High prevalence of Bordetella pertussis in children under 5 years old hospitalized with acute respiratory infections in Lima, Peru

Pertussis is an endemic vaccine-preventable disease with the highest morbidity and mortality in the youngest infants [1]. Worldwide, there are an estimated of 16 million cases of pertussis, 95 % of which occur in developing countries, resulting in about 195 000 children deaths per year [1, 2]. In the last years, an increase in reported cases of pertussis has been noted, even in countries with high vaccination coverage [3–6].

Bordetella pertussis is a fastidious gram-negative coccobacillus which causes Pertussis disease, a highly contagious infection of the human respiratory tract also known as “whooping cough” [3, 5]. Pertussis is characterized by three phases: catarrhal, paroxysmal, and convalescent; being the most infectious periods the catarrhal and early paroxysmal phases [7]. This classic presentation is well-known, but is observed less often since the start of immunization [4].

The standard diagnostic criteria for B. pertussis identification and epidemiological surveillance are culture and molecular techniques such as polymerase chain reaction (PCR). The DNA amplification techniques (e.g., PCR) for B. pertussis detection are faster, and have increase the sensitivity by approximately 19 % the overall percentage of laboratory-confirmed cases, being the preferred method [8, 9]. However, in clinical practice the diagnosis is generally reached without microbiological confirmation leading to a possible lack of clinical awareness to start early treatment and prevent complications [10].

Pertussis can be especially difficult to diagnose in children under 1 year of age during winter season, when other pathogens, such as respiratory syncytial virus (RSV) or Influenza, are prevalent. In these difficult cases, pertussis acute respiratory symptoms can overlap with those of bronchiolitis or other unspecific acute respiratory infections [10, 11]. This is especially worrisome in infants too young to be immunized in whom atypical and more severe presentations have been reported, often requiring hospitalization for respiratory or other complications [7, 12, 13].

In Peru, pertussis is a major health problem that has been raising in the last 5 years [14]. Furthermore, the most affected are infants under 1 year old representing 38 % of cases, despite a national immunization coverage of 92 % in this age group [15, 16]. Currently, the “whole cell” B. pertussis vaccine (DTwP) is the only available presentation in Peru; and the national coverage level for this vaccination is 88.3 % for the 3 doses of the pentavalent vaccine (DTwP-Hib-HepB) according to the 2014 epidemiology reports [17].

To study the Pertussis epidemiology in Peru is essential in order to understand the real impact of the disease, especially in the most vulnerable population. The aim of this study is to determine the prevalence, epidemiological and clinical characteristics of B. pertussis and Respiratory syncytial virus cases in infants under 5 years old hospitalized with acute respiratory infections in a Peruvian hospital between 2009 and 2010.

Positive samples for B. pertussis and RSV were analyzed according to age distribution, and infants under 3 months old were the most frequently affected in 43 % (49/114) and 35.9 % (37/103) respectively. A similar sex distribution was observed in both groups. Moreover, around 59 % of enrolled children had a previous contact with another patient with acute respiratory infections. An equivalent proportion of household contacts was observed for Bordetella pertussis and RSV positive samples (Table 1).

A higher prevalence of B. pertussis cases were registered between October and November 2009 and February to April 2010. (Figure 1) Seasonal indexes were calculated for B. pertussis and RSV positive samples separately. An increase of pcases was observed from February to March and from October to November with a Seasonal index between 1.32 and 1.51 and 1.24–3.5 respectively. A similar predominance was observed in RSV cases from November to December. However, RSV showed to be also frequent from April to June with a seasonal index between 1.09 and 2.00 (Fig. 2).

Bordetella pertussis is a strict human pathogen which causes whooping cough, an endemic illness responsible of significant morbidity and mortality, especially in infants under 6 months old [1, 2, 5]. Although regional differences exist, Pertussis represents a considerable global disease burden that has been increasing, even in countries with high vaccination coverage [2, 5, 13]. In Peru, an alarming increase of cases has been observed in the last 5 years, and 56 % of cases are reported in infants under 1 year old [12, 14–16]. This have raise especial concern since infants under 6 months old are more vulnerable to disease related complications and carry a higher mortality [7, 21, 22].

