For the anemia of kidney disease, trial after randomized trial has shown a lack of benefit and the presence of harm with higher hemoglobin targets. In a surprising shift, complete correction of anemia staved off the progression of chronic allograft nephropathy in a small randomized, controlled trial of kidney transplant recipients published in the Journal of the American Society of Nephrology (2012;23:360–368).
However, clinicians are counseled to sit tight, as more research is needed to determine what implications, if any, the results hold for patient care, experts in kidney disease-related anemia said in phone interviews.
“I thought the study was interesting and definitely hypothesis generating, but I don't think it should necessarily change our management or treatment of patients,” said Ajay K. Singh, MBBS, MBA, Chair of the Nephrology Times Editorial Board, Senior Nephrologist at Brigham and Women's Hospital, and Associate Professor of Medicine at Harvard Medical School.
The findings came from the open–label Correction of Anemia and Progression of Renal Insufficiency in Transplant patients (CAPRIT) study of 128 patients treated at 17 centers in France. Gabriel Choukroun, MD, PhD, Chief of the Nephrology, Dialysis, and Transplantation Department at Amiens University Hospital in Amiens, France, was the lead author.
“We found that in the group of patients with a hemoglobin level close to normal, which means that they were around 13 g/dL, the rate of decline of renal function was lower compared with the group of control patients, and the number of patients reaching end-stage renal disease and the number of graft failures was lower in this treatment group compared with the control group, so we clearly show that correcting anemia in transplant patients reduces the rate of decrease of renal function and reduces the number of grafts lost,” Dr. Choukroun said.
Still, given the conclusions of the much larger clinical trials of patients with chronic kidney disease (CKD), the CAPRIT findings should not yet be adopted in clinical practice, he agreed.
“I would like to have this result confirmed by another study to make sure that there is a real impact of correcting anemia,” Dr. Choukroun said.
Estimated Creatinine Clearance
In the CAPRIT trial, the mean age of participants was 49, and 87% were first kidney transplant recipients. The transplants had taken place at least 12 months before study enrollment, with a mean time from transplantation of eight years.
At baseline, the maximum creatinine clearance, as estimated using the Cockcroft-Gault formula, was 50 mL/min/1.73 m2, and hemoglobin level had to be less than 11.5 g/dL. Patients in both groups had a mean estimated creatinine clearance of 38.3 mL/min/1.73 m2 and a mean hemoglobin level of 10.5 g/dL at this time point.
Of the 128 patients randomized in the study, three from the control group were excluded in the first month because they had no hemoglobin measurement. This left 63 patients in the complete correction group, with a hemoglobin target of 13–15 g/dL, and 62 patients in the partial correction group, with a hemoglobin target of 10.5–11.5 g/dL. Epoetin-ß was the erythropoiesis-stimulating agent (ESA) used to target these levels.
Patients were followed for 24 months or until study withdrawal or loss to follow-up. The two-year follow-up assessment was not completed in six patients (9.5%) in the higher hemoglobin group and 18 patients (29.0%) in the lower hemoglobin group because they went back on dialysis, were not compliant, or died.
At the conclusion of the trial's correction phase, the mean hemoglobin level was 13.1 g/dL in the complete correction group and 11.4 g/dL in the partial correction group, with this statistically significant difference persisting.
The primary efficacy endpoint was the variation in Cockcroft-Gault creatinine clearance between study inclusion and the 24-month follow-up assessment. By this time point, the mean estimated creatinine clearance had decreased by 2.4 mL/min/1.73 m2 in the complete correction group and 5.9 mL/min/1.73 m2 in the partial correction group, a statistically significant difference.
One of the secondary outcomes was glomerular filtration rate (GFR) estimated by the Modification of Diet in Renal Disease study equation. The mean last available estimated GFR at study's end was 32.6 mL/min in the complete correction group and 28.0 mL/min in the partial correction group.
During the study, three patients in the complete correction group (4.8%) and 13 patients in the partial correction group (21%) progressed to end-stage renal disease (ESRD) and returned to dialysis. The two-year death-censored graft survival was 94.6% in the complete correction group and 80.0% in the partial correction group, and no acute graft rejections were observed.
“I was very surprised, as I've read this a couple times now, that eight years after transplant, on average, a relatively small intervention would have been associated with such a dramatic difference in outcome,” said Jeffrey S. Berns, MD, Professor of Medicine and Pediatrics at the Perelman School of Medicine at the University of Pennsylvania, when asked to comment on the results in a phone interview.
“I wondered what happened in the control group—because these people were presumably stable at the time of enrollment—that all of a sudden they sort of dropped off and had a rather substantial decline in kidney function compared with the other group for what I think is a relatively small clinical difference in terms of either ESA dose or hemoglobin. We're only talking about a hemoglobin difference of about a gram and a half or so.
“What they don't show in the study, which would have been interesting to see, is what was the rate of decline of kidney function prior to enrollment. All they really say is serum creatinine variation was less than 20% in the previous three months.”
Truncated Safety Analysis
The safety analysis was restricted to patients who received at least one dose of the study treatment—63 patients in the complete correction group and 49 in the partial correction group—rather than the accepted practice of including all randomized participants.
Adverse events were mostly mild or moderate, occurring in 65.1% of patients in the complete correction group and 67.3% of patients in the partial correction group. There were no strokes or cardiac events in the higher hemoglobin group.
One patient died in the complete correction group (esophageal carcinoma) and three patients died in the partial correction group (hemophagocytosis, amyloidosis, and severe polyradiculitis).
“There are not enough patients in the group and the follow-up wasn't long enough to be confident about mortality outcomes and cardiovascular disease risk outcomes, so I think really the conclusion is limited to the observations about rates of progression for chronic allograft nephropathy,” Dr. Berns said.
