Laparoscopic hepatectomy in cirrhotic patients: Safe if you adjust technique

David J Worhunsky, MD, Monica M Dua, MD, Bernard Siu, BS, George A Poultsides, MD, MS, Jeffrey A Norton, MD, Brendan C Visser, MD. Stanford University Medical Center.

Laparoscopic liver resection (LLR) in patients with cirrhosis adds significant complexity. Parenchymal transection is challenging due to fibrosis and portal hypertension. This study reports our initial experience with LLR in this high-risk population and evaluates whether comparable outcomes can be achieved to non-cirrhotic patients.

Methods and Procedures:
Patients undergoing LLR by a single surgeon between 2007 and 2013 were reviewed. Patients with cirrhosis were compared to those without to examine differences in demographics, surgical technique, intraoperative characteristics, and outcomes. Variables assessed included morbidity scores, liver-related complications, length of stay, and mortality.

A total of 127 patients underwent LLR during the study period. Forty-two (33%) had cirrhosis, of which 36 (86%) had hepatitis C. Thirty-three (79%) were Child’s A, with a median MELD of 9 (range: 6-21). Compared to non-cirrhotics, patients with cirrhosis were older with more comorbidities, higher ASA scores, and more likely to have hepatocellular carcinoma (Table). Four patients with cirrhosis underwent major hepatectomy (≥3 segments); however, the ratio of major to minor hepatectomy was similar between groups. Precoagulation before parenchymal transection was used more frequently in cirrhotic patients (69% vs 21%, P<0.001) and portal triad clamping was longer (mean 32 vs 23 minutes, P=0.03). Blood loss and rates of transfusion were greater in patients with cirrhosis though the absolute difference remained small (Table). Although there was a trend toward increased minor complications in cirrhotic patients (Clavien I-II), liver-related morbidity, major complications and mortality rates were similar.

LLR is safe for selected patients with cirrhosis but requires technical modifications, including precoagulation and liberal use of portal triad clamping. Despite greater pre-operative comorbidities in this patient population and the technical complexity presented by cirrhosis, outcomes comparable to non-cirrhotics can be achieved.

 CirrhosisNo CirrhosisP value
Age†60 (50-80)56 (23-84)0.029
Male sex39 (93%)42 (49%)<0.001
ASA score‡3.0 (0.4)2.7 (0.5)0.001

Cardiovascular disease

30 (71%)36 (24%)0.003
Diabetes11 (26%)10 (12%)0.046








37 (88%)

1 (2%)

4 (10%)


18 (21%)

3 (4%)

50 (60%)

13 (15%)

1 (1%)







Tumor size†  (mm)3.0 (1.4-9.0)2.4 (0.3-12.7)ns
Minor hepatectomy36 (86%)68 (80%)ns
Major hepatectomy4 (10%)15 (18%)ns
Precoagulation29 (69%)18 (21%)<0.001
Duration of portal triad clamping‡  (min)32 (17)23 (13)0.031
Operative time†  (min)230 (89-528)205 (55-562)ns
Blood loss†  (mL)200 (10-2000)100 (10-800)<0.001
Transfusion4 (10%)1 (1%)0.041
Mortality – 90-day1 (2%)ns

Liver-related morbidity

  Bile leak

  Liver failure


1 (2%)





Clavien Score I-II13 (31%)13 (51%)ns
Clavien Score III-V2 (5%)1 (1%)ns
Hospital stay‡  (d)4.7 (3.9)3.3 (2.4)0.016

Categorical variables: number (%)

† median (range)

‡ mean (standard deviation)

ASA, American Society of Anesthesiologists; HCC, hepatocellular carcinoma; CRLM, colorectal liver metastasis; NCRLM, noncolorectal liver metastasis; CC, cholangiocarcinoma

« Return to SAGES 2014 abstract archive