Predictors of outcome in dogs with subcutaneous or intramuscular hemangiosarcoma

Kai-Biu Shiu Department of Medical Sciences, School of Veterinary Medicine, University of Wisconsin, Madison, WI 53706.

Search for other papers by Kai-Biu Shiu in
Current site
Google Scholar
PubMed
Close
 BVMS, DACVIM
,
Andrea B. Flory The Animal Medical Center, 510 E 62nd St, New York, NY 10065.

Search for other papers by Andrea B. Flory in
Current site
Google Scholar
PubMed
Close
 DVM, DACVIM
,
Christie L. Anderson The Animal Cancer Center, Colorado State University, Fort Collins, CO 80523.

Search for other papers by Christie L. Anderson in
Current site
Google Scholar
PubMed
Close
 DVM
,
Jackie Wypij Department of Veterinary Clinical Medicine, College of Veterinary Medicine, University of Illinois, Urbana-Champaign, IL 61802.

Search for other papers by Jackie Wypij in
Current site
Google Scholar
PubMed
Close
 DVM
,
Corey Saba Department of Small Animal Medicine and Surgery, College of Veterinary Medicine, University of Georgia, Athens, GA 30602.

Search for other papers by Corey Saba in
Current site
Google Scholar
PubMed
Close
 DVM, DACVIM
,
Heather Wilson Department of Small Animal Clinical Sciences, College of Veterinary Medicine and Biomedical Sciences, Texas A&M University, College Station, TX 77843.

Search for other papers by Heather Wilson in
Current site
Google Scholar
PubMed
Close
 DVM, DACVIM
,
Ilene Kurzman Department of Medical Sciences, School of Veterinary Medicine, University of Wisconsin, Madison, WI 53706.

Search for other papers by Ilene Kurzman in
Current site
Google Scholar
PubMed
Close
 MS, EdD
, and
Ruthanne Chun Department of Medical Sciences, School of Veterinary Medicine, University of Wisconsin, Madison, WI 53706.

Search for other papers by Ruthanne Chun in
Current site
Google Scholar
PubMed
Close
 DVM, DACVIM

Abstract

Objective—To identify prognostic factors in a large group of dogs with subcutaneous or intramuscular hemangiosarcoma (HSA) or both.

Design—Multi-institutional retrospective cohort study.

Animals—71 dogs with subcutaneous or intramuscular HSA.

Procedures—Medical records of affected dogs were reviewed. The following factors were evaluated for an association with outcome: dog age and sex, clinical signs, anemia, thrombocytopenia, neutrophilia, tumor stage at diagnosis, achievement of complete excision, intramuscular involvement, presence of gross disease, tumor recurrence, and treatment.

Results—Of the 71 cases identified, 16 (29%) had intramuscular tumor involvement. For all dogs, median time to tumor progression and overall survival time (OST) were 116 and 172 days, respectively; 25% survived to 1 year. Univariate analysis identified presence of clinical signs or metastasis at diagnosis, dog age, tumor size, use of any surgery, and presence of gross disease as predictors of time to tumor progression and OST. There was no significant difference in survival time between dogs with respect to type of HSA. Multivariate analysis confirmed that adequate local tumor control, tumor diameter ≤ 4 cm, presence of metastasis at diagnosis, and presence of gross disease were significantly associated with OST.

Conclusions and Clinical Relevance—Subcutaneous and intramuscular HSA remains a heterogeneous group of tumors that generally carries a poor prognosis. Adequate local control of smaller tumors with no associated clinical signs or metastasis may provide the best chance of long-term survival.

Abstract

Objective—To identify prognostic factors in a large group of dogs with subcutaneous or intramuscular hemangiosarcoma (HSA) or both.

Design—Multi-institutional retrospective cohort study.

Animals—71 dogs with subcutaneous or intramuscular HSA.

Procedures—Medical records of affected dogs were reviewed. The following factors were evaluated for an association with outcome: dog age and sex, clinical signs, anemia, thrombocytopenia, neutrophilia, tumor stage at diagnosis, achievement of complete excision, intramuscular involvement, presence of gross disease, tumor recurrence, and treatment.

Results—Of the 71 cases identified, 16 (29%) had intramuscular tumor involvement. For all dogs, median time to tumor progression and overall survival time (OST) were 116 and 172 days, respectively; 25% survived to 1 year. Univariate analysis identified presence of clinical signs or metastasis at diagnosis, dog age, tumor size, use of any surgery, and presence of gross disease as predictors of time to tumor progression and OST. There was no significant difference in survival time between dogs with respect to type of HSA. Multivariate analysis confirmed that adequate local tumor control, tumor diameter ≤ 4 cm, presence of metastasis at diagnosis, and presence of gross disease were significantly associated with OST.

Conclusions and Clinical Relevance—Subcutaneous and intramuscular HSA remains a heterogeneous group of tumors that generally carries a poor prognosis. Adequate local control of smaller tumors with no associated clinical signs or metastasis may provide the best chance of long-term survival.

