Monocyte, Lymphocyte and Neutrophil Ratios – Easy-to-Use Biomarkers for the Diagnosis of Pediatric Tuberculosis : The Pediatric Infectious Disease Journal

Secondary Logo

Journal Logo

Advanced Search
Translational Medicine Reports

Monocyte, Lymphocyte and Neutrophil Ratios – Easy-to-Use Biomarkers for the Diagnosis of Pediatric Tuberculosis

Kissling, Mirjam BM*; Fritschi, Nora MD*,†; Baumann, Philipp MD‡,§; Buettcher, Michael MD¶,‖; Bonhoeffer, Jan MD†; Naranbhai, Vivek MBChB, PhD, Dphil**,††,‡‡; Ritz, Nicole MD, PhD*,†,¶,§§

Author Information

From the *Department of Clinical Research, Mycobacterial and Migrant Health Research Group, University of Basel, Switzerland

†University Children’s Hospital Basel, Switzerland

‡Department of Intensive Care Medicine and Neonatology, University Children’s Hospital Zurich, University of Zurich, Zurich, Switzerland

§Infectious Disease and Vaccinology Unit, University Children’s Hospital Basel, University of Basel, Basel, Switzerland

Paediatric Infectious Diseases Unit, Children’s Hospital, Lucerne Cantonal Hospital, Lucerne Switzerland

‖Paediatric Pharmacology and Pharmacometrics Research Center, University Children’s Hospital Basel, University of Basel, Basel, Switzerland

**Massachusetts General Hospital Cancer Center, Massachusetts General Hospital, Boston

††Dana-Farber Cancer Institute, Boston, Massachusetts

‡‡Center for the AIDS Programme of Research in South Africa, Durban, South Africa

§§Department of Pediatrics, The Royal Children’s Hospital Melbourne, The University of Melbourne, Australia.

Accepted for publication January 26, 2023

N.F. and N.R. were supported by the Thomi Hopf foundation and by the European Union’s Horizon 2020 research and innovation program. The CITRUS study was supported by the following foundations: Bangerter Rhyner Stiftung, Lunge Zürich, Nora van Meeuwen-Häfliger Stiftung, Rozalia Foundation, Schweizerische Lungenstiftung and Nikolaus and Bertha Burckhardt Bürgin Foundation.

The ProPAED data used in this study were obtained from an investigator-initiated study without external funding.

The authors have no conflicts of interest to disclose.

Mirjam Kissling and Nora Fritschi contributed equally to the article.

The CITRUS study was approved by the local ethics committee (Ethik Kommission Nordwestschweiz 2016-01094). The ProPAED study was approved by both, the ethics committee of the University Basel and Kanton Aargau.

This study was designed by N.R. and N.F. data collection was performed by P.B., M.B. and J.B.; data analysis was carried out by M.K. and N.F.; M.K., N.F. and N.R. wrote the first draft; all the authors revised the article and approved the final version.

Part of these results was presented at the 40th annual meeting of the European Society of Paediatric Infectious Diseases, 8-12 May 2022, Athens, Greece.

Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal’s website (www.pidj.com).

Address for correspondence: Nora Fritschi, MD, University Children’s Hospital Basel and Department of Clinical Research, University of Basel, Spitalstrasse 33, CH-4031 Basel, Switzerland. E-mail: [email protected].

This is an open-access article distributed under the terms of the Creative Commons Attribution-Non Commercial-No Derivatives License 4.0 (CCBY-NC-ND), where it is permissible to download and share the work provided it is properly cited. The work cannot be changed in any way or used commercially without permission from the journal.

The Pediatric Infectious Disease Journal 42(6):p 520-527, June 2023. | DOI: 10.1097/INF.0000000000003901

Abstract

Diagnosing tuberculosis (TB) in children is particularly challenging since a significant proportion of children with TB disease present with no or only minimal symptoms,1 has a paucibacillary disease, and collection of a microbiological sample is hampered by difficulties in obtaining adequate sputum.2 The diagnostic gold standard for TB disease is bacteriological confirmation by microscopy, culture or molecular testing which in the pediatric population is at best positive in 60%.3–7 In the absence of a microbiological confirmatory test, composite diagnostic definitions are used for research purpose and clinical care of TB in children. These consider combinations of immunodiagnostics [interferon-γ release assays (IGRA) and tuberculin skin test (TST)], imaging [chest radiography or computed tomography (CT)], clinical symptoms, a history of contact to a TB index case and clinical improvement with treatment.2,8

The host-pathogen interaction in TB is complex and evokes a variety of innate and adaptive immune mechanisms. These are still only partially understood but are increasingly thought to be specific for infection with Mycobacterium tuberculosis. Early studies from the 1920s in rabbits9 and children10 showed an association of TB disease with increased monocyte and decreased lymphocyte counts resulting in increased monocyte-to-lymphocyte ratios (MLR). It was believed that a high ratio suggests susceptibility for TB disease and a low ratio was observed in patients with healing or less extensive TB disease at a time before antituberculous drugs were available and the natural course of the diseases had to be observed. In recent years, accumulating evidence from genomic, transcriptomic and proteomic studies have highlighted the importance of the innate immune response in TB disease.11–14 An interferon-inducible gene profile dominates throughout the progression from infection to TB disease with subsequent upregulation of monocytes and downregulation of the lymphocyte cell subset.11–14 The same studies also show the upregulation of proteins involved in the neutrophil response associated with TB disease.11–14

