Table 5.
Multivariate Cox regression analysis of cardiovascular-related adverse events
| Factors | No. of patients | No. of patients with CV-related AEs (%) | HR (95% CI) | p value | |
|---|---|---|---|---|---|
| Age (years) | < 65 | 1232 | 79 (6.4) | Ref | |
| ≥ 65 to < 75 | 1459 | 171 (11.7) | 1.931 (1.258–2.966) | 0.0026 | |
| ≥ 75 | 2817 | 404 (14.3) | 2.457 (1.632–3.699) | < 0.0001 | |
| History of CV disease | No | 4782 | 502 (10.5) | Ref | |
| Yes | 726 | 152 (20.9) | 2.216 (1.656–2.966) | < 0.0001 | |
| Concurrent diabetes | No | 3162 | 343 (10.8) | Ref | |
| Yes | 2346 | 311 (13.3) | 1.561 (1.199–2.032) | 0.0009 | |
| Transfusion | No | 4995 | 547 (11.0) | Ref | |
| Yes | 475 | 105 (22.1) | 1.913 (1.367–2.676) | 0.0002 | |
| Baseline eGFR (mL/min/1.73 m2) | < 15 | 2227 | 255 (11.5) | 1.849 (1.122–3.046) | 0.0159 |
| ≥ 15 to < 30 | 2074 | 279 (13.5) | 1.857 (1.142–3.019) | 0.0125 | |
| ≥ 30 | 655 | 61 (9.3) | Ref | ||
| Baseline urine protein (mg/dL) | Continuous quantity | 2451 | 260 (10.6) | 1.001 (1.000–1.001) | 0.0026 |
| Dose of darbepoetin per week | < Median | 2754 | 309 (11.2) | Ref | |
| ≥ Median | 2754 | 345 (12.5) | 1.361 (1.049–1.766) | 0.0201 | |
| Use of antithrombotic agent | No | 5018 | 511 (10.2) | Ref | |
| Yes | 490 | 143 (29.2) | 2.377 (1.701–3.322) | < 0.0001 | |
Patients from the safety analysis set (n = 5547) without a history of cerebrovascular disease (n = 39) were included in this multivariate analysis (n = 5508)
The following variables were selected as explanatory variables: sex, age, body mass index, history of CV disease, hypertension, hyperlipidemia, diabetes, previous treatment with recombinant human erythropoietin, transfusion, baseline systolic blood pressure, baseline diastolic blood pressure, baseline eGFR, increase rate of hemoglobin level during darbepoetin administration for 4 weeks, baseline urine protein (spot urine), dose of darbepoetin administered per week, antihypertensive agent, antithrombotic agent, and lipid-lowering agent. As history of CV disease and cerebrovascular disease were strongly correlated (Spearman’s rank correlation coefficient < 0.7), only history of CV disease was selected for the model
AE adverse event, CI confidence interval, CV cardiovascular, eGFR estimated glomerular filtration rate, HR hazard ratio
P values for trend (Wald chi-square test [type 3]) were p < 0.0001 for age and p = 0.0375 for baseline eGFR
The median (25, 75 percentiles) of darbepoetin dose administered per week was 17.1 (11.4, 26.5) µg