Longitudinal Study to Find the Association of Serum Phosphorus Level with FGF23 in Three Different Hyperphosphatemia Management Groups of Stage 3 and 4 Chronic Kidney Disease (CKD) Patients

Introduction

Chronic kidney disease (CKD) is a global public health problem affecting 10%–16% of the world population with the stage-wise prevalence being stage 1, 13.4%; stage 2, 9.9%; and late stages (3–5), 4.5%.[12] Mineral bone disease develops at stage 3 due to impaired conversion of 25-(OH)D to 1,25-(OH)2D, reducing intestinal calcium absorption and increasing parathyroid hormone (PTH) level. Along with PTH, fibroblast growth factor (FGF23) reduces renal tubular reabsorption of filtered phosphorus in response to changes in gastrointestinal phosphorus uptake, thereby preventing major changes in serum phosphorus levels.[3] As CKD advances, there is progressive deterioration in mineral homeostasis, leading to hyperphosphatemia, hypocalcemia, hyperparathyroidism, and increase in FGF23. At this stage, composite measurements of serum phosphorus and FGF23 may become an integral and improved approach in therapeutic decision-making.[4]

It is unclear whether actual serum phosphorus levels and other markers of disordered bone and mineral metabolism (e.g., serum calcium, calcium–phosphorus product) are associated with progression of kidney disease in patients with CKD. Limited data suggest that phosphate retention may cause harm in the pre-dialysis stages. There is a huge gap with regard to prospective clinical trials assessing the effect of phosphate-lowering strategies on clinically meaningful endpoints. As per KDIGO 2017, it has been suggested that in patients with CKD G3a to G5D, individual values of serum calcium and phosphate, evaluated together, be used to guide clinical practice.[5] Further, in these patients, it is suggested that phosphorus has to be reduced toward normal range, but the exact value and its benefit are unclear (2C).[5] Hence, the present 1-year longitudinal study was planned to find the association of serum phosphorus level with FGF23 in three different hyperphosphatemia-treated CKD stage 3 and 4 patient groups.

Material and Methods

Screening and randomization

All CKD patients of age more than 18 years attending the outpatient department were screened for serum creatinine to determine the estimated glomerular filtration rate (eGFR) and serum phosphorus levels, keeping in mind the inclusion and exclusion criteria of the study. Out of 330 screened patients, a total 150 patients were randomized initially, but only 120 patients completed 1 year. These patients were subjected to computer-based randomization with an allocation ratio of 1:1:1 in threegroups equally [Figure 1] as follows: group 1(n = 40; normal 700–800 mg/day phosphorus diet): this group was subjected to normal phosphorus diet during thefirst 6 months, followed by low-phosphorus diet in the next 6 months; group 2 (n = 40) patients were given calcium-based phosphate binder (starting dose was tab calcium acetate 667mg BD) along with a low-phosphorus diet, that is, 500–700 mg/day; and group 3 (n = 40) patients were given non–calcium-based phosphate binder sevelamer (starting dose was 400 mg BD) with a low-phosphorus diet. Throughout the study, all patients were maintained on a constant dietary phosphorus intake by telephonic recall.

Close

Figure 1

Box plot showing mean iPTH/ FGF 23 at baseline among 3 groups

Export to PPT

Definitions

1. Stages of CKD: Stage 1: eGFR ≥ 90 ml/min/1.73 m²; stage 2: 60–89 ml/min/1.73 m²; stage 3a: 45–59 ml/min/1.73 m²; stage 3b: 30–44 ml/min/1.73 m²; stage 4: 15–29 ml/min/1.73 m²; stage 5: <15 ml/min/1.73 m² or end-stage renal disease.[5]

2. Bone disease in CKD: Presence of all the parameters including hyperphosphatemia (>5mg/dl), PTH >150pg/ml, FGF23 >50pg/ml, and vascular calcification was considered as bone disease.[56]

3. eGFR was calculated using the MDRD 1 formula as follows:[8] eGFR = 186 × (serum creatinine[mg/dl])−1.154 × (age [in years])−0.203 × (0.742 if female).

4. Hypertension (HTN): HTN was defined as per the JNC-8 guideline. Systolic blood pressure (SBP) ≥140 mmHg and diastolic blood pressure (DBP) ≥90 mmHg were used as the cut offs in patients with HTN, diabetes mellitus (DM), and CKD regardless of age and in adults younger than 60 years. In people of age 60 years or older, the cut off was equal to or more than 150/90.[9]

Results

Baseline demographic profile

The mean age of 120 patients who completed the 1-year study period was 52.51 ± 15.5 years and there was a slightly male preponderance (59.2%). All the clinical parameters were comparable among all three study groups [Table 1]. Majority of the patients were in CKD stage 3b and 4. Total protein and phosphorus intake were significantly higher in group 1, whereas total calories were the same. All the major biochemical and haematological parameters at baseline were comparable among the three groups, [Table 1]. The average serum calcium and phosphorus levels were reasonably well controlled and the PTH level was high in all the groups. The mean FGF23 levels were also similar among all groups [Figure 2].

