Incidental findings are common on thoracolumbar MRIs. These studies provide high-resolution images of the spine and surrounding viscera, so abnormalities of unclear significance are frequently encountered. One of the more common findings is vertebral marrow signal abnormality (VMSA), usually seen as marrow heterogeneity on T1 weighted images. The differential diagnosis for these findings is broad and includes hemangioma, marrow hyperactivity seen in anemia and thrombocytopenia, degenerative conditions, fractures, infection, myeloma, lymphoma, and metastatic disease. In order to better understand the prevalence of VMSA, how it is described by radiologists, how often they recommended further follow-up, and how often the ordering provider performed follow-up or further studies, Dr. Carlson and colleagues from Oregon reviewed over 1,500 thoracic and lumbar MRI reports from studies ordered for back and/or leg pain. They found VMSA was reported in 4% of patients (n=65). In 48% of cases, the radiologist recommended further follow-up, including additional imaging in 45%, labs in 16%, and clinical correlation in 55%. The ordering provider documented further follow-up or diagnostic study in two-thirds of the patients for whom it was recommended by the radiologist and in 18% of those without any recommendation for follow-up in the MRI report. One patient was found to have myeloma, with the remainder having benign or nonspecific diagnoses. The most common diagnoses were hemangioma (44%), anemia (12%), and fracture (8%). Approximately 1/3 were classified as non-specific based on the MRI findings and lack of further follow-up.
The authors have done a nice job studying a clinically relevant question faced by spine providers and neuroradiologists on a regular basis. While VMSA is relatively uncommon overall (4% of patients), spine clinicians seeing high volumes of patients will encounter them frequently. The variation in follow-up recommendations and the degree to which the recommendations were followed was striking and indicates the need for a diagnostic algorithm to assist clinicians in managing these findings. Some of the variation was likely due to the different appearance of the lesions, some of which likely appeared quite benign and others more aggressive. The authors did not review the images or classify them beyond the presence or absence of VMSA noted by the radiologist, and there is wide morphological variation in VMSAs. Additionally, the clinical history provided to or available to the radiologist also likely varied substantially, and this may have affected their follow-up recommendations. The fact that the ordering provider did not perform the recommended follow-up in 1/3 of patients is concerning and suggests they either did not read the reports or felt that the radiologist was being overly conservative in recommending further follow-up. When large variation is observed in spine care, it usually indicates that the spine community has not come to a consensus on the best approach to the problem. More study on this topic would be helpful and might identify risk factors for malignancy or infection in VMSA patients that would prompt more aggressive diagnostic investigation or closer follow-up. Busy clinicians could use an algorithm to help with the work-up of VMSA that is efficient and avoids unnecessary testing but prevents missing the diagnosis in dangerous conditions.
Please read Dr. Carlson's article on this topic in the March 15 issue. Does this change how you approach patients with the incidental finding of VMSA?
Adam Pearson, MD, MS
Associate Web Editor