Canine Gastric Adenocarcinoma and Leiomyosarcoma: A Retrospective Study of 21 Cases (1986–1999) and Literature Review

Introduction

Gastric neoplasia comprises <1% of all canine malignancies.¹ Most gastric tumors are malignant and of epithelial origin.¹² Adenocarcinomas have been reported most frequently and account for 42% to 72% of canine malignant gastric tumors.^3–13 These tumors have been diagnosed in middle-aged to older dogs.^{24–610–1214–17} Some breeds, including the rough-coated collie,¹⁰ Staffordshire bull terrier,³⁴¹⁰ and Belgian shepherd dog,¹¹¹² may be predisposed to gastric carcinomas. Other studies have not found a breed predisposition.²⁷ Historically, male dogs are more commonly affected.^{246–810–121416}

Gastric carcinomas are usually located in the pyloric antrum⁷ or along the lesser curvature.¹² Grossly, three variations of this tumor have been identified. Diffuse neoplastic infiltration of the gastric wall can produce a thickened, nondistensible stomach known as linitis plastica or “leather-bottle stomach.” Gastric adenocarcinoma may also manifest itself as a discrete, polyploid lesion. The final variation consists of mucosal plaques with ulceration.⁹ Histopathologically, gastric adenocarcinomas may be classified as a diffuse type with randomly arranged, malignant epithelial cells or an intestinal type with neoplastic cells that are well organized into distinct, glandular structures.⁸ Unfortunately, the clinical significance of these histopathological types is not currently known.

Clinical signs suggestive of a space-occupying gastric mass are often seen and include chronic vomiting, anorexia, and weight loss. In addition, hematemesis, melena, abdominal discomfort, and a microcytic, hypochromic anemia may be present.^{2–69–1114–20} Plain and contrast abdominal radiographs have been the principal imaging modalities used to document the presence of a suspected gastric mass. Abnormal gastric motility, delayed gastric emptying, gastric filling defects, and loss of the normal rugal fold pattern have been associated with gastric masses.^29–111320

Complete surgical excision is the treatment of choice for gastric adenocarcinomas. However, advanced disease with extensive gastric involvement and metastasis to regional lymph nodes and liver often makes complete resection more difficult. Survival times are usually <3 months after the onset of clinical signs in untreated dogs.²¹ The survival time may be increased to <8 months following tumor excision.^513–1518 Two dogs are reported to have survived 35 months¹¹ and 5 years¹³ after gastric resection. Adjuvant chemotherapy has not been evaluated in dogs with gastric carcinomas.

The purposes of this study are to retrospectively describe 21 cases of canine gastric adenocarcinoma and leiomyosarcoma and to compare the findings in these cases to those in the literature. Signalment, clinical signs, imaging modalities, treatment, outcome, metastatic rate, and metastatic pattern are the focuses of this study.

Materials and Methods

Results

From 1986 through 1999, 21 dogs were included in this study with a histopathological diagnosis of gastric adenocarcinoma (n=19) or gastric leiomyosarcoma (n=2). Age of dogs at the time of presentation to the hospital ranged from 3.5 to 13 years (mean, 8.40±2.82 years; median, 8.25 years). Four (19%) dogs were ≤5 years of age. Seventeen (81%) dogs were >5 years old. Sixteen dogs were purebred, and five were mixed-breed dogs. Purebred dogs consisted of shih tzus (n=2), Newfoundlands (n=2), golden retrievers (n=2), Labrador retrievers (n=2), keeshond (n=1), chow chow (n=1), Scottish terrier (n=1), Alaskan malamute (n=1), Norwich terrier (n=1), Norwegian elkhound (n=1), Siberian husky (n=1), and border collie (n=1). These dogs weighed 7.5 to 56.8 kg (mean, 25.86±13.92 kg; median, 19.60 kg). Four dogs weighed ≤15 kg, 12 dogs weighed between 15 and 30 kg, two dogs weighed between 30 and 45 kg, and three dogs weighed between 45 and 60 kg. Ten dogs were spayed females, eight were intact males, and three were castrated males. Overall, there were 11 males and 10 females [Table 1].

Nineteen (90.5%) dogs had histopathological confirmation of an adenocarcinoma. Two (9.5%) dogs suffered from gastric leiomyosarcomas. The adenocarcinomas were located in the pylorus (n=10), all regions of the stomach (n=5), the fundus and body (n=1), or were not specified (n=3). The leiomyosarcomas were located in the pylorus (n=1) or in the fundus, body, and pylorus (n=1).

