Radiographic and Ultrasonographic Findings of Uterine Neoplasms in Nine Dogs

Introduction

Tumors of the uterus, including horn, body and cervix, are relatively rare in dogs and comprise 0.3 to 0.4% of all canine tumors.^1,2 Mesenchymal, tumors constitute the vast majority of cases of uterine tumors, leiomyomas representing 85 to 90% and leiomyosarcomas representing 10% of all tumors.³ Uterine adenomas, adenocarcinomas, fibromas, and lipomas have been reported occasionally.^1,4 Uterine tumors are rarely associated with clinical signs and consistent laboratory abnormalities have not been reported. Abdominal radiographs may confirm the presence of a soft-tissue mass in the caudal aspect of the abdomen and ultrasonography may help determine the origin of the mass. A definitive diagnosis of uterine tumor is usually obtained by histopathological examination of surgically excised specimens.³ To the authors’ knowledge, no studies have been published describing the ultrasonographic characteristics of uterine tumors in dogs. The purposes of the current study were to describe the radiographic and ultrasonographic signs observed with uterine tumors in nine dogs and to evaluate the usefulness of ultrasonography in determining the origin of the masses.

Materials and Methods

Nine female intact dogs with histologically confirmed uterine tumors that presented to the School of Veterinary Medicine, Aristotle University of Thessaloniki, Greece, between October 1996 and August 2011 were identified from the medical records and included in the current study. Patient signalment (age and breed), clinical history, physical examination, radiographic and ultrasonographic findings, and histopathological results were reviewed for each dog.

Abdominal radiographs in right lateral recumbency and thoracic radiographs in right lateral and sternal recumbency were obtained in all dogs. Radiographic reports were reviewed for the location, size, opacity, and margination of the lesions detected. Ultrasonographic examination of the entire abdomen was recorded on videotape and performed using either a 7.5 MHz mechanical sector transducer^a or an 8 MHz electronic microconvex transducer^b. Ultrasonographic reports were reviewed for the location, size margination, and echogenicity of the lesions seen and for evidence of uterine and other abdominal visceral abnormalities. The masses observed were classified, according to their ultrasonographic pattern, either as a solid lesion, a solid lesion with cystic component, or a cystic lesion, based on the percentage of well-defined anechoic cavities in relation to the amount of solid parenchyma. The cystic nature of the anechoic cavities was ascertained by lack of signal with spectral and color Doppler ultrasonography. Solid lesions had <10% anechoic cavities, solid lesions with cystic component had from 10 to 50% of anechoic cavities, and cystic lesions had >50% anechoic cavities.⁵ Histopathological diagnosis was obtained following surgical ovariohysterectomy in eight dogs and necropsy in one dog. Tissues were fixed in formalin and routinely sectioned and stained with hematoxylin and eosin.

Results

The signalment, clinical, radiographic, ultrasonographic, surgical, gross pathologic, and histopathological findings are reported in Table 1. Four were mixed-breed dogs, whereas only a single dog for each of the remaining breeds was seen. The mean age of the dogs was 10.1 yr (range, 8–14 yr).

TABLE 1 Summary of Signalment, Clinical Signs, Radiographic, Ultrasonographic, Surgical, Gross Pathological, and Histopathological Findings in Nine Dogs with Uterine Neoplasms

View Fullscreen

TABLE 1 (Continued)

View Fullscreen

The clinical signs at the time of presentation included palpable abdominal mass in nine dogs, vulvovaginal discharge in three dogs, constipation in four dogs, anorexia in two dogs, stranguria in four dogs, respiratory distress in one dog, and mammary gland tumor in one dog. Laboratory results revealed no abnormalities, except for leucocytosis and neutrophilia in two dogs.

Abdominal radiographic findings included a soft-tissue opacity between the urinary bladder and colon in six dogs (cases 2, 4, 5, 6, 8, and 9). In those six dogs, the outer margin of the lesion was well defined in three (cases 2, 4, and 5) and ill defined in the other three (cases 6, 8, and 9). The size of the lesion in the six dogs varied from 4 to 12 cm in diameter (Figure 1). In two of the remaining three dogs, an ill-defined, soft-tissue tubular structure associated with reduced serosal detail was found in one dog (cases 3) and an ill-defined, mineralized soft-tissue opacity in another dog (case 7), both in the midventral part of the abdomen (Figure 2). In the last of the remaining three dogs (case 1), no abnormality was detected, except for mild dorsal displacement of the descending colon. Cranial displacement of the urinary bladder in four dogs (cases 2, 4, 5, and 9), dorsal displacement of the descending colon in seven (cases 1, 2, 4, 5, 6, 8, and 9) and mild uterine enlargement in one (case 7) were also noted. The thorax was radiographically normal in eight dogs, and there was evidence of miliary pulmonary pattern, compatible with pulmonary metastatic disease in one dog (case 7).

