II. Target Hemoglobin/Hematocrit


The initial patient experience with Epoetin came in a Phase I-II clinical trial in hemodialysis patients with the anemia of CKD. The target maintenance Hct for these patients was 35% to 40%, ie, at the lower range of normal.35 When investigators met to design the Phase III multicenter clinical trial, hematologists argued that the target Hct should be a normal Hct, while nephrologists proposed a lower level. A compromise target Hct of 35% was used in the trial. The final Hct levels for the more than 300 patients treated with Epoetin in the Phase III trial ranged from 33% to 38%. The results of this study,65 together with those of the Phase I-II clinical trial, were submitted to the FDA. The FDA approved Epoetin therapy in June, 1989, but the target Hct range recommended by the FDA was only 30% to 33%, for reasons that have never been clear. The FDA recommendation is probably responsible for the previously held belief that a target Hct of 30% to 33% is medically appropriate. In spite of the FDA’s decision in June 1994 to widen the target hematocrit range to 30% to 36%, the USRDS data derived from practice in 1993 (United States Renal Data System 1996 Annual Data Report)66 showed that the mean Hct for Epoetin-treated dialysis patients in the United States was still in the lower end of this target range (30.2%), with 43% of patients having Hct values <30%. By the end of 1997 the mean Hct increased to 32.4% (USRDS 1999 Annual Data Report).67

Most of the initial physiologic and quality of life studies of anemic predialysis and dialysis patients treated with Epoetin in the United States had target Hct values of 36%. Virtually all studies have shown that, with increased Hct, there is marked improvement in various physiologic measures–oxygen utilization [VO2]4-8; muscle strength and function68; cognitive and brain electrophysiological function15; cardiac function9,12-14,69,70; sexual function18; or quality of life.24 While two reports have cautioned that a target Hct greater than 30% could result in clotting of various arteries, as well as underdialysis,71,72 these predictions have failed to materialize or be substantiated. Two other groups of investigators reported that there were no differences in various physiologic and quality of life measures between hemodialysis patients with Hct (or Hgb) levels of 30% (9 to 10 g/dL) versus 36% (11 to 12 g/dL).73-75 However, a re-examination of these data allows for a different conclusion: the data are difficult to interpret in one study, and some physiological parameters were better at the higher Hgb/Hct in the other study (Endnote b). The authors of one of these studies have recently completed various physiological and quality of life studies in a small number of hemodialysis patients and have clearly shown that a normal Hgb (14 g/dL) is superior to a Hgb of 10 g/dL.76-79

Since 1989, when the FDA established its guidelines for the target Hct, and the Health Care Financing Administration (HCFA) established a policy under which it would not reimburse dialysis centers for the use of Epoetin when the Hct was above 36%, there have been few studies published in the United States which examine whether a Hct higher than 36% is more beneficial than a Hct of 30% in dialysis patients. While there are many studies that have shown the benefits of Hct values 36%, in most cases the comparison was made to outcomes of patients with an Hct level of <25% (see above). In order to formulate our recommendations regarding the target Hgb/Hct, the Anemia Work Group reviewed only peer-reviewed studies that compared baseline Hgb/Hct levels of 10 to 11 g/dL/30% to 33% (which is the current target level in the United States and most other countries) to higher values. Review of the literature which involved predialysis and dialysis patients within and outside the United States showed that, compared to higher Hgb/Hct values, Hgb/Hct values 11 g/dL/<33% are associated with increased morbidity and mortality. In addition, a number of recent United States and non-United States studies reported in abstracts indicate that patients with CKD function better at Hct levels that are near normal or normal and that improvement is continuous as the Hgb/Hct increases above 10 g/dL/30% to normal levels. The only exception to this has been a study sponsored by Amgen that involved more than 1,200 hemodialysis patients with documented heart disease. This study was discontinued when it appeared that those patients randomized to a target Hct in the normal range (42% 3%) were experiencing a greater incidence (30%, with a confidence interval of 0.9 to 1.9) of non-fatal myocardial infarctions or death than did the control group randomized to a target Hct of 30% 3%.80 The difference was not statistically significant at the time the study was terminated, however. Additional studies are needed to clarify the relationship between Hgb/Hct and outcomes in CKD patients, particularly those with heart disease. Such studies should be designed to determine the highest Hgb/Hct that provides incremental benefits without serious side effects. Several multicenter studies addressing this question are in progress outside the United States. A study determining whether the "prevention" of anemia and its associated adverse effects could also be of value, since one of the aims of treating anemia is to prevent or retard the development of heart disease.


