Comparisons between elastographic stiffness scores for benign versus malignant lymph nodes in dogs and cats

Lymph node evaluation is an important component of oncology staging in dogs and cats, however diagnosis of malignancy currently requires cytology or histopathology. Elastography is an ultrasound technique that allows estimation of tissue stiffness. In people, it has been shown to increase the sensitivity of sonography in detection of metastatic lymph nodes. The objective of this prospective, cross‐sectional study was to determine if elastographic stiffness differs for malignant versus benign canine and feline lymph nodes that were considered abnormal in gray‐scale ultrasound imaging. Animals scheduled for ultrasound‐guided fine needle aspirates of a lymph node at a single center were prospectively enrolled. Elastography was performed by a board‐certified veterinary radiologist (G.S.) prior to tissue sampling. Softness was scored qualitatively as 1–4 (1 = hard; 4 = soft) on an image depicted by a color scale (blue = hard, red = soft). Quantitative analysis was performed using custom‐made software. Fifty‐one lymph nodes were included in the evaluation, 21 had benign and 30 had neoplastic cytology. Benign lymph nodes were softer (median score of 2.5, range 1–4) than malignant lymph nodes (median = 2, range 1–3), and differences were significant in qualitative and quantitative analyses (P < 0.01). Of the lymph nodes with a score of 2 or below (hard), 19/25 (76%) were malignant, of the lymph nodes with a score over 2 (soft), 24/36 (67%) were benign. Findings indicated that elastographic stiffness scores differed between benign and malignant lymph node groups; however overlapping scores for the two groups limited the use of this technique for routine clinical diagnosis of malignancy.     

Cytochemical staining characteristics of lymph nodes from normal and lymphoma-affected dogs

Frozen sections and imprint smears were used to evaluate the presence and pattern of cytochemical staining reactions in the B- and T-cell regions of lymph nodes from normal dogs and dogs with lymphoma. Staining procedures evaluated included peroxidase (PER), Sudan black B (SBB), naphthol AS-D chloroacetate esterase (CAE), alpha-naphthyl butyrate esterase (NBE), acid phosphatase (ACP), and leukocyte alkaline phosphatase (LAP). In normal lymph nodes, macrophages and some lymphocytes within the interfollicular (T-cell) region and medulla stained positive with ACP and NBE. Smaller numbers of macrophages also occurred sporadically within the germinal follicles. Cells positive for PER, SBB, and CAE were scattered infrequently throughout all regions of the normal lymph node, consistent with granulocytes and mast cells. The LAP stained cells were predominantly and prominently located within the mantle zone of secondary follicles and to a much lesser extent within the germinal centers, compatible with B-cell lymphocytes derived from follicular center cells. Of the 12 dogs with lymphoma, 7 cases (4 immunoblastic, 2 large noncleaved, 1 small noncleaved) stained diffusely positive with LAP, 4 cases (all lymphoblastic) had numerous focally positive lymphocytes using ACP and NBE, and 1 case (immunoblastic) did not stain positive with any of the cytochemical reactions. Cytochemical staining of canine lymph nodes with NBE, ACP, and LAP proved useful in distinguishing between B- or T-cell regions and detecting different cell types of canine lymphoma.     

The sonographic evaluation of lymph nodes

Ultrasound can be used to detect and evaluate both normal and abnormal lymph nodes, as well as aid in biopsy sampling procedures, an important part of staging procedures in cancer patients. Several parameters can be evaluated using ultrasound; lymph node size, margins, echogenicity, echopattern (echotexture), acoustic transmission, presence and distribution of vascular flow, and vascular flow indices. The most diagnostically helpful include the short/long axis ratio of the lymph node, the pattern of distribution of the blood vessels within the lymph node, and to some extent the resistive and pulsatility indices. This review discusses the use of ultrasound for detecting, evaluating, and sampling peripheral, abdominal and thoracic lymph nodes in small animals.     