Hemangiosarcoma is a malignant neoplasm of vascular endothelial cells and, in most anatomic locations, is associated with an aggressive biological behavior and poor prognosis. This type of neoplasm can arise in any tissue; however, in dogs, it is most commonly found in the spleen.1 Tumors in all but dermal locations are generally associated with rapid metastasis and short survival. Lower stage of disease at the time of treatment and use of adjuvant anthracycline-based chemotherapy are associated with improved survival times.2–8

Little literature exists regarding HSAs that arise in the subcutaneous or intramuscular tissues, and this paucity of knowledge has led to empirically based treatment decisions and imprecise prognostication. Published reports of subcutaneous or intramuscular HSA typically involve studies of sparse numbers of dogs with various forms of HSA and do not allow for meaningful analysis. Three studies7–9 are among those currently used as references for prognostication. In 1 study8 of cutaneous HSA, dogs with hypodermal (subcutaneous) involvement that were treated with surgery had a median survival time of only 172 days. The median survival time of 10 dogs with subcutaneous HSA treated with surgery and chemotherapy in 2 other studies2,10 was 242 days. In contrast, a study9 of 17 dogs with subcutaneous HSA treated with adjuvant doxorubicin revealed a much longer median survival time of 1,189 days.

Numbers of dogs in the aforementioned studies were low, and survival times varied among investigations. The purpose of the study reported here was to evaluate a larger number of dogs with subcutaneous or intramuscular HSA than has been evaluated previously, describe their outcomes, and identify prognostic variables. We hypothesized that dogs with small tumors, an absence of hematologic abnormalities, and an absence of metastatic disease at diagnosis would have longer survival times than dogs without these variables.

Materials and Methods

Dogs—Records of dogs in which subcutaneous or intramuscular HSA was diagnosed between January 1995 and January 2008 were identified in the veterinary medical databases at the following institutions: University of Wisconsin-Madison, Colorado State University, University of Illinois, Texas A&M University, University of Georgia, and The Animal Medical Center, New York. Inclusion criteria were a histologic diagnosis of primary subcutaneous or intramuscular HSA (excluding cutaneous HSA that had invaded the hypodermis and subcutaneous HSA that the attending clinician believed may be secondary to a systemic metastatic process), and follow-up of at least 4 months or known death date beyond 5 days after diagnosis. Follow-up was achieved by evaluation of medical records or telephone contact with referring veterinarians or owners.

Information extracted from the medical record for each dog included signalment, clinical history, results of staging tests and clinicopathologic analyses at diagnosis, and follow-up information. Tumor location (axial or appendicular), longest recorded diameter (size), histologic characteristics (attending pathologist's diagnosis of histologic margins and mitotic index [number of mitotic figures per hpf]), and presence of any intramuscular involvement were recorded. Details regarding the type of treatment instituted (surgery, chemotherapy, radiotherapy, or some combination thereof), whether ALC of the tumor (defined as complete tumor resection or incomplete tumor resected followed by adjuvant radiotherapy) was achieved, and TTP were also recorded.

Statistical analysis—The TTP was defined as the interval from diagnosis to local progression or recurrence, metastasis, or death. Overall survival time was defined as the time from diagnosis to death. Dogs were censored in survival and progression analyses for the following reasons: death or progression had not occurred at time of analysis, dog was lost to follow-up, or dog died of a cause other than HSA.

Risk factors analyzed for their effect on TTP and OST included the presence of anemia, neutrophilia, or thrombocytopenia (as defined by the reporting laboratories' reference limits); presence of metastasis at diagnosis; tumor location (axial or appendicular); presence of histologic evidence of intramuscular involvement; attempted surgical resection; achievement of complete histologic margins (as determined by the reporting pathologist); presence of gross disease (at the primary tumor site only); use of radiotherapy; use of chemotherapy; dog age and body weight; and tumor size and mitotic index. Kaplan-Meier product limit estimates were computed to estimate TTP and OST for noncontinuous variables, and the log-rank test was used to compare them. To evaluate continuous variables (dog age, body weight, and tumor size) for an association with TTP and OST, Cox proportional hazards regression analysis was performed.

For tumor size, various categories of sizes were also statistically evaluated to identify a cutoff with prognostic value. Variables that were considered significant (P < 0.05) upon univariate analysis were used as candidate variables for a multivariate Cox regression analysis. Selection of the final multivariate Cox regression model was performed by use of forward stepwise selection. All analyses were performed with commercially available software.a

Results

Dogs—Seventy-one client-owned dogs with subcutaneous or intramuscular HSA that met the inclusion criteria were included in the study. Mean ± SD age was 9.0 ± 2.3 years, and mean body weight was 33.4 ± 12.2 kg (73.5 ± 26.8 lb). Thirty-six of the 71 (51%) dogs were neutered males, 27 (38%) were spayed females, and 8 (11%) were sexually intact males. Breeds represented included the following: mixed (n = 14 [20%]), Labrador Retriever (14 [20%]), Golden Retriever (12 [17%]), and German Shepherd Dogs (5 [7%]). The remaining dogs (n = 26 [37%]) were of other pure breeds.

