We conducted a single-center retrospectivestudy, from the time of January 2019 to May 2023. All patients with erythropoietin hyporesponsiveness were divided into the roxadustat group and ESAs group according to drug use. PD patients (n = 61) who previously were receiving ESAs and switched to roxadustat were defined as the roxadustat group, and patients (n = 20) were defined as the ESAs group which they had an increased dose of ESAs or frequency of administration. All patients had adequate iron levels serum ferritin (SF) > 100 ng/mL, and transferrin saturation (TSAT) > 20% at the initial treatment stage. We used ESA resistance index (ERI) as an evaluation measure, which was calculated as the weight-adjusted weekly ESA dose divided by Hb concentration. When ERI ≥ 1 IU/kg/w/g/L, we considered the presence of erythropoietin hyporesponsiveness. In addition, these patients had been on adequate ESA treatment for more than three months, and their anemia-related causes had been actively treated, but their hemoglobin (Hb) was still below 110 g/L. The inclusion criteria were: (1) The patients were aged between 18 and 75 years old; (2) Patients with CKD associated with renal anemia underwent regular PD for more than 12 weeks. The exclusion criteria were: (1) Patients who were transferred to HD, kidney transplantation during the follow-up period; (2) Patients with other diseases causing anemia during the follow-up period, such as hematological disorders, hemorrhagic diseases resulting from various causes, and so on; (3) Patients who underwent blood transfusion during the follow-up period; (4) Patients with other serious chronic diseases, such as severe liver function impairment, severe heart failure, malignancy, serious infections, malignant hypertension. The specific screening scheme is shown in (Fig. 1). The study was approved by the Medical Ethics Committee of the Second Affiliated Hospital of Xuzhou Medical University (approval number: [2023]040601), and also completed Chinese Clinical Trial Registration (registration number: ChiCTR2300076409). This study was exempted from obtaining informed consents due to the retrospective character.

Only PD patients were included in the study, and the follow-up period was 16 weeks. All patients were given an initial therapeutic dose of 50–150 IU/kg per week of ESAs before the presence of erythropoietin hyporesponsiveness was determined, given by subcutaneous injection 1–3 times per week. In the roxadustat group, when we considered patients with erythropoietin hyporesponsiveness, we changed their anemia correction medication to roxadustat and discontinued ESAs. We determined the initial dose of roxadustat according to the drug instructions, giving 100 mg three times a week orally for subjects weighing 45–60 kg and 120 mg three times a week for subjects weighing ≥ 60 kg. The medication dose during the follow-up process was adjusted according to the comprehensive assessment of the current Hb level, Hb changes one month after taking the medication, the patient’s nutritional status, and iron metabolism level, making the target Hb level 110–130 g/L. In the ESAs group, we increased the dose of ESAs by 20 IU/kg each time or increased the frequency of administration as appropriate after the presence of erythropoietin hyporesponsiveness was determined, but the maximum dose of ESAs was not more than 2 times the stable dose. We referred to KDOQI clinical practice guidelines and clinical practice recommendations for anemia in chronic kidney disease: 2007 update of hemoglobin target [13] and Clinical Practice Guidelines for Diagnosis and Treatment of renal anemia in China to determine the Hb target of this study between 110–130 g/L. During the whole treatment, we timely adjusted the dosage of both drugs, timely controlled inflammation, and followed up with patients monthly to ensure compliance. When SF ≤ 100 ng/mL or TSAT ≤ 20%, we started oral iron therapy, using ferrous succinate tablets 200 mg/d. The target value of SF was 100–500 ng/mL, and the target value of TSAT was 20–50%. Iron status was assessed at 1–3 months, and patients who did not reach the target value or could not tolerate it orally were switched to intravenous iron supplementation.

We collected basic patient information and laboratory test results from our hospital’s electronic medical record system. Regarding the adverse events of drugs, we collected relevant data through the electronic medical record system, and on the other hand, we supplemented relevant data through telephone follow-up. The primary indicators for adverse events were as follows: (1) hypertension, characterized by a substantial increase in blood pressure accompanied by a rapid elevation in Hb levels, which could be restored to baseline by reducing the dosage of the medication; (2) hyperkalemia, indicated by serum potassium levels exceeding 5.5 mmol/L; (3) hyperbilirubinemia, denoted by serum bilirubin levels surpassing 17.1 umol/L; (4) as well as other discomfort experienced after drug administration, including headache, insomnia, vomiting, diarrhea, and skin issues.

