Patients
In this prospective study,
24 patients having undergone 25 hepatic resections between September 2016 and September 2017 were included : 11 men and 13 women.
Their ages ranged from 36 to 78,
with a mean age of 56 years.
There were 9 patients with liver metastases,
9 patients with hydatid cysts,
3 patients with cholangiocarcinomas,
3 patients with bile duct injuries and one patient with a hepatocellular carcinoma (HCC).
All IOUS and surgeries were performed at the general surgery department of Mahmoud Matri Hospital (Ariana,
Tunisia).
Only laparotomic surgeries were included.
Patients with incomplete preoperative imaging studies were excluded.
Imaging protocol
For each patient,
a preoperative computed tomography and/or magnetic resonance imaging was performed to determine the initial surgical strategy.
CT scans were performed using a 16 slice CT scanner (General Electric BrightSpeedTM ,
GE Healthcare,
USA).
The protocol consisted of at least 3 series: a non-enhanced CT,
an early arterial phase using the automatic bolus tracking technique and a late portal phase (70-80s).
In some patients,
late arterial and delayed phases were added for a better lesion characterization.
MRI were performed using a 1.5 Tesla system (Philips IngeniaTM ,
Philips Healthcare,
Netherland).
The protocol consisted of T2-weighted and fat-suppressed T2-weighted fast spin-echo sequences,
DIXON sequences,
diffusion-weighted images with b-values of 0,
150,
600 and 800 s/mm2 and T1 weighted post Gadolinium sequences.
IOUS was performed in all patients by an experimented radiologist competent at hepatobiliary imaging using an Esaote MyLab™ Seven (Esaote,
Genova,
Italy) ultrasound machine (Figure 1). Two types of dedicated and sterilized probes were used : a standard 1- 8 MHz convex probe and a 3-11 MHz linear T-shaped probe (Figure 2).
Data analysis
Data were analyzed by a senior radiologist competent at hepatobiliary imaging and a radiology resident.
The size,
location and number of lesions were compared between IOUS and preoperative imaging as well as their association with the main vascular structures.
Vascular and biliary variants have been specified for each patient in both preoperative et intraoperative imaging.
We also compared the initial surgical strategy with the operative report to determine if IOUS changed the surgical plan.
Statistical analysis
Statistical analysis was performed using SPSS 22.0 (SPSS,
Inc.,
Chicago,
IL,
USA) software.
The results were expressed as means,
medians and percentages.