A major unmet need in the field of spondyloarthritis (SpA) is the lack of feasible and sufficiently responsive tools to measure objective features of inflammation and disease progression. A major development in the field has been highly discriminatory MRI-based scoring tools that can quantify the degree of bone marrow edema (BME) in the sacroiliac joints (SIJ) and spine. Formal methods for quantifying disease activity in the SIJ can be generally categorized as being primarily based on either a global scheme that focuses on the single most severely affected semi-coronal image or on a more detailed method that scores several consecutive semicoronal images that depict the synovial portion of the SIJ (SPARCC SIJ score). The presence and extent of BME in the cartilaginous portion of the joint is the primary MRI feature that is scored. In the SPARCC method, each SIJ is divided into quadrants and the presence of BME in each quadrant is scored on a dichotomous basis with additional weighting for intensity and depth. ASAS/OMERACT conducted a multi-reader exercise evaluating the different scoring methods for reliability and sensitivity to change. Agreement between readers and sensitivity to change was somewhat better for the more detailed SPARCC scoring method. This approach to scoring inflammation in the SIJ is to date the only method shown to discriminate between treatment groups in a randomized placebo-controlled trial. Systematic evaluation of the SIJ and the use of an online training module (www.arthritisdoctor.ca) have been shown to enhance reliability. Several methods have recently been described for scoring structural lesions in the SIJ. The SPARCCC SIJ structural lesion score (SSS) has recently been shown to discriminate between therapies and show that anti-TNF reduces erosions. Two primary approaches have been validated for the quantification of the extent of inflammation in the spine for the purposes of evaluation of new therapeutics in clinical trials. The first method, the ASspiMRI-a (AS spinal MRI activity) index, scores the severity of BME and erosions at each DVU according to a 0 to 6 scoring scheme with higher values being assigned to the presence of erosions. The score for BME is based on the total area involved in the DVU according to a <25%, 25-50%, and >50% grading scheme. The Berlin method does not score erosions but otherwise is the same as ASspiMRI-a. For both methods all 23 DVUs in the spine are scored in a single sagittal plane. This approach has been shown to be reproducible and to discriminate between treatment groups in trials of anti-TNF therapies. The second method, the SPARCC MRI spinal inflammation index, takes advantage of the ability of MRI to visualize lesions in three dimensions and scores a maximum of 6 of the most severely affected DVU for the purposes of clinical trials. Since inflammatory lesions observed on MRI are often asymmetrical there are theoretical advantages to using a scoring method that systematically assesses lesions in several dimensions. The rationale for the choice of 6 affected DVU is based on data that shows that the mean number of affected DVU in patients with AS is 3.2. However, this method has also been validated for scoring all 23 DVU where discrimination appears comparable to a more limited assessment of 6DVU. In the SPARCC method each DVU is divided into quadrants with bone edema in each quadrant being scored on a dichotomous basis as being present or absent with additional weighting for intensity and depth. Lesions in each affected DVU are scored in 3 consecutive sagittal slices so that the extent of the lesion can be defined in both the coronal and the sagittal planes and also in the lateral segments of the spine, which may be important for assessing inflammation in the thoracic spine. Recent work has shown that a major portion of spinal inflammation in the thoracic spine occurs in the lateral segments involving the costovertebral and costotransverse joints. A validation exercise has been conducted by ASAS/OMERACT using multiple readers to determine which method performs best with respect to feasibility and ability to discriminate between active and control therapies. Both scoring methods demonstrated high responsiveness after administration of anti-TNFα therapies although the SPARCC method was consistently more reproducible, particularly when evaluated by neutral observers with limited experience in either method. A recent report has described a scoring method for assessing fat lesions in the spine, the Fat SpA Spine score (FASSS), which reliably detects change in fat lesion score. Further study in clinical trials is now warranted for further validation of MRI-based structural lesion scores in both the spine and SIJ.
Disclosure of Interest None declared