The most common clinical manifestations of B. pertussis infections are prolonged and paroxysmal coughing, accompanied by inspiratory stridor [1, 3]. However, several factors are known to affect the disease presentation and Pertussis diagnosis may go unrecognized when other pathogens, such as respiratory syncytial virus (RSV) or Influenza virus circulate [10, 23, 24]. A retrospective study in Italy, from a group of infants hospitalized from October 2008 to April 2010 for acute respiratory symptoms reported that most of Pertussis cases were infants under 6 months with median of 71.5 days old and a male: female ratio of 6:13 [10]. In our study pertussis toxin and IS481 genes were detected in 19.12 % (114/596) of the patients admitted with an acute respiratory infection and infants under 3 months old were the most frequently affected in 43 % (49/114) with a similar sex distribution.

Co-infection between Bordetella pertussis and RSV has been previously described to cause severe infections [10, 11]. A study conducted in a group of infants hospitalized for RSV bronchiolitis showed that almost 2 % of patients were co-infected with B. pertussis [25, 26]. In our series, co-infections were observed in 14 patients between B. pertussis and RSV-A and 1 sample was positive for B. pertussis and RSV-B. Moreover, 6 out of 9 cases of co-infections were clinically diagnosed as Bronquiolitis and B. pertussis was not suspected at the time of admission. Influenza virus and B. pertussis co-infections have been also identified as a possible pathogen present in children with community-acquired pneumonia; and the pertussis toxin-mediated suppression have been postulate to be responsible to produce more sever presentations [27, 28].

Multiple studies have reported Paroxysmal cough (76.5–91.1 %), cyanosis (46.7–81.7 %) and respiratory distress (47.8–55.7 %) as the most common symptoms in children [13, 29, 30]. However, several clinical features might help to suspect the diagnosis of pertussis in infants hospitalized for acute respiratory symptoms. [10]

One study in 2013, compared infants with Pertussis and confirmed RSV bronchiolitis; and the clinical characteristics showed that the percentage of infants with paroxysmal cough was significantly higher in infants with B. pertussis. Additionally, cough at admission lasted longer in infants with pertussis than in control infants. Also, fever was significantly lower in infants with pertussis, and more common in patients with bronchiolitis. In our study population, a similar clinical symptoms frequency was observed between patients with B. pertussis and RSV. The most frequently reported symptoms were fever, cough, rhinorrhea and respiratory distress, in more than 60 % of cases. However, the presence of rhinorrhea 88.35 %, respiratory distress 76.70 % and pharyngeal congestion 33.98 % was more common among patients with RSV. This higher frequency of symptoms in our study may be related to fact that more than 52 % of our patients were hospitalized infants under 6 months old.

The clinical diagnosis of Pertussis in infants can be challenging, especially in children with incomplete immunizations, and some patients may be catalogued as acute viral respiratory infections, before laboratory confirmation. Thus delaying the appropriate antibiotic treatment and isolation measures [11, 24]. In our series, pneumonia was clearly the most frequent diagnosis in 26.32 % (30/114) of the patients with positive B. Pertussis. However, other diagnosis were considered in this group, such as rhinopharingitis, bronchiolitis and influenza infections. In contrast, the diagnosis of Bronchiolitis was more common in 20.37 % (21/103) of children with a positive sample for RSV.

For Bordetella pertussis seasonality, a pattern corresponding to the summer and spring months have been reported in the southern hemisphere [13]. Comparably, a previous study in infants under 6 month of age from 2003 to 2008 in Lima, registered more hospitalizations due to whooping cough during the months of February and September. In our study, a similar distribution was observed with an increase of B. pertussis cases from February to March and from October to November and a Seasonal index between 1.32 and 1.51 and 1.24–3.5 respectively.

Pertussis represents a considerable disease burden in Peru and the diagnosis is complicated by the limitations of currently available diagnostic tests. Therefore, the only diagnostic tests that are recommended for case confirmation in national reporting are culture and polymerase chain reaction (PCR) [7, 31]. However, in Peru the use of PCR for surveillance was started recently in 2012 and there is still evidence of a deficient report and registration of cases that limit the analysis of the real disease burden.

As in other Latin American countries, epidemiologic surveillance for B. pertussis is essential in Peru, especially in children that could most benefit from the vaccine. This study demonstrates a considerable incidence of B. pertussis in children previously diagnosed as acute respiratory infections and highlights the importance of possible co-infections that may difficult the diagnosis and prognosis of patients. There is an increasing need for further investigations to better establish the impact of the disease and improve vaccination programs especially in hospitalized children were more severe presentations have been reported.