A preliminary study to investigate the serum urea: creatinine ratio in canine babesiosis in South Africa

Abstract

In order to investigate the increased serum urea: creatinine ratio encountered in canine babesiosis in South Africa, packed cell volume (PCV), serum haemoglobin, serum total bilirubin, plasma ammonia, serum urea, serum creatinine, calculated serum urea: creatinine ratio, serum cystatin-C, total serum protein (TSP) and urine analysis were performed in this study. The model consisted of 28 canine babesiosis patients, 25 in group 1, the anaemic group with PCV < 20%, and 3 in group 3, the haemoconcentrated group with PCV > 50%. 13 canine babesiosis negative dogs made up group 2, the control group. Unpublished human malaria data from 363 patients was also analysed and the urea: creatinine ratio was calculated in order to determine whether elevated serum urea: creatinine ratios also occur in this disease. The serum urea: creatinine ratio was > 20 in 23/25 anaemic patients, 3/13 control patients and 2/3 haemoconcentrated patients. The mean and median serum urea: creatinine ratio for the anaemic babesia group was 46.32 and 41.36, for the haemoconcentrated babesia group 25.10 and 20.85, for the malaria patients 20.64 and 18.75, and for the control group 15.91 and 15.36 respectively. Mean and median serum urea: creatinine ratio was significantly elevated and the distribution varied significantly between the anaemic babesiosis group compared with the control group (p=0.000003). Serum urea: creatinine ratio was elevated in canine babesiosis patients, especially when they are anaemic. The serum urea: creatinine ratio was also found to be elevated in the human malaria patients. Serum urea was found to be elevated in 17/25 anaemic patients, 0/13 control patients and 3/3 haemoconcentrated patients. Mean and median values for serum urea were 12.83 mmol/l and 11.8 mmol/l and 5.09 mmol/l and 4.3 mmol/l for the anaemic group and the control group respectively. Mean and median serum urea was significantly elevated and the distribution varied significantly between the anaemic group compared with the control group (p=0.00003). Serum creatinine was elevated in 0/25 anaemic patients 0/13 control patients and 2/3 haemoconcentrated patients. Mean and median values for serum creatinine were 72.08 <font face="symbol">m</font>mol/l and 67 <font face="symbol">m</font>mol/l, and 78.69 <font face="symbol">m</font>mol/l and 75 <font face="symbol">m</font>mol/l for the anaemic group and the control group respectively. Mean, median and distribution of serum creatinine did not vary significantly between the control group and the anaemic babesia group (p=0.274488). Serum cystatin-C was elevated in 0/25 anaemic patients, 0/13 control patients and 2/3 haemoconcentrated patients. The two haemoconcentrated patients with elevated serum urea and serum creatinine also had elevated serum cystatin-C. One of these two patients was available for post mortem and had significant renal lesions on histopathology. These two patients also had a urine specific gravity (SG) < 1.030 (1.020 for both cases). Mean and median values for serum cystatin-C were 0.71 mg/l and 0.7 mg/l, and 0.67 mg/l and 0.7 mg/l for the anaemic group and the control group respectively. Mean, median and distribution of serum cystatin-C did not vary significantly between the control group and the anaemic babesia group, (p=0.450172). Using the Spearman-rank test for correlation coefficients serum creatinine but not serum urea had a significant positive correlation with serum cystatin-C; and serum urea had a significant positive correlation with the serum urea: creatinine ratio. Log serum creatinine was significantly correlated to log serum cystatin-C, (r2=0.52, p<0.001). Log serum urea was also significantly correlated to log serum cystatin-C, (r2=0.14, p=0.024), however, this correlation was relatively poor compared to that of serum creatinine with serum cystatin-C. Log serum urea: creatinine ratio was not significantly correlated with log serum cystatin-C, (r2=0.014, p=0.485). Given its resistance to interference from the elevated plasma and serum pigments found in canine babesiosis, an elevated serum cystatin-C was taken to indicate significant renal disease. Therefore we can deduce that the increased serum urea: creatinine ratio encountered in canine babesiosis is a result of disproportionately elevated serum urea concentrations, which is of non-renal origin. A similar event appears to be present in human falciparum malaria patients. The cause of this phenomenon in either disease is unknown; but various forms of prerenal azotaemia, hyperureagenesis and laboratory perturbations may play a role. Serum creatinine was a better measure of significant renal disease than serum urea in this study. There also appeared to be no added benefit of measuring serum cystatin-C. Mean and median values for PCV were 11.64% and 11%, and 46.08% and 48% for the anaemic