Introduction
Hemangiosarcoma in dogs is highly aggressive and incurable, and this tumor of blood vessel–forming cells most commonly originates in the spleen, with the second and third most common sites of origin being the heart and skin, respectively. It has been estimated that HSA comprises 45% (17/38)1 to 51% (122/241)2 of all splenic malignancies in dogs. Visceral forms of HSA are highly metastatic, with metastases often already present when HSA is diagnosed.3 Dogs with HSA have a poor to grave prognosis, with internal hemorrhage from tumor rupture or impacts from metastases being the most common causes of death. Reported MSTs are between 4 and 6 months for dogs that undergo surgical treatment and adjuvant chemotherapy for HSA,4,5 with 11% (17/154)6 of dogs alive 1 year following treatment.2,7,8,9,10,11,12 Given that a decision of whether to proceed with surgical treatment is often made rapidly owing to life-threatening hemorrhage occurring in affected dogs and that some such dogs die perioperatively, whereas others may live ≥ 1 year after treatment, we sought to identify variables that could help predict outcome in affected dogs, thereby helping to guide clinicians in their recommendations and owners in their decisions regarding whether to pursue treatment.
Currently, tumor stage and location are the only reliable prognostic parameters for HSA in dogs, with shorter MSTs in dogs with HSA of an advanced stage (ie, stage 2 or 3 [MST, 62 to 210 days] vs stage 1 [MST, 345 days])8,9,13 or in a visceral location (MST, 86 days vs 780 days for dogs with HSA in cutaneous locations).9,11 Dogs with splenic HSA frequently have vague clinical signs (eg, lethargy, inappetence, and weight loss) and nonspecific hematologic abnormalities; however, some hematologic findings are frequently associated with HSA.14 For example, even in the absence of hemoabdomen, anemia is frequently observed.15 Anemia may be regenerative or nonregenerative, depending on duration. Additionally, thrombocytopenia12,16 and leukocytosis are common findings.11 To date, no reliable prognostic indicators readily available at presentation have been identified for dogs with splenic HSA treated with surgery and chemotherapy.17 Therefore, in the study reported here, we retrospectively assessed findings from dogs with splenic HSA that received surgical treatment combined with chemotherapy. We aimed to identify physical examination or perioperative CBC variables in dogs with splenic HSA that could aid in predicting PFI and OST in affected dogs.
Materials and Methods
Animals
The medical records database at the UMNVMC was searched for records of dogs with splenic masses that survived splenectomy performed from September 2004 to October 2016. Dogs were included if splenic HSA had been histologically diagnosed and if information in their medical record indicated intent to follow surgery with systemic chemotherapy (eg, protocols with various chemotherapeutic agents, including doxorubicin, cyclophosphamide, chlorambucil, dacarbazine, piroxicam, lomustine, and vincristine, alone or in combination). Administration of herbal supplements and experimental treatments were not exclusionary. Dogs that had been enrolled in either of 2 previously completed clinical trials18,19 evaluating the safety and efficacy of a bispecific EGF-urokinase angiotoxin20 for use in treating HSA were included if they met the inclusion criteria. In both of those trials,18,19 the dogs had received doxorubicin (30 mg/m2, IV, q 3 wk for 5 doses). Dogs included in the present study were grouped on the basis of whether they had been enrolled in the initial18 or secondary19 trial evaluating the safety and efficacy of a bispecific EGF-urokinase angiotoxin (SRCBST1 or SRCBST2 treatment group, respectively) or had not been enrolled in either of those trials (control group).
Data collection
Data available for dogs at the time of splenectomy were collected from medical records and included signalment, estimated BSA, preoperative clinical laboratory findings (eg, Hct, WBC count, and platelet count), presence of hemoabdomen (yes or no), blood transfusion (yes [and the number of transfusion instances] or no), HSA stage (1, 2, or 3), chemotherapy protocol and any additional treatments, PFI (duration from splenectomy to progressive disease), and OST (duration from splenectomy to death). Outcome was used to describe PFI and OST together. Because of the retrospective nature of our study, CBCs from ≤ 1 week before splenectomy to ≤ 48 hours following surgery were evaluated. When the Hct was unavailable, the PCV was recorded instead. The number of transfusions was not based on the number of units of blood product required but rather defined as the number of separate instances that a transfusion was administered during hospitalization. Staging was in accordance with the modified World Health Organization guidelines,3,17 such that stage 1 HSA had disease confined to the spleen; stage 2 involved ruptured splenic masses, regional lymph node involvement, or both; and stage 3 had distant metastasis.