Studies evaluating neutrophil-, monocyte- and lymphocyte counts in patients with TB disease compared to healthy controls found increased neutrophil- and monocyte counts and decreased lymphocyte counts.10,15–18 Although these patterns seem to be consistent, the observed changes in absolute numbers of neutrophils, monocytes and lymphocytes were usually minor. Stronger associations were observed when ratios were used: namely the neutrophil-to-lymphocyte ratio (NLR), the neutrophil-to-monocyte plus lymphocyte ratio (NMLR) and the monocyte-to-lymphocyte ratio. Evaluation of those ratios as diagnostic biomarkers has demonstrated a discriminatory power for the MLR and NLR between TB disease and infection or healthy individuals (MLR,15,19 NLR20) at the time of presentation. In addition, the ratios were found to have prognostic value for disease progression.17,20–24 So far, only a few studies evaluating these ratios were performed in children,17,19,22 all of which originate from high-TB incidence settings. Few studies evaluated the performance of the ratios as diagnostic markers for TB disease when compared to sick controls such as other respiratory diseases,18,25,26 and none included children.

The aim of our study was to evaluate NLR, NMLR and MLR as diagnostic biomarkers for TB compared with other non-TB lower respiratory tract infections (nTB-LRTI) in children and with healthy TB-exposed children in a low-incidence setting. We assessed if the above ratios differ in children with TB disease compared with healthy TB-exposed, TB-infected children and children with nTB-LRTI.

METHODS

For this analysis, we included data from two prospective, multicenter Swiss studies: the ongoing ChIldhood TubeRcUlosis in Switzerland (CITRUS) study (NCT03044509)27 and the Procalcitonin guidance to reduce antibiotic treatment of lower respiratory tract infection in children and adolescents (ProPAED) study (ISRCTN 17057980).28–30

For the CITRUS study all children <18 years evaluated for TB infection, TB disease or TB exposure in one of the nine participating centers in Switzerland (Bern, Basel, Zurich, Lausanne, Geneva, Aarau, St. Gallen, Lucerne, Bellinzona) were eligible. Written and signed consent was obtained by the children’s guardians or by the adolescents and/or their guardians. All children recruited for the CITRUS study from May 2017 to August 2021 were included in this analysis. Children with a chronic disease that potentially impacts the complete blood count (CBC) values or known infection with HIV or children without CBC reported were excluded.

We collected demographics, BCG vaccination status, information on symptoms, diagnostics and treatment. All data were extracted into a centralized database (RedCap).31 The first routine CBC results were collected within 3 days of study inclusion. IGRA results (QuantiFERON-TB Gold or T-SPOT.TB) were collected and classified according to the manufacturer’s instruction (generally a positive result was defined as interferon-γ concentration > 0.035 IU for QuantiFERON-TB Gold and > 5 Spots for T-SPOT.TB). In line with the Swiss guidelines, a TST of ≥ 5 mm was considered as positive.32

In accordance with the 2015 published case definitions33 TB disease was diagnosed either if culture or molecular confirmation was obtained from at least one specimen or if two additional criteria, including TB suggestive symptoms, suggestive chest radiography, close contact to a TB index case or immunodiagnostic evidence (a positive IGRA and/or TST) were found. TB infection was defined by a positive IGRA and/or TST and the absence of TB disease. TB exposure was defined as exposure in the same room as the TB index case for cumulatively more than 8 hours in the previous 3 months.34

Children with nTB-LRTI recruited for the ProPAED study served as sick controls. Eligible for this study were children with nTB-LRTI presenting to the pediatric emergency departments in Basel or Aarau from January 2009 to February 2010. Inclusion criteria for ProPAED were as follows: age 1 month to 18 years, presentation with LRTI including fever and cough, regardless of previous antibiotic treatment history. Excluded were children with severe immune suppression or known HIV infection, immunosuppressive treatment, neutropenia, cystic fibrosis, viral laryngotracheitis, hospital stay within the previous 14 days and other severe infections. The following variables were extracted from the ProPAED database: demographics, polymerase chain reaction (PCR) results from the nasopharyngeal aspirate analysis, blood culture (BC) results, findings from chest radiography as well as CBC values at the date of study inclusion. All CBCs were performed in the routine clinical laboratory of the recruiting hospitals with CBC instruments undergoing regular quality controls.

NLR was calculated by dividing absolute neutrophil by lymphocyte count. NMLR was calculated by dividing neutrophil by monocyte plus lymphocyte count. MLR was calculated by dividing absolute monocyte by lymphocyte count. For comparisons and associations with age, children were grouped as follows: < 2 years, 2 to < 5 years, 5 to < 10 years and ≥10 years.35

STATISTICS

Absolute CBC and ratios were compared between the groups using Mann-Whitney-U-tests. In the case of pairwise comparison of multiple groups P-values were adjusted with Bonferroni correction for comparisons made between the 4 diagnostic groups.