View Figure

• Download as Powerpoint
• Download Figure

View Figure

• Download as Powerpoint
• Download Figure

Ultrasonographically, a mass at either the dorsal aspect of the abdominal urethra (cases 2, 4, 5, and 9) or urinary bladder (cases 1, 6, and 8) was observed in seven dogs in which a tentative diagnosis of uterine body/cervical mass was made (Figures 3, 4). In all seven dogs, a uterine body/cervical mass was confirmed at surgery. In the remaining two dogs (cases 3 and 7) an ill-defined mass in the midabdomen was detected. The origin of the mass was determined to be the uterine horn in one dog (case 3), whereas a mass in the uterine horn was included in the differential diagnosis of the other (case 7) as shown in Figure 5. In the latter two dogs, the uterine horn masses were confirmed either at surgery or postmortem examination. Ultrasonographically, the size of the masses varied from 3 to 12 cm in diameter. The masses were characterized as solid in three dogs (cases 5, 7, and 9) as shown in Figure 3, solid with cystic component in four dogs (cases 1, 4, 6, and 8), and cystic in two (cases 2 and 3) as shown in Figures 4 and 5. Diffuse hyperechoic foci within the masses were observed in three dogs (cases 2, 4, and 7), casting a clean shadow in one (case 7) as shown in Figure 4. There was ultrasonographic evidence of fluid accumulation in the uterus of two dogs (cases 3 and 8), thickened uterine wall (20 mm uterine diameter) with anechoic lesions embedded in the wall that was compatible with cystic hyperplasia in one (case 7), and small cysts in both ovaries in one (case 3). No evidence of abdominal metastatic disease or other abnormality of the abdominal viscera was noted.

View Figure

• Download as Powerpoint
• Download Figure

View Figure

• Download as Powerpoint
• Download Figure

View Figure

• Download as Powerpoint
• Download Figure

A uterine body/cervical mass and a uterine horn mass were confirmed after either exploratory surgery or necropsy in seven dogs and two dogs (cases 3 and 7), respectively. In three cases (cases 5, 7, and 9), in which the histopathological diagnosis was uterine leiomyoma (including case 7 that had also dystrophic mineralization and osseous metaplasia), gross pathologic examination of the tumor revealed a homogeneous firm cut surface). In case 7, necropsy was performed after the owner declined further treatment due to radiographic evidence of pulmonary metastatic disease from a mammary adenocarcinoma. In the remaining six dogs, the tumor had a heterogeneous cut surface with either hemorrhagic foci or mucin and blood filled cavities. In those six cases, histopathological diagnosis was uterine leiomyoma in three dogs (cases 2, 3, and 6), uterine adenocarcinoma in two (cases 1 and 8), and uterine fibroleiomyoma in one (case 4).

Discussion

A small number of dogs with uterine tumors were observed over a period of 15 yr. The incidence, breed, and age of the dogs in this study were similar to those reported in the literature. Uterine neoplasms were rare, affected middle-aged to older animals, and no breed predilection occurred.³

The clinical signs of uterine tumors depend on the tumor size, the presence of metastases, and other concurrent disease, such either a mucometra or pyometra.^6,7 Anorexia, constipation, stranguria, and vulvovaginal discharge, as observed in some dogs in this study, have also been reported in dogs with uterine tumors.⁸ A palpable abdominal mass that was found in all cases of this study may be considered the most consistent clinical finding of uterine neoplasms in dogs, but an accurate diagnosis could not be based on that finding alone.

