Introduction
Vaginal cancer is a rare malignancy affecting 6,230 patients per year and resulting in 1,450 annual deaths. Reference Siegel, Miller and Jemal1 Vaginal cancer is a challenge to treat and cure as surgical excision frequently requires life-changing exenterative procedures due to the proximity and frequent involvement of the rectum, anus and bladder. Reference Ball and Berman2–Reference Peters, Kumar and Morley6 Radiation, and more recently, chemoradiotherapy have become the mainstays of treatment with pelvic control rates of 75–85% for stage II lesions and 25–70% for stage III lesions. Reference Stock, Chen and Seski4,Reference Chyle, Zagars, Wheeler, Wharton and Delclos7–Reference Otton, Nicklin and Dickie10 We report the successful delivery of salvage stereotactic ablative radiotherapy for a patient with persistent vaginal cancer at the introitus involving the urethra and clitoris after previous radiotherapy.
Clinical history
Our patient was a 71-year-old white female at diagnosis, presenting with dyspareunia and irritative urinary symptoms. A vaginal biopsy at an outside facility confirmed the presence of well-differentiated squamous cell carcinoma (large cell keratinising) in a 2 cm lesion in the distal vagina near the urethral orifice. PET/CT performed for staging showed a hypermetabolic vaginal mass with a standard uptake value (SUV) of 29·7 as well as obturator and inguinal nodal metastases. An examination under anaesthesia (EUA) and fiducial marker placement with cystoscopy was performed at our facility in preparation for definitive chemoradiotherapy. The examination at that time showed a 3 cm lesion at the 12–1 o’clock position in the distal vagina and extending into paravaginal tissue, slightly to the left side of the urethral orifice and involving the clitoris. Cystoscopy showed extrinsic urethral compression, but no direct intraluminal invasion. According to the American Joint Committee on Cancer (AJCC), Reference Amin11 her final AJCC stage pretreatment was T2b N1 M0 (Stage III). She signed informed consent to participate in the NRG Oncology GY 006 trial, 12 which randomises patients with the primary pelvis-contained cervix or vaginal cancer to pelvic radiotherapy with radio-sensitising cisplatinum at 40 mg/m2 +/− triapine, a radiosensitizer, randomising to the arm with added triapine.
According to the protocol, her radiation was delivered using volumetric-modulated arc therapy (Rapid Arc, Varian Medical Systems, Palo Alto, CA, USA) and a simultaneous integrated boost technique with external beam to the involved nodes, delivering 47·6 Gy to the primary and uninvolved nodes and 59·36 Gy to the hypermetabolic nodal tissue in 28 fractions. Initially compliant with treatment, the first 25 fractions were delivered over 40 elapsed days with 1 missed fraction in the third week of treatment and a 6-day treatment interruption between fractions 20 and 21. She then began a protracted treatment interruption lasting 33 days during which she was removed from study participation, and an additional 23·4 Gy were prescribed to the residual primary area at 1·8 Gy per day using a four-field technique and three-dimensional planning (71 Gy total, a slightly increased dose to compensate somewhat for her treatment delay). She took an additional treatment break between fractions 29 and 30 for 3 and half weeks, finally completing 71 Gy to the primary in 41 fractions over 123 days.
Approximately 1 month later, she was seen by her referring gynaecologic oncologist with persistent pain at the primary site and a limited exam was suspicious for persistent disease. A repeat PET/CT was ordered and showed complete resolution of the nodal disease, but only a partial metabolic response at the primary site with a residual SUV of 15. Her case was discussed at the multidisciplinary gynaecologic oncology tumour board where a decision to offer palliative chemotherapy versus SABR versus hospice care was made. After conferring with the patient, she elected to have SABR in an attempt to achieve remission and possible cure.
Intolerant of examination, she was simulated under general anaesthesia after EUA and had additional fiducial marker placement at the periphery of the tumour. After patient alignment with the central axis laser of the CT simulator, a 25 mm Nucletron vaginal cylinder (Elekta AB, Stockholm, Sweden) was placed intra-vaginally, aligned vertically with the laser through its central axis and secured using a standard Nuclectron brachytherapy stabiliser board. Additional cylinder ring pieces from an identical cylinder were placed on the central rod for a total length of 17·5 cm in order for the cylinder to span the entire distance from the vaginal cuff to the introitus, vulva and beyond. The abdomen, pelvis and upper thighs were imaged at 2·5 mm slice thickness and the patient was awakened and discharged. Contouring of the tumour and avoidance structures was carried out in Eclipse (Varian Medical Systems, Palo Alto, CA, USA) as shown in Figure 1. Then, a stereotactic plan was generated to deliver 30 Gy in five fractions to the tumour with a 5 mm margin, subtracting the area of the planning target volume (PTV) that extended into the cylinder. A treatment plan with two arcs was made and delivered on a Varian TrueBeam SRS machine with 6X flattening filter-free beam. The planning images were sent to the BrainLab ExacTrac system (BrainLab, Munich, Germany) to aid in patient positioning at treatment.