Target Hemoglobin/Hematocrit for Epoetin Therapy

The target range for hemoglobin (hematocrit) should be Hgb 11 g/dL (33%) to Hgb 12 g/dL (36%). (Evidence) This target is for Epoetin therapy and is not an indication for blood transfusion therapy. (Opinion)

Rationale A Hgb of 11 g/dL (Hct 33%) is at the lower limit of the normal range for pre-menopausal females and pre-pubertal patients; a Hgb of 12 g/dL (Hct 36%) is just below the lower limit of the normal range for adult males and post-menopausal females (see Guideline 1: When to Initiate the Work-up of Anemia). Because the anemia literature in CKD patients does not distinguish between sexes, subsequent Hgb/Hct levels will apply to both males and females.

There are several pieces of evidence suggesting that patient outcomes are worse when the Hgb is 10 g/dL (Hct 30%):

1. Survival of dialysis patients declines as the Hct decreases below a range of 30% to 33%.25,81 Survival was also shorter in dialysis patients with chronic glomerulonephritis whose mean Hgb level was 9.9 g/dL, compared to patients with polycystic kidney disease whose mean Hgb level was 11.3 g/dL.82 Whereas one study failed to note any improved survival at a Hgb >11 g/dL compared to an Hgb 10 to 11 g/dL,83 several other reports have shown improved survival at higher Hgb/Hct levels. Survival was improved in Italian hemodialysis patients when the Hct exceeded 32%, either spontaneously or following Epoetin therapy, when compared to Hct <32%,84 and in the United States an Hct of 33% to 36% reduced the risk of death from any cause by 10% when compared to patients whose mean Hct was 30% to 33%.85 Survival has been noted in one study to be better in patients with cardiac disease who attained and maintained a normal Hct compared to similar patients who did not attain and maintain a normal Hct.80 In fact, within both the normal Hct group and the control group, the mortality decreased at higher Hct levels.80 In those 200 patients who achieved and maintained a normal Hct for 6 months, mortality decreased to approximately 15% per year, versus 40% per year in those maintained at an Hct of 30%. There were no convincing factors that appeared to explain why those patients that did not achieve and stabilize at a normal Hct had a greater incidence of non-fatal myocardial infarctions or death than did the control group.80

2. Left ventricular hypertrophy (LVH) is more likely in CKD patients with anemia (Hct 33%)86-88 and in patients with ESRD89; in such patients the risk of death is increased 2.9-fold (Endnote c).90 Partial correction of anemia (Hgb 6.3 0.8 to 11.4 1.5 g/dL) with Epoetin resulted in partial regression of LVH in dialysis-dependent patients.91 Angina was significantly decreased in patients with progressive CKD when Epoetin therapy increased the Hct to 31% 4% versus 23% 4%.92

3. Quality of life either is not improved, or improved only slightly, when the Hgb/Hct is increased from 8 g/dL/25% to a level no higher than 9 to 10 g/dL/28% to 30%.93-95 However, quality of life of dialysis patients, as assessed by standardized patient questionnaires, increases as the Hgb/Hct increases above 10 to >12 g/dL and 30% to >36%.96-98 When the results of the Amgen Phase III (mean Hct 35%) and Phase IV (mean Hct 30%) studies were compared, it was concluded that patients with Hct levels of 35% had better quality of life as measured by Karnofsky scores than those maintained at a Hct of 30%.99 Both quality of life and various physiological parameters in predialysis patients have been shown to be significantly better at a Hct of 36% to 39% than at 27% to 29%.100-107

4. In hemodialysis patients, exercise capacity (VO2) increased when the Hct increased from 30% to 35% to 40%.108

5. In hemodialysis patients, the incidence of hospitalization was lower when the Hct was 33% to 36% in comparison to patients with lower Hct values.109

6. There are a number of studies in dialysis patients (reported initially only in abstracts) that indicate that quality of life, maximum exercise capacity, number of meters walked in 6 minutes, cardiac output, cognitive function, amino acid levels, sleep dysfunction with daytime sleepiness, insulin resistance with hyperlipidemia, and survival improved when a normal Hct was achieved,76-78,110-118 and that there were no adverse effects observed at a normal Hct.110,111 Several of these studies have now been published in peer review journals, and demonstrate that a normal Hgb/Hct is associated with better physical performance,79 better cognitive function,119 improved brain oxygen supply,120 and improved sleep patterns121 compared to lower Hgb/Hct levels.

