SMALL ANIMALS 323 © 2009 The Authors Journal compilation © 2009 Australian Veterinary Association Australian Veterinary Journal Volume 87, No 8, August 2009 Blackwell Publishing Asia Ultrasonographic characteristics of soft tissue tumours in dogs ZHK Loh, GS Allan, RG Nicoll and GB Hunt* Objective To identify and describe the ultrasonographic features of soft tissue tumours in dogs. Procedure Superficial soft tissue tumours of various histological types, including mast cell tumours (MCTs) and soft tissue sarcomas (STSs), were evaluated. Ultrasound was used to visualise internal characteristics of the tumour, including vascularity. Tumours were categorised according to size, shape, margin definition, tissue plane mobility, echogenicity, echotexture, acoustic shadowing or enhancement and vessel distribution. Objective measurements of intratumoural blood flow included velocities and maximal perfused cross-sectional area (fractional area). Logistic regression models incorporating a variety of data were used in an attempt to predict the histopathological type of tumours. Results The logistic regression model defined by the parameters echotexture, margin definition and presence of subcapsular vessels was highly predictive of MCTs (> 73%; P = 0.024). Several other trends, including a larger size for STSs and less vascularity for both MCTs and STSs, were observed, but did not reach statistical significance. Conclusion This preliminary study has shown the potential diagnostic value of ultrasound in differentiating soft tissue tumours. However, at present, ultrasound cannot replace biopsy and histopathological evaluation for tumour diagnosis. Keywords diagnosis; mast cell tumours; soft tissue sarcomas; ultrasound; vascularity Abbreviations CFM, colour flow mapping; FA, fractional area; IMD, intratumoural microvessel density; MCT, mast cell tumour; RI, resistive index; STS, soft tissue sarcoma Aust Vet J 2009;87:323–329 doi: 10.1111/j.1751-0813.2009.00460.x U ltrasound is becoming increasingly more common in veterinary practice and advanced sonographic techniques such as pulsed and tissue Doppler can provide greater degrees of sensitivity in assessing soft tissue characteristics such as blood flow. Ultrasound is already used in some cases to evaluate tumour size, mobility and relationship to surrounding structures such as muscle bellies, subcutaneous fat and collagen-rich tissue planes, as an aid to surgical planning. However, it is becoming apparent that other sonographic characteristics may have a role in predicting tumour type and biological behaviour. In particular, vascularity has prognostic significance in humans, in particular for thyroid and smooth muscle tumours, with increasing vascularity indicating a higher risk of malignancy. 1,2 Angiogenesis correlates with the risk of distant metastasis and it has been shown by several authors that the intratumoural microvessel density (IMD), measured histologically, is predictive of the histological grading of mast cell tumours (MCTs) and soft tissue sarcomas (STSs) in dogs. 3–5 Recent studies of canine tumours have demonstrated the feasibility of using ultrasound as a non-invasive means of determining tumour vascularity. 6,7 Based on previous observations, it seems likely that sonography could become an important tool for predicting tumour type and behaviour, and assist with minimally-invasive evaluation of tumour patients prior to planned surgical excision or biopsy. The use of ultrasound to predict the histologic type of tumour in dogs has not been previously reported. The present study used Doppler ultrasonography to determine the vascularity of a range of soft tissue tumours in dogs, in particular MCTs and STSs, and attempted to differentiate them using ultrasonographic characteristics. Objectives were: to identify local tissue planes and tumour mobility with respect to these tissue planes; evaluate the basic sonographic characteristics of different canine soft tissue tumours; and evaluate the predictive value of tissue blood flow parameters in determining the identity of soft tissue tumours. An additional aim of the present study was to determine whether results in dogs are similar to those previously reported in humans. 8,9 For this study we recruited 29 client-owned dogs presented to the University Veterinary Centre, Sydney between August 2006 and April 2007 for evaluation of spontaneously-occurring cutaneous or subcutaneous masses. Selection criteria included a palpable mass that resembled a MCT or STS, and histological or cytological confirmation of tumour type. Signalment (breed, age, sex) was noted. A thorough physical examination was performed and weight, heart rate and tumour dimensions were recorded. Blood samples were collected to obtain a minimum database consisting of packed cell volume, complete blood count and biochemistry panel. Dogs were sedated with standard agents (e.g. methadone or morphine (0.1 mg/kg) with acepromazine (0.01 mg/kg) administered IV), in order to minimise motion artefacts during ultrasound examination. All ultrasound assessments were performed using a Philips HDI 5000 ultrasound unit (Philips Ultrasound, Bothell, USA). Initial B-mode ultrasonography of the mass was performed to document the tumour’s appearance. The tumours were evaluated for size and a size score was defined as the sum of three dimensions. The cross-sectional tumour border shape was described as spindle-shaped, ellipsoid or ovoid. Margination was assessed by determining the definition of *Correspondence author. University Veterinary Centre, Sydney, University of Sydney, New South Wales 2006, Australia; gbhunt@mail.usyd.edu.au Materials and methods