%0 Journal Article %A J A Livesey %A J E Jackson %A C L Shovlin %T S99 Multiple regression analyses in a cohort of hereditary haemorrhagic telangiectasia patients suggest a novel role for iron in thrombosis %D 2010 %R 10.1136/thx.2010.150938.50 %J Thorax %P A46-A46 %V 65 %N Suppl 4 %X Introduction/objectives Pulmonary emboli (PE) secondary to deep venous thromboses (DVT) are the immediate cause of 10% of hospital deaths. Elevated basal plasma coagulation factor (F) VIII levels, now known to be synthesised by the pulmonary endothelium,1 are the strongest predictor of recurrent PE/DVT, and also associated with elevated pulmonary artery pressure (PAP). The reasons for these associations are unknown. The study objectives were to determine whether novel patient-specific factors were associated with FVIII levels; PE/DVT; and/or PAP.Methods Previous studies from our group have presented data on the 1999–2005 hereditary haemorrhagic telangiectasia (HHT) cohort of patients. In the current study, a replicate data series was generated from individuals seen between 2006 and 2010, and analysed independently to the first. Basal plasma FVIII measurements taken at least 4 months from any thrombotic event/intercurrent illness; and 32 patient-specific variables from 221 individuals, including catheter-measured PAP and verified DVT/PE, were entered into an Excel chart. Following verification, data were analysed using STATA v11c (Statacorp LP) software. Univariate analyses identified potentially associated variables, which were analysed using stepwise multiple regression analyses, to generate best-fit models.Results For FVIII, the most significant independent variables were identified as fibrinogen, age, and iron replacement therapy (IRT) (model: 4df; p<0.0001; r2 0.2014). Fibrinogen showed the most significant association with FVIII (OR 0.193, p<0.001). FVIII showed further associations with age (OR 0.006, p=0.028) and IRT (OR 0.162; p=0.075). PE/DVT was associated with IRT (OR 1.416, p=0.028). Mean PAP was associated with age (OR 0.079, p=0.007) and FVIII (OR 1.449, p=0.009). Iron deficiency did not emerge as significant in any model. SIREs software predicted specific iron response elements (IREs) in FVIII mRNA splice-variants, including one in the 5′UTR of splice-variant 2.1Conclusions This study suggests that basal plasma FVIII levels, and associated clinical endpoints of PE/DVT and PAP, may be influenced by iron replacement therapy, possibly by modulating expression of specific splice-variants of the F8 gene. This study therefore suggests that iron replacement may increase thrombotic risk, with significant potential clinical implications for millions of people who take iron tablets. %U https://thorax.bmj.com/content/thoraxjnl/65/Suppl_4/A46.1.full.pdf