedures such as cholecystectomies were carried out mainly on LLR patients. Furthermore, median blood loss in the resection group, if the two converted patients are not taken into account, was extremely low, thus representing a good result. The conversion rate reported in our study for LLR patients was comparable to the previous series.[5] Until date, thermoablation of HCC has been always compared to hepatic resection performed through an open approach, carrying the bias of comparing a non-invasive technique to a quite invasive one.[1819] Significant differences in terms of morbidity are present between the two aforementioned techniques reporting a 32-33% rate of post-operative complications for open resections of HCC. Buell et al. reported the only multicentre study where RFA and resection of HCC, both performed through a laparoscopic approach, were compared.[6] They reported a morbidity rate in the LLR and LRFA group of 29% and 13.5%, respectively, without reaching statistical significance.[6] Furthermore in our study, morbidity was higher in the LLR (26.9%) than in the LRFA group (16.6%), with no statistical difference between the two procedures.

In the literature, when considering the complications of the open resection in details, ascites represents the main concern in half of the patients.[19] The laparoscopic approach dramatically reduces this occurrence to 5%.[17] In our series, only one patient (3.8%) suffered from ascites medically treated. The low incidence rate of post-operative ascites is ascribed to the preservation of abdominal collateral venous circulation, differently from what it happens when a subcostal incision is used.[219] Another advantage of laparoscopy over the open approach is the avoidance of peritoneal adhesions, valuable in these patients who may require repeat operations for tumour recurrence or undergo liver transplantation. Owing to this reasons, Rao et al. suggested extending the indication for LLR to selected Child B patients.[20] In our series, the oncologic results of LLR in terms of surgical margin are good and comparable to open and laparoscopic counterparts, being free of tumour cells in all patients, with a margin of <5 mm in only one case. In the LRFA group, persistence of tumoral disease was present in 6 of the 62 thermoablated HCC nodules (9.7%) at the 1-month post-treatment CT evaluation. Local recurrence at the RFA treatment site is not infrequent, and its incidence ranges from 5.7% to 39%.[21] Local failure may be attributed to several factors, including tumour characteristics and thermoablation techniques. Since data from literature report that LLR of HCC can achieve the same long-term oncologic results that the open approach and, in general, superior to thermoablation, a larger use of this approach should be considered.

Laparoscopic liver resection for small peripheral HCC in patients with chronic liver disease represents a valid alternative to LRFA in terms of outcome of the procedure and short-term results. It offers a minimally invasive approach to resection more similar to LRFA in terms of patient toleration. Furthermore, it also can be proposed in tumour locations with a difficult surgical access for maintaining a low morbidity rate.