Introduction

Feline vaginal neoplasia, either benign or malignant, is infrequently reported in the veterinary literature.^1–4 Such reports include vaginal leiomyoma,¹ vaginal adenocarcinoma,³ and vaginal polyp.⁴ Complete and subtotal vaginectomy are recognized surgical techniques for management of vaginal neoplasia in canine and feline patients.^4–8 To date, there is limited literature describing and assessing the surgical approach for these procedures. In dogs, reported techniques include a combined abdominal and vestibular approach,⁷ sagittal pubic osteotomy,⁹ and a combination of perineal, abdominal, and perivaginal approach.¹⁰ In conjunction with these approaches, and while not currently reported for resection of vaginal tumors in dogs, a bilateral pubic and ischial osteotomy for surgical management of caudal colonic and rectal masses has also been described.¹¹ This technique was reported to provide sufficient exposure for resection of intrapelvic tumors, with minimal complications. In feline patients, only a single case report⁴ of the use of a bilateral pubic and ischial osteotomy for resection of a benign vaginal tumor has been described.

Due to the lack of reports in the literature, the aim of this case series was to describe the clinical signs, diagnostic findings, surgical treatment, and outcomes of 3 cats diagnosed with extensive vaginal masses. A second aim was to provide a detailed description of the surgical technique of a bilateral pubic and ischial osteotomy, including documentation of intraoperative and postoperative complications.

Methods

Eleven referral clinics in the UK were invited to participate in the study. Ethical review was obtained from the University of Edinburgh ethical review committee (VERC 7.21). The surgical database of each institution was searched using the words “feline,” “cat,” “vaginal,” “neoplasia,” and “osteotomy” between the dates of January 2004 and June 2022. Medical records were received from 3 institutions that met the inclusion criteria of feline patients that underwent a bilateral pubic and ischial osteotomy for surgical management of a vaginal neoplasm.

Data collected included signalment, clinical history, presenting clinical signs, physical examination findings, MRI or CT preoperative imaging findings, anesthesia and analgesia protocol, surgical technique, histopathological diagnosis, intra- and postoperative complications, and outcome. Information regarding surgical outcome was obtained during postoperative recheck appointments, up to 12-week to 15-month postoperative follow-up telephone calls, and ongoing adjunctive chemotherapy treatment appointments (depending on the case).

Results

Three cases met the inclusion criteria; all cases were operated by the same surgeon (KF). There were 2 domestic shorthair cats (cases 1 and 2) and 1 Sphynx cat (case 3) reported with an age at presentation of 11.7, 5.8, and 6.3 years, respectively. Clinical presentation included tenesmus and constipation due to dorsal compression of the rectum in 2 cases (cases 1 and 3) and an intermittent mucopurulent, occasionally hemorrhagic vaginal discharge in the third (case 2). Clinical signs were reported to be noticed 30, 210, and 21 days prior to presentation, respectively, for cases 1, 2, and 3. On clinical examination, cases 1 and 3 presented with a palpable caudal abdominal mass, with a dilated colon containing feces. Case 3 also showed an externally visible perineal mass protruding from the vulva. In case 2, clinical examination was unremarkable.

CT imaging (thorax, abdomen, and pelvis) was performed in cases 1 and 2, while MRI and thoracic radiographs were performed in case 3. Either MRI or CT images confirmed the presence of a large mass within the pelvis, arising from the vagina. Thoracic and abdominal imaging were unremarkable in all 3 cases confirming pathology focused to the vagina, without evidence of metastatic disease. Dimensions of the vaginal tumor (height, width, and length) in relation to the largest dimension of the pelvis (height, width, and length), expressed in percentage, was obtained by measuring multiplanar CT or MRI images. The masses measured 53% X 62% X 63% (case 1), 50% X 100% X 60% (case 2), and 150% X 120% X 120% (case 3), respectively.

In case 1, in which a CT had been performed, a large volume intrapelvic mass (length 5 cm X width 4.2 cm X height 3.6 cm) occupying the majority of the pelvic canal was noted, with the mass protruding slightly cranially from the pelvic inlet. A larger proportion of the mass bulged caudally beyond the pelvis into the perineal region. Postcontrast, there was a markedly enhancing vestibule/caudal vagina that was continuous with the caudal aspect of the mass. Severe secondary dorsal displacement and flattening of the rectum was noted.