For the receiver operating characteristic (ROC) analysis, the area under the curve (AUC) was calculated with the DeLong method, the confidence interval of the curve was calculated with bootstrapping and sensitivity and specificity were defined at a cutoff based on maximal Youden index. For the ROC curves, we evaluated the performance of the ratios for the diagnosis of children with TB disease, when compared to all other children merged, sick controls, children with TB infection and TB-exposed children.

To assess age as a confounder a descriptive analysis of the healthy TB-exposed children of the absolute CBC values and ratios was analyzed using a stratified ROC analysis at cutoffs of 5 and 10 years. All calculations and graphs were carried out with R (version 4.1.2) and R Studio (version 2021.09.1).

RESULTS

Study Population

A total of 464 children were evaluated for inclusion in the CITRUS and ProPAED study. In the CITRUS study 63 of 127 (50%) and in the ProPAED 13 of 337 (4%) were excluded (Fig. 1). Of the excluded children in the CITRUS study, 58 (45%) had no CBC available. These were 8 children (14%) with TB disease, 23 (40%) with TB infection and 27 (47%) were healthy TB-exposed. A total of 389 children were included in the final analysis, 25 (6 %) with TB disease, 12 (3%) with TB infection, 28 (7%) healthy TB-exposed and 324 (83%) with nTB-LRTI.

F1
FIGURE 1.:
Flow diagram of the children included in the CITRUS and ProPaed study, with reason for exclusion. A total of 389 children were eligible for this analysis. CBC indicates complete blood count; nTB-LRTI, non-TB lower respiratory tract infections; TB, tuberculosis.

Of the children with nTB-LRTI, 22 had a bacterial infection detected via nasopharyngeal airway swabbing (PCR) or BC (n = 4 Chlamydia pneumoniae (PCR), n = 11 Mycoplasma pneumonia (PCR), n = 1 Bordetella pertussis (PCR), n = 4 Streptococcus pneumococcae (in BC), n = 1coagulase-negative staphylococci (BC), n = 1 Streptococcus pyogenes (BC). A further 167 children had a viral respiratory infection detected (all by nasopharyngeal airway swabbing PCR): n = 71 respiratory syncytial virus, n = 41 human metapneumovirus, n = 35 influenza virus, n = 6 adenovirus and n = 4 parainfluenza viruses. A total of 10 children had more than 1 virus detected, 10 had viral and bacterial pathogens identified, 119 had no pathogen identified and for 6 children pathogen testing was not performed.

Overall, the median [interquartile range (IQR)] age was 3.1 (1.4, 6.2) years and 226 (58%) were male. Notably, with a median age of 2.8 (1.2, 5.4) years children with nTB-LRTI were younger compared to children with TB disease [median 8.9 (2.6, 14.9) years] and children with TB infection [median 8.2 (3.5, 11.7) years]. Further baseline characteristics are displayed in Table 1.

TABLE 1. - Baseline Characteristics of the Study Population
TB disease TB infection Healthy exposed nTB-LRTI
Number of children 25 12 28 324
Male sex, n (%) 11 (44.0) 8 (66.7) 21 (75.0) 185 (57.4)
Median age, (yr)
IQR age, (yr)		 8.91
(2.60, 14.92)		 8.21
(3.52, 11.67)		 5.33
(2.97, 9.57)		 2.83
(1.20, 5.35)
Age, n (%)
 <2 yr 4 (16.0) 0 (0.0) 6 (21.4) 128 (39.5)
 2 to <5 yr 5 (20.0) 5 (41.7) 8 (28.6) 103 (31.8)
 5 to <10 yr 6 (24.0) 1 (8.3) 9 (32.1) 72 (22.2)
 ≥10 yr 10 (40.0) 6 (50.0) 5 (17.9) 21 (6.5)
Chronic disease, n (%) 4 (16.0) 0 (0.0) 1 (3.6) 0 (0.0)
Foreign-born, n (%) 12 (48.0) 10 (83.3) 5 (17.9) NA
Contact person
 Same household
  Yes, n (%) 11 (44.0) 5 (41.7) 22 (78.6) NA
  No, n (%) 9 (36.0) 8 (57.1) 6 (21.4) NA
 Parent, n (%) 1 (4.0) 3 (25.0) 6 (21.4) NA
 Parent + other family, n (%) 2 (8.0) 1 (8.3) 0 (0.0) NA
 Other family, n (%) 10 (40.0) 1 (8.3) 15 (53.6) NA
 Other contact, n (%) 1 (4.0) 3 (25.0) 7 (25.0) NA
Children with TB disease, TB infection or healthy exposed were recruited for the CITRUS study. Children recruited for the ProPAED study with non-TB lower respiratory tract infection (nTB-LRTI) served as sick controls.
NA indicates not applicable; nTB-LRTI, non-TB lower respiratory tract infection; TB, tuberculosis.

Absolute Neutrophil, Monocyte and Lymphocyte Count in Diagnosis Groups

The median (IQR) neutrophil count was 5.0 (3.1, 6.1) × 109/L in children with TB disease, 2.7 (2.2, 3.5) × 109/L in children with TB infection, 3.1 (2.4, 3.5) × 109/L in healthy TB exposed children and 0.8 (0.3, 2.1) × 109/L in children with nTB-LRTI (TB disease vs. TB infection P = 0.210; TB disease vs. TB exposure P = 0.012; TB disease vs. nTB-LRTI P < 0.001) (Table 2).