Vomiting, partial anorexia, and weight loss were the most common clinical signs. Twenty of 21 (95.2%) dogs were vomiting for 1 to 120 days (mean, 40 days; median, 30 days) prior to presentation. Anorexia in 11 of 21 (52.4%) dogs was present for 2 to 90 days (mean, 27.3 days; median, 5 days) prior to presentation. Weight loss over 7 to 90 days (mean, 57.3 days; median, 5 days) was noted in 11 of 21 (52.4% ) dogs. Other clinical signs consisted of melena (n=8; 38.1%), lethargy (n=6; 28.6%), abdominal discomfort (n=3; 14.3%), ptyalism (n=2; 9.5%), and abdominal distention (n=2; 9.5%).

Thoracic radiographs were taken in 17 cases. Thirteen of 17 (76.5%) dogs had normal thoracic radiographs. Abnormalities noted in the remaining four (23.5%) dogs consisted of pulmonary metastasis (n=1), aspiration pneumonia (n=1), microcardia associated with dehydration (n=1), and bony proliferation of the fourth sternebra (n=1).

Abdominal radiographs were performed in 17 dogs. Twelve of 17 (70.6%) dogs had radiographic abnormalities. Gastric abnormalities were noted in 11 of 17 (64.7%) dogs. These gastric changes included a mass (n=3), thickening (n=3), dilatation (n=3), dilatation-volvulus (n=1), and caudal displacement of the gastric axis on lateral radiographs (n=1). Hepatomegaly, multiple liver masses, and splenomegaly were documented once each in three separate dogs. Five (29.4%) dogs had normal abdominal radiographs.

Positive-contrast gastrograms were performed in nine cases. Eight of nine (88.9%) dogs had one or more documented gastric lesions consisting of mucosal filling defects (n=5), a mass (n=3), and wall thickening (n=1). One dog had a normal study. This dog had diffuse infiltration of the cardia and distal esophagus with gastric adenocarcinoma.

Sixteen dogs had abdominal ultrasound examinations. All dogs had one or more gastric abnormalities. Gastric thickening (n=12), a gastric mass (n=3), and an outflow obstruction (n=2) were observed. Hepatic nodules (n=3), splenic nodules (n=3), and mesenteric lymphadenopathy (n=3) were additional ultrasound findings.

A single, abdominal imaging modality (radiographs [n=1], ultrasound [n=4]) was utilized in five of 21 (23.8%) dogs. The diagnosis in 16 (76.2%) dogs relied on multiple abdominal imaging modalities. Abdominal radiographs and a positive-contrast gastrogram were performed in four dogs. Abdominal radiographs and ultrasound were carried out in seven dogs. All three diagnostics were employed in five dogs.

In three dogs, endoscopic evaluation was performed. Each dog had an abnormality. A gastric mass was visualized in one dog. The other two dogs had thickened gastric mucosa with multiple erosions. Two of these dogs had endoscopic biopsies, with one interpreted as normal and the other as an adenocarcinoma.

Two dogs with adenocarcinomas were treated with chemotherapy alone [Table 2]. The diagnosis in the first dog was made via an endoscopic gastric biopsy. This dog (case no. 4) then received 5-fluorouracil (150 mg/m² intravenously [IV]) and cyclophosphamide (50 mg/m² per os [PO] for 4 days) every 2 weeks for two cycles. The owner elected euthanasia 9 weeks after diagnosis because of progression of vomiting [Table 3]. The other dog (case no. 9) was diagnosed with gastric adenocarcinoma and metastatic disease involving the lungs, liver, and gastric lymph nodes. Chemotherapy consisted of eight cycles of the FAC protocol (doxorubicin, 25 mg/m² IV, and cyclophosphamide, 75 mg/m² PO for 4 days on week 1; 5-fluorouracil, 150 mg/m² IV given on weeks 2 and 3). Thoracic radiographs taken 2 weeks after starting chemotherapy revealed a 50% decrease in number and size of the pulmonary metastases. Abdominal ultrasound performed at the same time documented enlarged liver masses, multiple gastric masses, and gastric lymphadenopathy. Clinically, the dog was much improved with occasional vomiting and partial anorexia lasting 1 to 2 days. After eight cycles of the FAC protocol, there was ultrasonographic progression of the liver masses, both in size and in number. At this time, the chemotherapy protocol was changed to cis-platinum (60 mg/m² IV over 6 hours) every 3 weeks for two cycles. This dog died 7.5 months after diagnosis because of acute anuric renal failure.