Statistical analysis
Kaplan-Meier analysis was used to determine median PFI and OST, and Cox regression analysis was used to identify associations between variables of interest and PFI and OST. First, initial models were fit for each CBC variable separately, with HSA stage and treatment as covariatesa and HRs with 95% CIs and associated P values reported. Variables of interest from CBCs were Hct, WBC count, and platelet count, and potential covariates included dog age, sex, and BSA; hemoabdomen; transfusion; and hospital or clinic where the splenectomy and chemotherapy were performed. Those models were then refit in 2 additional multivariable models, one to include all 3 CBC variables with HSA stage and treatment as covariates and another to include all 3 CBC variables and all evaluated covariates with HSA stage and treatment again as covariates. In all models, age and BSA were treated as continuous predictor variables, and standardized HRs were reported; all other variables were treated as categorical variables. The facilities where splenectomy and chemotherapy were performed were coded as either UMNVMC or other (for dogs referred to the UMNVMC). Results for Hct, WBC count, and platelet count were divided into terciles because of variation between reference ranges from the numerous analyzers used. Therefore, anemia, leukocytosis, and thrombocytopenia were defined as values outside the reference limits for the laboratories where tests were performed. Population Pearson correlation coefficient (ρ) analysis was performed to identify potential associations between variables of interest. Platelet counts for blood samples recorded as having had clumped platelets were categorized on the basis of the minimum number given by either an analyzer (if used) or manual count (if performed). All analyses were performed with available software,a and values of P < 0.05 were considered significant.
Results
A search of the medical records database identified records of 86 dogs with splenic HSA diagnosed between September 2004 and October 2016. Included dogs survived splenectomy, and their medical records indicated an intent to treat the dogs with systemic chemotherapy. Sixteen dogs were excluded because CBC results were unavailable, leaving 70 dogs for inclusion in the study. Of these 70 evaluable dogs, 21 and 18 had been enrolled in an initial18 or secondary19 clinical trial of a sarcoma bispecific toxin for use in treatment of HSA (SRCBST1 and SRCBST2 treatment group, respectively), and 31 had not been enrolled in either of those trials (control group). Thirty-four of the 70 (49%) dogs were spayed females, 32 (46%) were castrated males, and 4 (6%) were sexually intact males (Table 1). Dogs were most commonly reported as Golden Retrievers (14/70 [20%]), Labrador Retrievers (12 [17%]), German Shepherd Dogs (10 [14%]), mixed-breed dogs (6 [9%]), or English Springer Spaniels (5 [7%]). Overall, the mean ± SD age and BSA were 9.36 ± 2.25 years (median, 9 years; range, 4 to 14 years) and 0.98 ± 0.29 m2 (median, 0.91 m2; range, 0.20 to 1.46 m2). Thirty-five of the 70 (50%) dogs had a BSA between 0.6 and 1.0 m2. Fifty-five of the 70 (79%) dogs had stage 2 HSA, whereas 8 (11%) had stage 1 HSA and 7 (10%) had stage 3 HSA.
Findings for variables from physical examinations and CBCs performed between 6 days before and 2 days after splenectomy in 70 client-owned dogs with splenic HSA treated with splenectomy and chemotherapy between September 2004 and October 2016.
| Variable | No. (%) of dogs | 95% CI |
|---|---|---|
| Tumor stage | ||
| 1 | 8 (11) | 5.1–21.3 |
| 2 | 55 (79) | 67.1–87.5 |
| 3 | 7 (10) | 4.1–19.5 |
| Reproductive status | ||
| Spayed female | 34 (49) | 36.4–60.8 |
| Castrated male | 32 (46) | 33.7–58.1 |
| Sexually intact male | 4 (6) | 1.6–14.0 |
| Breed or type of dog | ||
| Golden Retriever | 14 (20) | 11.4–31.3 |
| Labrador Retriever | 12 (17) | 9.2–28.0 |
| German Shepherd Dog | 10 (14) | 7.1–24.7 |
| Mixed-breed dog | 6 (9) | 3.2–17.7 |
| English Springer Spaniel | 5 (7) | 2.4–15.9 |
| Airedale Terrier | 2 (3) | 0.3–9.9 |
| Bichon Frise | 2 (3) | 0.3–9.9 |
| Miniature Poodle | 2 (3) | 0.3–9.9 |
| Vizsla | 2 (3) | 0.3–9.9 |
| Other* | 15 (21) | 12.5–32.9 |
| Hemoabdomen | ||
| Yes | 60 (86) | 75.3–92.9 |
| No | 10 (14) | 7.1–24.7 |
| Transfusion | ||
| Yes | 37 (53) | 40.6–64.9 |
| No | 33 (47) | 35.1–59.4 |
| Treatment facility | ||
| UMNVMC | 39 (56) | 43.3–67.6 |
| Other | 31 (44) | 32.4–56.7 |
| Anemia | ||
| Yes | 52 (74) | 62.4–84.0 |
| No | 18 (26) | 16.0–37.6 |
| Leukocytosis | ||
| Yes | 32 (46) | 33.7–58.1 |
| No | 38 (54) | 41.9–66.3 |
| Thrombocytopenia | ||
| Yes | 50 (71) | 59.4–81.6 |
| No | 20 (29) | 18.4–40.6 |
Consisted of 1 each of an American Bulldog, Australian Terrier, Boston Terrier, Brittany Spaniel, Cairn Terrier, Chihuahua, Dachshund, German Shorthaired Pointer, Great Dane, Mastiff, Newfoundland, Papillon, Pembroke Welsh Corgi, Rat Terrier, and Scottish Terrier.