The normal uterus is tubular, approximately 1 cm in diameter, located between the colon and the urinary bladder and usually not visible on survey radiographs.⁹ Uterine body neoplasms tend to achieve a large size between the descending colon and urinary bladder and may be seen radiographically as localized uterine enlargement that may displace the descending colon dorsally.¹⁰ That finding was observed in six of seven dogs with uterine body/cervical tumors in this series and was considered consistent with uterine body/cervical neoplasms in intact bitches. However, segmental pyometra, early pregnancy, stump pyometra, or granuloma in a neutered animal may appear similar on survey radiographs and should be included in the differential diagnosis.^11–13

The size and the location (cranial or caudal) of the uterine masses may determine the clinical and radiographic findings, and a mass at the level of the body of the urinary bladder may produce urinary incontinence. A mass at the level of the urinary bladder neck may cause hydroureter, and a mass adjacent to the urethra may produce urinary outflow obstruction of the urinary bladder and dorsal displacement of the cranial part of the rectum.¹⁴ As expected, the study authors found that the larger the uterine body/cervical mass, the bigger the displacement of adjacent viscera. Also, a mass smaller than the diameter of the adjacent viscera will not have a mass effect and may not be visible on plain radiographs, as was the case in one dog in the current study (case 1).

The radiographic examination of the abdomen in the cases of this study included a single lateral view. If uterine horn enlargement does not exceed the diameter of an intestinal loop, it is unlikely that a ventrodorsal view of the abdomen would show changes related to the uterus. In cases when uterine enlargement does exceed the diameter of an intestinal loop, both the lateral and ventrodorsal views are likely to show abnormalities. Although the authors agree that a ventrodorsal abdominal view may be particularly useful to characterize focal uterine horn enlargement and masses, in the context of the authors’ daily clinical procedures where most dogs suspected of uterine disease would be further assessed ultrasonographically, it was considered justified to perform a single lateral radiograph of the abdomen. Uterine horn masses usually appear on radiographs as either well or ill-demarcated, round lesions in the midabdomen that may displace intestinal loops.^15,16 Similar findings were also observed in dogs included in this study with uterine horn tumors (cases 3 and 7). However, those radiographic findings were not characteristic of uterine horn tumors because they have also been observed in a number of uterine horn diseases (i.e., cystic endometrial hyperplasia, localized pyometra, uterine adenomyosis) as well as in cases of mesenteric masses.^13,15,17

Ultrasonography was helpful in demonstrating the uterine body/cervical origin of a mass in all dogs, and the anatomic location of the mass between either the abdominal urethra or bladder neck and the descending colon enabled to confirm the tissue of origin.¹⁸ Although it is possible to differentiate the uterine body and cervix in the normal, nongravid uterus with ultrasound, it may be difficult to ascertain whether a uterine mass had either a body or cervical origin.

Ultrasonography was helpful in showing uterine horn tumors in two dogs included in this study (cases 3 and 7), but was able to detect the horn origin of the mass in only one dog (case 3). In the other dog, the horn origin of the mass was included in the list of differential diagnoses. Ultrasound was a reliable technique to determine the origin of an abdominal mass when there was recognizable parenchyma to which the mass could be traced.⁵ Normal uterine horns are difficult to identify by ultrasound because they become lost in small bowel echoes and mesenteric fat; however, fluid accumulation in the lumen of uterine horns makes them easily visible.¹⁹ It is possible that enlargement and fluid accumulation in the horns was helpful in tracing the origin of the mass in case 3, but in the dog in which the origin of the mass was not delineated precisely (case 7), there was no fluid accumulation in the uterine horns. In women, recognition of the origin of gynecologic tumors by ultrasound is difficult and gastrointestinal stromal tumors were misdiagnosed as uterine leiomyomas.²⁰ Similarly, uterine horn adenomyosis in a dog was misinterpreted as gastrointestinal tract tumor.¹⁵ The authors support the hypothesis that it may not be possible to determine that a mass originates in a uterine horn unless there is associated fluid distension of the uterus. Searching for a “sliding organ sign” may be useful for distinguishing a possible mass confluent with the uterus from one separate from it.^20,21

