The kinetic profiles of copeptin and mid regional proadrenomedullin (MR-proADM) in pediatric lower respiratory tract infections.

BACKGROUND: Kinetics of copeptin and mid regional proadrenomedullin (MR-proADM) during febrile pediatric lower respiratory tract infections (LRTI) are unknown. We aimed to analyze kinetic profiles of copeptin and MR-proADM and the impact of clinical and laboratory factors on those biomarkers.
METHODS: This is a retrospective post-hoc analysis of a randomized controlled trial, evaluating procalcitonin guidance for antibiotic treatment of LRTI (ProPAED-study). In 175 pediatric patients presenting to the emergency department plasma copeptin and MR-proADM concentrations were determined on day 1, 3, and 5. Their association with clinical characteristics and other inflammatory biomarkers were tested by non-linear mixed effect modelling.
RESULTS: Median copeptin and MR-proADM values were elevated on day 1 and decreased during on day 3 and 5 (-26%; -34%, respectively). The initial concentrations of MR-proADM at inclusion were higher in patients receiving antibiotics intravenously compared to oral administration (difference 0.62 pmol/L, 95%CI 0.44;1.42, p<0.001). Intensive care unit (ICU) admission was associated with a daily increase of MR-proADM (increase/day 1.03 pmol/L, 95%CI 0.43;1.50, p<0.001). Positive blood culture in patients with antibiotic treatment and negative results on nasopharyngeal aspirates, or negative blood culture were associated with a decreasing MR-proADM (decrease/day -0.85 pmol/L, 95%CI -0.45;-1.44), p<0.001).
CONCLUSION: Elevated MR-proADM and increases thereof were associated with ICU admission suggesting the potential as a prognostic factor for severe pediatric LRTI. MR-proADM might only bear limited value for decision making on stopping antibiotics due to its slow decrease. Copeptin had no added value in our setting.

Introduction

Lower respiratory tract infection (LRTI) is a leading cause of morbidity and mortality in children and adolescents worldwide [1, 2]. It may lead to severe complications like pleural effusions, empyema or necrotizing pneumonia [3, 4]. While some patients with LRTI benefit from antibiotic treatment, it is estimated that up to 90% are unnecessarily treated [5-7]. Biomarkers could help to rule-out LRTI requiring antibiotic treatment [8-10], but sensitivity and positive predictive value for ruling-in the need for antibiotic treatment are still low [11-13]. Ongoing research on new biomarkers with better test accuracy has generated promising results for two candidate markers, copeptin (CT-proAVP) and mid regional proadrenomedullin (MR-proADM). Copeptin is a 39-amino acid glycopeptide representing the C-terminal part of the vasopressin precursor molecule pre-provasopressin. It is secreted at equimolar concentrations from the neurohypophysis together with vasopressin and neurophysin II upon stimulation [14]. First studies on copeptin in pediatric LRTI demonstrated that copeptin might be used to diagnose community-acquired pneumonia (CAP) and to predict development of complications, but the results of these studies were contradictory [15-19] and kinetics over time have not been assessed yet. MR-proADM is a 48-amino acid fragment split from the adrenomedullin precursor pre-pro-adrenomedullin in a 1:1 ratio. MR-proADM reflects the adrenomedullin translation and was studied in pediatric LRTI in recent years in the context of assessing infectious disease severity [18, 20, 21]. Kinetic profiles on MR-proADM have been found to add prognostic value for predicting adverse outcome in adult LRTI [22], but have not been evaluated in children so far. Therefore, we aimed at (i) determining the kinetic profiles of copeptin and MR-proADM over five days in children and adolescents admitted to the emergency department for suspected LRTI, and (ii) investigating the influence of clinical and laboratory covariates as well as the development of LRTI complications on copeptin and MR-proADM course over time in a retrospective subanalysis of the ProPAED trial [9].

Methods

Study design and population

This study is a protocol-specified post-hoc analysis of a randomized controlled trial, evaluating PCT as a biomarker guiding antibiotic treatment of LRTI in children and adolescents [9]. Patients (1 month– 18 years) presenting with febrile LRTI to emergency departments of two Swiss pediatric hospitals (Basel and Aarau) between January 2009 and February 2010 were eligible. LRTI was defined by the presence of fever (≥38°C measured in the hospital or at home), and at least one of the following symptoms: tachypnea, dyspnea, wheezing, late inspiratory crackles, bronchial breathing, and/or pleural rub. Patients with severe primary or secondary immunodeficiency, immunosuppressive treatment, neutropenia (<1 G/L), cystic fibrosis, upper respiratory tract infection or hospital stay within previous 14 days were excluded. In the ProPAED trial patients were randomized to a PCT guided intervention group or to a standard care control group with treatment based on international guidelines [9, 23].

Variables

The following variables were prospectively recorded at inclusion or prior to these ancillary analyses as appropriate: age, days of fever before presentation, sex, antibiotic treatment, antibiotic pre-treatment before study inclusion, diagnosis, complications (parapneumonic effusion, empyema, acute respiratory distress syndrome (ARDS), sepsis, shock), body temperature, heart rate, breath rate, oxygen administration, hospitalization, admission to intensive care unit (ICU). C-reactive protein (CRP), PCT and cytokine (interferon (IFN)-γ, interleukin (IL)-1ra, IL-1β, IL-2, IL-4, IL-6, IL-10, IFN-γ-inducible protein (IP)-10 and tumor necrosis factor (TNF)-α) concentrations were recorded prospectively on day 1, 3 and 5 after inclusion [9, 24]. Blood cultures (BC) and nasopharyngeal aspirates (NPA) were performed at physician’s discretion. Microbiology results were classified in 4 categories: not performed, negative, systemic bacterial infection (blood culture positive for pneumococcus/streptococcus = invasive pneumococcal infection) and other. Copeptin and MR-proADM were measured for the present analysis, if the stored EDTA plasma sample volumes remaining from the ProPAED trial were large enough for three post-hoc biomarker measurements (day 1–3–5). Heart rate was classified as normal or elevated according to age-dependent references ranges, respiratory rate was classified as elevated if exceeding WHO age-dependent reference ranges for children up to 5 years of age, and if exceeding percentile 90 as reported by Fleming et al. for children >5 years of age (S1 Table) [25, 26].

Assays

EDTA plasma collected on day 1, 3 and 5 after inclusion was stored at– 80°C. Copeptin and MR-proADM levels were measured using the commercial sandwich immunoluminometric assays B·R·A·H·M·S™ Copeptin proAVP KRYPTOR™ and B·R·A·H·M·S™ MR-proADM KRYPTOR™ (B·R·A·H·M·S GmbH, part of Thermo Fisher Scientific, Henningsdorf, Germany). The lower limit of quantitation of the assays was 1.23 pmol/L and 0.23 nmol/L, the upper limit of quantitation was 2000 pmol/L and 100 nmol/L for copeptin and MR-proADM, respectively. For MR-proADM, the manufacturer reported for the values between 0.2–0.5 nmol/L and between 0.5–6 nmol/L intra assay precisions of ≤10% and <2–4% and inter assay precisions of ≤18% and <6–11%, respectively. For copeptin, the manufacturer reported for the values between 2–4 pmol/L and between 4–50 pmol/L intra assay precisions of ≤15% and <4–8% and inter assay precisions of ≤18% and <6–10%, respectively. According to maximum inter-assay coefficient of variation (CV%), an increase was set as significant of at least 18% and 20% for copeptin and MR-proADM concentrations, respectively.

Statistical analysis

Population characteristics and clinical findings were described by standard descriptive statistics (median and interquartile range (IQR) for continuous variables, count and proportion for categorical variables).

Non-linear mixed effect modelling

Copeptin and MR-proADM kinetics and their association with clinical parameters, laboratory variables, patient management, microbiology, and chest radiography were first investigated by visual inspection. Characterization of copeptin and MR-proADM kinetics, and quantification of inter- and intra-individual variability was performed using a nonlinear mixed effect modeling approach with NONMEM® (version 7.1.3, ICON Development Solutions, Ellicott City, MD, USA). Fitting was performed for each biomarker (copeptin and MR-proADM) separately. Model estimation was performed using the first-order conditional estimation with interaction (FOCE-I / LAPACE). Details on nonlinear effect modeling, parameter estimation, model selection and model evaluation are available in S1 Appendix.