Hematologic evaluations
The timing of hematologic evaluations was grouped on the basis of whether the evaluations were performed between 6 and 4 days before splenectomy (n = 1), between ≤ 96 and > 24 hours before splenectomy (8), ≤ 24 hours before splenectomy (58), or ≤ 24 hours after splenectomy (3). No dogs had a high Hct; however, 52 of the 70 (74%) dogs had a low Hct, and 18 (26%) had an Hct within reference limits (Table 1). Because only 3 dogs had low WBC counts or high platelet counts (alone or in combination), those findings were considered clinically normal and grouped along with findings within reference limits. Thirty-two of the 70 (46%) dogs had leukocytosis, and 50 (71%) had thrombocytopenia. With regard to platelet counts, the medical records indicated that 1 blood sample could not be accurately analyzed by a hematologic analyzerb because of platelet clumping; however, a board-certified veterinary clinical pathologist had visually evaluated the sample and subjectively estimated it to have had an adequate platelet count. Because the analyzer used to evaluate this clumped sample estimated a minimum of 105,000 platelets/μL and had a lower reference limit of 129,000 platelets/μL, we classified this sample as thrombocytopenic.
PFI and OST
No dogs were alive or lost to follow-up at the study end point (June 2019); thus, no censoring was necessary. Median OST for all 70 dogs was 173.5 days (range, 21 to 1,956 days). Sixty-five dogs had information on PFI in their medical records, and the median PFI was 147 days (range, 33 to 1,030 days).
We first validated the robustness of our cohort by evaluating the relationships between HSA stage and OST and PFI, with treatment group (SRCBST1, SRCBST2, or control) as a covariate. After accounting for treatment group, dogs with stage 3 HSA were significantly (P < 0.001) more likely to have had a shorter PFI (HR, 6.6; 95% CI, 2.6 to 17.0) and OST (HR, 4.5; 95% CI, 1.9 to 10.6) than were dogs with stage 1 or 2 HSA combined. Additionally, after accounting for treatment group and stage, dogs with (vs without) thrombocytopenia were significantly more likely to have had a shorter PFI (HR, 2.22; 95% CI, 1.18 to 4.15; P = 0.013) and OST (HR, 2.04; 95% CI, 1.09 to 3.81; P = 0.026; Figure 1; Table 2).
Estimated HR for a shorter PFI (A) or OST (B) in 70 client-owned dogs with splenic HSA treated with splenectomy and chemotherapy between September 2004 and October 2016 with the listed key variables of patient characteristics, tumor stage, and treatment as estimated through separate multivariable Cox proportional hazards models, controlling for HSA stage (1, 2, or 3) and treatment group (SRCBST1, SRCBST2, or control). Referent groupings of dogs for results shown are dogs with stage 2 versus 1 or 3 HSA, the control group versus the SRCBST1 or SRCBST2 group, females versus males, and those without versus with the individually listed abnormalities or treatments. For each continuous variable (age and BSA), the standardized HR associated with an increase of 1 SD in the value of the variable is shown. Each solid circle and horizontal line represent the HR and 95% CI, respectively. The vertical line represents an HR of 1 (ie, no effect).
Citation: Journal of the American Veterinary Medical Association 258, 6; 10.2460/javma.258.6.630
Estimated HR for a shorter PFI (A) or OST (B) in 70 client-owned dogs with splenic HSA treated with splenectomy and chemotherapy between September 2004 and October 2016 with the listed key variables of patient characteristics, tumor stage, and treatment as estimated through separate multivariable Cox proportional hazards models, controlling for HSA stage (1, 2, or 3) and treatment group (SRCBST1, SRCBST2, or control). Referent groupings of dogs for results shown are dogs with stage 2 versus 1 or 3 HSA, the control group versus the SRCBST1 or SRCBST2 group, females versus males, and those without versus with the individually listed abnormalities or treatments. For each continuous variable (age and BSA), the standardized HR associated with an increase of 1 SD in the value of the variable is shown. Each solid circle and horizontal line represent the HR and 95% CI, respectively. The vertical line represents an HR of 1 (ie, no effect).
Citation: Journal of the American Veterinary Medical Association 258, 6; 10.2460/javma.258.6.630
Estimated HR for a shorter PFI (A) or OST (B) in 70 client-owned dogs with splenic HSA treated with splenectomy and chemotherapy between September 2004 and October 2016 with the listed key variables of patient characteristics, tumor stage, and treatment as estimated through separate multivariable Cox proportional hazards models, controlling for HSA stage (1, 2, or 3) and treatment group (SRCBST1, SRCBST2, or control). Referent groupings of dogs for results shown are dogs with stage 2 versus 1 or 3 HSA, the control group versus the SRCBST1 or SRCBST2 group, females versus males, and those without versus with the individually listed abnormalities or treatments. For each continuous variable (age and BSA), the standardized HR associated with an increase of 1 SD in the value of the variable is shown. Each solid circle and horizontal line represent the HR and 95% CI, respectively. The vertical line represents an HR of 1 (ie, no effect).
