Primary Peritoneal Perivesical Hemangiosarcoma in 16 Dogs (2020–2024) Authors: Jose Roca, Xavier Escoda, Miguel García‐de la Virgen, Ignasi Rodríguez Published: 2026 ABSTRACT ABSTRACTBackgroundCanine hemangiosarcoma (HSA) commonly arises in the spleen, heart, and skin, whereas primary peritoneal presentations are rare and poorly characterized. A perivesical peritoneal location without involvement of adjacent organs has not been previously well described.ObjectiveThis study aims to characterize the clinical presentation, diagnostic features, treatment, and outcomes of dogs with primary peritoneal perivesical HSA.AnimalsThis study included 16 client‐owned dogs diagnosed with peritoneal perivesical HSA at 2 referral centres between 2020 and 2024.MethodsMethods consisted of retrospective multicentre case series. Medical records were reviewed for clinical signs, laboratory abnormalities, imaging findings, diagnostic procedures, treatment, metastatic status, and outcome measures, including progression‐free survival (PFS) and disease‐specific survival time (DSST).ResultsCommon presenting signs included lethargy, hyporexia, pale mucous membranes, and abdominal distension. Hemoperitoneum was identified in 50% of the dogs. Frequent laboratory abnormalities included anaemia, thrombocytopenia, neutrophilia, and azotaemia. All dogs had a mass adjacent to the urinary bladder without evidence of origin from surrounding organs. Diagnosis was achieved by cytology or histopathology. Metastatic disease was suspected at diagnosis in 56.25% of the dogs, most commonly affecting the lungs, and the overall metastatic rate was 87.5%. Treatment protocols included chemotherapy alone (n = 8), surgery with adjuvant chemotherapy (n = 7), and neoadjuvant chemotherapy followed by surgery (n = 1). All surgical procedures consisted of debulking surgery. Chemotherapy alone resulted in a 55.55% partial response rate. Mean PFS and DSST were both 125 days. Combination therapy, response to chemotherapy, and rescue treatment were significantly associated with improved DSST.ConclusionsPrimary peritoneal perivesical HSA may represent a distinct anatomical presentation of canine HSA. Multimodal treatment strategies may improve clinical outcomes and survival in affected dogs. Introduction Hemangiosarcoma (HSA) is a malignant and highly metastatic neoplasm that has historically been considered to originate from the vascular endothelium (Griffin et al. 2021). However, multiple studies suggest that it may instead arise from pluripotent progenitor cells in the bone marrow (J. H. Kim et al. 2015). Spleen is the most frequently affected organ (Vail 2020), but other common sites of involvement include the right atrium and auricle of the heart, as well as the dermis, subcutaneous tissue, and liver (Srebernik and Appleby 1991). Although less common, HSA has also been reported in other locations as the kidney or retroperitoneum (Brown et al. 1985,Mendez et al. 2020, Liptak et al. 2004, Martinez and Schulman 1988, Pirkey‐Ehrhart et al. 1995, Dennis et al. 2006, Tanaka et al. 2019, Haeussler et al. 2011, Paek et al. 2015, De Nardi et al. 2023, de la Fuente et al. 2014, Guinan et al. 2012, Burton et al. 2014). Advanced imaging techniques currently play a key role in tumour staging and the characterization of primary masses. Previous studies have aimed to describe the tomographic features of HSA (Tanaka et al. 2019, Fukuda et al. 2014, Mattolini et al. 2024), characterizing these lesions as large masses, with weak and heterogeneously contrast‐enhancing, particularly at their periphery (splash sign), often displaying areas of intense enhancement consistent with vascular channels filled with contrast medium. Surgery followed by adjuvant chemotherapy—most commonly with doxorubicin‐based protocols—remains the standard of care when metastases are not present (Griffin et al. 2021, Vail 2020). However, responses and improved outcomes have also been observed with other conventional chemotherapeutic agents or alternative therapies as immunotherapy, even in the setting of metastatic disease (Faulhaber et al. 2021, Kahn et al. 2013, Musser et al. 2022, Labé et al. 2025, Hillers et al. 2007, Vail et al. 1995). Regardless of treatment, the prognosis remains guarded (Griffin et al. 2021). Primary location has been associated with prognosis and varying median survival times (MST), with better outcomes observed in dermal HSA (MST 780–987 days) compared to muscular, subcutaneous, or visceral forms (MST < 1 year). Among abdominal visceral HSAs, primary renal HSA may have a better prognosis than primary splenic HSA. In contrast, retroperitoneal HSA is associated with the poorest outcomes (MST < 6 months) (Liptak et al. 2004, Szivek et al. 2012, Locke and Barber 2006, Walter