Pertussis, mainly caused by Bordetella pertussis, is a highly contagious disease affecting all age groups, with high morbidity and mortality among young infants.1,2 Several factors are believed to be involved in the resurgence of pertussis, including waning immunity, reduction in vaccine efficacy, genetic changes in the bacterium and increasingly high utilization of polymerase chain reaction (PCR) testing.3 Currently, PCR for B. pertussis is usually a component of an extensive respiratory panel. Accordingly, this study examines the incidence of B. pertussis, viral pathogen or combination of the 2 among outpatient and hospitalized infants up to 1 year of age with clinical suspicion of pertussis.
MATERIALS AND METHODS
This retrospective laboratory study analyzed 168 Nasopharyngeal samples (NP) samples of infants <1 year, throughout Israel, suspected of having pertussis by their healthcare provider, February 2015 to January 2016. Under Israel pediatric pertussis immunization policy, 1 of 2 acellular pertussis vaccine (aP) formulations—Infanrix, GlaxoSmithKline (Rixensart, Belgium) or Poliacel, Sanofi Pasteur (Toronto, Canada)—is administered at 2, 4, 6 and 12 months. In addition, diphtheria, tetanus and acellular pertussis vaccine (dTaP)-inactivated polio virus (Boostrix Polio, GlaxoSmithKline) is administered at 7 or 8 years of age and tetanus toxoid, reduced diphtheria toxoid and acellular pertussis vaccine dTap (Boostrix, GlaxoSmithKline) at 13 or 14 years of age. Immunization of all pregnant women with Tdap (primarily Boostrix), preferably at 27–36 weeks, is also advocated.
Nasopharyngeal samples were subjected to amplification testing via multiplex PCR targeting the B. pertussis IS481, Bordetella parapertussis, Bordetella holmesii IS1001, BHIS1001 and ptxS1 loci. To detect respiratory viruses, the Anyplex (South Korea) II 16 respiratory viruses (RV-16) detection assay (V1.1; Seegene, South Korea) was used. B. pertussis was cultured on Regan-Lowe Bordetella agar, incubated for a week at 37°C.
Among the 168 patients suspected of having pertussis, B. pertussis was detected in only 48 (28.5%) samples. By comparison, 1 or more viral agents, but not B. pertussis, were detected in 103 (61.3%) of the remaining 120 NP samples (χ2 = 67.75, P < 0.001). No B. parapertussis was detected. In 17 (10.1%) samples, no pathogen was detected. Of the 48 samples containing B. pertussis, 37 (77%) had a concomitant respiratory virus, while in 11 samples (22.9%) B. pertussis was the sole pathogen (χ2 = 14.08, P < 0.002).
The incidence rates for the various subgroups were: RV-16 only (n = 103); total B. pertussis (n = 48), comprising B. pertussis/RV-16 coinfection (n = 37) and B. pertussis only (n = 11); and No Pathogen (n = 17) (Fig. 1).
Overall, rhinovirus was most frequently detected in both the RV-16 only (56/103; 54.4%) and RV-16/B. pertussis coinfection (26/37; 70.2%) subgroups (χ2=2.82, P > 0.12), followed by adenovirus 24/103 (23.3%) versus 12/37 (32.4%) in the RV-16 only and coinfection subgroups, respectively (χ2 = 1.18, P = 0.28).
Patient age ranged from 10 days to 11.4 months, mean 3.1 ± 2.2 months (median, 2.4 months), with 88% of the infants <6 months. Sixty-three infants (37.5%) were <2 months, 65 (38.7%) were 2–4 months, 20 (11.9%) were 4–6 months and 20 (11.9%) were >6 months old. There was no statistically significant difference in positive B. pertussis PCR rates between infants <2, 2–4, 4–6 and >6 months of age (P > 0.37) or between infants >6 months (4/20, 20%) and <=6 months of age (44/148, 29.7%); χ2 = 0.37, P > 0.36. Forty-four (91.6%) of the total B. pertussis group represented infants <6 months. Hence, all of the B. pertussis only infants were <6 months, as were the vast majority of infants with B. pertussis/RV-16 coinfection, 33/37 (89.2%) versus 4/37 (10.8%) for the <6 months and ≥6 months groups, respectively (χ2 = 44.84, P < 0.0001). There was no statistically significant difference in age distribution between the B. pertussis only and B. pertussis/RV-16 coinfection groups, χ2=1.26, P = 0.56.
The RV-16 only group comprised 103 infants, of whom 8 (85%) were <6 months and 15 (15%) were ≥6 months. Rhinovirus was the most common virus found in this group, with no statistically significant difference between the younger and older infants, 46/88 (52.3%) versus 10/15 (66.7%), respectively (χ2=1.07, P > 0.30). B. pertussis cultures were not performed in 36 samples (21.4%). Of the remaining 132 samples, 24 (18.2%; 95% confidence interval: 12.3–25.5%) had positive cultures. All 24 samples were B. pertussis PCR positive as well. Twenty-one subjects with negative cultures had positive PCR.