Citation: Journal of the American Veterinary Medical Association 258, 6; 10.2460/javma.258.6.630
Results of Cox proportional hazards regression modeling used to identify potential associations between variables of interest and a shorter PFI or OST for the 70 dogs described in Table 1.
| Initial models* | Multivariable models with 3 CBC variables† | Multivariable models with all variables‡ | ||||||
|---|---|---|---|---|---|---|---|---|
| Outcome | Variable | No. of dogs | HR (95% CI) | P value | HR (95% CI) | P value | HR (95% CI) | P value |
| PFI | Tumor stage | 65 | ||||||
| 1 | 7 | 1.35 (0.59–3.08) | 0.482 | 2.07 (0.82–5.24) | 0.125 | 1.91 (0.27–13.35) | 0.516 | |
| 2 | 51 | Referent | Referent | Referent | ||||
| 3 | 7 | 6.98 (2.69–18.09) | < 0.001 | 7.65 (2.87–20.40) | < 0.001 | 7.51 (2.65–21.34) | < 0.001 | |
| Treatment group | ||||||||
| SRCBST1 | 19 | 0.76 (0.42–1.39) | 0.377 | 0.82 (0.43–1.58) | 0.559 | 1.17 (0.55–2.47) | 0.680 | |
| SRCBST2 | 16 | 1.53 (0.79–2.96) | 0.205 | 1.50 (0.71–3.17) | 0.291 | 2.14 (0.90–5.11) | 0.086 | |
| Control | 30 | Referent | Referent | Referent | ||||
| Age (+1 SD)§ | 65 | 1.15 (0.88–1.51) | 0.314 | — | — | 1.19 (0.87–1.62) | 0.276 | |
| Sex | ||||||||
| Male | 32 | 1.12 (0.65–1.91) | 0.690 | — | — | 1.27 (0.70–2.30) | 0.428 | |
| Female | 33 | Referent | — | — | Referent | |||
| BSA (+1 SD)§ | 65 | 0.91 (0.71–1.18) | 0.489 | — | — | 1.04 (0.73–1.46) | 0.839 | |
| Hemoabdomen | ||||||||
| Yes | 56 | 1.69 (0.39–7.37) | 0.483 | — | — | 1.05 (0.19–5.80) | 0.956 | |
| No | 9 | Referent | — | — | Referent | |||
| Transfusion | ||||||||
| Yes | 35 | 1.46 (0.81–2.64) | 0.204 | — | — | 1.48 (0.78–2.79) | 0.226 | |
| No | 30 | Referent | — | — | Referent | |||
| Treatment facility | ||||||||
| UMNVMC | 37 | 1.24 (0.71–2.18) | 0.450 | — | — | 1.34 (0.72–2.49) | 0.363 | |
| Other | 28 | Referent | — | — | Referent | |||
| Anemia | ||||||||
| Yes | 48 | 1.83 (0.96–3.49) | 0.065 | 1.01 (0.40–2.56) | 0.977 | 0.92 (0.34–2.53) | 0.878 | |
| No | 17 | Referent | Referent | Referent | ||||
| Leukocytosis | ||||||||
| Yes | 31 | 0.93 (0.55–1.57) | 0.775 | 1.01 (0.59–1.72) | 0.983 | 0.99 (0.52–1.90) | 0.985 | |
| No | 34 | Referent | Referent | Referent | ||||
| Thrombocytopenia | ||||||||
| Yes | 46 | 2.22 (1.18–4.15) | 0.013 | 2.20 (0.89–5.45) | 0.090 | 2.34 (0.85–6.45) | 0.102 | |
| No | 19 | Referent | Referent | Referent | ||||
| OST | Stage | 70 | ||||||
| 1 | 8 | 0.87 (0.41–1.86) | 0.724 | 1.59 (0.66–3.82) | 0.302 | 1.68 (0.27–10.24) | 0.576 | |
| 2 | 55 | Referent | Referent | Referent | ||||
| 3 | 7 | 4.42 (1.85–10.52) | < 0.001 | 6.37 (2.44–16.64) | < 0.001 | 9.35 (3.07–28.48) | <0.001 | |
| Treatment group | ||||||||
| SRCBST1 | 21 | 0.72 (0.41–1.29) | 0.271 | 0.79 (0.42–1.47) | 0.452 | 0.97 (0.50–1.89) | 0.924 | |
| SRCBST2 | 18 | 1.22 (0.66–2.26) | 0.516 | 1.28 (0.65–2.52) | 0.469 | 1.73 (0.82–3.68) | 0.153 | |
| Control | 31 | Referent | Referent | Referent | ||||
| Age (+1 SD)§ | 70 | 1.22 (0.95–1.58) | 0.119 | — | — | 1.34 (0.99–1.79) | 0.056 | |
| Sex | ||||||||
| Male | 36 | 0.91 (0.54–1.53) | 0.714 | — | — | 0.84 (0.47–1.52) | 0.570 | |
| Female | 34 | Referent | — | — | Referent | |||
| BSA (+1 SD)§ | 70 | 1.00 (0.79–1.26) | 0.995 | — | — | 1.15 (0.84–1.57) | 0.381 | |
| Hemoabdomen | ||||||||
| Yes | 60 | 1.55 (0.35–6.77) | 0.560 | — | — | 1.38 (0.25–7.56) | 0.713 | |
| No | 10 | Referent | — | — | Referent | |||
| Transfusion | ||||||||
| Yes | 37 | 1.22 (0.72–2.07) | 0.455 | — | — | 1.28 (0.69–2.39) | 0.432 | |
| No | 33 | Referent | — | — | Referent | |||
| Treatment facility | ||||||||
| UMNVMC | 39 | 1.34 (0.77–2.32) | 0.303 | — | — | 1.09 (0.61–1.93) | 0.771 | |
| Other | 31 | Referent | — | — | Referent | |||
| Anemia | ||||||||
| Yes | 52 | 1.13 (0.64–2.00) | 0.672 | 0.55 (0.25–1.22) | 0.142 | 0.44 (0.19–1.01) | 0.052 | |
| No | 18 | Referent | Referent | Referent | ||||
| Leukocytosis | ||||||||
