Introduction: Ultrasonography is associated with a high error rate in the evaluation of soft tissue masses. The purposes of this study were to examine the nature of the diagnostic errors and to identify areas in which reporting could be improved.
Methods: Patients who had soft tissue tumours and received ultrasonography during a 10-year period (1999–2009) were identiﬁed from a local tumour registry. The sonographic and pathological diagnoses were categorised as either ‘benign’ or ‘non-benign’. The accuracy of ultrasonography was assessed by correlating the sonographic with the pathological diagnostic categories.
Recommendations from radiologists, where offered, were assessed for their appropriateness in the context of the pathological diagnosis.
Results: One hundred seventy-ﬁve patients received ultrasonography, of which 60 had ‘non-benign’ lesions and 115 had ‘benign’ lesions. Ultrasonography correctly diagnosed 35 and incorrectly diagnosed seven of the 60 ‘non-benign’ cases, and did not suggest a diagnosis in 18 cases. Most of the diagnostic errors related to misdiagnosing soft tissue tumours as haematomas (four out of seven). Recommendations for further management were offered by the radiologists in 144 cases, of which 52 had ‘non-benign’ pathology.There were eight ‘non-benign’ cases where no recommendation was offered, and the sonographic diagnosis was either incorrect or unavailable.
Conclusions: Ultrasonography lacks accuracy in the evaluation of soft tissue masses. Ongoing education is required to improve awareness of the limitations with its use. These limitations should be highlighted to the referrers, especially those who do not have speciﬁc training in this area.
Key words: diagnostic error; haematoma; neoplasm, connective and soft tissue; ultrasonography.
Ultrasonography lacks accuracy in the evaluation of soft tissue masses due to the non-specific nature of many imaging findings. The present study has reaffirmed our previous observations that ultrasonography has a high error rate in distinguishing non-benign from benign lesions. Despite our earlier experiences and the increased awareness of the limitations of ultrasonography, no significant improvement in error rates was observed between the current and the previous study periods.
A common diagnostic error involves mistaking solid tumours for haematomas, sometimes resulting in diagnostic delay and suboptimal management. In a review of 31 cases of soft tissue tumours masquerading as haematoma, Ward et al. found that misdiagnosis was associated with diagnostic delays averaging 6.7 months; furthermore, neither ultrasonography nor magnetic resonance could reliably differentiate soft tissue tumours and haematoma. It may be useful to correlate with clinical history such as recent trauma and examination findings such as subcutaneous ecchymosis. However, it is important to note that a history of trauma may be incidental, and ecchymosis can also occur with tumoural bleeding.
Given these difficulties, it is not surprising that many radiologists err on the side of caution when confronted with a soft tissue mass on ultrasonography. Of the 115 patients with histologically benign lesions, ultrasonography suggested a suspicious diagnosis or recommended further evaluation in 92 cases (80%, positive likelihood ratio 1.1). A ‘positive’ ultrasonography result per se therefore adds little diagnostic value in the evaluation of patients with soft tissue masses. On the other hand, one should be cautious to assign a ‘negative’ result to a study without providing specific guidance or recommendation. Notably, diagnostic delays have been observed in false negative cases when such recommendation has not been explicitly made. This is particularly relevant in our local practice, where non-specialists (e.g. general practitioners) contribute up to 60% of the referrals for ultrasonography examination for soft tissue masses. In these circumstances, a short comment such as the following may help guide the referrers in appropriate cases:
… The findings are non-specific. If the lesion does not resolve rapidly, or if the radiological diagnosis does not fit the clinical picture, a referral to a specialist surgeon is recommended and further imaging such as MRI may be appropriate.
Despite its limitations, ultrasonography may serve specific roles in the work-up of soft tissue masses. First, ultrasonography can confirm the presence of a mass, which can sometimes be difficult to ascertain clinically. Second, ultrasonography can differentiate cystic lesions from solid lesions. Third, ultrasonography can often reliably diagnose lesions with certain well-characterised sonographic features. For instance, a cyst adjacent to a tendon may suggest a ganglion, and a superficial well-defined echogenic mass may suggest a lipoma. It should be noted that the majority of these lesions are satisfactorily managed in the community without being referred to the registry, and are therefore excluded in this review. Fourth, ultrasonography may be used to guide biopsy of the lesions. This is particularly valuable in targeted biopsy of large, heterogeneous tumours.
Ultrasonography has been utilised in tumour follow-up in the research setting. It has been used to detect tumour recurrence[6, 7] and monitor regression of tumour neovascularity induced by therapy for musculoskeletal sarcoma. Ultrasonography may also be used in conjunction with MRI when susceptibility artefacts from orthopaedic hardware prevent evaluation of specific areas following surgery. It has been suggested that colour Doppler flow imaging and spectral wave analysis may allow assessment of blood flow within soft tissue masses and, by inference, differentiate between malignant and benign tumours.
In summary, ultrasonography has specific roles in the evaluation of soft tissue masses. However, aside from the recognisable entities of ganglion, superficial lipoma and obvious peripheral nerve sheath tumour, ultrasonography of soft tissue masses remains non-specific with respect to malignancy. Ongoing education is prudent to improve our understanding of its limitations and pitfalls. In addition, it may be important to highlight these limitations to the referrers, especially if they have no specific training in the management of soft tissue masses.