Background

Whooping cough or pertussis is a vaccine –preventable disease (VPD) caused by Bordetella pertussis– an encapsulated Gram-negative coccobacillus. Since the 1930s, a whole-cell pertussis (wP) vaccine has been developed in combination with diphtheria and tetanus toxoids (DTwP). Massive worldwide coverage with DTwP reduced the infection rate (per 100,000 pop.) from 157 in the prevaccination era to less than 1 in the 1970s¹. Reports of B. pertussis toxin-mediated reactogenicity led to the development of an acellular pertussis (aP) vaccine in combination with diphtheria and tetanus (DTaP for children under 7 years of age and Tdap for children over 7 years of age, the latter with a lower concentration of antigens against diphtheria and pertussis). Trials in the 1990s demonstrated that DTaP combination vaccines are much less reactogenic than DTwP vaccines¹. However, despite the use of the vaccine, the vast majority of countries present outbreaks or an increase in the number of cases every 3 to 5 years².

Three recent studies have compiled and summarized the status of pertussis vaccination in Latin America^3–5. These reviews mentioned some challenges for the region: differences inherent in information systems, lack of standardized case definitions, and epidemiological adverse-event surveillance, passive in nature after vaccination^4,5. Furthermore, it was reported that the low geographical representation limits the generalizability of results and that the heterogeneity of the study methods limits the comparison of results³.

The Latin American Society for Pediatric Infectology (SLIPE) and the Latin American Pediatric Association (ALAPE) have taken a proactive role in recommending vaccination against B. pertussis. In a recent joint publication, ALAPE and SLIPE analyzed the continuity of the health services most affected by the pandemic in the region, particularly vaccination, and made recommendations based on the best practices reported in Latin American countries⁶.

In Latin America, the incidence of B. pertussis disease before the year 2000 was estimated at 17.8 (95% confidence interval [CI] from 5.9 to 29.7) cases per 100,000 population, while from of the year 2000, incidence was estimated between 2.5 and 8.5 cases per 100,000 population^3,4 with 90% of the cases reported in under-1-year-olds⁴. The case fatality ratio of pertussis in Latin America was estimated in 3.9% in the general population³ and in 4.8% in under-18-year-olds⁴, although again, most deaths occured in children too young to be vaccinated⁴.

Given the importance of B. pertussis disease in Latin America and the changes that have occurred in vaccination recommendations, this manuscript aims to review epidemiological data and recent changes in vaccination schedules and their impact on pediatric B. pertussis disease in Latin America.

Epidemiologic characteristics of B. pertussis disease

Pertussis is an endemic respiratory disease with global distribution and outbreak cycles every 3-5 years^7–10. Despite being a VPD, pertussis has high overall rates of hospitalization, complications, and deaths, especially in under-2-month-olds^7–11. Despite the global increase in vaccination coverage against pertussis, 151,000 cases were reported worldwide during 2018¹⁰.

To reduce its transmission, vaccination coverage must be greater than 90% and, to be effective, it must be homogeneous within a country and between countries in the region⁹. Otherwise, epidemic outbreaks appear that primarily affect the most vulnerable groups, especially under-1-year-olds.

Pertussis is underreported⁹, which is mainly due to regional variations in the case definition and to the different diagnostic methods used in each country, especially with the incorporation of the polymerase chain reaction (PCR), which significantly increases sensitivity for the detection of B. pertussis^12,13. Therefore, standardizing this definition for the region will allow adequate and comparable epidemiological surveillance between countries and regions. At a recent meeting, a working group of Latin American experts and the Pan American Health Organization (PAHO) proposed a homogeneous definition of cases (Table 1)¹⁴.

It is important to note that not all infected people meet all the criteria for a case definition. Patients, generally adolescents and adults, may present asymptomatic or with mild symptoms, such as a persistent cough, without neither other symptoms, nor a significant direct burden on morbidity. Since the disease has a reproductive number (R₀) between 15 and 17, it is considered highly contagious¹⁵. Atypical cases particularly put the youngest unvaccinated children or those with incomplete vaccination schedules at risk. A recent review of 16 studies found that the secondary attack rate is much more pronounced in infants (70%) than in the rest of the population (20%) or in those over 40 years of age (6%)¹⁶. This review also found that 42.6% of household contacts present with mild or atypical whooping cough and 55.6% of the contacts studied had laboratory-documented asymptomatic infection.

Current epidemiology of pertussis in Latin America

Recent data from PAHO, WHO, and UNICEF show a significant incidence rate in under-1-year-olds and that the groups 1-to-4-year-olds and adolescents and young adults should also be taken into account (Figure 1). For this reason, incidence rates should be registered by age groups through an active surveillance program, especially in the group of adolescents and young adults.