Influence of patient characteristics

Association with clinical parameters, laboratory variables, patient management, microbiology and, chest radiography, called covariates, were investigated on model parameters on which variability was identified. The stepwise covariate model building approach is explained in S1 Appendix.

Correlations between biomarkers

The correlation between copeptin and MR-pro-ADM concentrations on patients`admission was investigated by Pearson correlation test to evaluate the strength of relationship. The correlation between copeptin, MR-proADM, C-reactive protein (CRP), procalcitonin (PCT), and various cytokines (interferon (IFN)-γ, interleukin (IL)-1ra, IL-1β, IL-2, IL-4, IL-6, IL-10, IFN-γ-inducible protein (IP)-10, and tumor necrosis factor (TNF)-α) was investigated on log-transformed variables. Graphical inspection, descriptive statistics and Pearson correlation test were performed with R software (version 3.1.2; R Development Core Team, Vienna, Austria, http://www.r-project.org).

Ethics

Both pertinent ethics committees of the University of Basel and Kanton Aargau (Switzerland) approved the trial protocol. Written informed consent was obtained from all patients or their care takers. The trial was registered with the International Standard Randomized Controlled Trial Number Register (ISRCTN 17057980).

Results

Study population and clinical characteristics

The study population comprised 175 febrile LRTI pediatric patients (age: 1 month-18 years) for whom a sufficient quantity of blood plasma was in storage for copeptin and MR-proADM analysis for day 1, 3, and 5 after study inclusion. Blood sampling was performed in-house while patients were hospitalized or in the emergency department, to where the outpatients and the discharged patients were asked to return on day 3 and 5. Patients’ baseline characteristics are summarized in Table 1.

Table 1

Characteristics of the population included for present analyses and the ProPAED study population.

	Present analysis (n = 175)	ProPAED study (n = 337)
Demographic
Age, Median (IQR), years	4.1 (1.9–6.6)	2.8 (1.2–5.3)
<1 yr, n (%)	16 (9)	71 (21)
1–5 yr, n (%)	87 (50)	171 (51)
5–18 yr, n (%)	72 (41)	95 (28)
Male, n (%), gender	98 (56)	196 (58)
Clinical features at inclusion (day 1)
Antibiotic pre-treatment, n (%)	26 (15)	42 (12)
Days of fever before presentation, median (IQR)	3 (1–4)	3 (1–4)
Fever, n (%)	175 (100)	337 (100)
Day 1, n (%)	132 (75)	247 (73)
Day 3, n (%)	14 (8)	34 (10)
Day 5, n (%)	4 (2)	11 (3)
Fever before inclusion >1 day, n (%)	122 (70)	230 (68)
Cough, n (%)	174 (99)	336 (99)
Sputum production, n (%)	73 (42)	141 (42)
Poor feeding, n (%)	63 (36)	153 (45)
Pleuritic pain, n (%)	58 (33)	95 (28)
Clinical findings at inclusion (day 1)
Body temperature, median (IQR), °C	38.5 (38–39.1)	38.4 (37.9–39.1)
Respiratory rate, median (IQR), /min	38 (28–44)	40.0 (28.0–48.0)
Elevated breath rate*
Day 1, n (%)	52 (30)	103 (31)
Day 3, n (%)	5 (3)	21 (6)
Day 5, n (%)	8 (5)	16 (5)
Heart rate, median (IQR), b/min	136.0 (120.0–159.0)	142 (123–160)
Day 1, n (%)	143 (82)	269 (80)
Day 3, n (%)	70 (40)	120 (36)
Day 5, n (%)	66 (38)	103 (31)
Dyspnea, n (%)	96 (55)	217 (64)
Wheezing, n (%)	47 (27)	101 (30)
Late inspiratory crackles, n (%)	62 (35)	140 (41)
Reduced breathing sounds, n (%)	60 (34)	109 (32)
Laboratory findings at inclusion (day 1)
Procalcitonin, median (IQR), μg/L	0.28 (0.13–2.50)	0.24 (0.13–1.48)
Procalcitonin> 1 μg/L, n (%)	61 (35)	99 (29)
C-reactive protein, median (IQR) mg/L	26.05 (10.25–93.38)	20 (8–65)
C-reactive protein > 100 mg/L, n (%)	40 (23)	67 (20)
Antibiotic prescription and admission
Antibiotics within 14 days following randomization, n (%)	105 (60)	197 (58)
Antibiotics (AB)
iv, n (%)	50 (29)	83 (25)
oral, n (%)	55 (31)	114 (34)
no AB, n (%)	70 (40)	140 (41)
Hospitalization, n (%)	89 (51)	204 (60)
Admission to intensive care unit, n (%)	4 (3)	9 (3)
Supplemental oxygen, n (%)	22 (12)	79 (23)
Complications†, n (%)	5 (3)	6 (2)
Microbiology
Nasopharyngeal aspirate (NPA), n (%)	168 (96)	318 (94)
M. pneumoniae or C. pneumonia, n (%)	13 (8)	15 (4)
human Metapneumovirus, n (%)	17 (10)	42 (12)
Influenza virus, n (%)	20 (12)	36 (11)
Respiratory syncytial virus, n (%)	26 (15)	74 (22)
Blood culture, n (%)	87 (50)	148 (44)
positive for S. pneumonia or S. pyogenes, n (%)	5 (3)	6(2)
no blood culture and NPA negative, n (%)	38 (21)	69 (20)
blood culture negative and NPA negative	16 (9)	32 (9)
Diagnosis
Pneumonia
Day 1, n (%)	82 (47)	129 (38)
Day 3, n (%)	84 (48)	137 (41)
Day 5, n (%)	84 (48)	135 (40)
Bronchitis/-iolitis
Day 1, n (%)	54 (31)	122 (36)
Day 3, n (%)	79 (45)	148 (44)
Day 5, n (%)	82 (47)	165 (49)
Bronchitis/-iolitis + Pneumonia
Day 1, n (%)	39 (22)	86 (26)
Day 3, n (%)	12 (7)	48 (14)
Day 5, n (%)	9 (5)	31 (9)

IQR = interquartile range.

* For age specific reference values for breath and heart rate see S1 Table.

†Complications were defined as sepsis, shock, pleural effusion, pleural empyema or acute respiratory distress syndrome (ARDS). Further details on age-stratified distribution of copeptin and MR-proADM in S1 Fig and in S2 Table. Five patients developed complications of LRTI. All recovered well, there were no deaths (S3 Table).

Kinetics of copeptin and MR-proADM concentrations

Data exploration

Median biomarker concentrations decreased over the 5 study days (Fig 1, for absolute biomarker values S4 Table). Age-stratification of copeptin and MR-proADM did not reveal any differences (S1 Fig and S2 Table). One patient (<1%) had increasing MR-proADM but not increasing copeptin concentrations and 11 patients (6%) had increasing copeptin concentrations over the 5 study days.

Fig 1

Copeptin and MR-proADM concentrations over the study period.

Distribution and change in copeptin (pmol/L) and MR-proADM (nmol/L) concentrations for patients over 5 study days. Boxes represent the interquartile range (IQR). Solid lines are the median, 25th and 75th quantile and whiskers equal 25th quantile -1.5 IQR and 75th quantile +1.5 IQR.

Influence of patient characteristics on copeptin kinetics

The search for factors influencing the course of copeptin over time in the covariate model did not reveal any parameter to have a significant impact on copeptin plasma concentrations over time or on its kinetics. For details see S1 Appendix.

Influence of patient characteristics on MR-proADM kinetics

During forward covariate model building, antibiotic administration route, general complication, microbiology results and presence of fever had a significant influence on MR-proADM concentrations on day 1 (p<0.001). Admission to ICU was associated to an increase of MR-proADM and microbiology results were associated with a decrease in MR-proADM. During the backward process, association with fever became not significant and was thus excluded. Details on base and final model can be seen in S1 Appendix. In the final model, intravenous administration of antibiotics influenced MR-proADM concentration and was significantly associated with a higher MR-proADM concentration on day 1. Further, ICU admission was significantly associated with an increased MR-proADM concentration. In contrast, positive blood cultures, which were grouped together with negative NPA or no growth in blood culture because of the same effect when building the model, led to a significant decrease of MR-proADM (Table 2).

