Abstract

Sickle cell anemia (SCA) is characterized by end organ damage, especially in the kidney and heart, resulting in a high prevalence of chronic kidney disease (CKD), tricuspid regurgitation and left ventricular hypertrophy (LVH). In the general population, dysregulation of calcium/phosphate homeostasis by elevated plasma fibroblast growth factor 23 (FGF23) directly promotes LVH in end stage CKD without SCA. Accordingly, the high mortality associated with end stage renal disease is primarily cardiac, and is associated with ventricular remodeling and elevated plasma FGF23. However, FGF23 has not been investigated as a biomarker for CKD, LVH and mortality in SCA. Here, we assessed the relationships between CKD, LVH, plasma FGF23 and mortality in adults with SCA.

First, we determined that LVH is a common entity in SCA-CKD using a nested case control study of all SS subjects with stage 3-5 CKD (eGFR<60) enrolled in the Bethesda Sickle Cell Cohort Study (ClinicalTrials.gov identifier: NCT00011648) and age and sex-matched SCA subjects with normal renal function. Of those with SCA-CKD (n=42), LVH defined by echocardiogram was present in 74% compared to 41% in those with normal kidney function (odds ratio 3.79; 95%CI 1.57 - 10.06; p=0.0039).

We next measured plasma FGF23 levels in normal volunteers (n=37) and age and sex-matched SCA subjects with and without CKD (n=37), using ELISA. We observed a 3.5-fold increase in plasma FGF23 levels among subjects with SCA (p<0.0001) and a 22-fold increase in those with SCA-CKD (p<0.0001). FGF23 levels are elevated at baseline in adults with SCA which has not previously been reported. The much higher fold increase in SCA-CKD was not unexpected, based upon previously reported high plasma FGF23 associated with CKD in the absence of SCA. Spearman correlations between plasma FGF23 levels in adults with SCA were correlated with creatinine (r=0.52, p=0.0011), eGFR (r=-0.57, p=0.0003), left ventricular mass index (LVMI, r=0.48, p=0.0029), and tricuspid regurgitant velocity (TRV; r=0.44, p=0.01). These data suggest that as in CKD without SCA, plasma FGF23 is a biomarker associated with both a decline in renal function and an increase in the prevalence of LVH and tricuspid regurgitation.

We next sought to determine if plasma FGF23 levels are a biomarker associated with mortality in the overall SCA patient population. Univariate Kaplan-Meier analyses showed that categorical FGF23 levels (above versus below log mean; hazard ratio=1.74, 95% confidence interval 1.09-2.85, p=0.02] and elevated TRV (≥2.5 m/s versus all others, hazard ratio=4.42, 95% confidence interval 2.60-7.52, p<0.0001) are significantly associated with increased mortality in 478 adults with SCA. LVH defined by sex adjusted echocardiographic thresholds for LVMI was not associated with an increased risk of death (p=0.71). Multivariable Cox proportional hazard analysis indicated that log FGF23 is an independent risk factor for death (hazard ratio=1.46, 95% confidence interval 1.01-2.10, p=0.04), in addition, to the known mortality association with elevated TRV (hazard ratio=4.90, 95% confidence interval 2.86-8.39, p<0.0001). The Cox analysis suggests that plasma FGF23 is a biomarker for mortality in SCA, although the echocardiographic abnormality associated with death in this model is TRV and not LVMI.

In conclusion, we observed a high prevalence of LVH in adults with SCA-CKD (70.3%), and SCA-CKD is associated with a 22 fold increase in plasma FGF23 levels. One limitation of our study was the inability to assess associations between plasma FGF23, LVMI, TRV and mortality risk in SCA-CKD due to a limited number of CKD subjects. When FGF23 was investigated in all adults with SCA, plasma FGF23 was a significant risk factor for death, independent of TRV and other known risk factors. Larger SCA-CKD studies will be necessary in the future to better understand the relationships between TRV, LVH, FGF23 and mortality. Overall, plasma FGF23 is a potentially helpful biomarker to further elucidate the pathobiology of SCA-CKD end organ complications.

Disclosures

No relevant conflicts of interest to declare.

Author notes

*

Asterisk with author names denotes non-ASH members.