Of 70 dogs for which clinical signs at initial evaluation were recorded, 23 (33%) had signs associated with their tumors other than presence of a mass. Such signs included lameness (n = 12 dogs), anorexia (3), coughing (2), neurologic signs (2), bleeding mass (2), constipation (1), and vocal change (1). Complete blood counts were performed for 67 of 71 dogs at time of diagnosis. According to laboratory reference limits, 19 (28%) were anemic, 18 (27%) had neutrophilia, and 17 (25%) were thrombocytopenic. There were no consistent serum biochemical abnormalities. Sixty-four dogs underwent thoracic radiography, and 52 underwent abdominal ultrasonography at the time of diagnosis. Metastasis was identified in 23 dogs. Locations of metastases were known for 17 dogs and were as follows: lung (n = 6), local lymph node (3), spleen or liver (3), and heart (2). Three dogs had metastases to more than 2 sites at the time of diagnosis.

Tumors—Fifty-five (77%) dogs had subcutaneous HSA, and 16 (29%) had intramuscular HSA. Tumor location was recorded for 67 dogs. Forty-one of 67 (61%) dogs had tumors that were considered axial and were distributed as follows: thorax (n = 13), genital and inguinal (7), dorsum (7), neck (7), head (4), and axilla (3). The other 26 dogs had appendicular tumors that were distributed as follows: shoulder (n = 7), thigh (8), distal portion of the limb (4), unspecified site on limb (4), and hip (3). Three additional dogs had confirmed tumors at multiple sites, and location was not recorded for 1 dog.

Tumor size was recorded for 62 dogs. The median size was 6 cm (range, 0.5 to 34 cm). Nineteen (31%) dogs had tumors that were ≤ 4 cm in diameter, 13 (21%) dogs had tumors > 4 cm but ≤ 6 cm in diameter, and 30 (48%) dogs had tumors > 6 cm in diameter.

Treatment—In 51 (72%) dogs, surgical removal or debulking of their tumor was attempted. Margins of excision were known for 49 of these dogs, and in 22 of the 49 (45%), margins were complete. Mitotic index was available for 43 dogs. The median mitotic index was 2 (range, 0 to 14). Adequate local control (complete tumor resection and adjuvant radiotherapy for those with incompletely resected tumors) was achieved in 25 (35%) dogs. Thirty-two (45%) dogs continued to have gross signs of disease after initial treatment.

In total, 14 dogs received radiotherapy at their tumor sites, 9 of which underwent radiotherapy for unresected, measurable tumors. Radiotherapy for incompletely resected tumors was performed in 3 dogs (2 coarse fraction protocols of 4 × 8 Gy and 1 definitive protocol of 18 × 2.5 Gy). Two dogs received radiotherapy of histologically completely resected tumors. Overall, radiation protocols used included 3 × 8 Gy (n = 8 dogs), 4 × 8 Gy (4), 6 × 6 Gy (1), and 18 × 2.5 Gy (1).

Of the 36 dogs that received chemotherapy, 18 (50%) did not have local tumor control before administration of chemotherapy (9 dogs with incompletely excised HSA and 9 with gross evidence of HSA) and the rest (50%) were considered to have ALC (complete margins were achieved in 16, and adjuvant radiotherapy was used for incomplete resection in 2). For the 30 dogs for which the chemotherapy protocol was identified, 12 (40%) received doxorubicin alone and 11 (37%) received doxorubicin in combination with cyclophosphamide, dacarbazine, vincristine, or liposome-encapsulated muramyl tripeptide. Other chemotherapy drugs used included lomustine (n = 3 dogs), tumor necrosis factor α (1), and metronomic cyclophosphamide (1). Response to chemotherapy was not evaluated beyond recording the date of progression. Eleven (15%) dogs had no treatment for their tumors.

Outcome—Ten dogs were lost to follow-up, and 5 were still alive at the time of data compilation. Of the 61 dogs for which follow-up data were available, 10 (16%) developed local recurrence of HSA after a median of 241 days (range, 60 to 1,107 days). The median TTP for all dogs was 116 days (range, 0 to 1,449 days). The 17 dogs that received chemotherapy, with ALC and no evidence of metastasis at diagnosis, had a median survival time of 246 days (range, 73 to 1,107 days).

The nature of failure (progressive disease, death, or euthanasia) was recorded for 48 dogs. Thirty-seven (77%) had documented metastatic disease, and 11 (23%) had local disease progression. Follow-up consisting of thoracic radiography, abdominal ultrasonography, or both was performed every 3 months in 23 dogs; in the remainder, HSA was not restaged frequently or at all. Two dogs died of causes unrelated to HSA, and the remaining 54 dogs died of disease. The median OST for all dogs was 172 days (range, 8 to 1,449 days). The 1- and 2-year survival rates were 25% and 4%, respectively.