Table 2

MR-proADM: Variables effect on MR-proADM kinetics in multivariable analysis as compared to the typical patient.

	Definition of variable effect	Multivariable effect [95%CI]	p-value
MR-proADM concentration on day 1	Administration of IV antibiotics	+ 61% [39, 90]	<0.001
MR-proADM decrease over 5 days	ICU-admission	+ 107% [43, 150]	<0.001
	Positive blood culture, negative NPA or no growth in blood culture	- 85% [–45, –144]	<0.001

IV: intravenous; ICU: intensive care unit; NPA: nasopharyngeal aspirates.

Correlation between copeptin and MR-proADM concentrations and pro- and anti-inflammatory markers

There was only a moderate positive correlation between copeptin and MR-proADM concentrations with inflammatory biomarkers, most noteworthy being IL-6 (r = 0.42 and 0.47 respectively). MR-proADM and copeptin concentrations were not correlated to each other (r = 0.18). All correlations between copeptin, MR-proADM, CRP, PCT and various cytokines (interferon (IFN)-γ, interleukin (IL)-1ra, IL-1β, IL-2, IL-4, IL-6, IL-10, IFN-γ-inducible protein (IP)-10 and tumor necrosis factor (TNF)-α) on first admission to the emergency department (day 1) are presented in Fig 2.

Fig 2

Correlation between copeptin and MR-proADM concentrations and pro-and anti-inflammatory markers at study inclusion.

IL: interleukin; TNF: tumor necrosis factor; INF: interferon; IP-10: interferon-gamma induced protein 10 kD, CRP: c-reactive protein; PCT: procalcitonin; MR-proADM: mid regional proadrenomedullin.

Discussion

This is the first study that investigated the kinetics of copeptin and MR-proADM in pediatric LRTI. Our main aim was to determine the dynamics of these biomarkers over the study period of five days. Our additional aim was to test for impact of clinical and laboratory markers that might indicate severity of disease on the biomarkers kinetics. Our results suggest, that MR-proADM but not copeptin might be helpful in differentiating severe from not severe pediatric LRTI cases.

In the recent past biomarker guided antibiotic therapy of LRTI was one research focus in infectious disease [9, 27]. Copeptin and MR-proADM have shown to discriminate between mild and severe LRTI in children if measured on first presentation to an emergency department [16, 20]. Assuming that children with severe LRTI are more likely to have bacterial infection in need of antibiotic treatment, copeptin and MR-proADM could be used to start or withhold empiric antibiotic treatment. For monitoring during the course of disease and for the indication to stop antibiotics secondarily the course of these markers has to be known.

In this study the initial copeptin median was 6.3 pmol/L. To date, no internationally accepted normal copeptin values exist, but published values for healthy individuals are in the range 2–10 pmol/l [14, 28–30]. Our result could be seen as to be within the normal range and thus not being elevated, but because we observed a gradual decrease over time (-26%), it is suggestive that the median copeptin values found in our study were at least slightly elevated compared to the assumed baseline values in healthy children. The dynamics were not as anticipated, as we have seen in a cytokine analysis of the ProPAED cohort that pro- and anti-inflammatory cytokines decreased very rapidly over the study period [24], so a steeper copeptin decrease was expected. Especially, if the regulatory swiftness is taken into account, with which vasopressin and copeptin plasma values change when regulating body water homeostasis: copeptin decreases with a half-life of only 26 minutes after oral water challenge [14, 28, 31, 32]. Copeptin is actively released via the activation of the endocrine stress axis at equimolar concentrations from the posterior pituitary into circulation in situations with an increased plasma osmolality, hypovolemia and high individual stress level. The stress axis activation results in release of vasopressin, adrenocorticotropic hormone and cortisol [29, 32, 33]. In this context dehydration as a potential source of bias has to be taken into account. Hydration assessment of the included patients was not mandatory and as no blood gas analyses or metabolic assessments were recorded, a lack of adequate fluid intake cannot be excluded as possible confounder. However, an inflammation driven persistent and gradually decreasing stress level was more likely in our cohort than dehydration and we assume that the copeptin course in our study rather reflected the stress reaction of the body to the inflammatory state than body water homeostasis disturbance.

Besides kinetics, associations with potential markers of disease severity were of interest for this study but for copeptin we found none. Several studies have shown copeptin to be related to severity and complications of LRTI. Du et al. investigated plasma copeptin levels in 265 children and found median levels of 73.0 pmol/L in complicated pneumonia in 2013 [15]. This association of copeptin with complicated LRTI was later confirmed by two further studies [16, 19]. In our case we assume that the number of patients with LRTI complications (n = 5, S3 Table) was just too low to create significant influence in the model. Therefore, we cannot draw a rational conclusion on copeptin as marker of disease severity. Gender has been described as potential factor influencing copeptin levels [14, 34], but in our cohort, we did not find any gender disparity. Most likely, because the inflammatory state overruled a possible gender disparity in unstressed individuals.

The initial median MR-proADM concentrations in the present cohort were higher than the median values for healthy children reported by Michels et al. (0.22 nmol/L) and Hauser et al. (0.34 nmol/L) [35, 36], suggesting a possible role for identifying individuals with more severe LRTI and thus need of antibiotics. In contrast to copeptin, adrenomedullin is actively involved as a hormokine in inflammatory reactions and triggers cytokine release [37, 38], but is also triggered by cytokines themselves, most importantly by IL-1β and TNF [39]. However, the MR-proADM values stayed elevated longer than the cytokines measured in the ProPAED cohort and did consecutively not correlate with them [24]. When analyzing the relationship between cytokines and markers as copeptin and MR-proADM in the present cohort of pediatric LRTI, one has to keep in mind, that the children and adolescents attended the emergency room with a history of fever lasting (median) three days. Therefore, it is most likely, that the initial up-stream inflammatory activation happened, at least in part, before the first measurements of cytokines and inflammatory biomarkers were performed for the ProPAED study.

In recent studies MR-proADM was able to predict LRTI complications, such as pleural effusions and bacteremia [18, 20, 21]. In the present study IV administration of antibiotics was associated with higher MR-proADM concentrations on day 1, and ICU admission was associated with a slow increase in MR-proADM, indicating a persistent inflammatory state. Although IV administration of antibiotics and ICU admission can very well be triggered by other factors than severe invasive bacterial disease, these two markers can be seen as surrogate markers for disease severity. Our findings are in line with previous studies. Alcoba et al. found MR-proADM to be associated with bacteremia and empyema [18], and Sardà Sánchez et al. found MR-proADM related to pneumonia with pleural effusion (no information on empyema provided), but not to ICU admission, as in our case. However, in that study, only hospitalized patients were analyzed and in our study we included in- and outpatients, therefore our cohort might have been less sick than the cohort of Sardà Sánchez et al. [18, 21]. Further, the most recent study by Florin et al. [20], evaluating also in- and outpatients, was in line with our and former results: the association of MR-proADM with higher severity of disease was significant. In that study ICU admission was also significantly associated with higher MR-proADM values. Last, blood culture positivity was associated with a more distinct MR-proADM decrease in our study. This is most likely due to the effective antibacterial treatment of streptococci, the only pathogen found in blood cultures in our study, being highly susceptible to empirical treatment.

Limitations

The present study was performed in Switzerland, a highly developed country with an excellent medical safety netting, in two emergency departments of pediatric tertiary centers where generally the rate of complicated medical illness is low. The rate of complications from LRTI here was low and no strong correlation of complications to copeptin or MR-proADM concentrations could be established. In a different setting (e.g. different medical system or inclusion only of hospitalized children) with a higher prevalence of complications the result might be different and a stronger impact of LRTI complications on copeptin and MR-proADM kinetics could result. Another limitation is the lack of a healthy pediatric control group that would have been necessary to establish comparable pediatric normal values. In our case, we could only refer to previously published values in adults and children. Chest radiographs were not mandatory in the ProPAED study and only the more severe cases underwent chest radiographs. To avoid selection bias we did not perform any diagnostic accuracy analysis for the prediction of chest x-ray confirmed pneumonia or other types of radiographically diagnosed LRTI.

