Serum Cortisol Level and Monocyte HLA-DR Expression in Late Onset Neonatal Sepsis: A Case-Control Study

Data Availability

The datasets used and/or analyzed during this study available from the corresponding author on reasonable request.

Acknowledgements

The authors would like to thank the included neonates' parents and NICU crew of Minia University Hospital.

Author's contributions

Everyone who wrote the paper had a hand in coming up with the idea, carrying it out, and revising the final product. Data gathering and manuscript draught preparation were shared responsibilities among EM, NI, and MM. NI and MM interpreted the results of the laboratory work. EE analyzed the results statistically.  After reviewing the final manuscript, all authors gave their approval.

Funding

We had not received financial support for the study

Conflict of interest

We declared no conflict of interest concerning the study.

Date received: 24^th September 2024, accepted 30^th January 2025

Introduction

Subjects and Methods

Subjects

Collecting samples

Laboratory methods

Ethical consent

Data analysis using statistical methods

In this study, the Definite Sepsis group were 14 males and 16 females, they had a mean gestational age of 37±2.5 weeks and a mean postnatal age of 12.1±4.3 days, and the Probable Sepsis group were 17 males and 13 females, with a mean gestational age of 37.6±2.4 weeks and a mean postnatal age of 10.4±4.08 days. While the Control group were 16 males and 14 females, their mean gestational age was 38.4 ±1.3 weeks and mean postnatal age was 8.5±4.5 days. There were no statistically significant differences between the three groups regarding demographic data (Table 1).

In comparison between neonates with probable and definite sepsis regarding the clinical signs, there was no statistically significant difference in clinical signs, but only significant difference regarding blood culture results.

The isolated organisms were Klebsiella Pneumoniae from 14 neonates (46.6%) and Escherichia coli from 6 neonates (20%), staphylococcus pneumoniae and staphylococcus aureus each from 3 neonates, (10%) for each organism, Methicillin-resistant Staphylococcus aureus (MRSA) from 2 neonates (6.6%) and Pseudomonas aeruginosa was isolated from only one neonate (3.3%). Neonates with negative blood cultures (n = 30) were considered the probable sepsis group (clinical sepsis) (Table 2).

Regarding hemoglobin level and platelets count, there was statistically significant decrease in hemoglobin level and platelets count in the definite sepsis group when compared to the probable sepsis group and control group and statistically significant decrease in hemoglobin level and platelets count in the probable sepsis group when compared to the control group (P value= 0.0001* for all) and (P value = 0.0001* for all) respectively. While total leucocyte count (TLC), there was   statistically significant increase in total leucocytic count in the definite sepsis group when compared to the probable sepsis group and the control group and statistical significant increase in total leucocytic count in the probable sepsis group when compared to the control group (P value = 0.0001* for all) (Table 3).  

Concerning blood cells surface markers evaluated by  flow cytometry, mHLA-DR percent was  statistically significant decreased in definite sepsis group when compared to probable sepsis group and control group and statistical  significant  decrease in mHLA-DR percent in probable sepsis group when compared to control group ( P-value = 0.0001* for all) , while   statistically significant increase in cortisol level  and CRP level   in definite sepsis group when compared to probable sepsis group and control group and statistical  significant  increase in cortisol level and CRP level  in probable sepsis group when compared to control group ( P value = 0.0001* for all) and ( P-value = 0.0001* for all) respectively ( Table 4).

ROC curve analysis for detection of sepsis for comparison of definite to controls revealed that mHLADR at a cut-off value of   ≥23.5 had AUC (0.99±0.009) with highest sensitivity (96.7%) and specificity (90%), (P value = 0.0001*), while cortisol at a cut-off value of ≤25 had AUC (1±0.0) with highest both sensitivity and specificity (100%) for both, (P value = 0.0001*) (Table 5 and Figures  1 & 2).

ROC curve analysis for detection of sepsis for comparison of probable to controls revealed that mHLA-DR at a cut-off value of ≥ 42 had AUC (0.83±0.05) with highest sensitivity (80%) and specificity (73.3%%), P 0.0001*), while cortisol at a cut-off value of ≤ 22 had AUC (1±0.0) with highest sensitivity (100%) and specificity (100%), (P-0.0001*) (Table (6) and Figure (3 and 4). 

Discussion

Newborn sepsis diagnosis is still difficult. The signs of sepsis, the frequent sampling. Knowing the right biomarker to help diagnose sepsis is crucial due to the time it takes for blood culture results and the ideal timing of antibiotic treatment. Three groups of neonates  were used in this study: one group had definite sepsis based on a positive blood culture, another group had probable sepsis due to the presence of signs of sepsis but negative blood cultures, and the third group was a control group of healthy neonates. The invading  microorganisms released Cytokines that promote inflammation cause the liver to produce acute-phase proteins, such as C-reactive protein (CRP), which is an essential component of the humoral immune response to bacterial invasion [14].Our findings corroborate those of previous studies that found significantly increased CRP levels in neonates with clinically definite  sepsis; these studies served as a diagnostic tool for differentiating between healthy neonates and those with definite  or probable  sepsis [15, 16]. Both the existence and severity of sepsis can be indicated by CRP levels, which were found to be considerably greater in the definite sepsis group compared to the probable sepsis group. The size of the CRP response to sepsis depends on the underlying pathogen [14].

