97. Towards a dose-response correlation in radioiodine therapy of hyperthyroidism from nodular thyroid disease

Purpose Dosimetric treatment optimization is recommended for radioiodine therapy of hyperthyroidism. However, many clinicians still claim that fixed activity administrations yield analogous effectiveness, sparing time-consuming pretherapeutic studies. The possibility to establish a dose-response correlation was here investigated, highlighting the added value of patient-specific dosimetry. Methods 374 patients affected by autonomous thyroid nodule, or multinodular goitre – presenting abnormal blood tests (TSH, and/or FT3, FT4) – were treated. Post-therapy follow-up lasted for at least 1 year. 187 patients were treated with an empiric methodology (not basing on a pre-treatment dosimetric study). 187 patients underwent dosimetry-based treatments (collecting just three biokinetics points), including also 67 treatments based on a biologically effective dose (BED) prescription. The normality of blood tests at 1 year from the therapy, after the first treatment, was evaluated. The BED was also used in the result analysis for all dosimetry-based treatments. Statistical analysis was performed by Mann-Withney and Chi-squared test, univariate receiver operating characteristics (ROC) analysis, and interpolation of dose–response data by a logistic model. Results Dosimetric differences between the unhealed and the healed group (p = 0.0067) resulted from the 120 treatments based on absorbed dose prescriptions, also after conversion of the absorbed dose values in BED (p = 0.0049): the Area-under-curve (AUC) was 0.758 and 0.770, respectively. No significant differences were observed for the healing probability between the BED- or absorbed dose-based treatments (p = 0.423). Considering the total 187 dosimetry-based treatments, the absorbed dose/BED-response curves showed regular trends with r-value of 0.979 and 0.977, respectively. The absorbed dose associated to a healing probability of 50% and 95%, was 42.6 Gy and 242.3 Gy, respectively. The probability of recurrence was 9.6%, exactly half of that associated to the empiric treatment group (19.3%, p = 0.0081). Conclusions Patient-specific dosimetry has proven to increase therapeutic effectiveness, reducing the recurrence frequency and the need of repeated treatments.