et al. 2024). Regarding primary malignant peritoneal tumours, these are rare in veterinary medicine, with the majority of reported cases being soft tissue sarcomas (Vail 2020). Recently, 12 cases of HSA located within the pelvic cavity have been described, typically appearing unrelated to adjacent organs, although some cases were localized adjacent to the urinary bladder. In addition, a retrospective study of retroperitoneal HSA included several cases with lesions in the same location (Bourne et al. 2022, Ichimata et al. 2023). The primary objective of this case series is to report the clinical presentation, treatment, and outcome of 16 dogs with primary peritoneal perivesical HSA, and to explore whether it may represent a distinct anatomical presentation of HSA with potentially different behaviour and prognosis. Materials and Methods Medical records of dogs diagnosed with or suspected of having primary HSA adjacent to the urinary bladder between 2020 and 2024 at two referral centres in Spain were retrospectively reviewed. A total of 16 dogs met the criteria to be evaluated for this retrospective case series. Information collected from the medical records included signalment, weight, date of diagnosis and treatment initiation, clinical signs at diagnosis (including the presence of hemoperitoneum), laboratory test results (including complete blood count [CBC] and serum biochemistry), method of diagnosis, imaging findings, presence of metastasis at diagnosis, treatment type, response to treatment, follow‐up restaging methods, date of first progression, rescue treatment, and date and cause of death. Clinical staging was performed based on the modified World Health Organization (WHO) staging criteria for canine HSA. Staging diagnostics consisted of a CBC, biochemistry panel, abdominal ultrasound (US), thoracic X‐ray, and/or abdominal and/or pre‐contrast and post‐contrast (venous phase) thoracic computed tomography (CT). Cases with a cytological diagnosis of sarcoma (suspected HSA) or a histopathological diagnosis of HSA were included. Cases in which cytology was inconclusive required histopathology to confirm the diagnosis of HSA. Information was retrieved from the original cytology and histopathology reports. Immunohistochemistry (IHC) with CD31 and/or FVIII reports were collected when performed at diagnosis. Radiographic, CT, and US findings, including follow‐up studies when available, were also recorded. Response to treatment was determined in all cases through follow‐up. Response was classified using the Veterinary Cooperative Oncology Group (VCOG v1.1) RECIST criteria: complete response (CR) was defined as the disappearance of all target lesions; partial response (PR) as a ≥30% decrease in the sum of the longest diameters (SLD) compared to baseline; progressive disease (PD) as a ≥20% increase in SLD compared to the last value recorded (nadir); and stable disease (SD) as neither sufficient shrinkage for PR nor sufficient increase for PD. A minimum response duration of 21 days was required to determine response (CR and PR), and dogs that died or were euthanized before 21 days were considered non‐responders. Overall response rate (ORR) was defined as proportion of patients whose cancer shrinks (PR) or disappears (CR) after chemotherapy. For dogs with a positive response following any form of local or systemic treatment, outcomes were assessed by progression‐free survival (PFS), disease‐specific survival time (DSST), and overall survival time (OST). PFS was calculated (in days) from the date of treatment initiation to the date of local tumour progression/recurrence, new presence of hemoperitoneum, metastasis, or death related to the disease. DSST was defined as the survival time (in days) from the date of treatment initiation to the date of death related to the disease. OST was calculated (in days) from the date of treatment initiation to the date of death from any cause. Follow‐up information was obtained through phone calls or mail to referring veterinarians and owners when necessary. All data were recorded using Microsoft Excel 2024. IBM SPSS Statistics (version 26) was used to calculate means, medians, and generate Kaplan–Meier survival curves. The log‐rank test was applied to compare survival times between the groups, and Fisher's exact test was used to identify factors associated with chemotherapy treatment response. Hazard ratios were not calculated because all dogs died during the study. Baseline descriptive statistics are presented as means for normally distributed variables and medians for variables that were not normally distributed, based on the Shapiro–Wilk test. The groups used for statistical analysis to compare survival and chemotherapy treatment response were presence or