Background
Mycoplasma pneumoniae (M. pneumoniae) is the predominant pathogen of community-acquired pneumonia (CAP), which contributes to approximately 10 to 40% of CAP cases in children [
1‐
3]. Most pediatric cases of
M. pneumoniae pneumonia (MPP) are benign and self-limited, however, there still are some cases showing clinical and radiological deterioration despite of macrolides antibiotic therapy for 7 days or longer, which are defined as refractory
Mycoplasma pneumoniae pneumonia (RMPP) [
4,
5]. The exact mechanisms of RMPP are not fully clarified, reducing the incidence of RMPP and improving its prognosis remain challenges. Our previous study and other literatures demonstrated glucocorticoid therapy attenuated the clinical manifestations, radiological findings and length of stay (LOS) of RMPP children [
6,
7], indicating excessive inflammation involved in RMPP pathogenesis [
8].
Micronutrients share inter-dependent relationships with host’s infection immunity [
9,
10]. As acute infection affects concentrations of some micronutrients (including vitamin A, ferritin) [
11‐
13], those circulating concentrations should be adjusted by acute phase protein (APP) to eliminate the impact of infection and reflect real micronutrient status [
13‐
15]. The most two commonly used APPs are C-reactive protein (CRP) and α-1-acid glycoprotein (AGP) [
14,
15]. Vitamin A is an essential micronutrient governs broad range of biological processes [
16]. Recent studies highlight the interactions between vitamin A status and immune response [
9,
17], demonstrating vitamin A deficiency (VAD) may cause imbalance between pro- and anti-inflammatory factors and excessive immune response [
18], which emerged in RMPP. Serum retinol or retinol-binding protein (RBP) concentrations represent vitamin A status. High-performance liquid chromatography (HPLC) is recommended for serum retinol assessment, while it’s expensive and technically challenging. The method of RBP assessment takes the advantages of being more robust for sample collection and handling processes. Therefore, RBP is often substituted as an indicator of vitamin A status. Literatures have showed higher incidence of VAD in MPP children than healthy children, and VAD was associated with MPP severity, which indicated vitamin A levels could be associated with RMPP incidence [
19]. Based on all above, we hypothesis that vitamin A levels could be associated with RMPP incidence. Therefore, we constructed this retrospective study to investigate adjusted vitamin A concentrations in MPP children and clarify the association between adjusted vitamin A levels and RMPP incidence, trying to provide more evidence for RMPP intervention.
Methods
Study population
This study was retrospectively conducted in Children’s Hospital of Chongqing Medical University, a 1500-bed tertiary level III teaching hospital in Chongqing, China. The study was conducted in accordance with the Declaration of Helsinki and approved by the Ethics Committee of Children’s Hospital of Chongqing Medical University. Data from patients were analyzed anonymously for hospitalized mycoplasma pneumoniae pneumonia (MPP) children from 1 September 2018 to 31 December 2019 retrospectively. The inclusions had the following characteristics: (i) inpatients MPP children; (ii) age between 6 months and 18 years old. The exclusion criteria included any of the following: (i) patients who had an underlying organ dysfunction;(ii) patients coinfected with other pathogens. According to the diagnostic criteria of RMPP, patients were divided in to RMPP group and general M. pneumoniae pneumonia (GMPP) group. Besides, 65 children with micronutrients measurements in Physical Examination Center matched with age, gender and testing time were selected as the healthy control group.
Definitions
MPP was diagnosed based on the followings: (i) clinical presentation (fever, cough, etc); (ii) chest imaging with infiltrates; (iii) having the positive results for
MP polymerase chain reaction (PCR) tests of nasopharyngeal secretions with serum anti-
MP IgM titer ≥1:160. The diagnosis of refractory
M. pneumoniae pneumonia (RMPP) was based on the presence of persistent fever and clinical manifestations as well as radiological deterioration after regular macrolides treatment for 7 days or longer [
4,
5], the other cases were defined as general
M. pneumoniae pneumonia (GMPP). The body mass index (BMI) was calculated by weight in kilograms divided by height in meters squared (kg/m
2). Extrapulmonary presentations include liver function abnormalities, myocarditis, encephalitis, rash, proteinuria, hemolytic anemia and arthritis. VAD was defined as RBP concentration lower than 0.7 μmol/L (15 mg/L) [
20].
