Sequential ultrasound molecular imaging for noninvasive identification and assessment of non-alcoholic steatohepatitis in mouse models

Background and objective

Noninvasive non-alcoholic steatohepatitis (NASH) assessment is a clinical challenge to the management of non-alcoholic fatty liver disease. We aim to develop diagnostic models based on sequential ultrasound molecular imaging (USMI) for the noninvasive identification of NASH in mouse models.

Methods

Animal experiments were approved by the Animal Ethics Committee of South China Agricultural University. Forty-nine C57BL/6 mice were divided into normal control, non-alcoholic fatty liver, NASH, and hepatitis groups. Sequential USMI was implemented using CD36-targeted microbubbles (MBs-CD36) and intercellular adhesion molecule-1 (ICAM-1)-targeted microbubbles (MBs-ICAM-1) to visualize hepatic steatosis and inflammation. The targeting signal of USMI was quantified as the normalized intensity difference (NID) with the destruction-replenishment method. Correlation analysis was conducted between the NID-MBs-CD36 and pathological steatosis score and between the NID-MBs-ICAM-1 and pathological inflammation score. Finally, diagnostic models combining NID-MBs-CD36 with NID-MBs-ICAM-1 were established for NASH diagnosis.

Results

MBs-CD36 and MBs-ICAM-1 were successfully prepared and used for sequential USMI in all mice. NID-MBs-CD36 values increased with the progression of steatosis, while NID-MBs-ICAM-1 values increased in parallel with the progression of inflammation. A strong positive correlation was identified between NID-MBs-CD36 and pathological steatosis grade (r_s = 0.9078, P < 0.0001) and between NID-MBs-ICAM-1 and pathological inflammation grade (r_s = 0.9071, P < 0.0001). Among various sequential USMI-based diagnostic models, the serial testing model showed high diagnostic performance in detecting NASH, with 95% sensitivity, 97% specificity, 95% positive predictive values, 97% negative predictive values, and 96% accuracy.

Conclusions

Sequential USMI using MBs-CD36 and MBs-ICAM-1 allows noninvasive grading of hepatic steatosis and inflammation. Sequential USMI-based diagnostic models hold great potential in the noninvasive identification of NASH.