Aim: To establish the influence of different factors on the correlation between liver stiffness (LS) measurements by ARFI and liver fibrosis, evaluated by liver biopsy (LB). We assessed the following factors: the success rate (SR) and interquartile range (IQR) interval, the place where ARFI was performed, liver steatosis, the quality of the specimen obtained by LB.
Methods: We studied 471 patients: 82 with LB, 82 healthy volunteers and 307 with cirrhosis. We performed 10 valid ARFI measurements, a median value was calculated, expressed in meters/second.
Results: Valid measurements were not obtained in 11 patients. There was a direct, strong, correlation (r=0.694) between ARFI and fibrosis (p<0.0001). There were no statistically significant differences between the mean ARFI values obtained in segments V vs. VIII (p=0.89). Considering the IQR and SR, the correlation of ARFI with fibrosis was: for IQR<30% and SR ≥ 60%: r=0.722 and for IQR>30% and/or SR ≤ 60%: r=0.268 (p=0.0001). The quality of the liver specimen (2-3 cm long vs. >3 cm) did not influence the correlation of ARFI with fibrosis.
Conclusions: To obtain the best correlation between ARFI and fibrosis, IQR must be <30% and SR ≥ 60%. These technical parameters must be introduced to improve the ARFI value for LS evaluation.