Reclassification of small intestinal and cecal smooth muscle tumors in 72 dogs: clinical, histologic, and immunohistochemical evaluation

Objectives— To reclassify canine small intestinal and cecal leiomyoma (LM) and leiomyosarcoma (LMS) into smooth muscle and gastrointestinal stromal tumors (GIST) using histologic and immunohistochemical (IH) analysis and to report clinical findings and survival data.
Study Design— Retrospective review of cases.
Animals— Dogs (n=47) with small intestinal (40 LMS; 7 LM) and 25 dogs with cecal tumors (23 LMS; 2 LM).
Methods— Clinical and survival data were reviewed. Tissue sections were reevaluated for light-microscopic malignancy criteria and examined for expression of SMA, desmin, vimentin, S-100, and CD117 (KIT) by immunohistochemistry.
Results— Reclassification resulted in 2 LM, 9 LMS, 19 GIST, and 17 GIST-like tumors in the small intestine and 23 GIST and 2 GIST-like tumors in the cecum. GIST-like tumors were morphologic and IH identical to GIST but lacked KIT expression. No significant difference in survival was observed for tumor type, location, histologic, or IH characteristics; however, dogs with cecal tumors were significantly older in age, presented more commonly with intestinal perforation and peritonitis, and less commonly with weight loss. Cecal tumors had more histologic malignancy criteria than small intestinal tumors. After excision, 1 and 2 year recurrence-free periods were 80.1% and 67.2% for small intestinal and 83.3% and 61.9% for cecal tumors.
Conclusion— Prognosis for intestinal tumors with histologic smooth muscle appearance is good after excision and not related to tumor type, location, histologic, or IH characteristics.
Clinical Relevance— Clinical importance could not be demonstrated for reclassification, but may be for future treatment, of intestinal smooth muscle or stromal tumors.     

Clinical and immunohistochemical differentiation of gastrointestinal stromal tumors from leiomyosarcomas in dogs: 42 cases (1990-2003)

OBJECTIVE: To reexamine (via immunohistochemical techniques) canine tissue samples that had been previously classified as gastrointestinal leiomyosarcomas (GILMSs), identify and differentiate gastrointestinal stromal tumors (GISTs) from GILMSs, and compare the biological behavior and clinical course of GISTs and GILMSs in dogs. DESIGN: Retrospective case series. ANIMALS: 42 dogs. PROCEDURES: Medical records of 42 dogs for which a histologic diagnosis of GILMS was confirmed were reviewed for signalment, clinical signs, physical examination findings, results of initial diagnostic tests, surgical findings, adjunctive treatment, location of the tumor, completeness of resection, and outcome after surgery. Archived tumor tissue specimens from each dog were restained via immunohistochemical techniques to differentiate tumor types. Long-term follow-up information was obtained from the medical record or through telephone interviews with owners and referring veterinarians. RESULTS: On the basis of immunohistochemical findings, 28 of 42 tumors were reclassified as GISTs and 4 were reclassified as undifferentiated sarcomas; 10 tumors were GILMSs. In dogs, GISTs developed more frequently in the cecum and large intestine and GILMSs developed more frequently in the stomach and small intestine. Median survival times for dogs with GISTs and GILMSs were 11.6 and 7.8 months, respectively; if only dogs surviving the perioperative period were considered, median survival times were 37.4 and 7.8 months, respectively. These differences, however, were not significant. CONCLUSIONS AND CLINICAL RELEVANCE: In dogs, many previously diagnosed GILMSs should be reclassified as GISTs on the basis of results of immunohistochemical staining. The biological behavior of these tumors appears to be different.     

