We performed a single-center retrospective cohort study including all patients with multiple myeloma undergoing autologous HSCT at the Centro Hospitalar Universitário Lisboa Norte, EPE (CHULN, EPE) between January 2005 and December 2015. We excluded patients under the age of 18 years, patients who had received a previous HSCT, patients with chronic kidney disease already on renal replacement therapy, and patients who underwent renal replacement therapy within one week before HSCT.

Chemotherapy regimens before transplant included the following drug combinations: cyclophosphamide, bortezomib and dexamethasone; bortezomib, doxorubicin and dexamethasone; vincristine, doxorubicin and dexamethasone; doxorubicin, thalidomide and dexamethasone; thalidomide and dexamethasone; lenalidomide and dexamethasone.

Cyclophosphamide (3 g/m2) and Granulocyte colony-stimulating factor (10 to 16 μg/kg) were used to mobilize hematopoietic stem cells which were then collected from bone marrow or peripheral blood. Melphalan (dose 140 or 200 mg/m2 administered in single dose or divided in two days) was used as conditioning regimen and all patients received ciprofloxacin, acyclovir and fluconazole as prophylaxis.

We collected data from the patients’ medical records in our institution from diagnosis until five years after HSCT. We collected demographic variables (age, gender, race, body mass index (BMI), comorbidity variables (diabetes mellitus, hypertension, CKD, arrythmia, valvular heart disease, ischemic heart disease, cerebrovascular disease, chronic liver disease, intestinal inflammatory disease, peptic ulcer, connective tissue disease, chronic obstructive pulmonary disease, solid-organ cancer, psychiatric disease), multiple myeloma-related variables (subtype, light chain domain, concomitant amyloidosis, International Staging System classification, cytogenetic abnormalities, bone marrow plasma cells percentage, M-protein level, Serum B2-microglobulin level, number of previous lines of therapy, number of chemotherapy cycles, radiotherapy in the past), HSCT-related variables (blood results at hospital admission for HSCT, graft source, period of aplasia, sepsis, fever, shock, nephrotoxic drugs, mucositis, sinusoidal obstructive syndrome, thrombotic microangiopathy, tumor lysis syndrome, AKI, AKI stage).

Baseline eGFR was estimated according to the CKD-EPI equation [14] considering serum creatinine at the last medical appointment before HSCT.

Acute kidney injury diagnosis was made based on daily values of serum creatinine and 6-h urinary output from hospital admission for HSCT until hospital discharge, and on all hospital admissions and weekly evaluations in the outpatient clinic in the first 100 days after HSCT. Acute kidney injury was defined by KDIGO criteria [6] (any of the following: increase in serum creatinine by ≥ 0.3 mg/dl (≥ 26.5 µmol/l) within 48 h; or increase in serum creatinine to ≥ 1.5 times baseline, which is known or presumed to have occurred within the prior 7 days; or urinary output < 0.5 ml/kg/h for 6 h). Stage of AKI followed KDIGO classification considering the worst serum creatinine value and/or longest period of urinary volume reduction. Moderate to severe AKI was defined as AKI stage 2 and AKI stage 3.

Chronic kidney disease stage 3 or higher was defined according to the KDIGO definition [15] as a persistent decrease in eGFR to below 60 mL/min/1.73 m2.

The diagnosis of multiple myeloma was made according the International Multiple Myeloma Working Group (IMWG) Criteria [16]. Multiple myeloma staging was made using International Staging System for Myeloma (ISS) [17]. The associated‐amyloid light chain amyloidosis was diagnosed by demonstration of Congo red staining (kidney or subcutaneous fat biopsy). The hematopoietic cell transplantation—specific comorbidity index (HCT-CI) was calculated according to the validated version [18] considering the patient's comorbidities.

Obesity was defined as a BMI above 30 kg/m2.

Fever was defined as a measured body temperature of 38 °C (100.4 °F) or greater. Sepsis was diagnosed when patients presented with temperature ≥ 38 °C or < 36 °C, a white blood cell count > 10,000/mm3 or < 4000/mm3, and a positive blood culture for bacteria [19]. Shock was considered when patients presented with cardiac frequency > 90 bpm, systolic blood pressure < 90 mmHg and at least one lactate determination > 2 mmol/L or 22 mg/dL. Oral mucositis was graded according to the World Health Organization’s (WHO’s) Oral Toxicity Scale [20].

Nephrotoxic drugs considered in this report included gentamicin, amikacin, vancomycin, amphotericin B, and foscarnet.

The ethical committee of Centro Académico de Medicina de Lisboa approved this study (approval number 35337) according to institutional guidelines and waived the requirement to obtain informed consent due to its retrospective and non-interventional nature.

We present categorical variables as frequencies and continuous variables as median and interquartile range (P25 = 25th percentile; P75 = 75th percentile).

Cumulative incidence of AKI, univariable and multivariable analyses of factors predicting AKI were calculated using the Fine and Gray method [21] as survival analysis method considering competing events, and death was considered a competing risk event. The final multivariable model was created using backward stepwise regression.

In order to evaluate the prognostic impact of AKI, we considered type 1 right censoring for a period of five years after HSCT. The impact of AKI on relapse-free survival was calculated using the Fine and Gray method [21] and death was considered a competing risk event. The final multivariable model was created using backward stepwise regression.

The impact of AKI on overall survival was calculated using a Cox regression model including demographic variables and other variables with level of significance α < 0.2 in univariable analysis to create the final multivariable model. The Cox proportional hazards assumption was checked using formal statistical tests and graphical diagnostics based on the scaled Schoenfeld residuals.

The incidence of CKD and eGFR reduction > 25% was calculated at the end of the first year and at the end of 5 years, and the association with AKI in the first 100 days post-HSCT was evaluated.

Our approach followed the European Group for Blood and Marrow Transplantation guidelines on statistical methodology [22]. A level of significance α = 0.05 was considered for statistical significance. The statistical software package STATA 16.0 for Windows and R software (R Core Team (2017). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. URL https://www.R-project.org/.) were used for the data analysis.