Necropsies were performed on 8 dogs. Three of these had widespread metastatic disease involving > 3 tissue types, and 5 dogs had solitary sites of metastatic disease involving the lungs (n = 3 dogs), retroperitoneal space (1), and brain (1).

Statistical evaluation—The cutoffs for tumor size that had prognostic value were ≤ 4 cm, > 4 cm but ≤ 6 cm, and > 6 cm. Factors significantly associated with a shorter TTP included dog age, increased tumor size, presence of clinical signs, presence of metastasis at diagnosis, lack of complete surgical margins, and no surgery (Table 1). Body weight, anemia, thrombocytopenia, neutrophilia, local recurrence of tumor, tumor mitotic index, and use of chemotherapy or radiotherapy did not significantly affect TTP. Factors significantly associated with a shorter OST included dog age, tumor size, presence of clinical signs, presence of anemia at diagnosis (Figure 1), presence of metastasis at diagnosis (Figure 2), no surgery, no ALC (Figure 3), and presence of gross disease (Table 2). Body weight, thrombocytopenia or neutrophilia at diagnosis, intramuscular involvement, tumor mitotic index, use of radiotherapy or chemotherapy, and local recurrence did not significantly affect OST. Tumor > 6 cm in diameter was significantly associated with a shorter TTP and OST than smaller tumors (Figure 4). Tumor size ≤ 4 cm was significantly associated with a longer OST than for larger tumors.

Figure 1—
Figure 1—

Kaplan-Meier OST curves for dogs with subcutaneous or intramuscular HSA, with (n = 19) or without (51) anemia at initial diagnosis. Hatch marks signify censored dogs. Nonanemic dogs lived significantly (P = 0.01) longer than anemic dogs.

Citation: Journal of the American Veterinary Medical Association 238, 4; 10.2460/javma.238.4.472

Figure 2—
Figure 2—

Kaplan-Meier OST curves for dogs with subcutaneous or intramuscular HSA, with (n = 23) or without (48) metastasis at initial diagnosis. Hatch marks signify censored dogs. Dogs without metastasis at initial diagnosis lived significantly (P = 0.034) longer than dogs with metastasis at the same time point.

Citation: Journal of the American Veterinary Medical Association 238, 4; 10.2460/javma.238.4.472

Figure 3—
Figure 3—

Kaplan-Meier OST curves for dogs with subcutaneous or intramuscular HSA, with (n = 25) or without (46) ALC of tumors. Hatch marks signify censored dogs. Dogs with ALC lived significantly (P = 0.005) longer than dogs without ALC.

Citation: Journal of the American Veterinary Medical Association 238, 4; 10.2460/javma.238.4.472

Figure 4—
Figure 4—

Kaplan-Meier OST curves for dogs with subcutaneous or intramuscular HSA of various sizes. Hatch marks signify censored dogs. Dogs with tumors ≤ 4 cm in diameter lived significantly (P = 0.012) longer than dogs with larger tumors.

Citation: Journal of the American Veterinary Medical Association 238, 4; 10.2460/javma.238.4.472

Table 1—

Results of univariate analysis of factors predictive of TTP in dogs with subcutaneous (n = 55) or intramuscular (16) HSA.

 Median TTP (d)  
VariablePresentAbsentHazard ratioP value
Signalment    
  Age1.120.033
  Body weight0.990.312
Findings at initial evaluation    
  Tumor size1.050.044
  Clinical signs present491671.870.028
  Anemia421501.750.060
  Thrombocytopenia421301.230.504
  Neutrophilia931201.060.843
  Metastasis491782.080.010
Tumor size    
  ≤ 4 cm193820.540.056
  > 4 cm and ≤ 6 cm1501050.960.898
  > 6cm771791.790.043
Histologic findings    
  Intramuscular involvement921301.320.379
  Mitotic index1.130.060
  Complete surgical margins achieved2121610.460.017
Treatment    
  Surgery172110.27< 0.001
  Radiotherapy1301141.350.317
  Chemotherapy150430.730.233
Outcome    
  ALC202720.530.026
  Gross disease232884.46< 0.001
  Local recurrence2122401.270.556

A value of P < 0.05 was considered significant. For variables with continuous values, a significant hazard ratio > 1 suggests an increase in risk of progression for dogs per each unit increase in the continuous variable, and a significant value < 1 suggests a decreased risk. For variables with categorical values, a significant hazard ratio > 1 suggests an increased risk of progression for dogs with versus without that factor, and a significantyalue < 1 suggests a decreased risk.

– = Not applicable.

Table 2—

Results of univariate analysis of factors predictive of OST in dogs with subcutaneous (n = 55) or intramuscular (16) HSA.

