Article Text

FRI0624 Structural MRI-based connectomics in SLE: a pilot study
  1. GA Ramirez1,2,
  2. EP Bozzolo1,
  3. P Preziosa3,
  4. MA Rocca3,
  5. L Moiola3,
  6. LA Coletto2,
  7. E Tombetti1,2,
  8. P Rovere-Querini1,2,
  9. M Filippi2,3,
  10. AA Manfredi1,2
  1. 1Unit of Allergy, Immunology, Rheumatology and Rare Diseases, IRCCS Ospedale San Raffaele
  2. 2Università Vita-Salute San Raffaele
  3. 3Institute of Experimental Neurology and Unit of Neurology, IRCCS Ospedale San Raffaele, Milano, Italy


Background Neuropsychiatric manifestations are common in patients with systemic lupus erythematosus (SLE). Furthermore, subclinical brain damage occurs in an even higher fraction of patients. However, little is known about the effect of these phenomena on brain connectivity. MRI-based connectomics relies on graph analysis of structural and functional images to detect alterations of the topographic organization of the brain and has been successfully employed to dissect network disassembly in neuroinflammatory diseases such as multiple sclerosis (MS) and Devic's syndrome.

Objectives To investigate the topographic organization of the brain of patients with SLE with and without neuropsychiatric manifestations.

Methods Thirty-two patients with SLE (12 with overt neuropsychiatric involvement as per ACR criteria) were enrolled and compared with 32 healthy controls (HC) and 32 patients with relapsing-remitting MS, all matched for sex, age and disease duration (where applicable). Diffusion tensor (DT) and dual-echo MRI scans were obtained. Structural connectivity matrices between 116 cortical and subcortical brain regions were estimated and global and nodal network metrics were calculated.

Results Conventional MRI revealed that patients with SLE had significantly higher T2-lesional volumes when compared to controls (p<0.0001). Patients with definite NPSLE had a higher lesion burden (p=0.006). Network strength, transitivity and global efficiency were all significantly impaired in patients with MS and SLE when compared to HC (p<0.0001). MS and SLE were also characterized by higher average path length when compared to HC (p<0.0001). Global structural alterations were more significant in MS patients than in patients with SLE (p from 0.005 to 0.01 at multiple comparison). However, antiDNA-positive patients (n=24) showed a more severe phenotype when compared to antiDNA-negative patients (p from 0.026 to 0.041) and did not differ significantly from patients with MS. When regional hubs were analysed, patients with SLE and MS showed a reduced strength compared to HC (p from <0.0001 to 0.001 at multiple comparison). Hub strength impairment was more pronounced in MS when compared to SLE and preferentially involved hubs located in fronto-temporo-parieto-occipital cortices, subcortical nuclei (including the thalamus, caudate nucleus and putamen) and cerebellum (p from 0.001 to 0.05 at multiple comparison). No significant associations were found between global structural parameters, clinical diagnosis of neuropsychiatric SLE, other SLE sub-phenotypes, presence of antiphospholipid antibodies, antiphospholipid syndrome and SLE-related damage burden.

Conclusions Structural alterations of global and regional brain connectivity occur in patients with SLE, irrespectively of the clinical phenotype. AntiDNA-positive patients are characterized by a more severe phenotype, which is similar to that of patients with relapsing-remitting MS.


  1. Sabbadini MG et al, Lupus, 1999.

  2. Filippi M et al., Lancet Neurol, 2013.

  3. Rocca MA et al., Brain Struct Funct, 2016.


Disclosure of Interest None declared

Statistics from

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.