absence at diagnosis of analytical alterations such as anaemia, thrombocytopenia, monocytosis, azotaemia, elevations in liver enzymes, metastasis, and hemoperitoneum; treatment modality: surgery plus chemotherapy or chemotherapy alone; chemotherapy groups: doxorubicin only, doxorubicin plus other chemotherapeutics, or non‐doxorubicin chemotherapy; rescue treatment after progression: performed, not performed, and number of rescues when performed; mitotic count (MC) (mitoses/2.37mm2): <11, >10; nuclear or cellular atypia (pronounced pleomorphism, anisocytosis, anisokaryosis, multinucleation, coarse chromatin, or prominent or multiple nucleoli): presence or absence; and RECIST criteria in the chemotherapy‐only group: CR, PR, SD, and PD. Tumour size was not considered part of the analysis. p values of ≤0.05 were considered statistically significant. Results The study population included 16 dogs, with a mean age at diagnosis of 10.5 years (range: 5.1–14.1 years). The mean body weight was 27.65 kg, ranging from 7.4 to 54.5 kg. Of the 16 dogs, 12 (75%) were males (seven of which were castrated) and four (25%) were females (all of them were spayed). The study included 7 (43.75%) mixed‐breed dogs, four (25%) Boxers, and one (6.25%) each of the following: Golden Retriever, Beagle, Akita Inu, Cocker English Spaniel, and Husky. Of the 16 dogs, 8 (50%) had hemoperitoneum at presentation, which was detected using abdominal Point‐of‐Care Ultrasound (A‐POCUS) in all cases. Presenting complaints are included in Table 1. TABLE 1Signs and analytic findings summary.Signs (%)Analytic findings (%)Lethargy (56.2)Anaemia (56.2)Hyporexia (31.25)Thrombocytopenia (56.2)Pale mucoses (31.25)Neutrophilia (50)Abdominal distension (31.25)Eosinopenia (43.75)Urinary incontinence (18.75)Monocytosis (25)Abdominal pain (12.5)Lymphopenia (18.75)Fever (12.5)Azotaemia (37.5)Vomiting and/or diarrhoea (12.5)Increased ALT (31.25)Anorexia (12.5)Increased ALKP (25)Haematuria (6.25)Hypokalaemia (12.5)Collapse (6.25)Hyperkalaemia (6.25)Palpable abdominal mass (6.25)Hypoalbuminemia (6.25)John Wiley & Sons, Ltd. Signs and analytic findings summary. CBC and a biochemistry panel were performed in all dogs (100%) at the time of diagnosis, with only three presented with results within normal limits (18.75%). Laboratory abnormalities in the remaining cases are summarized in Table 1. All dogs underwent some form of imaging. Initial staging methods are detailed in Table 2. In all cases, a perivesical mass (adjacent to the urinary bladder), displacing it laterally, with preserved fat planes and no infiltration was observed. In these cases, a peritoneal origin was presumed, as the masses and bleeding were unrelated to abdominal organs or the retroperitoneum, and a urinary bladder origin was unlikely given the clearly defined bladder wall (Figure 1). On CT, the masses demonstrated poorly defined morphology, with irregular shape and indistinct margins toward surrounding fat and heterogeneous contrast enhancement predominantly in the periphery (splash sign). No vascular connections were identified in any case. Their main anatomical boundaries included the urinary bladder and transversus abdominis muscle laterally, the rectus abdominis muscle ventrally, and the descending colon or vascular structures such as the caudal vena cava, descending aorta, or iliac arteries dorsally. For dogs without CT, the lesion was described on US as heterogeneously echogenic and multilobulated with no apparent origin in adjacent organs. TABLE 2Characteristics of the 16 total dogs in the study, including staging method, metastasis at presentation, treatment, disease progression, rescue treatments and DSST.CaseInitial staging methodMetastasis at presentationTreatmentDisease progressionRescue treatments DSST (days) 1Thorax and abdomen CTLungNA DOX 30 mg/m2 IV every 2 weeks (x2 sessions); SD; Adj DOX 30 mg/m2 IV every 3 weeksMetastasis PDCARB 270 mg/m2 IV every 3 weeks; DOX 30 mg/m2 IV every 3 weeks3952Abdomen US + thorax and abdomen CTLungDOX 30 mg/m2 IV every 3 weeks + CTX 250 mg/m2 PO day 10Metastasis PDNo rescue753Abdomen US + thorax and abdomen CTNoSD; Adj DOX 30 mg/m2 IV every 3 weeksPeritoneal and spleen metastasisDOX 30 mg/m2 IV every 3 weeks5864Thorax and abdomen CTLiverDOX 27 mg/m2 IV every 2 weeksHemoperitoneumNo rescue105Thorax and abdomen CTMuscleDOX 27 mg/m2 IV every 2 weeksHemoperitoneumNo rescue196Abdomen US + Thorax X‐rays + Thorax and abdomen CTNoDOX 27 mg/m2 IV every 2 weeksHemoperitoneumNo rescue387Thorax and abdomen CTNoDOX 27 mg/m2 IV every 3 weeksEpileptic seizureNo rescue688Thorax and abdomen CTNoSD; Adj MTX 4 mg/m2 IV every 3 weeks + CTX 200 mg/m2 PO day10Head pressing and peritoneal metastasisVBL 2 mg/m2 IV every 2 weeks; weekly DOX 30 mg/m2 IV+ CTX 200 mg/m2 