Figure 1 Pertussis incidence rates in Latin America, 2006-2018. Source: 2006-2018 data from the Joint Report Form by PAHO/WHO and UNICEF, 2019. 

We will now review the 2018-2020 situation of some countries in the Latin American region, summarized in Figure 2. It should be noted that these data have previously described limitations^3–5, that include heterogeneity between information systems, lack of definitions standardized case reports and passive epidemiological surveillance, with little geographical representation, especially from smaller or less populated countries^3–5.

Figure 2 Age distribution of pertussis cases in selected Latin American countries, 2018-2020. Notes: Mexico reported the age groups in 3-year increments and they were adjusted to the groups that best correspond in the classification presented. Costa Rica and Colombia presented data for the first half of 2019. Sources: Argentina: Directorate for the Control of Immunopreventable Diseases in the SNVS 2.0 database. Ministry of Health. Brazil: SINAN, SVS / MS, as of January 25, 2020, obtained through the Electronic Citizen Service and Information System on 03/03/2020. Chile: ENO database and EPIVIGILA database as of January 10, 2020. Department of Epidemiology, DIPLAS. Ministry of Health. Colombia: National Institute of Health, Ministry of Health (Colombia). Pertussis: Colombia, first semester of 2019. Bogotá, Colombia: May 31, 2019. Document FOR-R02.4000-001. Costa Rica: Chanto, Grettel. Alert: Increase in cases of pertussis, Costa Rica, January-February 2019. Tres Ríos, Costa Rica: INCIENSA, 2019. Retrieved October 6, 2020 from https://www.inciensa.sa.cr/actualidad/Alertas/2019/Alert%20increase%20de%20cases%20de%20tosferina,%20Costa%20Rica%20January-February%202019.pdf. Mexico: Epidemiological Surveillance Bulletin. Secretariat of Health of Mexico. Peru: National Center for Epidemiology, Prevention and Control of Diseases. Ministry of Health. 

Vaccination against B. pertussis

There are two types of vaccines against B. pertussis: whole-cell (wP) and acellular (aP) vaccines. The wP vaccines use the whole bacteria, which can be of different strains, inactivated by physical or chemical methods. Hence, they contain many antigens and have variable immunogenicity, efficacy, and safety patterns, which we will present in this review.

With the development of aP vaccines, which contain fewer antigens (from 1 to 5), are better regulated and have less reactogenicity, developed countries have incorporated DTaP vaccines in their vaccination programs in pentavalent formulations (with the addition of Haemophilus influenzae type b and hepatitis B) or hexavalent (with the addition of inactivated polio vaccine [IPV]). The wP vaccines continue to be used in developing countries. Information on the immunogenicity, efficacy and safety of the wP vaccines used today is very limited⁹ but we will discuss the most recent data in this regard.

Within the Central American region, most of the countries use wP in their national schemes, except Costa Rica, Panama and Mexico, which already include aP for children. In the Andean region (Bolivia, Venezuela, Colombia, Peru) and most of the southern cone (Argentina, Brazil, Paraguay, Uruguay) only wP is used, with Chile being the exception when introducing the hexavalent aP vaccine since 2018. Many of the countries in the region already administer Tdap to pregnant women (Table 2).

The adoption of wP vaccines in the 1950s in both Japan¹⁷ and Sweden¹⁸ demonstrated a reduction in the incidence rate of pertussis, but reports of reactogenicity, which could explain, at least in part, the population's reluctance to vaccinate with the consequent decrease in vaccination coverage. In turn, this reduction in coverage translated into a resurgence of whooping cough that prompted the introduction of aP vaccines (Japan in the 1970s and Sweden in the 1990s), with adequate subsequent control^17,18.

Another explanation for the resurgences is the waning of immunity¹⁹. Countries that have introduced both aP and wP vaccination in different schedules show cyclical resurgences of pertussis. For these reasons, vaccination planning against B. pertussis should follow the life cycle with specific indications for infants, adolescents, adults, and pregnant women⁴.

Reactogenicity and effectiveness of wP and aP

Not all wP vaccines are the same. A study in Iran showed significant differences in local and systemic reactogenicity after vaccination with locally produced DTwP²⁰. Differences in relation to adverse events and immunogenicity may be associated with the strains used in the production of the vaccines, the methods of preparation or formulation of vaccines.

One study compared aP vaccines with wP vaccines having fewer adverse effects with less fever, pain and local swelling at the injection site²⁴. In addition, children vaccinated with aP have a lower risk of neurological adverse events, especially hypotonic-hyporesponsive episodes and seizures²⁵. The Canadian experience after the introduction of aP in 1997-1998 showed a significant reduction in episodes of febrile seizure (60%) and episodes of hypotonic-hyporesponsive (67%).²⁶ An analysis of data from 567,378 children born before and after the introduction of aP vaccines in Canada found a significant reduction in the relative risk of visiting an emergency room due to frequent adverse effects associated with vaccination with wP²⁷. This had 38 times more impact compared to a reduction in more rare events, such as seizures²⁷.

