A pattern may be emerging with the decision by the FDA to not approve vadadustat
The decision by the FDA not to approve vadadustat for patients on dialysis was surprising and disappointing to nephrologists hoping for a safe and effective oral alternative to erythropoiesis-stimulating agents for treatment of anemia.
Most patients on in-center dialysis have a satisfactory hemoglobin (Hb) response to the administration of erythropoiesis-stimulating agents (ESAs) via the extracorporeal circuit during the dialysis treatment. However, there are patients in whom escalating doses of ESAs do not produce a satisfactory Hb response for whom an alternate treatment would be welcome.
Likewise, the Advancing American Kidney Health Initiative and the ESRD Treatment Choices model have mandated an expanded percentage of patients receiving home dialysis with financial penalties to providers who fail to achieve set targets. Administration of ESAs to patients on home dialysis is cumbersome as it requires an injection that many patients find uncomfortable or must travel to a dialysis clinic to receive.
Availability of a safe and effective oral alternative to ESAs, such as Akebia Therapeutic’s vadadustat, would make anemia treatment for home dialysis more patient-friendly.
Vadadustat is the second investigational oral hypoxia-inducible factor prolyl hydroxylase inhibitor (HIF-PHI) turned down by the FDA in the last year. On July 15, 2021, the Cardiovascular and Renal Drugs Advisory Committee (CRDAC) of the FDA voted against approval of Fibrogen’s roxadustat for the treatment of anemia due to chronic kidney disease for patients not dialysis dependent (NDD) and dialysis dependent (DD). The FDA issued a complete response letter to FibroGen a month later indicating it would not approve the drug.
A third investigational HIF-PHI, daprodustat, is under development by GlaxoSmithKline. Researchers have completed phase 3 global clinical trials and published data comparing the efficacy of daprodustat vs. darbepoetin in NDD and DD patients and reported comparable efficacy and major adverse cardiovascular event (MACE) noninferiority in both populations.
GlaxoSmithKline announced on April 19 that the FDA had accepted its new drug application for daprodustat and a Prescription Drug User Fee Act action date of Feb. 1, 2023, has been set.
The basis for the decision by the FDA to not approve vadadustat for patients on dialysis is unclear and confusing. Vadadustat demonstrated comparable efficacy (maintenance of target Hb level) and MACE safety noninferiority to ESAs in its phase 3 INNO2VATE global clinical trial involving 3,950 patients on dialysis.
The hazard ratio (HR) for MACE (vadadustat vs. darbepoetin) was 0.96 (95% CI, 0.83-1.11) with a prespecified upper bound of 1.25. The HRs for expanded MACE, death from cardiovascular cause and death from any cause were 0.96, 0.96 and 0.95, respectively, with none of the 95% CI upper bounds exceeding 1.25.
Adverse events, drug-related adverse events, serious adverse events and serious drug-related adverse events were comparable between the groups receiving vadadustat and darbepoetin. There were more patients discontinuing vadadustat than darbepoetin due to adverse events and drug-related adverse events, but this may have been related to the fact that patients receiving vadadustat knew they were receiving an experimental drug because the study was not a double-dummy blinded design.
Reasons for rejection
The FDA decision regarding vadadustat did not include an evaluation by a CRDAC as did the prior evaluation of roxadustat by the FDA. Public deliberation of a CRDAC, along with the release of the committee’s available FDA briefing document regarding the candidate drug, led to transparency regarding the decision by the FDA to issue a complete response letter (which is not public) regarding roxadustat.
Without that CRDAC review, the decision by the FDA not to approve vadadustat for treatment of anemia in patients on dialysis was not transparent. A press release from Akebia detailing the basis for the FDA decision did not appear consistent with publicly available safety data regarding the use of the drug in DD patients.
The press release cited “vascular access thrombosis in dialysis patients” as a concern of the FDA, but a poster presentation at the the ASN Kidney Week in 2021 by Patrick S. Parfrey, MD, FASN, and colleagues showed no difference between vadadustat and darbepoetin-treated patients, with 6.6 vascular access thrombosis events/100 patient years in both groups of the INNO2VATE trial.
The press release from Akebia also cited FDA concerns regarding drug-induced liver injury with vadadustat. There is no mention of drug-induced liver injury in the publication of the INNO2VATE trial nor its online supplementary materials. Publication of the phase 3 PRO2TECT global clinical trial of vadadustat vs. darbepoetin in NDD patients with anemia involving 3,476 patients reported (in the online supplementary appendix) a 1.8% incidence of increased transaminases among previously ESA-untreated patients receiving vadadustat vs. 1% incidence of increased transaminases among previously ESA-untreated patients receiving darbepoetin.