In summary, this study evaluated kinetics of copeptin and MR-proADM and potential factors influencing the course over time. Our study supports the potential value of MR-proADM plasma concentrations as a prognostic factor for severe pediatric LRTI. Plasma copeptin concentration had no added value in our setting. These results call for larger studies confirming the clinical use of MR-proADM in pediatric LRTI and studies on copeptin in settings with more serious febrile LRTI.

(DOCX)

Click here for additional data file.

Age group and biomarkers.

Age-stratification and change in MR-proADM (nmol/L), and copeptin (pmol/L) concentrations over the study days 1, 3, and 5.

(DOCX)

Click here for additional data file.

Cut-off values to define normal or elevated breath and heart rates according to patient’s age.

Classification as normal or elevated rate were set according Fleming S, Thompson M, Stevens R et al. Normal ranges of heart rate and respiratory rate in children from birth to 18 years of age: a systematic review of observational studies. The Lancet. 2011; 377(9770):1011±8 and WHO.

(DOCX)

Click here for additional data file.

Age-stratification and change in MR-proADM (nmol/L), and copeptin (pmol/L) concentrations over the study days 1, 3, and 5. SD: Standard deviation.

(DOCX)

Click here for additional data file.

Type and day of complication experienced by five patients during the study period of 5 days.

ICU: intensive care unit; ARDS: acute respiratory distress syndrome.

(DOCX)

Click here for additional data file.

Copeptin, and MR-proADM, concentration profiles and relative change between study days.

MR-proADM: pro-adrenomedullin; IQR: inter-quartile range.

(DOCX)

Click here for additional data file.

11 Nov 2020

PONE-D-20-08646

Copeptin and mid regional proadrenomedullin (MR-proADM) in pediatric lower respiratory tract infections

PLOS ONE

Dear Dr. Baumann,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Please submit your revised manuscript by Dec 25 2020 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

We look forward to receiving your revised manuscript.

Kind regards,

Julia Robinson

Senior Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. One of the noted authors is a group or consortium [ProPAED study group]. In addition to naming the author group and listing the individual authors and affiliations within this group in the acknowledgments section of your manuscript, please also indicate clearly a lead author for this group along with a contact email address.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Partly

Reviewer #2: Yes

Reviewer #3: Partly

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: This is a well conducted and reported study. Nevertheless I have some comments and suggestions about how the manuscript could be further improved.

In the abstract, statements about differences in concentrations of MR-proADM in lines 43 to 45 should include the estimated difference and 95% confidence interval, not just a p-value. Additionally, the statements about association of copeptin with clinical characteristics or complications should be unambiguous, and terms such as "relevant association" and "correlated moderately" should be avoided.

The description of the non-linear mixed model in the methods has been done well. The descriptions in line 172 to 179 is unclear; 'parameter P' is mentioned without any introduction as to what it represents, and it is unclear what 'COV' means. Additionally there is a typo on line 172 ('median' not 'mdeian'?). Additionally, it is unclear what kind of models were used to address the classification question of predicting antibiotic treatment (lines 180 to 187). More details should be included here.

Most importantly, I think the study would benefit from a better approach to selecting the covariates to be included in the models for the various outcomes. This selection should consider whether the covariates would plausibly be related to the outcomes. For example it doesn't seem plausible to explore association between administration of intravenous antibiotics and baseline levels of the biomarkers, because the baseline implies before treatment i.e. pre-antibiotics, pre-blood culture, etc.

Lastly, the reporting of the results should focus on explaining what the various parameters of the non-linear model mean, rather than just citing their numerical values.

Reviewer #2: The authors of this manuscript reported the possible roles of two novel biomarkers (copeptin and MR-proADM) in predicting disease severity in pediatric lower respiratory infections (LRTIs). It was a post-hoc analysis of a randomized controlled trial evaluating procalcitonin (PCT) as a biomarker guiding antibiotic treatment in children and adolescents with LRTIs in the ProPAED trial. Biomarker research aimed at improving disease diagnosis and prognostication remains a relevant issue in scientific literature. Although the authors gave an in-depth report of their study which made a case for MR-proADM alone or in combination with PCT, cytokines or C-reactive protein (CRP) in predicting disease severity in LRTIs, there are few specific concerns that I request them to address to enhance the manuscript.

Reviewer #3: This study is a subanalysis of a prior RCT of children with LRTI that focuses on biomarker kinetics in pediatric LRTI. The strength of this manuscript lies in its novelty – the kinetics of most of these biomarkers in pediatric CAP has not been reported previously. There are several challenges to the approach and presentation of data in this manuscript that make the conclusions difficult to understand.

MAJOR ISSUES

1. This is a heterogeneous population – children with bacterial pneumonia require antibiotics and children with bronchiolitis do not. Therefore, to include use of IV antibiotics as one of the primary outcomes is not appropriate for half of the already small study population. This study requires stratifying data into those with and without pneumonia

2. The story is somewhat confusing as written – is this about severe outcomes? Use of antibiotics? Only copeptin and proADM or really about 4 biomarkers? I would recommend that this focus on kinetics of CRP, PCT, copeptin and proADM. The title suggests that it is only copeptin and proADM but there is important data presented on other biomarkers as well that should be accounted for in the title and results sections.

3. Results and tables require clarity on the meaning of the statistical parameters. The message is lost because of a focus on statistical rather than clinical terminology throughout.

ABSTRACT

1. I’d recommend that the results section include more numerical results. It would be more meaningful for readers to see effect estimates with confidence intervals rather than simply p-values.

2. If the conclusion is that there is additional clinical value using combinations of biomarkers, I would recommend giving some numerical results of the correlations in the results.

INTRODUCTION:

1. Would be more specific in the citation placement of references 1-6. The individual references should be placed after the prior sentences, as they all do not support the sentence that up to 90% are unnecessarily prescribed antibiotics. Would consider adding the following reference to support this statement: Florin TA et al. JPIDS 2020;9(2).

2. An important reference that should be included in Lines 55-57 is: Stockmann C et al. JPIDS 2018;7(1).

3. A more recent reference regarding MR-proADM and disease severity that should be included in the introduction is: Florin TA et al. Clin Infect Dis 2020 Aug 6; ciaa1138

METHODS

1. This data is now more than 10 years old and occurred during the H1N1 influenza pandemic. Can the authors comment on the age of the data and the possible influence of the pandemic on the results?

2. Line 109 – was pneumococcus the only bacterium considered for systemic bacterial infection? What about other known bacterial pathogens of pneumonia?

3. Line 114 – it is not clear what age-dependent reference standard is used for vital signs with the exception of respiratory rate. This is not laid out in the text or in Table S1.

4. Line 116 – the Fleming percentiles were derived from healthy children in the outpatient setting, thus I’m not certain they are appropriate in this sick hospitalized population. Others (Bonafide, Daymont, Oostenbrink, Thompson) have published vital sign curves for ED and hospitalized children which would be more appropriate to use.

5. Lines 145-179 – the section on modeling is not intuitive or easily understood by the average medical reader. I recommend that this entire section be written to summarize the statistical methods in prose that allows the average clinician reading this paper to understand what was performed in the body of the main manuscript. The details of the modeling can be expounded on in S1 Appendix.

6. Lines 166-169 – can the authors justify their use of a stepwise approach to covariate selection, rather than an approach based on biological and clinical plausibility?

7. Lines 182-183 – it appears that discriminating IV antibiotics vs oral or no antibiotics is a primary outcome. Were there standardized criteria for who received IV antibiotics? This is critical, as this otherwise becomes reflective more of an individual clinician’s gestalt or the fact that an IV line was present rather than a true need for IV antibiotics. This is what makes use of IV antibiotics a challenging primary outcome.

RESULTS

1. Line 207 – inclusion criteria makes it sound as if only those who had all measurements on day 1, 3, and 5 were included. If this is the case, I am concerned about selection bias for this population, as most children with LRTI who are hospitalized stay in the hospital for fewer than 5 days (median ~2-2.5 days per literature).

2. Line 215 – there are important data that are buried in Table S1. This table is labeled as a list of variables or data dictionary, when in fact the last column is important data to describe the study population and outcomes. This data should be in the main body of the paper.

