ObjectiveTo determine the prevalence of clinically and non-clinically relevant extra-spinal incidental findings (IF) in patients undergoing magnetic resonance imaging (MRI) of the lumbar spine and to evaluate the rate of undetected findings in archived radiological reports.MethodsA retrospective search of patients undergoing lumbar spine MRI from January 2006 to December 2010 was conducted. By means of randomisation, we retrospectively reviewed 3,000 lumbar spine MRI examinations. Extra-spinal abnormalities were classified according to a modified CT Colonography Reporting and Data System (C-RADS). We retrospectively compared our structured approach with the archived MRI reports as it regarded the detection of extra-spinal IF to estimate non-detection rates.ResultsBy means of the structured approach used, extra-spinal findings were detected in 2,060 (68.6 %) of the 3,000 lumbar spine MRI examinations; 362 (17.6 %) patients had indeterminate or clinically important findings (E3 and E4) requiring clinical correlation or further evaluation. After review of the original archived radiological reports, potentially important C-RADS E3 and E4 extra-spinal IF were respectively reported in 47 of the 265 (17.7 %) and in 8 of 74 (10.8 %) patients.ConclusionsOur study shows that incidental extra-spinal findings at conventional lumbar spine MRI are common but underestimated in radiological reports.

The frequency of the diagnosis of brain metastases has increased in recent years, probably due to an increased diagnostic sensitivity. Site predilection of brain lesions in oncological patients at the time of onset, may suggest mechanisms of brain-specific vulnerability to metastasis. The aim of the study is to determine the spatial distribution of intra-axial brain metastases by using voxel-wise statistics in breast and lung cancer patients. For this retrospective cross-sectional study, clinical data and MR imaging of 864 metastases at first diagnosis in 114 consecutive advanced cancer patients from 2006 to 2011 were included. Axial post-gadolinium T1 weighted images were registered to a standard template. Binary lesion masks were created after segmentation of volumes of interest. The voxel-based lesion-symptom mapping approach was used to calculate a t statistic describing the differences between groups. It was found that the lesions were more likely to be located in the parieto-occipital lobes and cerebellum for the total cohort and for the non small cell lung cancer group, and in the cerebellum for the breast cancer group. The voxel-wise inter-group comparisons showed the largest significant clusters in the cerebellum for the breast cancer group (p < 0.0008) and in the occipital lobe (p = 0.02) and cerebellum (p = 0.02) for the non small cell lung cancer group. We conclude a non-uniform distribution of metastatic brain lesions in breast and lung cancer patients that suggest differential vulnerability to metastasis in the different regions of the brain.

LSE is highly associated with spondylolisthesis, facet arthropathy and BMI, suggesting underestimation of its clinical impact as an integral component in chronic lumbar back pain. Longitudinal simultaneous X-ray/MRI studies should be conducted to test the relationship of LSE with lumbar spinal instability and low back pain.

When large cerebrospinal fluid loss is observed during spinal surgery, brain imaging study should be carried out. The pathogenetic hypothesis of microcirculation vessels tearing, the role of previous spinal surgery and of cerebellar atrophy should be considered and validated with further investigation.