Accuracy of four ultrasonography techniques in predicting histopathological classification of canine mammary carcinomas

Due to the importance of presurgical, noninvasive, and accurate diagnostic tools in mammary carcinoma characterization, this prospective secondary observational cohort study was designed to evaluate and compare the diagnostic accuracy of B‐mode, Doppler, contrast enhancement ultrasonography, or acoustic radiation force impulse–elastography in identifying mammary carcinomas types with high degree of malignancy. A total of 246 mammary carcinomas from 141 female dogs were analyzed using B‐mode, Doppler, contrast enhancement ultrasonography, and acoustic radiation force impulse ultrasonography prior to their histopathological classification according to types (simple, complex, or special) and grade (I, II, or III). Qualitative and quantitative variables were compared between carcinoma types and grades by Fisher's or analysis of variance. Diagnostic performance was estimated by receiver‐operating characteristic analysis, using histopathological classification as a reference. Deformability (acoustic radiation force impulse) had a diagnostic specificity of 100% and sensitivity of 12% in identifying special carcinomas. A width:length ratio greater than 0.53 can be suggestive of special carcinoma, with 80% sensitivity and 76% specificity. Contrast wash‐in and peak enhancement times lower than 7.5 and 13.5 s, respectively, were indicative of complex carcinoma at 62% sensitivity and 60% specificity. Contrast wash‐in, peak enhancement, and wash‐out times greater than 6.5, 12.5, and 64.5 s, respectively; were indicative of grade II and III carcinoma at 68% sensitivity and 62% specificity. In conclusion, B‐mode ultrasonography, contrast enhancement ultrasonography, and acoustic radiation force impulse–elastography enabled the identification of some of the characteristics of high‐grade mammary carcinoma types and grades in female dogs with limited accuracy. The findings from this study may contribute to oncology research and clinical management canine patients.     

犬瘟热病毒自然感染犬淋巴组织中T、B细胞变化的研究

为了调查患犬瘟热病犬淋巴组织中T、B细胞变化的特点及淋巴细胞减少的发病机制,试验通过免疫组织化学的方法观察了T细胞(用CD3和CD45RO检测T细胞)、B细胞(用IgG、IgM抗血清检测B细胞)和犬瘟热病毒(抗犬瘟热病毒抗体)在病犬淋巴组织中的分布。结果表明:在淋巴组织中的淋巴细胞、淋巴小结中树突状细胞和巨噬细胞中均检出了抗病毒阳性反应细胞。在骨髓组织的前髓细胞中也发现抗病毒阳性反应细胞和嗜酸性胞浆内及核内包涵体的存在。与对照组相比,CD3和CD45RO阳性细胞主要存在于T细胞的分布域;但CD3和CD45RO阳性T细胞的数量较少。位于淋巴组织中的巨噬细胞有的被CD45RO染成阳性。在B细胞分布的区域中,IgG、IgM阳性细胞的数量明显减少;一些位于淋巴组织的浆细胞也被IgG或IgM染成阳性。在淋巴组织中淋巴细胞减少的顺序为:IgG阳性细胞减少最明显,其次为IgM和CD45RO阳性细胞,再次为CD3阳性细胞。依据试验结果,作者认为病犬淋巴组织中淋巴细胞减少主要是由B细胞缺乏所引起的;淋巴细胞的增殖能力减弱是引起淋巴组织中淋巴细胞减少的重要原因。     

Sonographic characteristics of presumptively normal main axillary and superficial cervical lymph nodes in dogs

Objective-To evaluate the B-mode and Doppler ultrasonographic appearance of presumptively normal main axillary and large superficial cervical lymph nodes (MALNs and SCLNs, respectively) in adult dogs. Animals-51 healthy adult dogs (data from 1 dog were not analyzed). Procedures-For each dog, weight, distance from the cranial aspect of the first sternebra to the caudal aspect of the left ischiatic tuberosity, and thoracic height and width at the level of the xiphoid process were recorded. Via B-mode and Doppler ultrasonography, echogenic characteristics, size in relation to body size and weight, and vascular supply of the MALNs and the SCLNs were evaluated (1 SCLN in 1 dog was not ultrasonographically visible). Results-Most MALNs were clearly margined, solitary, and ovoid; echopatterns were homogenous or cortical and hypo- to isoechoic, compared with surrounding soft tissues. Size measurements of MALNs correlated with dogs' body length, thoracic width and height, and body weight. Most SCLNs were clearly margined, fusiform, and hypoechoic (compared with surrounding soft tissues) with a cortical or homogenous echopattern. Size measurements of SCLNs correlated with dogs' body length, thoracic width and height, and body weight. In 50 of the 100 MALNs, an intranodal vascular supply was detected; in contrast, an intranodal vascular supply in SCLNs was detected infrequently. Conclusions and Clinical Relevance-Results indicated that, in dogs, anatomically separate lymph nodes have different echogenic and vascular characteristics; body size (skeletal length, height, and width), along with body weight, were correlated with sizes of presumptively normal MALNs and SCLNs.     

