Introduction

Predictors of splenic neoplasia become an important factor in cases of nontraumatic hemoperitoneum in dogs. An owner’s decision for surgical intervention in cases of nontraumatic hemoperitoneum can be difficult because of the possibility of malignancy and reportedly poor expected survival times. The reported median survival time (MST) for large-breed dogs with splenic hemangiosarcoma (HSA) treated with splenectomy alone is 1–3 mo.^1–4 When adjuvant chemotherapy is used for large-breed dogs with splenic HSA, the MST is reported to be 3–8 mo.^3,5,6 Previous histopathologic studies report splenic neoplasia accounts for 48–65% of splenomegaly with HSA confirmed in 57–70% of these malignancies.^1,2,7

Multiple preoperative predictors for splenic neoplasia have been researched to date. In large-breed dogs, splenic malignancies are reported as the most common cause of nontraumatic hemoperitoneum.^1,8–10 Johnson et al. reported that the presence of coexisting preoperative anemia and hemoperitoneum was up to 69% accurate in predicting splenic neoplasia in their population of 100 dogs.² Research performed by Mallinckrodt and Gottfried showed that splenectomized dogs with a lower mass-to-spleen volume ratio had a higher chance of malignancy.¹¹ Abdominal ultrasonography is a common diagnostic tool used for preoperative evaluation; however, it has not been proved to be sensitive in differentiating between malignant and benign neoplastic lesions. In one study involving 20 dogs with splenic masses, contrast-enhanced ultrasonography did not allow distinction of splenic hematoma lesions from HSA.¹²

Previous studies investigating splenic lesions uniformly report larger-breed dogs having a higher risk for neoplasia with median weights of 35⁹, 33.6,³ 33.3,¹ 22.5 kg,¹³ and dogs weighing >21 kg.² The most common breeds reported to develop splenic neoplasia are golden retrievers, German shepherd dogs, Labrador retrievers, boxers, and standard poodles.^{1,2,5–7,9,11–13}

Anecdotally, it has been believed that small-breed dogs were more likely to be diagnosed with benign nodular splenic lesions. Recent reports have defined “small-breed dogs” as dogs weighing <15 kg in orthopedic-related studies.^14–16 A recent evaluation of small-breed dog splenomegaly by Corbin et al. reported results similar to previous large-breed dogs. In that study, 47% of small-breed dogs with splenomegaly had malignant lesions with 67% of malignant lesions being HSA.¹⁷ In that same study, malignancy was not significantly associated with hemoperitoneum.¹⁷ Given the small population size (45 dogs total) in the Corbin et al. study, further investigation with a larger population of small-breed dogs may provide additional information regarding predictive factors for splenic malignancy, incidence of varying histopathologic diagnoses, associations with malignancies, and survival times.

The purpose of this study was to evaluate a full range of nodular splenic lesions in a large population of small-breed dogs (<15 kg) that were splenectomized. MSTs and any predictive risk factors affecting survival were also evaluated. We hypothesized that splenic malignancies would be similar in frequency, with similar MSTs, when compared with previous large-breed dog studies. We also hypothesized that nodular splenic lesions in combination with hemoperitoneum would be positively associated with malignancy, as seen in previous large-breed dog studies.

Materials and Methods

Statistical Analysis

Nominal data was reported as frequency and percentages based on population and descriptors including diagnosis, malignancy, breed, and presence of hemoperitoneum (Tables 1, 2). A χ² test or Fischer exact test was used to evaluate associations between hemoperitoneum and nodular splenic lesion diagnoses. A χ² test or Fisher exact test was also used to evaluate breeds of interest for an association between breed and the presence of malignancy, hemoperitoneum, or malignant hemoperitoneum (present or absent). Odds ratio was used to evaluate risks of having malignancy with the presence of hemoperitoneum.

