Bordetella pertussis in children hospitalized with a respiratory infection: clinical characteristics and pathogen detection in household contacts

Describe the prevalence of Bordetella pertussis via PCR in children under 5 years old hospitalized as probable cases of pertussis and report the most common clinical features among them. A positive PCR result for B. pertussis was observed in 20.5% of our samples (18/88), one-third of them were from infants between 2 and 3 months old. The most common symptoms were paroxysms of coughing (88.9%), difficulty breathing (72.2%), cyanosis (77.8%) and fever (50%). The mother was the most common symptomatic carrier (27.8%), followed by uncles/aunts (22.2%) among children with pertussis.


Introduction
Pertussis, also known as 'whooping cough' , is an acute respiratory tract infection caused by the gram-negative bacteria Bordetella pertussis [1][2][3]. Worldwide, around 30 million cases of pertussis and 160,000 deaths in children younger than 5 years old are registered every year, 90% of them occurring in developing countries [4][5][6].
In the last years, despite a widespread vaccination, the resurgence of B. pertussis infections has been observed, primarily affecting low-income countries [7][8][9][10][11]. Despite the pertussis vaccine success, cyclical outbreaks are observed every 2-5 years as they did in the prevaccine era [12,13]. Furthermore, the vaccination calendar has left a window of vulnerability for newborns and infants in which high morbidity and mortality rates are observed [14][15][16][17]. In addition, in infants younger than 3 months and neonates the increasing incidence of this infection and the sustained mortality rates have encouraged further investigation about the efficacy and safety of pertussis vaccination for pregnant women as an attempt to reduce the disease burden [18].
In Peru, since 2017 a rapid increase of B. pertussis has been observed and by the first half of the year, the number of cases has doubled compared to the previous year suggesting the possibility of a new outbreak [19]. This study main objective was to describe the prevalence of B. pertussis via PCR in children under 5 years old hospitalized as probable cases of pertussis and report the most common clinical features among them.

Materials and methods Patients
the Andes Mountain Range and was the second most affected region by B. pertussis in 2016 [19].
Children under 5 years old hospitalized as probable cases of pertussis at the Hospital Regional de Cajamarca were consecutively studied from April 2016 to September 2017. Household contacts with similar respiratory complaints were also included in the study and were considered positive after PCR detection of B. pertussis.
Cases were defined as probable for pertussis in the absence of a more likely diagnosis of cough illness with one of the following symptoms: Paroxysms of coughing or inspiratory "whoop" or posttussive vomiting or apnea as per CDC case definition recommendations [20].
This study was approved by the Research Ethics Board of the Hospital Docente Regional de Cajamarca, Peru. A written informed consent was signed by parents or children caregivers before enrollment. Household contacts also signed a written informed consent before enrollment.

Samples
Nasopharyngeal samples were obtained inserted one swab into each nostril parallel to the palate (calcium alginate swab, USA) and submerged into transport solution (phosphate buffered saline).

DNA extraction
DNA was extracted from a volume of 200 μl of each samples using a commercial kit (High Pure Template Preparation Kit, Roche Applied Science, Germany) according to the manufacturer's instructions.

Real-time PCR assay detection Bordetella pertussis with the TaqMan probe
PCR was performed using a BHQ quencher probe at 125 and 250 nM of primers in a final volume of 20 μl. Five microliters of the extracted DNA were combined with 15 µl of the master mix. PCR conditions for B. pertussis were 95 °C for 10 s, 60 cycles of 5 s at 95 °C, 5 s at 57 °C and 30 s at 72 °C. All cycles were performed in Light Cycler ® 2.0 (Roche Diagnostic, Deutschland-Mannheim, Germany). The primers and the probe used were described by Kosters et al. [21].

Statistical analysis
Quantitative variables were described as frequencies and percentages for each group using the GraphPad Prism3 statistical (Graph Pad Sofware Inc., San Diego, USA).