Indirect computed tomography lymphangiography with aqueous contrast for evaluation of sentinel lymph nodes in dogs with tumors of the head

Sentinel lymph node evaluation is widely used in human medicine to evaluate the first lymph node(s) to which a tumor drains. Sentinel lymph node biopsy allows avoidance of extensive lymphadenectomies in cases where the sentinel lymph node is negative for metastasis, thereby reducing patient morbidity. It has been shown that regional lymph nodes are not always the sentinel lymph node, thus identification and sampling of sentinel lymph nodes allows for more accurate staging, which is critical for treatment and prognostication in dogs with cancer. The objective of this prospective, pilot study was to determine if indirect computed tomography (CT) lymphangiography with aqueous contrast agent would successfully allow identification of sentinel lymph nodes in dogs with masses on the head. Eighteen dogs underwent CT lymphangiography. The sentinel lymph node was successfully identified within 3 min of contrast injection in 16 dogs (89%). Compression of lymphatic vessels from endotracheal tube ties and/or the patient's own body weight delayed or prevented identification of sentinel lymph nodes in two dogs (11%). Computed tomography lymphangiography with aqueous contrast can be used successfully to rapidly identify sentinel lymph nodes in dogs with masses on the head.     

Assessment of cardiac troponin I (cTnI) and tissue velocity imaging (TVI) in 14 dogs with malignant lymphoma undergoing chemotherapy treatment with doxorubicin

Doxorubicin has been shown to be cardiotoxic at high doses but is an efficacious chemotherapeutic agent in the treatment of canine lymphoma. Echocardiographic measurements and serum ultrasensitive cardiac troponin I (cTnI) levels were obtained before and after doxorubicin administration in 14 dogs diagnosed with lymphoma. The aim of this prospective study was to evaluate changes in cTnI concentrations and tissue velocity imaging (TVI) values in dogs with lymphoma undergoing chemotherapy with doxorubicin. A total of 182 cTnI and 1017 TVI measurements were performed. Standard echocardiographic parameters, tissue Doppler indices and cTnI concentrations did not differ at any time point within a 12‐week cyclic combination protocol. In conclusion, the use of doxorubicin at standard doses in the treatment of canine lymphoma may not be associated with significant myocardial damage.     

Spectral waveform analysis of intranodal arterial blood flow in abnormally large superficial lymph nodes in dogs

OBJECTIVE: To evaluate pulsed-wave Doppler spectral parameters as a method for distinguishing between neoplastic and inflammatory peripheral lymphadenopathy in dogs. SAMPLE POPULATION: 40 superficial lymph nodes from 33 dogs with peripheral lymphadenopathy. PROCEDURES: 3 Doppler spectral tracings were recorded from each node. Spectral Doppler analysis including assessment of the resistive index, peak systolic velocity-to-end diastolic velocity (S:D) ratio, diastolic notch velocity-to-peak systolic velocity (N:S) ratio, and end diastolic velocity-to-diastolic notch velocity ratio was performed for each tracing. Several calculation methods were used to determine the Doppler indices for each lymph node. After the ultrasonographic examination, fine needle aspirates or excisional biopsy specimens of the examined lymph nodes were obtained, and lymphadenopathy was classified as either inflammatory or neoplastic (lymphomatous or metastatic) via cytologic or histologic examination. Results of Doppler analysis were compared with cytologic or histopathologic findings. RESULTS: The Doppler index with the highest diagnostic accuracy was the S:D ratio calculated from the first recorded tracing; a cutoff value of 3.22 yielded sensitivity of 91%, specificity of 100%, and negative predictive value of 89% for detection of neoplasia. Overall diagnostic accuracy was 95%. At a sensitivity of 100%, the most accurate index was the N:S ratio calculated from the first recorded tracing; a cutoff value of 0.45 yielded specificity of 67%, positive predictive value of 81%, and overall diagnostic accuracy of 86.5%. CONCLUSIONS AND CLINICAL RELEVANCE: Results suggested that noninvasive Doppler spectral analysis may be useful in the diagnosis of neoplastic versus inflammatory peripheral lymphadenopathy in dogs.     

