Comparing Interleukin-6, C-Reactive Protein, Erythrocyte Sedimentation Rate, and White Blood Cell Count for the Early Diagnosis of Acute Pyelonephritis in Children

Background : Urinary tract infection (UTI) is common in children. Differentiating lower UTI from acute pyelonephritis (APN) is of great importance in children since early diagnosis and timely management can prevent kidney scarring and associated complications. Objectives : We aimed to compare the diagnostic performance of serum C-reactive protein (CRP), erythrocyte sedimentation rate (ESR), and white blood cell (WBC) count with interleukin-6 (IL-6) for the diagnosis of APN in children. Methods: This cross-sectional study included 38 children with APN (based on clinical findings and positive urine culture) aged 1 month to 15 years admitted to Bandar Abbas Pediatric Hospital, Bandar Abbas, Iran, during 2019-2020. Serum CRP, WBC, IL-6, and WBC count were measured in all children. Age and sex of the participants were also recorded. Results: The mean ± SD age of the children was 65.82 ± 46.67 months, of whom 23 (60.5%) were girls. The sensitivity of WBC, ESR, CRP, and IL-6 for the early diagnosis of APN was 39.5%, 71.1%, 60.5%, and 71.1%, respectively. Taking the results of ESR, CRP, and WBC into account as the basis of diagnosis, altogether 9.98 pg/mL was the best cut-off of serum IL-6 for the diagnosis of APN with 70%-80% sensitivity and 56.5%-73.7% specificity. A significant correlation was found between serum IL-6 and CRP, ESR, and WBC with the strongest correlation between IL-6 and CRP (r = 0.451, P = 0.004). Age and sex did not affect these correlations. The sensitivity of serum IL-6 at the 5 pg/mL cut-off for the diagnosis of APN was higher in boys and children younger than 6 years. Conclusion: IL-6 can diagnose APN in children aged 1 month to 15 years with 71.1% sensitivity. Its sensitivity is superior to CRP and WBC and comparable with ESR in this regard.


Objectives
We aimed to compare IL-6, CRP, ESR, and WBC count for the early diagnosis of APN in children.

Participants
This cross-sectional study included children with APN (based on clinical findings and positive urine culture) aged 1 month to 15 years admitted to Bandar Abbas Pediatric Hospital, Bandar Abbas, Iran in from March 21, 2019 to March 20, 2020. Patients who had received any antibiotics within one week prior to admission, as well as those with any comorbidities, confirmed vesicourethral reflux, or renal dysfunction were excluded from the study. Sample size was calculated as at least 38 participants based on the correlation coefficient in a previous study (α = 0.05, β = 0.1) (1), using MedCalc statistical software, version 14.0.

Study Design
After obtaining written informed consent from the parents or guardians of the children, their age and sex were recorded. A complete medical history was taken from the patients and they were all thoroughly examined by an expert pediatric nephrologist.
Body temperature was measured using a standard tympanic thermometer and fever was defined as >38°C in children < 3 years, >37.8°C in children 3-11 years, and >37.6°C in children >11 years. Accordingly, all children in the study were febrile. Urinalysis and urine culture were performed for all participants and those with a positive urine culture were included.
Random venous blood samples were collected from all patients before the initiation of any treatments. WBC count was determined using an automated blood cell counter. ESR was measured using an ESR reader. CRP was measured quantitatively using Pars Azmoon kits (Pars Azmoon Inc., Tehran, Iran) and a biochemistry autoanalyzer (Technicon RA-1000, Technicon Instruments Corp., TerryTown, NY, USA) with the immunoturbidity method. Serum IL-6 was measured using the Roche kit (Roche Diagnostics, Germany) and the electrochemiluminescence immunoassay (ECLIA) method (by Roche's Elecsys® cobas e 411 analyzer for immunoassay tests). ESR >10 mm/h, CRP >10 mg/L, IL-6 >5 pg/mL, and WBC count >10 000/µL were regarded as elevated.

Data analysis
We used the Statistical Package for the Social Sciences (SPSS) software (version 25.0, Armonk, NY: IBM Corp.) for data analysis. For the description of quantitative variables, means and standard deviations were used and for qualitative variables, frequencies and percentages were used. Based on the results of Kolmogorov-Smirnov normality test for age, Spearman's correlation test was used to evaluate the association between IL-6 and other inflammatory markers. By taking the results of ESR, CRP, and WBC count as reference, sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and diagnostic accuracy (DA) of IL-6 >5 pg/mL was calculated. Also, receiver operating characteristic (ROC) curves of serum IL-6 were drawn for the diagnosis of APN. The area under the curve (AUC) was calculated for every curve and the optimal IL-6 cut-off was determined. P ≤ 0.05 was regarded as statistically significant.

