Pitfalls of breast MRI.

This article reviews some of the common pitfalls associated with breast MRI. Pitfalls can be broadly considered as "technical" (relating to patient factors or machine factors, which can influence image interpretation) or "non-technical" (relating to misinterpretation of imaging findings in the absence of technical problems). Awareness of potential pitfalls is important if MRI is to maximize its potential in breast imaging.

MRI screening for acoustic neuroma: a comparison of fast spin echo and contrast enhanced imaging in 1233 patients.

Gadolinium enhanced MRI is the gold standard investigation for the detection of acoustic neuroma. Non-contrast MRI sequences have been suggested as an alternative for screening examinations. In order to determine the utility of fast spin echo imaging, both gadolinium enhanced T1 weighted images and fast spin echo T2 weighted images were acquired in 1233 consecutive patients referred for exclusion of acoustic neuroma. Two radiologists independently recorded their findings. Fast spin echo T2 weighted images were evaluated with respect to the visibility of nerves within the internal auditory canals and allocated a confidence score for the presence or absence of acoustic neuroma. 33 acoustic neuromas were identified. Only 56% were confidently identified on fast spin echo T2 weighted images alone; gadolinium enhanced T1 weighted images were required to confirm the diagnosis in 44% of the cases, including 9 of the 10 intracanalicular tumours. However, when identification of two normal intracanalicular nerves is employed as the criterion of normality, the single fast spin echo T2 weighted sequence excluded acoustic neuroma in 59% of this screened population. It is concluded that an imaging strategy intended to identify small intracanalicular acoustic neuromas cannot rely on fast spin echo T2 weighted imaging alone. Gadolinium enhanced T1 weighted imaging could be restricted to patients where fast spin echo images do not exclude acoustic neuroma but this strategy requires continuous supervision by an experienced radiologist. In most practices the screening examination should continue to include a gadolinium enhanced sequence in order to optimize the detection of small acoustic neuromas.

MRI in routine breast cancer follow-up: correlation with clinical outcome.

AIM: Magnetic resonance imaging (MRI) of the breast has been shown to be useful as an additional imaging test in patients suspected of having recurrent disease, when clinical examination or mammography are equivocal. This study examines the utility of MRI as a routine test in the follow-up of treated breast cancer patients without equivocal clinical or imaging findings. MATERIALS AND METHODS: Contrast-enhanced breast MRI was performed as part of routine breast cancer follow-up in 26 patients with equivocal clinical or mammographic findings and 33 control patients in whom clinical and imaging findings were consistent with post-treatment changes only. Clinical outcome was assessed at 3 years post MRI. RESULTS: Four patients with equivocal clinical or mammographic findings had abnormal MRI: all MRI abnormalities were subsequently shown to represent benign disease. None of the equivocal group developed local recurrence, although two of 26 developed distant metastases. One of the control group had unsuspected distant metastases detected by MRI: the other 32 patients had no significant abnormality on MRI. Three patients subsequently developed local tumour recurrence. CONCLUSION: A single normal breast MRI examination during follow-up is a poor predictor of subsequent local recurrence.

MRI appearances of the axilla in treated breast cancer.

Differentiation between recurrent axillary disease and changes due to radiotherapy or surgery has major implications for management in patients following breast cancer treatment, but clinical examination of the axilla may be difficult. This study was undertaken to correlate the MRI appearances of the axilla following breast cancer treatment with clinical outcome. 74 women with treated breast cancer were evaluated by MRI (0.5 T) and the appearances defined by consensus. Outcome was assessed by long-term clinical follow-up. 62 women had symptoms related to the axilla while 12 were scanned to stage the axilla. None of the axillary staging group had abnormal MRI appearances and none of these subsequently developed recurrence. The 62 symptomatic women were subdivided according to MRI appearances. 22 had normal axillary appearances, 18 had an axillary mass and 22 women had abnormal axillary appearances (rated mild, moderate and severe) in the absence of a mass. Normal axillary appearances on MRI excluded recurrent disease as the cause of symptoms with a specificity of 94.7% and a positive predictive value (PPV) of 95.5%. The presence of an axillary mass was commonly but not exclusively due to recurrent disease (sensitivity 68.4%, specificity 88.4%, PPV 72.2%). Sensitivity for diagnosis of axillary recurrence was increased to 89.5% with a specificity of 76.7% if the criteria for recurrent disease were taken as either the presence of an axillary mass or severe axillary changes in the absence of a mass lesion.

An objective assessment of speech and swallowing following free flap reconstruction for oral cavity cancers.

In this prospective study, eight consecutive patients underwent excision of oral cavity cancer and reconstruction with a microvascular free flap. Six patients had one pre- and two postoperative assessments of speech and swallowing at four to six weeks and four to six months respectively. One patient could not attend for the preoperative assessment and another for the first postoperative assessment. The speech assessment consisted of an intelligibility score and an articulation score. The swallowing assessment consisted of a videofluoroscopic examination. Five patients had an excellent postoperative speech assessment score, two had a moderate and one a poor result. Videofluoroscopy demonstrated minor swallowing problems preoperatively in one patient. Postoperatively, three patients had severe or moderate loss of control of bolus. Significant aspiration was detected in two and mild aspiration in one patient at the second postoperative assessment. The abnormalities of speech and swallowing were minor in the majority of the patients and the overall outlook was good. This information should be of value for preoperative counselling.

Ultrasound demonstration of small breast cancers detected by mammographic screening.

The visibility on ultrasound (US) of 79 screen-detected small breast cancers was studied retrospectively. Overall visibility on US was 61% (48/79). Cancers presenting as microcalcification alone on mammography were rarely demonstrated (1 of 26 tumours). However, the US visibility of cancers with other mammographic appearances was 88.7% (47/53). The majority of cancers were impalpable but were demonstrated using free-hand ultrasound. All cancers visible on US appeared as hypoechoic masses with three quarters of these showing distal attenuation. US is useful in assessing and localizing small cancers detected by mammographic screening.