Identification by whole genome sequencing of genes associated with delayed postoperative hemorrhage in Scottish deerhounds

**Background:** Delayed postoperative hemorrhage (DEPOH) is a recognized health problem in sighthound breeds, including Scottish deerhounds and greyhounds. The condition is characterized by unexpected bleeding starting 36-48 hours after surgery, and prior work in greyhounds suggested reduced plasma alpha-2 antiplasmin activity as a potential mechanism. Antifibrinolytic drugs such as epsilon aminocaproic acid (EACA) have been shown to prevent DEPOH. This study aimed to identify genetic variants associated with DEPOH in Scottish deerhounds.

**Methods:** A retrospective case-control study was conducted using health surveys from owners of 269 Scottish deerhounds. DEPOH cases were defined as dogs with unexpected bleeding from a surgical wound starting 1-4 days after the procedure. Controls had undergone surgery without evidence of DEPOH and had not received prophylactic antifibrinolytic drugs. Of the 269 dogs, 10 cases and 62 controls met inclusion criteria and had DNA available. Whole genome sequencing was performed on a subset of 8 cases and 17 controls, with an average 25-fold genome coverage. Genome-wide association analysis (GWAA) was performed using an efficient mixed-model association approach with Bonferroni correction (P<0.05). All 72 dogs were genotyped for a missense SERPINF2 variant (c.605 C>T; p.A202V) using a custom TaqMan assay, and for F7 c.407 G>A by Sanger sequencing. Odds ratios and 95% confidence intervals were calculated.

**Key Results:** GWAA identified a single locus on chromosome 9 (41-47 Mb) spanning 40 genes that reached genome-wide significance (P<5×10⁻⁸) under an additive model. Of these, SERPINF2 (encoding alpha-2 antiplasmin) was identified as the most likely causal gene based on biological function. All 16 SERPINF2 variants were in complete linkage disequilibrium, forming a single haplotype. The nonsynonymous c.605 C>T variant (Ala202Val) was used as a haplotype marker. The T allele frequency was 80% in cases vs. 12% in controls (allelic OR=10; 95% CI=2.8-38; P=0.0004). None of the 10 cases had the C/C genotype, while none of the 62 controls had the T/T genotype. Compared to C/C dogs, the odds of DEPOH were significantly higher for C/T heterozygotes (OR=28; 95% CI=1.4-542; P=0.03) and T/T homozygotes (OR=1235; 95% CI=23-6752; P=0.0005). The F7 c.407 G>A variant showed no association with DEPOH (allelic OR=1.4; 95% CI=0.4-5.0; P=0.62).

**Clinical Implications:** This study provides strong evidence that SERPINF2 is associated with DEPOH in Scottish deerhounds. The very high odds ratios, particularly for T/T homozygotes, suggest that genetic testing could identify dogs at risk. Both heterozygous (C/T) and homozygous variant (T/T) dogs may benefit from prophylactic antifibrinolytic therapy before surgery. The lack of association with F7 c.407 G>A indicates that factor VII deficiency testing, while useful for other bleeding phenotypes, is not predictive of DEPOH. Limitations include the retrospective design, reliance on owner-reported outcomes, and the possibility that the identified variant is a marker rather than the causal variant. Additional studies are needed to confirm the functional effect of the Ala202Val substitution on alpha-2 antiplasmin activity and to determine whether this haplotype is predictive in other breeds such as greyhounds.