#2315 Impact of iron and erythropoiesis-stimulating agent dose on mortality of hemodialysis patients with cancer

Abstract

Abstract Background and Aims Anemia and iron deficiencies are frequent in hemodialysis (HD) patients and in patients with cancer. If specific guidelines on the management of erythropoiesis-stimulating agent (ESA) in HD patients with cancer have been published, the association between the iron status and the risk of morbidity and mortality in this population remains elusive. Method This retrospective, observational study include HD patients who had been diagnosed with cancer after the start of HD, between September 2009 and March 2021. Demographics, lab test, iron status and cumulative dose of iron and ESA were collected quarterly, from one year before cancer diagnosis to one year after. Univariate analyses were performed to assess the impact of iron or ESA exposure on mortality, after modelling exposure before and after diagnosis of cancer. Then, each model was adjusted, using the baseline variables that were found to be independently associated with mortality. Cox proportional-hazards modelling was used for all regression analyses. P-value < .05 is considered significant. Results Seventy-nine patients were included, with a median follow-up of 2.2 years (IQR: 0.9–5.3). At baseline, the median age was 72 years (IQR 63–79), 30.4% were women. The median dialysis vintage before the diagnosis of cancer was 3 years (IQR 2-7). The main cancers found were prostate (14%), colorectal (13%), bronchopulmonary (11%), urothelial (10%) and skin (10%). Lymph node involvement were observed in 18% and metastasis in 16% patients. The median hemoglobin level was 10.4 g/dL (IQR 9.9–11.7), the median ferritin level was 412 ng/mL (IQR 213–626). The median cumulative quarterly dose of iron was 600 (215–900) mg and the cumulative dose of ESA was 71 (IQR 40–148) UI/kg/week of darbopoetin alfa or equivalent. The year following the cancer diagnosis, the median dose of iron administered remained stable, but with significant variations between patients. There was an increase in the median dose of ESA, particularly after the 6th month, up to 134 (35–314) UI/kg/week. The median hemoglobin remained between 10 and 12 g/dL. The median ferritin level rises moderately during the first 6 months, then returns to values equivalent to those preceding the diagnosis of cancer. Mortality after 12 months of follow-up was high (mortality rate of 47.4%). Age, the presence of a atrial fibrillation, lymphatic invasion, metastatic invasion, or the need for more than one antitumor treatment were independently associated with mortality. In adjusted analysis (using baseline characteristics associated with death), an association with all-cause mortality was only observed with the cumulative dose of ESA in the year preceding the diagnosis of cancer (HR 1.41 [1.06–1.88], P = .017). In multivariate analysis, an association was observed between all-cause mortality and the cumulative dose of ESA in the year preceding the diagnosis of cancer, with an HR of 1.37 (1.03-1.83, P = .031). This association between all-cause mortality and cumulative dose of ESA was also found in the year following cancer diagnosis, but without reaching significance (HR 1.37 [0.95-1.98], P = .092). However, there was no association between mortality and the cumulative dose of iron administered before or after the diagnosis of cancer in multivariate analyses. Conclusion In HD patients diagnosed with cancer, the amount of iron administered was not associated with an increased risk of mortality, probably making it possible to increase the doses administered, if necessary, without any additional risk. However, it was not possible to guarantee the safety of increasing the dose of ESA in these patients. These findings should be confirmed in a larger population in order to define guideline for iron and ESA supplementation in this population.