| Yes | 32 | 1.16 (0.68–1.97) | 0.595 | 1.19 (0.69–2.03) | 0.536 | 1.17 (0.60–2.30) | 0.638 | |
| No | 38 | Referent | Referent | Referent | ||||
| Thrombocytopenia | ||||||||
| Yes | 50 | 2.04 (1.09–3.81) | 0.026 | 3.09 (1.32–7.23) | 0.009 | 3.50 (1.43–8.55) | 0.006 | |
| No | 20 | Referent | Referent | Referent | ||||
Initial models were fit for each CBC and physical examination variable separately, with HSA stage and treatment as covariates.
Multivariable models that refit the initial models to include all 3 CBC variables (Hct, WBC count, and platelet count) and with HSA stage and treatment as covariates.
Multivariable models that refit the initial models to include all CBC and physical examination variables of interest and with HSA stage and treatment as covariates
For continuous variables (age and BSA), the standardized HR associated with an increase of 1 SD in the value of the variable is shown.
— = Not assessed in the model.
Given the strong impact of HSA stage on outcome, the effects of thrombocytopenia on PFI and on OST were assessed in separate Kaplan-Meier curves for dogs with stage 1 or 2 HSA and for dogs with stage 3 HSA. For dogs with stage 1 or 2 HSA, the median PFI was shorter for dogs with thrombocytopenia (138 days; 95% CI, 104 to 172 days; n = 41) versus those with platelet counts within or above reference limits (203 days; 95% CI, 149 to 479 days; 17; Figure 2). Similarly, for dogs with stage 3 HSA, the median PFI was shorter for dogs with thrombocytopenia (60 days; n = 5) versus those with platelet counts within or above reference limits (77 days; 2). Confidence intervals were not reported for dogs with stage 3 HSA because of the small sample size. This same pattern was observed for OST in that for dogs with stage 1 or 2 HSA, the median OST was shorter for those with thrombocytopenia (159 days; 95% CI, 133 to 218 days; n = 45) versus those with platelet counts within or above reference limits (259 days; 95% CI, 195 to 537 days; 18; Figure 3). For dogs with stage 3 HSA, the median OST was shorter for those with thrombocytopenia (60 days; n = 5) versus those with platelet counts within or above reference limits (154 days; 2). Again, confidence intervals were not reported for dogs with stage 3 HSA because of the small sample size.
Kaplan-Meier curves of PFI (median, 147 days; range, 33 to 1,030 days) for the 65 of 70 dogs described in Figure 1 with PFI information in their medical records and grouped on the basis of whether they had stage 1 or 2 (A) versus stage 3 (B) splenic HSA and whether they had thrombocytopenia (dashed lines) versus platelet counts within or above reference limits (solid lines) evident on hematologic evaluation performed perioperatively (between 6 days before and 2 days after splenectomy).
Citation: Journal of the American Veterinary Medical Association 258, 6; 10.2460/javma.258.6.630
Kaplan-Meier curves of PFI (median, 147 days; range, 33 to 1,030 days) for the 65 of 70 dogs described in Figure 1 with PFI information in their medical records and grouped on the basis of whether they had stage 1 or 2 (A) versus stage 3 (B) splenic HSA and whether they had thrombocytopenia (dashed lines) versus platelet counts within or above reference limits (solid lines) evident on hematologic evaluation performed perioperatively (between 6 days before and 2 days after splenectomy).
Citation: Journal of the American Veterinary Medical Association 258, 6; 10.2460/javma.258.6.630
Kaplan-Meier curves of PFI (median, 147 days; range, 33 to 1,030 days) for the 65 of 70 dogs described in Figure 1 with PFI information in their medical records and grouped on the basis of whether they had stage 1 or 2 (A) versus stage 3 (B) splenic HSA and whether they had thrombocytopenia (dashed lines) versus platelet counts within or above reference limits (solid lines) evident on hematologic evaluation performed perioperatively (between 6 days before and 2 days after splenectomy).