3. Line 224 - Less than half the children in this study had pneumonia. The pathophysiology and microbiology of different lower respiratory tract infections are different, and therefore I’m concerned that this is a highly heterogeneous population (with a relatively small n, in addition). Children with bronchiolitis shouldn’t be requiring antibiotics to begin with and therefore it is not appropriate to include them with children with pneumonia in a primary outcome that includes IV antibiotic use.

4. Lines 246-252 – this section needs to be edited to explain the results in prose rather than relying on statistical parameters. The meaning of these results is lost in statistical jargon.

5. Lines 264-266 – same as above – rather than talking about ‘baseline parameter’ and ‘slope parameter,’ this would be easier to read and interpret if the text read: “the following variables were statistically associated with initial levels of MR-proADM…ICU admission and microbiology results were associated with the change in MR-proADM from baseline to Day 5.” (if I’m interpreting the statistical jargon appropriately)

6. Table 2 and 3 – it would be helpful to have footnotes explaining the parameters in this table. What is meant by ‘limit?’

7. Line 302 – predict is the wrong term here. The biomarker is not doing the predicting – I would stick with MR-proADM discriminated those children receiving IV antibiotics vs those who did not. If less than half of the children in this study had pneumonia and hospitalized children with pneumonia generally warrant antibiotics, aren’t these results really a proxy for discriminating pneumonia from no pneumonia? There should be an examination of the interaction of diagnosis and IV antibiotics in these analyses.

8. Figure 2 – requires labels

DISCUSSION

1. Line 333 – the discussion states that decreases were not as steep for CRP and PCT, yet the declines observed in Fig S1 do suggest that there were in fact declines in CRP and PCT. If this statement is going to be made, I would recommend some presentation of data in the Results section. In fact, I think that this manuscript would benefit from being reframed as biomarker kinetics in general, including CRP, PCT, copeptin and proADM, as that is really what this manuscript is about – not just copeptin and proADM.

2. Line 385 – include some discussion of Florin TA et al. Clin Infect Dis 2020 Aug 6; ciaa1138

3. Line 400-401 – just because the correlations were moderate here (some were not very good at all), one cannot conclude that combinations of biomarkers may be helpful. It would have been helpful if there was statistical analysis that focused on the outcomes of interest with combinations of biomarkers to be able to make this claim rather than simple correlations.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: Yes: Samuel Uwaezuoke

Reviewer #3: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

Submitted filename: Reviewers comments for PLOS ONE JOURNAL.docx

Click here for additional data file.

2 Apr 2021

Dear Julia Robinson,

thank you for considering the publication of our manuscript and the thorough review. Here we answer point-by-point to all aspects raised by the reviewers. Please see the corresponding corrections in the uploaded manuscript which we hope is now suitable for publication.

All line numbers in the revised manuscript correspond to the track change version.

Journal Requirements:

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

We carefully checked our manuscript again. All requirements that are provided in these two pdf-files are met.

2. One of the noted authors is a group or consortium [ProPAED study group]. In addition to naming the author group and listing the individual authors and affiliations within this group in the acknowledgments section of your manuscript, please also indicate clearly a lead author for this group along with a contact email address.

The lead author for this group is Dr. Jan Bonhoeffer, we added his contact details to the Acknowledgments section (lines 545/546oo).

Summary

Before we begin with the point-for-point answers for each reviewer we would like to introduce a short section-for-section summary, as some reviewers requested important major changes that affected the whole direction of the article and are likewise important to all three reviewers. We would like to summarize these changes here and come back to them in detail later.

Abstract and Introduction:

We thank the reviewers for the advice to focus on the clinical relevance of our work and to favor language that is more intuitively understandable also without biostatistical degree. Therefore the abstract was reworded with respect to reviewers` requirements. We trust that the abstract is now appropriate for clinical physicians. We avoided ambiguous terms and present more data in numerical detail.

The introduction was shortened, partly reworded and restructured to emphasize the focus on the clinical aspects and the primary outcome biomarkers` kinetics.

Methods/Results:

Parts of the methods/results sections were outsourced to S1 Appendix, language was adapted to become easier to read. ROC analysis for iv administration of antibiotics was deleted due to the rather vague outcome parameter “administration route”.

Discussion:

The discussion was in large parts reorganized and re-structured according to the requirements of reviewer #2. It is now also significantly shorter.

Conclusion:

The conclusion has been changed according to the modifications made to the article.

Funding: We added a funding part as test kits were provided by BRAHMS free of charge.

Reviewer #1: This is a well conducted and reported study. Nevertheless I have some comments and suggestions about how the manuscript could be further improved.

We are grateful to reviewer # 1 for his comments relevant to the improvement of this manuscript. According to his suggestion, we have changed the following sections.

In the abstract, statements about differences in concentrations of MR-proADM in lines 43 to 45 should include the estimated difference and 95% confidence interval, not just a p-value. Additionally, the statements about association of copeptin with clinical characteristics or complications should be unambiguous, and terms such as "relevant association" and "correlated moderately" should be avoided.

Abstract: The estimated differences and the change per day of the biomarker values have been entered into lines 47-52. Potentially ambiguous statements have been deleted (line 52 - 54).

Additionally there is a typo on line 172 ('median' not 'mdeian'?).

Methods: Typo was corrected, thank you.

Additionally, it is unclear what kind of models were used to address the classification question of predicting antibiotic treatment (lines 180 to 187). More details should be included here.

Methods: For the sake of clarity we deleted the ROC analysis (Fig 2) and the corresponding S6 Table. IV administration is a rather weak outcome parameter as it could be influenced by other factors (iv line in place, need of fluids, etc.) and as it was correctly stated by the reviewers.

Most importantly, I think the study would benefit from a better approach to selecting the covariates to be included in the models for the various outcomes. This selection should consider whether the covariates would plausibly be related to the outcomes. For example it doesn't seem plausible to explore association between administration of intravenous antibiotics and baseline levels of the biomarkers, because the baseline implies before treatment i.e. pre-antibiotics, pre-blood culture, etc.

Methods: For the selection of covariates we had different demographic, clinical and laboratory parameters at hand (age; sex; body temperature; breath rate; heart rate; days of fever before inclusion; complications including sepsis, pleural effusion, pleural empyema, ARDS; hospital or ICU admission; oxygen requirement; form of antibiotic administration and antibiotic pre-treatment; CRP; PCT; nasopharyngeal aspirates; blood culture; and diagnosis given by the attending physician), see Table 1. Our exploratory approach was to include these into the base model to evaluate broadly the impact on MR-proADM and copeptin kinetics. IV-administration stood out (among others) even though a part of the population was pre-treated (15%). A potential biomarker suggesting need or withhold of antibiotic treatment should be discriminative, even though some patients would have been pre-treated already. Therefore we retained this parameter in the final model.

We agree, that this approach is rather broad, but as published data for respective associations with copeptin and MR-proADM kinetics in children is scarce, we still believe, that this is appropriate.

Lastly, the reporting of the results should focus on explaining what the various parameters of the non-linear model mean, rather than just citing their numerical values

Results: In the results section, we would like to state just the numerical results and comment on these in the discussion. However, this also refers to other review comments criticizing the language we used. We changed this throughout the article and feel that also the results section is now easier to understand.

Reviewer #2: General comments

The authors of this manuscript reported the possible roles of two novel biomarkers (copeptin and MR-proADM) in predicting disease severity in pediatric lower respiratory infections (LRTIs). It was a post-hoc analysis of a randomized controlled trial evaluating procalcitonin (PCT) as a biomarker guiding antibiotic treatment in children and adolescents with LRTIs in the ProPAED trial. Biomarker research aimed at improving disease diagnosis and prognostication remains a relevant issue in scientific literature. Although the authors gave an in-depth report of their study which made a case for MR-proADM alone or in combination with PCT, cytokines or C-reactive protein (CRP) in predicting disease severity in LRTIs, there are few specific concerns that I request them to address to enhance the manuscript.

We are grateful to the reviewer #2 for his help to improve this article.

Specific comments

1. Abstract and Title: The full title of your manuscript does not convey the main objective of the study. Curiously, the title is the same as your short title. Based on the aim of your study, I suggest a modification of your full title to read- ‘The kinetic profiles of copeptin and mid-regional proadrenomedullin in pediatric lower respiratory tract infections.’

