Session Information
Session Type: ACR Concurrent Abstract Session
Session Time: 4:30PM-6:00PM
Background/Purpose:
Cardiac manifestations of neonatal lupus, comprising atrioventricular
conduction defects and cardiomyopathy, occur in fetuses exposed to anti-Ro/SSA
antibodies, and carry substantial mortality. There is strong evidence of a
genetic contribution to the risk (sibling
recurrence risk ratio is ~3,000).
Previously, in a case-control genome-wide association study, we reported that
the 17 strongest associations were within the Human Leukocyte Antigen (HLA)
region. Meisgen (2013) further underscored the potential importance of the HLA
region when they observed two-digit classic HLA associations using family-based
tests. Here, we report the first large-scale investigation of four-digit
classic HLA associations in mothers with cardiac neonatal lupus (CNL) diagnosed
children.
Methods:
Parent-CNL-affected-offspring (n = 182) trios of European ancestry were
genotyped using the Illumina Immunochip single nucleotide polymorphism (SNP) array
and merged with 1073 non-lupus out-of-study controls (635 female). Admixture
estimates were computed using ADMIXTURE. Classic four-digit HLA alleles in HLA-A,
-B, -C, -DPB1, -DQA1, -DQB1, and –DRB1 were imputed using high-quality SNPs
and the software HIBAG. Analyses were computed on HLA alleles with a best guess
allele count of at least 10 in cases or controls. Association analyses of CNL
mothers vs non-lupus female controls were computed using logistic regression,
adjusting for admixture. Transmission/disequilibrium tests (TDT) were computed using
best guess HLA alleles to test for differential transmission to affected
offspring.
Results:
A total of 119 unique HLA alleles met quality control and allele frequency
standards. Six HLA alleles met genome-wide significance: A*01:01(3.10, 1.47×10-12),
B*08:01(7.40, 2.56×10-24), C*07:01(4.28, 1.27×10-17),
DQA1*05:01(8.22, 8.80×10-26), DQB1*02:01(8.47, 6.03×10-26),
and DRB1*03:01(8.91, 1.03×10-26) and four additional alleles met an
HLA bonferoni threshold of 0.05/119=0.00042: DQA1*01:01(0.15,1.35 x10-5),
DQA1*03:01(0.15, 1.00 x10-4), DQB1*05:01(0.21, 1.41×10-5),
and DRB1*01:01(0.11, 9.61×10-5). To consider the joint effects
across HLA genes, stepwise modeling was computed (Table).
|
|
Allele Frequency |
Single-allele |
Stepwise |
||
|
HLA-allele |
Cases |
Controls |
P-value |
OR (94% CI) |
P-value |
|
B*08:01 |
0.382 |
0.121 |
2.56E-24 |
2.50 (1.41-4.45) |
1.84E-03 |
|
DQB1*04:02 |
0.037 |
0.022 |
1.65E-01 |
6.32 (2.49-16.00) |
1.02E-04 |
|
DQB1*06:02 |
0.189 |
0.132 |
3.60E-03 |
4.00 (2.57-6.23) |
8.71E-10 |
|
DRB1*03:01 |
0.408 |
0.121 |
1.03E-26 |
7.98 (4.26-14.96) |
8.94E-11 |
|
DRB1*11:01 |
0.061 |
0.046 |
2.42E-01 |
5.01 (2.17-11.60) |
1.66E-04 |
None of
the HLA alleles in the multilocus model exhibited differential transmission to
offspring via a TDT (p>0.15). In addition, none of the 119 HLA alleles
showed statistically significant evidence of differential transmission to the
offspring after adjusting for multiple comparisons.
Conclusion:
These analyses demonstrate the importance of the HLA region in mothers of CNL
children. However, the four-digit classic HLA allele associations found in the
mothers of CNL children were not differentially transmitted to the offspring.
These results suggest that the maternal HLA profile contributes to an
environment (e.g., anti-Ro antibodies) that increases the risk for CNL.
To cite this abstract in AMA style:
Ainsworth HC, Langefeld CD, Marion MC, Costedoat-Chalumeau N, Brucato A, Buyon JP, Clancy RM. HLA Associations in Mothers of Children with Cardiac Manifestations of Neonatal Lupus [abstract]. Arthritis Rheumatol. 2015; 67 (suppl 10). https://acrabstracts.org/abstract/hla-associations-in-mothers-of-children-with-cardiac-manifestations-of-neonatal-lupus/. Accessed .« Back to 2015 ACR/ARHP Annual Meeting
ACR Meeting Abstracts - https://acrabstracts.org/abstract/hla-associations-in-mothers-of-children-with-cardiac-manifestations-of-neonatal-lupus/