TABLE 2. - Absolute Values and Ratios of Complete Blood Counts
TB disease TB infection Healthy exposed nTB-LRTI
Number of children 25 12 28 324
Neutrophils × 109/L [median (IQR)] 5.01 (3.10, 6.12) 2.74 (2.19, 3.52) 3.06 (2.39, 3.53) 0.81 (0.34, 2.07)
Monocytes × 109/L [median (IQR)] 0.71 (0.47, 0.84) 0.57 (0.39, 0.65) 0.48 (0.41, 0.63) 0.90 (0.55, 1.38)
Lymphocytes × 109/L [median (IQR)] 2.74 (1.82, 3.83) 2.60 (1.94, 3.36) 3.30 (2.72, 4.80) 2.62 (1.55, 4.18)
NLR [median (IQR)] 2.03 (1.23, 2.23) 1.08 (0.67, 1.50) 0.81 (0.64, 1.29) 0.31 (0.11, 0.97)
NMLR [median (IQR)] 1.42 (1.20, 1.69) 0.89 (0.56, 1.20) 0.71 (0.56, 1.08) 0.24 (0.08, 0.61)
MLR [median (IQR)] 0.23 (0.18, 0.35) 0.23 (0.17, 0.26) 0.16 (0.11, 0.20) 0.34 (0.21, 0.58)
Values are displayed as median and interquartile range (IQR) stratified by diagnostic group.
MLR indicates monocyte-to-lymphocyte-ratio; NLR, neutrophil-to-lymphocyte-ratio; NMLR, neutrophil-to-monocyte-plus-lymphocyte-ratio; nTB-LRTI, non-TB lower respiratory tract infection; TB, tuberculosis.

The median (IQR) monocyte count was 0.7 (0.5, 0.8) × 1099/L for children with TB disease, 0.6 (0.4, 0.7) × 109/L for children with TB infection, 0.5 (0.4, 0.6) × 109/L for healthy TB exposed and 0.9 (0.6, 1.4) × 109/L for children with nTB-LRTI (TB disease vs. TB infection P = 0.336; TB disease vs. TB exposure P = 0.24; TB disease vs. nTB-LRTI P = 0.060) (Table 2).

The median (IQR) lymphocyte count was 2.7 (1.8, 3.8) × 109/L in children with TB disease, 2.6 (1.9, 3.4) × 109/L in children with TB infection, 3.3 (2.7, 4.8) × 109/L in healthy TB exposed children and 2.6 (1.6, 4.2) × 109/L in children with nTB-LRTI (TB disease vs. TB infection: P = 1; TB disease vs. TB exposure P = 0.324; TB disease vs. nTB-LRTI P = 1) (Table 2).

NLR, NMLR and MLR in Age and Diagnosis Groups

The median (IQR) NLR was highest in children with TB disease with 2.0 (1.2, 2.2) and lower in healthy TB-exposed children [0.8 (0.6, 1.3), P = 0.002] and children with nTB-LRTI with 0.3 [(0.1, 1.0), P < 0.001] (Fig. 2, see Supplemental Digital Content 1, https://links.lww.com/INF/E982). The median (IQR) NMLR was highest in children with TB disease with 1.4 (1.2, 1.7) and lower in healthy TB-exposed children [0.7 (0.6, 1.1), P = 0.003] and children with nTB-LRTI [0.2 (0.1, 0.6), P < 0.001). The median (IQR) MLR was 0.3 (0.2, 0.6) in children with nTB-LRTI and in a similar range compared to children with TB disease [0.2 (0.2, 0.4), P = 0.106] and children with TB infection [0.2 (0.2, 0.3), P = 0.024], and higher compared to healthy TB exposed children [0.2 (0.1, 0.2), P < 0.001].

F2
FIGURE 2.:
Boxplot with Tukey whiskers and scattered dot plots of absolute values and ratios of complete blood counts stratified by diagnostic groups. Groups were compared with pairwise Mann-Whitney-U-Tests and P values were adjusted with Bonferroni Correction. Results of hypothesis testing are given as * P ≤ 0.05, ** P ≤ 0.01, *** P ≤ 0.001, **** P ≤ 0.0001. MLR indicates monocyte-to-lymphocyte-ratio; NLR, neutrophil-to-lymphocyte-ratio; NMLR, neutrophil-to-monocyte-plus-lymphocyte-ratio.

The median (IQR) NLR, NMLR and MLR in children aged < 2 years, 2 to < 5 years, 5 to < 10 years and ≥ 10 years, respectively is shown in Supplemental Digital Content 1, https://links.lww.com/INF/E982.

Diagnostic Performance of the NLR, NMLR and MLR for TB Disease

ROC curve analysis of children with TB disease compared to all other groups showed that NLR and NMLR performed best. NLR at a cutoff of 1.16 had a sensitivity of 80% with a specificity of 76% (AUC: 0.81; 95% CI: 0.75–0.88) and NLMR at a cutoff of 1.19 had a sensitivity of 76% and specificity of 86% (AUC: 0.85; 95% CI: 0.79–0.92) (Fig. 3). MLR at a cutoff of 0.44 had a sensitivity of 92% but a considerably lower specificity of 32% (AUC: 0.6; 95% CI, 0.51–0.69). Similar findings were seen in the ROC curve analysis comparing children with TB disease to children with nTB-LRTI and healthy TB exposed (Fig. 3).