Comparison of B-mode and Doppler ultrasonographic findings with histologic features of benign and malignant superficial lymph nodes in dogs

OBJECTIVE: To compare and correlate B-mode and color Doppler ultrasonographic characteristics with histopathologic findings of benign and malignant superficial lymph nodes in dogs. STUDY POPULATION: 50 superficial lymph nodes that were normal, abnormally large on physical examination, or represented regional lymph nodes draining an area of suspected primary malignancy in 30 dogs. PROCEDURES: Before excision, lymph nodes were evaluated via B-mode and color Doppler ultrasonography to assess size, echogenicity, presence of a hilus, acoustic transmission, and vascular flow. Formalin-fixed, paraffin-embedded tissue sections of excised lymph nodes were stained with H&E and examined for the presence and extent of necrosis, fibrosis, fat, metastases, and tissue heterogeneity. To assess vascularity, the number and distribution of vessels stained by the Verhoeff van Gieson technique were recorded. RESULTS: In superficial lymph nodes, a varied echogenicity corresponded to tissue heterogeneity. The ultrasonographic detection of a hilus was associated with the presence of fibrous tissue, fat, or both in the hilar region. Acoustic enhancement corresponded to presence of areas of intranodal necrosis. There was significant correlation between both the distribution and the number of vessels detected via ultrasonography and that detected by histopathology. The amount of flow estimated via ultrasonography was typically higher than that estimated via histologic examination. CONCLUSIONS AND CLINICAL RELEVANCE: Results indicated that histopathologic changes in canine lymph nodes have associated ultrasonographic changes and suggest that lymph node ultrasonography has an important role in the evaluation of lymph nodes in dogs in general and in dogs with neoplastic disease in particular.     

Ultrasonographic assessment of abdominal lymph nodes in puppies

The jejunal and medial iliac lymph nodes of 53 clinically normal dogs between the age of 4 and 6 weeks were examined ultrasonographically. At least two jejunal and both left and right medial iliac lymph nodes were seen in all dogs. One hundred forty‐five jejunal, 53 right medial iliac and 53 left medial iliac lymph nodes in six litters of dogs, for a total of 251 lymph nodes, were measured for cross sectional maximum diameters. Mean jejunal lymph node length was 16.4 mm (range 6.4–34.9 mm) and mean width was 6.0 mm (range 2.3–15.7 mm). The mean medial iliac lymph node length was 13.6 mm (range 7.2–27.8 mm) and mean width was 4.4 mm (range 1.9–8.2 mm). Significant differences of lymph node size, noted between and within breeds, may not be of clinical significance. The mean size of the combined left and right medial iliac nodes was within previously published ranges for normal adult dogs. Lymph nodes were described in four litters of dogs (162 lymph nodes). Lymph nodes were either uniformly hypoechoic (108/163, 66%) or centrally hyperechoic with a hypoechoic rim (55/163, 34%). Although most (60%) lymph nodes were oval, a variety of shapes were seen, including vermiform and complex branching shapes. We concluded that in 4‐ to 6‐week‐old dogs, medial iliac lymph nodes are similar in size to adult dogs and jejunal lymph nodes are multiple, routinely seen, are larger than in adults and often have unconventional shapes.     

Ultrasonographic appearance of the superficial supramammary lymph nodes in lactating dairy cattle

The superficial supramammary lymph nodes of 54 lactating dairy cows were examined ultrasonographically with a 7.5 MHz linear transducer; each node was measured in two planes within 24 hours of recording the milk somatic cell count. In most cows, the nodes were well demarcated from the surrounding tissue. The parenchyma of the nodes ranged from hypoechoic to anechoic, with a central bright hyperechoic area, and a thin hyperechoic line surrounded the nodes. The size of the nodes varied, but their internal architecture remained relatively consistent. Their mean length was 7.4 cm (range 3.5 to 15 cm) and their mean depth was 2.5 cm (range 1.2 to 5.7 cm). They were significantly larger in cows with more lactations (P<0.05), but there were no correlations between their size and either the time calved or the milk somatic cell count. The lymph nodes on sides which were positive in a California milk test were significantly larger than those on sides which were negative (P<0.05).     

