Evaluation of the preimplantation worksheet in determining Calypso eligibility for men prescribed postprostatectomy radiotherapy with electromagnetic transponder guidance

This study aimed to assess the design and performance of the preimplant suitability worksheet in determining Calypso eligibility for prostate cancer patients prescribed postprostatectomy radiotherapy with electromagnetic transponder guidance. The medical records and radiotherapy planning datasets of 75 patients prospectively recruited between June 2015 and September 2016 to a Phase 2 trial evaluating electromagnetic transponder-guided postprostatectomy radiotherapy were retrospectively examined. Correlation and differences between computed tomography (CT)-defined greater trochanter and prostatic fossa landmarks were evaluated. Receiver operating characteristic curves were also generated to assess the expected and observed accuracy of the worksheet in determining Calypso eligibility. Strong correlation was demonstrated between anterior surface to planning CT-defined greater trochanter and prostate bed center distances (r = 0.95, p < 0.001), with a mean difference between measurements of 1.1 cm (95% confidence interval [CI]: 0.9 to 1.3). A similar correlation coefficient was found for surface to greater trochanter location and posterior beacon location (r = 0.92, p < 0.001) but with a reduced mean difference of 0.4 cm (95% CI: 0.1 to 0.6). Performance of the worksheet as assessed by planning CT data demonstrated excellent accuracy as a test to determine eligibility (area under the curve: 0.97; 95% CI: 0.92 to 1.00); however, this was not replicated using the same data captured clinically (area under the curve 0.83; 95% CI: 0.68 to 0.98). Although the greater trochanter is a good surrogate for the prostate bed center, it is better associated with the posterior beacon location. As a result, the worksheet will underestimate the truly eligible population if performed accurately and according to manufacturer guidelines. Theoretically, the worksheet could be improved if a cut off of 20 cm is used and the greater trochanter is accurately identified; however, the latter appears to be difficult to achieve in practice.