MO532: Cardiovascular Events in Patients With Non–Dialysis-Dependent Chronic Kidney Disease and Anemia: Regional Analysis of Patients Previously Treated With Erythropoiesis-Stimulating Agents in the PRO2TECT Trial

BACKGROUND AND AIMS

Vadadustat is an oral hypoxia-inducible factor prolyl hydroxylase inhibitor being investigated for the treatment of anemia due to chronic kidney disease (CKD). In global phase 3 trials, vadadustat has demonstrated non-inferiority to darbepoetin alfa for time to major adverse cardiovascular event (MACE) in patients with dialysis-dependent CKD (INNO2VATE trials), but not with non–dialysis-dependent CKD (NDD-CKD; PRO2TECT trials) [1, 2]. In a prespecified subgroup analysis of the PRO2TECT trials, a difference in the relative safety between vadadustat and darbepoetin alfa was observed between the US and non-US regions [1]. We investigated regional differences in MACE in patients with NDD-CKD-administered vadadustat and darbepoetin alfa who were previously treated with erythropoietin-stimulating agents (ESAs) in the PRO2TECT trials.

METHOD

Two phase 3, open-label, randomized, active-controlled clinical trials comparing vadadustat with darbepoetin alfa were conducted in North America, Latin America, Europe, Africa, and the Asia-Pacific region. This post hoc analysis evaluated MACE in patients in the PRO2TECT trial who were actively maintained on ESAs at study entry with ≥1 dose received within 6 weeks before or during screening (Conversion trial; NCT02680574) and who received ≥1 dose of trial drug, stratified by region (USA versus Europe versus non-US/non-Europe). MACE was defined as a composite of death from any cause, nonfatal myocardial infarction or nonfatal stroke. Expanded MACE was defined as MACE plus hospitalization for heart failure or thromboembolic event, excluding vascular access failure.

RESULTS

A total of 1723 ESA-treated patients were randomized in the conversion trial of the PRO2TECT program and received ≥1 dose of study drug, including 665 patients in the USA, 444 in Europe, and 614 outside the USA and Europe. Europe had a lower proportion of patients with diabetes and more patients who had received intravenous (IV) iron, while patients in the non-US/non-European region were younger and had a lower prevalence of cardiovascular (CV) disease. A higher proportion of patients in Europe were using darbepoetin alfa at baseline than in other regions (59%–63% versus 14%–28%), though ESA dose was lower (57–61 versus 93–149 IV epoetin equivalent U/kg/week). The hazard ratio for MACE in the overall population for vadadustat versus darbepoetin alfa was 1.16 (95% CI 0.93–1.45). Across regions, event rates were similar in the vadadustat groups, but event rates in the darbepoetin alfa group were lower in Europe compared with the US and non-US/non-Europe (Table 1), which was driven by fewer total deaths (non-CV and CV) reported in Europe in the darbepoetin alfa arm (n = 24/220) compared with the vadadustat arm (n = 38/224) (Table 2). Additional post hoc analyses of MACE accounting for several baseline characteristics, including ESA dose, did not alter the outcomes of the study.

CONCLUSION

Regional differences in time to first MACE were observed in patients with NDD-CKD who were treated with ESA and randomized to receive vadadustat or darbepoetin alfa as part of the PRO2TECT program. It remains unclear if the lower risk of MACE observed in Europe with the darbepoetin alfa group was related to differences in unobserved baseline characteristics, regional treatment practices, methodological reasons or chance.

MO536: Cardiovascular Events in Patients With Anemia Associated With Non–Dialysis-Dependent Chronic Kidney Disease: Regional Analysis of Patients not Previously Treated With Erythropoiesis-Stimulating Agents in The PRO2TECT Trial

BACKGROUND AND AIMS

Vadadustat is an oral hypoxia-inducible factor prolyl hydroxylase inhibitor being investigated for treatment of anemia due to chronic kidney disease (CKD). In global phase 3 trials, vadadustat has demonstrated non-inferiority to darbepoetin alfa for time to major adverse cardiovascular event (MACE) in patients with dialysis-dependent CKD (INNO2VATE trials), but not with non–dialysis-dependent CKD (NDD-CKD; PRO2TECT trials) [1,2]. In a prespecified subgroup analysis of the PRO2TECT trials, no difference in cardiovascular safety was observed in the USA, but a higher risk of MACE was found for patients treated with vadadustat outside the USA.1 We investigated regional differences in MACE in patients previously untreated with erythropoietin-stimulating agents (ESAs) in the PRO2TECT trial.

