Hemoglobin Variability and Mortality in ESRD

Although multiple studies suggest higher hemoglobin (Hb) levels improve patient quality of life, physical function, and exercise capacity, randomized trials have not demonstrated an association between higher Hb levels and increased survival. One explanation for this observation is that the measurements of absolute Hb level may be insufficient to capture the effects of Hb level on clinical outcome. Other indices of Hb status, such as rate of change or difference between observed level and expected level, may provide better prediction of clinical outcome. Therefore, the authors used a linear regression model to define the independent association between patient outcome and Hb status, described by either (1) absolute Hb level (intercept), (2) rate of Hb change (slope), or (3) Hb variability (residual standard deviation).

The Fresenius Medical Care database was used to extract records of patients receiving hemodialysis (HD) treatment. Inclusion of subjects was based on having at least two Hb measurements during the six-month exposure window (needed to measure the absolute change in Hb level) and at least three Hb measurements during the exposure window (needed to define residual standard deviation (SD)). The primary analysis considered patients with no missing Hb data or missing covariate values, who remained enrolled throughout the Hb exposure window. A review of 34,693 medical records of HD patients yielded 8797, 5246, and 2523 subjects who satisfied the inclusion criteria and had a minimum Hb level greater than 9.5 g/dL, greater than 10.0 g/dL, and greater than 10.5 g/dL, respectively.

After adjusting for multiple covariates and confounders, the analysis found that each 1 g/dL Hb level increase in the residual SD was associated with a 33% increase in rate of death. As expected, longer survival was also associated with a higher absolute Hb level and a positive value for Hb temporal trend. For the groups with minimum Hb level greater than 9.5 g and Hb level greater than 10.0 g/dL, the association between Hb-Variability and survival was of the same magnitude and the same direction as the overall population. However, for the group with Hb level greater than 10.0 g/dL, the confidence intervals were wide because of the small sample size.

The use of a regression-based metric to analyze a historic national cohort was able to identify an association between Hb variability and mortality that was independent of either hemoglobin level or Hb trend. Although the linear regression method identified an association between several patient variables and greater Hb variability (possibly due to large sample size), the overall R2 for the model was 0.019, indicating that the collective impact of these factors accounted for very little of the observed variation in Hb variability among subjects.

Please reference the source article:
Hemoglobin variability and mortality in ESRD. Yang W, Israni RK, Brunelli SM, Joffe MM, Fishbane S, Feldman HI. J Am Soc Nephrol. 2007 Dec;18(12):3164-70.

NAAC Expert Commentary:
Hemoglobin (Hb) cycling and other forms of Hb variability has recently been reported in HD patients receiving epoetin treatment. The factors contributing to this variability are probably multiple, and are thought to include such things as changes in epoetin and intravenous iron doses, hemodilution due to variable interdialytic weight gains, acute blood loss, and intercurrent medical events such as surgery, infection, and other acute medical illnesses. The clinical implications of Hb variability are being investigated, with the intent of ascertaining whether Hb variability is: (1) just a phenomenon of epoetin therapy that is itself of no clinical consequence; (2) a reflection of variability in any of several patient-specific factors, but not a direct cause of adverse clinical sequelae; or (3) a direct cause of adverse patient outcomes.

Historically, our clinical focus, and that of studies attempting to associate Hb levels with outcome, has been on the absolute Hb level at any time. This study and a few others extend this discussion to the impact of changes and trends in Hb level. This study uses sophisticated analytic methods to study the relationships between absolute Hb level, the Hb trend, the extent of variability in Hb level, and other factors affecting patient mortality.

Their conclusion that greater Hb variability is associated with a decrease in survival is intriguing, and begs for additional studies. Unfortunately, observational studies such as this are incapable of proving causation. Although observational studies can only suggest associations, these studies cannot completely eliminate the possibility that unmeasured confounding factors may interfere with outcomes. Despite claims to the contrary (not by these authors, but by others), there is very little evidence to suggest that the use of darbepoetin alfa or other longer-acting agents are associated with less Hb variability or with any different outcomes, compared to the use of short acting epoetins.

