MO311: Acute Kidney Injury in Leukaemia Patients Submitted to Allogeneic Haematopoietic Stem Cell Transplant—Incidence and Risk Factors: A Cohort Analysis

Abstract

Abstract BACKGROUND Studies on acute kidney injury (AKI) in hematopoietic stem cell transplant (HSCT) consider several haematologic diagnoses in their cohorts and heterogeneous definitions for AKI not taking into consideration urinary output (UO). We aimed to evaluate the incidence and risk factors of the first AKI episode in patients with leukaemia submitted to allogeneic HSCT occurring in the first 100 days post-HSCT, considering both SCr and UO criteria in KDIGO classification. METHODS We conducted a single-center retrospective cohort study, including patients with leukaemia admitted for allogeneic HSCT between 2005 and 2015. KDIGO classification was used for AKI diagnosis considering daily values of SCr and 6-h UO from admission day for HSCT until hospital discharge, and weekly evaluations within the first 100 days beyond hospital discharge. To estimate AKI cumulative incidence and to establish AKI risk factors, survival analysis methods considering death as a competing risk were used. RESULTS A total of 164 patients, 54.3% female, 91.5% Caucasian, median age 39.1 (P25 = 28.1–P75 = 50.4), body mass index 23.2(20.9–25.3), HCT-CI (haematopoietic cell transplantation-specific comorbidity index) < 2 in 85.4% of patients. Of them, 55.5% had acute myeloid leukaemia, 33.5% had acute lymphoblastic leukaemia, 8.5% had chronic myeloid leukaemia and 2.4% had rare forms of leukaemia. A total of 6.7% had been submitted to radiotherapy in the past, median number of chemotherapy cycles was 3 (2–4). A total of 71.3% were prescribed a reduced-intensity conditioning regimen, 56.1% had a related donor and peripheral blood was the source of progenitor cells in 13.4%. In the 100 days post-HSCT, 89.6% developed sepsis, 82.9% were exposed to nephrotoxic drugs, 39.0% had hypovolemia, 25.0% complicated with shock and 7.9% went to intensive care unit. A total of 71.3% developed graft versus host disease, 32.3% registered cytomegalovirus infection. The cumulative incidence of AKI was 58.5% at 30 days post-HSCT and 63.4% at 100 days post-HSCT. AKI diagnosis was first made by SCr criteria in 76.9%, by UO criteria in 15.4% and by both in 7.7%. The highest stage of AKI was 1 in 61.8%, 2 in 21.6% and 3 in 16.7%. A total of 12.5% underwent renal replacement therapy. In our multivariable model, variables independently associated with a higher incidence of AKI were: HCT-CI < 2 [HR 1.88, 95% confidence interval (95% CI) 1.13–3.11; P = .015], previous radiotherapy (HR 2.07, 95% CI 1.06–4.03; P = .034), shock (HR 1.57, 95% CI 1.02–2.39; P = .039), LDH (HR 1.51, 95% CI 1.03–2.21; P = .035) and sepsis (HR 3.36, 95% CI 1.22–9.24; P = .019). CONCLUSIONS AKI incidence affects almost two-thirds of patients with acute and chronic leukaemia submitted to allogeneic HSCT. More than 25% of AKI patients reach stages 2–3. HCT-CI, previous radiotherapy, shock, LDH and sepsis are important independent AKI risk factors. This should bring the attention to the need of developing prevention strategies towards these aspects. This is the first study considering both SCr and UO for defining AKI by KDIGO classification in HSCT, as well as focusing specifically on patients with leukaemia. This approach contributes for a more accurate evaluation and bias reduction.