Severe Anemia As an Early Predictor of Kidney Injury

Adult patients with Sickle Cell Disease (SCD) develop overt proteinuria; microalbuminuria precedes overt proteinuria in Pediatrics. Prior Pediatric cross-sectional studies have evaluated the prevalence and associations of laboratory variables with the presence of microalbuminuria (MiA). This study utilized a prospective SCD nephropathy cohort study to understand the incidence, risk factors, and predictors for the development of MiA. Then, we tested the hypothesis that severe anemia as an infant or toddler, independent of future blood counts, clinical course, or therapies, would be predictive of developing microalbuminuria during adolescence.

Methods: We are conducting an IRB approved Pediatric SCD Nephropathy cohort of patients with HbSS or SB0 thalassemia (n=152). We record every standard of care urine albumin/creatinine level obtained since birth (n=595) for each participant in our cohort along with their blood pressure, SCD therapy, complete blood count, and estimated glomerular filtration (cystatin C) from that visit. Patients were classified as having microalbuminuria if their urine albumin/creatinine was ≥ 30 mg/g using a Siemens DCA Vantage Analyzer. T-tests and chi-squared tests were used to compare individual demographic and clinical characteristics for continuous and categorical, respectively, of those patients who developed at least one episode of MiA using SAS 9.4. Next, we used a combination of left, right and interval censored data in a time to development of MiA model. To examine the incidence of MiA and its predictors, we utilized univariate and multivariate parametric survival regression models assuming a logistic distribution. To examine the predictive ability of a single CBC early in life for the development of MiA, we utilized logistic regression and computed the area under the receiving operating characteristics curves (AUC). Based on Baby HUG data, we used a hemoglobin <8.0 g/dL on their earliest Hb after 9 months and prior to their 2^nd birthday for our classification of severe anemia.

Results: Among 152 participants in our prospective cohort, 48 (32%) developed at least one episode of MiA. The mean age of the 48 patients that developed MiA was 13.2 years (s.d 4.3) as compared to the mean age of 13.9 years (s.d 4.4) among 104 patients that have not yet developed MiA. Participants with lower hemoglobin (p<0.0001), higher reticulocyte percent (p=0.003), higher white blood cell (WBC) count (<0.0001), higher absolute neutrophil count (p=0.003) and higher platelet counts (p=0.03) were significantly more likely to develop MiA at an earlier age. Nineteen of 75 patients (25%) were on HU at the time of their first identified episode of MiA, 20 of 53 (37%) were on transfusion and 10 of 24 (42%) were on no SCD modifying therapy. No statistical difference in time to first MiA was identified based on SCD modifying therapy (p=0.08), abnormal systolic blood pressure (0.6), diastolic blood pressure (0.3), eGFR (0.6), or bilirubin (p=0.08). In multivariate analysis, patients that were more anemic (p=0.002) and had a higher WBC (p=0.01) were more likely to develop MiA at an earlier age. To determine if an early CBC could predict future development MiA, we evaluated each patient's CBC by two years of age. Severe anemia (Hb<8 g/dL)(p<0.0001), but not WBC (p=0.2) or platelet count (p=0.2), was identified as a risk factor for progressing to MiA in our cohort. Thirty one of 50 participants (62%) with Hb<8g/dL at two years of age eventually developed MiA in our cohort. Seventeen out of 98 participants (17%) with Hb≥8g/dL progressed to MiA. The AUC for a model of severe anemia alone as a predictor of MiA was 0.7 which was statistically similar to the model for development of MiA using severe anemia, WBC, and platelets (AUC 0.75).

Conclusion: This cohort supports the finding that at the time adolescent SCD patients are identified with microalbuminuria, they have a greater degree anemia and leukocytosis than adolescent patients without microalbuminuria. Of great importance, severe anemia identified prior to two years of age is predictive of patients that will develop MiA during adolescence. Additional research into the pathophysiology of kidney injury and its association with acute anemic events or hemolysis is vital to modifying the progression to CKD.