Elli Kalliopi Papantoniou MBBS, BSc, Presenter: Nothing to Disclose

Ruchi Kabra MBBS, MRCS, Abstract Co-Author: Nothing to Disclose

Paul David Humphries MRCP, FRCR, Abstract Co-Author: Nothing to Disclose

Shonit Punwani MBBS, Abstract Co-Author: Nothing to Disclose

To compare Short Tau Inversion Recovery – Half Fourier Single Shot Turbo Spin Echo (STIR-HASTE) MRI used for whole body lymphoma staging against optimised Periodically Rotated Overlapping ParallEL Lines with Enhanced Reconstruction (PROPELLER) MRI for detection of chest wall disease extension.

A local database was retrospectively interrogated for patients < 18 years, with histologically proven lymphoma that had undergone MRI and CT as part of their routine pre-treatment disease staging. In total 20 respiratory and ECG triggered axial STIR-HASTE and 21 optimised non-fat saturated multi-breath-hold axial PROPELLER together with the pre-treatment chest CT were identified and retrieved for analysis. A radiologist (3 years experience of lymphoma MRI staging) unaware of patient and sequence details evaluated each study for the presence or abscence of chest wall extension. The radiologist scored the presence or abscence of chest wall disease on a 4 point scale (1 - no disease, 2 - likely no disease, 3 - likely disease, 4 - definite disease). Subsequently all CT studies were assessed in consensus by two separate radiologists to establish a reference standard. Sensitivity and specificity for STIR-HASTE and PROPELLER sequences were calculated against the reference standard. Reciever operator characteristic curves were determined for each sequence and area under the curve calculated.

The sensitivity and specificity for detection of chest wall extension on STIR-HASTE MRI was 100% and 82%; and 100% and 100% for PROPELLER imaging. Area under the reciever operator characteristic curve was 1.0 for PROPELLER  and  0.97 (95% CI: 0.89 to 1.05) for STIR-HASTE imaging.  

Optimised multi-breath-hold PROPELLER MRI maintains sensitivity for chest wall disease detection whilst improving specificity. There is only a marginal improvement in the overall performance for wall disease detection of using PROPELLER as compared with STIR-HASTE imaging.

Identification of chest wall disease extension is critical for accurate disease staging and treatment selection. PROPELLER has a marginal improved performance over STIR-HASTE for chest wall disease.

Papantoniou, E, Kabra, R, Humphries, P, Punwani, S, STIR-HASTE and PROPELLER MR Sequences: Which is Better for Detecting Chest Wall Extension of Disease in Lymphoma?.  Radiological Society of North America 2010 Scientific Assembly and Annual Meeting, November 28 - December 3, 2010 ,Chicago IL. http://archive.rsna.org/2010/9008531.html