Haemoglobinuria is associated with chronic kidney disease and its progression in patients with sickle cell anaemia

Abstract

To evaluate the association between haemoglobinuria and chronic kidney disease (CKD) in sickle cell anaemia (SCA), we analysed 356 adult haemoglobin SS or Sβ(o) thalassaemia patients from the University of Illinois at Chicago (UIC) and 439 from the multi-centre Walk-Treatment of Pulmonary Hypertension and Sickle Cell Disease with Sildenafil Therapy (Walk-PHaSST) cohort. CKD was classified according to National Kidney Foundation Kidney Disease Outcomes Quality Initiatives guidelines. Haemoglobinuria, defined as positive haem on urine dipstick with absent red blood cells on microscopy, was confirmed by enzyme-linked immunosorbent assay in a subset of patients. The prevalence of CKD was 58% in the UIC cohort and 54% in the Walk-PHaSST cohort, and haemoglobinuria was observed in 36% and 20% of the patients, respectively. Pathway analysis in both cohorts indicated an independent association of lactate dehydrogenase with haemoglobinuria and, in turn, independent associations of haemoglobinuria and age with CKD (P < 0·0001). After a median of 32 months of follow-up in the UIC cohort, haemoglobinuria was associated with progression of CKD [halving of estimated glomerular filtration rate or requirement for dialysis; Hazard ratio (HR) 13·9, 95% confidence interval (CI) 1·7-113·2, P = 0·0012] and increasing albuminuria (HR 3·1, 95% CI: 1·3-7·7; logrank P = 0·0035). In conclusion haemoglobinuria is common in SCA and is associated with CKD, consistent with a role for intravascular haemolysis in the pathogenesis of renal dysfunction in SCA.

Keywords: albuminuria; haemoglobinuria; haemolysis; kidney disease; sickle cell.

Figure 1

Validation of haemoglobinuria defined by dipstick urinalysis. Patients with haemoglobinuria on dipstick urinalysis (n=17) had significantly higher concentrations of urine haemoglobin directly measured by enzyme-linked immunosorbent assay than those without haemoglobinuria (n=26) (23.1 vs. 11.5ng/ml, P<0.0001).

Figure 2

Figure 2A-D: Haemoglobinuria association with laboratory markers of haemolysis. In the UIC cohort lower haemoglobin concentration (80 vs. 90 g/l, P<0.0001), higher lactate dehydrogenase (LDH, 426 vs. 279 u/l, P<0.0001), higher aspartate transaminase (AST, 49 vs. 37 u/l, P<0.0001), and higher reticulocyte percentages (16.1% vs. 11.7%, P<0.0001) were observed in patients with haemoglobinuria. Similarly, in the Walk-PHaSST cohort, lower haemoglobin concentration (81 vs. 89 g/l, P<0.0001), higher LDH (540 vs. 428 u/l, P<0.0001), higher AST (49 vs. 42 u/l, P=0.0052), and a trend for higher reticulocyte percentage (9.6% vs. 8.5%, P=0.07) were observed in patients with haemoglobinuria. UIC = University of Illinois at Chicago; Walk-PHaSST = Walk-Treatment of Pulmonary Hypertension and Sickle Cell Disease with Sildenafil Therapy. Figure 2E: Haemoglobinuria association with increased albuminuria. In both the UIC and Walk-PHaSST cohorts, the presence of haemoglobinuria was associated with higher levels of albuminuria (UIC: 248 vs. 20 mg/g creatinine, P<0.0001; Walk-PHaSST: 109 vs. 18 mg/g creatinine, P<0.0001). UIC = University of Illinois at Chicago; Walk-PHaSST = Walk-Treatment of Pulmonary Hypertension and Sickle Cell Disease with Sildenafil Therapy. Figure 2F: Haemoglobinuria association with decreased GFR. In both the UIC and Walk-PHaSST cohorts, with adjustment for age the presence of haemoglobinuria was associated with decreased eGFR (UIC: 120 vs. 131 ml/min/1.73 m², P=0.0014; Walk-PHaSST: 124 vs. 131 ml/min/1.73 m², P=0.035). UIC = University of Illinois at Chicago; Walk-PHaSST = Walk-Treatment of Pulmonary Hypertension and Sickle Cell Disease with Sildenafil Therapy.