Association of p53 Gene Mutation with the Prognosis of Canine Lymphoid Tumors

Introduction:  Many dogs with lymphoid tumors develop resistance to chemotherapy. As a mechanism of drug resistance in canine lymphoma, ATP‐dependent drug efflux by P‐glycoprotein was reported, however, inhibition of apoptosis mediated by P53 inactivation has not been investigated. In this study, we investigated the relationship between p53 gene mutation and clinical drug resistance in canine lymphoid tumors.
Methods:  Tumor specimens were obtained from 44 dogs with lymphoid tumors. Mutations of p53 gene at exon 4–8 of these tumor tissues were examined by PCR‐SSCP (single strand conformational polymorphism) analysis, followed by nucleotide sequencing of the abnormal bands. The cases were treated with UW‐Madison protocol, and its response was evaluated by the tumor size or the number of peripheral leukemic cells.
Results:  Of the 44 dogs, 15 dogs (34%) had p53 mutation, whereas 29 dogs (66%) were devoid of p53 mutation, before or during the chemotherapeutic protocol. Rate of good response (CR and PR) to chemotherapy was significantly lower in the dogs with p53 mutation (20%) than those without p53 mutation (55%)(p = 0.022). Median overall survival duration after examination of p53 mutation was significantly shorter in dogs with p53 mutation (101days) than those without p53 mutation (223days)(p = 0.008).
Conclusions:  Lymphoid tumors with p53 mutations were shown to have worse prognosis than those without p53 mutation.     

Serum and Plasma Vascular Endothelial Growth Factor Levels in Various Cat Populations

Introduction:  Vascular endothelial growth factor (VEGF) is one of the most important mediators of angiogenesis. Elevated levels within tumors, and in serum, plasma, and tumor effusion have been correlated with the development of metastatic disease, recurrence, and poor prognosis in many tumors in humans. Canine VEGF has been sequenced as homologous with the human form, and elevated serum and plasma VEGF have been found in dogs with hemangiosarcoma. Feline VEGF has also been sequenced, and shares homology with the human and canine forms.
Materials and Methods:  Stored serum and plasma samples from normal cats, cats with various neoplasms, and cats with non‐neoplastic disease were evaluated with a commercial ELISA kit (R&D Systems, Minneapolis MN). Samples were run in duplicate, and a standard curve was performed for each plate. The data were analyzed for differences between populations, and between serum and plasma levels in the same patient to determine the optimal sample for evaluating VEGF in cats.
Results:  In seven apparently healthy cats mean plasma VEGF was 95.6 pg/mL. In non‐neoplastic disease (7 cases), mean plasma VEGF was 117.3 pg/mL and mean serum VEGF level was 219.7 pg/mL. In ten tumor‐bearing patients mean plasma VEGF was 247.1 pg/mL, and mean serum VEGF was 322.3 pg/mL. Statistical analysis showed no significant difference between mean serum and plasma VEGF concentrations within each group or between groups (p > 0.05).
Discussion:  Serum and plasma VEGF levels could not be used to distinguish between healthy cats and cats with neoplastic or non‐neoplastic disease.     

Characterization, Expression and Function of c‐MET in Canine Spontaneous Cancers

Introduction:  Aberrant expression of the proto‐oncogene c‐Met has been noted in a variety of human cancers. In dogs, inappropriate Met expression has been identified in canine osteosarcoma (OSA) tumor samples. To better define the potential role of Met dysregulation in canine cancer, we cloned canine Met, HGF, and HGF activator and evaluated their expression patterns in a variety of canine tumor cell lines.
Methods:  Canine Met, HGF, and HGF activator were cloned from normal canine liver and canine OSA cell lines using primers based on regions of homology between mouse and human sequences as well as 5' and 3' RACE.
Results:  Inappropriate expression of Met was found in canine cell lines derived from OSAs, mast cell tumors, histiocytic sarcomas, hemangiosarcoma, and melanomas. Both HGF and HGF activator were found to be expressed in several of these tumor cell lines, providing evidence of a possible autocrine loop of Met stimulation. Incubation of canine tumor cell lines with rhHGF resulted in Met autophosphorylation and activation of the downstream signaling elements Gab1, Akt and Erk1/2. Scattering of tumor cells in response to HGF occurred under conditions of cell stress, such as serum starvation. Lastly, the Met inhibitor PHA‐665752 blocked HGF induced phosphorylation of canine Met and Gab1.
Conclusions:  These studies provide evidence that similar to the case in human tumors, aberrant Met expression may play an important role in the biology of canine cancer. As such, inhibition of Met function may represent a potentially useful novel therapeutic approach.     