“I think it's important that this be confirmed, maybe with more precise measures of kidney function, because they really used surrogates of kidney function here and creatinine clearance and some estimating formulas, as opposed to actual measurements of GFR, but that adds a substantial burden to any type of study.”
The study was funded in part by grants from Roche France and the Clinical Research Department of Amiens University Hospital.
Different Patient Populations
While the results of the CAPRIT study don't match those of trials in patients with kidney disease, it's important to note the distinctions between the transplant and CKD populations.
“Kidney failure is a different animal in kidney transplant patients than it is in people with kidney disease,” said Donald Hricik, MD, Division Chief of Nephrology & Hypertension at University Hospitals Case Medical Center, when asked to comment on the study in a phone interview.
“First of all, most patients with kidney disease have a single disease: it's polycystic disease; it's diabetic nephropathy; it's hypertensive nephrosclerosis. In kidney transplant patients, this is a mixed bag.
“We lose allografts because of recurrent episodes of acute rejection, recurrence of the original disease, viral infections, new diseases, or, in the long run now, the latest concerns are loss of kidneys from antibody-mediated disease. It has nothing to do with the original disease. Now having said that, I'm not sure if that explains the disparity.”
Certain characteristics of the trial complicate its interpretation, Dr. Hricik noted.
“The patients in the study had actually been transplanted for quite a while,” he said. “To get into this study, you had to be more than 12 months out from transplant, but, on average, they were eight years posttransplant.
“This is a group of patients who've already demonstrated that their graft survival is pretty good, and I'm not sure if the results of the study are relevant to dealing with newly transplanted patients, where anemia's so common and we deal with the same kinds of issues.”
Another factor is that 69.8% of patients in the complete correction group and 71.0% of patients in the partial correction group were on an angiotensin-converting enzyme inhibitor or an angiotensin II receptor blocker.
“That's important because those agents are well known to cause anemia in transplant recipients,” Dr. Hricik said.
“They stratified for that so that the percentage of patients receiving those drugs was the same in each group, but it just adds more noise to interpreting the data. Are there other things going on than just correcting the hemoglobin and giving more EPO [erythropoietin] that account for the excellent results that they say? Having said that, I still think it's an important paper.”
In terms of immunosuppressive agents, most patients were on double (46%) or triple (54%) combination therapy. Mycophenolate mofetil was used in 76% of patients, cyclosporine in 74%, and corticosteroids in 68%.
“There was a variety of medicines used, although they point out carefully that the distribution of the agents was about the same in each of the groups,” Dr. Hricik said. “Still, some immunosuppressive agents are more likely to cause anemia than others.”
Sirolimus use was not recommended in the study.
“That's fine except there are a lot of patients out there on sirolimus, and that's another drug that's known to cause significant anemia in many patients,” Dr. Hricik said.
“The results of the study don't apply, I don't think, to patients who are on sirolimus because they often require huge doses of erythropoietin products and are often very resistant to even high doses.”
Dr. Choukroun commented on the exclusion of this patient group.
“We didn't want to include patients with sirolimus or any mTOR [mammalian target of rapamycin] inhibitor because we wanted to have all the patients on a CNI [calcineurin inhibitor],” he said. “CNIs are supposed to be nephrotoxic.”
Dr. Choukroun and colleagues proposed a potential mechanism for the results.
“We think that anemia is responsible for hypoxia, and, in the kidney, hypoxia may be one of the causes of an increase in fibrosis, so by increasing hemoglobin level, we are able to increase oxygen transport in the kidney, decrease hypoxia, and decrease fibrosis,” Dr. Choukroun said. “That's probably the main explanation for the result.
“Another one can be a link to the role of EPO therapy directly on the prevention of cyclosporine toxicity. There is some experimental data showing in animals that when you give cyclosporine to rats you induce kidney injury—this has been known for 20 years—but if you use a high dose of EPO you can decrease this lesion and protect the kidney.”
The anemia experts commenting on the CAPRIT results were not convinced of this mechanism, particularly given the results from the Trial to Reduce Cardiovascular Events with Aranesp Therapy (TREAT; Pfeffer MA et al: N Engl J Med 2009;361:2019–2032), Correction of Hemoglobin and Outcomes in Renal Insufficiency (CHOIR; Singh AK et al: N Engl J Med 2006;355:2085–2098), and Cardiovascular Risk Reduction by Early Anemia Treatment with Epoetin Beta (CREATE; Drüeke TB et al: N Engl J Med 2006;355:2071–2084).
“Eighty-nine percent versus  percent of patients got treated with epoetin, so I can't conceive of a mechanism by which a mildly higher hemoglobin or somewhat higher EPO dose would uniquely protect a transplant kidney when we have over 6,000 patients in CHOIR and CREATE and TREAT showing that using these drugs and targeting higher hemoglobin don't improve kidney survival,” said Daniel W. Coyne, MD, Professor of Medicine at Washington University School of Medicine in St. Louis.
But there may be players in the observed relationship between anemia correction and kidney protection that are not yet clear.
“Transplant patients are exposed to different drugs, and it's possible that the mechanism by which EPO confers protection of kidney function might be different,” Dr. Singh said.
“It's a different setting than in a patient who's not a transplant patient, so certainly there may be some subtle mechanistic factors that we don't understand well that may account for a benefit here and not a benefit in the non-transplant population.
“The second is that this study used lower doses, and it's possible that that's why you saw the benefit, but in the other trials you saw a risk because they were using much higher doses.
“I think there are reasons why this might be different, but the study definitely, in my mind, merits the need for an additional study. That's what the authors concluded, which I agree with.”