Citation: Journal of the American Veterinary Medical Association 258, 6; 10.2460/javma.258.6.630
Kaplan-Meier curves of OST (median, 173.5 days; range, 21 to 1,956 days) for the 70 dogs described in Figure 1 grouped on the basis of whether they had stage 1 or 2 (A) versus stage 3 (B) splenic HSA and whether they had thrombocytopenia (dashed lines) versus platelet counts within or above reference limits (solid lines) evident on hematologic evaluation performed perioperatively (between 6 days before and 2 days after splenectomy).
Citation: Journal of the American Veterinary Medical Association 258, 6; 10.2460/javma.258.6.630
Kaplan-Meier curves of OST (median, 173.5 days; range, 21 to 1,956 days) for the 70 dogs described in Figure 1 grouped on the basis of whether they had stage 1 or 2 (A) versus stage 3 (B) splenic HSA and whether they had thrombocytopenia (dashed lines) versus platelet counts within or above reference limits (solid lines) evident on hematologic evaluation performed perioperatively (between 6 days before and 2 days after splenectomy).
Citation: Journal of the American Veterinary Medical Association 258, 6; 10.2460/javma.258.6.630
Kaplan-Meier curves of OST (median, 173.5 days; range, 21 to 1,956 days) for the 70 dogs described in Figure 1 grouped on the basis of whether they had stage 1 or 2 (A) versus stage 3 (B) splenic HSA and whether they had thrombocytopenia (dashed lines) versus platelet counts within or above reference limits (solid lines) evident on hematologic evaluation performed perioperatively (between 6 days before and 2 days after splenectomy).
Citation: Journal of the American Veterinary Medical Association 258, 6; 10.2460/javma.258.6.630
Results for Hct correlated significantly (ρ = 0.58; P < 0.001) with results for platelet count. However, after including all 3 CBC variables (Hct, WBC count, and platelet count) together in multivariable regression analysis with HSA stage and treatment as covariates, significant associations between thrombocytopenia and shorter PFI (HR, 2.20; 95% CI, 0.89 to 5.45; P = 0.09) and OST (HR, 3.09; 95% CI, 1.32 to 7.23; P = 0.009) were still identified (Table 2).
On the basis of initial analysis after accounting for only HSA stage and treatment, we could not rule out a meaningful association between anemia and shorter PFI (HR, 1.83; 95% CI, 0.96 to 3.49; P = 0.065); however, no significant association between anemia and OST (HR, 1.13; 95% CI, 0.64 to 2.0; P = 0.670) was apparent (Table 2). Similarly, there was no relevant association between WBC count and outcome. When the combination of anemia, leukocytosis, and thrombocytopenia was considered, a modest association with shorter OST (P = 0.044) was evident; however, no significant association was evident with PFI (P = 0.078).
Results for variables of age, reproductive status, BSA, presence of hemoabdomen, transfusions, and clinical treatment facility were not significantly associated with PFI or OST (Figure 1; Table 2). Analysis to identify potential associations between breed and outcome were not performed because most breeds were represented by too few individuals (Table 1).
Finally, full models were fit with all variables and covariates. Adding covariates did not meaningfully change the estimated HR or significance of any of the 3 CBC variables, nor were any of the covariates significant in this full model (Table 2).
Discussion
To our knowledge, the present retrospective study was the first to identify an association between perioperative thrombocytopenia and shorter OST in dogs with splenic HSA. Consistent with previous reports,8,9,12,21,22 our results also indicated that stage 3 HSA at presentation was associated with shorter PFI and OST.
The demographics of dogs in the present study were generally as expected,5,7,21 and the study population consisted predominantly of older (median age, 9 years) Golden Retrievers, German Shepherd Dogs, and Labrador Retrievers. Although 50% (35/70) of the dogs in the present study had a BSA between 0.6 and 1.0 m2, equivalent to 15 to 31 kg (33 to 68.2 lb),17,21,22 BSA was not meaningfully associated with outcome. This finding indicated that even if certain large breeds might have a higher risk of developing HSA, they do not appear to develop a more aggressive form of the disease.
Dogs with stage 3 HSA consistently have worse outcomes, compared with dogs with a lower stage of disease,6,8,9,12,13 and our finding that stage 3 HSA was a negative prognostic indicator for dogs in the present study further supported this. The poorer prognosis for dogs with stage 3 (vs stage 1 or 2) HSA is mostly because, despite splenectomy, affected dogs have continued presence of macroscopic tumor that increases their risk of death from tumor-related causes, as tumor growth can lead to subsequent hemorrhage and metastasis. Conversely, dogs with stage 1 or 2 HSA often have most or all of their macroscopic tumors excised during splenectomy and thereby minimize their immediate risk of death from tumor-related causes.
When considered individually, WBC count, hemoabdomen, and blood transfusions were not associated with poor outcome in the dogs of the present study. To our knowledge, the effect of leukocytosis on outcome in dogs with splenic HSA had not been previously evaluated; however, the lack of significant association between hemoabdomen or transfusion and outcome in the present study was consistent with another report.21
Although anemia during treatment of HSA in dogs was associated with a shorter disease-free interval in a previous study,7 our findings did not indicate a significant association between anemia and shorter PFI (HR, 1.83; P = 0.065) but could not rule one out. Previous studies14,15 show a strong association between anemia and OST, as seen in the present study. Anemia may serve as a surrogate marker of tumor hypoxia and, consequently, chemoresistance. In solid tumors, many cells in hypoxic areas are shielded from doxorubicin exposure because the drug does not penetrate these regions well.23 This protective effect may allow these chemoresistant cells to survive, proliferate, and ultimately contribute to the progression of disease.