F3
FIGURE 3.:
ROC analysis of the diagnostic performance of NLR, NMLR and MLR. Children with TB disease were compared to all children merged, nTB-LRTI, TB infection and healthy TB exposed. The corresponding area under the curve (AUC) with 95% CI, cutoff and the according sensitivity and specificity at this cutoff are given for every comparison and ratio. CI indicates confidence interval; MLR, monocyte-to-lymphocyte-ratio; NLR, neutrophil-to-lymphocyte-ratio; NMLR, neutrophil-to-monocyte-plus-lymphocyte-ratio; nTB-LRTI, non-TB lower respiratory tract infection; ROC, receiver operating characteristic; Sens, sensitivity, spec, specificity TB, tuberculosis.

In ROC analysis stratified into two age groups at cutoffs of 5 and 10 years, we found only a minor influence of age in the diagnostic performance of the ratios to diagnose TB disease compared to all other children merged and sick controls (see Figure, Supplemental Digital Content 2, https://links.lww.com/INF/E983).

Diagnostic Performance of Absolute Neutrophil Count

ROC analysis of the neutrophil count of children with TB disease compared to all other groups showed a cutoff of 1.57 a sensitivity of 100% with a specificity of 62% (AUC: 0.85; 95 % CI: 0.8–0.9). The diagnostic performance of the neutrophils to distinguish children with TB disease from children with nTB-LRTI was in the same range as the performance to distinguish children with TB disease from all other groups and is shown in Figure, Supplemental Digital Content 3, https://links.lww.com/INF/E983).

DISCUSSION

The TB diagnostic potential of ratios from CBC including NLR, NMLR and MLR has been suggested many decades ago but has attracted renewed interest in recent years. Our study in children is unique as control with other nTB-LRTI was included, which is the most common clinical scenario. This is particularly important as in many settings TB is only suspected in symptomatic children but symptoms, such as fever, cough, and failure to thrive, on their own, have low specificity. The results in this study show that NLR and NMLR are higher in children with TB disease with a sensitivity above 88% and specificity above 76% to distinguish children with TB disease from children with other nTB-LRTI or healthy children and were superior to the performance of MLR. The absolute neutrophil count alone showed a higher sensitivity but lower specificity compared to the ratios.

These findings are consistent with results from a few prior studies in high-incidence TB settings, notwithstanding the absence of control groups in those settings. This suggests that these ratios may be used as easy-to-obtain biomarkers for the diagnosis of TB disease in children, which have comparable performance to the currently used, more specific testing recommended for TB. For example, data from meta-analyses on the performance of immune-diagnostic tests as a supportive tool for the diagnosis of TB disease in children have shown that if IGRA and TST are pooled the sensitivity is approximately 60%–90% and with a specificity of 85% or greater.36–39

Findings from earlier studies in adults have also shown the association of higher ratios in TB disease when compared to healthy controls.15–18,20–23,40–46 One study from Spain including 60 adults (21 with TB disease, 19 with TB infection and 20 healthy TB exposed) reported a sensitivity of 77% and a specificity of 81% (cutoff of 2.85 and AUC of 0.83) to classify patients with TB disease versus exposed uninfected individuals.40 Another study from Turkey including 134 adults (51 with TB disease, 40 with sarcoidosis and 43 controls) reported a sensitivity of NLR of 88% and a specificity of 80% (cutoff of 2.16 and AUC of 0.92) to classify patients with TB disease versus controls.45 In a further study in 173 adults in Italy even higher sensitivity of 91% and a specificity of 94% (at a cutoff of 0.28) was reported to distinguish adults with TB disease from healthy controls.15

MLR in our study had a lower sensitivity and specificity than NLR and NLMR to distinguish between children with TB disease and nTB-LRTI. However, monocytosis and neutrophilia may be associated with a viral infection of the lower respiratory tract. It is therefore possible that the high rate of children with nTB-LRTI with viral pathogenesis resulted in many children with monocytosis and thus driving the lower sensitivity and specificity to distinguish between TB disease and nTB-LRTI controls.47

The challenge remains to define optimal cutoffs for these ratios. As evidence was limited and cutoffs are highly variable in the literature, we did not use predefined cutoffs but calculated ROC curves and analyzed which cutoffs would best be suitable for our study population. Using this approach, we were able to detect TB disease in children referred for TB investigation using NLR and NMLR with sensitivities of 76%–80% and specificities of 76%–86%. Validation of these cutoffs proposed here in other settings may be useful.

The immune mechanisms explaining the findings in our study originate from recent evidence showing that innate immunity plays an important role in TB disease.11–14 In human in vitro and mouse in vivo models, persistent infection with M. tuberculosis is shown in hematopoietic stem cells in the bone marrow. As a result, M. tuberculosis itself favors the development of the progenitor cells towards the myeloid lineage stimulating the generation of monocytes and neutrophils over the lymphoid lineage.48,49 These underlying mechanisms may lead to increased neutrophil and monocyte counts, and decreased lymphocyte counts, resulting in elevated NLR, MLR and NMLR associated with TB disease, as found in our and other studies.