In case 2, in which an MRI had been performed, a large, long, bilobed mass lesion was reported in the cranial half of the pelvis with approximate dimensions of length 5 cm X width 1.5 cm. The cranial margin of the mass was just caudal to the urinary bladder neck, and caudally the mass extended to the caudal aspect of the coxofemoral joints. The mass was hyperintense on T2-weighted and STIR and isointense to the surrounding musculature on T1-weighted, showing moderate contrast enhancement. In summary, a large, bilobed concentric vaginal mass was reported.

In case 3, in which a CT had been performed (Figure 1), a large-volume, smoothly marginating, hypoattenuating intrapelvic vaginal mass (length 7 cm X width 4 cm X height 4 cm) was described. The mass was minimally contrast enhancing and occupied the entire pelvic canal. A significant proportion of the mass bulged beyond the pelvis into the perinium and protruded from the vulva caudally. Significant dorsal compression of the rectum within the pelvic canal was evident, with a dilation of the colon cranial to the mass, within the abdomen.

View Full Size

Case 3. Sagittal and transverse contrast-enhanced CT, showing a vaginal mass occupying 150% of the length of the pelvis, with vulval protrusion, and 120% of the width of the pelvis.

Citation: Journal of the American Veterinary Medical Association 261, 12; 10.2460/javma.23.05.0260

• Download Figure
• Download figure as PowerPoint slide

Cytological evaluation of the vaginal mass in case 1 revealed mesenchymal cells, and a low-grade soft tissue sarcoma was reported. Cytological evaluation of case 2 confirmed a vaginal mycetoma, and due to the extent of the vaginal mass, a combination of surgical and medical management was recommended as previously described.¹² Cytological evaluation of case 3 confirmed a vaginal polyp. For all 3 cases, staging investigations prior to surgery did not show any evidence of metastatic disease. In light of the size and location of all 3 of the reported vaginal masses, a bilateral pubic and ischial osteotomy was performed.

Surgery time ranged from 120 to 200 minutes, and anesthesia time ranged from 200 to 265 minutes. The preoperative anesthesia and analgesia protocol was similar for all 3 cases involving a premedication of methadone (0.2 mg/kg, IV) and midazolam (0.3 mg/kg, IV; cases 1 and 2) or methadone (0.2 mg/kg, IV) and acepromazine (0.05 mg/kg, IV; case 3). Anesthesia was induced with propofol (5 mg/kg, IV) and maintained with isoflurane in oxygen. In cases 1 and 3, an epidural of morphine (0.1 mg/kg) and bupivacaine (0.5 mg/kg) was given. In all 3 cases, amoxicillin–clavulanic acid (Augmentin; 20 mg/kg, IV) was administered 30 minutes prior to surgery and continued every 90 minutes during surgery.

The surgical procedure was similar for all 3 cases (Figures 2–4; case 3). A caudal midline celiotomy was performed extending from the umbilicus to the most caudal aspect of the ischium. The subcutaneous tissues and linea alba were incised to approach the caudal abdomen. The incision extended along the midline pubis with a No. 10 scalpel blade, and the abductor longus and brevis muscles were subsequently elevated from the midline pubis with a Freer periosteal elevator (Figure 2). At the cranial pelvis, the prepubic tendon was cut using a No. 11 blade as its insertion on the pubis. At the caudal pelvis, the ligamental attachment to the ischium was also cut, again using a No. 11 blade, taking care not to traumatize the urethra at this level. Any remaining adductor and external obturator musculature attachment on the ventral pubis was elevated from the obturator foramen bilaterally, taking care to identify and carefully retract the obturator nerve bilaterally. While identification and catheterization of the urethra would be recommended prior to vaginal mass resection, the distortion of the urethral orifice precluded this in all 3 patients. Avoidance of iatrogenic damage of the urethra during osteotomy was achieved through placement of a periosteal elevator dorsal to the osteotomy site and ventral to the urethra.

View Full Size

Midline approach to the pelvis showing reflection of adductor muscles bilaterally and obturator foramen. Cranial is to the top of the photo. The vaginal mass protruding through the vagina is visible to the bottom of the photo. Predrilling of the pelvis (1.1 mm) bilaterally in the pubis and ischium prior to osteotomy cut midline to the predrilled holes (yellow stars). The black arrow denotes the obturator nerve.

Citation: Journal of the American Veterinary Medical Association 261, 12; 10.2460/javma.23.05.0260

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Vaginal mass (yellow star) bluntly dissected following removal of pubis and ischium. The urethra is retracted laterally with a yellow vessel loop. The cervical stump following ovariohysterectomy is evident (black arrow).