Surgery alone was performed in 13 dogs with gastric adenocarcinomas. Eight dogs had surgical treatments, while five dogs received surgical biopsies. One dog (case no. 1) suffered from a gastric adenocarcinoma and concurrent gastric dilatation-volvulus. At surgery, the stomach was repositioned, and a partial gastrectomy (the fundus and a third of the stomach were excised) and splenectomy were performed. This dog was euthanized 2 days postoperatively because of disseminated intravascular coagulation and worsening ventricular arrhythmias.

Three dogs (case nos. 2, 5, 8) with gastric adenocarcinoma were not treated after completion of their diagnostics and were euthanized at the owners’ request because of worsening clinical signs and the perceived poor prognosis associated with this disease.

In six dogs (case nos. 3, 6, 7, 10, 11, 21), metastasis was detected for the first time at surgery [Table 4]. An incisional biopsy was taken from the primary lesion and suspected metastatic sites. All biopsies confirmed the presence of metastatic gastric adenocarcinoma. Metastatic sites included gastric lymph nodes (n=6), omentum (n=2), and esophagus (n=1). These six dogs were euthanized at the owners’ request at the time of surgery (n=3) or when the biopsy results were available 1 to 3 days after surgery (n=3).

In two dogs (case nos. 14, 18), the gastric carcinoma was deemed by the surgeon to be too extensive for successful excision. The primary lesion was biopsied, and then a diversionary procedure (a side-to-side gastrojejunostomy) was performed. Both dogs had an uneventful recovery and were discharged from the hospital. The referring veterinarians euthanized these dogs 4 and 5 weeks postoperatively because of recurrence of vomiting and anorexia.

In four dogs (case nos. 12, 15–17), the gastric adenocarcinoma was locally confined to the pylorus. Each dog received a Billroth I (i.e., gastroduodenostomy) and gastric biopsy. One of these dogs was euthanized 3 days postoperatively because of persistent vomiting. The other three dogs were asymptomatic until the time of euthanasia at 6 weeks (n=2) and 10 months (n=1). At these times, vomiting and anorexia had recurred.

One dog (case no. 13) with a gastric carcinoma and metastasis to the right limb of the pancreas and descending duodenum received a Billroth I, partial pancreatectomy, cholecystojejunostomy, and jejunostomy tube. The dog did not have any postoperative complications and was discharged from the hospital 1 week after surgery. The dog remained clinically normal 2 and 4 weeks postoperatively. Four weeks after surgery, chemotherapy (doxorubicin, 25 mg/m² IV, and cyclophosphamide, 50 mg/m² PO for 4 days) was initiated. This dog did not return to the hospital for additional chemotherapy and was euthanized 9 weeks after diagnosis because of recurrent vomiting [Tables 2, 3].

Overall, 14 of 19 (73.7%) dogs with gastric adenocarcinoma had metastasis [Table 4]. Metastatic sites included gastric and hepatic lymph nodes (n=10), omentum (n=3), duodenum (n=2), liver (n=2), pancreas (n=1), spleen (n=1), esophagus (n=1), adrenal glands (n=1), and lungs (n=1). The presence of metastasis was detected by diagnostics done prior to surgery (n=4) and was later confirmed by postmortem findings in two of these four cases. Surgical biopsies were necessary in 10 dogs to document the presence of metastasis. Three of the 10 dogs requiring surgical confirmation of metastasis also had necropsy evidence of metastasis. Five of 19 (26.3%) dogs with adenocarcinomas did not have evidence of metastasis at the time of initial evaluation [Table 4].

Both dogs (case nos. 19, 20) with gastric leiomyosarcoma had a Billroth I performed. Abdominal metastases involving the liver (n=2) and duodenum (n=1) were recognized at the time of surgery and later confirmed histopathologically. These dogs were clinically normal and were discharged from the hospital 1 week later. At the time of suture removal 2 weeks later, both dogs remained asymptomatic. Four weeks after surgery, one dog developed a hemoabdomen with ultrasonographic progression of his liver masses. The owners elected euthanasia at this time. The other dog developed melena and new liver masses that were detected with ultrasound performed 3 weeks after surgery. Chemotherapy (doxorubicin, 25 mg/m² IV, and cyclophosphamide, 50 mg/m² PO for 4 days) was instituted. This dog received two cycles of chemotherapy and then was lost to follow-up 2.5 months after surgery.