TABLE 1 Pathologic Causes of Splenic Lesions >1 cm in 370 Small-Breed Dogs (<15 kg)

View Fullscreen

TABLE 2 Hemoperitoneum at the Time of Splenectomy in 370 Small-Breed Dogs (<15 kg) with Nodular Splenic Lesions (>1 cm)

View Fullscreen

Survival was defined as the time from the date of surgery to the date of known death. Dogs reported as alive at the last known follow-up date of >2 mo postoperatively were censored for survival time analysis. Dogs with no known follow-up (lost to follow-up after routine 2 wk postoperative recheck) were excluded from the survival analysis. Survival times of different histopathologic lesions (with or without hemoperitoneum) were examined by Kaplan-Meier plot and compared by means of a log-rank (Mantel-Cox) test. Cox proportional hazards regression was used to evaluate associations of age, weight, sex, altered status, lesion type, presence of malignancy, presence of hemoperitoneum, and breed with survival.

Statistical analyses were performed using statistical computing programs^a,b. For all data, a value of P ≤ .05 was used to determine statistical significance.

Results

Survival and Risk Factors

Follow-up data was available for 213 (58%) dogs. One hundred fifty-seven (42%) dogs were excluded from survival time analysis as a result of being lost to follow-up. Known date of death was available for 134 (36%) dogs. The most common splenic lesions in this study with available survival data include HSA, hematoma, and benign nodular hyperplasia/extramedullary hematopoiesis. Comparison of survival data was evaluated, and significant differences (P < .05, Df 1) existed between different diagnoses of splenic lesions. Log-rank (Mantel-Cox) test showed a significant difference in survival for dogs with malignant nodular splenic lesions versus those with benign and nonneoplastic (all nonmalignant lesions combined) nodular splenic lesions with P = .0001. The MST for nonmalignant nodular splenic lesions was 983 days (range, 0–3620 days; 95% CI, 671–1295 days) and for malignant nodular splenic lesions was 143 days (range, 0–1439 days; 95% CI, 96–189 days). The hazard ratio for this particular survival analysis is 2.8 (95% CI, 1.95–4.12; P = .0001), meaning that at any given postoperative period, there is a 2.8 times greater likelihood of death in dogs with malignant nodular splenic lesions. There was a significant difference in survival between splenic HSA versus all other nodular splenic lesions (P = .0001; Figure 1). The MST of splenic HSA in this population of dogs was 93 days (range, 3–614 days; 95% CI, 69–117 days) compared with all other nodular splenic lesions (benign and non-HSA neoplastic) with an MST of 686 days (range, 0–3620 days; 95% CI, 471–901 days). The MST of non-HSA malignant lesions was 369 days (range, 0–1284 days; 95% CI, 179–559 days).

View Figure

• Download as Powerpoint
• Download Figure

There was a significant difference in survival (P < .05, Df 1) for dogs with hemoperitoneum and for those without hemoperitoneum (Figure 2). The MST for dogs with nodular splenic lesions and presence of hemoperitoneum was 127 days (range, 2–1735 days; 95% CI, 54– 200 days), and the MST for dogs without hemoperitoneum was 481 days (range, 0–3620 days; 95% CI, 286–675 days). There was a significant difference in survival (P = .005, Df 1) when comparing dogs who had malignant hemoperitoneum with dogs who had malignant splenic lesions without hemoperitoneum. The MST of dogs with malignant hemoperitoneum was 95 days (range, 2–614 days; 95% CI, 51–139 days), and the MST of dogs with malignant nonhemoperitoneum was 309 days (range, 0–1439 days; 95% CI, 139–479 days).

View Figure

• Download as Powerpoint
• Download Figure

Of the 100 dogs diagnosed with HSA, 60 had follow-up data and were evaluated for survival analysis. Nineteen of those received adjuvant chemotherapy, with the most common agents used being doxorubicin, adriamycin, cyclophosphamide, and chlorambucil. The MST for dogs with HSA who received adjuvant chemotherapy (127 days [range, 42–614 days]) was not significantly different (P = 1.57) compared with dogs not receiving chemotherapy (63 days [range, 3–290 days]; Table 3). There was no significant difference in survival times when comparing all other histopathologic diagnoses and the following groups of splenic lesions (P > .05, Df 1): hematoma, non-HSA splenic malignancies, and benign neoplastic lesions. There was no significant difference in survival when comparing different breeds (P > .05, Df 1).

