On the reliability of clinical criteria for the diagnosis of hepatic veno-occlusive disease.

Among 217 patients who received an allogeneic (136 cases) or autologous (81 cases) bone marrow transplant, the diagnosis of hepatic veno-occlusive disease (VOD) was established in 38 according to Seattle clinical criteria. Thirty-two underwent a transjugular liver biopsy and measurement of the hepatic venous pressure gradient (HVPG). The study was completed in 30 patients with no serious complications. Hepatic VOD was histologically confirmed in 18 patients (60%); the remaining 12 were classified as non-VOD. An increased HVPG discriminated well between VOD and non-VOD cases. Thus, hemodynamic data can considerably reinforce the accuracy of histological diagnosis. The predictive value of two vs. three clinical data of the Seattle criteria was analyzed. Among the 19 cases fulfilling two clinical data VOD was confirmed in only eight (42%), whereas VOD was proved in ten of 11 cases (91%) (p = 0.02) suspected on the basis of three clinical data. When reliability of the Baltimore clinical criteria was analyzed, the result was identical to that observed when three Seattle clinical data were present. The specificity of the latter classification was high (92%) while its sensitivity was relatively low (56%). In conclusion, clinical criteria are not reliable for either recognizing or excluding the diagnosis of VOD. Thus, a transjugular liver biopsy, associated with hemodynamic evaluation, is strongly recommended when VOD is clinically suspected.