Figure 2

Figure 2A-D: Haemoglobinuria association with laboratory markers of haemolysis. In the UIC cohort lower haemoglobin concentration (80 vs. 90 g/l, P<0.0001), higher lactate dehydrogenase (LDH, 426 vs. 279 u/l, P<0.0001), higher aspartate transaminase (AST, 49 vs. 37 u/l, P<0.0001), and higher reticulocyte percentages (16.1% vs. 11.7%, P<0.0001) were observed in patients with haemoglobinuria. Similarly, in the Walk-PHaSST cohort, lower haemoglobin concentration (81 vs. 89 g/l, P<0.0001), higher LDH (540 vs. 428 u/l, P<0.0001), higher AST (49 vs. 42 u/l, P=0.0052), and a trend for higher reticulocyte percentage (9.6% vs. 8.5%, P=0.07) were observed in patients with haemoglobinuria. UIC = University of Illinois at Chicago; Walk-PHaSST = Walk-Treatment of Pulmonary Hypertension and Sickle Cell Disease with Sildenafil Therapy. Figure 2E: Haemoglobinuria association with increased albuminuria. In both the UIC and Walk-PHaSST cohorts, the presence of haemoglobinuria was associated with higher levels of albuminuria (UIC: 248 vs. 20 mg/g creatinine, P<0.0001; Walk-PHaSST: 109 vs. 18 mg/g creatinine, P<0.0001). UIC = University of Illinois at Chicago; Walk-PHaSST = Walk-Treatment of Pulmonary Hypertension and Sickle Cell Disease with Sildenafil Therapy. Figure 2F: Haemoglobinuria association with decreased GFR. In both the UIC and Walk-PHaSST cohorts, with adjustment for age the presence of haemoglobinuria was associated with decreased eGFR (UIC: 120 vs. 131 ml/min/1.73 m², P=0.0014; Walk-PHaSST: 124 vs. 131 ml/min/1.73 m², P=0.035). UIC = University of Illinois at Chicago; Walk-PHaSST = Walk-Treatment of Pulmonary Hypertension and Sickle Cell Disease with Sildenafil Therapy.

Figure 2

Figure 2A-D: Haemoglobinuria association with laboratory markers of haemolysis. In the UIC cohort lower haemoglobin concentration (80 vs. 90 g/l, P<0.0001), higher lactate dehydrogenase (LDH, 426 vs. 279 u/l, P<0.0001), higher aspartate transaminase (AST, 49 vs. 37 u/l, P<0.0001), and higher reticulocyte percentages (16.1% vs. 11.7%, P<0.0001) were observed in patients with haemoglobinuria. Similarly, in the Walk-PHaSST cohort, lower haemoglobin concentration (81 vs. 89 g/l, P<0.0001), higher LDH (540 vs. 428 u/l, P<0.0001), higher AST (49 vs. 42 u/l, P=0.0052), and a trend for higher reticulocyte percentage (9.6% vs. 8.5%, P=0.07) were observed in patients with haemoglobinuria. UIC = University of Illinois at Chicago; Walk-PHaSST = Walk-Treatment of Pulmonary Hypertension and Sickle Cell Disease with Sildenafil Therapy. Figure 2E: Haemoglobinuria association with increased albuminuria. In both the UIC and Walk-PHaSST cohorts, the presence of haemoglobinuria was associated with higher levels of albuminuria (UIC: 248 vs. 20 mg/g creatinine, P<0.0001; Walk-PHaSST: 109 vs. 18 mg/g creatinine, P<0.0001). UIC = University of Illinois at Chicago; Walk-PHaSST = Walk-Treatment of Pulmonary Hypertension and Sickle Cell Disease with Sildenafil Therapy. Figure 2F: Haemoglobinuria association with decreased GFR. In both the UIC and Walk-PHaSST cohorts, with adjustment for age the presence of haemoglobinuria was associated with decreased eGFR (UIC: 120 vs. 131 ml/min/1.73 m², P=0.0014; Walk-PHaSST: 124 vs. 131 ml/min/1.73 m², P=0.035). UIC = University of Illinois at Chicago; Walk-PHaSST = Walk-Treatment of Pulmonary Hypertension and Sickle Cell Disease with Sildenafil Therapy.