To our knowledge, this was the first report of an association between thrombocytopenia and shorter OST in dogs with splenic HSA. Given the positive correlation identified between Hct and platelet count and the independent impact of thrombocytopenia on OST in dogs of the present study, we believe that platelets are involved in a unique mechanism relative to OST in dogs. Worse outcome in dogs with HSA and thrombocytopenia may be associated with hemostatic disturbances. For instance, coagulopathy11 (including disseminated intravascular coagulation5) and thromboembolism are common complications in dogs with HSA.21,24,25 In dogs with pulmonary thromboembolism, the development of acute respiratory distress syndrome has a significant negative impact on survival time.21 Because our study did not assess coagulation data, thrombocytopenia as an indicator of long-term prognosis warrants additional evaluation.
A previous study24 shows that thrombocytopenia is associated with perioperative death in dogs with HSA undergoing splenectomy. Because all dogs in the present study survived splenectomy, perioperative death attributable to thrombocytopenia was not considered. Interestingly, a study26 links thrombocytopenia of < 200,000 platelets/μL with a poor outcome in dogs receiving doxorubicin chemotherapy for presumptive cardiac HSA. However, none of the dogs with thrombocytopenia in that study26 had manual blood films evaluated; thus, the possibility that platelet clumping might have artifactually resulted in low platelet counts cannot be excluded.
Beyond their role in hemostasis, platelets are increasingly appreciated for their role as regulators of the immune system.27,28 Therefore, it is also possible that thrombocytopenia could lead to reduced systemic inflammatory and immune responses and thus to worse outcomes. Because our study was exploratory, we encourage further investigation of the effect of thrombocytopenia and platelet immunoregulation in veterinary medicine.
Limitations of the present study included its retrospective nature along with the use of numerous blood analyzers with individual reference intervals necessitating qualification of hematologic results. In addition, our findings indicated that platelets were clumped in a blood sample analyzed from 1 dog, and although the best approximation was used, it might have been an underestimation. Further, not all dogs underwent full staging after splenectomy, and lesions consistent with metastasis (eg, cavitated or actively bleeding masses or pulmonary nodules) were assumed to represent disease progression. Although no significant association was identified between clinical treatment facility and PFI or OST, differences among clinicians and geographic locations might still have been sources of bias.
Our results suggested that an inverse relationship exists between perioperative thrombocytopenia and outcome for dogs with splenic HSA. This finding warrants prospective study to understand causal and mechanistic connections and ultimately may prove valuable in guiding treatment recommendations.
Acknowledgments
Funded in part by the Animal Cancer Care and Research Program of the University of Minnesota. Dr. Modiano is supported by the Alvin and June Perlman Chair in Animal Oncology.
The authors declare there were no conflicts of interest.
The authors thank Dr. Laura Garrett for her intellectual insights.
Abbreviations
| BSA | Body surface area |
| HR | Hazard ratio |
| HSA | Hemangiosarcoma |
| MST | Median survival time |
| OST | Overall survival time |
| PFI | Progression-free interval |
| UMNVMC | University of Minnesota Veterinary Medical Center |
Footnotes
R: A language and environment for statistical computing, version 3.6.2, R Foundation for Statistical Computing, Vienna, Austria. Available at: r-project.org. Accessed Jan 15, 2020.
Advia 120, Siemens Healthcare Diagnostics, Deerfield, Ill.
References
- 1. ↑
Day MJ, Lucke VM, Pearson H. A review of pathological diagnoses made from 87 canine splenic biopsies. J Small Anim Pract 1995;36:426–433.
- 2. ↑
Spangler WL, Kass PH. Pathologic factors affecting postsplenectomy survival in dogs. J Vet Intern Med 1997;11:166–171.
- 3. ↑
Thamm DH. Hemangiosarcoma. In: Withrow SJ, Vail DM, Page RL, eds. Withrow and MacEwen's small animal clinical oncology. 5th ed. St Louis: Elsevier-Saunders, 2013;679–688.
- 4. ↑
Lana S, U'ren L, Plaza S, et al. Continuous low-dose oral chemotherapy for adjuvant therapy of splenic hemangiosarcoma in dogs. J Vet Intern Med 2007;21:764–769.
- 5. ↑
Hammer AS, Couto CG, Filppi J, et al. Efficacy and toxicity of vac chemotherapy (vincristine, doxorubicin, and cyclophosphamide) in dogs with hemangiosarcoma. J Vet Intern Med 1991;5:160–166.
- 6. ↑
Wendelburg KM, Price LL, Burgess KE, et al. Survival time of dogs with splenic hemangiosarcoma treated by splenectomy with or without adjuvant chemotherapy: 208 cases (2001–2012). J Am Vet Med Assoc 2015;247:393–403.