Citation: Journal of the American Veterinary Medical Association 261, 12; 10.2460/javma.23.05.0260

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Polydioxanone suture preplaced and then tied through predrilled holes in the pelvis for replacement of the pubis and ischium. Care was taken to resuture the prepubic tendon and tendinous attachments caudal to the ischium to the pelvis.

Citation: Journal of the American Veterinary Medical Association 261, 12; 10.2460/javma.23.05.0260

• Download Figure
• Download figure as PowerPoint slide

Once the ventral pelvis was clearly visualized, the pubis and ischium were predrilled using a 1.1-mm drill bit. Two holes were drilled in the left pubis and left ischium. The process was repeated on the right side. The holes were drilled at a distance carefully accounting for the width of an oscillating saw blade between them (Figure 2). An oscillating saw was then used to cut between the predrilled holes; 4 cuts were made in total. After releasing its intrapelvic muscular attachments from the internal obturator muscle, the ventral pelvis was carefully removed and placed in a saline-soaked swab.

The vaginal mass and urethra were identified, and a vascular loop was placed around the urethra (Figure 3). The vaginal mass was bluntly dissected from the rectum and bladder, taking care not to damage their respective nerve supplies. In case 3, a standard ovariohysterectomy was performed first, allowing for easier dissection of the vaginal mass around the bladder. In cases 1 and 2, ovariohysterectomy had previously been performed. Once the tumor was fully dissected, the caudal vagina was transected at the level of the urethral orifice and the remainder of the vestibular mucosa closed with 1.5-m polydioxanone in a simple continuous pattern (PDS 4/0; Ethicon Inc).

Polydioxanone 3-m (PDS 2/0; Ethicon Inc) suture was preplaced through the predrilled holes in the pelvis, and the pubis and ischium was replaced. The sutures were tied, taking care to place equal pressure on each suture as they were tied to replace the pelvis in its precise anatomical position presurgery (Figure 4). Overtightening of the cranial sutures before tightening of the caudal sutures would result in a gap at the osteotomy sites. In case 2, orthopedic wire was initially used to replace the pubis/ischium rather than polydioxanone. However, this resulted in a fracture of a single drill hole through to the osteotomy site, and in all subsequent sutures, polydioxanone was used instead.

The prepubic tendon was carefully reattached to the pubis with 3-m polydioxanone (PDS 2/0; Ethicon Inc), placing each simple interrupted suture through the tendon and around the pubis/obturator foramen to allow for a firm suture anchorage and prevent any postoperative abdominal herniation. At the caudal pelvis, the tendinous attachments were also sutured in a similar fashion around the ischium/obturator foramen; again, allowing for a firm suture anchorage and preventing any caudal herniation of pelvic content. The adductor muscles were carefully apposed, again with 3-m polydioxanone in a simple continuous pattern (PDS 2/0; Ethicon Inc) and the caudal abdomen closed routinely.

Postoperatively, all 3 cases showed signs of mild intermittent splaying of the pelvic limbs and 2 of the 3 cases (cases 1 and 2) presented with mild stranguria. Case 1 also presented with intermittent hematuria for 6 weeks postsurgery, with a negative urine culture on 3 separate submissions. In case 2, a urinary catheter was placed postoperatively for 3 days and the patient was discharged on prazosin hydrochloride (1 mg/kg, PO, q 12 h for 5 days) and itraconazole (10 mg/kg, PO, for 4 weeks). In case 1, the stranguria self-resolved 24 hours postsurgery and intermittent hematuria resolved 6 weeks postsurgery. The patients remained hospitalized for 3 (case 3) to 4 days (cases 1 and 2). In all 3 cases, adequate postoperative analgesia was achieved with a combination of meloxicam (0.1 mg/kg) and methadone (0.2 mg/kg); the latter was changed to buprenorphine (0.02 mg/kg) 24 hours postoperatively in cases 1 and 3 and 48 hours postoperatively in case 2. Patients were discharged from the hospital on meloxicam (0.05 mg/kg) for 10 days.

In cases 2 and 3, full resolution of clinical signs was noted 14 days postoperatively, with no signs of splaying of the pelvic limbs. In case 1, mild pelvic limb splaying was evident for 8 weeks postsurgery. A 3-month telephone update with owners of cases 2 and 3 again confirmed no stranguria or pelvic gait abnormalities. This was again confirmed with a 15-month telephone update with the owner of case 3. Case 1 re-presented to the oncology service at 12 weeks postsurgery, and no pelvic gait abnormalities were noted at this time.

