327 - Simplified Imaging-Based Dosimetry of Organs and Inherently Heterogeneous Tumors for <sup>131</sup>I-MIBG Therapy Using Combined Pretherapy <sup>123</sup>I-MIBG and Posttherapy <sup>131</sup>I-MIBG SPECT Imaging
Presenter(s)
Y. Wang1, Y. Huh2, K. K. Matthay3, K. Karolczuk4, S. G. Dubois5, and Y. Seo1,6; 1University of California, San Francisco, San Francisco, CA, 2Seoul National University Hospital, Seoul, Korea, Republic of (South), 3University of California San Francisco Benioff Children's Hospital, San Francisco, CA, 4Imaging and Radiation Oncology Core (IROC), Lincoln, RI, 5Dana-Farber/Boston Children's Cancer and Blood Disorders Center, Boston, MA, 6University of California, Berkeley, Berkeley, CA
Purpose/Objective(s): 131I-metaiodobenzylguanidine (MIBG) is an effective therapy for neuroblastoma. Imaging-based dosimetry of 131I-MIBG therapy, such as via SPECT, plays an important role in accurate dose estimation and personalized treatment planning. Yet, dosimetry typically requires multiple posttherapy imaging sessions, increasing the logistical burden and complexity. We hypothesize that a simplified and more efficient imaging-based dosimetry can be achieved by combining the pretherapy 123I-MIBG imaging and a single posttherapy 131I-MIBG imaging for both normal organs and tumors with heterogenous biokinetics within individual subjects.
Materials/Methods: Fifteen SPECT imaging studies of neuroblastoma subjects were acquired, with pretherapy 123I-MIBG imaging performed ~24 hours postinjection and posttherapy 131I-MIBG imaging 2-6 days post-injection. A retrospective phantom-based calibration method was applied to convert the unitless pre- and posttherapy SPECT images into Bq/mL. A two-time-point activity dataset of organs and tumors were obtained by combining activity values obtained from the pre- and posttherapy images. We proposed a kinetic modeling-based method to estimate the time-integrated activity (TIA) based on the two-time point dataset. S-values of the organs and tumors in the 131I-MIBG therapy were determined through Monte Carlo simulation using a simulation toolkit. Subsequently, the absorbed dose was estimated as the product of TIA and S-values.
Results: SPECT images of human subjects in Bq/mL were generated using the phantom-based calibration method. The estimated whole-body 123I-MIBG and 131I-MIBG activities ranged 1.8-55 MBq and 38-1400 MBq, respectively. The kinetic modeling method, utilizing the two-time-point data for TIA estimation, was validated using an independent cohort of 124I-MIBG PET datasets and achieved a lower relative RMSE (14%) compared with the conventional mono-exponential method (relative RMSE = 36%). Organ segmentation and Monte Carlo-based S-value estimations were performed, yielding results consistent with previous studies. Our next step is to combine the pre- and posttherapy SPECT data by rescaling the pretherapy image and to estimate the doses delivered to organs/tumors using the kinetic modeling-based TIA estimation and the S-values. Further, tumor heterogeneity will be studied.
Conclusion: This study proposes a simplified imaging-based dosimetry approach for 131I-MIBG neuroblastoma treatment by combining pretherapy 123I-MIBG and posttherapy 131I-MIBG imaging. Associated techniques, including SPECT image calibration, TIA estimation, and S-value determination, were developed and preliminarily validated. This method could reduce the number of imaging sessions required, simplify the clinical workflow, and enhance treatment planning.