 Median OST (d)  
VariablePresentAbsentHazard ratioP value
Signalment    
  Age1.1260.034
  Body weight0.9940.625
Findings at initial evaluation    
  Tumor size1.0410.030
  Clinical signs982372.0750.013
  Anemia982261.9110.034
  Thrombocytopenia1301931.2940.415
  Neutrophilia1221931.2470.476
  Metastasis1152462.0890.012
Tumor size    
  ≤ 4 cm3641310.4260.014
  > 4 cm and ≤ 6 cm2121670 9310.842
  > 6 cm1302942.2840.006
Histologic findings    
  Intramuscular involvement1362121.4300.268
  Mitotic index1.1160.090
  Complete surgical margins achieved3991300.4600.017
Treatment    
  Surgery246490.194< 0.001
  Radiotherapy1661671.0650.844
  Chemotherapy2261110.6810.162
Outcome    
  ALC3991200.4370.007
  Gross disease893994.325< 0.001
  Local recurrence3993281.1820.702

See Table 1 for key.

Factors identified as independent prognostic factors for OST upon multivariate analysis included ability to achieve ALC, tumor size ≤ 4 cm, presence of metastasis at diagnosis, and presence of gross disease (Table 3).

Table 3—

Results of multivariate Cox proportional hazard analysis of factors predictive of OST in dogs with subcutaneous (n = 55) or intramuscular (16) HSA.

VariableHazard ratio95% Confidence intervalP value
Metastasis2.6801.114–6.446< 0.001
Tumor size ≤ 4 cm0.3120.107–0.9100.033
ALC0.1540.059–0.4030.028
Gross disease5.8152.396–14.116< 0.001

Discussion

The present study was designed to identify factors that predicted TTP and OTP in dogs with subcutaneous or intramuscular HSA. The median TTP in the study dogs was 116 days, and the median OST was 172 days. The 1-year survival rate was 25%, and only 4% of dogs were alive 2 years after initial evaluation. These results suggested that intramuscular and subcutaneous HSA carried a poor prognosis and the likelihood of long-term survival was low. Tumor location had no effect on outcome.

Dogs with tumor-associated clinical signs at the time of evaluation had a shorter TTP (P = 0.013) and OST (P = 0.028) than dogs without these signs. The most common clinical sign was lameness. Lameness may indicate intramuscular involvement, advanced disease, or compartment syndrome (defined in this context as high pressure confined within the muscle compartment).11 Dogs that had no tumor-associated clinical signs may have received treatment earlier in the course of disease; however, this possibility was not evaluated.

The finding that the presence of anemia at the time of diagnosis predicted a significantly shorter OST than in dogs without anemia (226 vs 98 days) was not surprising. Anemia in dogs with subcutaneous or intramuscular HSA may be due to chronic inflammatory disease, hemolysis, hemorrhage, or bone marrow disease.12 Microangiopathic hemolysis (identified by the presence of schistocytes) has been detected in dogs with experimentally induced HSA, and in 1 study,13 45% of dogs had anemia attributed primarily to this process. In our study, the presence of schistocytosis was not evaluated; furthermore, bone marrow evaluations were not performed and reticulocyte counts were rarely available. For these reasons, it was impossible to determine whether the anemia detected in the dogs was a marker of metastasis, hemolysis, local hemorrhage, or chronic disease.

In a previous study14 of visceral HSA in dogs, thrombocytopenia and neutrophilia were associated with poorer outcome in affected versus unaffected dogs. In the present study, the associations were not detected. Only 18 (26%) dogs had neutrophilia and 17 (25%) had thrombocytopenia at initial evaluation. Neutrophilia can be associated with, among other causes, inflammation and tissue necrosis.15 At our institution, it has been observed that histologic evidence of necrosis is less prominent in subcutaneous HSA than visceral HSA.b This may explain a lower incidence of neutrophilia than previously reported in dogs with visceral HSA.14 A study16 of hemostatic abnormalities in 24 dogs with HSA identified thrombocytopenia in 75% of the dogs and disseminated intravascular coagulation in 50%. None of the dogs in the present study were known to have widespread disease at diagnosis, which is a condition associated with disseminated intravascular coagulation and thrombocytopenia in dogs with HSA.16

Initial histologic evaluation of tumors revealed 16 dogs in the present study had intramuscular involvement; however, because there was no standardized second histologic review of tumors, this number may have been underestimated. In addition, we were unable to differentiate between primary intramuscular disease and subcutaneous disease that infiltrated the musculature. Hemangiosarcoma commonly does not respect tissue boundaries and can be primarily cutaneous in nature with subcutaneous infiltration or primarily subcutaneous with intramuscular involvement.8 The tissue involved can have prognostic importance; cutaneous HSA confined to the dermis is associated with a much better prognosis than subcutaneous infiltration by cutaneous HSA.8 Although most studies8,17 have failed to demonstrate a significant difference in survival times between subcutaneous and intramuscular HSA, a recent study9 showed that 4 dogs with intramuscular HSA had significantly worse median survival time (272.5 days) than did 17 dogs with subcutaneous HSA (1,189 days). The same was not true for our study.

