Résumé
Survival after liver transplantation (LT) is 74% at 5 years and 63% at 10 years. Liver biopsy (LB) is the gold standard for post-transplant complications diagnosis. However, the use of LB as a monitoring tool is limited due to invasiveness, cost and risk of sampling error. In many indications, Vibration Controlled Transient Elastography (VCTE) has replaced biopsy for assessing liver stiffness. To date, some preliminary studies have demonstrated the feasibility and diagnosis performance of VCTE for the detection of fibrosis and steatosis following LT. However, there is no existing data regarding the value of VCTE for medium- and long-term follow-up after LT compared with a protocol biopsy in asymptomatic patients. The main objective of our study was to compare VCTE with protocol LB in LT patients. Methods: patients 5 to 10 years post-LT with both protocol LB and VCTE were included. Protocol liver allograft biopsies were defined as LB carried out at specific time points according to predetermined guidelines and performed in the absence of clinical or biological evidence of a post-LT complication. The biopsy was considered satisfactory if it included at least 8 portal tracts. VCTE was used to measure liver stiffness measurement (LSM) and steatosis via CAP (Controlled Attenuation Parameter) using two available probes (M or XL). The primary endpoint was the performance of the VCTE in predicting an abnormal LB. Isolated steatosis < 33% without another anomaly was not considered as abnormal biopsy. Results: among the 622 patients who underwent LT between 2012 and 2019 in our center, 138 individuals had both LB and VCTE assessment. 90 patients’ data could be analysed (excluding non-interpretable LB and VCTE). The majority of these patients were men (n=62, 68.9%), with a mean age of 60.7±11.2 years, and the main indications for LT were alcohol-related cirrhosis (n=37, 41.1%) and hepatitis B or C virus (n=19, 21.1%). The median time between LT and LB was 6.3 years (IQR 5-8) and the median time between LB and VCTE was 38 days (IQR 21.2-73.5). Of the 90 LB performed, 44 (48.8%) were normal and 17 (18.8%) showed steatosis with no other abnormalities. 29 LB (32.2%) were abnormal: acute cellular rejection (n=9), recurrence of initial disease (n=9), steatohepatitis (n=8), vascular disease (n=2). In total, 13 LB (45%) had fibrosis. Mean LSM was 8.07±8,6 kPa (range 3.1-74.5) with a mean CAP of 245±58 dB/m. Patients with abnormal biopsies had higher elastography (12.7±13.9 vs 5.9 ±1.6); p<0.001). Liver tests were more frequently abnormal in patients with high elastography: AST (28.8 (±14.4) vs 21.4 (±10.3); p=0.013), ALT (25.5 (±16.1) vs 19.2 (±13.1); p=0.043) total bilirubin (13.4 (±7.6) vs 8.9 (±4.3); p=0.001). There were no differences in clinical parameters, indication for LT or time since LT. Elastography (AUC: 0.816) had the best statistical performance in predicting an abnormal biopsy compared with ALT (AUC: 0.640) and total bilirubin (AUC: 0.712). For elastography, the threshold with the highest Youden index was 6.9 kPa. Elastography ≤5.6 kPa ruled out an abnormal biopsy (Se 100%, Sp 52.8%, NPV 100%, PPV 34.6%) and a value above 12.9 kPa predicted an abnormal biopsy (Se 22%, Sp 100%, NPV 83.7 %, PPV 100%). The best CAP threshold for the detection of steatosis was 285 dB /m (AUC: 0.892).Conclusion: for patients in 5 to 10-years post-LT period, elastography is a valuable tool to assess if patients should benefit from a protocol LB (>12.9kPa) or not (<5.6 kPa), particularly in the absence of biological abnormalities. These results underscore the appropriateness of employing VCTE in long-term LT recipients to determine the need for a LB.