Correspondence
Effects of Normal as Compared with Low Hematocrit Values in Patients with Cardiac Disease Undergoing Hemodialysis and Receiving Epoetin
N Engl J Med 1998; 339:2023-2024December 31, 1998
- Article
To the Editor:
In their assessment of normal as compared with low hematocrit values in patients with cardiac disease who were undergoing hemodialysis and receiving epoetin, Besarab et al. (Aug. 27 issue)1 report seemingly paradoxical results. Although the mortality rate among the patients assigned to the normal-hematocrit group was higher than that among those in the low-hematocrit group, within each group, the mortality rate decreased with increasing hematocrit values.
The authors point out that these disparate outcomes are not explained by differences in base-line demographic and clinical characteristics or laboratory values or by differences in the dose of epoetin, and they emphasize mechanisms associated with the correction of anemia in patients undergoing dialysis that may lead to their premature death (e.g., higher doses of parenteral iron dextran). An alternative hypothesis is that refractoriness to high-dose epoetin may select out patients with ill-defined but nevertheless pernicious factors leading to death, such as infection and other inflammatory states, low values for Kt/V (where K is the rate of urea clearance by the artificial kidney, t is the duration of each hemodialysis session, and V is the volume of distribution of urea in the patient), malnutrition, and conditions conferring a predisposition to chronic blood loss.2 Do the investigators have data that would allow them to explore the possibility that resistance to epoetin simply serves as a marker for the presence of underlying conditions in these patients that may lead to their premature death?
2 ReferencesAllan H. Sklar, M.D.
United Health Services Hospitals, Binghamton, NY 13903Sri Narsipur, M.D.
State University of New York Health Science Center, Syracuse, NY 132101
Besarab A ,Bolton WK ,Browne JK , et al. The effects of normal as compared with low hematocrit values in patients with cardiac disease who are receiving hemodialysis and epoetin. N Engl J Med 1998;339:584-590
Full Text | Web of Science | Medline2
NKF-DOQI guidelines. Anemia of chronic renal failure. V. Inadequate epoetin response. (Guidelines 20-22.) New York: National Kidney Foundation, 1997.
To the Editor:
It is difficult to understand why the patients in the normal-hematocrit group had a higher mortality rate than those in the low-hematocrit group and yet within both groups, the mortality rate decreased with higher hematocrit values. The authors attribute the results mainly to the potentially deleterious effects of excess iron doses given the patients in the normal-hematocrit group and downplay the importance of the difference in the intensity of dialysis between the two groups, with more patients in the normal-hematocrit group undergoing inadequate dialysis.
Perhaps more important than the decrease in the adequacy of dialysis was extracellular volume expansion. The authors note that there were no significant differences in the use of six categories of cardiovascular drugs between the two groups before or during the study, but they do not mention whether the drug doses were changed differently in the two groups, or whether there were differences in the New York Heart Association (NYHA) functional class or in the postdialysis weight of the patients in the two groups during the study. Patients with cardiac dysfunction are susceptible to acute and chronic increases in plasma volume, especially in situations associated with an increase in total vascular resistance such as occurs during epoetin therapy.1 Perhaps subtle volume overload in the patients in the normal-hematocrit group explains the higher mortality rate in this group.
1 ReferencesFernando González, M.D.
Universidad de Chile, Santiago, ChileAuthor/Editor Response
The authors reply:
To the Editor: Drs. Sklar and Narsipur ask whether coexisting conditions explain the higher mortality rate among the patients with lower hematocrit values. As we stated in our article, regression analysis revealed a 30 percent decrease in the risk of death or myocardial infarction per 10-point increase in the hematocrit, after adjustment for age, sex, race, adequacy of dialysis (Kt/V), NYHA class, type of vascular access, and presence or absence of diabetes mellitus, hypertension, congestive heart failure, ischemic heart disease, and peripheral vascular disease. A lower hematocrit value remained a significant risk factor after the serum albumin concentration was added to this analysis, as a surrogate marker for malnutrition. We cannot exclude the possibility that bleeding, infection, or inflammation contributed to both anemia and mortality. It is still important to recognize that within each group, higher hematocrit values were associated with a lower mortality rate, and it does not appear that higher hematocrit values in themselves accounted for the overall difference in mortality between the groups.
In reply to Dr. González: in our discussion, we noted the differences in the adequacy of dialysis (Kt/V values) between the groups as an explanation of the results. He proposes alternative hypotheses to explain the results with respect to mortality. Though formal measurements of extracellular volume were not performed, predialysis weight was recorded monthly, and optimal postdialysis weight was estimated quarterly. Neither increased during the study in either group. Blood pressure did not increase. We have no evidence that patients in the normal-hematocrit group had more complications of volume overload, and the finding that a higher hematocrit was associated with a lower mortality rate does not support the theory that increased blood volume underlies the differences in mortality.
With respect to cardiac drugs, many patients had divergent changes in their drug regimens, making exact comparisons difficult. Hence, we analyzed the proportion of patients in each group who took any quantity of each of six categories of drugs. There were no differences between the groups. Finally, the NYHA class did not differ between the groups at base line or during the study.
Allen R. Nissenson, M.D.
University of California, Los Angeles, Los Angeles, CA 90024Steve J. Schwab, M.D.
Duke University Medical Center, Durham, NC 27710David A. Goodkin, M.D.
Amgen, Thousand Oaks, CA 91320- Citing Articles (1)
Citing Articles
1
Dennis J. Cotter, Kevin Stefanik, Yi Zhang, Mae Thamer. (2004) Improved Survival with Higher Hematocrits: Where is the Evidence?. Seminars in Dialysis 17:3, 181-183
CrossRef