Expression of E‐cadherin in Canine Anal Sac Apocrine Gland Adenocarcinoma and its Association with Survival

Introduction:  Inverse associations have been shown between E‐cadherin expression and degree of differentiation in canine mammary tumours and with the acquisition of malignancy in human epithelial cancers. The purpose of this study was to examine whether there was an association between prognosis, and the distribution or intensity of immunohistochemical staining for E‐cadherin in anal sac adenocarcinoma.
Methods:  Formalin‐fixed paraffin‐wax embedded canine anal sac adenocarcinoma specimens were obtained for 36 patients for whom clinical data, including survival statistics, were available. Canine mammary adenoma tissue was employed as the positive and negative controls with antibody diluent replacing the primary antibody in the latter. Samples were scored according to proportion of cells showing positive cytoplasmic membranous staining, intensity of cytoplasmic membranous staining and proportion of nuclei staining. As quality control, a sample of 10 slides was randomly selected for repeat evaluation; results exactly matched those obtained in the first assessment.
Results:  All anal sac adenocarcinoma cells expressed E‐cadherin to varying degrees. There was no evidence of an association between survival time and the intensity of staining or the proportion of nuclei staining. Patients whose tumours exhibited >75% positive cytoplasmic membranous staining survived significantly longer than those exhibiting <75% staining with median survival times of 1100 and 479 days respectively (log rank test: χ² = 3.89, 1 DF, p = 0.049).
Conclusion:  Immunohistochemical evaluation of the proportion of cells in anal sac adenocarcinoma samples exhibiting positive cytoplasmic membranous staining for E‐cadherin may allow improved determination of prognoses and could aid clinical decision making.     

Efficacy and Toxicity of Intracavitary Administration of Pegylated Liposomal Encapsulated Doxorubicin (DOXIL) in Dogs with Hemangiosarcoma

Introduction:  Canine hemangiosarcoma (HSA) is a fatal malignancy and most dogs die within 6–8 months of diagnosis. The spleen is a common primary site, representing 50% of all cases. These dogs typically present with clinical signs due to tumor rupture and intra‐abdominal dissemination; the abdomen is also the main site of disease progression when these patients fail. Direct delivery of chemotherapy into the abdominal cavity may therefore be a rational approach in this malignancy.
Methods:  14 dogs with stage 2 or 3 splenic HSA were recruited. Doxil at a dose of 1 mg/kg was diluted in saline and administered via ultrasound‐guidance into the abdominal cavity. The dogs were scheduled to receive 4 treatments every 3 weeks. Samples of plasma and abdominal fluid were collected for pharmacokinetic analysis. All dogs were monitored for recurrence and complete necropsies were requested at death.
Results:  8 dogs with stage 3 and 6 dogs with stage 2 HSA were enrolled. All 14 dogs have died, 12/14 due to tumor and 2 from other causes. There was no difference in median survival days between stages (stage 2: 244, stage 3: 125, p = .22). All 12 dogs that died due to tumor‐related causes failed with intra‐abdominal recurrence. Necropsies showed that the dogs in this study had relatively fewer extra‐abdominal metastasis compared to dogs treated with systemic chemotherapy. Pk analysis showed detectable plasma doxorubicin 1 and 2 weeks after treatment.
Conclusion:  Direct abdominal administration of Doxil did not prevent intra‐abdominal recurrence; however, it appeared to provide effective systemic coverage.     

Cyclooxygenase ‐2 Expression in Mammary Tumors in Dogs and its Correlation to Histologic and Biologic Behavior

Introduction:  Cyclooxygenase‐2 (Cox‐2) is the inducible form and the rate‐limiting enzyme, for conversion of arachidonic acid to prostaglandins. Cox‐2 overexpression, common in carcinomas, is associated with increased growth rate, resistance to apoptosis, angiogenesis, and overall, both local and distant aggressive behavior. Cox‐2 overexpression has been detected in human and canine mammary tumors (MTs). Histopathology of canine MT is not always predictive of biologic behavior, and anecdotally, only 50% of the malignant MTs are expected to metastasize. We hypothesize that Cox‐2 expression correlates with aggressive behavior.
Methods:  This retrospective study evaluated 48 bitches, presented for excision of MT between 2000 and 2003 at FMVZ de Botucatu‐UNESP, Brasil. Follow‐up varied from 18 months to 24 months and included physical examination and thoracic radiographs. Histopathologic examination was performed in all tumors, as well as in metastatic lesions when detected in the follow‐up period. Immunohistochemistry was used to detect expression of Cox‐2 in paraffin blocks (Rabbit polyclonal anti‐PGHS‐2. Oxford Biomedical). 10 adenomas, 10 carcinomas, 10 benign mixed tumors, 10 malignant mixed tumors and 8 cases of primary carcinomas and their metastatic lesions.
Results:  Expression of Cox‐2 varied among groups. Adenomas (32.1%), mixed benign tumors (38%), carcinomas (60.3%), malignant mixed tumors (65.8%), and metastatic carcinomas (81.25%) and their metastatic lesions (84.35%). Statistically significant differences (p < 0.05) were observed between the benign and malignant counterparts and between carcinomas and metastatic carcinomas.
Conclusions:  Cox‐2 expression correlates with both histologic and biologic behavior in mammary carcinomas, and may serve as a predictor of metastatic potential.     

