This was a prospective, observational, single-center study conducted between January and July 2023 at Dialysis Center Avericum SL (Las Palmas de Gran Canaria, Spain). All patients had previously received phosphate binders. After a standardized 2-week washout period, patients were switched to sucroferric oxyhydroxide (SFO) due to persistent hyperphosphatemia despite prior therapy, with the goal of achieving guideline-recommended serum phosphate targets. The study followed the STROBE guidelines for reporting observational studies.10

The protocol complied with the Declaration of Helsinki and was approved by the Institutional Review Board of Hospital Universitario de Gran Canaria Dr. Negrín (Approval Code: 2021-414-1). Written informed consent was obtained from all participants.

Fifty-three adult patients undergoing maintenance HD three times per week for more than three months were screened. Inclusion criteria were: age ≥18 years; serum phosphate >5.0mg/dL despite use of conventional phosphate binders (both calcium-based and non-calcium-based agents other than SFO); and a clinical indication to initiate SFO following binder washout. Thus, all patients had prior exposure to phosphate binders but remained hyperphosphatemic.

Exclusion criteria included active infection, malignancy, autoimmune disease, recent hospitalization (<3 months), life expectancy <6 months, or ongoing treatment with systemic corticosteroids or other immunosuppressive agents. Regarding mineral metabolism therapies, patients were excluded only if they had initiated or undergone dose adjustments of calcimimetics or pharmacologic doses of vitamin D analogues within the preceding three months, while those on stable chronic maintenance regimens were permitted to participate.

Eligibility for switching to SFO was established at screening based on documented, persistent hyperphosphatemia under previous phosphate binders. The 2-week washout period was performed exclusively to avoid pharmacologic carryover and to establish a clean biochemical baseline, following methodological standards used in phosphate-binder studies. After the washout, SFO was initiated at 500–1500mg/day with meals and titrated to achieve serum phosphate levels below 4.5mg/dL, in accordance with guideline-based clinical practice and previous evidence on SFO effectiveness.11–13 All patients received standard dietary counseling. Clinical and laboratory parameters were assessed at baseline and after six months. Treatment adherence and adverse events were recorded prospectively.

Vitamin D therapy was recorded and maintained stable throughout the study period. Treatment included nutritional vitamin D supplementation (cholecalciferol) and active vitamin D receptor activators (calcitriol and/or paricalcitol), prescribed according to routine clinical practice. No systematic initiation or dose escalation of vitamin D analogues occurred during follow-up unless clinically indicated.

Fasting blood samples were collected pre-dialysis at baseline and at six months using BD SST™ serum tubes and PAXgene™ Blood RNA tubes (PreAnalytiX, Switzerland). The PAXgene system preserves RNA integrity for accurate gene expression analysis.14

Routine biochemical measurements included serum calcium, phosphate, intact parathyroid hormone (iPTH), magnesium, 25-hydroxyvitamin D3 [25(OH)D3], hemoglobin, hematocrit, transferrin saturation (TSAT), ferritin, albumin, and bicarbonate. C-reactive protein (CRP), cytokines (TNF-α, IL-6, IL-10), and fibroblast growth factor-23 (FGF-23) were measured by enzyme-linked immunosorbent assay (ELISA) using commercial kits (R&D Systems and Millipore).

Erythropoiesis-stimulating agent (ESA) therapy was recorded throughout the study, and weekly epoetin-equivalent doses were collected. ESA hyporesponsiveness was assessed clinically, based on the absence of adequate hemoglobin response despite stable or increasing ESA requirements, as well as the need for high ESA doses to maintain target hemoglobin levels.