PO + VCR 0.6 mg/m2 IV; IMA 10 mg/Kg PO SID + LOS 10 mg/kg BID4839Abdomen US + Thorax and abdomen CTNoSD; Adj DOX 1 mg/kg IV every 3 weeks + CTX 250 mg/m2 PO day 10Lung metastasisWeekly DOX 1 mg/kg IV + CTX 200 mg/m2 PO + VCR 0.6 mg/m2 IV; CHL 4 mg/m2 PO every 48 h + LOS 10 mg/kg BID43910Abdomen US + Thorax X‐rays + Thorax and abdomen CTNoSD; Adj DOX 30 mg/m2 IV every 3 weeks + CTX 250 mg/m2 PO day 10Peritoneal metastasisNo rescue12511Abdomen US + Thorax X‐rays + Thorax and abdomen CTLungWeekly DOX 1 mg/kg IV + CTX 200 mg/m2 PO + VCR 0.6 mg/m2 IVMetastasis PDVBL 1.6 mg/m2 IV every week27812Thorax X‐Rays + abdomen USNoWeekly DOX 30 mg/m2 IV + CTX 200 mg/m2 PO + VCR 0.6 mg/m2 IVNo progressionNo rescueNon‐tumour related death13Thorax X‐Rays + abdomen and thorax CTMuscleSD; Adj DOX 30 mg/m2 IV every 3 weeks + CTX 50 mg/m2 PO EODPeritoneal and lung metastasisVBL 1.6 mg/m2 IV every 2 weeks18614Thorax X‐rays + abdomen USSpleenSD; Adj DOX 1 mg/kg IV every 3 weeks + CTX 250 mg/m2 PO day 10Peritoneal metastasis and epileptic seizuresMTX 4 mg/m2 IV every 3 weeks30115Thorax and abdomen CTLeft adrenal gland, lung, and vertebraeDOX 1 mg/kg IV every 2 weeksHemoperitoneumNo rescue8516Thorax and abdomen CTLung and peritoneumSD + peritoneal metastectomy; Adj weekly DOX 30 mg/m2 IV + CTX 200 mg/m2 PO + VCR 0.6 mg/m2 IVHead pressing, ataxia and no facial sensibilityNo rescue111Abbreviations: Adj, adjuvant; BID, twice a day; CARB, carboplatin; CHL, chlorambucil; CT, computed tomography; CTX, cyclophosphamide; DOX, doxorubicin; EOD, every other day; IMA, imatinib; IV, intravenous; LOS, losartan; MTX, mitoxantrone; NA, neoadjuvant; PD, progressive disease; PO, oral; SID, once a day; SD, surgery debulking; US, ultrasound; VBL, vinblastine; VCR, vincristine.John Wiley & Sons, Ltd. Characteristics of the 16 total dogs in the study, including staging method, metastasis at presentation, treatment, disease progression, rescue treatments and DSST. DSST (days) Abbreviations: Adj, adjuvant; BID, twice a day; CARB, carboplatin; CHL, chlorambucil; CT, computed tomography; CTX, cyclophosphamide; DOX, doxorubicin; EOD, every other day; IMA, imatinib; IV, intravenous; LOS, losartan; MTX, mitoxantrone; NA, neoadjuvant; PD, progressive disease; PO, oral; SID, once a day; SD, surgery debulking; US, ultrasound; VBL, vinblastine; VCR, vincristine. CT‐scan images of primary perivesical peritoneal hemangiosarcoma (arrows) adjacent to the urinary bladder (UB) and with SPLASH sign (red asterisk). (A) Transversal image of case 3. (B) Transversal image of case 4. (C) Transversal image of case 5. (D) Transversal image of case 6. (E) Transversal image of case 7. (F) Transversal Image of case 11. Metastases were suspected at diagnosis in nine cases (56.25%), most commonly in the lungs (55%). Other sites included muscle (22%), and one case each (11%) in the liver, spleen, vertebrae, peritoneum, and left adrenal gland. Fine‐needle aspiration (FNA) of the primary mass was performed as the sole sampling method in four cases (25%), histopathology alone in eight cases (50%), and both FNA and histopathology in the remaining four cases (25%). An incisional biopsy was obtained in five of 12 cases in which histopathological examination was performed. No complications were reported during the sample collection process. Of the eight animals where cytology was carried out, descriptions were available in six cases, all of which were compatible with malignant mesenchymal neoplasia, most likely HSA. In most reports, the samples had low‐to‐moderate cellularity with abundant blood spread throughout the field. The main population was described as mesenchymal, basophilic cells with indistinct borders, intracytoplasmic colourless vacuoles and evidence of erythrophagia. The most representative samples were obtained when FNA was performed in the periphery of the mass. Histopathology reports were available in all cases where performed, with a definitive diagnosis of HSA in 10 cases. The MC was reported in nine cases, with a mean of 10.6 mitoses/2.37 mm2 (range: 0–23). Nuclear or cellular atypia were present in 10 of the 11 histopathology reports. In the two cases where a definitive diagnosis was not obtained, IHC for FVIII was requested in one case, and CD31 in the other, with positive and negative expression, respectively. In the CD31‐negative case, a second opinion by a board‐certified pathologist was obtained, yielding a result supportive of HSA. None of the dogs were euthanized at diagnosis. The median time from first consultation to initiation of treatment was 2 days (range: 0–20 days). Eight cases (50%) received chemotherapy alone. Mass debulking surgery followed by adjuvant chemotherapy was performed in seven cases (43.75%). The remaining dog received neoadjuvant