Cancer Related Anemia in the Developing World: Risk Factors and Treatment Patterns

Abstract

Abstract 4747 Introduction:Cancer related anemia (CRA) is an important problem in limited resource settings due to highly prevalent rates of anemia and malnutrition in the population. Due to a high baseline prevalence of anemia in India (56.2% in females and 24.3% in males), we hypothesized that CRA would have a higher prevalence in India compared to the West. Methods:We used a cross sectional design to evaluate the prevalence of CRA in cancer patients admitted to a 1100 bedded tertiary care centre located in Bangalore, India between January 2005 and December 2007. Using a standardized questionnaire, we collected data from the medical records of patients that fulfilled the inclusion criteria of this study. Patient charts were identified by the ICD code and the data so obtained was further validated using histopathological records and cancer registry information. A total of 579 patients with cancers of the breast, cervix, prostate, head and neck, lung, Hodgkin’s and non Hodgkin’s lymphomas were included in this study. We excluded those patients with known hemoglobinopathies, HIV infection, patients on myelosuppressive therapy for indications other than cancer, patients with chronic renal disease, and pregnant women. Anemia was defined as Hb < 12g/dl and was further categorized as mild (10-11.9g/dl), moderate (8.0-9.9g/dl) and severe (<8.0g/dl) according to WHO criteria. Descriptive statistics were used to characterize anemic and non anemic patients. The chi square test of association was used to identify significant differences between the groups. Binary logistic regression was used with the presence of anemia as the dependant factor to calculate the crude and adjusted odds ratios for anemia. Results:Of the 579 patients admitted to the hospital, 300 patients (52%) had anemia. The majority of patients had mild anemia (31%) with 14% and 6% having moderate and severe anemia respectively. On bivariate analysis, gender (p=0.010), reason for admission (p<0.001), first admission for cancer (p=0.003), and fatigue as a presenting complaint (p=0.001) were significant differentiating factors between the anemic and non anemic groups. A previous history of or ongoing chemotherapy (p=<0.001), ongoing radiotherapy (p=0.017) and history of recent surgery (p=0.032) were factors that differentiated the two groups.After adjustment for potential confounders, on binary logistic regression analysis only gender, admission for surgery, and a presenting complaint of fatigue were significant factors for anemia. When compared to males, females cancer patients were more likely to be anemic (AOR=2.017, 95% CI 1.084–3.753; p=0.027). Compared to patients who did not present with fatigue, patients reporting fatigue during admission had a significantly higher risk for anemia (AOR=1.703, 95% CI 1.022–2.838; p=0.041). Surprisingly, patients being admitted for surgery had significantly lower odds of having anemia compared to patients admitted for diagnosis (AOR=0.469, 95% CI 0.228–0.966; p=0.040), although this could indicate prophylactic blood transfusion prior to surgery in this group.Only 33% of the 300 patients with cancer related anemia received treatment for the anemia. The most favored treatment method was blood transfusion (17%), while hematinic therapy (Iron tablets) and the combination of transfusion and hematinics were used in the remaining patients. Of note, none of the patients in this setting received erythropoietin stimulating agents (ESAs). Conclusions:CRA is a major complication in the management of cancer; however, its prevalence in South India is not higher than in developed countries. Female sex is associated with greater odds of having CRA which may reflect the higher baseline incidence of anemia in women in India. Since micronutrient deficiencies could contribute to high rates of CRA, the role of oral and parenteral iron for treatment independent of ESAs could be further explored. Additional studies to identify the etiology of CRA in these patients will help aid decisions on treatment in limited resource settings. [Display omitted] Disclosures:No relevant conflicts of interest to declare.