RStudio version 1.3 was used for statistical analysis and plotting. When the measurement data conformed to a normal distribution, it was represented by mean ± standard deviation, when it was not normally distributed, it was described by quartile M (P25, P75), statistical analysis was conducted by t-test or Wilcoxon-test. Tegorical variables are expressed as a percentage (%); statistical analysis was conducted by chi-squared test or Fisher’s exact test. The intra-group difference was compared by analysis of variance. Pearson correlation analysis was used to analyze its correlation. p < 0.05 indicated a statistically significant difference.

A total of 81 patients diagnosed with PD were included in the research, with 61 individuals in the roxadustat group and 20 individuals in the ESAs group. Among the participants, there were 52 males and 29 females, with an average age of 57.1 ± 15.2 years in the roxadustat group and 60.0 ± 11.3 years in the ESAs group. The average duration of PD was 18.0 ± 8.68 months for the roxadustat group and 21.6 ± 8.98 months for the ESAs group. The initial demographic and clinical characteristics of both groups did not exhibit any significant differences (p > 0.05). The basic information and the baseline values of laboratory indicators are shown in (Table 1).

The baseline Hb levels of roxadustat and ESAs groups were 89.8 ± 18.9 g/L and 95.2 ± 16.0 g/L, respectively, with no significant difference between groups (p = 0.218). And over time, the mean Hb levels of roxadustat group increased to 105 ± 29.5 g/L at week 4, 107 ± 21.8 g/L at week 8, 112 ± 20.3 g/L at week 12, and 118 ± 19.8 g/L at week 16, respectively, showing significant differences from baseline values at all time nodes (p < 0.05). However, in the ESAs group, the mean Hb value was 96.4±11.4 g/L at week 4, 96.8 ± 17.1 g/L at week 8, 98.2 ± 19.1 g/L at week 12, and 101 ± 19.3 g/L at week 16, all of which did not change significantly from baseline (p > 0.05) (Fig. 2).

The laboratory of our hospital determined that the upper limit of the normal value for hs-CRP is 5 mg/L. We categorized the patients into two subgroups based on their hs-CRP baseline levels: hs-CRP < 5 mg/L and hs-CRP ≥ 5 mg/L. We then compared the changes in Hb levels within each subgroup. The results indicated that in the roxadustat group, both subgroups (hs-CRP < 5 mg/L and hs-CRP ≥ 5 mg/L) had significantly higher Hb values at week 16 compared to their baseline values (122 ± 22.4 g/L vs. 93.8 ± 17.3 g/L and 114 ± 15.8 g/L vs. 85.3 ± 19.9 g/L, respectively), with statistical significance (p < 0.05) (Fig. 3A). However, in the ESAs group, there were no significant changes in Hb values within the two subgroups at week 16 compared to their baseline values p > 0.05) (Fig. 3B).

Additionally, we referred to the 2003 K/DOQI clinical practice guidelines for bone metabolism and disease in chronic kidney disease. According to the guideline’s suggested target for iPTH, we divided iPTH into three subgroups: < 150 pg/mL, 150–300 pg/mL, and > 300 pg/mL. We then analyzed the changes in Hb values within each subgroup. The results showed that in the roxadustat group, Hb values at week 16 were significantly higher than baseline values in all three subgroups (123 ± 14.0 g/L vs. 93.1 ± 17.4 g/L, 117 ± 23.1 g/L vs. 85.4 ± 18.1 g/L, and 115 ± 20.5 g/L vs. 92.1 ± 21.3 g/L), with statistical significance (p < 0.05) (Fig. 3C). However, no significant differences were observed within the subgroups of the ESAs group (Fig. 3D).