Anemia Management for Patients with Lymphoid Malignancies

Lymphoid malignancies are the leading cause of new cancer cases (4.6%) and cancer deaths (3.6%). Because many lymphoid malignancies have a prolonged disease course, patients typically undergo several courses of intensive anticancer treatments to maintain clinical and/or molecular remission. These may include radiation, cytostatic drug combinations, high-dose chemotherapy with bone marrow or stem cell support, or monoclonal antibody therapy. Anemia is associated with both the malignancy and treatment of Hodgkin's disease and non-Hodgkin's lymphoma, and anemia negatively impacts all aspects of patient quality of life, treatment outcomes, and overall survival. Anemia remains vastly under-recognized and under-treated. Thus the authors felt, an update of the current guidelines is warranted, including the definition, identification, and optimal management of anemia.

Throughout several years of experience, data has continued to show that anemia can be corrected by epoetin treatment, with either epoetin alfa or darbepoetin alfa. In addition, epoetin beta has been approved in countries outside of the United States. Among various organizations, treatment guidelines for optimal dosage vary only slightly. Organizations such as the National Comprehensive Cancer Network (NCCN), the American Society of Hematology (ASH), and the American Society of Clinical Oncology (ASCO) recommend treatment guidelines that suggest intervention with erythropoietic therapy when hemoglobin (Hb) levels fall below the 10-11 g/dL range or based on appropriate clinical symptoms.

High rates of anemia have been associated with patients with lymphoid malignancies, with some studies reporting 22-82% at presentation, and 55-88% following chemotherapy. In patients with non-Hodgkin’s lymphoma, the severity of chemotherapy-induced anemia was grade 1-2 in 55-63% of patients and grade 3-4 in 9-79% of patients, depending on drug regimen. In patients with Hodgkin's disease, the rates of anemia were 5-31% (grade 1-2) and 0-13% (grade 3-4). It is important that the goals of treatment remain the same for all patients with anemia, regardless of whether the anemia is mild or moderate-to-severe.

The etiology of anemia in patients with lymphoid malignancies may include neoplastic cell infiltration of the bone marrow, nutritional deficiencies, hemolysis, and most often, treatment-induced bone marrow suppression.

Anemia has been reported in one survey as the leading cause of fatigue, and the survey also reported that fatigue was the most frequent and long-lasting symptom in patients with hematological malignancies. Findings from the Functional Assessment of Cancer Therapy-Anemia Scale have shown that quality of life correlates directly with the degree of anemia in cancer patients.

Despite the high prevalence and broad impact of anemia, and the availability of current and new treatment options, rates of anemia treatment remain as low as 50% in symptomatic or anemic patients with hematological malignancies. Anemia treatments under development include CERA, a continuous erythropoietin receptor activator, CNTO an erythropoietin receptor agonist, Hematide™, a synthetic peptide-based erythropoiesis-stimulating agent, and FG-2216, and orally active, small molecule-inhibitor of hypoxia-inducible factor-prolyl hydroxylase.

This article also discusses the choice of appropriate method for providing iron supplementation, which is a key issue for anemic cancer patients. Issues of poor iron absorption, defective iron storage, and adverse events associated with supplements must be considered when providing iron supplementation.

Please reference the source article:
Guidelines and recommendations for the management of anaemia in patients with lymphoid malignancies. Henry DH. Drugs 67(2): 175-94.

NAAC Expert Commentary:
For hematologist and oncologists immersed in the clinical trenches, 2007 will be remembered for conflict, consternation, and confusion regarding the management of anemia with recombinant erythropoietin. This treatment, so widely embraced and commonly employed, recently fell into high level and very public regulatory scrutiny. Praised primarily for its demonstrable improved quality of life in responding anemic patients, EPO has become controversial as evidence emerges indicating higher risks for thromboembolic events and enhanced tumor growth. The review by Dr. Henry provides a thorough, balanced, and well-referenced overview of the rationale behind current recommendations from the various advisory panels including the American Society of Hematology/American Society of Clinical Oncology (ASH/ASCO), the National Cancer Center Network (NCCN) and the European Organization for Research and Treatment of Cancer (EORTC). In fact, the reader is referred to a recent update of the ASH/ASCO guidelines(1) published after Dr. Henry’s review. Most agree that anemia negatively affects lymphoma patients and also that treatment of anemia is often associated with improved quality of life. Yet, most would also agree, as Dr. Henry points out, additional research is warranted to provide greater confidence regarding safety and more precision about optimal dose and schedule.

Reference in the commentary:
(1) Use of epoetin and darbepoetin in patients with cancer: 2007 American Society of Hematology/American Society of Clinical Oncology clinical practice guideline update. Rizzo JD, Somerfield MR, Hagerty KL, Seidenfeld J, Bohlius J, Bennett CL, Cella DF, Djulbegovic B, Goode MJ, Jakubowski AA, Rarick MU, Regan DH, Lichtin AE. Blood. 2008 Jan 1;111(1):25-41.