Vitamin D deficiency was defined as 25(OH)D3 <20ng/mL,15 and persistent inflammation as CRP >5mg/L at six months.16 Neutrophil-to-lymphocyte ratio (NLR) and platelet-to-lymphocyte ratio (PLR) were calculated from absolute leukocyte counts.17 For extended analyses (cytokines and leukocyte KL expression), an exploratory subgroup was formed based on feasibility, specifically the availability of appropriately collected PAXgene RNA tubes and complete cytokine panels. All 42 patients completed the 6-month clinical follow-up and were included in the primary endpoints. The reasons for missing extended molecular data in the remaining 24 patients (such as poor tolerance or lack of adherence) refer strictly to non-compliance with the specific blood-sampling protocol required for molecular processing, without affecting their participation or follow-up in the main clinical cohort. Reasons for missing molecular data in the remaining patients are summarized in Fig. 1.

Fig. 1.

Study flowchart. All 42 patients completed the 6-month clinical follow-up under sucroferric oxyhydroxide (SFO) therapy. Extended inflammatory and leukocyte gene-expression analyses were feasible in 18 patients. In the remaining 24, these assessments could not be performed due to poor tolerance (n=9), lack of adherence (n=8), kidney transplant (n=4), or withdrawal of consent (n=3).

Total RNA was extracted from whole blood using the PAXgene Blood RNA Kit, following the manufacturer's instructions. Complementary DNA (cDNA) was synthesized using the High-Capacity RNA-to-cDNA Kit (Applied Biosystems). Quantitative real-time polymerase chain reaction (qRT-PCR) was performed with PerfeCTa FastMix II (QuantaBio) on a QuantStudio 5 platform.

KL, TNF-α, IL6, IL10, and GAPDH expression were quantified using TaqMan Gene Expression Assays: Hs00183100_m1 (KL), Hs00174128_m1 (TNF), Hs00174131_m1 (IL6), Hs00961622_m1 (IL10), and Hs99999905_m1 (GAPDH). Relative expression was normalized to GAPDH and calculated using the 2−ΔΔCt method.18 All samples were run in triplicate, and RNA quality was verified spectrophotometrically.

Continuous variables were expressed as mean±SD or median [IQR], as appropriate. Normality was evaluated using the Shapiro–Wilk test. Paired t-tests or Wilcoxon signed-rank tests were applied for pre–post comparisons. Linear regression assessed associations between changes in 25(OH)D3 and inflammatory indices (ΔNLR, ΔPLR).

Persistent inflammation (CRP>5mg/L) was modeled using logistic regression. Because of the limited number of events, Firth's penalized likelihood estimation was used, with exact logistic regression as sensitivity analysis. Event-per-variable ratios were calculated to minimize overfitting, and results are reported as odds ratios (OR) with 95% confidence intervals (CI).

The primary endpoint was the change in serum phosphate at six months. Secondary endpoints included changes in iPTH, 25(OH)D3, NLR, PLR, cytokines, and KL expression.19 No formal sample size calculation was performed given the exploratory design. A post hoc power analysis indicated >80% power to detect the observed changes in phosphate, 25(OH)D3, NLR, and PLR. Statistical analyses were conducted using SPSS version 26.0 (IBM, Armonk, NY), and significance was defined as p<0.05 (two-sided).

Forty-two patients (mean age 63.4±12.7 years; 56.3% [n=24/42] men) were enrolled (Fig. 1). The median dialysis vintage was 47 [28-82] months. Diabetes mellitus was present in 64.3% (n=27/42) of participants, and 100% (n=42/42) had hypertension (Table 1).

Regarding CKD etiology, the relatively high proportion of glomerular disease and autosomal dominant polycystic kidney disease likely reflects the referral pattern and case-mix of our center, as well as the specific clinical profile of patients selected for phosphate binder switch due to persistent hyperphosphatemia. Therefore, this etiologic distribution may introduce selection bias and should be interpreted with caution when extrapolating our findings to broader HD populations.

After six months of treatment with SFO, serum phosphate significantly decreased from 6.12±0.81 to 4.44±0.67mg/dL (−27.5%, p<0.001). Intact parathyroid hormone (iPTH) declined from 478 [325–639] to 295 [208–431]pg/mL (−38.3%, p<0.001). The calcium–phosphorus product also improved markedly. Serum calcium levels remained stable throughout the study (9.1±0.4 vs. 9.2±0.3mg/dL; p=0.44) (Fig. 2).