One key challenge is that ratios such as NLR, NMLR and MLR are nonspecific and may also be influenced by other infections. Therefore, a comparison of the diagnostic potential of these ratios must be evaluated against sick controls as performed in our study by including children with nTB-LRTI. Only a few prior studies, however, included sick controls and none – to our knowledge – were done in children. One study evaluating adults with TB disease and other respiratory infections, including viral pneumonia, bacterial pneumonia, aspiration pneumonia and empyema, found lower NLR, NMLR and MLR in patients with TB compared to the other respiratory infections.18 Two further studies looked at adults with TB disease compared to community-acquired bacterial pneumonia: with lower25 and higher NLR26 shown in patients with TB disease compared to the sick controls. In comparison to these conflicting results in adults, our study showed higher ratios in children with TB disease compared to children with nTB-LRTI. This might be explained by different ages, differences in pathophysiology of TB disease in children and adults and differences of causative pathogens for respiratory tract infections in younger age groups. In particular, the high rate of viral infections in our sick control group may have led to only a limited increase in neutrophil count in our control group. This has been shown in a study in children comparing NLR in respiratory infections, showing viral infections had lower NLR compared to bacterial infections.50–52 Viral infections are the most common reason for childhood respiratory infections, also in high TB-endemic regions, and thus our study setting represents a real-life comparison. In this study, an association between NLR, NMLR and MLR and TB disease was shown. These biomarkers could thus be used early in the diagnostic evaluation of children with respiratory symptoms to assess the likelihood of TB disease or in children with a positive immunodiagnostic test and only minor or absent symptoms to distinguish between TB disease and infection.

Age may also influence the ratios and therefore may be a confounding factor, as the age distribution varied in the diagnostic groups in our study. We, therefore, stratified our analysis by age at 2 cutoffs, 5 and 10 years. The performance of the ratios as diagnostic markers for TB disease was similar in our analysis. Nevertheless, other studies showed associations of monocytes, neutrophils and lymphocytes and NLR with age: for NLR a significant change over age was shown, with a considerable reduction until the age of 6 months, followed by a nearly 3-fold increase from 2 to 18 years of age and with higher rates in females during puberty.53 We are therefore unable to exclude an age-specific effect on the ratios because we had a too low number of healthy children.

Our study has some other limitations: due to the limited sample size, we were not able to analyze differences between microbiologically confirmed and unconfirmed TB. Further, we did not include children living with HIV infection or malnutrition. Our study included data from multiple centers in Switzerland, which is a low TB incidence setting and the translational value for high TB incidence settings may be limited.

In conclusion, this study shows that NLR and NMLR are promising easy-to-obtain diagnostic biomarkers to differentiate children with TB disease from other lower respiratory tract infections. These results require validation in a larger study sample and low and high TB incidence settings.

ACKNOWLEDGMENTS

We thank all the local PI’s of the CITRUS study: Sara Bernhard, Lisa Kottanattu, Andrea Duppenthaler, Anne Morand, Jürg Barben, Christoph Berger, Christa Relly, Isabelle Rochat, Marie Rohr. We thank Noemi Meier. We thank all the investigators of the ProPAED study: Gurli Baer, Ulrich Heininger, Gerald Berthet, Julia Schäfer, Heiner Bucher, Daniel Trachsel, Jaques Schneider, Muriel Gambon, Diana Reppucci, Jessica Bonhoeffer, Jody Stähelin-Massik, Philipp Schuetz, Beat Mueller, Gabor Szinnai, and Urs Schaad, and all the children and their parents, who participated in this study. Further, we would thank Andrew Aktinson for his support in the data analysis.