The percentage of previously ESA-treated NDD patients experiencing transaminase elevations was slightly less among patients receiving vadadustat vs. darbepoetin; the percentage of patients (previously ESA-untreated or treated) experiencing alanine aminotransferase increase or aspartate aminotransferase increase was equal or lower among patients treated with vadadustat vs. darbepoetin.
In any event, the FDA may have conflated any concern regarding drug induced liver injury from the NDD population with the DD population. Without more granular publicly available data upon which the FDA based its decision not to approve vadadustat for DD patients, the rationale for the FDA decision remains a matter for speculation.
Patients not on dialysis
The rationale for the FDA decision not to approve vadadustat in NDD patients is more clear-cut and was expected by most observers. Vadadustat failed to meet the prespecified upper 95% CI bound of 1.25 for MACE vs. darbepoetin in the PRO2TECT study (HR=1.17; 95% CI, 1.01-1.36), although it came closer to the upper bound when only the U.S. population in this global study was examined (HR= 1.06; 95% CI, 0.87-1.29).
The greater unmet need for an oral anemia therapy is in the NDD population. Winkelmayer and colleagues reported that older adults (67 years or older for whom Medicare data are available) with CKD in the 2 years prior to initiating dialysis were more likely to receive a blood transfusion (40.3%) vs. ESA (35.0%). In this group, 39% of these patients received no anemia treatment at all.
Transfusions are associated with allosensitization, which results in longer waiting time for transplantation because of decreased availability of compatible donors, increased risk of rejection and poorer graft survival. Highly sensitized transplant recipients require more costly immunosuppressive treatments.
Future of HIF-PHIs
In the wake of the FDA decision not to approve vadadustat, one must wonder what happens next in the United States to the HIF-PHI drug class. Five HIF-PHIs (including roxadustat and vadadustat) have been approved in Japan; roxadustat has also been approved in China, South Korea, Chile and the European Union.
While the FDA has accepted GlaxoSmithKline’s new drug application for daprodustat, there are areas of concern that might jeopardize its approval given the disposition of the FDA for roxadustat and vadadustat.
Daprodustat demonstrated MACE noninferiority among 3,871 NDD patients randomized to the study drug vs. darbepoetin followed for a median of 1.9 years in the primary intention-to-treat (ITT) analysis. With the HR of 1.03 (95% CI, 0.89-1.19, prespecified upper bound 1.25) in the supportive analysis of first occurrence on-treatment MACE in NDD patients, daprodustat failed to achieve non-inferiority (HR=1.40; 95% CI, 1.17-1.68). NDD patients treated with daprodustat had a significant increase in esophageal or gastric erosions (relative risk [RR]=1.70; 95% CI, 1.16-2.49, P = .005) and in cancer-related death or tumor recurrence (RR 1.47; 95% CI, 1.03-2.10, P = .04) compared with darbepoetin-treated controls.
Given the possibility the FDA may extrapolate adverse events observed in the NDD population to the DD population, these findings make FDA approval of daprodustat less than assured even for DD patients among whom these safety signals were not reported.
Nephrology lags far behind other medical specialties in pharmaceutical and technological innovation. This has become a barrier to the recruitment of trainees into our fellowship programs and results in the loss of opportunities to improve the care, outcomes and quality of life of our patients.
There is no reason HIF-PHIs should not be approved in the United States with appropriate safety warnings in the labels as has been done in other countries. Post-marketing registries of adverse events can be established to assure pharmacovigilance. The FDA should trust practitioners and patients to responsibly avail themselves of new therapeutic choices to advance American kidney health.
- Chertow GM, et al. N Engl J Med. 2021;doi:10.1056/NEJMoa2035938.
- Eckardt K-U, et al. N Engl J Med. 2021;doi:10.1056/NEJMoa2025956.
- Parfrey PS, et al. Abstract PO0463. Presented at: ASN Kidney Week; San Diego.
- Singh AK, N Engl J Med.2021;doi:10.1056/NEJMoa2113380.
- For more information:
- Jay B. Wish, MD, is professor of clinical medicine at Indiana University School of Medicine and chief medical officer for outpatient dialysis at Indiana University Health. He is also vice-chair of the Editorial Advisory Board for Nephrology News & Issues. He can be reached at firstname.lastname@example.org.