Could liver involvement be early detected in sickle cell disease?

Background: Liver damage is a severe and frequent complication in Sickle Cell Disease (SCD), mainly characterized by intra-hepatic cholestasis. So far, no effective approaches to prevent or treat this condition are established. Aims: Clinical, laboratory and imaging findings are evaluated longitudinally in SCD patients, comparing different sickle-genotypes, in order to identify possible early predictors of liver involvement. Methods: Sixty-eight SCD patients were studied: 17 Sickle Cell Anemia (SCA, median age 42.8±10.3yrs, M:F 4:13), 38 Sickle Cell Thalassemia (HbS-βThal, 45.2±9.4yrs, M:F 14:24) and 13 HbS/HbC (HbSC, 35.6±8.7yrs, M:F 5:8). Patients with at least two Stiffness data (Transient Elastography TE) (T0 and T1), measured out of sickle crisis, were retrospectively evaluated (2007-2016). Liver function tests, HBV, HCV, iron status, and hemolytic indices, were recorded. Abdominal ultrasound (US) and Magnetic Resonance Imaging (MRI) T2* were also collected. Results: In SCA pts Hb were 9±0.92 g/dL, HbS% 67.9±18.2% and HbF% 8.08±5.36; in HbS-βThal Hb 10±1.59, HbS 63.4±14.2% and HbF% 12.1±9.03; in HbSC Hb 11.9±1.1 (25.9-83.1), HbS 46.4±1% and HbF% 1.53±1.3 (SCA vs HbSC and HbS-βThal vs HbSC <0.0001). Considering clinical manifestations, 76.5% of SCA, 60.5% of HbS-βThal and 30.8% of HbSC pts had >1 vaso-occlusive crisis during the decade (VOCs)/yr (SCA vs HbSC p=0.02; SCA vs HbS-βThal p=0.36; HbS-βThal vs HbSC p=00.1). Occasional transfusions (<4 RBCs Units/yr) occurred in 88.2% of SCA, 84.2% of HbS-βThal and 61.5% of HbSC pts. HydroxyCarbamide was prescribed to 58.8% of SCA, 65.8% of HbS-βThal and 15.4% of HbSC pts and iron-chelators to 23.5% of SCA, 23.7% of HbS-βThal and none of HbSC pts. At T0 AST, ALT, LDH were statistically higher in SCA pts than in HbS-βThal and HbSC (ALT p<0.0001) and in HbS-βThal compared to HbSC (ALT p=0.01). GGT, ALP were higher in SCA than in HbS-βThal and HbSC (GGT p=0.013; ALP p=0.006), but without statistical significance in HbS-βThal compared to HbSC (GGT p=0.23; ALP p=0.44). Liver synthesis indices were similar in the three subgroups; none was neither HbsAg nor HCV-RNA positive. No differences were found comparing laboratory indices at T0 and T1. TE Stiffness was statistically higher in SCA (KPa 8.3±6.86) than in HbSC pts (KPa 5.33±2.15; p=0.014). In HbS-βThal (KPa 6.17±2.58) was increased but not statistically significant compared to either SCA and HbSC pts (p=0.2). Liver Iron Concentration (LIC) (derived from MRI T2*) was higher in HbS-βThal than in SCA and HbSC pts (HbS-βThal vs HbSC p=0.0145) and in SCA comparing HbSC (p=0.018). Univariate analysis was performed to correlate GGT with ferritin (p=0.02), TE (p=0.002), US (p=0.107) and LIC (p=0.511) in all SCD pts. A good correlation between GGT and US liver echogenicity was present in SCA and HbS-βThal pts, with GGT values respectively 20% and 160% higher than normal. TE and US (p=0.045) in all SCD pts correlated positively. No differences were found in TE and MRI T2* at T0 and T1. US showed significant differences at T0 compared with T1 in HbS-βThal (p=0.04) and in HbSC pts (p=0.001), but not in SCA pts (p=0.46) probably because of higher Stiffness since T0. Multivariate analysis showed as independent risk factors: sex (male), low HbF values, high ferritin values, more severe sickle genotype as predictors of liver involvement. Summary and Conclusions: Function liver tests associated with US, TE and when possible to MRI T2* taking into account sex, percentage of HbF, and SCD genotypes, are important to early detect and follow the sickle hepatopathy.