Title: We changed the title according to his suggestion.

In the background of your abstract, you mentioned that both novel biomarkers may be predictive of complications of pediatric LRTIs. Which complications are you referring to?

Abstract: In fact, we did not explain LRTI-complications in the abstract. We defined complications in the variables section of the methods (parapneumonic effusion, empyema, ARDS, sepsis, shock) in lines 119-120. While rewording with the focus on biomarkers` kinetics and the impact of clinical and laboratory features, we choose to temper the part of predicting complications and omitted this sentence in the final version of the abstract.

Again, the phrase-‘The change over time of these biomarkers during LRTIs’ may better be rendered thus- ‘The kinetics of these biomarkers during LRTIs.’

Abstract: We changed this according to this suggestion (line 31).

The second part of the aim read- ‘…and to investigate the relationship between copeptin, MR-proADM and demographic, clinical and laboratory characteristics of pediatric LRTI.’ I suggest you rephrase this for clarity.

Abstract: We changed this sentence according to his and the overall suggestion to focus more on clinical relevant aspects, it now reads: “We aimed to analyze kinetic profiles of copeptin and MR-proADM and the influence of clinical and laboratory aspects on biomarkers` progression.” (lines 34-37).

2. Introduction: Study aims stated in lines 79-84 appear slightly different from the aims under Abstract. Can you please harmonize them?

We changed the phrases accordingly.

Abstract: “We aimed to analyze kinetic profiles of copeptin and MR-proADM and the impact of clinical and laboratory factors on biomarkers` progression.”

Introduction: “Therefore, we aimed at (i) determining the kinetic profiles of copeptin and MR-proADM over five days in children and adolescents admitted to the emergency department for suspected LRTI, and (ii) investigating the influence of clinical and laboratory covariates as well as the development of LRTI complications on copeptin and MR-proADM course over time in a retrospective subanalysis of the ProPAED trial.”

3. Methods: Every study methodology should have clarity and easy reproducibility as key features. Why did you adopt a post-hoc analysis of an RCT in your study? Post-hoc analysis conducted and interpreted without sufficient consideration of the ‘multiple testing problem’ is criticized by some scholars who believe the statistical associations are often spurious. In fact, you to admitted this methodological flaw of using a retrospective approach in one of your listed study limitations (lines 428-430)

Methods: We totally agree on this comment. We acknowledge this possible flaw, but can state, that our main aim was to demonstrate the kinetics of Pro-ADM and copeptin over 5 days in a well-defined population. Comparable data was missing so far. Besides the plain demonstration of the kinetics we wanted to go beyond and try to finds answers to the question of possible causality. Obviously this approach can only produce loose associations and no clear causal reasons for biomarkers` kinetics.

To clarify this approach, we inserted a statement on this in the beginning of the discussion, lines 367-369:

“Our main aim was to determine the dynamics of these biomarkers over the study period of five days. Our additional aim was to test for impact of clinical and laboratory markers that might indicate severity of disease on the biomarkers kinetics.”

4. Discussion: There are many redundant statements in your discussion which made it difficult to follow and understand. For instance, there are repetition of results in several paragraphs. I suggest a discussion with this frame work for clarity and conciseness:

Discussion: We restructured the discussion according to the frame work the reviewer suggested it below. We added a small introductory paragraph for those readers who directly start reading in the discussion. Throughout, we deleted any information that was already present in the other paragraphs of the article (e.g. lines 374-382) except for direct comparisons of our values with published values.

paragraph 1: the research problem or gap and the significance of addressing or filling it

Discussion: We formulated the need for this study and inserted this paragraph into lines 383-389.

paragraph 2: Critical analysis of your major findings +

paragraph 3 additional findings and how they fit into the existing literature (paragraph 3),

Discussion:

We choose to combine these paragraphs, so that we could discuss first copeptin (major + additional findings) and then MR-proADM (major + additional findings). As such, the requested structure could be achieved, just for the two biomarkers one following the other. The whole paragraph was inserted for copeptin in lines 390-443 and for MR-proADM in lines 445-478.

study limitations (paragraph 4),

Discussion: Was presented before already as an own paragraph, is now slightly modified in the lines 510-526.

future directions (paragraph 5)

This was briefly integrated into the summary (conclusion, lines 533-534) at the end of discussion.

and overall conclusion and major impact of the study (paragraph 6).

Discussion: Conclusion has been adjusted according to the changes made throughout the article (now lines 528-534.

Your study limitations may appear illogical to some readers. In lines 423-425, you tend to suggest that severity of the presentation of LRTIs is determined by the location of medical practice. However, it is obvious there are underlying conditions such as immunosuppression that could worsen disease presentation and serve as confounders. In any case, your study exclusion criteria lent credence to this fact (lines 94-95).

Limitations: This is true, underlying conditions like immunosuppression could definitively influence the disease progress and severity. We excluded these cases and did not have any patients with known immune insufficiency in the cohort. However, in our region of Europe it is obvious, that parents can seek medical advice within minutes and around the clock. This is definitively not the case in other parts of the world. Therefore, we are careful with reproducibility of our study in other medical systems.

Reviewer #3: This study is a subanalysis of a prior RCT of children with LRTI that focuses on biomarker kinetics in pediatric LRTI. The strength of this manuscript lies in its novelty – the kinetics of most of these biomarkers in pediatric CAP has not been reported previously. There are several challenges to the approach and presentation of data in this manuscript that make the conclusions difficult to understand.

We thank also reviewer #3 for the extensive review of our article and the amount of time the reviewer invested into it. We have redirected the general orientation of the article to meet the requested requirements.

MAJOR ISSUES

1. This is a heterogeneous population – children with bacterial pneumonia require antibiotics and children with bronchiolitis do not. Therefore, to include use of IV antibiotics as one of the primary outcomes is not appropriate for half of the already small study population. This study requires stratifying data into those with and without pneumonia

In general: Our primary objective was to display biomarkers` kinetics and not biomarker guided antibiotic therapy. We are now aware that this was not as clear as we intended it to be. We did a thorough workover to emphasize kinetics and temper the statements about the associations with clinical and other lab values. We wanted to display biomarkers` kinetics for the whole cohort of pediatric patients with LRTI as we think, that the differentiation of LRTI into subgroups (pneumonia, bronchiolitis, bronchitis) has several flaws and diagnoses can relevantly overlap or dynamically develop from one into another. However, for the sake of completeness, in the article version submitted firstly, we actually did stratify the patients into diagnosis groups (presented in the text and formerly S1 Table, now integrated in Table 1 in main article body) but the stratification did not reveal any relevant associations and was not retained in the final model. Further, only febrile children were included into the initial study. Fever is still a strong driver for antibiotic prescription because of concerns for primary/secondary infection and therefore we wanted to include all children into the final model.

This said, the reviewer is correct, that IV antibiotics should not be a primary outcome and we did not intended it to be primary outcome. It was an association that we found when examining factors that could potentially impact biomarkers` course over time and we think, it is important to mention it. However, as just being an association and as IV administration can indeed be driven by factors other than severe illness, we agree to not overemphasize this.

2. The story is somewhat confusing as written – is this about severe outcomes? Use of antibiotics? Only copeptin and proADM or really about 4 biomarkers? I would recommend that this focus on kinetics of CRP, PCT, copeptin and proADM. The title suggests that it is only copeptin and proADM but there is important data presented on other biomarkers as well that should be accounted for in the title and results sections.

In general: This article is about copeptin and pro-ADM kinetics and their association with clinical and laboratory covariates. In fact, we initially added PCT and CRP kinetics for comparative reasons, but did not explore them the same way as we did with copeptin and pro-ADM. This indeed was confusing. We deleted these graphs and focused the whole article entirely on copeptin and pro-ADM. PCT and CRP will be evaluated along with other factors in a separate article, those factors would be out of scope of this article.

3. Results and tables require clarity on the meaning of the statistical parameters. The message is lost because of a focus on statistical rather than clinical terminology throughout.

Results: This was mentioned by reviewer #1 as well and we are grateful for these comments. We changed the jargon throughout the article and tables to make it more accessible for the broad readership.

ABSTRACT

1. I’d recommend that the results section include more numerical results. It would be more meaningful for readers to see effect estimates with confidence intervals rather than simply p-values.

