Refining risk assessment in 177Lu-DOTATATE extravasation: the value of 3D Monte Carlo dosimetry

Introduction Extravasation of radiopharmaceuticals is a significant radiation safety event in Molecular Radiotherapy (MRT). Timely and accurate dosimetric evaluation is essential to anticipate normal tissue toxicity and guide personalized clinical management. This paper presents a case of 177 Lu-DOTATATE extravasation in which standard dosimetric calculations predicted severe tissue necrosis, whereas the patient experienced no permanent clinical sequelae. This study aimed to improve current risk assessment by comparing standard dosimetric calculations with patient-specific three-dimensional Monte Carlo dosimetry. Methods Following the extravasation of approximately 2.7 GBq of 177 Lu, detected 15 min after the start of the injection, the French Nuclear Safety Authority (ASNR) performed a dosimetric assessment using the MIRD formalism (IDAC-Dose 2.1), modeling the extravasation as a spherical volume that decreased from 10 mL to 6 mL over the first 4 h. To complement this approach, voxel-based dosimetry was performed using the OpenGATE Monte Carlo platform. Patient-specific SPECT-CT images acquired at 4, 7, and 70 h after injection were corrected for partial volume effects and registered to the corresponding CT images. A bi-exponential washout model was used to account for the surgical lavage performed 6.5 h after injection. Results The standard ASNR calculation estimated a mean absorbed dose of 56 Gy within the extravasation volume, exceeding the 25 Gy necrosis threshold for acute single-dose exposure. Monte Carlo simulation estimated a mean absorbed dose of ( 44.09 ± 3.53 ) Gy within the highest-dose 1 mL volume. However, the three-dimensional dose distribution revealed marked heterogeneity. The highest-dose region was confined to the deep subcutaneous adipose tissue, while the cutaneous and underlying muscle tissues received substantially lower doses. Clinical follow-up over 12 months confirmed the absence of tissue necrosis, supporting the hypothesis that the depth of the extravasation protected these radiosensitive tissues. Conclusion Although the MIRD and Monte Carlo approaches yielded similar mean absorbed dose estimates, the spatial information provided by three-dimensional Monte Carlo dosimetry offered essential clinical insight into the irradiated tissues and their dose distribution. This additional information enabled a more accurate interpretation of the expected clinical outcome. Furthermore, this case suggests that, when surgical lavage is considered, it should be performed as early as possible after extravasation is detected to provide any meaningful dosimetric benefit.