REFERENCES

1. Fritschi N, Wind A, Hammer J, et al. Subclinical tuberculosis in children: diagnostic strategies for identification reported in a 6-year national prospective surveillance study. Clin Infect Dis. 2022;74:678–684.
2. Perez-Velez CM, Roya-Pabon CL, Marais BJ. A systematic approach to diagnosing intra-thoracic tuberculosis in children. J Infect. 2017;74:S74–S83.
3. Guidance for National Tuberculosis Programmes on the Management of Tuberculosis in Children. 2nd ed. World Health Organization; 2014.
4. Fritschi N, Schmidt AJ, Hammer J, et al. Pediatric tuberculosis disease during years of high refugee arrivals: a 6-year national prospective surveillance study. Respiration. 2021;100:10501–11059.
5. Tebruegge M, Ritz N, Curtis N, et al. Diagnostic tests for childhood tuberculosis: past imperfect, present tense and future perfect? Pediatr Infect Dis J. 2015;34:1014–1019.
6. Ritz N, Curtis N. Novel concepts in the epidemiology, diagnosis and prevention of childhood tuberculosis. Swiss Med Wkly. 2014;144:w14000.
7. Oesch Nemeth G, Nemeth J, Altpeter E, et al. Epidemiology of childhood tuberculosis in Switzerland between 1996 and 2011. Eur J Pediatr. 2014;173:457–462.
8. Gunasekera KS, Vonasek B, Oliwa J, et al. Diagnostic challenges in childhood pulmonary tuberculosis—optimizing the clinical approach. Pathogens. 2022;11:382.
9. Sabin FR, Sugiyama S, Kindwall JA, et al. The role of the monocyte in tuberculosis. B Johns Hopkins Hosp. 1925;4:231–280.
10. Rogers PM. A study of the blood monocytes in children with tuberculosis. N Engl J Med. 1928;198:740–749.
11. Scriba TJ, Penn-Nicholson A, Shankar S, et al.; other members of the ACS cohort study team. Sequential inflammatory processes define human progression from M. tuberculosis infection to tuberculosis disease. PLoS Pathog. 2017;13:e1006687.
12. Berry MP, Graham CM, McNab FW, et al. An interferon-inducible neutrophil-driven blood transcriptional signature in human tuberculosis. Nature. 2010;466:973–977.
13. Joosten SA, Fletcher HA, Ottenhoff THM. A helicopter perspective on TB biomarkers: pathway and process based analysis of gene expression data provides new insight into TB pathogenesis. PLoS One. 2013;8:e73230.
14. Ottenhoff THM, Dass RH, Yang N, et al. Genome-wide expression profiling identifies type 1 interferon response pathways in active tuberculosis. PLoS One. 2012;7:e45839-e.
15. La Manna MP, Orlando V, Dieli F, et al. Quantitative and qualitative profiles of circulating monocytes may help identifying tuberculosis infection and disease stages. PLoS One. 2017;12:e0171358-e.
16. Wang W, Wang LF, Liu YY, et al. Value of the ratio of monocytes to lymphocytes for monitoring tuberculosis therapy. Can J Infect Dis Med Microbiol. 2019;2019:3270393.
17. Rakotosamimanana N, Richard V, Raharimanga V, et al. Biomarkers for risk of developing active tuberculosis in contacts of TB patients: a prospective cohort study. Eur Respir J. 2015;46:1095–1103.
18. Jeon Y, Lee WI, Kang SY, et al. Neutrophil-to-monocyte-plus-lymphocyte ratio as a potential marker for discriminating pulmonary tuberculosis from nontuberculosis infectious lung diseases. Lab Med. 2019;50:286–291.
19. Choudhary RK, Wall KM, Njuguna I, et al. Monocyte-to-lymphocyte ratio is associated with tuberculosis disease and declines with anti-TB treatment in HIV-infected children. J Acquir Immune Defic Syndr. 2019;80:174–181.
20. Miyahara R, Piyaworawong S, Naranbhai V, et al. Predicting the risk of pulmonary tuberculosis based on the neutrophil-to-lymphocyte ratio at TB screening in HIV-infected individuals. BMC Infect Dis. 2019;19:667.
21. Naranbhai V, Hill AVS, Abdool Karim SS, et al. Ratio of monocytes to lymphocytes in peripheral blood identifies adults at risk of incident tuberculosis among HIV-infected adults initiating antiretroviral therapy. J Infect Dis. 2014;209:500–509.
22. Naranbhai V, Kim S, Fletcher H, et al. The association between the ratio of monocytes: lymphocytes at age 3 months and risk of tuberculosis (TB) in the first two years of life. BMC Med. 2014;12:120.
23. Yin Y, Kuai S, Liu J, et al. Pretreatment neutrophil-to-lymphocyte ratio in peripheral blood was associated with pulmonary tuberculosis retreatment. Arch Med Sci. 2017;13:404–411.
24. Fritschi N, Vaezipour N, Buettcher M, Portevin D, Naranbhai V, Ritz N. The monocyte to lymphocyte and neutrophil to lymphocyte ratio as diagnostic, treatment response and prognostic biomarker for tuberculosis: a systematic review. Manuscript in review.
25. Yoon NB, Son C, Um SJ. Role of the neutrophil-lymphocyte count ratio in the differential diagnosis between pulmonary tuberculosis and bacterial community-acquired pneumonia. Ann Lab Med. 2013;33:105–110.
26. Berhane M, Melku M, Amsalu A, et al. The role of neutrophil to lymphocyte count ratio in the differential diagnosis of pulmonary tuberculosis and bacterial community-acquired pneumonia: a cross-sectional study at Ayder and Mekelle Hospitals, Ethiopia. Clin Lab. 2019;65:527−533.
27. Meier NR, Sutter TM, Jacobsen M, et al. Machine learning algorithms evaluate immune response to novel mycobacterium tuberculosis antigens for diagnosis of tuberculosis. Front Cell Infect Microbiol. 2021;10:594030.
