Can decision analysis help in the management of giant hemangioma of the liver?

We explore the trade-off between the risk and expected benefits from resection of giant liver hemangioma (GLH), one larger than 4 cm in diameter. We searched the English-language literature for studies of the postoperative mortality after resection of GLH and of the outcome of expectantly treated patients. The results of the data synthesis showed an operative mortality of 1.5% (90% confidence intervals, 0.1-3.0%) and an early surgical morbidity of up to 13%. Mortality among the 37 reported cases with ruptured GLH was 78%. There were no cases of spontaneous or traumatic rupture of unresected GLH during a follow-up of a total of 282 patient years. The main source of ambiguity regarding the management of GLH is the uncertain risk of its rupture. Rough estimates of this risk based on published data suggest that surgical resection is not justified in asymptomatic GLH. Yet, although rare, rupture of GLH does occur with disastrous consequences. Future research may attempt to define patient subsets whose GLHs are at higher risk of rupture, and in whom preventive resection may improve survival.