On the other hand, we found that tumors < 4 cm had a significantly longer OST, whereas tumors > 6 cm were associated with a shorter OST and TTP. Tumor size is reportedly of prognostic value for many tumors, including oral melanoma, mammary carcinoma, and perianal and anal sac adenocarcinoma and is therefore included in many cancer staging schemes.18–22 It is conceivable that a larger tumor represents a longer duration or faster proliferation of disease than a small one does and thus increases the likelihood of metastatic disease.

Proliferation indices have been evaluated for various tumors and have been shown to have prognostic value for many3,14,23–26 Histologic grading schemes for HSA have been evaluated in dogs, but such schemes did not appear to be of prognostic value.2,3 Mitotic index is a measure of the proportion of cells in the mitosis phase of the cell cycle as observed via light field microscopy.27 The technique does not require special staining and is a convenient way to quantify tumor proliferation and thus infer the degree of biological aggressiveness. One of 2 studies3,14 that involved evaluation of the mitotic index in dogs with HSA found the index could be used to predict outcome, but the index had not been evaluated in dogs with subcutaneous or intramuscular HSA until now. In the 43 dogs in our study in which mitotic index was recorded, there was no association with TTP (P = 0.06) or OST (P = 0.09). Because standardized pathological examinations were not performed, it is possible that a type II error could exist, particularly with the P values so close to significance.

Results of univariate regression analysis indicated study dogs that underwent surgery (debulking or attempted resection) had a longer OST than dogs that did not have surgery (246 vs 46 days, respectively); however, surgical intervention was not significant upon multivariate analysis, suggesting that it is not an independent prognostic factor. It is likely that tumor size and presence of metastasis were confounding factors, as dogs with large tumors or metastasis at initial evaluation may have been less likely to have surgery recommended as part of their treatment protocol. Complete resection of tumors was also associated with longer survival than presence of incomplete surgical margins (399 vs 130 days, respectively); however, local tumor recurrence was not associated with a shorter TTP or OST than was no recurrence. Because metastatic disease was the major cause of failure (ie, progressive disease, death, or euthanasia; nature of failure was recorded for 37/48 dogs), we hypothesize that complete resection is more important for reducing the risk of development of metastatic disease than obtaining local tumor control. Nevertheless, dogs in which gross disease was evident after receiving no treatment or after having undergone surgery had a much shorter OST than those that had no gross evidence of neoplasia after surgery (89 vs 399 days). There is, however, a possibility that these findings are biased for reasons such as owners not being willing to pursue further treatment for measurably diseased dogs. These dogs with gross disease had a 5.8-fold higher risk of death than the other dogs, as revealed by multivariate analysis, which suggests that resection (whether complete or incomplete) is more beneficial than allowing gross disease to remain.

The use of radiotherapy did not have an impact on TTP or OST; however, the fact that various radiotherapy protocols were used renders interpretation of these results difficult. A study9 of dogs with subcutaneous and intramuscular HSA found that radiotherapy for treatment of microscopic local disease was associated with a significantly shorter disease-free interval than surgery alone. Another study28 demonstrated palliative coarse fraction radiotherapy is of benefit for treatment of gross, nonvisceral HSA. It also found coarse fraction (eg, 3 × 8 Gy) was the most common radiotherapy approach for incompletely resected tumors (in 3 dogs) and gross disease (in 9 dogs). Dogs with ALC of tumors in the present study included 3 with incompletely resected tumors treated with radiotherapy (coarse or definitive). Although none had local tumor recurrence, insufficient evidence exists to promote radiotherapy for incompletely resected subcutaneous HSA as acceptable for ALC.

The use of chemotherapy did not affect TTP (P = 0.233) or OST (P = 0.162) in the study reported here. Multiple chemotherapy protocols were used; however, most were doxorubicin based. Given the improvement in outcome associated with chemotherapy for treatment of visceral HSA,1–3,5,6 chemotherapy would likely be of benefit for the subcutaneous and intramuscular variants. A retrospective study9 of 21 dogs with subcutaneous or intramuscular HSA included dogs with ALC of their tumors and no evidence of metastasis at initial staging that received adjuvant doxorubicin chemotherapy. Although tumor size was not evaluated, the median survival time was 1,189 days for the 17 dogs with subcutaneous HSA. The dogs in our study did not have such a favorable outcome; 17 dogs with subcutaneous HSA, no metastases at presentation, ALC, and doxorubicin-based chemotherapy had a median survival time of only 246 days. However, 8 of these 17 dogs with ALC had tumors that were > 6 cm in diameter, which may account for the shorter survival time.