Studies in patients with anemia due to conditions other than CKD also indicate that an Hct of 30% is harmful. Patients undergoing peripheral vascular surgery had more cardiac ischemia when their Hct decreased from 39% to 27% to 30% compared to others whose Hct decreased only to 32%.122 Patients with lupus nephritis have increased mortality, unrelated to kidney dysfunction, as Hgb decreases to below 11 g/dL.123 Patients with anemia related to cancer have improved quality of life when Hct is increased from 29% to >35% with Epoetin.124 Finally, an Hgb <10 g/dL increases prematurity rates to almost twice normal in otherwise healthy pregnant women.125 Since anemia is treated in these conditions to improve patient well-being and survival, patients with CKD should not be deprived of the same therapeutic goal.

Despite this plethora of data, there has been much controversy as to what Hgb/Hct is best in CKD patients. Prior to the availability of Epoetin, evidence was provided that, during neurosurgical procedures, relative oxygen transport capacity for the human brain was optimal at an Hct of 30% to 33%, as achieved by hemodilution (usually phlebotomy combined with dextran infusion).126 This was based on in vitro hemodilution studies involving blood flow through glass capillaries.127 This long-standing belief among anesthesiologists, neurologists, and neurosurgeons was supported by data showing that hemodilution increased cerebral blood flow. This belief has been challenged by more recent studies, which indicate that, while there is an inverse relationship between Hct and cerebral blood flow, there is a linear relationship between Hct and oxygen delivery to brain tissue, with the maximal level of oxygen delivery occurring within an Hct range of 40% to 45% (Endnote d).128 There also was concern during the early experience with Epoetin about possible adverse effects when the Hct was increased above 30%. These adverse effects are not, in fact, seen today, except for an increase in the need for antihypertensive medications in 23% of patients with CKD, whose blood pressure can be controlled with appropriate clinical care (see Guideline 24: Possible Adverse Effects Related to Epoetin Therapy: Hypertension). Two studies of small numbers of unselected dialysis patients found no adverse effects when target Hct was increased to a mean of 38.3%110 and 42%.111 Hypertension was no more frequent in those cardiac patients who attained a normal Hct than patients maintained at an Hct of 30.4%.129 In a large multicenter study of 1,200 hemodialysis patients with documented cardiac disease (congestive heart failure or ischemic heart disease), hypertension was no more frequent at a normal Hct (42% 3%) than at an Hct of 30% 3%.80 There was no association between adverse outcomes and either (a) the average Hct during the study or (b) the Epoetin dose administered. No conclusions can be drawn from the results of this trial with regard to CKD patients who do not have heart disease (as defined in the Amgen study).

The only published investigation relating Epoetin therapy to increased cardiovascular disease is from Okinawa, Japan.130 The authors reported that the annual incidence of stroke and acute myocardial infarction increased following the use of Epoetin in CKD patients (Endnote e). However, these findings differ from recent European data, in which cardiovascular morbidity and mortality were decreased by 15% to 20% after 2 years of Epoetin therapy in CKD patients.36

Clinicians should bear in mind that approximately 5% of hemodialysis patients attain an Hct 40%131 without receiving Epoetin. In addition, some Epoetin-treated patients may have a spontaneous increase in Hct to above 36% after Epoetin therapy has been discontinued. There have been no published reports of an increased incidence of deaths and/or non-fatal myocardial infarctions in such patients. In view of recent abstracts76-78,111-118 and peer-reviewed articles79,119,120,122 showing beneficial effects from raising the Hgb/Hct to normal, it is possible that a target Hgb/Hct higher than what the Anemia Work Group recommends now may ultimately prove to be appropriate.

In summary, based on currently available data, the Anemia Work Group recommends that the Hgb/Hct be maintained between 11 and 12 g/dL (33% and 36%). In striving to maintain the Hgb/Hct within this target range, the Hgb/Hct will likely rise above this range. The reasons why some patients will temporarily exceed an Hgb/Hct of 12 g/dL (36%) is that the response to Epoetin varies amongst patients, the interplay between IV iron supplementation and epoetin dosing may be unpredictable, and it is impossible for the bell-shaped distribution of Hgb/Hct for all patients be limited to between 11 and 12 g/dL or 33% and 36%, respectively. As of January 2000, HCFA will continue to provide reimbursement for the cost of Epoetin alfa even if the Hgb/Hct temporarily rises above their target range, as long as the rolling 3-month average Hgb/Hct is <12.5 g/dL (37.5%). Medical justification is needed for maintaining the Hgb (Hct) above 12 g/dL (36%).





© 2001 National Kidney Foundation, Inc

web version created by cyberNephrologyTM and The Nephron Information Center