Abstract: This corresponds well to the comment of reviewer #1, who also requested more numerical values. We have entered the estimated differences between biomarker and decrease values as numerical data into lines 47-52.

2. If the conclusion is that there is additional clinical value using combinations of biomarkers, I would recommend giving some numerical results of the correlations in the results.

Abstract: We agree, that correlation data were missing, but as correlations were rather weak and as we stepped back from the statement on usefulness of combined biomarkers, we deleted the two respective sentences.

INTRODUCTION:

1. Would be more specific in the citation placement of references 1-6. The individual references should be placed after the prior sentences, as they all do not support the sentence that up to 90% are unnecessarily prescribed antibiotics. Would consider adding the following reference to support this statement: Florin TA et al. JPIDS 2020;9(2).

Introduction: References have been changed accordingly and the cited reference was inserted into line 65.

2. An important reference that should be included in Lines 55-57 is: Stockmann C et al. JPIDS 2018;7(1).

Introduction: This reference was inserted into line 66.

3. A more recent reference regarding MR-proADM and disease severity that should be included in the introduction is: Florin TA et al. Clin Infect Dis 2020 Aug 6; ciaa1138

Introduction: Indeed, this article was not published, when we submitted our article. We are grateful to pick this up in line 79 and in the discussion in line 471.

METHODS

1. This data is now more than 10 years old and occurred during the H1N1 influenza pandemic. Can the authors comment on the age of the data and the possible influence of the pandemic on the results?

First, the analyses were performed in 2016 (6 years storage at -80°C) and as there is no hint for long-term degredation at this temperature (Copeptin: Heida JE, Boesten LSM, Ettema EM, Muller Kobold AC, Franssen CFM, Gansevoort RT, et al. Comparison of ex vivo stability of copeptin and vasopressin. Clin Chem Lab Med. 2017;55(7):984-992; MR-proADM: Morgenthaler NG, Struck J, Alonso C, Bergmann A. Measurement of Midregional Proadrenomedullin in Plasma with an Immunoluminometric Assay. Clin Chem. 2005;51(10):1823-1829) we think, that the samples were still usable. Second, H1N1 is assumed to drive inflammatory biomarkers as usual in influenza viruses: they are associated with severe outcome (Vasileva D, Badawi A. C-reactive protein as a biomarker of severe H1N1 influenza. Inflamm Res. 2019;68(1):39-46). This was shown in particular for MR-proADM in inflenza patients (Valero Cifuentes S, García Villalba E, Alcaraz García A, Alcaraz García MJ, Muñoz Pérez Á, Piñera Salmerón P, et al. Prognostic value of pro-adrenomedullin and NT-proBNP in patients referred from the emergency department with influenza syndrome. Emergencias : revista de la Sociedad Espanola de Medicina de Emergencias. 2019;31(3):180-184).

Thus, we think, that a patient enrolled during the pandemic H1N1-influenza situation at that time should be examined for biomarkers the same way as any other influenza patient in different years. Severe disease is expected to drive inflammatory biomarkers.

2. Line 109 – was pneumococcus the only bacterium considered for systemic bacterial infection? What about other known bacterial pathogens of pneumonia?

Methods: No, pneumococcus was not the only bacterium considered for systemic bacterial infection. In lines 108 – 110 of the original version, we mentioned, that we classify microbiology findings into 4 categories: “Microbiology results were classified in 4 categories: not performed, negative, systemic bacterial infection (blood culture positive for pneumococcus/streptococcus = invasive pneumococcal infection) and other.” Indeed, we had overall 6 patients with positive blood culture, in one patient growth of coagulase negative Staphylococcus was present, this was deemed contamination (no intravascular access or foreign materials present) and the patient was treated as blood culture negative. We did consider, of course, many bacterial pathogens that could have grown from blood culture, but we just found the above mentioned ones.

3. Line 114 – it is not clear what age-dependent reference standard is used for vital signs with the exception of respiratory rate. This is not laid out in the text or in Table S1.

S1 Table: This is correct. Reference values were laid out in S2 Table, however the link was indeed missing in S1 Table. S1 Table was merged with Table 1 as it was required to be part of the main text body in one of the next comments, there we refer to S1 Table in line 238.

4. Line 116 – the Fleming percentiles were derived from healthy children in the outpatient setting, thus I’m not certain they are appropriate in this sick hospitalized population. Others (Bonafide, Daymont, Oostenbrink, Thompson) have published vital sign curves for ED and hospitalized children which would be more appropriate to use.

Methods: Thank you for the comment and the recommendations for alternative percentiles. We would like to stick to the Fleming percentile because of the following reasons:

• The Fleming cohort is large (Fleming: 150 080 HR and 7565 RR measurements in healthy children, Bonafide (2013): 77 825 HR and 77 825 RR measurements in children hospitalized in tertiary care, Daymont (2015): 60 863 observations (HR only) in children hospitalized in tertiary care, Oostenbrink (2012): 1555 observations in ED (RR only), Thompson (2008): 1933 observations in primary care (HR only).

• Most important, we wanted healthy children as reference to which we could compare vital sign values from the study.

• In our cohort 51% of patients were hospitalized at Day 1 and 49% returned home and were seen in the outpatient clinics on Day 3 and Day 5. Many (21%) of the patients hospitalized at Day 1 returned home before Day 3 or Day 5. Thus, more vital sign measurements were recorded in the ambulatory setting and not in the hospitalized setting.

5. Lines 145-179 – the section on modeling is not intuitive or easily understood by the average medical reader. I recommend that this entire section be written to summarize the statistical methods in prose that allows the average clinician reading this paper to understand what was performed in the body of the main manuscript. The details of the modeling can be expounded on in S1 Appendix.

Methods: We agree on this. In fact, the information on base model was already redundantly in place in S1 Appendix, so it was deleted from the article. Further, the explanations on covariate modelling have been moved to S1 Appendix as well, leaving a plain language summary in the article text.

6. Lines 166-169 – can the authors justify their use of a stepwise approach to covariate selection, rather than an approach based on biological and clinical plausibility?

Methods: A preliminary selection of covariates based on biological and clinical plausibility was first performed. The selected covariates were then assessed for statistical significance using a standard stepwise approach.

7. Lines 182-183 – it appears that discriminating IV antibiotics vs oral or no antibiotics is a primary outcome. Were there standardized criteria for who received IV antibiotics? This is critical, as this otherwise becomes reflective more of an individual clinician’s gestalt or the fact that an IV line was present rather than a true need for IV antibiotics. This is what makes use of IV antibiotics a challenging primary outcome.

Methods: Biomarkers` kinetics and not IV antibiotic use was the primary outcome. We reworded the article to clearly state this. IV administration was performed according to attending physicians´ discretion. We are aware, that IV use is not a very precise marker for disease severity because of several points that can influence the decision to give antibiotics intravenously (IV line in place, in need of IV fluids etc.). Therefore, we do not want to overemphasize this. However, it stays in the pool of possible associations to test for, because children that get antibiotics intravenously can generally be considered to be sicker than those without IV administration. Further, we deleted the ROC analysis for the same reason.

RESULTS

1. Line 207 – inclusion criteria makes it sound as if only those who had all measurements on day 1, 3, and 5 were included. If this is the case, I am concerned about selection bias for this population, as most children with LRTI who are hospitalized stay in the hospital for fewer than 5 days (median ~2-2.5 days per literature).

Yes, all patients included had blood withdrawn on days 1, 3, and 5, but they were not necessarily inpatients (only 51% admitted to hospital). We asked outpatients to come back on day 3 and 5 to the ED. Therefore we could generate complete datasets also for those patients not (anymore) in need of hospitalization or IV antibiotic. Of 339 randomized patients in the RCT, only two withdrew their consent, they were excluded from analyses. The rest of the patients had blood sampled on day 1, 3, 5. However, in this case “complete data sets” also refers to sufficient quantity of remaining stored plasma volume for this biomarker analysis. Therefore, only 175 datasets with enough stored blood plasma for copeptin and pro-ADM analysis were included in this analysis. Thus, we see no selection bias here. As we did not include this as explanation into the manuscript so far, we were happy to have changed the lines 227-232 to:

“The study population comprised 175 febrile LRTI pediatric patients (age: 1 month-18 years) for whom a sufficient quantity of blood plasma was in storage with febrile LRTI of whom a complete set of 3 consecutive blood samples were available for copeptin and MR-proADM measurements analysis for on day 1, 3, and 5 after study inclusion. of clinical presentation. Blood sampling was performed in-house while patients were hospitalized or in the emergency department, to where the outpatients and the discharged patients were asked to return on day 3 and 5.”