- 7. ↑
Kahn SA, Mullin CM, De Lorimier L-P, et al. Doxorubicin and deracoxib adjuvant therapy for canine splenic hemangiosarcoma: a pilot study. Can Vet J 2013;54:237–242.
- 8. ↑
Sorenmo KU, Baez JL, Clifford CA, et al. Efficacy and toxicity of a dose-intensified doxorubicin protocol in canine hemangiosarcoma. J Vet Intern Med 2004;18:209–213.
- 9. ↑
Matsuyama A, Poirier VJ, Mantovani F, et al. Adjuvant doxorubicin with or without metronomic cyclophosphamide for canine splenic hemangiosarcoma. J Am Anim Hosp Assoc 2017;53:304–312.
- 10. ↑
Ward H, Fox LE, Calderwood-Mays MB, et al. Cutaneous hemangiosarcoma in 25 dogs: a retrospective study. J Vet Intern Med 1994;8:345–348.
- 11. ↑
Pintar J, Breitschwerdt EB, Hardie EM, et al. Acute nontraumatic hemoabdomen in the dog: a retrospective analysis of 39 cases (1987–2001). J Am Anim Hosp Assoc 2003;39:518–522.
- 12. ↑
Wood CA, Moore AS, Gliatto 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;34:417–421.
- 13. ↑
Kim SE, Liptak JM, Gall TT, et al. Epirubicin in the adjuvant treatment of splenic hemangiosarcoma in dogs: 59 cases (1997–2004). J Am Vet Med Assoc 2007;231:1550–1557.
- 14. ↑
Hammond TN, Pesillo-Crosby SA. Prevalence of hemangiosarcoma in anemic dogs with a splenic mass and hemoperitoneum requiring a transfusion: 71 cases (2003–2005). J Am Vet Med Assoc 2008;232:553–558.
- 15. ↑
Moore AS, Rassnick KM, Frimberger AE. Evaluation of clinical and histologic factors associated with survival time in dogs with stage II splenic hemangiosarcoma treated by splenectomy and adjuvant chemotherapy: 30 cases (2011–2014). J Am Vet Med Assoc 2017;251:559–565.
- 16. ↑
Hammer AS, Couto CG, Swardson C, et al. Hemostatic abnormalities in dogs with hemangiosarcoma. J Vet Intern Med 1991;5:11–14.
- 17. ↑
Batschinski K, Nobre A, Vargas-Mendez E, et al. Canine visceral hemangiosarcoma treated with surgery alone or surgery and doxorubicin: 37 cases (2005–2014). Can Vet J 2018;59:967–972.
- 18. ↑
Borgatti A, Koopmeiners JS, Sarver AL, et al. Safe and effective sarcoma therapy through bispecific targeting of EGFR and uPAR. Mol Cancer Ther 2017;16:956–965.
- 19. ↑
Borgatti A, Fieberg A, Winter AL, et al. Impact of repeated cycles of EGF bispecific angiotoxin (eBAT) administered at a reduced interval from doxorubicin chemotherapy in dogs with splenic haemangiosarcoma. Vet Comp Oncol 2020;18:664–674.
- 20. ↑
Schappa JT, Frantz AM, Gorden BH, et al. Hemangiosarcoma and its cancer stem cell subpopulation are effectively killed by a toxin targeted through epidermal growth factor and urokinase receptors. Int J Cancer 2013;133:1936–1944.
- 21. ↑
Lux CN, Culp WTN, Mayhew PD, et al. Perioperative outcome in dogs with hemoperitoneum: 83 cases (2005–2010). J Am Vet Med Assoc 2013;242:1385–1391.
- 22. ↑
Johnson KA, Powers BE, Withrow SJ, et al. Splenomegaly in dogs. Predictors of neoplasia and survival after splenectomy. J Vet Intern Med 1989;3:160–166.
- 23. ↑
Primeau AJ, Rendon A, Hedley D, et al. The distribution of the anticancer drug doxorubicin in relation to blood vessels in solid tumors. Clin Cancer Res 2005;11:8782–8788.
- 24. ↑
Wendelburg KM, O'Toole TE, McCobb E, et al. Risk factors for perioperative death in dogs undergoing splenectomy for splenic masses: 539 cases (2001–2012). J Am Vet Med Assoc 2014;245:1382–1390.
- 25. ↑
Maruyama H, Miura T, Sakai M, et al. The incidence of disseminated intravascular coagulation in dogs with malignant tumor. J Vet Med Sci 2004;66:573–575.
- 26. ↑
Mullin CM, Arkans MA, Sammarco CD, et al. Doxorubicin chemotherapy for presumptive cardiac hemangiosarcoma in dogs. Vet Comp Oncol 2016;14:e171–e183.
- 27. ↑
Sonmez O, Sonmez M. Role of platelets in immune system and inflammation. Porto Biomed J 2017;2:311–314.
- 28. ↑
Morrell CN, Aggrey AA, Chapman LM, et al. Emerging roles for platelets as immune and inflammatory cells. Blood 2014;123:2759–2767.
