Understanding time-activity curve and time-integrated activity variations in radiopharmaceutical therapy challenge:...

The process of determining/calculating the time-activity curve (TAC) for radiopharmaceutical therapy (RPT) is generally heavily dependent on user- and site-dependent steps (e.g., the number and schedule of measurement points to be used, selection of the fit function), each having a notable effect on the determination of the time-integrated activity coefficient (TIAC) and thus on the calculated absorbed dose. Despite the high clinical importance of absorbed doses, there is no consensus on the methodology for processing time-activity data or even a clear understanding of the influence of various uncertainties and user-dependent variations in personalized RPT dosimetry on the accuracy of TAC calculations.

To address this critical unmet need, the time-activity curve and time-integrated activity variations (TACTIC) AAPM Grand Challenge was designed to explore the variations in TAC modeling for RPT applications.

Launched in January 2023, the TACTIC challenge consisted of three phases: i) warm-up phase (phase 0, to gain familiarity with the logistics and the modalities of the challenge), ii) TAC fitting based on data from individual patients (phase 1, rated to determine winner 1), and iii) TAC fitting using population-based data (phase 2, rated to determine winner 2). Based on the distributed synthetic biokinetic data of [¹⁷⁷Lu]Lu-PSMA-617 RPT (kidney, blood, and tumor), participants were asked to model the TAC and calculate the TIAC values for each of these tissues to the best of their ability. In addition, participants were requested to submit information about the fit function and fit optimization parameters. The best-performing team in each phase was determined on the basis of total root-mean-square error (RMSE) value across all three tissues.

A total of 132 teams from over 30 countries registered for this data-driven challenge, of which 95 individual groups submitted their results throughout the challenge. By presenting participants with an identical set of measurement points previously generated from measured biokinetic data and providing additional a priori information about the procedure at different stages of the challenge, we could assess the degree of variation within the TIAC estimation. We investigated which of the commonly used TIAC estimation methods performs best and could therefore be used to harmonize TAC modeling in RPT dosimetry.

The results of the TACTIC challenge demonstrate the large variability in TAC fitting despite the participants receiving identical input data. This highlights the fundamental role of TAC fitting methodology selection in the calculation of absorbed doses in RPT and successfully raises awareness of the need for greater harmonization in dosimetric approaches.