Therefore, our findings may be attributable to age-related liver fibrosis which does not manifest as decreased L:S on CT. Disclosures: Claude B. Sirlin – Advisory Committees or Review Panels: Bayer; Grant/Research Support:

XL765 in vitro GE, Pfizer, Bayer; Speaking and Teaching: Bayer Rohit Loomba – Consulting: Gilead Inc, Corgenix Inc, Janssen and Janssen Inc; Grant/Research Support: Daiichi Sankyo Inc, AGA, Merck Inc The following people have nothing to disclose: Kathleen Jacobs, Sharon S. Brouha, Ricki Bettencourt Liver iron overload, measured histologically or using serum ferritin (SF) levels, is associated with NAFLD severity. In this study we evaluated the association of hepatic iron measured using T2* MRI, and disease severity in NAFLD. Patients (n=60; 38 male) having a liver biopsy for suspected NAFLD were recruited to have a MR scan for the quantification of liver fat (proton magnetic resonance spectroscopy; 1H-MRS) and liver iron (T2* mapping). Liver biopsies were assessed for fibrosis (Ishak stage; 0-6), steatohepatitis (NAS score; 0-8) and iron deposition (Perl’s staining; 0-4). SF was measured at the time of the MR study (available in 57 cases). A T2* cut off of 19ms (corresponding to

a liver iron concentration of 1.3mg/g) was CRM1 inhibitor used to stratify patients into 3 groups: (a) Normal Iron (T2*>19ms, and a negative Perl’s stain; n=20), (b) High MR iron (T2*<19ms) and negative Perl's stain (High MR iron-Perl's neg; n=29); and (c) High MR iron (T2*<19ms)

and positive Perl’s stain (High MR iron-Perl’s pos; n=11). T2* was negatively associated with SF (r=−0.67; p<0.0001) and liver fat measured by 1H-MRS (r=−0.65; p<0.0001). SF was respectively, >1.5 times the Upper Limit of Normal (>1.5×ULN) in 0%, 22% and 91% of those with Normal Iron, High MR iron-Perl’s neg, and High MR iron-Perl’s pos. The median ferritin concentrations for patients with Normal Iron, High MR iron-Perl’s neg and High MR iron-Perl’s pos were 67μg/L, 194μg/L and 1104ng/L (p<0.0001) respectively. The mean 1H-MRS liver fat in those with Normal Iron, High MR iron-Perl's neg and High MR iron-Perl's pos were 6.6%, Olopatadine 19.1% and 29.8% (p<0.0001) respectively. A diagnosis of definite NASH (NAS>5) was made in 30%, 69% and 64% (p=0.02) of patients with Normal Iron, High MR iron-Perl’s neg and High MR iron-Perl’s pos, respectively. Patients who had iron overload by MR criteria only (T2*<19ms, Perl’s negative, SF<1.5×ULN) were also compared to patients with Normal Iron. Overall, 21 (35%) patients had liver iron overload that could only be detected by MRI, and significantly more of these (71%) had definite NASH, compared to those with Normal Iron (26%; p=0.004). In conclusion, our study demonstrates that MRI is more sensitive than histology or SF concentration in identifying liver iron overload, and may improve the risk stratification of patients with NAFLD.