Figure 1. Separation of organs at risk such as rectum from the gross tumour volume (orange) and PTV (green) using the vaginal cylinder as shown on an axial and sagittal views of the CT images.
This vaginal cylinder is made from non-metallic materials, which has made it compatible for X-ray radiography and CT imaging. The dummy seeds made from titanium in the inner channel of the cylinder showed well on the radiographs. All five treatment fractions were delivered under general anaesthesia. For each fraction, the patient was repositioned on the robotic couch with the central axis laser at midline and the vaginal cylinder was inserted and secured to the stabilisation board. AP and lateral kV radiographic images were acquired and the cylinder was repositioned to replicate in three axes as closely as possible to the original position at simulation as shown in Figure 2. The ExacTrac images were acquired using the vaginal cylinder as a surrogate to localise the nearby tumour. The patient was positioned with six degrees of freedom that included three translational shifts and three rotational shifts of the robotic couch. Then, a second set of ExacTrac images were acquired with the corrected patient position, confirming less than 1 mm and 1° deviation in all directions of the cylinder position for treatment from its initial position at the time of the simulation.
Figure 2. Radiographic image obtained from X-ray imaging system in the first column and digitally reconstructed radiograph in the second column by the ExacTrac system from the CT images.
Results and Discussion
Vaginal examination at the conclusion of the fifth fraction showed near-complete involution of the tumour at the introitus. Oral narcotic analgesia between fractions and after treatment was necessary for patient comfort. The SABR plan delivered an additional 2.04 Gy mean dose to the rectum, with a maximum point dose of 10.4 Gy. The cumulative rectal dose for all treatment showed a maximum point dose of 74.57 Gy, a mean dose of 59.06 Gy, a 2 cc dose of 70.37 Gy and a 5 cc dose of 68.19 Gy as shown in Figure 3. Although higher than we would like, we felt these doses were acceptable given the surety of disease progression without treatment and the patient’s ardent wish to achieve remission.
Figure 3. Dose distributions from the boost SABR plan with the vaginal cylinder to spare the rectum from excess dose.
Retreatment of recurrent or persistent disease in the vagina is challenging because of the proximity of bladder, rectum, anus, urethra and clitoris. In this patient, the distal urethra and clitoris were involved and could not be spared, but the presence of the cylinder physically distanced the posterior vaginal wall and rectum/anus an additional 2.5 cm from the central axis of the SABR volume, allowing the delivery of a tumouricidal dose with diminished risk to the rectum, anus and posterior vaginal wall. Multichannel cylinder vaginal brachytherapy has been reported to accomplish similar ends for small volume disease that is not thicker than 5 mm in the vaginal wall, allowing asymmetric dose delivery to volumes of vaginal tissue with relative sparing of the rest of the organ. Reference Glaser and Beriwal13–Reference Beriwal, Demanes and Erickson15 However, this was not an option for our patient due to the size and thickness of her persistent tumour.
To our knowledge, this is the first report of a vaginal cylinder used to physically distance organs at risk from the treatment target, and also as a localising device for the delivery of SABR. This case is also illustrative of radiobiological principles of dose fractionation. Accelerated simultaneous integrated boost delivered sequentially to the involved nodes as prescribed during the initial period of compliance with treatment schedule was able to sterilise the involved nodes. Protracted and delayed treatment of the primary lesion allowed its persistence and required multiservice intervention, and demanded additional risk and manpower such as anaesthesia delivery and SABR.
Conclusion
Patients should be encouraged to adhere to timely delivery of their prescribed radiotherapy treatments, but we have reported here a novel approach to extirpating persistent disease in a non-compliant patient. This study demonstrated the successful use of a vaginal cylinder to achieve both goals of physically separating organs at risk from the PTV and employment of the cylinder as a localising device for accurate patient setup with image guidance in the delivery of SABR. This combination of external beam radiation therapies is an alternative and effective treatment option for patients who do not tolerate standard of care therapies.
Acknowledgements
None.
Financial support
None.
Conflict of interest
None.
Open access
