Increased alloreactive and autoreactive antihuman leucocyte antigen antibodies associated with systemic lupus erythematosus and rheumatoid arthritis

Abstract

Objectives: Rheumatoid arthritis (RA) and systemic lupus erythematosus (SLE) disproportionately affect women during and following childbearing years. Antihuman leucocyte antigen (HLA) alloantibody responses are common in healthy parous women, and as these diseases are both linked with HLA and immune dysregulation, we sought to evaluate anti-HLA antibodies in RA and SLE.

Methods: Anti-HLA antibodies were measured among parous SLE cases (n=224), parous RA cases (n=202) and healthy parous controls (n=239) and compared with each other as well as with nulliparous female and male controls. Antibody specificities were identified and compared against subject HLA types to determine autoreactivity versus alloreactivity. The association of anti-HLA antibodies with clinical outcomes was evaluated.

Results: Levels and frequencies of anti-HLA antibodies were significantly higher among parous females with SLE (52%) or RA (46%) compared with controls (26%), and anti-HLA antibodies were also found among nulliparous females and males with SLE and RA. Autoreactive anti-HLA antibodies were observed among SLE and RA antibody-positive subjects, but not healthy controls, with the highest frequency of autoreactive anti-HLA antibodies found in the SLE subjects. Higher levels of anti-HLA antibodies were associated with nephritis among the nulliparous SLE cases (p<0.01). The presence of anti-class I HLA antibodies was associated with younger age at diagnosis among both the RA and SLE nulliparous cases.

Conclusions: Both autoreactive and alloreactive anti-HLA antibodies were found at high levels in RA and SLE subjects. These occurred even in the absence of alloexposure, particularly among SLE subjects and may be linked with disease severity.

Keywords: alloantibodies; autoantibodies; human leukocyte antigen (HLA); rheumatoid arthritis (RA); systemic lupus erythematosus (SLE).

Figure 1

Increased anti-HLA antibodies among SLE and RA subjects. (A) Class I and (B) class II anti-HLA antibody NBG ratios are plotted by group for healthy males (n=1148), healthy nulliparous females (n=1757), healthy parous females (n=239), males with SLE (n=48), nulliparous females with SLE (n=48), parous females with SLE (n=224), males with RA (n=48), nulliparous females with RA (n=48) and parous females with RA (n=202). Dashed lines indicate cut-offs used to identify antibody positive samples. Boxes display IQR with median indicated with a line. Whiskers display fifth to 95th percentiles. SLE, RA and healthy controls were compared by exposure group (eg, males, nulliparous females and parous females) by analysis of variance with Dunnett’s post-test used to compare each to the matched healthy control. (C) The frequency of class I and (D) class II anti-HLA antibodies is plotted by the same groups as above. Frequencies of anti-HLA antibodies were compared between male, nulliparous female and parous female groups among SLE, RA or healthy subjects and between healthy, SLE and RA subjects for the parous females by χ² test. ***P<0.001, ****P<0.0001, ^nsP>0.05. HLA, human leucocyte antigen; NBG, normalised to background; SLE, systemic lupus erythematosus; RA, rheumatoid arthritis.

Figure 2

Presence of autoreactive anti-HLA antibodies among SLE and RA subjects. (A) A subset of antibody positive parous female SLE and RA subjects was selected for antibody specificity screening. Selected samples are plotted side by side with their parent populations to compare NBG ratio distributions for class I (left panel) and class II (right panel) anti-HLA antibodies. Samples with autoreactive anti-HLA antibodies are indicated in red. Dashed lines indicate cut-offs used to identify antibody positive samples. (B) The frequency of one or more autoreactive class I (left panel) or class II (right panel) antibodies is plotted for parous female subjects by disease category—healthy, SLE or RA. Three increasingly stringent cut-offs were used to detect particular antibody specificities (X2 is the least stringent and X6 the most stringent). Groups were compared using the X6 cut-off by χ² test. (C) The same analysis was used as in (B) above, this time evaluating the specificities of the male, nulliparous female and parous female SLE subjects. These three groups were compared for the X6 cut-off by χ²-test. **P<0.01, ****P<0.0001, ^nsP>0.05.

Figure 3

Distribution of autoreactive and alloreactive epitopes of anti-HLA antibodies. The number of anti-class I and anti-class II HLA epitopes detected at the X6 cut-off are plotted by subject with autoreactive epitopes shaded black. Each bar represents the distribution of epitopes for an individual subject. HLA, human leucocyte antigen; SLE, systemic lupus erythematosus; RA, rheumatoid arthritis.

Figure 4

Associations between HLA antibodies and clinical outcomes (A) Plots showing the class I or class II antibody max NBG ratios for parous female and combined nulliparous SLE cases with and without lupus nephritis. (B) Plots showing the class I or class II antibody max NBG ratios for parous female and combined nulliparous RA cases with and without erosions. Dashed lines indicate cut-offs used to identify antibody positive samples. **P<0.01. HLA, human leucocyte antigen; NBG, normalised to background; SLE, systemic lupus erythematosus; RA, rheumatoid arthritis.

Figure 5

Associations between the HLA antibodies and age at diagnosis. Plots showing age of diagnosis for parous female and combined nulliparous SLE (A) or RA (B) cases with and without class I or class II HLA antibodies. *P <0.05, **P <0.01. HLA, human leucocyte antigen; SLE, systemic lupus erythematosus; RA, rheumatoid arthritis.