Impact of sucroferric oxyhydroxide on the oral and intestinal microbiome in hemodialysis patients

Abstract

Hyperphosphatemia is a consequence of chronic kidney disease associated with mineral/bone impairment, increased cardiovascular events and mortality. Therapeutically, most dialysis patients have to take phosphate binders. Here, we investigated effects of the Fe(3+)-based phosphate binder sucroferric oxyhydroxide (SFOH) on the oral and gastrointestinal microbiome of 11 hemodialysis patients. Saliva, dental plaque and stool were collected at baseline, one and four weeks of SFOH intake and subjected to 16S rRNA gene (V3-V4 region) directed Illumina MiSeq-based analysis. Total Fe, Fe(2+) and Fe(3+) were determined in stool and saliva. Overall, the microbiome did not change significantly. However, some patient-, sample- and taxon-specific differences were noted, which allowed patients to be divided into those with a shift in their microbiome (6/11) and those without a shift (5/11). Total Fe and Fe(2+) were highest after one week of SFOH, particularly in patients who exhibited a shift in microbiome composition. Eight bacterial taxa showed significant unidirectional changes during treatment. In-depth microbiome analysis revealed that taxa that significantly benefited from iron plethora had no iron-binding siderophores or alternatives, which was in contrast to taxa that significantly declined under iron plethora. Patients with microbiome-shift were significantly younger and had higher serum phosphate concentrations. In conclusion, this study sheds light on the impact of iron on the microbiome of hemodialysis patients.

Figure 1

CONSORT flow diagram of the study.

Figure 2

Analysis of β-diversity. Left: multidimensional scaling (MDS): over all taxa, the beta-diversity was not significantly different between the saliva, biofilm, and fecal samples collected at three different time points, t0 (before switch to SFOH), t1 (one week after SFOH switch), and t2 (four weeks after SFOH switch). Right: Cluster analysis and intra-patient dynamical change of microbiome. Changes that appear significant are followed by arrows. Patients with such a shifting microbiome (named shifters) were A01 (saliva, biofilm), A02 (biofilm, feces), A09 (biofilm), A12 (saliva, biofilm, feces), A13 (saliva, biofilm), and A16 (biofilm, feces, with data for t0/1 only). Microbiomes of patients A05, A08, A15, A17, A19 (non-shifters) appeared quite stable, independent of specimen. In a subgroup here, however, two specimens (saliva of A15 and feces of A08) changed from t0 to t1, but t2-microbiome clustered again with t0, following a boomerang move.

Figure 3

Changes in the oral (a–b saliva and c–d teeth-attached biofilm) and intestinal (e–f) microbiome from baseline (t0) to one (t1) and four (t2) weeks of SFOH administration. Phylum composition is presented on the left (a, c, e) and genera composition on the right side (b, d, f).

Figure 4

Changes in the oral (a–b saliva and c–d teeth-attached biofilm) and intestinal (e–f) microbiome from baseline (V0, t0) to one (V1, t1) and four (V2, t2) weeks of SFOH administration. The patient-specific dynamic is presented on the left (a, c, e) whereas the relative abundance of each genus is presented on the right side (b, d, f).

Figure 5

Statistically significant changes in certain taxa of the microbiome in saliva, biofilm, and fecal samples after SFO.

Figure 6

Dynamic of (a) Fe(2+) and (b) Fe(3+) detected at t0, t1 and t2 in the fecal supernatant of all 11 patients. The statistical analysis was performed using the non-parametric one-way ANOVA (Kruskal–Wallis) as non-pairing due to the missing measurements of patients A16, A17 and A19 at t2. The significance threshold was set at p = 0.05. For Fe(2+) ** = 0.0028 and *** = 0.0003; while for Fe(3+) * = 0.0424 and ** = 0.0085.

Figure 7

Statistically significant differences of patients’ characteristics and clinical chemistry data (from blood serum or whole blood in case of pH and HCO3), grouped by shifters and non-shifters. DW = dry weight on dialysis, P = phosphate, iCa = ionized calcium, Ca = total calcium, wPTH whole PTH, Hb = hemoglobin, Fe = iron, TSAT = transferrin saturation, HCO3 = bicarbonate (mmol/l). Shifters presented a lower age and higher serum phosphate concentration at baseline. Wilcoxon signed rank sum test. p < 0.05 (*) was considered significant. Differences by height and body mass index were also not significant between both groups.