MRI Findings Help Differentiate SLE-Related Neuropsychiatric Events from Others

Magnetic resonance imaging (MRI) findings such as inflammatory-type lesions and myelopathies are significantly associated with attribution of neuropsychiatric (NP) events to systemic lupus erythematosus (SLE), supporting their use in distinguishing NP-SLE from mimicking conditions, according to study results published in Lupus Science & Medicine.

Researchers conducted a retrospective cohort analysis to evaluate whether specific elementary brain MRI lesions are associated with (1) the attribution of NP events to SLE (using clinical judgment [CJ] and a validated attribution algorithm [AA]), (2) the type or cluster of NP manifestations, and (3) the pharmacologic treatment choices made. The analysis was conducted across 4 academic centers (Italy, Greece, Brazil); 154 patients (mean age at diagnosis, 37 years; 90% women) with SLE who experienced a first NP event between 1999 and 2015 and underwent conventional brain MRI were included.

Among the 154 NP events analyzed, 57% were attributed to SLE by CJ and 55% by an AA. Diffuse central nervous system (CNS) events, which accounted for 55% of all NP manifestations, were significantly more likely to be nonattributed to SLE (P =.005 for CJ; P <.001 for AA). Headache was the most common NP manifestation. Most nonattributed events occurred after SLE diagnosis (71% for CJ; 77% for AA; P =.005).

The MRI data showed that 37% of patients had normal scans, while 46% had T2/FLAIR hyperintense lesions, and 11% had large infarcts. Inflammatory-type lesions and myelopathies were significantly associated with attribution to SLE per AA (P =.027 and P =.043, respectively), while cerebral atrophy was more common among nonattributed cases per CJ (12% vs 2.3%; P =.020).

In multivariable models, inflammatory-type lesions were associated with AA-based attribution (odds ratio [OR], 3.91; 95% CI, 1.15-18.1) and myelopathies remained significant for both CJ and AA. Normal MRI was associated with nonattribution (CJ: OR, 0.42; 95% CI, 0.21-0.82; AA: OR, 0.27; 95% CI, 0.13-0.52).

Excluding MRI from the AA or omitting peripheral nervous system events in sensitivity analyses did not materially change results. Inflammatory-type lesions were more frequent among patients with early disease (<5 years duration, P =.041) and younger age (<50 years, P =.041), particularly for attributed events.

Normal MRI was more common among patients with diffuse CNS manifestations (52%; P <.001), while large infarcts, lacunes, and myelopathies were seen almost exclusively among those with focal CNS events. No significant treatment trends emerged by lesion type, although corticosteroid use was more frequent among patients with inflammatory lesions and focal CNS events.

Study limitations include the retrospective design, a lack of access to potentially influential clinical data (eg, comorbidities and corticosteroid exposure), and the exclusion of patients with rare neuropsychiatric manifestations. The lack of detailed characterization of T2/FLAIR lesion volume, shape, and location and the absence of centralized, blinded MRI review were cited as further limitations.

The study authors concluded, “Clinicians should perform brain MRI not only to exclude secondary causes but also to refine their critical approach to NP symptoms, until more specific (neuroimaging) biomarkers could modify the actual diagnostic paradigm in NP-SLE.”

This article originally appeared on Rheumatology Advisor