Color and power Doppler ultrasonography for characterization of splenic masses in dogs

Benign and malignant splenic masses can appear similar on B-mode imaging, making ultrasound sensitive but not specific in their diagnosis. Our goal was to characterize color and/or power Doppler characteristics of vasculature within and adjacent to a splenic mass, which would distinguish benign vs. malignant lesions. The hypothesis was that malignant splenic masses will have altered vascular patterns compared with benign masses. Color and power Doppler cineloops evaluating the vasculature within the mass and normal splenic parenchyma were obtained in sagittal and transverse planes using a standardized protocol. Categories of evaluation included presence of peritoneal effusion, a large aberrant or tortuous vessel within the mass, relative blood flow within the mass compared with normal parenchyma, and path of vessels in the adjacent parenchyma entering into the mass. All patients had histopathologic or definitive cytologic diagnosis. Thirty-one dogs were included. There were 13 malignant masses and 18 were benign. Peritoneal effusion was significantly associated with malignancy (P = 0.0007). Presence of an aberrant or tortuous vessel within the mass was nearly significant (P = 0.059). There was no significant difference in any of the color or power Doppler blood flow evaluations. Ultrasonographic findings of a splenic mass and peritoneal effusion may indicate malignancy. The presence of an aberrant vessel within a splenic mass could suggest malignancy; however more data are needed.     

Abdominal ultrasonographic findings at diagnosis of osteosarcoma in dogs and association with treatment outcome

The purpose of this study was to describe abdominal ultrasonographic findings present at diagnosis of osteosarcoma (OSA) in dogs and to investigate for associations with treatment outcome. Medical records from 118 dogs diagnosed with OSA that had abdominal ultrasonography performed as part of their initial evaluation were reviewed. Fifty‐seven percent had ultrasonographic abnormalities identified. The organ with the highest frequency of ultrasonographic changes was the spleen. While most sonographic changes were considered to be either benign or of unknown clinical consequences, metastases were identified in three dogs (2.5%), two of which (1.7%) did not have other evidence of metastasis. Dogs with any ultrasonographic abnormality were less likely to receive definitive therapy (P = 0.005) and exhibited shorter survival, although the latter observation was not statistically significant (P = 0.071). However, the identification of lesions in either the liver (P = 0.021) or the kidney (P = 0.003) was statistically associated with shorter survival.     

Minimal residual disease in lymph nodes after achievement of complete remission predicts time to relapse in dogs with large B‐cell lymphoma

Most dogs with large B‐cell lymphoma (LBCL) that undergo chemotherapy and achieve clinical complete remission (CR) eventually relapse. However, time to relapse (TTR) is unpredictable. The aims of this prospective study were to assess the influence of post‐chemotherapy lymph node (LN) infiltration by large CD21+ cells using flow cytometry (FC) on TTR, and to establish a cut‐off value of prognostic significance. Dogs with newly‐diagnosed, completely staged LBCL in CR after treatment were enrolled. Minimal residual disease (MRD) analysis by FC was performed on LN aspirates. TTR was calculated between MRD and relapse. Thirty‐one dogs were enrolled: 4% had stage V disease, and diffuse large B‐cell lymphoma was the most common histotype (74%). Based on LN infiltration at MRD evaluation, three groups were created: (a) acellular samples, (b) ≤0.5% infiltration and (c) >0.5% infiltration. Overall median TTR was 154 days (range, 31‐1974): 22 (71%) dogs relapsed during the study period, whereas 9 (29%) dogs did not. The difference among the three groups was significant (P = 0.042 log‐rank test): median TTR was not reached for dogs with LN infiltration ≤0.5% (range, 195‐429 days), 164 days (range 63‐1974) for dogs with acellular LN samples, and 118 days (range, 31‐232) for dogs with LN infiltration >0.5%. These results demonstrate that MRD assessment by FC on LN aspirates in dogs with LBCL in clinical CR predicts TTR. LN infiltration by >0.5% large CD21+ cells after treatment is an unfavourable prognostic factor.     

