Pre-plan technique feasibility in multi-interstitial/endocavitary perineal gynecological brachytherapy

Purpose To present the implementation of a magnetic resonance imaging (MRI) pre-planning technique in multi-interstitial perineal and endocavitary gynecological brachytherapy. Material and methods We used a new fully MRI-compatible applicator that is capable to engage titanium needles, and an intrauterine tandem, developed in our department for the treatment of gynecological cervical cancer patients. This applicator is an attempt to combine the technical advantages of the Martinez universal perineal interstitial template (MUPIT) with the improvement in dose distribution by adding an intrauterine probe with the imaging advantages of MRI-based brachytherapy, thus preserving the stability, geometry, and robustness of the implant, avoiding possible errors of free-hand needle placement. A pre-brachytherapy MRI T2 acquisition is carried out with the template in place 3-5 days before the implant. On this image set, clinical target volume (CTV) is drawn. The required needles and their depths are selected accordingly to encompass the CTV (as conformal as possible). To facilitate this task, a Java based application linked to the treatment planning system has been developed. From this procedure, each needle identification and its depth are obtained previously to the implantation. With this information, the radiation oncologist proceeds with implant and then, a post-implant MRI is carried out, in which the contouring, needles, tandem reconstruction, and optimization are established. Results This pre-planning procedure has been successfully applied in 10 patients. An excellent reproduction of the virtual pre-planning has been achieved. Conclusions We describe a virtual pre-planning technique using a multi-interstitial and endocavitary perineal template. It is based on a virtual work with MRI images. This procedure has shown to be feasible and efficient in clinical practice by facilitating the work of specialists, and reducing uncertainties of the application.