Baseline volumetric multiparametric MRI: Can it be used to predict survival in patients with unresectable intrahepatic cholangiocarcinoma undergoing transcatheter arterial chemoembolization?

Purpose: To evaluate whether baseline MRI can help predict survival in patients with unresectable intrahepatic cholangiocarcinoma (ICCA) undergoing transcatheter arterial chemoembolization (TACE). Materials and Methods: This retrospective study was compliant with HIPAA and approved by the institutional review board. The requirement to obtain informed consent was waived. The study included 111 patients (mean age, 62 years 6 12; range, 29–86 years), with 44 men (mean age, 61 years 6 12; range, 29–81 years) and 67 women (mean age, 63 years 6 12; range, 34–86 years). Patients underwent TACE after baseline MRI, which included contrast-enhanced and diffusion-weighted imaging with apparent diffusion coefficient (ADC) mapping between 2003 and 2016. The single largest tumor was assessed independently by a radiologist for anatomic and functional (viable tumor volume, percentage viable tumor volume [100 × viable tumor volume/whole tumor volume], viable tumor burden [100 × viable tumor volume/whole liver volume], and ADC) parameters. Survival analysis was performed with Kaplan-Meier and Cox regression analysis. Results: Overall survival (OS) was higher with a baseline ADC of 1415 3 102⁶ mm²/sec or less compared with greater than 1415 3 102⁶ mm²/sec (P = .005; 25th percentile of OS, 17 months vs 7 months, respectively), percentage viable tumor volume greater than 90% compared with 90% or less (P = .001; 25th percentile of OS, 20 months vs 7 months, respectively), and viable tumor burden greater than 6.6% compared with 6.6% or less (P = .09; 25th percentile of OS, 17 months vs 7 months, respectively). Baseline ADC greater than 1415 3 102⁶ mm²/sec (hazard ratio [HR]: 2.176 [95% confidence interval: 1.217, 3.891]; P = .009) and percentage viable tumor volume greater than 90% (HR: 0.319 [95% confidence interval: 0.148, 0.685]; P = .003) were associated with OS independent of clinical confounders (age and sex). At multiparametric MRI risk stratification (with low ADC and high percentage viable tumor volume considered favorable for survival), differences in OS were noted (P = .002; 25th percentile of OS for low vs intermediate vs high risk, 22 months vs 10 months vs 7 months, respectively). Conclusion: Baseline multiparametric MRI assessment including volumetric ADC, percentage viable tumor volume, and viable tumor burden can help predict mortality risk among patients with intrahepatic cholangiocarcinoma undergoing transcatheter arterial chemoembolization.