Sinusoidal Obstructive Syndrome Diagnosed With Superparamagnetic Iron Oxide–Enhanced Magnetic Resonance Imaging in Patients With Chemotherapy-Treated Colorectal Liver Metastases

Janice Ward(St James's University Hospital), J. Ashley Guthrie(St James's University Hospital), Maria Sheridan(St James's University Hospital), Sheila Boyes(St James's University Hospital), Jonathan Smith(St James's University Hospital), Daniel J. Wilson(St James's University Hospital), Judy Wyatt(St James's University Hospital), Darren Treanor(St James's University Hospital), Philip Robinson(St James's University Hospital)
Journal of Clinical Oncology
September 8, 2008
Cited by 85

Abstract

PURPOSE: To assess the predictive value of superparamagnetic iron oxide (SPIO) -enhanced T2-weighted gradient echo (GRE) imaging to determine the presence and severity of sinusoidal obstructive syndrome (SOS). PATIENTS AND METHODS: Sixty hepatic resection patients with colorectal metastases treated with chemotherapy underwent unenhanced magnetic resonance imaging (MRI) followed by T2-weighted GRE sequences obtained after SPIO. The images were reviewed in consensus by two experienced observers who determined the presence and severity of linear and reticular hyperintensities, indicating SOS-type liver injury, using a 4-point ordinal scale. Sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) with 95% CIs for the detection of SOS were calculated. RESULTS: Twenty-four of 60 patients had moderate to severe SOS on MRI. MRI achieved a sensitivity of 87% (95% CI, 66% to 97%), specificity of 89% (95% CI, 75% to 97%), PPV of 83% (95% CI, 63% to 95%), and NPV of 92% (95% CI, 77% to 98%). SOS was never found at surgery or histology in patients whose background liver parenchyma was normal on SPIO-enhanced MRI. CONCLUSION: SOS is present in a significant proportion of patients with treated colorectal metastases and is effectively detected on SPIO-enhanced T2-weighted GRE images.


Related Papers

No related papers found

Powered by citation graph analysis