Fig. 2.

Percentage changes after 6 months of SFO therapy. Data are presented as mean percentage change relative to baseline values. Mineral metabolism: phosphate, iPTH, and Ca×P decreased; 25(OH)D3 increased. Inflammation: NLR and PLR decreased, whereas CRP showed a modest, non-significant decline. Gene expression: KL expression declined markedly. iPTH, intact parathyroid hormone; Ca×P, calcium–phosphorus product; 25(OH)D3, 25-hydroxyvitamin D3; KL, Klotho; NLR, neutrophil-to-lymphocyte ratio; PLR, platelet-to-lymphocyte ratio; CRP, C-reactive protein; SFO, sucroferric oxyhydroxide.

Hematologic inflammatory indices improved following SFO therapy. The NLR decreased from 4.4 [3.1–6.9] to 2.5 [1.8–3.8] (−44.5%, p<0.001), and the PLR declined from 167 [132–216] to 136 [118–188] (−18.9%, p=0.02). In contrast, C-reactive protein (CRP) levels and cytokines (IL-6, TNF-α, IL-10) did not show significant changes (p>0.05 for all) (Table 2).

Persistent inflammation, defined as CRP >5mg/L, was observed in 33% of patients at baseline and 28% after six months (p=0.47).16,17 Fibroblast growth factor-23 (FGF-23) levels remained stable.

ESA requirements remained stable over the 6-month follow-up. No patient fulfilled clinical criteria for ESA hyporesponsiveness, and no sustained increase in ESA requirements was observed during the study period.

Despite improvements in biochemical and hematologic parameters, leukocyte Klotho (KL) gene expression significantly decreased after six months (median fold change −0.39 [IQR −0.65 to −0.14]; p=0.01). This reduction was consistent across subgroups stratified by baseline phosphate or vitamin D status.

In exploratory logistic regression models, higher baseline 25(OH)D3 concentrations demonstrated a significant protective effect against persistent inflammation in unadjusted assessments (OR [95% CI]). However, due to the limited sample size and low event-per-variable ratio, this predictive value did not reach consistent statistical significance across all fully adjusted multivariate models. These findings emphasize the exploratory nature of our regression data. No associations were observed between serum phosphate or iPTH changes and inflammatory indices (Fig. 3). Logistic regression identified no predictors of persistent inflammation at six months.

Fig. 3.

Exploratory analysis of predictors of persistent inflammation at 6 months. Odds ratios (OR, 95% CI) for baseline and follow-up parameters. Only serum 25(OH)D3 showed a protective effect, with each 1ng/mL increase associated with lower odds of persistent inflammation. The red dashed line represents the null value (OR=1). 25(OH)D3, 25-hydroxyvitamin D3; NLR, neutrophil-to-lymphocyte ratio; PLR, platelet-to-lymphocyte ratio; CRP, C-reactive protein; IL-6, interleukin-6; CI, confidence interval.

Regarding the primary endpoint, 86% of participants showed a reduction in serum phosphate concentration. Secondary endpoints (iPTH, vitamin D, and hematologic indices) also showed significant improvements, while cytokine and KL results indicated complex regulation beyond mineral metabolism.

The observed reduction in serum phosphate (−27.5%) and intact parathyroid hormone (−38.3%) confirms that SFO is an effective non-calcium phosphate binder in routine clinical use, consistent with results from randomized and real-world studies.11–13,19–21 The parallel 42.7% increase in serum 25(OH)D3 represents an important and novel finding. Improvement in vitamin D status may reflect decreased intestinal phosphate burden and improved nutritional intake, as previously reported in patients treated with iron-based binders.22–24 In contrast to calcium-based agents, SFO minimizes the risk of hypercalcemia, which may preserve endogenous vitamin D metabolism.24,25