group and the control group respectively. As expected the mean and median PCV was significantly lower and the distribution varied significantly between the control group and the anaemic babesia group (p=0.000001). The mean and median values for PCV in the haemoconcentrated group were 54.33% and 54% respectively. Serum haemoglobin was > 1.6 g/l in 8/25 anaemic patients, 0/13 control patients and 3/3 haemoconcentrated patients. Mean and median values for serum haemoglobin were 1.72 g/l and 1.3 g/l, and 0.88 g/l and 0.8 g/l for the anaemic group and the control group respectively. Mean and median serum haemoglobin was significantly elevated and distribution of serum haemoglobin varied significantly between the control group and the anaemic babesia group (p=0.002364). Serum total bilirubin was > 15 <font face="symbol">m</font>mol/l in 11/25 anaemic patients, 0/3 haemoconcentrated patients and 0/13 control patients. Mean and median values for serum total bilirubin were 22.42 <font face="symbol">m</font>mol/l and 10.6 <font face="symbol">m</font>mol/l, and 10.63 <font face="symbol">m</font>mol/l and 11.0 <font face="symbol">m</font>mol/l for the anaemic group and the control group respectively. The median and the distribution of serum total bilirubin did not vary significantly between the control group and the anaemic babesia group (p=0.355888). TSP was elevated in 5/25 anaemic patients, was below the normal laboratory reference range in 5/25 anaemic patients, was elevated in 1/13 control patients, and was normal in all the haemoconcentrated cases. Mean and median values for TSP were 58.60 g/l and 54 g/l, and 60.92 g/l and 60.0 g/l for the anaemic group and the control group respectively. The mean, median and distribution of TSP did not vary significantly between the control group and the anaemic babesia group (p=0.130661).Plasma ammonia was elevated in 16/25 anaemic patients, 0/13 control patients and 2/3 haemoconcentrated patients. Mean and median values for plasma ammonia were 78.77 <font face="symbol">m</font>mol/l and 53.8 mol/l, and 23.18 <font face="symbol">m</font>mol/l and 21.5 <font face="symbol">m</font>mol/l for the anaemic group and the control group respectively. Mean and median plasma ammonia was significantly elevated and the distribution of plasma ammonia varied significantly between the control group and the anaemic babesia group (p=0.009562). Using the Spearman-rank test for correlation coefficients PCV had a significant negative correlation with plasma ammonia, serum urea and serum urea: creatinine ratio. Serum haemoglobin and serum bilirubin had a significant positive correlation with plasma ammonia, serum urea and serum urea: creatinine ratio. Plasma ammonia also had a significant positive correlation with serum urea and serum urea: creatinine ratio. These correlations may be explained by substrate loading as a result of haemolysis. I speculate that various protein by-products of haemolysis may cause the hyperammonaemia encountered via deamination of these proteins. This in turn may lead to ureagenesis, the physiological process of ammonia clearance. This ureagenesis may be excessive and therefore may account for the elevated serum urea encountered in patients with normal serum creatinine and serum cystatin-C concentrations. Alternatively the presence of an elevated serum haemoglobin concentration with its high intrinsic absorptive capacity in the 300 nm to 500 nm wave length range, may positively bias the measurement of serum urea and plasma ammonia concentrations. Serum urea is measured at 340 nm and plasma ammonia is measured at 365 nm. Thus a non-physiological process could also explain the above correlations. Urine analysis of the canine babesiosis patients showed mild evidence of renal disease as portrayed by proteinuria, renal tubular epithelium (RTE) celluria and granular casts. However, these findings, as well as the presence of an elevated serum urea, haemoglobinaemia and haemoglobinuria, bilirubinaemia and bilirubinuria, were present in both patients with mild renal disease and patients with overt renal disease. Therefore the benefit of single serum urea concentrations and random urine sediment analysis in canine babesiosis is questioned. Serial urine analysis monitoring was not investigated. This study concluded that serum urea is often elevated due to non-renal factors in canine babesiosis patients. This causes an elevated serum urea: creatinine ratio in this disease, with a similar event appearing to occur in human malaria patients. The cause of these elevated ratios remains undetermined, but is likely to be as a result of hyperureagenesis or laboratory error. The measurement of serum creatinine, urine SG and hourly urine production is advocated to detect significant renal disease in these patients. There appeared to be no added benefit derived from the measurement of serum cystatin-C concentrations in canine babesiosis patients in this preliminary study.