The Biologic Behavior of Mast Cell Tumors in Pug Dogs

Introduction:  The purpose of this study was to quantitate the risk and to describe the behavior of mast cell tumors (MCT) in Pugs.
Methods and Materials:  Using the Veterinary Medicine Database, the frequency of MCT in Pugs was compared to the frequency in other dogs using a Chi‐square test. To describe the biologic behavior of MCT in pugs, cases with histologically confirmed diagnosis were identified through the University of Minnesota (UMN) Diagnostic Laboratory and Veterinary Medical Center. Histology was reviewed by a single pathologist. Survival analysis was performed to determine the impact of clinical and histologic data on survival.
Results:  The frequency of MCT diagnosis in Pugs was significantly increased compared to other dogs (OR = 2.28, 95% CI = 1.81–2.86). Twenty‐five purebred Pugs with a diagnosis of MCT were identified through UMN. Multiple cutaneous tumors were documented in 14 (56%) of the dogs. Most tumors were low to intermediate grade. Only three dogs have died of their disease. Sixteen are still living (median follow‐up = 660 days). The only factors predicting survival were grade, mitotic index and tumor size.
Discussion:  Our data confirms MCT predisposition in Pugs and suggests that mast cell tumors in Pugs are relatively benign, despite the presence of multiple cutaneous tumors in most cases. Multiple tumors in breeds with predisposition to MCT may indicate separate primaries rather than advanced stage disease.
This work was supported in part by NCI grants R03‐CA101030 and K08‐CA89530.     

Characterization of the Biological Behavior of Appendicular Osteosarcoma in Rottweilers with a Comparison to Other Breeds

Introduction:  The clinical impression of many oncologists is that Rottweilers with appendicular osteosarcoma (OSA) have an unfavorable prognosis despite standard therapy (amputation and chemotherapy). The purpose of this study was to compare clinical characteristics, biological behavior and outcome of OSA in Rottweilers to that of other breeds.
Methods:  107 OSA cases were studied: 56 clinical cases (17 Rottweilers) and 54 necropsy submissions (13 Rottweilers) of OSA (3 cases overlapping). Information regarding signalment, prognostic factors, treatment and outcome was obtained from medical records and follow‐up with owners. Data were analyzed by chi‐square test, t‐test, and log rank test.
Preliminary Results:  Compared to other breeds Rottweilers were more likely to have femur (23% vs. 16%) and ulna (15% vs. 3%) and less likely to have tibia (8% vs. 27%) as their primary tumor location (p = 0.04). In the necropsy group, Rottweilers had a younger age at death (6.4 vs. 10 years, p = 0.0002) and a greater body weight at diagnosis (49 vs. 37 kg, p = 0.002). No statistical differences in the age at diagnosis or survival time were noted in Rottweilers vs. other breeds. Metastasis to lymph nodes (31% vs. 10%, p = 0.08) and lungs (64% vs. 49%, p = 0.24) was more frequent in Rottweilers. There was no difference in chemotherapy drugs given; however, Rottweilers were 20 times more likely to receive NSAIDs (p = 0.0004).
Conclusion:  Greater body weight with a predilection for certain tumor locations may contribute to OSA development in Rottweilers. A younger age at death may suggest a more aggressive phenotype.     

