SIR 2024
Interventional Oncology
Kurt Pianka, BS (he/him/his)
Medical Student
University of California San Diego
Financial relationships: Full list of relationships is listed on the CME information page.
Mark Barahman, MD, PhD
Resident Physician, Radiology
University of California San Diego
Disclosure information not submitted.
Claudio B. Ghetti, BS
Medical Student
University of California San Diego
Disclosure information not submitted.
Jeet Minocha, MD
Division of Interventional Radiology
University of California San Diego
Financial relationships: Full list of relationships is listed on the CME information page.
Jonas Redmond, MD
Associate Professor of Clinical Radiology
UCSD
Financial relationships: Full list of relationships is listed on the CME information page.
Gabriel Schnickel, MD
Faculty hepatobiliary and transplant surgery physician
University of California San Diego
Disclosure information not submitted.
Zachary Berman, MD
Assistant Clinical Professor
University of California, San Diego
Disclosure information not submitted.
The absorbed dose for transarterial radioembolization (TARE) treatment with glass microspheres is traditionally calculated using the Medical Internal Radiation Dose (MIRD) formula, which assumes uniform distribution to the perfused target area. This study correlates the tumor-specific absorbed dose and dose-volume histograms with tumor explant pathology in order to identify dose thresholds to predict pathologic complete response.
Materials and Methods:
All patients with HCC treated with Y-90 TARE using glass microspheres at a single institution between 2015 – June 2023 who then underwent liver transplant were eligible. The Y-90 distribution and dose-volume histograms were determined using the commercially available software Simplicity90 (Mirada Medical, Oxford UK). A complete response was assigned if explant pathology showed complete necrosis and the patient did not undergo additional treatments to the same lesion after the index TARE based on multidisciplinary evaluation using Liver Imaging Reporting and Data System (LI-RADS) treatment response criteria. Logistic regression was used to model response as a function of dose and t-tests to compare average dose by response. Receiver operator characteristic (ROC) curves were constructed to evaluate dose thresholds for complete response.
Results:
Forty patients were included, representing 52 HCC lesions. Thirty lesions (~58%) met criteria for complete response. Of the other 22 lesions, six underwent further locoregional therapy prior to transplant due to residual viability. Lesions with a complete response had a significantly higher mean predicted MIRD dose: 483 versus 275 Gy (p < 0.001).
Cumulative dose-volume histograms were obtained to assess dose thresholds for complete response. Dose administered to 50 (D50), 70 (D70), and 95% (D95) of the tumor volume were all independently predictive of complete response (p < 0.05). For lesions with complete response, the mean D95 was 844 Gy versus 294 Gy (p < 0.001), mean D70 was 1168 versus 432 (p < 0.01), and mean D50 was 1384 versus 520 (p < 0.01). ROC curve analysis demonstrated that D95 had the largest area under the curve (AUC), 0.84. Based on a 70% probability of complete response, a D95 threshold dose of >679 Gy had 57% sensitivity and 91% specificity (71% accuracy).
Conclusion:
Voxel based dosimetry is well-correlated with tumor explant pathology across several dose-volume thresholds. The dose to 95% of the tumor volume had the highest accuracy, with a proposed cutoff of >679 Gy. These data suggest that the D95 dose threshold may be a relevant target for multi-compartment model dosimetry.