Estimation of absorbed radiation doses to skin and S-values for organs at risk due to nasal administration of PET agents using Monte Carlo simulations.

Purpose: The intranasal (IN) administration of radiopharmaceuticals is of interest in being a viable route for the delivery of radiopharmaceuticals that do not ordinarily cross the blood-brain barrier (BBB). However, to be viable in a patient population, good image quality as well as safety of the administration should be demonstrated. This work provides radiation dosimetry calculations and simulations related to the radiation safety of performing such experiments in a human cohort.

Methods: We performed Monte Carlo (MC) simulations to estimate radiation dose to the skin inside a cylindrical model of the nasal cavity assuming a homogenous distribution layer of C and F and calculated a geometry conversion factor (F ) which can be used to convert from a planar geometry to a cylindrical geometry using more widely available software tools. We compared radiation doses from our simulated cylindrical geometry with the planar dose estimates employing our geometry conversion factor from VARSKIN 6.1 software and also from an analytical equation. Furthermore, in order to estimate radiation dosimetry to surrounding organs of interest, we performed a voxelized MC simulation of a fixed radioactivity inside the nasal cavity and calculated S-values to organs such as the eyes, thyroid, and brain.

Results: MC simulations of contamination scenarios using planar absorbed doses of 15.50 and 8.60 mGy/MBq for F and C, respectively, and 35.70 and 19.80 mGy/MBq per hour for cylindrical geometries, leading to determination of an F of 2.3. Planar absorbed doses (also in units of mGy/MBq) determined by the analytical equation were 16.96 and 8.68 ( F and C) and using VARSKIN were 16.60 and 9.26 ( F and C), respectively. Application of F to these results demonstrates values with a maximum difference of 9.41% from the cylindrical geometry MC calculation, demonstrating that when accounting for geometry, more simplistic techniques can be utilized to estimate IN dosimetry. Voxelized MC simulations of radiation dosimetry from a fixed source of 1 MBq of activity confined to the nasal cavity resulted in S-values to the thyroid, eyes, and brain of 1.72 x 10 , 1.93 x 10 , and 3.51 x 10  mGy/MBq·s, respectively, for F and 1.80 × 10 , 1.95 × 10 , and 3.54 × 10  mGy/MBq·s for C.

Conclusion: Dosimetry concerns about IN administrations of PET radiotracers should be considered before clinical use. Values presented in the simulations such as the S-values can be further used for assessment of absorbed doses in cases of IN administration, and can be used to develop and adapt specific study protocols. All three presented methods provided similar results when considering the use of a geometry conversion factor for planar to cylindrical geometry, demonstrating that standard tools rather than dedicate MC simulations may be used to perform dose calculations in nasal administrations.