Cone-Beam Computed Tomography Tumor Enhancement Predicts Treatment Response and Survival in Patients with Hepatic Malignancies Following Hepatic Artery Embolization

Purpose: To examine the prognostic value of cone-beam computed tomography (CBCT)–based tumor enhancement parameters at predicting treatment response and overall survival after hepatic artery embolization (HAE)

Materials and Methods: Thirteen patients with 29 primary or secondary hepatic tumors were enrolled in this study and underwent bland embolization. All patients received preprocedural computed tomography (CT) and magnetic resonance imaging (MRI). All patients underwent intraprocedural pre- and postembolization subtracted (noncontrast and contrast-enhanced) CBCT. Semiautomatic contouring of the target tumor(s) combined with automatic extraction of the perfusing vascular tree allowed quantification of the pre- and postembolization tumor enhancement adjusted by the arterial input function (before and after adjusted tumor enhancement [ATE]). The decrease of ATE after embolization was subsequently calculated (ATE). All patients had 1-month CT or MRI for assessment of treatment response using Modified Response Evaluation Criteria in Solid Tumors (mRECIST) criteria. For analysis purposes, tumors were grouped into complete response (CR) and noncomplete response (non-CR) group. Local progression-free survival (PFS) and overall survival (OS) were calculated from the time of first embolization. A receiver operator curve (ROC) of ATE relative to CR was performed to analyze the sensitivity and specificity of our method and determine the most appropriate ATE cut-off.

Results: A total of 17 of 29 (58%) tumors showed CR, and 12 of 29 (42%) showed non-CR (8 of 29 [27.5%] partial response [PR], 1 of 29 [3%] stable disease [SD], and 3 of 29 [10.3%] progressive disease [PD]). Pre-ATEs were 38.5% ± 10.6% in the CR group and 30.4% ± 11.0% in the non-CR group (P = 0.023). ATE in the CR group was 38.7% ± 12.1% after embolization compared with 29.6% ± 11.1 in the non-CR group (P = 0.009). Because ATE was found to be predictive of favorable treatment response, for survival analysis, the tumor with the lowest ATE (minimum ATE) was selected for each patient. Patients with a minimum ATE greater than 33% have a longer time to local recurrence or progression (TTR or TTP) (P = 0.04) and a longer median overall survival period (P = 0.04).

Conclusions: This limited study suggests that subtraction CBCT-derived tumor enhancement parameters might be a useful method to predict treatment response, TTR or TTP, and OS after HAE. The magnitude of decrease in ATE is a significant predictor of CR. ATE is a significant predictor of longer TTR or TTP and longer median OS.