Close

Figure 2

Flow Diagram

Export to PPT

Table 1 Baseline Demographic and radiological parameters

Parameters	Group 1 (n=40) Mean±SD/%	Group 2 (n=40) Mean+SD/%	Group 3 (n=40) Mean±SD/%	P
	Age (in years)	54.50±13.41	51.93±16.26		51.10±16.80	0.60
Mean BP (mm Hg)	102.59±14.77	102.43±11.15	99.79±11.61	0.50***
Gender (male)	25 (62.5%)	23 (57.5%)	23 (57.5%%)	0.87**
Smoking (Smoker)	3 (7.5%)	1 (2.5%)	0 (0%)	0.12 *
Alcohol (Alcoholic)	1 (2.5%)	1 (2.5%)	0 (0%)	0.70*
CKD Stage
3a	15 (37.5%)	7 (17.5%)	12 (30.0%)	0.206**
3b	7 (17.5%)	14 (35.5%)	13 (32.5%)
4	18 (45.0%)	19 (47.5%)	15 (37.5%)
Diabetic’s	15 (37.5%)	16 (40.0%)	12 (30.0%)	0.624**
Hypertensive’s	32 (80.0%)	35 (87.5%)	29 (72.5%)	0.245**
H/o CVS disease	3 (7.5%)	5 (12.5%)	2 (5.0%)	0.601 *
H/o CVA	0	2 (5.0%)	1 (2.5%)	0.772 *

*Fisher’s exact test,**-Chi square, ***-kruskalwallis test, CKD- chronic kidney disease, CVS- cardiovascular disease, CVA- cerebrovascular disease. Data presented as Mean±SD or number (percentage)

Relationship of FGF23 and phosphorus level

Based on FGF23, iPTH, and vascular calcification, patients in all the groups were further stratified into category 1, 2, and 3 according to the serum phosphorus levels of upto 3.9, 4–5, and >5mg/dl, respectively [Figure 3]. It was found that there was no significant relationship of FGF23 with different target values of serum phosphorus (P > 0.05).

Close

Figure 3

Bar diagram showing relation of different phosphorus levels with progression of disease

Export to PPT

Discussion

In the present study, we assessed the target serum phosphorus level and its association with FGF23 in three different hyperphosphatemia management groups of stage 3 and 4 CKD patients. All three study groups, namely, phosphorus-restricted diet, calcium-based phosphorus binder, and non–calcium-based phosphorus binder, showed significant phosphorus reduction over 1 year follow-up. There was significant decline of FGF23 in non–calcium-based phosphate binder group, but not in the diet- and calcium-based phosphate binder groups. Serum phosphorus at the level of 5 mg/dl was associated with iPTH at baseline and with eGFR at 1 year; however, it was not associated with FGF23. Further, serum FGF23 levels were significantly correlated with iPTH and correlated negatively with eGFR. The latter may be pointing to a disrupted feedback loop resulting in high levels of serum FGF23.

The present study comprised middle-aged patients with slight male predominance. Age and gender did not show any correlation with FGF23 as per the literature, which also showed no age-related changes in intact FGF23 or in C-terminal FGF23.[10] Here, eGFR was on higher side (34.38 ± 11.84ml/min) compared to the Chronic renal Insufficiency Cohort (CRIC) study population's eGFR (32 ± 15 ml/min/1.73 m²).[11] Whether FGF23 contributes to CKD progression or is only a marker of its risk remains an important and unresolved question. However, some authors believe that FGF23 can be a novel predictor for the progression of CKD.[12] Similar to the previous study, the diabetic population formed 35.8% of the patients in our study.[11] It has been suggested by Titan et al.[13] that serum FGF23 is related to the risk of CKD progression in macroalbuminuric diabetic nephropathy (DN). But it is still unclear whether this relationship is causal. The baseline mean serum phosphorus levels were similar to a previous study in which the serum phosphorus was 4.2 ± 0.4 mg/dl.[14] Researchers are of the opinion that phosphate retention, and not FGF23, is the pathologic mechanism majorly involved in aging, glucose metabolism, insulin sensitivity disturbances, and oxidative stress.[15]

It was observed that there was a statistically significant reduction in serum phosphorus levels over 1 year in patients who were given non–calcium-based phosphate binders. In a previous study, it has been shown that sevelamer can diminish serum phosphate in dialysis patients.[16] In the study presented here, sevelamer treatment appeared to be effective in lowering serum FGF23 levels. It is possible that sevelamer's increased ability to reduce FGF23 occurs through an unknown or pleiotropic effect that is unrelated to the reduction of serum phosphorus.

