Influence of brachytherapy applicators geometry on dose distribution in cervical cancer

Although the relationship between the dose delivered to adjacent organs (urinary bladder and rectum) and the frequency and severity of treatment complications has been reported in many series, the factors influencing pelvic dose distribution are not well defined. The aim of the study was to assess retrospectively the influence of the size of cervical cancer brachytherapy applicators (ovoids and uterine tandems) on pelvic dose distribution and the impact of various therapy-dependent factors on patient anatomy and on dose distribution in particular applications.

The subject of this study were 356 cervical cancer patients treated with Selectron LDR as a part of their radical radiotherapy. Analysed factors included preceding external beam radiotherapy (EBRT) or brachytherapy applications, use of general anaesthesia for application and the system of pellet loading.

Significant correlation was found between the size of applicators and doses to bladder, rectum and points B: larger vaginal applicators produced lower dose in bladder and rectum and higher dose in point B (all p < 0.0001), longer uterine tandems produced lower dose in rectum and higher dose in point B (both p < 0.0001). Significant decrease in the frequency of use of large applicators (ovoids: p < 0.0001, tandems: p = 0.055) and worsening of dose distribution, i.e. higher doses to critical organs (respectively: bladder p = 0.0012, rectum p = 0.02) and lower point B dose (p = 0.0001) were observed at consecutive brachytherapy applications. Similar situation occurred in patients, who received EBRT prior to brachytherapy (ovoids: p < 0.001, tandem: p = 0.04, bladder dose: p = 0.009, rectal dose: p = 0.073, point B dose: p = 0.059). Vaginal applicators were larger (p = 0.026) and the dose distribution was better (bladder: p = 0.023, rectum: p = 0.002, point B: p = 0.0001) in patients who had their insertions performed under general anaesthesia. The comparison of 2 consecutively used systems of pellet loading revealed more favourable dose distribution: lower dose for bladder (p = 0.014) and higher dose for point B (p < 0.0001) for the system, which utilised more sources in ovoids and in the distal part of the uterine tandem, in spite of more frequent use of smaller applicators in this group of patients. In multivariate analysis ovoid size was related to preceding external beam radiotherapy (p = 0.025). Uterine tandem length was dependent on the number of preceding intracavitary applications (p < 0.001) and preceding external beam radiotherapy (p = 0.007). Bladder dose was related to preceding brachytherapy (p = 0.011) and the pattern of pellet loading (p = 0.031). Rectal dose was dependent only on the use of general anaesthesia during application (p = 0.001) and point B dose was dependent on the pattern of pellet loading (p < 0.001) and marginally-on the use of preceding external beam radiotherapy (p = 0.06).

The results of this study allow for identification of treatment-related factors determining pelvic dose distribution in cervical cancer brachytherapy and may potentially enable optimisation of this distribution in particular clinical situation.