Clinical Outcomes of 30 Cases (1997–2004) of Canine Gastrointestinal Lymphoma

Introduction

Lymphoma is one of the most common neoplasms affecting dogs.^1,2 It has been reported as being the most common type of gastrointestinal neoplasia in both dogs and cats.^3,4 In dogs, the gastrointestinal form of lymphoma is second in frequency only to the multicentric form.⁵

Despite gastrointestinal lymphoma being the most common extranodal form of lymphoma, the veterinary literature regarding the clinical outcomes of dogs afflicted with this disease is sparse. Much of the literature currently available on lymphoma in dogs is either centered on the multicentric form or does not distinguish between different anatomical forms.^6,7 Therefore, direct interpretation of the information regarding the gastrointestinal form of lymphoma is exceedingly difficult. In fact, only one published case series directly examining clinical cases of canine gastrointestinal lymphoma has been published, and results of chemotherapy were discouraging, with most dogs succumbing to the disease or being euthanized 3 to 14 weeks after diagnosis.⁸

Numerous advances in veterinary oncology have been made in the 18 years since the last study of canine gastrointestinal lymphoma was published. Multi-agent therapy has substantially increased survival times in multicentric canine lymphoma.⁹ Despite these advances, survival times in cases of canine gastrointestinal lymphoma have improved little in recent, small case series.^10,11 The authors report the clinical outcomes and survival times of 30 dogs with gastrointestinal lymphoma.

Materials and Methods

Information for this study was obtained from medical records at the Veterinary Teaching Hospital at Michigan State University for dogs diagnosed with gastrointestinal lymphoma from June 1997 to April 2004. Case selection was limited to dogs with a definitive histological and/or cytological diagnosis of lymphoma. The following information was obtained from the medical records: sex, breed, age at diagnosis, presenting complaints, physical examination findings, clinical laboratory findings, how the diagnosis was made, location of lymphoma in the gastrointestinal tract, treatments, and survival time (i.e., date of diagnosis to date of death). For all histological samples that were available (n=19), immunophenotyping was performed by a veterinary pathologist to determine cell line of origin.

Owners and referring veterinarians were contacted to obtain follow-up information, including whether the dogs were currently alive or deceased, type of treatment received before death, date of death, and how the dogs died (i.e., euthanasia or natural death).

Results

Of the 30 dogs identified for inclusion in this study, 15 were neutered males, four were intact males, 10 were spayed females, and one was an intact female. Eighteen breeds were represented, with the golden retriever (n=4) and mixed-breed dogs (n=3) being most common. The median age at presentation was 8.2 years (range 3.2 to 13.2 years). The most common presenting complaints included vomiting (n=18), diarrhea (n=14), weight loss (n=14), anorexia (n=14), and lethargy (n=10). The most common abnormalities identified on physical examination included thin body condition (n=7), pale mucous membranes (n=7), lethargy (n=6), increased lymph node size (n=6), tacky mucous membranes (n=5), and a tense abdomen on palpation (n=5) [Table 1].

A complete blood count (n=30) and serum biochemical profile (n=29) were performed at admission. Urinalyses were also performed in most cases (n=25). The most common abnormalities on the complete blood count included neutrophilia (n=19), monocytosis (n=22), leukocytosis (n=13), and anemia (n=11). The most common abnormalities on the serum biochemical panel included hypoalbuminemia (n=24), hypoproteinemia (n=20), and hypocalcemia (n=20) [Table 2]. In 19 of the 20 dogs with hypocalcemia, the calcium concentrations were actually within the reference range (9.0 to 11.9 mg/dL) when adjusted for concurrent hypoalbuminemia. In the remaining dog with hypocalcemia, the calcium concentration was actually mildly increased (12.8 mg/dL) after adjustment for hypoalbuminemia.

Diagnoses were made in 22 dogs based on histopathology and in eight dogs based on cytology. The lymphoma was localized within the gastrointestinal tract to the stomach only (n=5), small intestine only (n=13), large intestine only (n=5), or two or more sites (n=7). Tissue for histopathology was obtained either via exploratory laparotomy with full-thickness biopsies (n=4) or mass resections (n=8) or via endoscopy (n=10). For 19 of the 22 cases that had been diagnosed via histopathology, tissue was available for immunophenotyping to determine if the lymphoma was of B-cell or T-cell origin. Results revealed a T-cell origin in 12 cases and a B-cell origin in the other seven cases.

