Trial design
This was a randomized, controlled, multicenter, pragmatic clinical trial. Patients were randomly assigned in a 1:1 ratio to one of two groups. The intervention group received the EPO treatment and the control group received standard care without EPO. This study aimed to assess the superiority of EPO use in acute kidney injury with anemia against no use of EPO.
Participants and eligibility criteria
Patients included in this trial were admitted to one of the five hospitals where the investigators of this study were practicing: Saint-George Ajaltoun Hospital, Bellevue Medical Center, Serhal Hospital, Sacre-Coeur Hospital, Middle-East Institute of Health.
All adult patients > 18 years old hospitalized with acute kidney injury and anemia were eligible. Acute kidney injury was defined based on the RIFLE, AKIN and KDIGO criteria, as an increase of serum creatinine of 0.3 mg/dL within 48 hours or 1.5 times the baseline level. Anemia was defined in this trial as requiring erythropoietin if Hb <11 g/dl. Since the decrease in hemoglobin levels can be very rapid in acute settings due mainly to inflammatory causes and since the onset of action of erythropoietin takes several days, we started erythropoietin before the patient reaches lower levels of hemoglobin.
Eligible patients were included after giving their informed consent to participate.
Exclusion criteria were: pregnant women, terminally ill patients, active bleeding, patients with major or minor thalassemia, patients on dialysis and patients who were receiving rHuEPO or any erythropoiesis-stimulating agent (ESA) before admission.
Data collection
Data for presumed cause of acute kidney injury, comorbidities, medications and laboratory results were collected from the patients' medical records. The following variables were studied: age, gender, home altitude, body mass index (BMI), diabetes, current smoking status, hypertension, hyperlipidemia, previous cardiovascular disease, chronic inflammatory disease, previous chronic obstructive pulmonary disease (COPD), baseline serum creatinine (Scr) if available (with corresponding estimated glomerular filtration rate (eGFR) using the CKD-EPI equation), Scr, hemoglobin level and C-Reactive Protein (CRP) at the time of AKI diagnosis (T1) and before discharge or death (T2). We collected as well serum phosphate, calcium, albumin, bicarbonate, white blood cells, platelets, ferritin, transferrin saturation (TSAT), LDH, vitamin B12, folic acid, reticulocyte count, uric acid and CPK. Data on previous medications intake were collected: iron, non-steroidal anti-inflammatory drugs (NSAIDs), antihypertensive medications specifically renin-angiotensin-aldosterone system inhibitors (RAASi) such as angiotensin-converting enzyme (ACE) inhibitors and angiotensin receptor blockers (ARBs), antiplatelet and anticoagulant agents, urate lowering therapy, antibiotics, immunosuppressive treatment and corticosteroids. Medications administered during the hospitalization were collected particularly vasopressors such as noradrenaline, dopamine, furosemide, antibiotics, anticoagulants, vitamins, enteral or parenteral nutrition and proton-pump inhibitors. Data including average hospital length of stay (LOS), oligo-anuria at any stage of the AKI, need for dialysis and number of days till renal recovery were collected. Adverse events were also noted such as any thrombotic event.
Ethical considerations
The study got the approval from the ethics committee of Saint-Joseph University number CE-HDF1115 and is in agreement with the Helsinki Declaration of 1975. The patients signed an informed consent before entering the trial. The trial is registered on ClinicalTrials.gov (NCT03401710, 17/01/2018). Informed consent was signed by the patient or a member of his family. Each participant was assigned two numbers, one for the unit and another for each individual. The data analyst was blinded regarding group allocation.
Interventions
Patients were randomly assigned to one of two groups: Group 1 received erythropoietin (EPO) at least 150 UI/Kg/week subcutaneously divided into three doses per week and EPO treatment was stopped if Hb reached 12 g/dl and above. Group 2 received the usual treatment. Treatment was started within 24 hours of the diagnosis of concomitant AKI and anemia.
Outcomes
Primary outcome:
Need for red blood cell transfusion during the hospitalisation.
Secondary outcomes:
-Renal recovery whether full recovery defined as a decrease of serum creatinine at discharge to the patient’s baseline or to less than 1.5 mg/dl or partial recovery defined by any reduction in serum creatinine at discharge.
-Need for dialysis.
-All-cause mortality.
Sample size calculation
If we consider a two-sided alpha of 5% and power of 80% and an effect size = 0.4 (Cohen’s effect size, i.e., standardized mean), the total sample size needed would be 198 patients, 99 patients in each arm.
Randomization
Patients were assigned to receive EPO or not, using a 1:1 allocation ratio. We used the randomization plan from the www.randomization.com to generate the random allocation sequence. Each time one of the investigators enrolled a new participant, the others were informed. Being a pragmatic trial, the investigators and patients were not blinded to treatment.
Pragmatic trial: After randomization, the investigators were free to treat and manage the patient based on their usual real-world practice.
Participant timeline
Every patient was followed from first day of acute kidney injury until discharge or transfer or death.
Statistical analysis
Continuous variables are presented as mean ± standard deviation (SD) if normally distributed and as median and interquartile range (IQR) if skewed. Categorical variables are reported as numbers and percentages. Differences between the two groups of the trial were compared using Chi Square test for categorical variables and Mann-Whitney or t independent test for continuous variables. The risk ratio was calculated for each outcome with the confidence interval (CI). A categorical regression analysis was performed to assess the factors associated with each outcome in the two groups of the trial. Statistical analysis was performed using the Statistical Package for the Social Sciences (IBM SPSS, version 24). A P-value of <0.05 was considered statistically significant.
As of 25/08/2021, a total of 134 patients had been recruited into the study, accounting for 68% of the planned sample size. No further patients could be recruited for several months due to Covid-19 outbreak and the disruption of the usual management of the patients. A decision was reached by the PI, the investigators, the statistician and the ethics committee to terminate prematurely the trial, given the forced zeroing of recruitment rate and the quasi-impossibility of reaching the planned sample size within a reasonable time.
A post-hoc power analysis was performed for the primary outcome for transparency purposes, calculating the power to detect the initially planned effect size of 0.4 with the effective sample size (ref: The 20% Statistician: Observed power, and what to do if your editor asks for post-hoc power analyses (daniellakens.blogspot.com)). Power analysis was performed using GPower software v3.192 (ref: Faul, F., Erdfelder, E., Lang, A.-G., & Buchner, A. (2007). G*Power 3: A flexible statistical power analysis program for the social, behavioral, and biomedical sciences. Behavior Research Methods, 39, 175-191).