Figure 2

Figure 2A-D: Haemoglobinuria association with laboratory markers of haemolysis. In the UIC cohort lower haemoglobin concentration (80 vs. 90 g/l, P<0.0001), higher lactate dehydrogenase (LDH, 426 vs. 279 u/l, P<0.0001), higher aspartate transaminase (AST, 49 vs. 37 u/l, P<0.0001), and higher reticulocyte percentages (16.1% vs. 11.7%, P<0.0001) were observed in patients with haemoglobinuria. Similarly, in the Walk-PHaSST cohort, lower haemoglobin concentration (81 vs. 89 g/l, P<0.0001), higher LDH (540 vs. 428 u/l, P<0.0001), higher AST (49 vs. 42 u/l, P=0.0052), and a trend for higher reticulocyte percentage (9.6% vs. 8.5%, P=0.07) were observed in patients with haemoglobinuria. UIC = University of Illinois at Chicago; Walk-PHaSST = Walk-Treatment of Pulmonary Hypertension and Sickle Cell Disease with Sildenafil Therapy. Figure 2E: Haemoglobinuria association with increased albuminuria. In both the UIC and Walk-PHaSST cohorts, the presence of haemoglobinuria was associated with higher levels of albuminuria (UIC: 248 vs. 20 mg/g creatinine, P<0.0001; Walk-PHaSST: 109 vs. 18 mg/g creatinine, P<0.0001). UIC = University of Illinois at Chicago; Walk-PHaSST = Walk-Treatment of Pulmonary Hypertension and Sickle Cell Disease with Sildenafil Therapy. Figure 2F: Haemoglobinuria association with decreased GFR. In both the UIC and Walk-PHaSST cohorts, with adjustment for age the presence of haemoglobinuria was associated with decreased eGFR (UIC: 120 vs. 131 ml/min/1.73 m², P=0.0014; Walk-PHaSST: 124 vs. 131 ml/min/1.73 m², P=0.035). UIC = University of Illinois at Chicago; Walk-PHaSST = Walk-Treatment of Pulmonary Hypertension and Sickle Cell Disease with Sildenafil Therapy.

Figure 2

Figure 2A-D: Haemoglobinuria association with laboratory markers of haemolysis. In the UIC cohort lower haemoglobin concentration (80 vs. 90 g/l, P<0.0001), higher lactate dehydrogenase (LDH, 426 vs. 279 u/l, P<0.0001), higher aspartate transaminase (AST, 49 vs. 37 u/l, P<0.0001), and higher reticulocyte percentages (16.1% vs. 11.7%, P<0.0001) were observed in patients with haemoglobinuria. Similarly, in the Walk-PHaSST cohort, lower haemoglobin concentration (81 vs. 89 g/l, P<0.0001), higher LDH (540 vs. 428 u/l, P<0.0001), higher AST (49 vs. 42 u/l, P=0.0052), and a trend for higher reticulocyte percentage (9.6% vs. 8.5%, P=0.07) were observed in patients with haemoglobinuria. UIC = University of Illinois at Chicago; Walk-PHaSST = Walk-Treatment of Pulmonary Hypertension and Sickle Cell Disease with Sildenafil Therapy. Figure 2E: Haemoglobinuria association with increased albuminuria. In both the UIC and Walk-PHaSST cohorts, the presence of haemoglobinuria was associated with higher levels of albuminuria (UIC: 248 vs. 20 mg/g creatinine, P<0.0001; Walk-PHaSST: 109 vs. 18 mg/g creatinine, P<0.0001). UIC = University of Illinois at Chicago; Walk-PHaSST = Walk-Treatment of Pulmonary Hypertension and Sickle Cell Disease with Sildenafil Therapy. Figure 2F: Haemoglobinuria association with decreased GFR. In both the UIC and Walk-PHaSST cohorts, with adjustment for age the presence of haemoglobinuria was associated with decreased eGFR (UIC: 120 vs. 131 ml/min/1.73 m², P=0.0014; Walk-PHaSST: 124 vs. 131 ml/min/1.73 m², P=0.035). UIC = University of Illinois at Chicago; Walk-PHaSST = Walk-Treatment of Pulmonary Hypertension and Sickle Cell Disease with Sildenafil Therapy.

Figure 2

Figure 2A-D: Haemoglobinuria association with laboratory markers of haemolysis. In the UIC cohort lower haemoglobin concentration (80 vs. 90 g/l, P<0.0001), higher lactate dehydrogenase (LDH, 426 vs. 279 u/l, P<0.0001), higher aspartate transaminase (AST, 49 vs. 37 u/l, P<0.0001), and higher reticulocyte percentages (16.1% vs. 11.7%, P<0.0001) were observed in patients with haemoglobinuria. Similarly, in the Walk-PHaSST cohort, lower haemoglobin concentration (81 vs. 89 g/l, P<0.0001), higher LDH (540 vs. 428 u/l, P<0.0001), higher AST (49 vs. 42 u/l, P=0.0052), and a trend for higher reticulocyte percentage (9.6% vs. 8.5%, P=0.07) were observed in patients with haemoglobinuria. UIC = University of Illinois at Chicago; Walk-PHaSST = Walk-Treatment of Pulmonary Hypertension and Sickle Cell Disease with Sildenafil Therapy. Figure 2E: Haemoglobinuria association with increased albuminuria. In both the UIC and Walk-PHaSST cohorts, the presence of haemoglobinuria was associated with higher levels of albuminuria (UIC: 248 vs. 20 mg/g creatinine, P<0.0001; Walk-PHaSST: 109 vs. 18 mg/g creatinine, P<0.0001). UIC = University of Illinois at Chicago; Walk-PHaSST = Walk-Treatment of Pulmonary Hypertension and Sickle Cell Disease with Sildenafil Therapy. Figure 2F: Haemoglobinuria association with decreased GFR. In both the UIC and Walk-PHaSST cohorts, with adjustment for age the presence of haemoglobinuria was associated with decreased eGFR (UIC: 120 vs. 131 ml/min/1.73 m², P=0.0014; Walk-PHaSST: 124 vs. 131 ml/min/1.73 m², P=0.035). UIC = University of Illinois at Chicago; Walk-PHaSST = Walk-Treatment of Pulmonary Hypertension and Sickle Cell Disease with Sildenafil Therapy.