131‐I therapy for Advanced, Unresectable Thyroid Tumors in Dogs

Introduction:  Greater than 50% of dogs with thyroid tumors present with surgically unresectable disease for which external beam radiotherapy has been reported to prolong survival. The success of ¹³¹I for control of thyroid tumors in cats and in humans suggests such therapy may also play a role in the management of canine thyroid cancer.
Methods:  Thirty‐nine dogs with WHO stage II/III (invasive or ectopic; n = 32) or IV (metastatic; n = 7) thyroid tumors were treated with ¹³¹I alone. Changes in thyroid function, ^99MTc‐pertechnetate (^99MTc) scintigraphic changes, and tumor response were recorded. Dogs with ventral cervical tumors were evaluated for feasibility of surgical resection following ¹³¹I.
Results:  Median overall survival was 839 days and 366 days for dogs with stage II/III and stage IV tumors, respectively. Thyroid hormone status, site and surgical resection were not associated with outcome in dogs with stage II/III tumors. Three dogs developed severe bone marrow suppression.
Conclusions:  These findings suggest ¹³¹I should be investigated more thoroughly in dogs with thyroid tumors not considered surgical candidates to more clearly characterize the indications for therapy and followup recommendations. ¹³¹I dosimetry in dogs with thyroid tumors remains problematic. Administration of ¹³¹I is currently based on empiric recommendations and, in general, the treatment is well tolerated although additional studies are indicated to optimize response and minimize toxicity.     

An hGM‐CSF DNA‐Cationic Lipid Complexed Autologous Tumor Cell Vaccine did not Improve Remission Duration in Dogs with Lymphoma

Introduction:  Lymphoma is one of the most common cancers in dogs and while clinical remission can be induced using chemotherapy, very few dogs are cured. Since cytokine‐adjuvanted autologous whole‐tumor‐cell vaccines (ATCV) can induce potent antitumor immune responses against otherwise non‐immunogenic cancers we initiated a study of such an approach in dogs with lymphoma.
Methods:  Following achievement of a complete remission using a 19‐week CHOP‐based chemotherapy protocol, 51 dogs with B‐cell lymphoma were randomized to receive 8 treatments (4 weekly, then 4 additional at q2wk intervals) of vaccine or lipid‐equivalent placebo. Dogs were followed monthly for assessment of remission duration and survival. Surrogate indices of immune response (delayed‐type hypersensitivity, interferon‐γ quantitative RT‐PCR, lymphocyte proliferation, and flow cytometry for lymphoma‐specific antibodies) were also investigated before and after vaccination.
Results:  No significant difference in median remission duration was observed between dogs receiving vaccine (277 days) or placebo (258 days); the Kaplan‐Meier curves were virtually super‐imposable. No significant differences in surrogate indices of immune response were noted pre‐ and post‐vaccination.
Conclusions:  In this context, an hGM‐CSF DNA‐cationic lipid complexed ATCV vaccine did not enhance remission duration in dogs with lymphoma, likely due to lack of vaccine‐induced tumor‐specific immunity.     

The effect of deracoxib and piroxicam on viability of canine osteosarcoma cells in vitro

Introduction:  Cyclooxygenase‐2 (COX‐2) inhibitors are being used increasingly in cancer therapy. Although the effects of COX‐2 inhibitors have been evaluated extensively in carcinomas, less is known about their effects in sarcomas. Since the majority of dogs with appendicular osteosarcoma (OSA) are treated for pain with a non‐steroidal anti‐inflammatory drug (some COX‐2 selective) prior to definitive treatment, it is important to determine the effects that commonly used NSAIDS have on tumor cell growth.
Methods:  Established canine osteosarcoma (POS, HMPOS and COS31) and canine fibroblast cell lines were maintained in culture under standard conditions. Cells were incubated with either deracoxib (1 uM to 500 uM) or piroxicam (1 uM to 1000 uM). Cell viability was assessed at 72 hours by cell counts and the MTT assay. The DNA fragmentation analysis was utilized to assess for apoptosis induction.
Results:  Deracoxib concentrations ≥100 uM and piroxicam concentrations ≥500 uM significantly reduced mean cell viability of all three OSA cell lines (lowest cell viability percentages 20% and 32%, respectively). Deracoxib concentrations ≥250 uM and piroxicam concentrations ≥500 uM also reduced viability of fibroblasts; however, the cell viability percent was reduced to only 54% and 68%, respectively, of the control value. Exposure of OSA cells to cytotoxic concentrations of deracoxib and piroxicam did not result in DNA fragmentation.
Conclusions:  Deracoxib and piroxicam demonstrated a cytotoxic effect on canine osteosarcoma cells. There was no evidence of apoptosis induction at the concentrations evaluated. Further investigation will need to be performed to determine whether either drug exhibits anti‐tumor effects in vivo .     