Robust data is lacking on the role of dietary phosphate restriction in pre-dialytic CKD population. The present study showed a significant decline in serum phosphate levels in patients who were given phosphate-restricted diet, thereby reflecting that serum phosphorus is under the control of dietary intake. Similar findings were also reported in a study by Tsai et al.,[17] where a very low phosphate diet significantly lowered serum phosphate.

There was significant reduction in serum phosphorus levels in patients who were given calcium-based phosphate binders. The ability of calcium salts to effectively bind intestinal phosphate and reduce serum phosphate concentration has been recognized for decades. Similarly, in a controlled trial of three commercially available phosphate binders, Block et al.[14] reported a slight reduction in serum phosphate levels (mean serum phosphorus declined from 4.2 to 3.9 mg/dl over 9 months). Interestingly, we found that there was significant decline is serum calcium levels over a period of 12 months in patients who were given non–calcium-based phosphate binders, but similar decline was not seen in the group given calcium-based phosphate binder, thereby reflecting the impact of non–calcium-based phosphate binders on the levels of serum calcium in CKD population.

Serum FGF23 is considered an early biomarker of disordered mineral and bone metabolism in patients with CKD, as observed by Fliser et al.[12] They showed that FGF23 level was increased before an increase in PTH and phosphate was observed. However, in our study, secondary hyperparathyroidism was more prevalent than FGF23 excess. There was insignificant initial decline in serum FGF23 in group 1 patients over a period of 6 months, but not thereafter. The main explanation for this is that dietary phosphate-to-protein ratio restriction was not large enough in the study to achieve a dose–response FGF23-lowering effect. Additionally, the compliance of the patients to dietary restrictions cannot be ascertained. Chang et al.[18] also did not find any effect of a diet with low phosphate additives (11 mg/day) on FGF23 levels in patients with CKD stage 2. Isakova et al.[19] stated that phosphate-restricted diet did not significantly reduce FGF23 levels in early CKD. There were no significant treatment effects on the iPTH levels. Indeed, it may be because of relative hypocalcemia, which has led to elevated PTH levels. Similarly, in a 12-month study in CKD patients of stages 3a, 3b, and 4, there were no treatment effects on the iFGF23 or iPTH levels.[20]

FGF23 increased insignificantly from baseline to 12 months in group 2, possibly due to the exogenous calcium load that may have contributed to its elevation. Our results are as per a study of dietary phosphate restriction and therapy with lanthanum carbonate, in which the levels of FGF23 were not reduced.[19] Similarly, comparisons of lanthanum carbonate and placebo have yielded negative results in normo-phosphatemic, non-dialyzed patients with CKD.[21]

We determined the association between clinical variables of CKD progression and serum FGF23 concentration in patients with CKD in our study. FGF23 levels showed significantly strong correlation with eGFR. Though the correlation of FGF23 with intact PTH was significant and positively correlated, the correlation of FGF23 with calcium, phosphorus, ALP, and mean BP was not found to be significant. These results were as per a previous study which showed that FGF23 negatively correlated with eGFR (P-0.000) and iPTH (P-0.009) at baseline.[3] Similar to a previous study, there was no correlation of age, phosphorus, and nutritional phosphorus intake with FGF23 at early CKD.[7]

In our study, the correlation between iFGF23 and vascular calcification was not significant, which may be due to the younger population in the study in contrast to an observational French study. The authors in that study showed that aortic and coronary artery calcification scores could predict all-cause and cardiovascular mortalities in CKD patients.[22] Yuvaraj et al.[23] showed that raised intact-FGF23 was associated with cardiovascular mortality.

As an additional finding, there was progressive decline in SGOT/SGPT levels in patients who were given phosphate binders along with dietary phosphorus restriction. In our own study, SGOT/SGPT levels were low in hemodialysis patients; pyridoxine deficiency, increased levels of hepatocyte growth factor, uremic toxins, and hemodilution are the possible postulations.[24] Pleotropic effects of phosphate binder may be one of its etiologies; however, it needs further study.

This is thefirst Indian randomized controlled study conducted among three hypophosphatemia management groups taking dietary intervention to assess the phosphorus level based on FGF23 level. All participants were closely observed during the washout period before the intervention, and all were contacted telephonically to prevent any dietary deviations. The study cohort was small; it was from a single hospital, which may not extrapolate to other geographic regions. We have assessed bone disease based on noninvasive FGF23 levels, but not used the gold standard bone-biopsies, to assess bone turnover.

Conclusion

Serum phosphorus level upto 5mg/dl had no effect on FGF23 at early CKD stages; however, it was positively corelated with iPTH. Serum FGF23 levels were significantly positively correlated with iPTH and negatively correlated with eGFR. Although different treatment groups showed significant phosphorus reduction, non-calcium phosphate binders had a major impact on FGF23 reduction. Further studies are required considering dietary intervention without any phosphate binder for hyperphosphatemia management and to assess appropriate phosphorus level at every stage of early CKD in a larger long-term cohort.