Antemortem diagnosis of metastatic endometrial adenocarcinoma using computed tomography in a Vietnamese Pot-bellied sow

A 13-y-old Vietnamese Pot-bellied sow (Sus scrofa) was presented for evaluation because of a 5-d history of mucohemorrhagic vulvar discharge, lethargy, polydipsia, respiratory difficulty, and a 4-mo history of weight loss and intermittent coughing. Urination and bowel movements were reportedly normal. The sow was acquired as a 4-y-old, and its medical history included a dog attack resulting in the traumatic amputation of both pinnae several years prior, severe dental malocclusion, and severe endodontic and periodontal disease for which she had been receiving extra-label gabapentin and carprofen daily for 2 y. To the owners’ knowledge, the sow had never farrowed.

On presentation, the sow was quiet and alert with normal vital findings and bilateral forelimb lameness. The most recent weight measurement was recorded 5 y prior; she had lost ~26 kg since that time, and she appeared moderately underconditioned. Notable physical examination findings included mildly increased bronchovesicular sounds in all lung fields, with markedly increased respiratory effort and an intermittent dry cough when placed in lateral recumbency. Mucous membranes were pink and moist with rapid capillary refill, and membranes remained pink when in lateral recumbency. Cardiac auscultation revealed a regular rhythm with no appreciable murmur. She had an excessively pendulous and guarded abdomen, although no discrete mass was found on abdominal palpation. No feces were obtained on digital rectal palpation, which was otherwise within normal limits. Vulvar discharge was not observed.

Given the lack of patient cooperation, blood could not be collected for a CBC. Serum biochemistry was prioritized to assess major organ function and metabolic stability, revealing azotemia (creatinine 361 μmol/L; RI: 53–203 μmol/L), hypercalcemia (3.4 mmol/L; RI: 2.2–2.7 mmol/L), and hyperkalemia (5.0 mmol/L; RI: 3.4–4.9 mmol/L). Metastatic neoplasia originating from the reproductive tract was suspected, and other considerations included benign reproductive neoplasms, dynamic tracheal and bronchial collapse, bacterial or fungal pneumonia, chronic kidney disease, pyelonephritis, and lower urinary tract infection. Abdominal and thoracic radiographs and abdominal ultrasound were offered as initial imaging tests, although computed tomography (CT) was ultimately elected to minimize the number of sedation events and stress while gaining the most detailed information.

CT of the abdomen, which included the caudal thorax, was acquired (SOMATOM Sensation 40 slice helical CT scanner; Siemens) in sternal recumbency under brief general anesthesia. CT was performed with 120 kVP, 250 mAs, image matrix 512 × 512, and 3-mm slice thickness. Images were obtained in a soft tissue (WL 30, WW 300) algorithm. Positive-contrast medium was not injected because venous access could not be secured. A large, multilobulated, cranially bifurcated, tubular, soft-tissue attenuating structure with variably sized fluid-attenuating structures, consistent with the uterus, occupied most of the caudal abdomen, extending into the pelvic canal and situated between the urinary bladder and descending colon. Associated with the left cranial aspect of the uterus was a large, rounded, mixed soft-tissue and fluid-attenuating mass with striated mineral attenuation and a mildly irregular lateral margin ( Fig. 1 ). The ovaries were not identified definitively. Both kidneys had mildly lobulated margins, and their internal architecture was almost entirely composed of many variably sized, fluid-attenuating cystic structures. There was a poorly defined, mildly lobulated, soft-tissue attenuating structure in the region of the left medial iliac lymph node and a poorly defined, mildly lobulated conglomeration of soft-tissue attenuating material adjacent to the portal vein and caudal vena cava at the level of the transverse fissure of the liver. The urinary bladder was small and compressed by the uterus. A small, rounded, mineral-attenuating structure was associated with the right liver lobe. The caudal aspects of the lungs contained innumerable, variably sized, soft-tissue attenuating nodules (Fig. 1). The cumulative findings were highly suggestive of metastatic neoplasia. Given the animal’s age, deteriorating body condition, and poor long-term prognosis, euthanasia was elected.

OPEN IN VIEWER

Figure 1.

