Prospective Comparison of Transient Elastography Using Two Different Devices: Performance of FibroScan and FibroTouch.

PURPOSE

Transient elastography (TE) using FibroScan (FS) has been established to non-invasively assess liver fibrosis and steatosis. The aim of this study was to compare the recently introduced FibroTouch (FT) device with the established FS with respect to liver stiffness and CAP.

PATIENTS AND METHODS

Thirty-nine patients with and without liver disease were included. All patients were measured three times with FS (FibroScan 530 compact, Echosens, France) and FT (FibroTouch-FT100, Wuxi Hisky Med, China). For FS, M and XL probe were used according to the manufacturer's specifications. For steatosis, CAP and the comparable FT equivalent UAP (ultrasound attenuation parameter) was determined. Finally, FT and FS were explored in liver tissue-mimicking phantoms.

RESULTS

LS between FS and FT correlated well with r=0.91. Root-mean-square (RMS) of the coefficient of variation for LS was better in FS (11.1% vs 27.4%). Bland-Altman analysis showed a 3.1 kPa mean overestimation of LS by FT. In addition, UAP strongly and linearly depended on the BMI following UAP=3.02 × BMI+186. In phantoms, a similar relation was found with UAP (phantom)= 3.78 × BMI + 146 suggesting that UAP is directly calculated from entered BMI instead of assessing shear-wave attenuation. Consequently, RMS-CV was lower for FT (6.0% vs 9.7%). However, if using different BMI, CV-RMS for FT increased to 12.7%. LS of a patient with manifest liver cirrhosis and ascites was 38.8 kPa using the FS-XL probe but almost normal with FT (7.2 kPa).

CONCLUSION

Although LS by FT shows good correlation with LS-FS, it has larger variation, continuously overestimates LS and completely fails in ascites. Moreover, FT-UAP seems to be a misleading parameter for steatosis assessment because it is at least in part calculated from mandatory entered patient data. In conclusion, novel LS cut-off values need to be defined for LS-FT and usage of UAP is not recommended.