There is a paucity of reports on the ultrasonographic characteristics of uterine tumors in dogs. Leiomyomas and fibroleiomyomas are composed of spindle-shaped smooth muscle cells arranged in whorl-like patterns separated by variable amounts of fibrous connective tissue and have a firm homogeneous cut surface in gross pathologic examination. This histological pattern corresponds to a typical ultrasonographic appearance of a solid echogenic mass. The echogenicity of a mass depends on the relative ratio of connective tissue to smooth muscle, the extent of degeneration, and the presence of either necrosis or dystrophic mineralization. The presence of necrosis, degeneration, and/or mineralization within the tumor is likely to cause a mass to have a heterogeneous echogenicity and echotexture. With an increased fibrous component in a mass, there is increased echogenicity of the mass. In this study, a solid mass was found in three leiomyomas for which the gross pathologic examination demonstrated a firm, homogeneous cut surface. Similarly, in another case report, a uterine leiomyoma was observed as isoechoic to the surrounding uterine tissue, projecting in to the uterine lumen.²² As a tumor enlarges, it may outgrow its blood supply, resulting in ischemia, cystic degeneration and necrosis, producing blood-filled spaces in the tumor that appear ultrasonographically heterogeneous, with anechoic cyst-like cavities into the solid mass.²³ In this study, three leiomyomas and one fibroleiomyoma demonstrated ultrasonographically anechoic, cyst-like lesions, which, on gross pathological and histopathological examination, were confirmed as blood-filled cavities due to necrosis. However, a similar ultrasonographic pattern has been reported with a leiomyosarcoma, and it would therefore appear that it is not possible to differentiate leiomyomas, fibroleiomyomas, and leiomyosarcomas from their ultrasonographic appearance alone.¹⁶ In women, leiomyomas are commonly solid masses; hypoechoic or isoechoic to normal surrounding uterine tissue; but different forms of internal degeneration, hemorrhage, or proteolytic liquefaction cause anechoic areas in the tumor.²³ If anechoic areas are extensive enough, they may have a cystic appearance and can mimic predominantly cystic or multiseptated adnexal or ovarian masses.²⁴ Furthermore, this variable ultrasonographic pattern may also mimic numerous pelvic conditions and add to diagnostic confusion.²⁴

A solid mass with cystic component due to mucin and blood-filled cavities in the tumor was also demonstrated ultrasonographically in two adenocarcinomas in this study. The ultrasonographic pattern may resemble the pattern described for leiomyomas with cystic component. In women, most endometrial adenocarcinomas are either diffusely or partially echogenic with either irregular or poor endometrial thickening that usually measures >5 mm.²⁵ Although rare, cystic changes in adenocarcinomas are probably due to associated endometrial hyperplasia and necrosis.^26,27 Because cystic changes may be present in either benign or malignant endometrial tumors, it is not possible to use ultrasound, without biopsies, to exclude a malignancy.^16,27

Many neoplastic diseases appeared to have hyperechoic foci due to either fibrosis or dystrophic mineralization within the tumors, with related acoustic shadowing if there was enough sound attenuation.¹⁹ The hyperechoic foci seen in a leiomyoma (case 2) and in a fibroleiomyoma (case 4) in this study may reflect the abundant fibrous connective tissue that those neoplasms were found to contain.²⁸ In women, many leiomyomas may demonstrate areas of acoustic attenuation or shadowing due to dense fibrosis within the substance of the tumor.²³ Hyperechoic foci casting a clean shadow indicative of mineralization and/or osseous metaplasia, as confirmed by histopathology in one leiomyoma in this study (case 7), have also been reported in a uterine leiomyosarcoma in the dog and in uterine leiomyomas in women.^16,23,27 In older women, calcification in a uterine tumor may appear as a curvilinear echogenic rim, which may simulate the outline of a fetal head.²³

Uterine tumors are easily differentiated from segmental pyometra and pregnancy because the latter have a typical ultrasonographic appearance. The presence of a hyperechoic capsule is compatible with a granuloma; however, uterine granulomas may have a similar appearance to neoplastic tumors ultrasonographically. The only way to differentiate granulomas from tumors is either by fine-needle aspiration or biopsy, both of which can be performed under ultrasound guidance.^15,18,29

Conclusion

Ultrasonography was a useful method of detecting either uterine body or cervical tumors in dogs. However, it was not possible to ascertain that a mass originated in a uterine horn unless there was also associated evidence of involvement of the uterine horn to which the mass could be traced. The ultrasonographic appearance of uterine tumors in dogs was variable and the only way to differentiate the type of neoplasm was by taking a biopsy of the mass. Ultrasonographic techniques, such as power Doppler imaging, contrast ultrasound, and elastography, as well as contrast enhanced computed tomography and magnetic resonance imaging could all be used for further investigation of the different tumor types.