Transvaginal Ultrasound-Guided Transperineal Interstitial Brachytherapy for Vaginal Tumors: A Fixed Template Technique

Lower vaginal tumors can be difficult to manage because of the vulvar morbidity associated with external beam radiotherapy (EBRT). We report our experience with development of a fixed-template technique for transvaginal ultrasound-guided transperineal interstitial brachytherapy (BT) for lower vaginal tumors. Seven patients with primary vaginal cancer or recurrent endometrial cancer of the lower vagina were treated with EBRT followed by interstitial BT. All patients underwent pre-treatment magnetic resonance (MR) imaging with a 2 cm vaginal obturator to simulate the shape of the vagina with the transvaginal ultrasound (TVUS) probe in place. A pre-treatment TVUS was performed and these images were fused with the MR images, matching on soft tissue, bones and the obturator. The high risk CTV (HRCTV) was identified on both MR and TVUS images to guide needle insertion. A custom template provided a radial array of potential needle positions at 8 mm spacing around the central hole for the TVUS probe. A stabilizer and stepper system identical to that used for transperineal prostate brachytherapy fixes the relationship between the probe and template. On the day of BT, the transperineal insertion of the needles was performed under TVUS guidance. At the completion of needle placement, with verification in transverse and saggital planes, each needle was individually locked into place in the template. The TVUS probe was removed and replaced by an obturator and the template was sutured to the perineum. The patient was transferred from the recovery room to the CT simulator. The HRCTV and organs at risk were contoured on CT and MR images and a 3D plan was generated. Four patients with primary squamous carcinoma of the vagina (T1-T3, N0) and 3 patients with recurrent endometrial carcinoma were treated. The median dose of EBRT was 45 Gy delivered to the whole pelvis, vagina and bilateral groins using 3D conformal radiotherapy. A single implant was done for 6/7 patients, with 2 fractions of 6.5 Gy BID with a minimum 6 hour interval between fractions. One patient previously treated with EBRT received BT alone with 22 Gy in 4 fractions BID in 2 implants. Mean HRCTV volume was 22.3 cm3 and the median number of needles was 11. Mean D90% was 111% and mean V95% was 96.7%. The mean (range) V150% and V200% were 51.2% (31.1-75.8%) and 20.4% (7-49.2%), respectively. The mean EQD2(3Gy) D2cc of EBRT + BT for the bladder, urethra and rectum was 46 Gy, 46.1 Gy and 48.7 Gy. After a median followup of 30.1 months, 1 patient (T3 vaginal primary) had an out of field second primary of the vulva. MR and physical exam showed no evidence of tumor in the previously treated vagina. The most common acute toxicity was grade 1 vaginal discharge (4 patients). Only 1 case of > grade 3 toxicity was seen (vaginal stenosis). In our experience, transvaginal ultrasound-guided transperineal interstitial BT using a fixed template technique is feasible. Our custom template allowed for precise placement, good parallelism and spacing, and stability of the needles. We were able to identify the tumor on TVUS in the majority of the cases when compared to the HR-CTV done on MRI. We are moving forward with ultrasound-based intraoperative planning for these patients, delivering a single fraction per implant.