28. Baer G, Baumann P, Buettcher M, et al. Procalcitonin guidance to reduce antibiotic treatment of lower respiratory tract infection in children and adolescents (ProPAED): a randomized controlled trial. PLoS One. 2013;8:e68419.
29. Fuchs A, Gotta V, Decker ML, et al. Cytokine kinetic profiles in children with acute lower respiratory tract infection: a post hoc descriptive analysis from a randomized control trial. Clin Microbiol Infect. 2018;24:1341.e1–1341.e7.
30. Baumann P, Fuchs A, Gotta V, et al.; ProPAED study group. The kinetic profiles of copeptin and mid regional proadrenomedullin (MR-proADM) in pediatric lower respiratory tract infections. PLoS One. 2022;17:e0264305.
31. https://www.project-redcap.org/.
32. SwissLungAssociation. Tuberculosis in Switzerland - Guidance for Healthcare Professionals. 2021. Available at: www.tbinfo.ch. Accessed October 30, 2021.
33. Graham SM, Cuevas LE, Jean-Philippe P, et al. Clinical case definitions for classification of intrathoracic tuberculosis in children: an update. Clin Infect Dis. 2015;61(Suppl 3):S179–S187.
34. Lungenliga Schweiz. Manual of Tuberculosis - Revised version January 2021. 2021.
35. Cuevas LE, Browning R, Bossuyt P, et al. Evaluation of tuberculosis diagnostics in children: 2. Methodological issues for conducting and reporting research evaluations of tuberculosis diagnostics for intrathoracic tuberculosis in children. Consensus from an expert panel. J Infect Dis. 2012;205(Suppl 2):S209–S215.
36. Auguste P, Tsertsvadze A, Pink J, et al. Comparing interferon-gamma release assays with tuberculin skin test for identifying latent tuberculosis infection that progresses to active tuberculosis: systematic review and meta-analysis. BMC Infect Dis. 2017;17:200.
37. Chiappini E, Accetta G, Bonsignori F, et al. Interferon-γ release assays for the diagnosis of Mycobacterium tuberculosis infection in children: a systematic review and meta-analysis. Int J Immunopathol Pharmacol. 2012;25:557–564.
38. Laurenti P, Raponi M, de Waure C, et al. Performance of interferon-γ release assays in the diagnosis of confirmed active tuberculosis in immunocompetent children: a new systematic review and meta-analysis. BMC Infect Dis. 2016;16:131.
39. Mandalakas AM, Detjen AK, Hesseling AC, et al. Interferon-gamma release assays and childhood tuberculosis: systematic review and meta-analysis. Inter J Tuberc Lung Dis. 2011;15:1018–1032.
40. Estevez O, Anibarro L, Garet E, et al. Multi-parameter flow cytometry immunophenotyping distinguishes different stages of tuberculosis infection. J Infect. 2020;81:57–71.
41. Wang J, Yin Y, Wang X, et al. Ratio of monocytes to lymphocytes in peripheral blood in patients diagnosed with active tuberculosis. Braz J Infect Dis. 2015;19:125–131.
42. Handayani I, Massi MN, Leman Y, et al. Composite bacterial infection index and serum amyloid a protein in pulmonary tuberculosis patients and their household contacts in Makassar. Open Access Maced J Med Sci. 2021;9:557–562.
43. Abakay O, Abakay A, Sen HS, et al. The relationship between inflammatory marker levels and pulmonary tuberculosis severity. Inflammation. 2015;38:691–696.
44. Abedini A, Naderi Z, Kiani A, et al. The evaluation of interleukin-4 and interleukin-13 in the serum of pulmonary sarcoidosis and tuberculosis patients. J Res Med Sci. 2020;25:24.
45. Iliaz S, Iliaz R, Ortakoylu G, et al. Value of neutrophil/lymphocyte ratio in the differential diagnosis of sarcoidosis and tuberculosis. Ann Thorac Med. 2014;9:232–235.
46. Valizadeh Ardalan P, Servatyari K, Kashefi H, et al. The relationship between inflammatory markers extracted from complete blood count and active pulmonary tuberculosis. Rev Med Microbiol. 2019;30:18–25.
47. Bicer S, Giray T, Çöl D, et al. Virological and clinical characterizations of respiratory infections in hospitalized children. Ital J Pediatr. 2013;39:22.
48. Tamburini B, Badami GD, Azgomi MS, et al. Role of hematopoietic cells in Mycobacterium tuberculosis infection. Tuberculosis (Edinb, Scotland). 2021;130:102109.
49. Belay M, Tulu B, Younis S, et al. Detection of Mycobacterium tuberculosis complex DNA in CD34-positive peripheral blood mononuclear cells of asymptomatic tuberculosis contacts: an observational study. Lancet Microbe. 2021;2:e267–e275.
50. Bekdas M, Goksugur SB, Sarac EG, et al. Neutrophil/lymphocyte and C-reactive protein/mean platelet volume ratios in differentiating between viral and bacterial pneumonias and diagnosing early complications in children. Saudi Med J. 2014;35:442–447.
51. Naess A, Nilssen SS, Mo R, et al. Role of neutrophil to lymphocyte and monocyte to lymphocyte ratios in the diagnosis of bacterial infection in patients with fever. Infection. 2017;45:299–307.
52. Tamelytė E, Vaičekauskienė G, Dagys A, et al. Early blood biomarkers to improve sepsis/bacteremia diagnostics in pediatric emergency settings. Medicina (Kaunas). 2019;55:99.
53. Li K, Peng YG, Yan RH, et al. Age-dependent changes of total and differential white blood cell counts in children. Chin Med J (Engl). 2020;133:1900–1907.

Supplemental Digital Content

Copyright © 2023 The Author(s). Published by Wolters Kluwer Health, Inc.