2. Line 215 – there are important data that are buried in Table S1. This table is labeled as a list of variables or data dictionary, when in fact the last column is important data to describe the study population and outcomes. This data should be in the main body of the paper.

Results: We inserted this data into Table 1 and deleted S1 Table as both table overlapped in content.

3. Line 224 - Less than half the children in this study had pneumonia. The pathophysiology and microbiology of different lower respiratory tract infections are different, and therefore I’m concerned that this is a highly heterogeneous population (with a relatively small n, in addition). Children with bronchiolitis shouldn’t be requiring antibiotics to begin with and therefore it is not appropriate to include them with children with pneumonia in a primary outcome that includes IV antibiotic use.

Results: We refer to the third general reply to the major issues of this review further up. 1. Primary outcome was biomarkers` kinetics, secondary outcome were associations with other factors that could impact biomarkers` course over time. IV antibiotics was one of the few that showed a relevant association, not more, it should not be primary outcome. 2. We shy the differentiation of LRTI especially in young children, because case definitions are not very clear (e.g. CXR with highly variable interpretability are included for pneumonia), pathophysiology may overlap and antibiotic prescription is driven by fever. We rather would like to divide the cohort into those in need of antibiotics and those without, regardless of clinical diagnosis. Therefore we wanted to explore biomarkers` kinetics and associated factors in all febrile LRTI.

For the sake of completeness diagnosis stratification was included into primary model, but showed no significant association.

4. Lines 246-252 – this section needs to be edited to explain the results in prose rather than relying on statistical parameters. The meaning of these results is lost in statistical jargon.

Results: The NLME base model description was moved to S1 Appendix. The results are presented now in a more accessible fashion and focused on the end results rather than including the description of the model.

5. Lines 264-266 – same as above – rather than talking about ‘baseline parameter’ and ‘slope parameter,’ this would be easier to read and interpret if the text read: “the following variables were statistically associated with initial levels of MR-proADM…ICU admission and microbiology results were associated with the change in MR-proADM from baseline to Day 5.” (if I’m interpreting the statistical jargon appropriately)

Results: We agree on this and have changed the entire result section and the table to become more accessible. It now reads:

“During forward covariate model building, antibiotic administration route, general complication, microbiology results and presence of fever had a significant influence on MR-proADM concentrations on day 1 (p<0.001). Admission to ICU and microbiology results were associated with the steepness of MR-proADM decrease. During the backward process, association with fever became not significant and was thus excluded. Details on base and final model can be seen in S1 Appendix. In the final model, intravenous administration of antibiotics influenced MR-proADM concentration and was significantly associated with a higher MR-proADM concentration on day 1. Further, ICU admission was significantly associated with a flatter MR-proADM concentration decrease from day 1 to day 5. In contrast, positive blood cultures, which were grouped together with negative NPA or no growth in blood culture because of the same effect when building the model, led to a significant decrease of MR-proADM (Table 1).”

6. Table 2 and 3 – it would be helpful to have footnotes explaining the parameters in this table. What is meant by ‘limit?’

Results: The Tables have been modified and shifted in parts to the S1 Appendix, leaving a simplified single table in the main text body. Limit means the biomarker concentration at the end of the study. “Limit” was exchanged by “Copeptin/MR-proADM on day 5” to clarify this in Table 1 and 2 in S1 Appendix.

7. Line 302 – predict is the wrong term here. The biomarker is not doing the predicting – I would stick with MR-proADM discriminated those children receiving IV antibiotics vs those who did not. If less than half of the children in this study had pneumonia and hospitalized children with pneumonia generally warrant antibiotics, aren’t these results really a proxy for discriminating pneumonia from no pneumonia? There should be an examination of the interaction of diagnosis and IV antibiotics in these analyses.

Results: We do not want to include the interaction between diagnosis and IV antibiotics here, because this is planned to be part of a separate article that includes much more interactions not related to copeptin and pro-ADM and which be outside the scope of this present article. In the new version of this article we would keep kinetics as primary outcome and the factors impacting kinetics as secondary outcomes. For clarity, the antibiotic treatment in the ProPAED interventional group was PCT guided regardless of diagnosis, so only IV administration could serve as potential marker of severity.

8. Figure 2 – requires labels

Results: Figure 2 and the corresponding S6 Table were deleted due to the rather vague outcome parameter antibiotic administration route.

DISCUSSION

1. Line 333 – the discussion states that decreases were not as steep for CRP and PCT, yet the declines observed in Fig S1 do suggest that there were in fact declines in CRP and PCT. If this statement is going to be made, I would recommend some presentation of data in the Results section. In fact, I think that this manuscript would benefit from being reframed as biomarker kinetics in general, including CRP, PCT, copeptin and proADM, as that is really what this manuscript is about – not just copeptin and proADM.

Discussion: We did reframe this article as a biomarker article, focusing on kinetics and factors impacting kinetics. We excluded PCT and CRP from this version. We are planning a further article evaluating additional factors that go beyond the scope of this article. Further, PCT and CRP values have partly been published already, so we choose to exclude them.

2. Line 385 – include some discussion of Florin TA et al. Clin Infect Dis 2020 Aug 6; ciaa1138

Discussion: Included in lines 471.

3. Line 400-401 – just because the correlations were moderate here (some were not very good at all), one cannot conclude that combinations of biomarkers may be helpful. It would have been helpful if there was statistical analysis that focused on the outcomes of interest with combinations of biomarkers to be able to make this claim rather than simple correlations.

Discussion: We agree on this and have tempered our conclusions in line 528-533.

We are grateful to the staff and editors at PLOS ONE for the work they have done on our manuscript. This process helped to improve the quality of the manuscript. Please feel free to contact us anytime if further questions should arise.

Kind regards,

Philipp Baumann Jan Bonhoeffer

14 Jan 2022

30 Jan 2022

Revised Manuscript PONE-D-20-08646R1: The kinetic profiles of copeptin and mid regional proadrenomedullin (MR-proADM) in pediatric lower respiratory tract infections

Dear Yu Ru Kou,

thank you very much for considering the publication of our manuscript. Here we answer point-by-point to the aspects raised by reviewer #1. Please see the corresponding corrections in the uploaded manuscript which we think is now suitable for publication. All line numbers in the revised manuscript correspond to the track-change version.

Reviewer #1: Please abbreviate the 95% confidence intervals as 95%CI (not CI95%) in the abstract and everywhere else in the text.

Authors` response: “CI95%” has been replaced by “95%CI” in Manuscript in lines 42, 43, and 46, and also in S1 Appendix.

Please replace all mentions of 'multivariate' analysis with 'multivariable' - all analyses here focused on one outcome at a time, not multiple outcomes in the same model - they are therefore multivariable NOT multivariate.

Authors` response: “Multivariate” has been replaced by “multivariable” in Manuscript in lines 197/198 (Table 2 caption) and in Table 2 itself.

Please add 95% confidence intervals to the estimates of effect listed in Table 2 and wherever else they are cited in the text.

Authors` response: 95%CIs have been inserted into Table 2 as requested.

We are grateful to the reviewer and editors at PLOS ONE for the work they have done on our manuscript. Please feel free to contact us anytime if further questions should arise.

Kind regards,

Philipp Baumann

Submitted filename: Response to Reviewers.docx

Click here for additional data file.

9 Feb 2022

The kinetic profiles of copeptin and mid regional proadrenomedullin (MR-proADM) in pediatric lower respiratory tract infections

PONE-D-20-08646R2

Dear Dr. Baumann,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Yu Ru Kou, PhD

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

1 Mar 2022

PONE-D-20-08646R2

The kinetic profiles of copeptin and mid regional proadrenomedullin (MR-proADM) in pediatric lower respiratory tract infections

Dear Dr. Baumann:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Yu Ru Kou

Academic Editor

PLOS ONE