Dorsal (A) and transverse (B–D) computed tomographic images, including the caudal thorax and abdomen in a soft-tissue window in a Vietnamese Pot-bellied sow. The patient’s right side is to the left. Variably sized soft-tissue attenuating nodules are visible within the caudal lung fields (white arrowheads; Figs. 1A, 1B). The uterus occupies most of the caudal abdomen and appears multilobulated with multifocal fluid-attenuating structures (arrow; Fig. 1A). A mixed soft-tissue and fluid-attenuating mass with striated mineral attenuation is at the cranial aspect of the left uterine horn (black arrowheads; Figs. 1A, 1C).

On postmortem examination, a 20 × 9.5 × 9.5-cm, white, firm nodule was present in the mid-body of the left uterine horn. On cut section, the mass was tan, solid, homogeneous, and transmural ( Fig. 2A ). Similar smaller masses infiltrated the remaining uterine wall and the left broad ligament, spleen, liver, aorta, and diaphragm. Masses were also found throughout the lungs ( Fig. 2B ) and along the dorsal aspect of the thoracic cavity, adhering to the aorta. Upon uterine dissection, the endometrial mucosa was diffusely irregular with numerous variably sized fluid-filled cysts (Fig. 2A). Other findings included severe dental malocclusion, renal cysts, and bilateral degenerative joint disease in the radiocarpal, carpometacarpal, and elbow joints.

OPEN IN VIEWER

Figure 2.

Gross and histologic findings of endometrial adenocarcinoma in a Vietnamese Potbellied sow. A. A tan, solid uterine mass (endometrial adenocarcinoma; yellow arrow) surrounded by numerous fluid-filled cysts (cystic endometrial hyperplasia; white arrow). B. Numerous tan, solid nodular metastases of endometrial adenocarcinoma in the lungs. C. Neoplastic cells—in small nests, cords, and ill-defined tubular structures supported by thin stromal trabeculae—arising from the endometrium and invading the myometrium (asterisk). H&E. D. Neoplastic cells and endometrial epithelium have strong, diffuse, cytoplasmic immunoreactivity to pan-cytokeratin (asterisk). On the right, the endometrial glands are severely dilated and variably sized, supported by branching endometrial stroma, characteristic of cystic endometrial hyperplasia. Immunohistochemistry, pan-cytokeratin.

Histologically, the uterine mass was a proliferation of polygonal neoplastic cells that originated from the endometrium and invaded the myometrium. The neoplastic cells were in small nests, cords, and ill-defined tubular structures supported by thin trabeculae of stroma ( Fig. 2C ). The cells had a large amount of eosinophilic cytoplasm. Nuclei were round with frequent indentation, vacuolated to finely stippled chromatin, and multiple small nucleoli. Anisocytosis and anisokaryosis were marked, with frequent karyomegaly, binucleation, and multinucleation. Mitotic figures, some of which were atypical, were 9 in 2.37 mm². Neoplastic cells frequently invaded lymphatic vessels and had strong, diffuse immunoreactivity to pan-cytokeratin ( Fig. 2D ). In the remaining endometrium, the endometrial glands were severely dilated and variably sized, supported by branching endometrial stroma, characteristic of cystic endometrial hyperplasia (CEH). The glands were frequently filled with a large amount of amorphous eosinophilic material. The same neoplastic cells were also evident in the spleen, liver, aorta, diaphragm, and lung. Gross and histologic findings confirmed the clinical diagnosis of disseminated endometrial adenocarcinoma with concurrent severe CEH.

Uterine neoplasms have been reported in the aged sow, but adenocarcinomas appear to be uncommon.^8–10,17,24 One study reported uterine neoplasms in 20 of 32 (62.5%) Pot-bellied sows undergoing ovariohysterectomy. ¹⁰ Most of the tumors in that study were of smooth muscle origin (leiomyomas or leiomyosarcomas), with only 2 of 20 being adenocarcinomas. Other reported uterine neoplasms include adenoma, fibroma, fibroleiomyoma, carcinosarcoma, and choriocarcinoma. The pathology findings in our case are similar to those in previously reported cases of uterine adenocarcinoma in miniature pigs.^4,7,8 In the 4 reported cases, 3 cases had evidence of metastasis or vascular invasion, and metastatic foci were noted in the liver, lung, and lymph nodes. Metastasis was more widespread in our case. Concurrent CEH was clearly noted in 3 of the 4 previously reported cases. CEH appears to be common in miniature pigs. In the cited study of 32 pigs, 24 (75%) animals had CEH. ¹⁰ In humans, endometrial hyperplasia is a known risk factor for the development of endometrial carcinoma. The pathogenesis involves long-term exposure to unopposed estrogen and a variety of molecular alterations, including microsatellite instability and mutations in the PTEN, k-RAS, PIK3CA, FGFR2, and beta-catenin genes. ¹