METHOD

Two phase 3, open-label, randomized, active-controlled clinical trials comparing vadadustat with darbepoetin alfa were conducted in North America, Latin America, Europe, Africa, and the Asia-Pacific region. This post hoc analysis evaluated MACE in patients in the PRO2TECT trial not treated with ESAs within 8 weeks of enrollment (correction trial; NCT02648347) and who received ≥1 dose of trial drug, stratified by region (USA versus Europe versus non-USA/non-Europe). MACE was defined as a composite of death from any cause, nonfatal myocardial infarction, or nonfatal stroke. Expanded MACE was defined as MACE plus hospitalization for heart failure or thromboembolic event, excluding vascular access failure.

RESULTS

A total of 1748 ESA-untreated patients receiving ≥1 dose of study drug were enrolled in the PRO2TECT trial, including 1058 patients in the USA, 139 in Europe and 551 outside the USA and Europe. Patients in the non-USA/non-European countries were younger and had a lower mean eGFR than patients in the USA or Europe. Furthermore, 34.7% of patients randomized to vadadustat had a baseline eGFR <10 mL/min/1.73 m2 versus 24.0% of patients randomized to darbepoetin alfa in the non-USA/non-European countries. The hazard ratio for MACE in the overall population for vadadustat versus darbepoetin alfa was 1.16 [95% confidence interval (CI) 0.96–1.41]. When analyzed by region, higher event rate for MACE was observed in the vadadustat arm in the non-USA/non-European countries compared with the USA and Europe (Table 1). The higher event rates for MACE in the non-USA/non-European countries were driven by 21 excess MACEs reported in the vadadustat group. Many deaths in the non-USA/non-European countries were related to kidney failure (n = 25/43 in the vadadustat group; n = 20/30 in the darbepoetin alfa group; Table 2), and were concentrated in Brazil and South Africa, countries that enrolled a higher proportion of patients with end-stage kidney failure who may not have had access to dialysis. The adverse event profiles for vadadustat and darbepoetin alfa were similar across regions.

CONCLUSION

Regional differences in time to first MACE were observed in patients with NDD-CKD who were not previously treated with ESAs and randomized to receive vadadustat or darbepoetin alfa as part of the PRO2TECT trial. The higher event rate in the vadadustat group in non-USA/non-European countries may have been related to randomization imbalances and/or design and methodological issues. These findings should help inform care providers as they assess the overall safety of vadadustat for the treatment of anemia associated with NDD-CKD.

MO541HEMATOLOGIC EFFICACY OF VADADUSTAT FOR ANEMIA IN PATIENTS WITH NON--DIALYSIS-DEPENDENT CHRONIC KIDNEY DISEASE

Background and Aims

Vadadustat is a small-molecule inhibitor of hypoxia-inducible factor prolyl hydroxylases under development to treat anemia associated with chronic kidney disease (CKD). The vadadustat phase 3 program includes four efficacy and cardiovascular safety outcome trials of vadadustat versus the erythropoiesis-stimulating agent (ESA) darbepoetin alfa. Here we describe detailed results on hematologic efficacy in two phase 3, randomized trials (the PRO2TECT trials) in adult patients with non–dialysis-dependent (NDD) CKD and anemia, in which vadadustat met prespecified noninferiority criteria compared to darbepoetin alfa, with respect to hematologic efficacy (correction/maintenance of hemoglobin [Hb] target concentrations).