TABLE 3 Results of Univariable Kaplan-Meier Analysis of Survival Data Available for 213 Small-Breed Dogs (<15 kg) Who Underwent Splenectomy for Nodular Splenic Lesions >1 cm in Size

View Fullscreen

Discussion

The median age for the small-breed dogs in the present study was 11 yr, which is similar to previous studies on dogs with splenic lesions reporting mean and median ages ranging from 9.8 to 11 yr.^{1–7,11,13,17,19} The median weight at time of splenectomy of 9.5 kg (20.9 lb) is similar to the recent small-breed study that reported a median weight of 10.3 kg (22.7 lb).¹⁷ We found no significant association between malignancy and patient sex, age, or weight in the current study. Previous large-breed studies have reported sex associations with splenic HSA.^4,20 The current study did not identify associations between splenic HSA and sex, age, or weight. The most common pure-breeds with nodular splenic lesions in our study were miniature schnauzers (n = 30), dachshunds (n = 28), and beagles (n = 26), representing 8, 8, and 7%, respectively, of the total population. These results differ from the Corbin et al. study, in which the most common reported breeds were wheaten terriers (n = 6), bichon frise, cocker spaniels, and Pembroke Welsh corgis.¹⁷ These breeds were not as common in our study of 370 small-breed dogs, collectively representing <9%. Interestingly, beagles as an individual breed and terrier breeds as a group were found to be associated with malignant splenic lesions in the current study. Furthermore, the terrier breeds as a group were also associated with the presence of hemoperitoneum and malignant hemoperitoneum at the time of splenectomy. This has not been previously reported.

In the present study, the near equal distribution of neoplastic (56%) to nonneoplastic (44%) nodular splenic lesions was comparable to previously reported data. A 1997 study by Spangler and Kass evaluated splenic lesions in 500 large-breed dogs and revealed a near equal distribution of neoplastic (48%) and nonneoplastic (51%) nodular splenic lesions.⁷ A recent study by Corbin et al. evaluated splenic lesions in 45 small-breed dogs (<16 kg) and reported that 47% had a diagnosis of a malignant splenic lesion.¹⁷ The diagnosis of splenic HSA in the present study population of dogs (n = 370) was 27%. Previous large-breed dog studies report a diagnosis of splenic HSA in 24–70% of cases.^1,2,7 In the recent small-breed study, splenic HSA was diagnosed in 31% of the population.¹⁷ The landmark large-breed study by Johnson reported that of dogs who present with splenomegaly, 65% have splenic neoplasia, and of this group subset, 66% were diagnosed with HSA: the “two-thirds, two-thirds rule.”² In the present study of small-breed dogs weighing <15 kg with nodular splenic lesions, we report 56% (approximately one-half) have splenic neoplasia, and of this group subset, 54% (one-half) were diagnosed with HSA (Table 1).

Additional malignancies reported in this study included nodular lymphoma (9/186 [5%]) and other mesenchymal tumors (8 cases [4%] for histiocytic sarcoma, nonlymphomatous nonangiomatous sarcoma, stromal tumor, and spindle cell sarcoma) as categorized in Table 1. Nodular lymphoma included histopathologic diagnoses of nodular B-cell lymphoma, marginal zone lymphoma, and indolent type lymphoma. Four of the splenic nodular lymphoma cases in our study had follow-up dates available, and all four were alive >12 mo postoperatively. Although this is a small number, this data supports a favorable prognosis with splenectomy alone for nodular splenic lymphoma as reported in a previous study.²¹ In this larger study, 29 dogs with splenic marginal zone lymphoma treated with splenectomy alone resulted in an MST of 383 days overall.²¹ If immediate perioperative deaths were excluded, MST was reported as 571 days.²¹