doxorubicin every 3 weeks, followed by surgical debulking after two chemotherapy cycles. Among the eight dogs that underwent surgery, metastasis was suspected at diagnosis in four cases. Treatment approaching is summarized in Table 2. Imaging methods to evaluate response according to RECIST criteria included thoracic X‐ray and abdominal US in eight cases (50%), abdominal US in four cases (25%), and thoracic and abdominal CT in four cases (25%). The ORR in dogs receiving chemotherapy as the sole treatment or in the neoadjuvant setting was 55.55% with CR, PR, SD, and PD achieved 0% (n = 0), 55.55% (n = 5), 22.22% (n = 2), and 22.22% (n = 2), respectively. Progressive disease was observed in all cases except one (6.25%), in which the dog died from an unrelated condition (dilated cardiomyopathy). In five cases (31.25%), PD was detected during routine follow‐up staging, with metastatic progression involving the peritoneum (2 cases) and one case each involving the peritoneum and lungs, lungs alone, and spleen. Three cases (18.75%) developed respiratory signs (coughing and respiratory distress), and pulmonary metastatic progression was confirmed in all three. Additionally, three cases (18.75%) developed neurological signs (head pressing, ataxia, loss of facial sensation, and seizures). Although advanced imaging was not performed to confirm central nervous system (CNS) involvement in these cases, the neurological signs were presumed to be associated with disease progression. The overall metastatic rate was 87.5%. Finally, hemoperitoneum was confirmed in four cases (25%) following lethargy and collapse. The overall median PFS was 125 days (range: 10–340 days). Once progression was suspected or confirmed, euthanasia was performed in eight cases (50%). In contrast, rescue chemotherapy treatment was chosen in the seven cases (43.75%). Rescues summaries are detailed in Table 2. No dogs were still alive at the time of writing this study. The median OST was 155 days (range: 10–586 days) for all dogs. Median DSST was 125 days (range: 10–565 days). Statistical analysis Kaplan–Meier tables are shown in Figure 2. Among dogs treated with chemotherapy alone, the presence of hemoperitoneum was the only factor that reached statistical significance in RECIST. Specifically, 83.3% of the dogs with hemoperitoneum had PR, compared to 0% of the dogs without hemoperitoneum (p = 0.048). Factors showing a significant difference in PFS were the combination of surgery and chemotherapy as a therapeutic approach (81 days for chemotherapy only vs. 195 days when combined with surgery; p = 0.019) and the response to chemotherapy (in chemo‐only group) according to RECIST criteria (132 days for PR, 76 days for SD, and 14 days for PD). Factors showing a significant difference in DSST were the combination of surgery and chemotherapy as a therapeutic approach (81 days for chemotherapy only vs. 328 days when combined with surgery; p = 0.001), the response to chemotherapy (in chemo‐only group) according to RECIST criteria (196 days for PR, 76 days for SD, and 14 days for PD; no cases showed CR with chemotherapy alone; p = 0.013), and switching to rescue therapies upon the detection of progressive disease (381 days with rescue therapy vs. 66 days without; p = 0.000). Survival analysis Kaplan–Meier (KM) with significant survival factors. (A) DSST KM. (B) DSST KM for RECIST chemotherapy groups (p = 0.013). (C) DSST KM for rescued or no rescued groups (p = 0.000). (D) DSST KM for dogs treated with surgery and chemotherapy or one method treatment (p = 0.01). Discussion This study reports the presenting complaints, treatment modalities, outcomes, and prognostic factors of 16 dogs diagnosed with primary peritoneal HSA adjacent to the urinary bladder from two Spanish referral institutions over a period of 4 years. To the authors’ knowledge, peritoneal HSA has only been reported in the literature as peritoneal metastasis and not as a primary location (Brown et al. 1985, Mendez et al. 2020, Liptak et al. 2004, Martinez and Schulman 1988, Pirkey‐Ehrhart et al. 1995, Dennis et al. 2006, Tanaka et al. 2019, Haeussler et al. 2011, Paek et al. 2015, De Nardi et al. 2023, de la Fuente et al. 2014, Guinan et al. 2012, Burton et al. 2014). Additionally, a few cases of primary HSA near the urinary bladder have been described in a case series of 12 intrapelvic HSA cases and in a retrospective study of retroperitoneal HSA (Bourne et al. 2022, Ichimata et al. 2023). However, the retroperitoneal space is anatomically defined as the region of the abdomen located ventral to the spine and dorsal to the intestines, lying outside the peritoneal cavity, and includes organs such as the adrenal glands and kidneys, whereas the urinary bladder is considered an