Beyond improved phosphate control, several complementary mechanisms may contribute to the observed increase in serum 25(OH)D3 following SFO therapy. Sustained reduction of intestinal phosphate burden may alleviate FGF-23-mediated suppression of vitamin D metabolism, thereby favoring higher circulating 25(OH)D3 concentrations. In addition, iron-based phosphate binders such as SFO have been associated with improved gastrointestinal tolerance and nutritional status, which could indirectly enhance dietary vitamin D intake and absorption. Finally, avoidance of calcium overload may preserve endogenous vitamin D hydroxylation pathways. Importantly, these mechanisms remain speculative, and this observational study was not designed to establish mechanistic causality. These interpretations should be considered hypothesis-generating and require confirmation in controlled mechanistic studies.

Although traditional inflammatory biomarkers such as C-reactive protein and cytokines remained unchanged, both the NLR and PLR improved significantly. These hematologic indices have been validated as sensitive predictors of inflammation and cardiovascular risk in chronic kidney disease.17,26 This apparent discrepancy is biologically plausible. Hematologic indices such as NLR and PLR can reflect early shifts in leukocyte dynamics and innate immune activation, whereas soluble inflammatory markers like CRP and cytokines show substantial intra-individual variability in hemodialysis and limited sensitivity to detect short-term changes in low-grade inflammation. In our cohort, NLR and PLR decreased markedly while IL-6 and TNF-α did not change significantly, a pattern consistent with prior reports questioning the sensitivity of CRP and cytokines and supporting the prognostic usefulness of hematologic indices in dialysis populations. The small molecular subgroup (n=18) further limits the statistical power to detect changes in cytokine concentrations. The strong inverse correlation between 25(OH)D3 and both NLR and PLR suggests that improved vitamin D status may exert anti-inflammatory effects through modulation of leukocyte activity and cytokine signaling, supporting prior translational studies linking vitamin D sufficiency to reduced systemic inflammation.27–29 However, these findings should not be interpreted as being exclusively driven by phosphate control, as multiple factors – including vitamin D status, iron-related effects, and overall clinical context – may contribute to the observed changes in hematologic inflammatory indices.

Importantly, ESA therapy remained stable during follow-up, and no signal of ESA hyporesponsiveness was observed. This reduces the likelihood that changes in hematologic inflammatory indices were driven by variations in anemia management rather than underlying biological effects.

Despite biochemical and hematologic improvement, leukocyte Klotho (KL) gene expression decreased after SFO therapy. This finding is consistent with previous reports describing persistent suppression of KL in dialysis patients even under optimized mineral metabolism.9,24,30 Downregulation of KL may reflect epigenetic silencing, oxidative stress, or chronic exposure to uremic toxins that are not corrected by phosphate control alone.31,32 The disconnect between improved clinical parameters and suppressed KL expression underscores that CKD-associated inflammation has a multifactorial origin extending beyond phosphate retention.33,34

Taken together, these data suggest that SFO provides biochemical control and improvement in surrogate inflammatory indices, yet fails to normalize KL transcriptional activity. This paradox highlights the complexity of CKD-mineral and bone disorder (CKD-MBD), where mineral dysregulation, oxidative stress, and vitamin D metabolism interact at molecular and cellular levels. Future studies integrating transcriptomic and metabolomic approaches could elucidate whether iron-mediated phosphate binding directly affects pathways governing KL expression and vitamin D bioactivation.35,36

The present work provides novel insight into the interrelationship between phosphate control, vitamin D status, and inflammation in a real-world dialysis cohort. Its strengths include the prospective design, homogeneous patient population, and molecular analysis of leukocyte KL expression. Limitations include the single-center design, small sample size, and absence of a control group, which may restrict external validity. In addition, inflammatory findings are based primarily on hematologic indices (NLR and PLR), which, although clinically relevant, represent indirect markers of inflammation and may be influenced by multiple factors. Nevertheless, post hoc power analysis supports adequate statistical sensitivity for the primary outcomes. The observational design also precludes causal inference, and additional mechanistic studies are warranted.