At a frequency of 18.5%, AFA is the second most common anti-nuclear antibody among AA patients with SSc (second to ATA). The present report is the first study investigating the genetic associations, clinical manifestations, and survival impact of AFA in a large population of AA SSc patients.
Distinct HLA class-II allelic associations of SSc-specific autoantibodies in different ethnic groups have been demonstrated in several studies (5
). In a large sample of Caucasian patients, we have previously reported that the HLA DRB1*13:02
haplotype correlated with AFA (14
). In the current study, we did not observe a similar pattern among AA patients with AFA. Our results indicated that HLA DRB1*08:04
is strongly associated with AFA in AA patients with SSc, compared to either unaffected individuals or AFA negative AA patients with SSc.
Previous studies investigated potential association of HLA DRB1*08:04
with other rheumatic conditions like SLE (29
) and RA (30
). Reveille et al.
detected no difference in frequency of HLA DRB1*08:04
between 88 AA patients with SLE or 88 unaffected AA controls. In another study by Hughes LB et al.
no difference was reported in frequency of HLA DRB1*08:04
between 321 AA patients with RA and 564 unaffected individuals (30
). Previously, we showed that HLA DRB1*08:04
might be a susceptibility gene for SSc among AA (14
); while the results of the current study demonstrated that the higher frequency of HLA DRB1*08:04
with SSc in AA patients is mainly driven by its strong association with AFA in this ethnic group. Furthermore, through the Binding Peptide Prediction Server for HLA DRB1*0804
, we identified four potential binding peptides from the human fibrillarin protein that could serve as potential binding sites within the antigen binding groove. The large effect sizes ( and supplement table 1
) and predicted binding peptides should prompt more studies to investigate their potential causal and/or environmental relationship of these autoantibodies.
In this regard, an animal model for induction of AFA has been extensively studied and may provide clues to an environmental trigger in humans with AFA-positive SSc. Certain mouse strains possessing specific H2 (the murine counterpart for HLA) haplotypes develop a non-SSc autoimmune disease and high titer AFA following administration of mercuric chloride or silver nitrate (31
). Of note, one study of urinary mercury levels in SSc patients noted higher levels in those with AFA. However, this observation did not maintain statistical significance following corrections (35
). Interestingly, heavy metals have previously been noted to be highly concentrated in the nucleolus (36
). It was noted by Pollard et al.
) that most if not all of the SSc-specific autoantigens were at sometime during their life cycle localized to the nucleolus. Clearly, larger and more targeted studies of heavy metal and other environmental exposures in AFA-positive SSc patients, perhaps selected for the associated HLA-class II alleles (DRB1*13:02
in Caucasians and DRB1*08:04
in AAs) and AFA negative SSc patients, as well as well-matched normal controls are warranted.
Confirming our previous findings (10
), we showed that HLA DQB1* 03:01
has higher frequency among AFA positive AA patients compared to AA unaffected individuals. However, there is no difference between AFA positive and negative AA patients with SSc.
AA SSc patients are younger at SSc onset compared to other ethnic groups (1
). Moreover, other studies have shown that SSc patients with AFA have younger age of onset (2
). In support of these findings, we further demonstrated that AA SSc patients with AFA had younger age of onset in comparison to AA patients without AFA.
Higher Medsger peripheral vascular severity index and prevalence of digital ulcers in AA patients with AFA compared to those without AFA are novel findings. These findings were also present when we compared AFA vs.
ATA positive AA patients with SSc. Previous studies have shown higher rate of digital ulcers among AA patients with SSc compared to Caucasians (2
). Higher frequency of AFA in AAs might contribute to this finding. In another study, Steen et al
noted higher frequency of digital ulcers in AFA positive patients (12
). However, these findings were not stratified for ethnic background.
In agreement with previous studies reporting more severe GI involvement in AFA positive patients (regardless of ethnicity) (10
), we observed a higher frequency of SSc-associated diarrhea in AFA positive AA patients. The higher frequency of lower GI tract involvement was more significant when AFA positive AA patients were compared to ATA positive patients. It is possible that AFA positive patients have more severe lower GI tract hypomotility and bacterial overgrowth that contribute to diarrhea.
Our results imply a less severe lung involvement among AFA positive AA patients with SSc, as assessed by lower Medsger lung severity index. The comparison of AFA vs. ATA positive AA scleroderma patients further demonstrated less severe lung involvement (higher FVC and DLCO % predicted values and lower Medsger lung severity index). In agreement with our findings, in an ethnically homogenous cohort of Japanese patients with SSc, AFA positive patients had less severe lung involvement (17
). While data from several multi-ethnic cohorts suggested a higher frequency of isolated PAH and/or pulmonary fibrosis in the SSc patients with AFA (9
), these comparisons were adjusted neither for ethnicity nor for other antibodies i.e. ATA, as potential confounders. Therefore, the higher frequency of lung fibrosis and PAH might be due to a sizeable AA population in AFA positive group and large number of Caucasian SSc patients in AFA negative group. More severe SSc-associated lung involvement in AA patients with SSc compared to other ethnic groups has been reported in several studies (2
Based on the current findings, AFA positive AA patients with SSc have a higher prevalence of pericarditis, compared to AFA negative as well as and ATA positive patients. This is in agreement with former studies indicating higher frequency of cardiac involvement in AFA positive (10
) and AA patients with SSc (39
Our study did not confirm previous reports of worse (13
) or better (11
) survival in AFA positive patients with SSc. The worse survival of AFA positive patients in a previous report (13
) might be attributable to the confounding or modifying effects of ethnicity in studies that are not stratified by ethnicity. As AA ethnicity is associated with AFA positivity as well as poorer survival (1
This study has some limitations. Although potentially important, the data on heavy metal exposure were not collected in the current study. The Medsger severity indices were only available in the patients from the longitudinal GENISOS cohort. High-resolution computed tomography scan (HRCT) and echocardiography were not performed on all patients, which might have lead to underreporting of pulmonary involvement, despite being the largest genetic study ever reported in AA patients with SSc, we might be underpowered to detect more subtle HLA associations with AFA in AA population.
In conclusion, AFA was the second most common anti-nuclear antibody in African Americans with SSc. Presence of AFA was strongly associated with the HLA DRB1*08:04 in the AA SSc patients. In addition, AA SSc patients with AFA had a younger age of disease onset, higher frequency of digital ulcer and pericarditis, more severe lower GI involvement, and less severe pulmonary involvement. Studies in the future should focus on environmental factors, such as heavy metal exposure, that may influence the B cell response and the immunopathology of the disease.