We divided the roxadustat group and ESAs group into statin and non-statin subgroups based on whether or not they were using statins, and compared changes in total cholesterol levels, low-density lipoprotein (LDL), and high-density lipoprotein (HDL) over time. The results showed that total cholesterol in the roxadustat statin subgroup decreased significantly at weeks 12 and 16 compared with the baseline values (3.78 ± 0.951 mmol/L vs. 4.93 ± 1.92 mmol/L and 3.64 ± 0.799 mmol/L vs. 4.93 ± 1.92 mmol/L). The specific values at each time node are shown in Figure 4A. The p values at week 12, and week 16 compared with the baseline were 0.048 and 0.011, respectively. Similarly, total cholesterol decreased significantly from baseline at weeks 12 and 16 in the non-statin subgroup of roxadustat (3.68 ± 1.04 mmol/L vs. 4.31 ± 0.833 mmol/L and 3.62 ± 1.06 mmol/L vs. 4.31 ± 0.833 mmol/L), p values were 0.031 and 0.013 compared with the baseline at week 12 and week 16, respectively. However, in the ESAs group, only the statin subgroup had a significant decrease in total cholesterol level at week 16 compared with the baseline value (3.56 ± 0.768 mmol/L vs. 4.80 ± 1.37 mmol/L, p = 0.049), and the total cholesterol value at each time node was listed in detail in Figure 4B.

LDL values in the non-statins and statins subgroups of the roxadustat group at week 16 were significantly lower than the baseline value (2.38 ± 1.37 mmol/L vs. 3.39 ± 1.48 mmol/L and 2.37 ± 1.19 mmol/L vs. 3.56 ± 1.80 mmol/L), with statistically significant differences (p = 0.024 and 0.04), and the specific values at each time node were shown in Figure 4C. However, in the ESAs group, only the statin subgroup had a significant decrease in LDL level at week 16 compared with the baseline value (2.43 ± 0.925 mmol/L vs. 3.64 ± 1.09 mmol/L, p = 0.029), and the LDL value at each time node was listed in detail in Figure 4D. In relation to HDL, the HDL values within the subgroups of both the roxadustat and ESA groups remained relatively stable throughout the study period, with no statistically significant differences (p > 0.05), as illustrated in Figure 4E and Figure 4F.

A statistical analysis was conducted to examine the impact of roxadustat on cardiac function. The evaluation indicators used in this study included brain natriuretic peptide (BNP), cardiac ejection fraction, left ventricular end-systolic dimension (LVESD), and left ventricular end-diastolic dimension (LVEDD). In the group receiving roxadustat, there was a gradual increase in left ventricular ejection fraction as anemia improved. This increase was significant compared to the baseline measurements at week 12 and week 16 (54.1 ± 4.4% vs. 49.4 ± 7.08% and 58.5 ± 2.74% vs. 49.4 ± 7.08%), with statistically significant differences (p < 0.0001 and < 0.0001). However, no significant change in left ventricular ejection fraction was observed in the group receiving ESAs. The values of left ventricular ejection fraction at each time point for both groups are presented in Figure 5A. The BNP values for both groups did not show any significant changes throughout the follow-up period (p > 0.05), as depicted in Figure 5B.

Although there were no significant changes in LVESD values in both groups throughout the follow-up period, it is important to note that in the roxadustat group, LVESD gradually decreased as anemia improved and nearly approached statistical significance compared to baseline at week 16 (33.0 ± 4.80 mm vs. 35.3 ± 4.72 mm, p = 0.056). The specific LVESD values at each time point for both groups are presented in Figure 5C. Similarly, the LVEDD values did not show significant changes during the entire follow-up period for both groups, but there was an overall downward trend in the roxadustat group, and a near statistical difference compared to baseline was observed at week 16 (49.2 ± 4.25 mm vs. 51.0 ± 2.83 mm, p = 0.057). The specific values at each time point are displayed in Figure 5D.

During the entire follow-up period, one patient in the roxadustat group died of severe peritonitis at day 65, but we did not consider this to be due to adverse drug reactions. No serious thromboembolic disease and cardiovascular adverse events occurred. It is worth noting that three patients developed hyperbilirubinemia and returned to normal after appropriate treatment, but we do not have sufficient evidence to suggest that it was caused by adverse reactions to roxadustat. Five individuals encountered a swift elevation in Hb levels and a notable rise in blood pressure from their initial state, and subsequently, their blood pressure gradually reverted back to the initial levels following a reduction in drug dosage. Four patients exhibited symptoms of headaches, which subsided after the reduction in dosage. A limited number of patients experienced insomnia, vomiting, diarrhea, and pruritic skin; however, these symptoms vanished after receiving treatment. In general, there was no substantial disparity in the occurrence of unfavorable events between the roxadustat and ESA groups. The adverse events are shown in (Table 2).