Retropective Evaluation of Survival Time of Dogs with Stage III Osteosarcoma that Undergo Treatment

Introduction:  Stage III osteosarcoma generally carries a grave prognosis. Despite this, some owners elect to treat using palliative or curative intent protocols. The purpose of this study was to determine the survival times for dogs with stage III osteosarcoma that undergo treatment and to evaluate the variables that affect survival time.
Methods:  Retrospective study using the CSU Animal Cancer Center osteosarcoma data base. Search criteria included dogs diagnosed with osteosarcoma 1985–2004 with metastasis at the time of presentation. Dogs were excluded if they were euthanized at the time of diagnosis (13 dogs) or lost to follow‐up (10 dogs). There were 90 cases for analysis. The survival times were compared based on the primary tumor site, site of metastasis, treatment given, age, sex and breed.
Results:  Survival times in days ranged from 0 to 1583 days. The overall median survival time was 76 days. The one‐year, two‐year and three‐year percent survival were 7%, 4.7% and 3.5%, respectively. Treatment included various combinations of chemotherapy, surgery, radiation therapy, bisphosphonates and NSAIDs. Findings included an increased survival time with surgery and adjuvant therapy compared with surgery alone and a decreased survival time with metastasis to the lung or lymph node.
Conclusions:  Treatment of dogs with stage III osteosarcoma can result in variable survival times. Multimodality therapy appears to result in longer survival times. Metastasis to the lung or lymph node correlated with a decreased survival time.     

Results of the Combined Treatment with Piroxicam and Carboplatin in Canine Oral Non‐Tonsillar Squamous Cell Carcinoma

Introduction:  Over‐expression of COX‐2 has been observed in several human and animal malignancies and is implicated in carcinogenesis through the conversion of arachidonic acid to PGE‐2. Use of platinum‐containing cytostatic agents and/or (non‐)specific COX‐2 inhibitors, has been reported as a treatment option for canine oral non‐tonsillar squamous cell carcinomas (ONT‐SCC). However, no study describes the effect of a combination of carboplatin and piroxicam on this tumor type.
Methods:  7 dogs with a T3 (WHO‐TNM) ONT‐SCC were treated with piroxicam and carboplatin. Five had bone involvement and no detectable metastasis. Two dogs without bone involvement had metastasis in the regional lymph nodes. Piroxicam was given orally 0.3 mg/kg s.i.d. Each dog was scheduled to receive between 6 and 12 carboplatin infusions (300 mg/m² i.v.) at 3 week intervals. Ondansetron and metoclopramide were used as anti‐emetic agents. The dogs are planned to receive piroxicam on a lifelong basis.
Results:  Complete response (CR) without adjuvant surgery was achieved in 4 of the 7 dogs. Two dogs needed adjuvant surgery to achieve CR. One dog had progressive disease and was euthanised 231 days after start of therapy. All the others were still alive and in CR at date of analysis. Median follow‐up was 335 days (107–689 days).
Conclusions:  Our study suggests that a combination of piroxicam and carboplatin is a useful treatment option for canine ONT‐SCC. All dogs tolerated therapy well and the 57% response rate for reaching a complete and durable remission without adjuvant surgery is promising.     

