Use of USPIOs for Clinical Lymph Node Imaging

2008 
The detection of lymph node metastases is critical for the choice of preoperative chemo-radiotherapy, surgical treatment, and patient prognosis. Ultrasmall superparamagnetic iron oxide (USPIO) enhanced magnetic resonance imaging (MRI) is now being used as a potential biomarker for the detection of lymph node metastases. USPIO is transported into the interstitial space and reaches the lymph nodes via the lymphatic circulation acting as a ‘negative contrast’ agent, which can potentially identify metastases independent of lymph node size. We assessed here the diagnostic precision and application of magnetic resonance imaging (MRI) in conjunction with USPIO contrast as a biomarker for detecting LNMs pre-operatively, compared with the gold-standard post-operative histopathology. A total of 19 prospective studies have been published between 1994 and 2005, comprising 3,004 lymph nodes in 631 patients who underwent comparable MR imaging with and without USPIO. The overall sensitivity and specificity for MRI with USPIO were 88.1% and 96.2% respectively, with an area under the ROC curve of 84.2% and a diagnostic odds ratio of 123.1. When unenhanced MRI was evaluated, there was a significant reduction in the overall sensitivity and specificity (63% and 92.7% respectively) and a diagnostic odds ratio of 26.7. USPIO-enhanced MRI had a higher sensitivity and specificity for lymph node status in the abdomen and pelvis compared with the chest or head and neck. Similarly, the diagnostic precision was better when a 3T MRI scanner was used at 2.6 mg/kg contrast dose, than when using a 1.5T field strength. Thus, USPIO is a promising nanomarker, which may potentially increase the precision for the pre-operative diagnosis of lymph node metastases. It may be used for providing guidelines for selecting patients for targeted neo-adjuvant chemotherapy and extended lymphadenectomy thus possibly reducing the incidence of local recurrence.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    27
    References
    1
    Citations
    NaN
    KQI
    []