Results
From the 38 children included in this study with a mean ± SD age of 65.82 ± 46.67 months, 23 (60.5%) were girls and 15 (39.5%) were boys. General characteristics and laboratory finding of the participants are demonstrated in Table 1. Elevated ESR, CRP, WBC count, and serum IL-6 were observed in 71.1%, 60.5%, 39.5%, and 71.1% of the children, respectively. Since APN was diagnosed in all the children, these figures are also reflective of the sensitivity of each marker for the diagnosis of APN.
Serum IL-6 was significantly correlated with CRP (P = 0.004), ESR (P = 0.018), and WBC count (P = 0.043); however, when controlled for age and sex, the correlations were no longer statistically significant ( Table 2). Elevated serum IL-6 (>5 pg/mL) was found in 11 boys and 16 girls as well as 16 patients < 6 years of age and 11 aged 6-15 years. Since all patients in the current study had APN, the sensitivity of serum IL-6 was 73.3% in boys, 69.6% in girls, 84.2% in children aged < 6 years, and 57.9% in children aged 6-15 years ( Table 4).
The ROC curve of IL-6 for the diagnosis of APN is shown in Figure 1, taking CRP >10 mg/L as reference. Consistently, the AUC was calculated as 0.772 with an optimal cut-off of 9.98 pg/mL for serum IL-6, having 78% sensitivity and 73.3% specificity. Figure 2 shows the ROC curve of IL-6 for the diagnosis of APN, taking ESR>10 mm/h as reference. Appropriately, the AUC was calculated as 0.702 with an optimal cut-off of 9.98 pg/ mL for serum IL-6, having 70% sensitivity and 72.7% specificity. Figure 3 demonstrated the ROC curve of IL-6 for the diagnosis of APN, taking WBC count >10 000 /µL as reference. The AUC was 0.623 with an optimal cut-off of 9.98 pg/mL for serum IL-6, having 80% sensitivity and 56.5% specificity.

Discussion
In the current study we found 71.1% sensitivity for serum IL-6 for the diagnosis of APN in children aged 1 month to 15 years. We also found a significant correlation between serum IL-6 and ESR, CRP, and WBC count in children with APN. The sensitivity of serum IL-6 was comparable with ESR (71.1%) but superior to CRP (60.5%) and WBC count (39.5%) for the diagnosis of APN. On the other hand, the sensitivity of IL-6 >5 pg/mL for the diagnosis of APN was higher in boys compared to girls and in children < 6 years of age compared to 6-15 years of age.
Multiple studies have assessed serum and urine biomarkers for the differentiation of upper from lower UTI with contradictory results. In the most recent systematic review by Shaikh and colleagues, performed on 25 studies, the average sensitivity of CRP at the 20 mg/L cut-off and ESR at the 30 mm/h cut-off for this purpose were 93% and 83%, respectively. They concluded that ESR is not sufficient for distinguishing between cystitis and APN in children, while CRP < 20 mg/L decreased the possibility of APN to lower than 20% (1). The foremost reason for the discrepancy between their study and ours is that Shaikh and colleagues used the results of dimercaptosuccinic acid (DMSA) scanning as reference (1). Moreover, the larger sample size, the sensitivity of diagnostic kits and measurement tools, as well as demographic characteristics of the participants might have contributed to this discrepancy. Also, they evaluated CRP and ESR at different cut-offs compared with our study.
Contrary to our findings, one study showed that IL-6 is not an appropriate marker for the diagnosis of UTI and the differentiation of upper from lower UTI. They explained the wide range of IL-6 values and its inappropriateness by the variety of APN symptoms onset, sex distribution, and the virulence of the responsible organisms (12). While serum IL-6, especially at the 10 pg/mL cut-off, had a relatively suitable sensitivity for the diagnosis of APN in our study. However, our findings have to be confirmed in larger studies. In another study, Sheu et al found that Abbreviations: CC, correlation coefficient; ESR, erythrocyte sedimentation rate; CRP, C-reactive protein; WBC, white blood cell; IL-6, interleukin-6. *Analyzed by Spearman's correlation test. Abbreviations: CC, correlation coefficient; ESR, erythrocyte sedimentation rate; CRP, C-reactive protein; WBC, white blood cell; IL-6, interleukin-6. in smaller children with APN and high fever, an acute elevation in serum and urine IL-6 increases the risk of renal scarring (13). This is quite consistent with our findings, as we also found a higher sensitivity for serum IL-6 in children < 6 years of age compared to those age 6-15 years. The results of an older study showed the utility of serum and urine IL-6 for the early diagnosis of APN in febrile children (7).
In other studies on the evaluation of biomarkers to differentiate upper from lower UTI, procalcitonin has been proposed as a useful inflammatory marker. For instance, one study showed a higher sensitivity and specificity for procalcitonin, compared to CRP, for the early diagnosis of APN. Additionally, they showed a significant correlation between procalcitonin values and the degree of renal parenchymal involvement (14). This shows that other biomarkers such as procalcitonin should be investigated along with other markers evaluated in our study, for the simultaneous comparison of their utility in the early diagnosis of APN. Furthermore, serum CRP is an effective and valuable marker for the diagnosis of APN which can help prevent APN-associated complications (9). Other researchers found that CRP, ESR, and WBC count are not sensitive or specific enough for the diagnosis of UTI and determination of the infection location and that where advanced laboratory facilities are available, other laboratory tests should also be performed along with CRP, ESR and WBC count (15). Their findings question what we did to ascertain the diagnostic value of serum IL-6. We took the results of CRP, ESR, and WBC count as reference. A major limitation of the current study was that because of limited resources, patients did not undergo DMSA scan, which is the gold standard for the diagnosis of renal parenchymal involvement, indicative of APN. Therefore, we calculated diagnostic indices based on ESR, CRP, and WBC count results, which may not have been as sensitive or specific as DMSA scanning for the diagnosis of APN.

Conclusion
In the current study, we found that the sensitivity of WBC count, ESR, CRP, and serum IL-6 was 39.5%, 71.1%, 60.5%, and 71.1% for the diagnosis of APN in children. We also showed that by taking the results of ESR, CRP, and WBC count as reference, serum IL-6 levels higher