Tag: iron

Posted on

The state of anemia research

When we last left the anemia wars, the erythropoietin stimulating agents (ESA, which is Epo, Darbe, and CERA) were in full retreat. The ESAs had conquered ESKD and were (and still are) the standard of care to treat anemia. From that position they made a strategic strike on pre-dialysis CKD, targeting anemia in a larger slice of the world. There are 30 million patients with CKD, the market is big but relatively few of them have hemoglobin that is really low. Only 15% of patients with CKD have anemia, and that is using <12 g/dl in woman an < 13 g/dl in men.

One of the prominent theories at the time was to increase the hemoglobin to around 13 to reduce the cardiovascular disease that is so prevalent in CKD (remember the vast majority of these CKD 3 patients will die of heart disease long before they are even smelling a dialysis center).

This strategy was actually not a new idea. We had already tried this is dialysis patients. Besarab’s Normalization of Hematocrit was published in 1998. This showed that bringing the hemoglobin to 13 was bad in hemodialysis. It increased access failure, it nearly increased total mortality, and there was no signal of any benefit.

Despite Besarab’s red flag three randomized controlled trials were done to show that a high hemoglobin was beneficial in CKD, CHOIR, CREATE, and TREAT. All of them failed to show a benefit from the high hemogbin and all of them showed various, and not always consistent safety signals. The bottom line lesson of these three studies (four, if you include Besarab’s) was not to target a high (or even normal) hemoglobin with these drugs. The FDA recommended only using these drugs to avoid transfusions. KDIGO guidelines suggest a hemoglobin from 10 to 11.5 g/dl.

Nobody missed the top line result but the failure of normalization of hemoglobin also shined a light on the remainder of the ESA data. Erythropoietin was approved in the late 80’s for its ability to reduce the need for transfusions. But look as hard as you can you won’t find any data showing that treating anemia improved mortality. This is why the FDA limited use of the drug to prevent transfusion.

  • Those are quarterly transfusion rates, not annual.
  • The transfusion rate fell by two-thirds in the first year of widespread ESA use.

So that is how we left the anemia wars. ESAs retreated back to dialysis and there was a smaller amount of anemia treatment in pre-dialysis CKD to much more modest targets.

Since publishing TREAT and the revised FDA and KDIGO guidelines, anemia has slipped from the forefront of nephrology. The science has continued to mature and there is a new target, HIF-stabilizers (Hypoxia-Inducible Factor Prolyl Hydroxylase Inhibitors). he HIF-stabilizers have been winding their way from the bench to the bedside for the last 20 years and are currently in the midst of large, long phase three trials. These trials are different than the approval trials for CERA, Darbe and Epo. Those drugs needed to show the ability to correct and maintane a stable hemoglobin. The HIF-stabilizers have to be at least as safe as erythropoietin. Just like diabetes drugs after Rosiglitazone, the FDA is no longer satisfied with an improved number. the drugs need to show cardiovascular safety.

That was the landscape in which two high impact trials on anemia were announced at Kidney Week 2018.

The first was Cardiovascular Safety of Methoxy Polyethylene Glycol-Epoetin Beta in Treatment of Anemia of CKD by Francesco Locatelli. This is a cardiovascular safety study of Mircera. It looks like the same type of trial the HIF-stabilizers are now undergoing to demonstrate CV safety. This is clearly a government request and so my criticism about the study may need to be directed at the FDA/EMA. What makes me uncomfortable with this study is the choice of control. Remember these are cardiovascular safety studies, but the epoetin was originally approved not based on cardiovascular safety but on the ability to avoid transfusion. In the only placebo controlled, cardiovascular trial of ESAs, darbepoetin alfa fell short of placebo.

In this latest Locatelli trial, Mircera was found to be as safe as ESAs that were previously shown to be inferior to placebo (at least at high hemoglobin and in diabetics, with pre-dialysis CKD). It is essential, if we are going to do studies assessing cardiovascular safety that they are compared to appropriate controls. 

We should not be having patients spend years in a randomized trial trying to show non-inferiority to a drug that looks pretty inferior.

This is in opposition to the Intravenous Iron in Patients Undergoing Maintenance Hemodialysis study by Iain C. Macdougall. They showed that  an aggressive iron schedule resulted in fewer transfusions, no safety signal and some real patient oriented outcome advantages (mainly hospitalization for heart failure and transfusions, which are especially important for patients waiting on a transplant list, i.e. almost everyone on dialysis). 

Daniel Coyne, Author of the DRIVE and DRIVE II trials had a nice series of tweets about this trial:

https://twitter.com/i/moments/edit/1059278971030921216
Posted on

IV Iron in Hemodialysis

Steven Fishbane, MD from The University of Pennsylvania, came to our fellowship program to discuss Fe and hemodialysis. He began by talking about hepcidin and then went on to discuss the iron targets in light of the DRIVE trial and then touched on IV Fe and proteinuria and finished with a discussion of platelets and mortality and its relationship to recent anemia trials.

He started by talking about why Fe deficiency is so prevelant among CKD and dialysis. As an explanation he did a brief overview of Hepcidin.

Fe is a powerful growth factor for bacteria.
During an infection the body decides to withold Fe.

Inflammation leads to increases in hepcidin which increases ferritin.
Hepcidin decreases intestinal iron absorption and increases Fe sequestration in RE system. The end result is a decrease in serum Fe and TSAT.

DRIVE Study by Daniel Coin and Morin was designed to see if giving Fe in patients with a high ferritin and low TSAT was effective.

