br Outcomes br The primary outcome measure was the
The primary outcome measure was the development of AKI. Kaplan-Meier cumulative incidence curves of AKI were plotted at 2 weeks, 3 months, and 1 year after diagnosis by each cancer group. The Kaplan-Meier cumulative incidence curve was also plotted for stage 3 AKI. The maximum stage and number of episodes of AKI were assessed for each cancer group. In the comparison analysis of patients who Chloramphenicol developed AKI and those who did not, the correlation between AKI and baseline patient characteristics, use of CT scans, and treatments were assessed. The application of RRT was described separately.
The secondary outcome measure was the development of renal impairment or proteinuria in childhood cancer survivors. For the risk factor analysis, patient characteristics,
May 2019 ORIGINAL ARTICLES
time interval from cancer diagnosis to first onset of AKI, number of AKI episodes, maximum stage of AKI, and use of RRT were assessed.
Continuous variables were presented as a median with IQR regarding their normality, and categorical variables were presented as counts and percentages. Groups were compared with a t test, c2 test, or Mann-Whitney U test, as appropriate. All variables with a P value of <.25 in the bivariate analysis were included in the multivariable analysis. A P value of <.05 was considered statistically significant. Statistical analysis was done by using R-project version 3.3.2 (R core team, Vienna, Austria), and cumulative incidence of AKI was calculated with the package survival.
A total of 2170 children were diagnosed with cancer during the study period, and 1877 of them had their serum Cr levels measured at least twice in the first year after cancer diagnosis. Of these patients, 9 patients with CKD stage V were excluded. A total of 1868 patients (1059 male and 809 female) were eligible for analysis (Figure 1; available at www.jpeds.com). These patients were diagnosed with cancer at a median age of 7.9 years old. Brain tumor, acute leukemia (ALL and AML), and lymphoma were the most common cancer types in our study population (Table I; available at www.jpeds. com). The median initial eGFR of the eligible patients was 90.0 mL/minute/1.73 m2 (IQR 74.6-110.4). It is notable that 145 patients (7.8%) had an initial eGFR below 60 mL/minute/1.73 m2. These patients were younger (median age of 1.7 years old), but distribution of their cancer types were not significantly different from that of those without low initial eGFR.
Incidence of Acute Kidney Injury
Risk Factors of Acute Kidney Injury
In the bivariate analysis, a greater proportion of patients who experienced AKI had hematologic malignancies (ALL, AML, and lymphoma), lower eGFR at diagnosis, experienced TLS, had undergone a greater number of CT scans, had nephrotoxic chemotherapeutic agents administered (especially cyclophosphamide, ifosfamide, and methotrexate), had received HSCT, and had experienced cancer relapse. In the current study, nephrectomy and the use of carboplatin and cisplatin did not increase the risk of AKI. In the multivariable analysis, statistically significant risk factors of AKI were cancer group (especially ALL and AML), lower eGFR at diagnosis, occurrence of TLS, use of methotrexate, administration of HSCT, and greater number of CT scans performed (Table V).
Long-Term Renal Outcome
Acute Kidney Injury in Pediatric Cancer Patients 245 r> THE JOURNAL OF PEDIATRICS www.jpeds.com Volume 208
Figure 2. A, Kaplan-Meier cumulative incidence curve of AKI according to cancer group; B, 2-week, 3-month, and 1-year cumulative incidence of AKI according to cancer group, with 95% CI; C, Kaplan-Meier cumulative incidence curve of stage 3 AKI, according to cancer group. WT, Wilms tumor.