Discussion

Gastric neoplasia is an uncommon diagnosis in canine patients. Published reports of gastric adenocarcinomas and leiomyosarcomas are scarce. Most are case reports and small case series with little clinical information on treatment and outcome. Studies to date have focused on the histopathological features and postmortem findings in these cases. Most neoplasms are malignant and of epithelial origin.¹² Gastric adenocarcinoma has been the single most common diagnosis, accounting for 42% to 72% of all documented cases.^3–13 In this study, adenocarcinoma (19/21 dogs) predominated, with only two dogs diagnosed with gastric leiomyosarcoma. Leiomyosarcomas are the second most common intestinal tumor and the most common intestinal sarcoma in dogs.⁷

Gastric adenocarcinoma is considered a neoplasm of older dogs.^{24–610–1214–18} In the literature, the ages of 77 dogs with gastric adenocarcinoma range from 3.5 to 15 years (mean, 9.42±2.55 years; median, 9.5 years). The findings of this study are similar to those in the literature, with ages ranging from 3.5 to 13 years (mean, 8.40±2.82 years; median, 8.25 years). However, this disease can occur in young dogs. In this study, 19% of the dogs were <5 years of age. In contrast, only four (5%) dogs in the literature were this young. This difference, perhaps, is due to the owners’ desires to pursue a diagnosis sooner.

A breed predilection was not appreciated in this study. Dogs of all sizes were affected, with weights ranging from 7.5 to 56.8 kg (mean, 25.86±13.92 kg; median, 19.60 kg). Others have found the rough-coated collie,¹⁰ Staffordshire bull terrier,³⁴¹⁰ Belgian shepherd dog,¹¹¹² and chow chow¹⁸ to be at increased risk for the development of this tumor. A genetic mechanism is suggested in the pathogenesis of the disease in Belgian shepherd dogs, since pedigree evaluation revealed eight related dogs.¹¹¹²

The dogs in this study presented with chronic histories and signs compatible with a space-occupying mass. Vomiting was the most consistent finding in 20 (95.2%) dogs. Other classic signs included anorexia, weight loss, melena, lethargy, abdominal pain, and ptyalism. These observations mirror those in the literature.

Thoracic radiographs are an important diagnostic tool when staging any neoplasm. Thoracic radiographs were abnormal in four (23.5%) dogs. Interestingly, only one dog had radiographic evidence of pulmonary metastasis at the time of initial evaluation.

Multiple abdominal imaging modalities were used to evaluate the majority (16/21; 76.2%) of dogs. Abdominal radiography followed by abdominal ultrasound was the most common combination used in seven of 21 (33%) dogs. Abdominal radiographs alone demonstrated a gastric abnormality in 11 of 17 (64.7%) dogs. A discrete gastric mass was visualized in three dogs. A positive-contrast gastrogram delineated a gastric abnormality in eight of nine dogs. A gastric mucosal filling defect was noticed in five dogs, followed by a mass in three dogs, and gastric thickening with a loss of the normal rugal pattern in one dog.

Abdominal ultrasound was performed in 16 dogs. All dogs had abnormal gastric findings. Dr. Penninck et al.¹⁸ described pseudolayering of the gastric wall in 14 of 16 dogs with gastric epithelial tumors. This ultrasonographic feature was thought to correlate to the unevenly layered tumor distribution described on histopathology. In addition, an added advantage of ultrasound is the ability to assess the remaining abdominal structures. Liver masses were identified in three dogs. Splenic masses were seen in three dogs. In two dogs, splenic nodular hyperplasia was confirmed with surgical biopsies. Postmortem examination in the remaining dog documented gastric adenocarcinoma with splenic and adrenal metastases. Mesenteric lymphadenopathy was detected in three dogs. Lymph-node metastasis was confirmed by necropsy in one dog and by surgical biopsy in another dog. In another study,¹⁸ regional lymphadenopathy was identified on ultrasound examination in 15 of 16 dogs.

