Background/Purpose: Scleroderma renal crisis (SRC) is a severe complication
of systemic sclerosis (SSc). Most SSc cases demonstrate a disease-specific
antinuclear antibody including anti-RNA polymerase III (ARA), anti-fibrillarin (AFA),
anti-topisomerase-1 (ATA) or anticentromere (ACA). ARA defines a distinct sub-phenotype characterised by diffuse skin
disease and risk of complications including SRC and pulmonary arterial
hypertension. We used the strong association between ARA and SRC and the predominant
occurrence of SRC early in disease to develop an extreme phenotype strategy for
defining genetic factors in susceptibility to renal crisis.

Methods:
First we analysed SRC in a well characterised group of SSc cases followed at
our centre (n=415). Informed by this, 50 patients with confirmed SRC and ARA+
and another 50 SSc ARA+ that had never developed SRC were identified from our
larger SSc cohort. These cases, all with Northern European ancestry, were
genotyped across approximately one million SNPs using the Illumina Human
Omni-express bead array chip. All data underwent quality control checks for
Hardy-Weinberg equilibrium and genotyping rate in PLINK (HWE p < 0.001, and
genotyping rate > 90%). After filtering of SNPs, a logistic regression was
performed in PLINK comparing patients with or without SRC to determine the
genetic signature difference between these two groups of patients.

Results:
Initial analysis confirmed a strong association between ARA and SRC with 30%
ARA+ SSc developing SRC compared with 7% AFA, 4% ATA and 1% ACA (p < 0.001).
Moreover in ARA+ cases almost all SRC occurred within 18 months of disease
onset, and SRC after 5 years of follow up was very rare. Thus we could define a
group with SRC and another at very low risk. This dichotomy formed the basis of
our genetic analysis comparing ARA+ patients with SRC history (Group A) to the
control group, who had been followed for > 60 months without SRC (Group B).
We performed GWAS analysis on these two groups. Quality control checks removed
2309 SNPs for missingness (GENO > 0.1) and 77122 failed MAF filters (MAF
< 0.01). In total 641,489 SNPs were analysed. The logistic regression
analysis identified a number of SRC associated SNPs within genes and gene
regions. Top associations were found in the complement region (P= 1.66×10-5),
and in other genes including EPHA5 (P= 1.87×10-5), GRIA3 (P= 2.16×10-5), HECW2
(P= 2.71×10-5) and CTNND2 (P= 2.92×10-5).

Conclusion:
We present a novel study using extreme phenotypes of ARA+ SSc to identify
genetic association of SRC in cases that are serologically and clinically
otherwise homogeneous. We identified genes including Caterin
cadherin-associated protein delta 2 (CTNND2) which is known to regulate
adhesion molecules relevant to fibrosis. Genes identified from this analysis
may have general relevance to SSc vasculopathy or other forms of hypertensive
thromobotic microangiopathy. Additional functional and genetic replication
studies are needed.