Erythropoiesis-Stimulating Agents in Elderly Patients with Anemia of Unknown Etiology: Treatment Response and Cardiovascular Outcomes

Background: Anemia of unknown etiology (AUE) is a common category of anemia in the elderly where investigations cannot identify a specific cause. We have previously shown that AUE patients exhibit lower erythropoietin (EPO) levels which may be associated with either decreased EPO production or a blunted EPO response to anemia. Erythropoiesis-stimulating agents (ESAs) mimic the effect of endogenous EPO and may play a role in treating AUE. In this study, we investigated the response to ESA treatment in patients with AUE compared to other causes of anemia. We also examined the effect of ESAs on cardiovascular outcomes in our cohort. To our knowledge, no previous study has specifically assessed ESAs in AUE.

Patients and methods: We conducted a retrospective cohort study including all consecutive hematology patients referred to our centre and who had EPO levels determined between 2005 and 2013. We included patients 60 years or older who met the World Health Organization criteria for anemia (hemoglobin [Hb] <130 g/L in men, <120 g/L in women) excluding patients with insufficient electronic medical records. The cohort was subdivided into a group treated with ESAs and an untreated group. Three reviewers independently adjudicated each patient's anemia to one of four diagnostic groups: chronic kidney disease (CKD), myelodysplastic syndrome (MDS), AUE, or other miscellaneous etiologies. The etiology reported by at least two of the three reviewers was used in the analysis, with differences resolved by consensus. Inter-observer agreement was assessed using Fleiss' Kappa statistic. Treatment response was defined by at least a 15 g/L increase in the Hb level from baseline, or a decrease of at least 4 transfusions over 8 weeks, compared to pretreatment. We performed logistic regression to measure the association between the anemia etiology and treatment response while controlling for the following potential confounders: sex, age, weight, Charlson's comorbidity index, Hb, EPO level, estimated glomerular filtration rate (eGFR), and the presence of additional cytopenias. To evaluate safety we identified each documented cardiovascular event in the cohort including ischemic stroke, myocardial infarction, pulmonary embolism, deep vein thrombosis or portal vein thrombosis. We generated Kaplan-Meier curves comparing cardiovascular events and cardiovascular event-free survival between the treated and untreated groups. Hazard ratios were calculated using Cox regression analysis.

Results: The inclusion criteria were met by 570 of 1511 potentially eligible patients. Of the 113 patients treated with an ESA, data was adequate to assess treatment response in 101 patients. Of the patients treated with an ESA, the mean age was 75.1 years and 60% were male. The mean pretreatment hemoglobin was 88.5 g/L. Inter-observer agreement for diagnostic categories was adequate. Eighty-two patients were treated with epoetin alfa and 19 patients received darbepoetin alfa. Treatment response was better in the CKD and AUE groups (54% and 47%, respectively) compared to the other groups. Compared to the group of other etiologies logistic regression analysis showed a 3.6 and 3.3 adjusted odds ratio (OR) for response for CKD and AUE respectively, although this was not statistically significant (Table 1). A baseline EPO level <200 IU/L was associated with a response to ESAs (OR 9.3; 95% CI 1.1-75.4). There was no significant difference in cardiovascular events or cardiovascular event-free survival between the treated and untreated groups, even after adjusting for confounders (Table 2).

Conclusion: Our results suggest that ESAs can be used to treat anemia of unknown etiology, and responses may be similar to those in chronic kidney disease. This supports the notion that a relative EPO deficiency is probably related to the pathogenesis of AUE. Although treatment may be associated with increased cardiovascular events, this was not found to be significant in our cohort. Limitations of this study include its retrospective nature and a relatively small sample size. Further studies exploring the safety and efficacy of ESAs in the treatment of AUE are warranted.

Table 1

View largeDownload slide

Odds ratio of treatment response in unadjusted and adjusted logistic regression models

Table 1

View largeDownload slide

Odds ratio of treatment response in unadjusted and adjusted logistic regression models

Close modal

Table 2

View largeDownload slide

Hazard ratios for cardiovascular outcomes in patients receiving ESAs in unadjusted and adjusted Cox regression models

Table 2

View largeDownload slide

Hazard ratios for cardiovascular outcomes in patients receiving ESAs in unadjusted and adjusted Cox regression models

Close modal