Nine dogs in this study were not treated once a gastric mass was identified, because of the perceived poor prognosis. In two other dogs, a side-to-side gastrojejunostomy was performed to bypass the gastric adenocarcinoma. These dogs survived 4 and 5 weeks before the return of clinical signs. Five dogs (four with an adenocarcinoma and one with a leiomyosarcoma) received only a Billroth I. None of these dogs experienced postoperative complications. The recurrence of clinical signs and suspected recurrence of the primary lesion prompted all owners to have their dogs euthanized. Survival times for these dogs were 3 days, 4 weeks (leiomyosarcoma), 6 weeks in two dogs, and 10 months. Of the 140 cases in the literature, information about treatment and outcome exists in only 17 cases.^{5691113–161819} The survival times ranged from 2 hours to 5 years (mean, 8.2±15.7 months; median, 2 months). The reasons for euthanasia in previously reported cases included persistent vomiting and anorexia (n=5), local recurrence and abdominal metastasis observed at a second surgery (n=2), endometritis (n=1), and unknown (n=9). In the literature, there are occasional cases in which surgical resection resulted in long-term survival of 35 months¹¹ and 5 years.¹³ This was not the authors’ experience, and advanced disease at the time of surgery probably contributed to the short survival times in the population of dogs reported here. However, relief of the gastric outflow obstruction resulted in improvement of clinical signs in the immediate postoperative period. In addition, none of the dogs experienced any postoperative complications.

The use of chemotherapy alone or adjuvant chemotherapy following a Billroth I procedure for gastric neoplasia has not been investigated in the veterinary literature. In this study, chemotherapy alone (either a 5-fluorouracil and cyclophosphamide combination) or the FAC protocol followed by cis-platinum resulted in survival times of 7 weeks and 7.5 months, respectively. The first dog was euthanized at 9 weeks because of persistent vomiting. The other dog was euthanized because of acute renal failure 7.5 months after diagnosis. Adjuvant chemotherapy consisting of doxorubicin and cyclophosphamide was given to two dogs (one with adenocarcinoma and one with leiomyosarcoma) following a Billroth I due to the presence of metastasis at the time of surgery. The dog with the metastatic adenocarcinoma involving the pancreas and duodenum was euthanized 9 weeks after diagnosis because of recurrent vomiting. The other dog with gastric and hepatic leiomyosarcoma received two cycles of the previously mentioned chemotherapy and was lost to follow-up 2.5 months after initial evaluation.

Dogs with either a gastric adenocarcinoma or leiomyosarcoma tended to present to the hospital with advanced disease. Fourteen of 19 (74%) dogs in this study with a gastric adenocarcinoma had metastasis. Metastatic sites included regional lymph nodes, omentum, duodenum, liver, pancreas, spleen, esophagus, adrenal glands, and lungs. Unfortunately, in 10 dogs, the presence of metastasis was not known until the time of surgery. In the literature, information about the metastatic potential of gastric adenocarcinoma is available for 132 cases.^{3–68–1618–21} One-hundred (76%) of these cases had metastasis. Reported metastatic sites were widespread and included regional lymph nodes (n=81), peritoneum (n=27), liver (n=19), spleen (n=15), lungs (n=13), adrenal glands (n=6), pancreas (n=5), duodenum (n=4), kidneys (n=4), omentum (n=3), diaphragm (n=2), myocardium (n=2), long bones (n=2), and one each involving the pituitary gland, brain, testicles, and bile duct.

Both dogs with a gastric leiomyosarcoma had metastatic disease at the time of surgery. Affected sites were the liver (n=2) and duodenum (n=1). Kapatkin et al.²² described 13 cases of canine gastric (n=2) and intestinal (n=11) leiomyosarcoma. None of the gastric leiomyosarcomas had metastasis at surgery. Two dogs with jejunal leiomyosarcomas had diffuse abdominal metastases and died of cardiac arrest immediately after surgery or of suspected sepsis the day after surgery. Four other dogs survived <2 weeks. Two were euthanized because of severe peritonitis and adherence of the mass to the mesentery. A third dog died of septic shock during surgery. The final dog died 9 days after surgery because of refractory atrial fibrillation. The remaining seven dogs survived from 7 months to 5.3 years (median, 1.1 years).

There are some inherent limitations of this retrospective study. Postmortem examinations were not available for any of the cases in which treatment was attempted. The recurrence of clinical signs was therefore assumed to be caused by progression of the gastric neoplasia. Given the high metastatic potential of these tumors, every effort should be made to accurately stage suspected gastric neoplasms prior to surgery. This study and literature review document the generally poor prognosis of dogs with primary gastric neoplasia. In light of this information, confirmation of the diagnosis prior to surgery by endoscopic or ultrasound-guided biopsy would provide owners with important information. The accuracy of such gastric biopsy techniques in veterinary patients is not documented and deserves further study.