In the current study, we categorized splenic hematoma, nodular extramedullary hematopoiesis, and lymphoid (or similar) hyperplasia together in the nonneoplastic group of nodular splenic lesions because of the theoretical excellent expected long-term outcome for dogs with these diagnoses. If nonneoplastic and neoplastic splenic lesions were present within individual spleens, the most malignant lesion was used to categorize these cases. The MST of nonneoplastic nodular splenic lesions in our group of small-breed dogs was 983 days (32.8 mo), which supports the theory of an expected long-term survival. A recent study showed that the presence of lymphoid hyperplasia might support a diagnosis of splenic hematoma. In 27% (27/100) of splenic hematomas, lymphoid hyperplasia was noted.²² In contrast, no diagnosis (0/120) of splenic HSA contained lymphoid hyperplasia in dogs.²² In our study, 21/81 (26%) of splenic hematomas had concurrent nodular lymphoid hyperplasia lesions, which may be supportive of a diagnosis of a nonneoplastic lesion. Eighty-one dogs in the current study had a diagnosis of splenic hematoma. Forty of those dogs were alive at the 2 wk recheck and were then lost to follow-up. For the remaining 41 dogs with a diagnosis of splenic hematoma, 21 dogs were alive at follow-up. Of the remaining 20 dogs, 13 were euthanized for unrelated reasons, 5 had suspected postoperative thromboembolisms within 8 days from surgery, 1 case arrested 7 days postoperatively from complications relating to a concurrent pheochromocytoma, and 1 case was euthanized 14 days postoperatively as a result of the progression of a trigonal transitional cell carcinoma. The reported MST for splenic hematoma in our study was 686 days (23 mo), with 48% of the cases surviving past 12 mo postoperatively. A recent study showed excellent long-term outcome of 35 dogs diagnosed with splenic hematomas with an MST of 674 days.²³ In this same study, 4/35 (11%) were euthanized because of reported or clinical evidence of metastatic disease, which raises the concern of occult splenic HSA being misdiagnosed as splenic hematomas.²³ A previous study reviewing 125 histopathologic diagnoses of splenic hematomas reclassified 3 (2%) of those cases as HSA.²⁰ Although misdiagnoses are uncommon, additional sectioning may improve the accuracy of a diagnosis.

The current study found significant associations between hemoperitoneum and malignancy, as well as between hemoperitoneum and a diagnosis of HSA. Of the 115 hemoperitoneum cases with splenic lesions, 66% (76/115) were diagnosed with a malignant splenic lesion, with the majority being HSA (60/115 [52%]; Table 2). When evaluating all splenic HSA cases in this population of small-breed dogs, 60% (60/100) were presented with hemoperitoneum, which is comparable to other large-breed dog studies that reported 54–80% of splenic HSA presented with hemoperitoneum.^1,9-11,22 The recent small-breed splenomegaly study did not find a significant association between the presence of hemoperitoneum at the time of surgery and malignancy.¹⁷ This was likely because of the small study population. The risk of a diagnosis of a malignant splenic lesion was 2.6 times greater with the presence of hemoperitoneum at the time of surgery in our current small-breed dog population of 370 dogs. Based on our analysis, we accept our hypothesis that the presence of hemoperitoneum is positively associated with malignancy in small-breed dogs undergoing splenectomy.

Follow-up data was available for 213 (58%) cases in this study. The MST for small-breed dogs (<15 kg) with splenic HSA (n = 60) in this study was 93 days (3 mo), which is comparable to the previously reported MST of 1–3 mo in large-breed dogs.^1–4,20 Splenic HSA can be divided into three clinical stages, with stage I defined as a solitary nodule, stage II presence of concurrent hemoperitoneum, and stage III presence of gross metastatic disease.² Previous reports found a correlation between clinical stage of disease at time of splenectomy and MST.^2–4,19 For accurate clinical staging, a confirmation of peritoneal metastatic disease with histopathology is necessary, which was not consistently available in the present retrospective study. Additional studies with small-breed dogs are recommended to evaluate the correlation of clinical staging of splenic HSA and MST after splenectomy.