Figure 3

UIC and Walk-PHaSST pathway analysis for CKD stage. Standardized betas are indicated in the model. Comparative fit index=1.0; goodness-of-fit test χ²/degress of freedom≤0.8; Root Mean Square Error of Approximation of model<0.0001. According to this analysis 1) higher LDH is associated with haemoglobinuria, 2) higher age and presence of haemoglobinuria are associated with the stage of CKD. UIC = University of Illinois at Chicago; Walk-PHaSST = Walk-Treatment of Pulmonary Hypertension and Sickle Cell Disease with Sildenafil Therapy; CKD = chronic kidney disease; LDH = lactate dehydrogenase.

Figure 4

Figure 4A: Reliability of haemoglobinuria based on repeat dipstick urinalysis. Sixty-one percent (33/54) of patients with haemoglobinuria on initial assessment remained positive for haemoglobinuria on repeat urinalysis while 90% (8/85) of patients without haemoglobinuria on initial assessment remained negative on repeat uirnalysis. Figure 4B: CKD progression based on the absence or presence of haemoglobinuria on initial urine dipstick analysis. CKD progression, defined by a reduction of eGFR by 50% or requiring haemodialysis/kidney transplant, was observed in 13% (7/56) of patients with haemoglobinuria versus 1% (1/108) of patients without haemoglobinuria (HR 13.9, 95% CI: 1.7–113.2; logrank P=0.0012). CKD = chronic kidney disease. Figure 4C: Albuminuria category progression rate based on the absence or presence of haemoglobinuria on initial urine dipstick analysis. Forty-two percent (11/26) of patients with haemoglobinuria had progression of albuminuria category versus 13% (9/67) of patients without haemoglobinuria (HR 3.1, 95% CI: 1.3-7.7; logrank P=0.0035

Figure 4

Figure 4A: Reliability of haemoglobinuria based on repeat dipstick urinalysis. Sixty-one percent (33/54) of patients with haemoglobinuria on initial assessment remained positive for haemoglobinuria on repeat urinalysis while 90% (8/85) of patients without haemoglobinuria on initial assessment remained negative on repeat uirnalysis. Figure 4B: CKD progression based on the absence or presence of haemoglobinuria on initial urine dipstick analysis. CKD progression, defined by a reduction of eGFR by 50% or requiring haemodialysis/kidney transplant, was observed in 13% (7/56) of patients with haemoglobinuria versus 1% (1/108) of patients without haemoglobinuria (HR 13.9, 95% CI: 1.7–113.2; logrank P=0.0012). CKD = chronic kidney disease. Figure 4C: Albuminuria category progression rate based on the absence or presence of haemoglobinuria on initial urine dipstick analysis. Forty-two percent (11/26) of patients with haemoglobinuria had progression of albuminuria category versus 13% (9/67) of patients without haemoglobinuria (HR 3.1, 95% CI: 1.3-7.7; logrank P=0.0035

Figure 4

Figure 4A: Reliability of haemoglobinuria based on repeat dipstick urinalysis. Sixty-one percent (33/54) of patients with haemoglobinuria on initial assessment remained positive for haemoglobinuria on repeat urinalysis while 90% (8/85) of patients without haemoglobinuria on initial assessment remained negative on repeat uirnalysis. Figure 4B: CKD progression based on the absence or presence of haemoglobinuria on initial urine dipstick analysis. CKD progression, defined by a reduction of eGFR by 50% or requiring haemodialysis/kidney transplant, was observed in 13% (7/56) of patients with haemoglobinuria versus 1% (1/108) of patients without haemoglobinuria (HR 13.9, 95% CI: 1.7–113.2; logrank P=0.0012). CKD = chronic kidney disease. Figure 4C: Albuminuria category progression rate based on the absence or presence of haemoglobinuria on initial urine dipstick analysis. Forty-two percent (11/26) of patients with haemoglobinuria had progression of albuminuria category versus 13% (9/67) of patients without haemoglobinuria (HR 3.1, 95% CI: 1.3-7.7; logrank P=0.0035