Efficacy of C/B‐OPP rescue for the treatment of relapsed canine lymphoma (1998–2004)

Introduction:  MOPP chemotherapy is useful for relapsed canine lymphoma. This study evaluates the efficacy of this protocol after substitution of CCNU (lomustine) or BiCNU (carmustine) for mechlorethamine (C/B‐OPP).
Methods:  Patient signalment, response to chemotherapy, toxicity and survival data were abstracted from medical records of dogs from receiving C/B‐OPP between 1998 and 2004.
Results:  Fifty‐eight dogs received C/B‐OPP rescue chemotherapy during the study period. The median remission duration after initial chemotherapy, consisting of CHOP‐based therapy in 91% of dogs, was 133 days (range, 10 to 932 days). Thirty‐eight of fifty‐eight dogs (66%) responded to C/B‐OPP rescue after relapse (22 CR, 16 PR), for a median of 48 days (range, 2 to 359 days). Overall, C/B‐OPP extended survival by a median of 90 days (range, 2 to 426 days). Twenty‐four dogs (41%) experienced one or more episodes of Grade II or higher gastrointestinal toxicity. Forty‐one dogs (71%) experienced one or more episodes of Grade II or higher hematologic toxicity. Twelve dogs (20%) developed regenerative anemia with diarrhea consistent with gastrointestinal hemorrhage. Treatment delays due to hematologic toxicity occurred in 37 dogs (63%). There were 16 nonfatal treatment‐related episodes of sepsis requiring hospitalization. 5 dogs died due to sepsis and/or chemotherapy‐related complications.
Conclusions:  C/B‐OPP chemotherapy has activity against relapsed canine lymphoma which is similar to that of traditional MOPP rescue therapy. Moderate to severe hematologic toxicity was observed. Further work is warranted to optimize drug doses and scheduling.     

Increase in Regulatory T Cells in Dogs with Cancer Undergoing Chemotherapy

Introduction:  Regulatory T cells (Treg) are a naturally occurring population of T cells phenotypically identified by co‐expression of CD4 and the IL‐2 receptor (CD25). Theyplay a critical role in the control of tolerance and autoimmunity and have also been implicated in impairment of anti‐tumor responses. We hypothesized that levels of Treg would be higher in cancer‐bearing dogs than in normal dogs and that they would decrease with chemotherapy.
Methods:  Serial PBMC were isolated from twenty cancer‐bearing dogs receiving either single‐agent doxorubicin or the Madison‐Wisconsin protocol. The following time points were studied: pre‐treatment, day 2, week 1, week 3, 3 months and 6 months after initial treatment. Ten age‐matched, normal dogs were also studied. PBMC were immunostained with directly conjugated antibodies to CD4, CD8, CD44 and IL‐2 receptor and then evaluated by flow cytometry.
Results:  Low numbers of lymphocytes with the CD4+/IL‐2R+ phenotype were detectable in both normal and cancer‐bearing dogs. A statistically significant increase in the percentage of IL2‐R+/CD4+ T cells was observed in the cancer‐bearing dogs beginning two days after chemotherapy and persisting throughout treatment. The percentage of IL‐2R+/CD4+ T cells was also increased in pre‐treated cancer‐bearing dogs compared to control dogs.
Conclusion:  The percentage of IL‐2R+/CD4+ T cells was generally higher in dogs with cancer than in healthy dogs. Unexpectedly, the percentage of IL‐2R+/CD4+ cells increased during chemotherapy which suggests that chemotherapy may exert immunosuppressive effects through a previously undescribed mechanism. The identity of these CD4+/IL‐2R+ T cells as true Treg awaits additional characterization studies.     

A Candidate Second‐Generation Photosensitizer for Photodynamic Therapy in Veterinary Medicine

Introduction:  Photodynamic therapy (PDT) involves the light activation of a drug within a tumor causing selective tumor cell death. Unfortunately, some photosensitizing drugs have been associated with adverse reactions in veterinary patients. Zinc phthalocyanine tetrasulfonate (ZnPcS4) is a promising second‐generation photosensitizer for use in veterinary medicine, however, it cannot be applied clinically until safety and efficacy data are available.
Methods:  Increasing intraperitoneal doses of ZnPcS4 were given to Swiss Webster mice to assess acute toxicity. Based on mouse toxicity data, a phase I clinical trial of ZnPcS4‐based PDT in tumor‐bearing dogs was designed, using an accelerated titration scheme starting at 0.5% of the minimum toxic dose in mice. 24‐hours after ZnPcS4 administration tumors were irradiated with 675 nm light and dogs were evaluated by routine hematology and serum biochemistry at regular intervals after PDT.
Results:  Doses >125 mg/kg were associated with acute toxicity and mortality in Swiss Webster mice, suggesting the minimum toxic dose is 120–125 mg/kg. One dog, a Golden retriever with a massive malignant fibrous histiocytoma, has been entered into the phase I clinical trial. No deleterious effects were noted after ZnPcS4 administration. Within 48 hours of PDT, the tumor was dark and necrotic, with no grossly visible changes to the surrounding normal tissues. Histological examination of the PDT‐treated tumor confirmed widespread necrosis and thrombosis consistent with PDT‐mediated damage. The owner reported no adverse effects after treatment.
Conclusions:  Although preliminary data are encouraging, additional evaluation of ZnPcS4‐based PDT is required to determine its role in veterinary medicine.     