Enrollment:

  • Hgb less than 11
  • ESA over 22,500 U/week
  • Ferritin 500-1200

Ave TSAT 18, Ferritin 761, CRP 27, Hgb 10.3, ESA 34,000 U/wk

All patients had their ESA bumped 25% and randomized to either 1g of Fe or placebo.

The primary end-point was a change in hemoglobin over 6 weeks. The hemoglobin rose 10.2 to 11.3 in control and 10.4 to 11.9 in the Fe group over 6 weeks.

47% of the treatment group had an increase in hemoglobin of 2g versus 29% in the control group.

DRIVE-II by Kapoian

6 week follow-up after the randomized trial. Observational
Hgb increase was sustained despite the Fe group getting the Epo cut by 20% with no change in the in the epo dose of the control group.

Ferritin initially rose to 934 from 760+.

No different in responce in the patients with ferritin 500-800 vs 800-1200.

TSAT was not predictive of response low TSAT and high initial TSAT had the same response.

Prospectively studies cannot find link between IV Fe and infection.

Here is some criticism of the DRIVE study by the National Anemia Action Council:

The purpose of the DRIVE study and its six week extension, DRIVE II, was to determine the efficacy and safety of supplemental intravenous iron in anemic hemodialysis patients receiving recombinant erythropoietin who had a transferrin saturation <> 500 ng/mL. The intravenous iron group did increase their hemoglobin levels slightly more than control patients not given intravenous iron without additional toxicity, leading the authors to conclude that intravenous iron in this situation was both safe and effective. Unfortunately, the design and power of both studies were not sufficient for the investigators to reach these conclusions.

First, the decision to increase the dose of recombinant erythropoietin in each group by 25 % confuses the response to ESA and iron. In a chronic inflammatory state, such as chronic renal disease, the problem is not impaired iron availability but insufficient erythropoietin, which is required to mobilize iron and upregulate transferrin receptor expression. It is well recognized that erythropoietin trumps hepcidin in this situation and the authors merely confirmed that phenomenon.

Second, in the control groups of both DRIVE and DRIVE II, there were a disproportionate number of women, who are more likely to be iron deficient, and their response to recombinant erythropoietin proved this, reducing the effectiveness of comparisons.

Third, both DRIVE and DRIVE II were open label observational studies and in addition physician discretion was also allowed with respect to erythropoietin dosing and iron administration. This discretion can introduce significant bias, weakening the conclusions of the studies.

Fourth, no attempt to estimate blood loss, iatrogenic or otherwise, was made for either experimental group. Fifth, the difference in the hemoglobin level achieved with supplemental iron was not striking and also pushed the hemoglobin level above that currently recommended for safety reasons. Finally, since the serum ferritin and transferrin saturation increased in the iron-supplemented group, a state of iron overload was achieved that was unnecessary and the 12 week observation period was certainly not long enough to exclude the possibility of iron-induced organ toxicity.

It is clear that more data derived from larger prospective trials that are conducted for longer periods are needed. Until this data becomes available, anemic hemodialysis patients not responding to conventional doses of recombinant erythropoietin, in whom the serum ferritin is greater than 500 ng/mL, should first be evaluated for a source of blood loss or infection. Then the patient should be given a higher dose of recombinant erythropoietin for a minimum of 6 weeks with serial transferrin saturation and ferritin measurements before resorting to intravenous iron supplementation.

Fishbane then spoke about the possibility of proteinuria induced by iron sucrose. He showed the Agarwal data on proteinuria and IV iron. He stated that with out knowing how long the proteinuria lasted it was impossible to guess if this was important clinically.

Platelets in Nephrology

Fe deficiency causes reactive throbocytosis
ESAs raises platelets
ESAs cause fe deficiency

Increased RBC push platelets along the walls of the blood vessel. So treatment may cause a lot of atherothrombotic complications because they have both more platelets and red cells.

In DRIVE 20% drop in platelets with Fe treatment.
Streja Et al. used the Davita retrospective database to show a nice human association of TSAT and platelets with thrombosis. Importantly he showed no increase in mortality for Hgb over 13 when the data was controlled for the platelet count. Dr. Fishbane was unable to suggest a how one could design a prospective RCT to test this theory.

Posted on

My fellow just answered a question I have had for years.

Part of the dogma of evaluating iron deficiency in patients with anemia is that after a transfusion the iron indices are altered. Over and over the question comes up…

How long after a transfusion do you have to wait before checking to see if the patient is iron deficient?

Well Jabri, my current fellow went to the literature and found this reference. It looks like the answer is you should be safe 48 hours after the transfusion. This surprised me, I expected the acute effects of the transfusion to persist longer than that.

Abstract from the paper:

The effect of transfusion of packed red blood cells on serum iron level, total iron-binding capacity, and transferrin saturation was studied. Samples of blood from 37 hemodynamically stable patients were obtained for analysis at various intervals following the transfusion of packed red blood cells. In 10 patients with possible iron deficiency, a significant rise in serum iron level and transferrin saturation occurred during the 24 hours following transfusion, which persisted at a marginally significant level up to 36 hours. In the remaining 27 patients, a significant rise was also noted in serum iron level and transferrin saturation results, but the rise did not persist beyond the 24 hours after transfusion. No change in total iron-binding capacity was noted in either group. These data show that the diagnosis of iron deficiency (based on a transferrin saturation of < 0.16) might be missed if iron studies are performed on patients within 24 hours following packed red blood cell transfusion. Therefore, if serum iron studies are obtained for patients suspected of having iron deficiency anemia, these studies are best done on blood samples obtained before blood transfusion.