Selection bias and confounding factors are difficult to control for in retrospective studies, which should prevent them from being the basis for changes in the standard of care for a disease process. An example of a potential source of selection bias in the present study was the criteria for inclusion. These inclusion criteria were chosen to focus attention upon patients that would help test the study hypothesis (ie, not those that were euthanatized upon diagnosis or that were lost to follow-up within 4 months). As a result, an unknown proportion of dogs with subcutaneous or intramuscular HSA may have had progressive disease within 4 months after their initial evaluation, so the reported TTPs and OSTs may be longer than those of dogs in the target population (all dogs with subcutaneous or intramuscular HSA). However, our inclusion of dogs with a recorded death date beyond 5 days after initial evaluation should somewhat temper concerns regarding dogs with early progressive disease and early death from their disease. Furthermore, there is no way to determine whether our attempt to exclude primary visceral HSAs, while retaining subcutaneous tumors that appeared to have metastasized to the viscera, was accurate. This means that tumors considered to be primary subcutaneous HSA with metastasis to the viscera (usually characterized as small multifocal lesions rather than large masses) could, in fact, have been primary visceral HSA with metastasis to the subcutaneous tissues. The impact of this possibility on the outcome is unknown.

Retrospective studies are also limited in that non-uniform and infrequent reevaluation and follow-up of patients can result in a false impression of an increase in TTP. Another limitation of our study was the lack of a methodical review of histologic samples from all institutions and their referring veterinarians' pathologists. Tumors were not graded, and methods of determining the mitotic index and completeness of excision may have differed between pathologists, adding an element of uncertainty to our results. Lastly, the low number of dogs, particularly when categorized by specific treatments, reduced the statistical power of the study.

Overall, the study reported here revealed that dogs with nonmetastatic subcutaneous and intramuscular HSA at diagnosis and small tumors in which ALC can be achieved may be expected to have an improved outcome relative to other dogs with the same disease. However, these types of HSA should be considered aggressive because the OST was only 172 days. Additional studies are needed in which data are collected prospectively, tumor size is determined, and the importance of histologic criteria such as proliferation indices and grade is evaluated in a standardized manner.

ABBREVIATIONS

ALC

Adequate local control

HSA

Hemangiosarcoma

OST

Overall survival time

TTP

Time to tumor progression

a.

R software, version R2.10.1, R Foundation for Statistical Computing, Vienna, Austria.

b.

Steinberg H, Department of Pathobiological Sciences, School of Veterinary Medicine, University of Wisconsin, Madison, Wis: Personal communication, 2009.

References

  • 1.

    Brown NOPatnaik AKMacEwen EG. Canine hemangiosarcoma: retrospective analysis of 104 cases. J Am Vet Med Assoc 1985; 186: 5658.

  • 2.

    Hammer ASCouto CGFilppi J, et al. Efficacy and toxicity of VAC chemotherapy (vincristine, doxorubicin, and cyclophosphamide) in dogs with hemangiosarcoma. J Vet Intern Med 1991; 5: 160166.

    • Search Google Scholar
    • Export Citation
  • 3.

    Ogilvie GKPowers BEMallinckrodt CH, et al. Surgery and doxorubicin in dogs with hemangiosarcoma. J Vet Intern Med 1996; 10: 379384.

    • Search Google Scholar
    • Export Citation
  • 4.

    Prymak CMcKee LJGoldschmidt MH, et al. Epidemiologic, clinical, pathologic, and prognostic characteristics of splenic hemangiosarcoma and splenic hematoma in dogs: 217 cases (1985). J Am Vet Med Assoc 1988; 193: 706712.

    • Search Google Scholar
    • Export Citation
  • 5.

    Vail DMMacEwen EGKurzman ID, et al. Liposome-encapsulated muramyl tripeptide phosphatidylethanolamine adjuvant immunotherapy for splenic hemangiosarcoma in the dog: a randomized multi-institutional clinical trial. Clin Cancer Res 1995; 1: 11651170.

    • Search Google Scholar
    • Export Citation
  • 6.

    Wood CAMoore ASGliatto JM, et al. Prognosis for dogs with stage I or II splenic hemangiosarcoma treated by splenectomy alone: 32 cases (1991–1993). J Am Anim Hosp Assoc 1998; 34: 417421.

    • Search Google Scholar
    • Export Citation
  • 7.

    Hargis AMIhrke PJSpangler WL, et al. A retrospective clinicopathologic study of 212 dogs with cutaneous hemangiomas and hemangiosarcomas. Vet Pathol 1992; 29: 316328.

    • Search Google Scholar
    • Export Citation
  • 8.

    Ward HFox LECalderwood-Mays MB, et al. Cutaneous hemangiosarcoma in 25 dogs: a retrospective study. J Vet Intern Med 1994; 8: 345348.

    • Search Google Scholar
    • Export Citation
  • 9.

    Bulakowski EJPhilibert JCSiegel S, et al. Evaluation of outcome associated with subcutaneous and intramuscular hemangiosarcoma treated with adjuvant doxorubicin in dogs: 21 cases (2001–2006). J Am Vet Med Assoc 2008; 233: 122128.

    • Search Google Scholar
    • Export Citation
  • 10.