Respiratory viruses frequently mimic B. pertussis and are more likely to be the causative agent in infants under 1 year of age with suspected pertussis. In about three-quarters (71.4%) of the study group, PCR was negative for B. pertussis; instead, either respiratory viruses were detected (61.3%) or no pathogen was detected at all (10.1%). Note that 77% of infants with positive B. pertussis PCR had a concomitant respiratory virus. Additionally, B. pertussis infection and B. pertussis/RV-16 coinfection were just as prevalent with younger infants as older ones. Finally, in terms of the pathogens potentially mistaken for B. pertussis or co-occurring with B. pertussis, rhinovirus was the most common virus detected followed by adenovirus.
Past studies also highlight that pertussis-like cough may be caused by several respiratory pathogens.4–6 Respiratory outbreaks in New Hampshire, Massachusetts and Tennessee initially thought to be caused by pertussis found that pertussis was unlikely to have been the primary etiology; rather, respiratory viruses or atypical bacteria were the etiologic agent.7 The converse is also described: Greenberg et al8 found that pertussis can mimic viral lower respiratory tract infections in infants; 15% of infants with severe lower respiratory tract infections admitted to the pediatric intensive care unit were found retrospectively to be PCR positive for B. pertussis. Accordingly, it behooves the practitioner to make the correct diagnosis as early as possible and initiate treatment to reduce B. pertussis spread to susceptible contacts. Moreover, instituting chemoprophylaxis only for individuals exposed to pertussis, and not those exposed to a respiratory virus mimicking pertussis, should be a primary aim in assuring the judicious use of antibiotics.
In this study, rhinovirus was the most common pathogen detected with B. pertussis. Still, it should be noted that while rhinovirus was once thought to be nearly ubiquitous and a frequent colonizer in asymptomatic children, its role as a major cause of acute respiratory illness, like pneumonia in young children, has only recently been appreciated.9
Another important factor to consider in pertussis epidemiology is the patient’s age. Ministry of Health data show DTap vaccination coverage in the study period to be 99.2%, 98.3%, 96.9% and 91.9% for vaccines at 2, 4, 6 and 12 months, DTap1 through -Dtap4, respectively; furthermore, vaccination rate of pregnant women with Tdap (primarily Boostrix) is 45%–53% (Haas, Division of Epidemiology, Israel Ministry of Health, April 11, 2018, personal communication). Interestingly, the incidence of pertussis observed in this study was similar among the different age groups. A possible explanation may be the high pertussis incidence rate (5202) in 2015, the year when most of the nasopharyngeal samples were collected.
This study’s limitations include its retrospective nature with samples collected over a period when B. pertussis incidence was relatively high, thereby limiting the generalizability of the findings. Additionally, important clinical data were not available.
Although respiratory viruses frequently mimic the presentation of B. pertussis in young infants suspected of having pertussis, healthcare providers should have a low threshold for employing PCR pertussis testing, given the epidemiologic implications of undiagnosed pertussis. Coinfection of B. pertussis and respiratory viruses seems to be common, but the significance for clinical severity remains controversial. Further studies should be performed to investigate the effect of coinfection of B. pertussis and respiratory viruses on disease severity.
1. Greenberg DP, von König CH, Heininger U. Health burden of pertussis in infants and children. Pediatr Infect Dis J. 2005;24(5 suppl):S39–S43.
2. Masseria C, Martin CK, Krishnarajah G, et al. Incidence and burden of pertussis among infants less than 1 year of age. Pediatr Infect Dis J. 2017;36:e54–e61.
3. Tan T, Dalby T, Forsyth K, et al. Pertussis across the globe: recent epidemiologic trends from 2000 to 2013. Pediatr Infect Dis J. 2015;34:e222–e232.
4. Ferronato AE, Gilio AE, Vieira SE. Respiratory viral infections in infants with clinically suspected pertussis. J Pediatr (Rio J). 2013;89:549–553.
5. Nicolai A, Nenna R, Frassanito A, et al. Respiratory viruses
and B. pertussis
co-infections: a frequent occurrence in children hospitalized with B. pertussis
. Eur Respir J. 2016;48:PA1275.
6. Frassanito A, Nenna R, Nicolai A, et al; Pertussis study group. Infants hospitalized for Bordetella pertussis
infection commonly have respiratory viral coinfections. BMC Infect Dis. 2017;17:492.
7. Centers for Disease Control and Prevention (CDC). Outbreaks of respiratory illness mistakenly attributed to pertussis –New Hampshire, Massachusetts, and Tennessee, 2004–2006. MMWR Morb Mortal Wkly Rep. 2007;56:837.
8. Greenberg D, Bamberger E, Ben-Shimol S, et al. Pertussis is under diagnosed in infants hospitalized with lower respiratory tract infection in the pediatric intensive care unit. Med Sci Monit. 2007;13:CR475–CR480.
9. Toivonen L, Schuez-Havupalo L, Karppinen S, et al. Rhinovirus infections in the first 2 years of life. Pediatrics. 2016;138:e20161309.