In case 1, histopathology postvaginectomy did not support the preoperative cytological diagnosis and a low-grade T-cell–rich B-cell lymphoma (Feline Hodgkin-like lymphoma) was reported following immunohistochemistry. Histopathology in cases 2 and 3 supported the initial cytological diagnosis of mycetoma (case 2) and vaginal polyp with the uterus reported as endometrial hyperplasia and hydro/mucometra (case 3).

In case 1, in which a T-cell–rich B-cell lymphoma was diagnosed postoperatively through histopathology and immunohistochemistry, adjunctive chemotherapy was recommended to achieve clinical remission but declined initially. The patient re-presented 12 weeks postsurgery with a pleural effusion. Abdominal ultrasound and thoracic radiographs identified multiple enlarged mediastinal lymph nodes and a pulmonary mass. Abdominal lymphadenopathy was also confirmed. These imaging findings were absent on preoperative assessment. Cytology of the pulmonary mass revealed a mesenchymal proliferation, similar to the cytological presentation documented for the initial vaginal mass, and hence, while the cytology did not confirm lymphoma, this could not be excluded. Cytology of the abdominal lymph nodes was suggestive of lymphoma. The patient received L-asparaginase (400 IU/kg, SC), vincristine (0.5 mg/m², IV), cyclophosphamide (230 mg, PO), and prednisolone (2 mg/kg, q 24 h, PO) but was euthanized 5 days following initiation of treatment due to failure of response to chemotherapy.

Discussion

Selection of an appropriate surgical approach to the vagina for management of vaginal masses in feline patients has been infrequently discussed in the veterinary literature. The most commonly utilized surgical approaches in the canine patient include a combined abdominal and vestibular approach,⁷ a caudal celiotomy or episiotomy alone, perineal, perivaginal, or a combination of the above.¹⁰ Less-utilized techniques include a pubic symphysiotomy⁹ and a bilateral pubic and ischial osteotomy.¹¹ Only a single report⁴ has discussed the use of bilateral pubic and ischial osteotomy in the cat, and the short case series presented here further highlights its successful use in feline patients and promising surgical outcomes.

In these 3 reported feline cases, the technique of bilateral pubic and ischial osteotomy was selected to address the large tumor size in relation to the pelvis (cases 1 and 3) or improve access to resect a multilobular elongated mass (case 2). Selection of a bilateral pelvic osteotomy allowed adequate exposure for meticulous dissection, which was not expected to be sufficient via a standard episiotomy approach for circumferential dissection of such large masses through the vagina in these feline patients. An assumption that a bilateral osteotomy creates significant trauma and hence increased postoperative complications such as bruising, pelvic limb ataxia, and splaying of the limbs due to obturator nerve damage is certainly a possibility with such invasive surgery. However, the 3 cases presented in this study showed only mild splaying of the pelvic limbs postsurgery, hospitalization of only 3 to 4 days, and no splaying of pelvic limbs at 14 days postsurgery. Two of the 3 cases presented with mild stranguria postoperatively. While it is possible that the bilateral pelvic osteotomy may have resulted in inflammation of the intrapelvic musculature causing stranguria (due to inflammation/compression of the urethra in contact with these tissues), it is likely that this stranguria was simply related to the nature of the surgery and resection of the vaginal mass rather than the selected surgical approach per se. Only 1 minor intraoperative complication of a fractured drill hole (case 2) was reported in these 3 cases, further highlighting the success of this approach for resection of vaginal masses in cats.

The reported surgical technique here differed from that of the single previously reported osteotomy technique in several ways. In the description by Saitoh et al,⁴ a smaller K wire (0.9 mm) was used to drill the pubic and ischial holes and the ischial/pubic bony “shelf” was reflected rather than removed completely. Orthopedic wire was also used to reoppose the bone fragments, and an 8-week follow-up was reported for this patient. In our study, a larger 1.1-mm drill bit was used and the bony “shelf” removed completely to maximize access to the tumor. Due to fracture of a single drill hole, all osteotomy sites were realigned with polydioxanone suture rather than orthopedic wire and a minimum of a 12-week follow-up was reported for these 3 cases. While it is true that reflection of the bony shelf and use of orthopedic wire may well result in better bone healing, we did not identify any clinical examination abnormalities at 12 weeks postsurgery to suggest any complications with bone healing. We acknowledge that this can only be confirmed radiographically, but performing pelvic radiographs could not be justified in patients with no clinical signs. Equally, it has been suggested that patients having unremarkable recoveries from other osteotomy surgeries, such as tibial plateau leveling osteotomy and medial patella luxation, are unlikely to benefit from follow-up radiography should they present clinically sound at postoperative clinical evaluations.^13,14 In recognition of these findings, we feel justified in not performing postoperative radiography in these clinically sound cats.