Results
A total of 88 children under 5 years old hospitalized as probable cases of pertussis were prospectively studied from April 2016 to September 2017. In our study population, more than 70% of patients were infants under 3 months old, with infants between 29 days to 2 months old being the most predominant age group in 31.8% of patients. No significant difference between gender was observed among our patients as 53.4% were male vs 46.6% females (Table 1).
A positive PCR result for B. pertussis was observed in 20.5% of our samples (18/88), one-third of them were from infants between 2 and 3 months old followed by three positive cases on infants between 29 to 2 months old and three cases in children between 1 and 5 years old. Most of our patients with a positive sample were male infants in 72.2% of cases ( Table 1).
Our patients were hospitalized as probable cases of 'whooping cough' and the most common presenting symptoms were paroxysms of coughing (76.1%), difficulty breathing (75%), cyanosis (67%) and fever (52.3%). The same symptoms were also observed among patients with positive samples for B. pertussis in which paroxysms of coughing (88.9%), difficulty breathing (72.2%), cyanosis (77.8%) and fever (50%) were the most frequent complaints. Furthermore, clinical symptoms were compared by age groups showing that paroxysmal coughing was the most common symptom across all ages, except in neonates were cyanosis was the most common presentation. Complications during hospitalization were also registered, pneumonia was the most frequent outcome in 33% of our study population and 38.9% of our patients with B. pertussis (Table 2). In addition, 83% (15/18) of our patients with B. pertussis received antibiotic on day 1 of hospitalization, the same day we enrolled them but before we sampled and reported our results. High leukocyte and lymphocyte count were assessed in our patients. We observed that both leukocytosis and lymphocytosis were present in four cases with B. pertussis. On the other hand, leukocytosis and lymphocytosis were also observed in three and four patients with negative results for B. pertussis respectively (Additional file 1: Table S1).
From our patients with a positive result for B. pertussis, most cases older than 2 months old didn't receive any vaccination 76.9% (10/13), one patient had the incomplete vaccination and only one patient with a positive sample for B. pertussis received the two vaccine doses (Additional file 1: Table S2).
Household contacts who presented with respiratory symptoms were also evaluated for the presence of B. pertussis. We found that mothers and siblings under 7 years old were the most common family members with positive samples for B. pertussis in our population. This predominance was more evident among patients with B. pertussis in which 27.8% (5/18) of their mother were also positive for the bacteria (Table 1).
We enrolled patients with a clinical presentation compatible with whooping cough through the whole study period. However, most of our patients with positive samples for B. pertussis were observed during 2017, being January and June the months with the higher case-rate distribution (Fig. 1).
In our study, a total of 18 cases (20.5%) had positive samples for B. pertussis, being 70% of them infants under 3 months old. This proportion of positive samples is slightly lower than a previous study we conducted in 2015, where 38.4% of probable cases for pertussis were also positive for the bacteria [27]. However, these variations are expected as the clinical features have shown to be insufficient to establish a diagnosis and it is estimated that without PCR testing, the overall percentage of missed cases would range from 9 to 26% per year in infants under 6-month-old [29].
Even though pertussis clinical presentation tends to be unspecific, there are some clinical features that seem to be more common in different age groups [14,30]. In our population, we observed that difficulty breathing and cyanosis were present in most of our patients between 29 days to 3 months old; however, we were surprised that paroxysmal coughing was also a common symptom in this age group. Additionally, it also has been reported that post-tussive vomiting is common at all ages [14]; in Table 2    It has been demonstrated that a whooping cough alone is not enough to start antibiotics immediately, especially in infants younger than 4 months [5]. However, in rural areas where laboratory resources are limited physician usually give macrolides when there is high suspicious of pertussis. In our population, 17% (15/88) of patients received antibiotics on day 1 of hospitalization before we took samples. In addition, 83% (15/18) of patients with B. pertussis were covered before we report their results as positive, and the other three cases were started on antibiotics the same day we sent our results.
One study in Mexican infant showed that 59% of patients with pertussis had leukocytosis and 64.4% presented lymphocytosis, with a mean age of 87.1 days old [31]. In our study, we found that leukocytosis and lymphocytosis were each observed only in 4 (22.2%) pertussis cases. However, these findings were not exclusive among patients with B. pertussis, as we found three patients with leukocytosis and four with lymphocytosis in the group of patients with negative samples.
Bordetella pertussis is a highly contagious disease acquired through direct contact or inhalation and the source of infection is usually the mother has been identified as the most common source in up to 63% of cases, followed by fathers, siblings and other family members [11,17,31]. In our study, among children with B. pertussis, we found that mothers were the most common symptomatic carriers in 27.8% (5/18) of patients, followed by uncles or aunts in six cases. Notably, we also found nine mothers with B. pertussis among infants with negative samples of the bacteria demonstrating the presence of carriers even in patients without the disease.
Evidence suggests that the households contacts transmission occurs due to the limited immunity to B. pertussis infection whether acquired naturally or by immunization [32]. In adult carriers, who usually do not receive booster protection, the levels of protective antibodies almost negligible with more than 50% of them lost in the first 6 years after vaccination [33]. On the other hand, contacts between 0 and 18 years old, primarily represented by the patients' siblings, are usually carriers without vaccination or incomplete schedules who did not receive chemoprophylaxis [34].
The most recognized strategy is the maternal pertussis immunization supported by the World Health Organization (WHO) which has been found to be highly effective at preventing severe disease in infants [34][35][36][37]. However, the effectiveness of pertussis immunization in preventing transmission in other household contacts is mostly unknown [33,34]. Nevertheless, it is recommended that all contacts who have an incomplete immunization should be vaccinated in addition to receiving chemoprophylaxis. In addition, adolescents and adults, especially those in close contact with children, may benefit from a booster dose of the acellular vaccine. Although, this immunity will wane considerably fast having no impact in a distant future [34].
We analyzed pertussis vaccination status in our population. Not surprisingly, we observed that 76.9% (10/13) our patients older than 2 months old with B. pertussis didn't receive any vaccination. Thus, indicating that children with incomplete vaccination are prone to be

Conclusions
Bordetella pertussis is an important cause of respiratory tract infection in children under 5 years of age, primarily affecting infants under 6 months old in whom the diagnosis is not always suspected. The vaccination schedule leaves a window of vulnerability for B. pertussis infection among infants under 2 months old. In rural areas of Peru, limited laboratory resources may contribute to the underdiagnosis of the disease and an increased use of macrolides upon clinical criteria. In our study, mothers were the most common symptomatic carrier in patients with and without pertussis.

Limitations
The present study had one important limitations. We designed the study for the detection of B. pertussis we cannot exclude the presence of other etiologies that may be responsible for the patients' clinical presentation.

Authors' contributions
JdVM conceived the study; JdVM, PW, and WS designed the study protocol; MAAL, RAO, CdV, and JML performed real-time PCR, analysis and interpretation of these data. JBM, EC, WS and AS responsible for the collection of samples and for completing the clinical information of the epidemiological filled. JdVM, CdV, and PW drafted the manuscript; JdVM and PW critically revised the manuscript for intellectual content. All authors critically revised the manuscript for intellectual content. All authors read and approved the final manuscript.