Method

The mean screening Hb level for the ESA-untreated NDD-CKD trial (NCT02648347) had to be <10.0 g/dL, and for the ESA-treated NDD-CKD trial (NCT02680574), the range had to be from 8.0-11.0 g/dL in the United States (US) and from 9.0-12.0 g/dL non-US. In the ESA-untreated trial, patients received no ESA within 8 weeks before randomization; in the ESA-treated trial, patients were maintained on ESA therapy, with ≥1 dose received within 6 weeks prior to or during screening. The vadadustat starting dose was 300 mg/day for all patients, whereas the initial darbepoetin alfa dose depended on each patient’s prior dose or the product label. Both vadadustat and darbepoetin alfa doses were titrated according to prespecified dosing algorithms to achieve target Hb concentrations (US: 10-11 g/dL; non-US: 10-12 g/dL) both during the primary (PEP; weeks 24-36) and secondary (SEP; weeks 40-52) evaluation periods. Herein, we present topline results from the PEP and SEP endpoints, in addition to more detailed erythrocyte parameters.

Results

A total of 3,476 patients (1751 ESA-untreated and 1725 ESA-treated) were randomized 1:1 to vadadustat or darbepoetin alfa. In both trials, vadadustat was noninferior to darbepoetin alfa with regard to the difference of mean change in Hb concentrations between baseline and PEP, as well as between baseline and SEP. The respective proportions of patients (vadadustat vs. darbepoetin alfa) with an average Hb value within the geography-specific target range in the PEP and SEP were 50.4% versus 50.2% and 43.1% versus 43.5% in the ESA-untreated trial and 60.1% versus 60.7% and 50.7% versus 49.0% in the ESA-treated trial. The proportion of patients (vadadustat vs darbepoetin alfa) who achieved an Hb increase >1.0 g/dL from baseline to week 52 was assessed only for the ESA-untreated trial and was 87.7% (95% CI: 85.4%, 89.8%) for vadadustat versus 88.0% (95% CI: 85.6%, 90.0%) for darbepoetin alfa.

Hematologic parameters at time points within the PEP and SEP are presented in Table 1. In both the ESA-untreated and ESA-treated trials, the reticulocyte count trended up from baseline through week 52 for vadadustat and trended down from baseline for darbepoetin alfa. Trends in erythrocyte mean corpuscular volume and erythrocyte mean corpuscular Hb were largely unremarkable by week 52 in both treatment groups.

Conclusion

Vadadustat demonstrated similar profiles across erythrocyte parameters compared with darbepoetin alfa in the treatment of adults with anemia in CKD not on dialysis, whether ESA-untreated or ESA-treated at study entry.

MO114NEPHROLITHIASIS AS CAUSE OF KIDNEY FAILURE AND MAJOR CARDIOVASCULAR OUTCOMES IN INCIDENT DIALYSIS PATIENTS

Background and Aims

Symptomatic kidney stone formers experience excess rates of cardiovascular events and are at increased risk of chronic kidney disease and its progression. Little is known about the cardiovascular outcomes of persons with kidney failure in whom nephrolithiasis or urolithiasis was listed as their presumed cause of kidney failure.

Method

We used the United States Renal Data Service (USRDS), a national kidney failure registry, to identify all persons initiating dialysis between 1/1/1996 and 9/30/2015. Patients were required to have Medicare fee-for-service coverage on day 90 of dialysis, which served as index date. The presumed cause of kidney failure was abstracted from information reported on the Medical Evidence Report (form CMS-2728) and categorized as: nephro-/urolithiasis, diabetes, hypertension, glomerulonephritis, cystic kidney disease, other urologic cause, and other cause. Eligible patients were then followed for the occurrence of a major adverse cardiovascular event (MACE; nonfatal myocardial infarction, nonfatal stroke, or cardiovascular mortality) using claims-based algorithms and causes of death reported in the Death Notification (form CMS-2746). Multivariable Cox regression models were fit while controlling for incident year, sociodemographic characteristics, initial dialysis modality, reported comorbidities and disabilities, biometric data (body mass index [BMI], eGFR, serum albumin, hemoglobin). Multiple imputation was used for missing data. Both cause-specific and Fine-Gray sub-distribution hazard ratios (HR) with corresponding 95% confidence intervals (CI) were estimated.