Sensitivity is 60% during sepsis due to the slow initial concentration increase and the need for serial measurements; furthermore, it is not specific because elevated concentrations can be observed in other conditions like tissue necrosis, surgery, recent vaccination, and meconium aspiration [7, 14].

The platelet count was substantially lower in the definite septic group compared to the probable septic group, according to the hematological parameters examined in our study. Consistent with our findings, Omran et al. (2021) discovered a significantly reduced platelet count in both the septic and non-septic groups [17]. Tosson et al., (2021) found no statistically significant change in platelet count between the septic and non-septic groups (P=0.47) [18], which contradicts our findings. Additionally, TLC was found to be considerably greater in the definite septic group compared to the probable -septic group, which is in agreement with the findings of Mubaraki et al., (2023) , who found that TLC was significantly higher in septic group [19].

Omran et al. (2021) and Tosson et al. (2021) disagree with our findings, since they failed to detect a statistically significant difference between the septic and non-septic groups [17, 18]. The most reliable way to diagnose bacterial sepsis is with a blood culture, but the results  can take a long time due to factors including the low bacteremia that is typically expected and the fact that antibiotic therapy is commonly started before the blood culture is taken. And to get the most out of the blood culture, it's best to take it out as the temperature is rising, according to hospital regulations and guidelines [20]. Thirty neonates, who did not show signs of blood culture growth, were included in this investigation. Although thirty neonates showed positive results in the blood culture, Klebsiella (46.6%), Escherichia coli (20%), streptococcus pneumoniae (10%), staphylococcus aureus (10%), methicillin-resistant Staphylococcus aureus (6.6%), and Pseudomonas aeruginosa (3.3%) were the most frequently found microbes in blood cultures. Among the gram-negative bacteria isolated, klebsiella pneumoniae accounted for nearly half (46.6%). These findings corroborated those of with Elmashad et al. (2019) [21] and Ramavath et al., (2023) [22]. Our findings are at odds with those of Tosson et al. (2021) who reported that MRSA was the most frequently isolated bacterium in their cases [18]. The majority of LOS pathogens were Coagulase-negative staphylococci (CONS), according to other researchers like Hammoud et al., 2017 [23]. There is a wide range of organism findings from one NICU to another and from one region to another, thus it is necessary for each hospital to modify its antibiotic policy accordingly. It is still challenging to diagnose sepsis at an early stage; so, new indicators for neonatal sepsis are needed to improve diagnostic sensitivity and specificity and, frequently, to track treatment progress. However, at this time, no biomarker exists that can distinguish between systemic inflammation and sepsis.

When diagnosing sepsis, flow cytometry can be useful. You only need a small amount of blood, and you'll see effects quickly. We can start treating patients based on their immune systems, which are evaluated by flow cytometric expression of cell surface markers [24]. There was a strong correlation between mHLA-DR and the likelihood of sepsis in this investigation. In the peripheral blood of neonates with septic shock, Winkler et al. (2017) found a greater number of monocytes but decreased expression of HLA-DR [25].  In neonates with sepsis, Genel et al. (2010) found a decreased HLA-DR level that had predictive value [26]. In contrast, HLA-DR expression on monocytes did not differ significantly across infected, non-infected, and control groups in a study by Ng et al. (2006) [27].

Das et al,. (2002) also discovered that neonates with sepsis had a higher mean serum cortisol level. This could be because sepsis causes a marked increase in cortisol production by the adrenal cortex, which is driven centrally [28]. The cortisol level was also significantly higher in the definite sepsis group compared to the probable sepsis group. An increase in cortisol levels was also found in neonatal septic shock by Bhat et al., (2022) [29]. 

The ROC curve for mHLA-DR demonstrated a high level of sensitivity and specificity with an area under curve of 0.99±0.009 and 0.83±0.05, respectively. In contrast, the ROC curve for cortisol showed an area under curve of (1±0.0) and a high level of sensitivity and specificity for both confirmed and suspected cases compared to controls (P = 0.0001*). Therefore, it would be more effective to combine the two tests when diagnosing late-onset neonatal sepsis.

Limitation: This study does have a few caveats. Some examples include a bigger sample size, continuous monitoring of cortisol levels, and HLA-DR expression on monocytes as the disease progress.

Conclusion

Serum cortisol levels rise and monocyte HLA-DR expression in blood decreases in cases of definite neonatal  sepsis compared to those of probable  neonatal sepsis, these markers are more sensitive and specific than C-reactive protein (CRP) when used to diagnose late-onset neonatal sepsis. Early detection of late-onset neonatal sepsis can improve outcomes and shorten neonatal hospital stays by allowing for the prompt initiation of appropriate treatment.