Impact of cell geometry, cellular uptake region, and tumour morphology on 225Ac and 177Lu dose distributions in prostate cancer.

Background: Radiopharmaceutical therapy with ²²⁵Ac- and ¹⁷⁷Lu-PSMA has shown promising results for the treatment of prostate cancer. However, the distinct physical properties of alpha and beta radiation elicit varying cellular responses, which could be influenced by factors such as tumour morphology. In this study, we use simulations to examine how cell geometry, region of pharmaceutical uptake within the cell to model different internalization fractions, and the presence of tumour hypoxia and necrosis impact nucleus absorbed doses and dose heterogeneity with ²²⁵Ac and ¹⁷⁷Lu. We also develop nucleus absorbed dose kernels for application to autoradiography images.

Methods: We used the GATE Monte Carlo software to simulate three geometries of LNCaP prostate cancer cells (spherical, cubic, and ovoid) with activity of ²²⁵Ac or ¹⁷⁷Lu internalized in the cytoplasm or bound to the extracellular membrane. Nucleus S-values were calculated for each geometry, source region, and isotope. The cell models were used to create nucleus absorbed dose kernels for each source region describing the dose to each nucleus in a cell layer, which were applied to simulated tumours composed of normoxic, hypoxic, or necrotic cancer cells to obtain dose rate maps. Absorbed doses within the tumours and dose heterogeneity were analyzed for each tumour morphology and isotope. Cell geometry made a minimal impact on S-values to the nucleus, however internalization resulted in higher nucleus doses. Applying the kernels to the simulated tumour maps showed that doses to each cell type varied between ²²⁵Ac and ¹⁷⁷Lu depending on tumour morphology. Dose heterogeneity within tumours was slightly higher with ²²⁵Ac, however the tumour morphology made a larger impact on dose heterogeneity compared to the choice of isotope, with hypoxic and necrotic tumours having very heterogeneous dose distributions.

Conclusions: Cell geometry simplifications may still allow robust results in simulation studies. Furthermore, the morphology of the tumour itself may make a larger impact on treatment response compared to other variables such as ratio of internalization. Finally, nucleus absorbed dose kernels were created that could enable microdosimetric studies with autoradiography.