Serum TSP2 level linked to advanced liver fibrosis in patients with T2DM and NAFLD

A recent study by researchers from the University of Hong Kong (HKU) reveals an association between serum thrombospondin-2 (TSP2) level and the presence and development of advanced (F3) fibrosis in patients with concomitant type 2 diabetes mellitus (T2DM) and non-alcoholic fatty liver disease (NAFLD), highlighting the potential for serum TSP2 level to be used as a prognostic biomarker for the development and progression of liver fibrosis in these patients. 

“Our study is the first to demonstrate the clinical relevance of serum TSP2 level as a fibrosis biomarker of NAFLD in T2DM using cross-sectional as well as prospective approaches,” said the authors. [Diabetes Care 2021;doi:10.2337/dc21-0131]

“Given its high negative predictive value [NPV, 92.1 percent] and significant improvement of area under the ROC curve [AUROAC] in identifying ≥F3 fibrosis [ie, hepatic fibrosis progression] on vibration-controlled transient elastography [VCTE], serum TSP2 test maybe particularly useful in identifying high-risk T2DM patients harbouring ≥F3 fibrosis in outpatient DM clinics where VCTE may not be readily available, for referrals to hepatologists for further evaluation,” they suggested. “Future studies with an increased sample size and extended follow-up period are encouraged to further validate our findings, and to explore the potential of modulating TSP2 level as a novel strategy to prevent hepatic fibrosis progression in T2DM.”

The researchers recruited 1,053 patients with T2DM from the Hong Kong West Diabetes NAFLD Cohort, who were regularly monitored at the diabetes clinic of Queen Mary Hospital. All patients received VCTE after fasting for ≥8 hours, as well as measurement of controlled attenuation parameter (CAP) and liver stiffness (LS) at baseline and every 12–18 months thereafter.

At baseline, 820 patients (77.9 percent) had hepatic steatosis (CAP ≥248 dB/m), of whom 138 (16.8 percent, mean age, 57.8 years; male, 52.2 percent) had ≥F3 fibrosis. Patients with vs without ≥F3 fibrosis had significantly increased median serum TSP2 level (4.17 ng/mL vs 2.33 ng/mL; p<0.001), while serum TSP2 levels were comparable between men and women (p=0.055).

Multivariable logistic analysis showed a significant association between baseline serum TSP2 level and ≥F3 fibrosis (odds ratio [OR], 5.13; 95 percent confidence interval [Cl], 3.16 to 8.32; p<0.001) in patients with ≥F3 fibrosis, with consistent association demonstrated in all cohort patients (OR, 5.57; p<0.001) and in the subgroup of patients (n=58) with VCTE performed using an XL probe (OR, 10.0; p=0.015).

Addition of serum TSP2 level to a clinical model consisting of body mass index (BMI) and aspartate transaminase (AST), two other independent determinants of ≥F3 fibrosis, led to significant improvements in AUROC by 0.03 (p=0.01), as well as integrated discrimination improvement (IDI) (8.1 percent; 95 percent CI, 5.0 to 11.0; p<0.001) and category-free reclassification improvement (NRI) (60.7; 95 percent CI, 53.1 to 78.3; p<0.001; 24.6 for events and 36.1 for non-events).

An optimal serum TSP2 cut-off level of 3.6 ng/mL on its own achieved sensitivity of 83.6 percent and specificity of 64.5 percent in identifying ≥F3 fibrosis. The positive predictive value and NPV were 44.3 percent and 92.1 percent, respectively.

At a median follow-up of 1.5 years, 43 of 491 participants (8.8 percent) who underwent VCTE reassessment developed ≥F3 fibrosis. Baseline serum TSP2 level was significantly associated with development of incident ≥F3 fibrosis (OR, 2.82; 95 percent Cl, 1.37 to 5.78; p=0.005), as were BMI (OR, 1.12; p=0.007), platelet count (OR, 0.992; p=0.010) and CAP (OR, 1.02; p=0.005).