intraperitoneal pelvic organ (O'Brien and Barr 2009, Barr and Gaschen 2011). Moreover, haemorrhage associated with retroperitoneal organs may remain confined to the retroperitoneal space, resulting in haemoretroperitoneum, in contrast to hemoperitoneum, which occurs when bleeding originates from a peritoneal structure. To date, no study has identified this presentation as a potentially distinct entity arising from the peritoneal vasculature adjacent to the urinary bladder, rather than from the retroperitoneal space or from organs within the abdominal and pelvic cavities (Figure 1). Patient characteristics in this case series share similarities with findings reported in the visceral HSA literature. For instance, HSA has been associated with older and large‐breed dogs. In our study, the mean age at diagnosis was 10.5 years, and the mean body weight was 27.65 kg. Additionally, a male predisposition has been suggested for visceral HSA; in this series, 75% of the dogs were male. The most commonly represented breeds in our study were mixed‐breed dogs and Boxers, whereas German Shepherds and Golden Retrievers are more frequently reported (Vail 2020, Srebernik and Appleby 1991, Brown et al. 1985). Presenting complaints in visceral HSA are typically associated with hemoperitoneum, intratumoral haemorrhage, or blood loss into other body cavities. The initial signs include weakness, lethargy, pallor, anorexia, and abdominal distension (Aronsohn et al. 2009). These findings were also the most observed at the time of initial consultation in our study. As previously mentioned, due to the pre‐test probability of disease and the nature of the clinical signs, an A‐POCUS was performed in all cases, revealing abdominal bleeding in 50% of the cases. The study involving 12 cases of intrapelvic primary HSA reported hemoperitoneum in only one case, in which the primary mass was located cranial to the pelvic inlet (Bourne et al. 2022), as well as in retroperitoneal HSA study (Ichimata et al. 2023), where three cases were present with haemoperitoneum, a significantly lower frequency than observed in our population. One possible explanation for this discrepancy is the intrapelvic or retroperitoneal location of the lesions. The smaller volume of these cavities compared to the abdominal cavity may lead to earlier onset of clinical signs due to space constraints, prompting earlier detection of the mass before spontaneous bleeding occurs. In those studies, the most common presenting complaints were anorexia and stranguria, followed by dyschezia and pelvic limb pain (Bourne et al. 2022, Ichimata et al. 2023). Clinicopathological findings in our population were consistent with previous reports, with anaemia, thrombocytopenia, and neutrophilia being the most common abnormalities (Childress 2012). These findings, likely resulting from tumour‐associated bleeding or paraneoplastic syndromes, were observed in 56.25% of the cases for both anaemia and thrombocytopenia, while neutrophilia was present in half of the cases in this study. Thrombocytopenia and neutrophilia have been identified as negative prognostic indicators in dogs with splenic HSA (S. E. Kim et al. 2007). However, in the present study, statistical analysis revealed no significant association between these abnormalities and either survival time or response to chemotherapy. Due to the retrospective and multidisciplinary nature of this study, a standardized staging protocol was not implemented. Imaging modalities varied depending on clinical judgment and owner preferences. In some cases, the available imaging reports were limited. All dogs underwent staging using one or more imaging modalities and, in every case, a heterogeneous mass located in the caudal abdomen and adjacent to the urinary bladder was identified. In our study, the primary masses exhibited heterogeneous contrast enhancement (splash sign), with a peripheral distribution and multiple focal areas of intense uptake as previously reported in other studies (Tanaka et al. 2019, Fukuda et al. 2014, Mattolini et al. 2024). This imaging pattern may reflect intratumoral haemorrhage and the chaotic arrangement of blood vessels within the tumour. As previously mentioned, HSA has a high metastatic rate, reaching up to 100% of cases, through hematogenous spread or intra‐abdominal dissemination following tumour rupture (Vail 2020). In this study, the metastatic rate at the time of diagnosis was 56.25%. Among these cases, 55.5% showed evidence of pulmonary involvement. Metastatic progression during follow‐up was observed in 11 cases (68.75%), most commonly affecting the lungs, peritoneum, and spleen. Among these 11 cases, CNS metastasis was suspected in three dogs due to the development of neurological signs. Although advanced