Anti‐Tumor Effect of Adenovirus‐Mediated P53 Gene Therapy on the Growth of Canine Osteosarcoma Transplanted into Nude Mice

Introduction:  It has been reported that 40–50% of canine osteosarcoma cases have p53 mutations. The p53 tumor supressor gene plays a central role in cell cycle regulation and induction of apoptosis. We previously showed that adenoviral vector expressing canine P53 (AxCA‐cp53) inhibited growth of cultured canine osteosarcoma cell lines. Here, we evaluated anti‐tumor effect of adenovirus‐mediated p53 gene therapy on the growth of canine osteosarcomas transplanted into nude mice.
Methods:  Nine nude mice were subcutaneously injected with cells of a canine osteosarcoma cell line (POS) having p53 gene mutation. The transplanted tumors formed into nude mice were injected with AxCA‐cp53, AxCA‐LacZ (adenovirus vector expressing LacZ) or PBS (3 mice each) 7 times during 15 days. Tumor sizes were measured every 3 days for 27 days after injection with the adenovirus vectors. Expression efficiency of the adenovirus‐mediated gene transfer was examined by X‐gal staining and P53 immunostaining. Effects of the P53 expression on cell cycle control were examined by RT‐PCR for expression of p21 gene downstream of P53.
Results:  Significant differences in the tumor size was observed between the transplanted osteosarcoma tissues injected with AxCA‐cp53 and those injected with AxCA‐LacZ or PBS. Expressions of LacZ and P53 were confirmed at the injection sites of the tumors. Moreover, p21 mRNA expression was shown to be induced in the AxCA‐cp53‐injected tumors, indicating the funciton of P53 to induce cell cycle arrest.
Conclusions:  Adenoviral vector expressing canine P53 inhibited the growth of canine ostersarcoma transplanted into nude mice.     

Xenogeneic Human Tyrosinase DNA Vaccination Induces Specific Immune Response in Dogs with Advanced Malignant Melanoma

Introduction:  Xenogeneic melanosomal differentiation antigens, delivered in the form of a plasmid DNA vaccine, can overcome host immune ignorance/tolerance in preclinical animals to melanoma by: 1) generating humoral and cytotoxic T cell responses and 2) inducing protection from tumor challenge. Initial trials of human tyrosinase (huTyr) DNA vaccination of dogs with advanced malignant melanoma (Bergman et al , 2003) demonstrated safety and prolongation of survival with this therapeutic modality. We investigated antigen‐specific immunity in dogs receiving huTyr DNA vaccination.
Methods:  Three cohorts of three dogs each with advanced (WHO stage II‐IV) canine malignant melanoma (CMM) received four biweekly IM injections (dose levels 100, 500, or 1,500 ug, respectively/vaccination) of huTyr plasmid DNA via the Biojector2000 jet delivery device. Sera samples were taken before and after vaccination to detect specific antibody formation to huTyr by Enzyme‐Linked ImmunoSorbent Assays (ELISAs) and flow‐cytometry.
Results:  Three dogs have measurable, 2 to 4 fold huTyr‐specific antibody titers. Preliminary studies by flow‐cytometry have confirmed antibody response to huTyr by positive binding to endogenous human tyrosinase in SK‐Mel188 cells using post‐vaccinate serum.
Conclusions:  Xenogeneic (huTyr) DNA vaccination generates antigen‐specific humoral responses, which may partially explain the previously reported clinical efficacy and anti‐tumor responses. Ongoing studies include: 1) a comprehensive analysis by flow‐cytometry to detect huTyr‐specific antibodies and 2) a quantitative measure of potential cytotoxic T‐cell responses in these dogs by the DNA vaccination.