    Sorenmo KUJeglum KAHelfand SC. Chemotherapy of canine hemangiosarcoma with doxorubicin and cyclophosphamide. J Vet Intern Med 1993; 7: 370376.

    • Search Google Scholar
    • Export Citation
  • 11.

    Bar-Am YAnug AMShahar R. Femoral compartment syndrome due to haemangiosarcoma in the semimembranosus muscle in a dog. J Small Anim Pract 2006; 47: 286289.

    • Search Google Scholar
    • Export Citation
  • 12.

    Aird B. Clinical and hematologic manifestations of anemia. In: Bernard FFeldman JGZJain NC, eds. Schalm's veterinary hematology. 5th ed. Hagerstown, Md: Lippincott Williams & Wilkins, 2000;140142.

    • Search Google Scholar
    • Export Citation
  • 13.

    Rebar AHHahn FFHalliwell WH, et al. Microangiopathic hemolytic anemia associated with radiation-induced hemangiosarcomas. Vet Pathol 1980; 17: 443454.

    • Search Google Scholar
    • Export Citation
  • 14.

    Kim SELiptak JMGall TT, et al. Epirubicin in the adjuvant treatment of splenic hemangiosarcoma in dogs: 59 cases (1997–2004). J Am Vet Med Assoc 2007; 231: 15501557.

    • Search Google Scholar
    • Export Citation
  • 15.

    Schultze AE. Interpretation of canine leukocyte responses. In: Bernard FFeldman JGZJain NC, eds. Schalm's veterinary hematology. 5th ed. Hagerstown, Md: Lippincott Williams & Wilkins, 2000;366381.

    • Search Google Scholar
    • Export Citation
  • 16.

    Hammer ASCouto CGSwardson C, et al. Hemostatic abnormalities in dogs with hemangiosarcoma. J Vet Intern Med 1991; 5: 1114.

  • 17.

    Schultheiss PC. A retrospective study of visceral and nonvisceral hemangiosarcoma and hemangiomas in domestic animals. J Vet Diagn Invest 2004; 16: 522526.

    • Search Google Scholar
    • Export Citation
  • 18.

    Yamagami TKobayashi TTakahashi K, et al. Prognosis for canine malignant mammary tumors based on TNM and histologic classification. J Vet Med Sci 1996; 58: 10791083.

    • Search Google Scholar
    • Export Citation
  • 19.

    Vail DMWithrow SJSchwarz PD. Perianal adenocarcinoma in the canine male: a retrospective study of 41 cases. J Am Anim Hosp Assoc 1990; 26: 329334.

    • Search Google Scholar
    • Export Citation
  • 20.

    Hahn KADeNicola DBRichardson RC, et al. Canine oral malignant melanoma: prognostic utility of an alternative staging system. J Small Anim Pract 1994; 35: 251256.

    • Search Google Scholar
    • Export Citation
  • 21.

    MacEwen EGPatnaik AKHarvey HJ, et al. Canine oral melanoma: comparison of surgery versus surgery plus Corynebacterium parvum. Cancer Invest 1986; 4: 397402.

    • Search Google Scholar
    • Export Citation
  • 22.

    Polton GABrearley MJ. Clinical stage, therapy, and prognosis in canine anal sac gland carcinoma. J Vet Intern Med 2007; 21: 274280.

  • 23.

    Bostock DECrocker JHarris K, et al. Nucleolar organiser regions as indicators of post-surgical prognosis in canine spontaneous mast cell tumours. Br J Cancer 1989; 59: 915918.

    • Search Google Scholar
    • Export Citation
  • 24.

    Scase TJEdwards DMiller J, et al. Canine mast cell tumors: correlation of apoptosis and proliferation markers with prognosis. J Vet Intern Med 2006; 20: 151158.

    • Search Google Scholar
    • Export Citation
  • 25.

    Seguin BBesancon MFMcCallan JL, et al. Recurrence rate, clinical outcome, and cellular proliferation indices as prognostic indicators after incomplete surgical excision of cutaneous grade II mast cell tumors: 28 dogs (1994–2002). J Vet Intern Med 2006; 20: 933940.

    • Search Google Scholar
    • Export Citation
  • 26.

    Kuntz CADernell WSPowers BE, et al. Prognostic factors for surgical treatment of soft-tissue sarcomas in dogs: 75 cases (1986–1996). J Am Vet Med Assoc 1997; 211: 11471151.

    • Search Google Scholar
    • Export Citation
  • 27.

    Cross SSStart RD. Estimating mitotic activity in tumours. Histopathology 1996; 29: 485488.

  • 28.

    Hillers KRLana SEFuller CR, et al. Effects of palliative radiation therapy on nonsplenic hemangiosarcoma in dogs. J Am Anim Hosp Assoc 2007; 43: 187192.

    • Search Google Scholar
    • Export Citation
All Time Past Year Past 30 Days
Abstract Views 1372 0 0
Full Text Views 4552 2887 190
PDF Downloads 3491 1740 170
Advertisement