Vaginal neoplasia in cats is sparsely reported in the veterinary literature.^1,2 While T-cell–rich B-cell lymphoma^15,16 and mycetomas^17–23 have both been reported in cats, they have never been identified within the vagina, making this an unusual presentation for these types of lesions. Vaginal polyps are frequently reported canine vaginal masses,^{5–8,23–25} yet they remain infrequently reported in feline patients and their description is limited to a single case report.⁴ The case series reported here describes only the second documented case of a feline vaginal polyp.

The lack of literature surrounding feline vaginal tumors is likely a reflection of the neuter status in feline patients. In canine patients, leiomyomas are often reported in older entire patients, with vaginal leiomyomas and leiomyosarcomas contributing 13 of 20 (65%) cases.^{5–8,23–25} Older, intact queens are less commonly represented in the animal population.

Vaginectomy is not a commonly performed surgery, and selection of an appropriate surgical approach can be challenging when presented with sagittal and transverse CT/MRI/ultrasonographic images. Excessively large vaginal masses, such as those occupying > 100% of the width, length, and height of the pelvis, may not be amenable to a bilateral caudal abdominal and vestibular approach,⁷ and this case series highlights the use of the bilateral pubic and ischial osteotomy in these cases of excessively large vaginal masses in feline patients.

All 3 feline patients developed mild and temporary splaying of the pelvic limbs, associated with abductor muscle resection/resuture rather than obturator nerve damage. This resolved shortly after surgery, and ongoing complications associated with this were not reported in any case. No urinary or fecal incontinence was reported with the bilateral pubic/ischial osteotomy approach in any of these cases. It is possible that a bilateral pelvic osteotomy technique, as reported in feline patients in this study, allows for a more meticulous dissection and reduction in urinary and fecal incontinence associated with inflammation/iatrogenic nerve damage.

The limitations of this study included those inherent to a retrospective case series. Errors in data collection, accurate initial data documentation, and retrieval will always remain a challenge. The 3- and 15-month follow-up telephone conversations were dependent on the owner’s opinion, and this was clearly subject to a possible error in the evaluation of the postoperative complications. In addition, recognition is made to the short 3-month postoperative follow-up in 2 of the 3 cases in this case series. While this follow-up is longer than that reported by Saitoh et al,⁴ a longer time period would be beneficial to allow for complications associated with bony healing. It is true, however, that no gait abnormalities or complications were reported at 12 weeks postsurgery in all 3 cases and at 15 months postsurgery in case 3. Vaginectomies in cats are not frequently performed; case numbers are limited despite a multicentric study aimed to negate this. Surgeon experience is likely to have a reflection on the clinical outcome.

The authors recognize, with complete transparency, that T-cell–rich B-cell lymphoma is a medically managed (and not a surgically managed) disease. However, in light of the initial cytological presentation, with no evidence of metastatic disease, surgery was deemed appropriate with the information presented. Within the aims and scope of this case series, inclusion of the case into a description of surgical technique and outcome appears relevant, yet the authors accept a limitation that the final diagnosis of case 1 could well affect the postoperative outcome of the surgical procedure. Indeed, case 1 showed splaying of the pelvic limbs and hematuria for a longer period than either case 2 or 3. While the case series here represents a heterogenous tumor neoformation type, the report aims to address a standardized technique for surgical management of extensive vaginal tumors in cats.

Bilateral pubic and ischial osteotomy for resection of vaginal tumors in cats is a successful surgical approach, offering good exposure for resection of large vaginal tumors, with minimal short-term postoperative complications. In these 3 cases, no urinary or fecal incontinence was noted. Considering the positive outcome seen with vaginectomies in cats in this case series, the use of bilateral pubic and ischial osteotomies in canine patients for vaginectomies in which tumors occupy a significant percentage of the height/width/length dimension of the pelvis may well be appropriate.

Disclosures

The authors have nothing to disclose. No AI-assisted technologies were used in the generation of this manuscript.

Funding

The authors have nothing to disclose.