Treatments consisted of surgery combined with chemotherapy (n=8), surgery alone (n=4), chemotherapy alone (n=15), or none (n=3). In those dogs that had surgery combined with chemotherapy, either the masses were resected (n=4) or biopsies were taken (n=4). Chemotherapeutic medications, administered either as a single agent or as part of a multi-agent therapeutic protocol, included prednisone (n=23), L-asparaginase^a (n=7), doxorubicin^b (n=4), vincristine^c (n=9), lomustine^d (n=1), and cyclophosphamide^e (n=8) [Table 1].

Of the 30 cases, four dogs died during initial presentation, 24 were eventually euthanized at the owner’s request, and two are currently alive. In both of these dogs, the gastrointestinal lymphoma was large intestinal in origin.

Overall median survival time for all 30 dogs was 13 days. The median survival times for dogs with lymphoma localized to the stomach, small intestine, large intestine, and two or more sites were 17 days, 13 days, 61 days, and 4.5 days, respectively. There was no significant difference in median survival time between dogs with lymphoma localized to the stomach or small intestine (P=0.7) or to two or more sites (P=0.2). Using the stomach as the reference site, the hazard rate ratios (HRR) and 95% confidence intervals (95% CI) for death caused by gastrointestinal lymphoma were 1.26 for the small intestine (95% CI=0.44 to 3.61), 0.28 for the large intestine (95% CI=0.050 to 1.51), and 1.3 for two or more sites (95% CI=0.38 to 4.41). Although no significant difference was seen between the large intestine and other sites (P=0.13), this may be more a function of the small study size, given the large disparity in overall median survival times [Figure 1].

Overall median survival time was not significantly associated with sex (P=0.53), age at diagnosis (P=0.71), low serum albumin level at presentation (P=0.27), low total protein level at presentation (P=0.89), surgery (P=0.58), or cell type of origin (P=0.21). Using univariate analysis, median survival time was significantly associated with treatment with prednisone (HRR=0.19, 95% CI=0.062 to 0.57, P=0.003, n=23), vincristine (HRR=0.39, 95% CI=0.16 to 0.94, P=0.036), and cyclophosphamide (HRR=0.32, 95% CI=0.12 to 0.81, P=0.017). Controlling for prednisone therapy, multivariate analysis showed no significant effect on survival time for treatment with vincristine (HRR=0.50, 95% CI=0.19 to 1.28, P=0.15, n=9) or cyclophosphamide (HRR=0.39, 95% CI=0.15 to 1.05, P=0.063, n=8). Survival time was not significantly associated with treatment with L-asparaginase (P=0.42, n=7), Adriamycin (P=0.46, n=4), or lomustine (P=0.28, n=1). The HRR of multimodal therapy (i.e., prednisone with other chemotherapy) versus prednisone alone was 0.40 (95% CI=0.15 to 1.09). Survival with prednisone versus multimodal therapy was not significantly different (P=0.074).

Discussion

Currently, no clear consensus exists regarding sex predilection for lymphoma. In some studies, the percentage of males affected with gastrointestinal lymphoma was higher than that of females, while no difference was found in other studies. ^6,12 In the current study, no significant difference in the sexes was identified. This plausibly may be due to low statistical power because of the low case number, or perhaps there truly is no sex predilection with this disease.

The clinical presentations of dogs in this study were similar to those of dogs in previous reports, ranging from non-specific signs such as anorexia and lethargy to vomiting and hemorrhagic diarrhea.^8,10,11

Many of the dogs in this study also had hypoalbuminemia (n=23). In a study by Price et al., hypoalbuminemic dogs being treated for multicentric lymphoma had 50% shorter remission times than dogs that were presented with normal albumin levels.¹³ Albumin plays a role in many functions of the body, including drug binding and transport, as well as maintenance of vascular oncotic pressure. ^14,15 Albumin also has a substantial role as a free radical scavenger, which functions in the pathogenesis of many neoplastic diseases.¹⁶ Although a trend was seen between increased albumin level at presentation and survival, no significant difference was identified. The effect of low albumin on survival time for gastrointestinal lymphoma remains unclear.