Radiography and transabdominal ultrasonography are commonly used in swine to investigate the reproductive tract when there is a clinical suspicion of neoplasia, although both modalities are limited in their ability to define tumor margins, size, and extent of metastasis, regardless of species.^5,16,18,24 Intrapelvic masses are challenging to visualize ultrasonographically, and copious adiposity in the Vietnamese Pot-bellied patient adds an obstacle to acquiring information and diagnostic image quality. ²⁴ In humans, transvaginal ultrasonography can be used to circumvent these hindrances; however, given the lack of patient cooperation, unavailability of specialized probes, and anatomic differences, transvaginal and transrectal ultrasonography are not commonplace in porcine practice. ¹² In dogs, CT has been demonstrated to be superior to radiography for defining intrapelvic masses. ²² Because the patient’s disposition may warrant general anesthesia to facilitate imaging, CT may be the more cost-effective and time-efficient modality to image the entire patient, especially if there is intrapelvic localization and extrapelvic metastasis. ⁶

The literature contains little information about CT features of endometrial adenocarcinoma and other neoplasms of the reproductive tract in swine. We identified only one case report documenting CT findings of a mixed CEH and endometrial adenocarcinoma in a Vietnamese Pot-bellied sow with no distant metastases in a search of PubMed using the search terms “potbellied sow” and “endometrial adenocarcinoma”. ⁴ Similar to our case, the mass had mixed soft-tissue and fluid attenuation with irregular mineral attenuation. In humans, contrast-enhanced CT examination of endometrial adenocarcinoma reveals a thickened endometrium or dilated endometrial cavity that is hypoattenuating and hypoenhancing and may or may not have solid nodules.^5,6,15,23 This appearance is not specific to endometrial adenocarcinoma; other neoplasms, both malignant and benign, can have a similar appearance.^2,5,15 In any species, no pathognomonic CT findings can definitively distinguish one malignant uterine neoplasm from another.^1,20 Based on the CT appearance in our case, other differentials included leiomyosarcoma, fibrosarcoma, lymphoma, leiomyoma, and CEH without concurrent malignancy.^4,5,10,19,23

Administration of positive-contrast medium may have helped further characterize the CT architecture of the uterine neoplasm in our case. Multifocal non-enhancing cystic areas were reported in 2 miniature pigs with uterine adenocarcinoma and CEH. ⁴ In dogs, malignant pelvic neoplasms were more likely to have heterogeneous contrast enhancement, whereas benign masses were more likely to have more uniform contrast enhancement, attributable to necrotic foci from tumor growth beyond its blood supply. ²² Positive-contrast media may have improved sensitivity for identifying lymphatic involvement and distant metastasis.^13,15 Even without contrast enhancement in our case, a malignant etiology was prioritized, given the nodules in the lungs. ¹⁴ In humans, endometrial adenocarcinoma has a proclivity for pulmonary metastasis.^13,21 A similar metastatic pattern was reported in an 8-y-old sow with endometrial adenocarcinoma. ³

CT has limitations. In female dogs, uterine and vaginal diameter and contrast enhancement patterns may vary with the stage of the estrous cycle, ¹¹ which may also be true in sows. Masses may distort normal anatomy and impede localization of the cervix and vagina, warranting additional imaging techniques, such as CT vaginography, to confirm that a tumor is associated with the uterus. ¹¹ MRI is more accurate than CT in assessing myometrial invasion in humans, particularly when dynamic contrast enhancement is utilized.^6,15 MRI in humans has also been reported to produce details that would enable discrimination between various tumor types.^5,20 Unfortunately for the swine practitioner, accessibility, cost, and a longer anesthetic episode make MRI less feasible than CT. ¹⁶ We found no literature on the MRI features of uterine neoplasms in swine.

Additional investigation into the CT features of endometrial adenocarcinoma in swine is needed. For now, the swine practitioner is left to extrapolate from CT features of other species. Histologic confirmation is ultimately necessary to confirm diagnosis; however, CT may increase confidence in diagnosis, prognostication, surgical planning, and surveillance for recurrence in similar cases.