Results

Of 2,000,072 persons with incident kidney failure, 1,048,006 (52.4%) were alive and satisfied all inclusion criteria on day 90 (71.9% of the excluded due had no Medicare coverage). Among those, 2207 (0.2%) had nephro-/urolithiasis as cause of kidney failure, while 47% had diabetes, 30.4% had hypertension, 8.1% had GN, and 1.6% had cystic kidney disease listed. Persons with nephro-/urolithiasis tended to be older, more likely to be female and non-Hispanic white, had lower rates of most comorbidities, higher serum albumin and hemoglobin concentrations and lower BMI and eGFR at dialysis initiation. The composite cardiovascular event rate (MI, stroke, cardiovascular death; 721 events) was 74.4/1000 person years for nephro-/urolithiasis. Compared with those whose cause was nephro-/urolithiasis, persons whose cause of kidney failure was listed as diabetes had fully adjusted 57% (95% CI, 46%-69%) higher hazards of MACE and those in whom it was attributed to hypertension the hazards were 33% (95% CI, 24%-44%) higher. While no difference in the hazards of MACE was observed compared with persons with glomerulonephritis (HR 0.97; 95% CI, 0.90-1.05), those with cystic kidney disease had a 14% (95% CI, 7%-20%) lower hazards of MACE than those with nephro-/urolithiasis. Qualitatively similar results were obtained when examining the individual components of MACE as well as when considering kidney transplant and non-cardiovascular death as competing risks. However, competing risk analysis substantially attenuated the magnitude of the associations with diabetes and hypertension as reported cause of kidney failure. Limitations include heterogeneity of the underlying pathology among patients with nephro-/urolithiasis as well as the possibility that some patients with nephro-/urolithiasis were not captured if a different cause of kidney disease was listed.

Conclusion

Patients with kidney failure presumably from nephro- or urolithiasis have distinct cardiovascular risk profiles, with lower major cardiovascular event rates compared with patients whose kidney failure was due to diabetes or hypertension, but slightly higher rates compared with those with cystic kidney disease.

MO539HEMATOLOGIC EFFICACY OF VADADUSTAT FOR ANEMIA IN PATIENTS WITH KIDNEY FAILURE ON DIALYSIS

Background and Aims

Vadadustat is a small-molecule inhibitor of hypoxia-inducible factor prolyl hydroxylase being developed for treatment of anemia associated with chronic kidney disease (CKD). The vadadustat phase 3 program includes four efficacy and cardiovascular safety outcome trials of vadadustat versus the erythropoiesis-stimulating agent (ESA) darbepoetin alfa. Here we describe detailed results on hematologic efficacy in two of the four phase 3, randomized, open-label, sponsor-blind trials (the INNO2VATE trials) in adult patients with dialysis-dependent (DD) CKD and anemia, where vadadustat met prespecified noninferiority criteria compared with darbepoetin alfa with respect to cardiovascular safety and correction/maintenance of hemoglobin (Hb) target concentrations.

Method

The mean screening Hb range for the incident DD-CKD trial (NCT02865850) was 8.0-11.0 g/dL; for the prevalent DD-CKD trial (NCT02892149), it was 8.0-11.0 g/dL in the United States (US) and 9.0-12.0 g/dL for non-US. Patients in the incident and prevalent DD-CKD trials had initiated dialysis within <16 weeks with limited or no prior ESA exposure and >12 weeks with established ESA treatment prior to screening, respectively. Vadadustat starting dose was 300 mg/day for all patients, whereas initial darbepoetin alfa dose depended on each patient’s prior dose or product label. Both vadadustat and darbepoetin alfa doses were titrated according to prespecified dosing algorithms to achieve target Hb concentrations (US: 10-11 g/dL; non-US: 10-12 g/dL) during the primary evaluation period (PEP; weeks 24-36) and the secondary evaluation period (SEP; weeks 40-52). Herein, we present topline results from PEP and SEP endpoints, as well as other, more detailed hematologic erythrocyte parameters.

