Objective : Cone-beam computed tomography (cone-beam CT) has become standard practice for tumor targeting during radioembolization procedures. Intraprocedural parenchymal blood volume (PBV) assessment using cone-beam CT is a new technique that might be useful for patient selection as well as in assessing adequacy of liver cancer treatment intra-procedurally. The purpose of this study is to investigate the baseline prognostication and subsequent changes in tumor and liver parenchymal blood volume (PBV) derived from intraprocedural cone-beam CT, correlating these findings to early tumor response following radioembolization. We hypothesized that higher baseline PBV might be associated with better response and there would be no significant reduction in PBV post radioembolization given its microembolic in nature.
Methods : Between November 2016 and December 2018, 20 patients with hepatocellular carcinoma (HCC) undergoing radioembolization with glass microspheres were enrolled in the study. This subanalysis includes nine patients who had appropriate baseline and post-radioembolization PBV cone-beam CT studies for evaluation. Cone-beam CT was performed on an Artis zee flat-panel detector angiography system (Siemens, Forchheim, Germany). The liver perfused by the injected artery was imaged in the steady-state by obtaining a two-sweep cone-beam CT during a single breath-hold. The mask was acquired initially. The contrast injection was initiated followed by a 6-8 second pause, and a 6-second cone-beam CT acquisition during the pseudo-steady state of parenchymal enhancement. PBV maps (mL per 1000 mL) were generated using the arterial input function of the injected artery. Tumor and adjacent non-tumoral parenchyma PBVs were compared before and after Y90 radioembolization using the Wilcoxon signed-rank test. We also used the Wilcoxon signed-rank test to compare the ratio of PBV values between tumors and adjacent parenchyma. The Mann-Whitney test was used to compare between PBV maps of patients who showed treatment response and patients who did not respond to treatment within 6-month of follow-up. Tumor Response was assessed at 1, 3 and 6 months using mRECIST criteria.
Results : All nine patients (6 males) successfully completed perfusion dynaCT before and after radioembolization. Median age was 62 (range: 46-70) years. Five patients were Child-Pugh A, while 3 & 1 were child-pugh B and C respectively. Four patients were BCLC A, while 1, 4 and 1 were BCLC B, C and D respectively. Mean tumor parenchymal blood volume (PBV) was 176 (Range: 32.6 – 352) mL/1,000 mL, which was significantly reduced after Y90 yielding a mean tumor PBV post-Y90 of 108.5 (range: 5.6 – 209) mL/1,000 mL. (P=0.0039) Meanwhile, PBV of the adjacent parenchyma significantly increased from a mean of 89.8 pre-Y90 to a mean of 165.4 post Y90 (P=0.0078). The ratio between tumor PBV to adjacent parenchyma significantly decreased after Y90, from a mean of 6.2 pre-Y90 to a mean of 0.3 post-Y90 (p=0.0039).
Three patients exhibited complete response within 6-month from Y90, 3 patients showed partial response while the other 3 patients maintained stable disease. When comparing tumor PBV of patients who exhibited partial or complete vs those who had stable disease, patients who showed response to treatment had significantly higher baseline tumor PBV than those who did not respond (p=0.02). Also, when comparing tumor PBV post-Y90 of responders vs non-responders, patients who responded had significantly higher PBV than those who failed to show response within 6 months of treatment (p=0.039).
Conclusions : The results of this study demonstrate the technical feasibility of employing PBV on Cone-Beam Computed Tomography for patients with HCC undergoing radioembolization. Patients who responded to radioembolization had higher baseline tumor PBV, possibly due to the inherent advantage of tumor hypervascularity for arterial-delivered therapy. Surprisingly, PBV values decreased significantly immediately following radioembolization. This was associated with increased PBV values in the adjacent non-tumor parenchyma. These findings need further validation.