imaging was not performed to confirm CNS involvement, HSA is recognized as one of the tumours with the highest propensity for brain metastasis (Snyder et al. 2008). Lesions that were not confirmed by histopathology were still considered metastatic to avoid underestimating the metastatic potential of peritoneal hemangiosarcoma. Given the high prevalence of primary hepatic or splenic HSA and their high metastatic rate (>60%) (Queiroga et al. 2026), in cases where lesions were identified in these organs, a primary peritoneal origin was assumed based on the characteristic and unique presentation described in this study, as this is not the most common site or presentation for HSA metastasis. The similarity in metastatic rates at diagnosis between our cases and those reported in other visceral HSA may be explained by the fact that diagnosis is often triggered by the occurrence of hemoperitoneum or other clinical signs. This delay between disease onset and diagnosis may allow for metastatic dissemination. Due to the retrospective and multidisciplinary nature of the study, CT scans were not performed in two cases; therefore, the metastatic rate could be underestimated. The overall metastatic rate observed during the course of the disease was 87.5%, which is consistent with previous reports (Vail 2020). The most reliable technique for the definitive diagnosis of HSA is histopathology. FNA of suspected HSA lesions often has a low diagnostic yield, primarily due to blood contamination (Bertazzolo et al. 2005); however, cytological characteristics of HSA neoplastic cells are described (Barger et al. 2012). In this case series, six dogs received a cytologic diagnosis consistent with malignant mesenchymal neoplasia, most likely HSA. In the authors' experience, the most representative cytologic samples were obtained when the FNA was targeted at the periphery of the lesion. This may be because malignant cells tend to be more densely aggregated and structurally intact within the tumour capsule. In contrast, central areas of the lesion often contain intratumoral haemorrhage and necrosis. When histopathology was performed, the mean MC was 10.6 mitoses/2.37 mm2 and nuclear or cellular atypias were frequent (10 cases). MC and nuclear pleomorphism have been proposed as prognostic indicators for HSA at various anatomical sites (S. E. Kim et al. 2007, Schultheiss 2004, Sorenmo et al. 2000, Moore et al. 2017). However, statistical analyses revealed no significant association between either MC or the presence of atypia (nuclear or cellular) and survival time or response to chemotherapy. In one case, CD31 immunohistochemical staining yielded a negative result. A 2017 study reported that loss of CD31 and FVIII expression may occur in atypical HSA cells, thereby complicating immunohistochemical diagnosis (Maharani et al. 2018). Considering the clinical presentation with distant muscular metastases, biopsy findings consistent with a sarcoma exhibiting marked cellular atypia, a second‐opinion biopsy strongly supportive of HSA, and the clinical course progressing with hemoperitoneum, this case was included in the study. Due to the multidisciplinary and retrospective nature of the study, treatment and restaging protocols were not uniform or standardized. ORR for chemotherapy alone was 55.55%, and none of them achieved a CR. Hemoperitoneum was the only factor that showed statistical significance: 83.3% of the dogs with hemoperitoneum achieved a PR, compared to 0% of those without. The better response in dogs with hemoperitoneum could be explained by the possibility that intratumoral bleeding initially overestimated the lesion size. Once chemotherapy was administered, the bleeding ceased, allowing for redistribution of abdominal and intratumoral blood, which may have contributed to a reduction in the apparent size of the primary mass. The median PFS was 125 days across all treatments. Significantly longer PFS was observed with combined surgery and chemotherapy. Among the eight cases treated with chemotherapy alone, better RECIST responses were associated with longer PFS. The favourable outcomes in dogs who underwent surgery may be attributed to selection bias, as surgical intervention was pursued mainly when metastasis was not evident at diagnosis. Furthermore, all dogs received chemotherapy either as a sole treatment or as adjuvant therapy, and no control group was available in which surgery was performed without chemotherapy. Therefore, it was not possible to assess the independent effect of surgery on outcome. However, previous studies have demonstrated improved outcomes combining surgery with doxorubicin‐based chemotherapy compared to surgery alone (Sorenmo et al. 2000). These PFS are consistent with those reported in