Results

A total of 3923 patients (369 with incident DD-CKD and 3554 with prevalent DD-CKD) were randomized 1:1 to vadadustat or darbepoetin alfa. Vadadustat was noninferior to darbepoetin alfa in achieving target-range Hb concentrations (primary efficacy endpoint) among patients who were new to, or established on, dialysis. The respective proportions of patients (vadadustat vs. darbepoetin alfa) with an average Hb value within the geography-specific target range in the PEP and SEP were 43.6% versus 56.9% and 39.8% versus 41.0% in the incident trial and 49.2% versus 53.2% and 44.3% versus 50.9% in the prevalent dialysis trial. The proportion of patients who achieved an Hb increase >1.0 g/dL from baseline to week 52 was assessed only for the incident trial and was 84.0% (95% CI: 77.8%, 89.0%) for vadadustat versus 89.9% (95% CI: 84.7%, 93.8%) for darbepoetin alfa.

Hematologic erythrocyte parameters at time points within the PEP and SEP are presented in Table 1. In the incident trial, reticulocyte count was slightly increased from baseline at 28 and 52 weeks for vadadustat, whereas for darbepoetin alfa, reticulocyte count was slightly decreased or unchanged in both trials. Erythrocyte mean corpuscular volume and erythrocyte mean corpuscular Hb showed increases by week 52 for both groups.

Conclusion

Vadadustat demonstrated similar profiles across erythrocyte parameters compared with darbepoetin alfa in the treatment of anemia associated with CKD in adults in both incident dialysis and prevalent dialysis settings.

Pharmacokinetics of apixaban in patients with end stage renal disease on hemodialysis and atrial fibrillation: results from the RENAL-AF trial

Background/Introduction

Apixaban use is increasing for stroke prevention in patients with atrial fibrillation (AF) and end stage renal disease (ESRD) on hemodialysis. There is uncertainty as to the optimal dose in this population in part related to the limited available pharmacokinetic (PK) data.

Purpose

We comprehensively evaluated the PK of apixaban collected over 1 month of apixaban dosing in 63 patients with AF and ESRD on hemodialysis.

Methods

Patients with AF and ESRD on hemodialysis were randomized to warfarin versus apixaban within the RENAL-AF trial with 5 mg BID dosing, except for 2.5 mg BID in those age ≥80 years or weight ≤60 kg. The 5 mg BID dose could be reduced to 2.5mg BID for minor bleeding. Day 1 PK data was collected on all patients pre- and post-hemodialysis. Day 3 and 1 month pre- and post-hemodialysis PK samples were collected in 49 patients. The timing of apixaban dosing and hemodialysis relative to PK samples was recorded. Dosing history, hemodialysis, and PK samples were chronologically integrated with patient specific data such as body size, age, race and gender. This dataset was combined with the ARISTOTLE dataset, and the published PK model from ARISTOTLE describing exposures in the AF population was updated to incorporate an additional clearance term for hemodialysis. The model estimated apixaban exposures (AUC) in RENAL-AF were compared to ARISTOLTE AUC values.

Results

There were 285 PK concentrations collected among 63 patients in the RENAL-AF trial. Patients had median age 69 years with 41% women (N=26) and a median weight of 84 kg (49, 157). The median AUCs for patients with ESRD on hemodialysis were 5,452 and 2,990 for patients treated with 5mg BID and 2.5mg BID doses, respectively. The median AUCs for patients treated with 5mg BID from ARISTOTLE increased from 2,802 for patients with class 1 CKD to 5,863 for class 4 CKD, while they increased from 2,392 for class 1 CKD to 2,881 for class 4 CKD in patients treated with 2.5mg BID. The median AUC for patients with ESRD on hemodialysis were within 50% of the exposure of patients from ARISTOTLE for all classes of CKD for the 2.5mg BID dose and for classes 2, 3A, 3B, and 4 CKD for the 5mg BID dose (Figure).

Conclusions

The steady state apixaban exposure data in patients with AF and ESRD on hemodialysis were modestly higher but consistent with the results of non-ESRD patients from ARISTOTLE, using 5 mg BID unless patients had age ≥80 years or weight ≤60 kg. Additional clinical outcomes data on the use of apixaban in patients with AF and ESRD on hemodialysis are needed.

Funding Acknowledgement

Type of funding source: Private grant(s) and/or Sponsorship. Main funding source(s): Investigator sponsored grant from Bristol-Myers Squibb and Pfizer