Iron Deficiency Anemia and Stroke in Young Children

Childhood ischemic stroke occurs at a frequency of 2-3 cases per 100,000 children each year in North America. Although rare, it can lead to devastating long-term consequences and death. Childhood ischemic stroke includes arterial ischemic stroke (AIS) and sinovenous thrombosis (SVT). Several risk factors for childhood ischemic stroke have been consistently reported in large cohort studies. In addition, iron-deficiency anemia (IDA) has emerged as a likely culprit. IDA is relatively common in developed countries, occurring with a prevalence of 4-8%. In fact, one large cohort study found anemia in 25% of 115 previously healthy children with AIS. To investigate whether IDA is a risk factor for childhood ischemic stroke, Maguire et al conducted a case-control study, comparing the prevalence of IDA in otherwise healthy children at the time of stroke with the prevalence of IDA in age-matched healthy children.

Previously healthy children age 12-38 months were selected for the study sample because nutritional IDA is highly prevalent in this age group. Case patients were initially selected from the Canadian Pediatric Ischemic Stroke Registry and became eligible for the study upon presenting to the Hospital for Sick Children with AIS or SVT. Control subjects were selected from a database of children prospectively enrolled in an observational study of iron deficiency. Baseline variables and characteristics of the two groups were compared to determine the association between stroke and IDA and thrombocytosis.

From the period of 1992 through 2004, 56 previously healthy children presented with stroke; 15 were generally healthy before the stroke and were included as case patients. For the control group, 332 healthy children met full eligibility criteria, of which 143 of these had available blood work. IDA was more common among case patients than control subjects (53% vs. 9%; OR: 12; 95% CI: 4–37). Thrombocytosis was also more common among case patients and control subjects (47% vs. 16%; OR: 5; 95% CI: 2–14). After using logistic regression to control for platelet count, the association between stroke and IDA remained significant (adjusted OR: 10; 95% CI: 3-33) There was no significant interaction between IDA and thrombocytosis.

According to the data, previously healthy children who develop vaso-occlusive stroke are 10 times more likely to have IDA then healthy children who do not develop stroke. More than half of the stroke cases among children without underlying medical illness consisted of children with IDA, suggesting that IDA is a significant risk factor for stroke in otherwise healthy children.

Please reference the source article:
Association between iron-deficiency anemia and stroke in young children. Maguire J, deVeber G, Parkin PC. Pediatrics 120(5): 1053-7.

NAAC Expert Commentary:
This is a case-controlled study that examines the association of iron-deficiency anemia (IDA) and stroke in children between the ages of 1 and 3 years. The authors report that among young children with stroke (n=15, after exclusion criteria were applied), 53% were found to have IDA, whereas only 9% (13 of 143) case-control subjects had IDA. While there are a number of reasons to interpret the results cautiously, the study adds more fuel to the fire of the already-recognized adverse effects of iron-deficiency in the young, such as impaired learning and developmental delay. The authors mention—but do not discuss in detail—how IDA could predispose to stroke: ‘a hypercoagulable state directly related to iron deficiency/anemia; thrombocytosis secondary to IDA; and anemic hypoxia.’ In a NAAC expert's opinion, only ‘anemic hypoxia’ seems a likely predisposition to stroke. However, there may be a ‘hypercoagulable’ feature of IDA erythrocytes that warrants mention. Several investigators(1-3) have reported that iron-deficient red cells have impaired deformability and, therefore, would be expected to have more difficulty traversing the microcirculation. Although not uniformly agreed upon, this is an important consideration. Regardless of the eventual risk of stroke attached to IDA in the young child, iron deficiency and anemia, either alone or together, are associated with adverse health consequences, further emphasizing the need to recognize and appropriately treat these common problems.

Reference in the commentary:
(1) The influence of iron deficiency on erythrocyte deformability. Reinhart WH. Br J Haematol. 1992 Apr;80(4):550-5.
(2) Red cell membrane stiffness in iron deficiency. Yip R, Mohandas N, Clark MR, Jain S, Shohet SB, Dallman PR. Blood. 1983 Jul;62(1):99-106.
(3) Deformability of erythrocytes in iron deficiency anemia. Tillmann W, Schröter W. Blut. 1980 Mar;40(3):179-86.