Background/Purpose: The type 1 interferon (IFN) cytokine family is key to the pathogenesis of SLE, evidenced by the expression of IFN stimulated genes (ISGs) in most patients. Phenotypic differences between SLE patients who express ISGs (IFN high) and those who do not (IFN low) are not well understood. With the emergence of IFN blocking drugs, it is important to examine the clinical utility of IFN status testing.  We report the results of longitudinal analysis of IFN and clinical status in SLE.

Methods: 205 patients meeting ACR criteria for SLE were recruited at a tertiary centre where extensive clinical data are collected prospectively on consenting subjects. Whole blood RNA samples were collected in PAXgene tubes. IFN status was determined using the IFN module from Modular Immune Profile Test (DxTerity Diagnostics).

Results: 729 results (205 patients, 142 with longitudinal samples) were analysed. At baseline, 62.9% of patients were IFN high, 30.2% IFN low and 6.8% borderline. Compared to IFN low patients, IFN high were more likely to be of Eastern Asian ethnicity (45.0 vs 25.8%, p = 0.01), were younger at SLE onset (median [IQR] 27[18-35] vs 33 [25-47] years, p = 0.0001), and were more likely to be positive for anti-Ro (65.3 vs 34.7% p = < 0.0001), anti-La (29.2 vs 11.3% p = 0.01) and anti-RNP (36.1% vs 15.1% p = 0.007) antibodies. In patients with multiple samples, 87.3% had stable ISG status. In longitudinal follow up (median 630 [459-707] days), IFN high patients had higher disease activity (median [IQR] time adjusted mean SLEDAI2K (4.2 [2.8-5.7] vs 2.0 [1.3-4.2], p=< 0.0001), more flares (mild/moderate 53.5 vs 25.8%, p = 0.0003; severe 26.4 vs 6.5%, p = 0.0009) and spent less observed time in lupus low disease activity state (LLDAS) (median 55.5% [26.3-85.5] vs 84.0 [53.0-100]%, p = 0.0003). IFN high patients were more likely to have active arthritis (21.7 vs 6.5%, p = 0.007), skin disease (38.8 vs 14.5%, p = 0.007), leukopenia (26.4 vs 6.5%, p = 0.0009) and haematuria (20.2 vs 8.1%, p = 0.03). More IFN high patients required DMARD therapy (74.6 vs 55.9%, p = 0.02) and IFN high patients had higher prednisolone exposure (median time-adjusted mean dose 1.7 [0-5.7] vs 0.0[0-2.9] mg, p = 0.004). There was no difference in overall SLICC damage index, however there were differences in damage domains. IFN low patients had more malignancy (11.6 vs 2.4%, p = 0.01) and diabetes (13.3 vs 1.6%, p = 0.002). IFN high patients trended towards more mucocutaneous damage (19.2 vs 8.3%, p=0.08). However, in individual patients, there was no association between ISG level and any disease activity marker over time. 
Of 18 patients with varying IFN status over time, 12 had ISG fluctuations that were < 2SD from the mean fluctuation seen in stable ISG patients.  Of the remaining 6 patients with larger ISG fluctuations, 5 had high dose ( ≥25mg) prednisolone or cyclophosphamide therapy temporally associated with ISG reduction.

Conclusion: IFN high SLE patients had significantly more disease activity than IFN low patients. They were more likely to be of Asian than Caucasian ethnicity, had younger onset of disease and displayed more auto-antibodies. The majority of patients had stable IFN status, and major changes in ISG expression related to high dose immunosuppression.

« Back to ACR Convergence 2020

ACR Meeting Abstracts - https://acrabstracts.org/abstract/longitudinal-analysis-of-ifn-status-and-disease-characteristics-in-sle/