Unresectable hepatocellular carcinoma: MR imaging after intraarterial therapy. part ii. response stratification using volumetric functional criteria after intraarterial therapy

Purpose: To assess whether volumetric functional magnetic resonance (MR) results 3-4 weeks after initial intraarterial therapy can aid accurate distinction between responders and nonresponders, to determine whether overall survival (OS) is improved, and to compare volumetric functional MR response with anatomic response criteria (Response Evaluation Criteria in Solid Tumors [RECIST], modified RECIST [mRECIST], European Association for the Study of the Liver [EASL]), as well as a1-fetoprotein [AFP] level. Materials and Methods: In this single-institution HIPAA-compliant retrospective, institutional review board-approved study, informed consent was waived; 143 patients with hepatocellular carcinoma underwent intraarterial therapy between October 2005 and February 2011. Volumetric functional MR response (25% or more increase in apparent diffusion coefficient, 65% or more decrease in enhancement) was stratified as follows: Dual-parameter responders fulfilled both criteria, single-parameter responders fulfilled one criterion, and those with stable disease (SD) fulfilled neither. RECIST, mRECIST, EASL, and AFP response criteria were determined. Kaplan-Meier technique, log-rank tests, and the Cox proportional hazards model were used to test whether OS was different per response. Results: OS differed significantly between single-parameter responders and dual-parameter responders (P = .01) and between singleparameter responders and those with SD (P = .001). Dualparameter responders' response improved OS compared with single-parameter responders; risk of death decreased (hazard ratio [HR] = 0.28, P = .01). In those with SD compared with single-parameter responders, risk of death increased (HR = 2.09, P = .001). RECIST, mRECIST, and EASL stratification was short of significant; most lesions were classified as SD. Baseline AFP level increased in 55 patients; AFP responders versus AFP nonresponders had decreased risk of death (HR = 0.36, P = .002). Agreement between anatomic response criteria and volumetric functional MR findings (k = 0.06-0.12) and between AFP response and imaging criteria (k = 20.04 to 0.14) was low. Conclusion: Volumetric functional MR response 3-4 weeks after initial intraarterial therapy showed improved OS. Volumetric functional MR was superior to current imaging (RECIST, mRECIST, and EASL) and biochemical (AFP level) response criteria.