The use of adjuvant chemotherapy has been shown to improve survival times for splenic HSA in large-breed dogs, with MST of 3.4–8 mo.^3,5,6 Dose-intensified doxorubicin chemotherapy has been shown to improve reported survival times for dogs with stage I and II HSA, with MST of 257 and 210 days, respectively.¹⁹ No improvement in survival times were observed in cases with stage III HSA in a group of 20 large-breed dogs.¹⁹ The present study did not report statistical improvement in MST with the use of adjuvant chemotherapy in small-breed dogs with splenic HSA (Table 3). The small number of splenic hemangiosarcoma cases receiving chemotherapy (n = 19) likely contributes to the lack of statistical significance (type I error). Further research to evaluate the use of adjuvant chemotherapy in small-breed dogs may provide additional survival information.

For small-breed dogs (<15 kg) with available follow-up data in this study, the MST for malignant nodular splenic lesions was 143 days (4.7 mo), compared with an MST of 983 days (32.8 mo) for all other nodular splenic lesions. This data includes cases with and without the presence of hemoperitoneum. The current data suggests a correlation between the presence of hemoperitoneum and splenic malignancy (66%) in small-breed dogs. The risk of a diagnosis of a splenic malignancy is 2.6 times higher with the presence of hemoperitoneum. Previous studies in large-breed dogs undergoing splenectomy report that the presence of hemoperitoneum is considered a poor prognostic indicator.^1,2,9,10 In a recent report of splenectomized large-breed dogs for incidentally found nonruptured splenic nodules, the MST for malignant lesions was 110 days and for benign lesions was 436 days.¹³ As such, we would expect the MST for dogs with hemoperitoneum to be less (127 days) compared with the nonhemoperitoneum cases (481 days) as supported by the survival analysis in this study. These results are similar to the recent small-breed study in which 23 small-breed dogs who presented with hemoperitoneum had an MST of 123 days (range, 0–1756).¹⁷ The MST in small-breed dogs with splenic lesions and concurrent hemoperitoneum reported in the current study is longer than other previous studies of large-breed dogs, which were 50 and 57 days.^1,10

There are several limitations to the current study. Given the retrospective nature of the study, inherent deficiencies exist with regard to incompleteness of medical records and follow-up data. Data concerning splenic lesion size and amount of peritoneal hemorrhage may have provided additional associations with malignancy. Because of the small number of dogs receiving chemotherapy, we were unable to provide support regarding the benefit of postoperative adjuvant chemotherapy to improve survival time with malignant lesions. Survival data was obtained by review of the medical records. The survival analysis was performed using dogs with known dates of death and those alive at least 2 mo postoperatively (censored data). Clinical staging of abdominal disease may have provided additional information with regard to survival data in this study.

Conclusion

In summary, we noted several risk factors and associations with nodular splenic lesions in small-breed dogs (<15 kg). Our small-breed dog study supports many factors previously reported in large-breed dog studies, including MST of splenic HSA, correlation of hemoperitoneum with malignancy, and histopathologic distribution of nodular splenic lesions. We conclude that dogs with splenic lesions >1 cm weighing <15 kg undergoing splenectomy have an ∼50/50 (56%) chance of splenic neoplasia and a 1 in 4 (27%) chance of splenic HSA. Beagles and small-breed terriers were more likely to have malignant splenic lesions, and small-breed terriers were more likely to present with hemoperitoneum. Small-breed dogs who present with hemoperitoneum are 2.6 times more likely to have a diagnosis of a malignant splenic lesion. Malignant nodular splenic lesions had a strong correlation to poor survival. Additional studies would be beneficial to provide benefit of adjuvant chemotherapy, effect of clinical staging with MST, and associations of nodular size with malignancy in small-breed dogs undergoing splenectomy.