M. pneumoniae detection
The specific antibodies against M. pneumoniae (IgG and IgM) were detected with passive particle agglutination (SERODIA-MYCO II, Japan) in nearly 2 ml serum samples of children on admission, MP antibody > 1:160 is a positive finding. Nasopharyngeal aspirate (NPA) was used for M. Pneumoniae DNA detection. In accordance with the manufacturer instructions, NPA was centrifugated 12,000 g for 5 min at 4 °C, the sediment was collected for DNA extraction with a real-time PCR commercial kit (Daan Gene Co. Ltd., Guangzhou, China). The DNA was then amplified using PCR primers and probes. Quantification curves were plotted using several concentrations of standard control samples (Daan Gene Co. Ltd., Guangzhou, China).
Micronutrients detection
Blood samples were collected from all inpatients during the first 24 h of admission. Serum micronutrients included ferritin, vitamin A, vitamin D, folate and vitamin B12 productions. Vitamin A concentrations were measured by retinol-binding protein (RBP). Vitamin D productions were measured by 25-hydroxy vitamin D (25(OH)D). Ferritin, folate, vitamin B12 and 25(OH)D concentrations were evaluated by Chemi Luminescence (Siemens, Germany), RBP levels were evaluated by Turbidimetric inhibition immunoassay (Homa Biological, Beijing, China). For accurate estimation, vitamin A and ferritin concentrations were adjusted by CRP, using regression correction (RC) approach [
14,
21].
Adjustment approach
Adjusted vitamin A concentrations were obtained through RC approach [
14,
21]. The RC approach was applied according to BRINDA methods articles, which uses linear regression to adjust RBP by the concentration of CRP. Briefly, the adjusted RBP equation was calculated by subtracting the influence of CRP, and RMPP as follows:
RBP adjusted = RBP unadjusted -β1 (CRP observe - CRP reference) -β2 (RMPP),
According to available data, β1 is the CRP regression coefficient, β2 is the RMPP regression coefficient, the reference of non-logged CRP is 8 mg/L, the minimum threshold of CRP measurement. CRP and RBP are all ln transformed, CRP and RBP are continuous variables, and RMPP is a dichotomous variable. The correction was only applied to individuals with CRP > 8 mg/L to avoid over adjustment. The same approach was applied to adjust ferritin.
BALF cytokines measurement
Mycoplasma pneumoniae pneumonia patients who received bronchoalveolar lavage, the bronchoalveolar lavage fluid (BALF) was extracted and collected, then delivered to the Center Laboratory Medicine immediately (no more than half an hour), kinds of cytokines including IL-2, IL-4, IL-6, IL-10, TNF-α, IFN-γ, IL-17a were measured by Cytometric Bead Array (CBA) (Cell-Gene, Hangzhou, China).
Data collection
Demographic characteristics (age, gender, weight, BMI), extrapulmonary manifestation, serum inflammatory factors (CRP, PCT, LDH, prealbumin), BALF inflammatory cytokines (IL-2, IL-4, IL-6, IL-10, TNF-α, IFN-γ, IL-17a), serum micronutrients (ferritin, vitamin A, vitamin D, folate and vitamin B12), oxygenation, fever duration and length of stay (LOS) were retrospectively collected from all children who were included in the study.
Statistical analysis
The continuous variables were presented as medians with 25 and 75% quartiles (interquartile range, IQR); the non-parametric test and Mann-Whitney U test were used for analysis. The categorical variables were presented as counts (percentages), and assessed by Kruskal-Wallis or the Fisher exact test. Correlations between variables were analyzed by Pearson Correlation. Univariate and multivariate logistic regression tests were used to evaluate the association between RMPP incidence and other variables. All the statistical analyses were conducted using SPSS 25.0 and Graphpad Prism 7.0 for Windows.
Discussion
M. pneumoniae is a leading cause of CAP, some MPP children could progress to RMPP. Researches have already noticed the important role of vitamin A in respiratory infection [
22,
23], and emphasized accurate estimation of vitamin A deficiency [
14]. Serum retinol or RBP concentrations provide vital information of vitamin A status and vitamin A deficiency (VAD) severity. Serum vitamin A concentrations could be affected by infection, for accurate estimation, adjusting it by CRP and/or AGP was recommended [
14,
15]. As compared with other adjustment approaches, RC approach adjusting vitamin A in a continuous manner that better reflects the association between vitamin A and APPs. Thus, vitamin A concentrations were adjusted by CRP with RC approach in this study. To our knowledge, this is the first study to investigate the association between the real vitamin A status and RMPP incidence. We demonstrated serum vitamin A concentrations were significantly lower in RMPP children than those in GMPP patients. Insufficient serum vitamin A concentration was independently associated with RMPP incidence.