Many of the dogs (n=20) in this study had low serum calcium concentrations at presentation, but because calcium concentrations were within the normal range after adjustment for hypoalbuminemia in all but one of these dogs, this finding was of little clinical significance.

Views differ regarding the accuracy of cytological diagnosis of neoplasia. In one study done by Bonfanti et al., results of fine-needle biopsy of deep abdominal organs were consistent with the diagnosis of inflammatory versus neoplastic disease in 89% of cases.¹⁷ However, in another study by the same authors, cytological examination of gastrointestinal tumors was consistent with histological findings in only 72% of cases.¹⁸ Although histology is preferred for accurate cell typing and grading, cytological examination has been identified as an accurate means to diagnose canine lymphoma.⁹

In this study, the role of surgery was primarily to confirm the diagnosis or to treat spontaneous perforation, rather than to serve as a cornerstone of treatment. However, these treatment findings must be interpreted with caution, because treatment decisions were not randomized in this group of dogs. This lack of randomization could potentially bias survival times among treatment groups if a certain treatment (e.g., surgical intervention) was selected more often for dogs with more severe clinical disease or with a more advanced-staged disease than dogs in other treatment groups.

Numerous protocols are currently used in treatment of multicentric lymphoma with reasonable success.^6,13,19 Reported median survival times range from 6 to 12 months, with complete remission achieved in 60% to 90% of cases.⁹ Canine gastrointestinal lymphoma has been treated empirically with many of these same protocols.^10,11 Couto and colleagues documented a poor clinical response to chemotherapy in dogs with gastrointestinal lymphoma, reporting that all but one dog had died or was euthanized within 14 weeks after diagnosis.⁸ This has been attributed to the use of single-agent therapy and the lack of good multi-agent protocols at the time. In the current study, hazard analysis showed that dogs treated with multi-agent therapy had a death rate of 0.40. Although close, this rate was not significantly different from that of dogs treated with prednisone alone (P=0.074).

In this study on canine gastrointestinal lymphoma, the median survival time was very short, and the responses to various therapeutic regimens were poor. Several potential explanations exist for this. Commonly, most cases of canine gastrointestinal lymphoma have been assumed to be of B-cell origin, likely based on the fact that almost all human, non-Hodgkin’s lymphomas are of B-cell origin.^9,20 However, it is now known that canine gastrointestinal lymphoma is more commonly of T-cell origin.^10,11,21 In this current study, for the cases in which immunophenotyping was performed, T-cell lymphoma was diagnosed in 63% of cases and B-cell lymphoma was diagnosed in 37% of cases. Because T-cell lymphomas have a more aggressive biological behavior and are less responsive to chemotherapeutic intervention, dogs with gastrointestinal lymphoma may be less responsive to chemotherapy.²² In the current study, subtype showed no influence on survival time. Additional studies investigating a potential link between cytological subtype and prognosis are warranted.

Interestingly, the two dogs with colorectal lymphoma in this study are currently alive (current survival times are 31 and 84 months). Couto et al. (1989) had one similar case of colorectal lymphoma that survived long term.⁸ One of the two dogs in the present series was treated with surgical removal of the rectal mass followed by a multi-agent chemotherapeutic protocol for lymphoma. Even more interesting, the other surviving dog was treated only with prednisone. This dog was originally presented with hematochezia and a rectal mass of 2 months’ duration. An incisional biopsy of the mass was taken for diagnosis, and the dog was placed on prednisone therapy. Although short-term remissions of 1 to 2 months are possible, most dogs treated only with prednisone succumb to the disease within a couple of months.²³ Unfortunately, no tissue was available from this case for immunohistochemistry. Despite these two long-term survivors, their influence on the survival rate with large-intestinal lymphoma was not significant; however, a trend toward this may be seen in the data (P=0.13). Low case numbers make valid statistical interpretation of this variable difficult.

Conclusion

Canine gastrointestinal lymphoma is a serious disease with a grave prognosis.Although chemotherapy reduces the mortality rate, only minor differences were seen between currently used treatment modalities with regard to survival time. The data within this study suggest that cases of colorectal origin may have potentially better survival outcomes than cases of lymphoma located elsewhere in the gastrointestinal tract. Large, multi-institutional studies are warranted to ascertain whether lymphoma of large-intestinal origin may be a favorable prognostic factor for survival.