Comparison of three ultrasound guided approaches to the lumbar plexus in dogs: a cadaveric study

ObjectiveTo assess the accuracy of contrast material injection and the dispersion of injectate following ultrasound guided injections at the level of L6 and L7, in canine cadavers.Study designProspective, randomized, experimental study.AnimalsTwenty nine mixed breed canine cadavers (28.9 ± 6.0 kg).MethodsThree ultrasound-guided approaches to the lumbar plexus (LP) were compared: 1) a dorsal pre-iliac approach at the level of L6; 2) a lateral paravertebral approach at mid-L6; and 3) a lateral paravertebral approach at mid-L7. An isovolumic mixture of iodine-based contrast with new methylene blue (0.1 mL kg⁻¹) was injected bilaterally in the juxta-foraminal region along the L6 or L7 nerve root. Computed tomography was performed followed by segmentation and 3D reconstruction of the lumbar spine and contrast material volumes using dedicated software. Distances between contrast material and the fifth through seventh lumbar foraminae, and length of femoral (FN) and obturator (ON) nerve staining were measured and compared between approaches (p < 0.05).ResultsInjectate moved cranial and caudal to the site of injection, and dispersed into an ovoid shape between the quadratus lumborum, iliopsoas and psoas minor muscles. Injections at L7 resulted in significantly closer contrast proximity to the L6 and L7 foraminae (p < 0.001). Femoral nerve staining was similar for all approaches, ON staining was more consistent after L7 injections (p < 0.001).Conclusion and clinical relevanceAn ultrasound-guided lateral paravertebral approach to the LP proved very practical and accurate, with easy visualization of the plexus and associated nerves. To ensure that the ON is covered by injectate, an approach at the level of L7 is recommended. Further studies are necessary to determine if this correlates with clinically effective local anesthesia.     

Ultrasound biomicroscopic findings of the iridocorneal angle in live healthy and glaucomatous dogs

By using ultrasound biomicroscopy (UBM), the cross-sectional structures of the entire iridocorneal angle (ICA) which are unable to assess with gonioscopic examination were evaluated objectively and quantitatively in live healthy and glaucomatous dogs. The ICAs of normotensive eyes in healthy dogs with normal open angle (NOR), a predisposition to primary closed angle glaucoma (PCAG) (PREDIS) and suffering from unilateral PCAG (UNI), as well as the ICAs of hypertensive eyes with acute and chronic PCAG (ACG and CRG), were assessed. The opening of the ciliary cleft in PREDIS was smaller than that in NOR. In UNI, the opening and area of the ciliary cleft were significantly decreased compared with those of NOR and PREDIS. ACG had widespread structural abnormalities including marked decrease in the ciliary cleft and scleral venous plexus, and a thinner sclera than those in normotensive eyes, whereas the ICA collapsed in CRG with the thinnest sclera. Medical therapy-responsive glaucomatous cases had wider ciliary cleft and scleral venous plexus than unresponsive ones. These findings suggest that the ciliary cleft and scleral venous plexus of the ICA are key structures contributing to not only the pathophysiology of canine glaucoma but also the responsiveness to medical therapy in glaucomatous eyes, and cross-sectional entire structures of the ICA should be evaluated quantitatively with UBM when diagnosing and managing canine glaucoma.     

Efficacy and toxicity of carboplatin in the treatment of macroscopic mesenchymal neoplasia in dogs

Palliative chemotherapy options for dogs with macroscopic non-osseous mesenchymal tumours are limited. The purpose of this study was to assess the response rate of these tumours to carboplatin chemotherapy. Medical records of 28 dogs treated with carboplatin for macroscopic mesenchymal neoplasia between 1990 and 2022 were retrospectively reviewed. Sixteen dogs with soft tissue sarcoma and 12 dogs with haemangiosarcoma were included. Responses observed included one complete response and three partial responses, for an overall response rate of 14.2% (4/28) and median time to progression of 42 days (range 21–259 days). Responses were only seen in patients with haemangiosarcoma, for a response rate of 33.3% (4/12) and median time to progression for responders of 103 days (range 39–252 days). Median time to progression for dogs with metastatic disease was similar to those with only local disease (distant median: 44 days; local median: 23 days, p = 0.56). Dogs with chemotherapy-naïve disease were compared to dogs having received previous chemotherapy treatment and had a median time to progression of 75 days and 40.5 days respectively (p = 0.13). Twenty-two dogs experienced 48 adverse events, with most being grade 1 or 2 (79%). Carboplatin was well tolerated, with variable macroscopic anti-tumour activity and short response duration. Carboplatin may be an acceptable rescue option for dogs with macroscopic haemangiosarcoma, especially those patients that cannot receive doxorubicin.