The overall incidence of RMPP in MPP patients was 16.02% in this study, which was similar with previous studies [
24,
25]. However, we found the prevalence of VAD was 35.92% in this study, which was relatively lower than others’ reports [
23,
26]. As serum vitamin A levels could be affected by infection [
13], using vitamin A without adjusted by APPs could overestimate the prevalence of VAD in MPP children. The adjustments estimated VAD by mathematically removing or reducing the effect of elevated CRP in this study, which is important for decisions regarding nutrition interventions, programs, and policies.
Another important finding of our study is that sufficient serum vitamin A served as an independently protective factor for RMPP, every one unit decrease of adjusted-vitamin A (mg/L) resulted in 0.205 odds increase in RMPP incidence. Vitamin A is essential for the airway epithelium integrity [
27], lung immune function and inflammation regulation [
28], VAD may result in impaired mucosal barrier [
29], disordered immune response [
29,
30] and excessive cytokines release [
31]. As we known,
M. pneumoniae adhere to the host airway epithelium during MPP, followed by local airway epithelium damage and inflammatory cytokines release. Therefore, the decreased vitamin A during
M. pneumoniae infection could deteriorate pulmonary injuries and clinical manifestations [
32], which contributes to RMPP development together with longer fever duration and LOS. In malnourished children, vitamin A supplementation showed beneficial effects in acute lower respiratory infection (ALRI) children [
23,
33,
34], those are evidences indicated the protective role of vitamin A in RMPP development. However, some studies indicated vitamin A supplementation in ALRI children had no benefits or modestly adverse effect in well-nourished children, which demonstrated the importance of accurate estimation of vitamin A status in planning and implementing interventions.
We also documented that one unit increase of CRP (mg/L) resulted in 0.05 odds increase in RMPP incidence, the median CRP concentrations in RMPP children were significantly higher than those in GMPP patients, which was in line with other studies [
35,
36]. CRP is wildly known as a kind of acute phase protein, which rises rapidly and acutely in response to an inflammatory stimulus and reflect the individual immune response, translating into unfavorable conditions such as RMPP development. Meanwhile, prealbumin was found to be significantly lower in RMPP than GMPP patients, which correlated with RMPP incidence. Prealbumin is a carrier protein synthesized in the liver, it serves as an nonspecific host defense substance by eliminating toxic metabolites during infection [
37]. Hrnciarikova [
38] et al. found it negatively correlated with CRP and could serve as a negative acute phase protein, suggesting the reduction of prealbumin has similar significance with the increase of CRP in RMPP development.
In addition, we found LDH was significantly higher in RMPP children, univariate regression test also found its correlation with RMPP incidence [
35,
39,
40]. It was confirmed that LDH elevated in many kinds of pulmonary diseases and reported to be associated with RMPP. LDH is released from cells after cell damage and can be used to monitor cell and tissue damage. Lung parenchymal cells, local inflammatory cells, including alveolar macrophages and neutrophils might be potential sources of LDH in pulmonary disorders [
41,
42]. Thus, the higher level of LDH could translated into excessive inflammatory cell infiltration and severe lung injury, indicating increased RMPP incidence.
Besides, there were trends for correlations with ferritin and TNF-α, which agreed with other’ studies [
43,
44]. In the linear correlation test, we found a significantly negative correlation between IL-6 and vitamin A.
M. pneumoniae infections are closely related to stimulation of macrophages via toll-like receptors that release immunomodulatory and inflammatory cytokines and chemokines [
33]. Ferritin is a kind of non-specific marker of inflammation induced by activated macrophages, which could also produce TNF-α [
45] and interplay with IL-6 [
46]. Thus, the increased level of ferritin and cytokines can reflect excessive inflammation and RMPP development. However, no significance was found between TNF-α and vitamin A in linear correlation analysis, this may relate to the small sample of bronchoalveolar lavage, which could underestimate the correlation between vitamin A and BALF cytokines.
There are potential limitations of this study. First, this is a single-center study, the data were collected from one academic teaching hospital in China, the results may not easily extrapolate to patients admitted to other regions. Second, the relatively small sample size of our study may reduce the ability to determine the statistical significance of the variables. Third, as the data were collected from the records retrospectively, some information was unfortunately missed, which may lead to imbalanced group sample size. A larger prospective study could help to evaluate the role of vitamin A for RMPP in different age groups, geographical locations.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.