studies of HSA at other anatomical locations (Griffin et al. 2021, Vail 2020, De Nardi et al. 2023, Bourne et al. 2022, Ichimata et al. 2023). No dogs were alive at the end of the study period. The median DSST and OST were 125 and 155 days, respectively. These outcomes are comparable to previously reported survival times for visceral HSA (Griffin et al. 2021, Vail 2020, De Nardi et al. 2023), including the intrapelvic and retroperitoneal studies, despite only a few cases in those studies involved masses adjacent to the urinary bladder (Bourne et al. 2022, Ichimata et al. 2023). Statistically significant factors associated with increased survival in the present study included (1) combination of surgery and chemotherapy, (2) a favourable response to chemotherapy (RECIST criteria) when performed as the sole treatment, and (3) switching to rescue therapy upon detection of progression. Improved survival with combined treatment may be influenced by a selection bias, as previously discussed. Moreover, it is reasonable that dogs with longer PFS also achieve longer OST. But when evaluating whether metastases at diagnosis influenced outcomes, no significant differences were found, but the prognostic value of clinical stage appears to vary among studies (Chu et al. 2023, Wendelburg et al. 2015, Masyr et al. 2021). In addition, owners whose dogs responded well to initial treatment may have been more inclined to pursue rescue therapies, also associated with improved outcomes. While most induction chemotherapy protocols were based on doxorubicin, other agents have shown efficacy against HSA (Faulhaber et al. 2021, Pasquier et al. 2016). Therefore, transitioning to a chemotherapeutic agent with a different mechanism of action is advisable upon disease progression. Finally, response to chemotherapy was a prognostic factor in the non‐surgical group. Tumour resistance to chemotherapy is a challenging and unpredictable variable in cancer treatment (Vail 2020), and as such, dogs who exhibited better responses to chemotherapy experienced longer PFS and, consequently, extended survival times. Limitations This retrospective study is subject to several inherent limitations. First, the lack of standardized diagnostic and staging protocols across the two participating referral centres led to inconsistencies in imaging modalities, cytologic interpretation, and restaging evaluations. Incomplete histopathological or immunohistochemical data in some cases may have limited diagnostic precision; however, cases with a cytological diagnosis of sarcoma were included when characteristics of neoplastic cells were consistent with HSA and this highly characteristic clinical presentation was present. Furthermore, the relatively small sample size (n = 16) restricts the statistical power of the study and limits the generalizability of the findings. Treatment decisions were made based on clinician discretion and owner preferences rather than uniform protocols, resulting in heterogeneity in therapeutic approaches and response assessments. Additionally, the possibility of selection bias must be acknowledged, particularly in the surgical group, as candidates for resection were likely those without overt metastases. Finally, due to the absence of a control group not receiving chemotherapy, the independent contribution of surgery to survival outcomes could not be accurately assessed. Conclusion This study identifies primary peritoneal hemangiosarcoma localized adjacent to the urinary bladder as a potentially distinct clinical entity in dogs. Despite aggressive multimodal treatment, overall prognosis remains guarded. However, the combination of surgical resection and doxorubicin‐based chemotherapy was associated with significantly prolonged progression‐free and disease‐specific survival compared to chemotherapy alone. Moreover, response to chemotherapy—as measured by RECIST criteria—was a critical prognostic indicator, particularly in non‐surgical cases. The study also highlights the potential benefit of rescue chemotherapy following disease progression. These findings underscore the importance of early diagnosis, comprehensive staging, and individualized treatment planning to optimize clinical outcomes in canine dogs with this unfrequented neoplastic presentation. However, more studies are needed to further characterize this possible entity. Author Contributions Xavier Escoda: supervision, data curation. Ignasi Rodríguez: supervision, data curation. Jose Roca: conceptualization, investigation, writing – original draft. Miguel García‐de la Virgen: data curation, supervision. Funding The authors have nothing to report. Ethics Statement This study did not require ethical approval. Conflicts of Interest The authors declare no conflicts of interest.