Various studies have shown that registered aP vaccines have proven efficacy in children, regardless of the number of components^8,21–23. A recently published systematic review with meta-analysis and mathematical modeling²⁸ described the effectiveness of aP vaccines (VE) after 5 doses of the infant series (VE: 91%; 95% confidence interval [CI]: 87-95%) and after the sixth dose in adolescents (VE: 70%; 95% CI: 54-86%). Mathematical modeling demonstrated an annual decrease in VE after 5 and 6 doses (9.6 and 45.3% per year, respectively)²⁸. After booster, the absolute VE was estimated at 85% (95% CI: 84-86%) with a decrease of 11.7% (95% CI: 11.1-12.3%) annually to 28.2% at age 18²⁸. The study concluded that aP vaccines have high Initial efficacy that decreases over time²⁸. This is fundamental, since observational studies on VE fail to recognize that they measure relative efficacy data and not absolute VE, as was the case in the meta-analysis presented. It is important to note that neither vaccine, aP or wP, nor natural infection by B. pertussis is capable of conferring lasting protection.

In the United States of America, the California Kaiser Permanente system study, which involved 469,982 children between 3 months and 11 years old, reported 738 cases of whooping cough, distributed as follows: 99 were not vaccinated, 36 were under-vaccinated, and 515 were fully vaccinated. Compared with fully vaccinated children, unvaccinated children had 13 times higher risk and under-vaccinated children had a 1.9 times higher risk of developing pertussis²⁹. Sub-analyses by age found that in children 19-84 months, the risk of pertussis increases five times in vaccinated people older than 3 years compared to children under 1 year and in children aged 84-132 months, the risk of pertussis increases twice in vaccinated people older than 6 years compared with children under 3 years²⁹.

During the transition from wP to aP in Turkey, a prospective study was conducted with more than 1,400 infants³⁰. The study found significant differences in local reactions (9.1 vs. 0.9%), systemic (12 vs. 1%) and fever (26 vs. 2.6%). Infants vaccinated with DTaP at 2 months had more systemic reactions (1.9%) compared to those vaccinated at 4 (0.92%), 6 (0.6%), and 18 months (0.13%). The wP vaccine was more reactogenic at all doses³⁰.

Patterson and colleagues³¹ analyzed 47 studies, mostly controlled clinical trials, that reported adverse events after vaccination with wP or aP. The analysis found significantly higher relative risk (RR) with wP in terms of local reaction (RR 2.73), pain (RR 4.15) or edema (RR 4.38) at the injection site, fever (RR 9, 21), drowsiness (RR 1.34), and vomiting (RR 1.28)³¹. Health authorities may consider local adverse events as minimally relevant in making vaccine decisions. However, this study demonstrates that local adverse events can interfere with vaccine acceptance and result in low coverage or delay in childhood vaccination. Furthermore, the WHO considers the reduction of local pain associated with vaccines to be a good vaccination practice³².

Despite the relatively high vaccination coverage against B. pertussis, the bacterium continues to circulate in the community, since neither wP nor aP prevent pharyngeal colonization.

Experience with aP in Latin America

We present next the Latin American experience with the countries that have introduced aP vaccination in the region, in the sequence in which they have introduced the vaccine.

Decreased coverage and delay in timely vaccination

The WHO recommends that vaccination against pertussis should be started early (between 6 and 8 weeks of age) and that coverage should be greater than 90%. In the Americas, the coverage for the three-dose primary series (DTP3) was close to 90% in 2018, but few countries present high and homogeneous coverage, with enormous regional variations in the coverage of the primary series between countries (provinces, states, municipalities or districts). It is important to note that vaccination coverage is a “hard” indicator that does not allow identifying the risk for children with delayed vaccination schedules. Therefore, indicators must be developed to take into account the impact on pediatric populations. On the other hand, coverage with four completed doses of diphtheria, pertussis and tetanus (DPT4) compared to coverage with three doses (DPT3) shows that DPT4 is below the optimal 80% and that coverage is not homogeneous for the region. (Figure 3).

Figure 3 HIstoric series of pertussis cases () and vaccination coverage with DPT3 () and DPT4 () in the Americas, 1978-2018. Source: 2006-2018 data from the Joint Report Form by PAHO/WHO and UNICEF, 2019. 

Despite the persistence of factors affecting pertussis vaccine coverage in Latin America (for instance, anti-vaccine movements, population reluctance, adaptation of strains through antigenic changes, reemergence, and decreased immunity over time), it is generally agreed that vaccination against pertussis is safe and effective. The greatest impact is observed in infants under 1 year old who receive the complete series, reducing the burden of disease. However, vaccination coverage must be high and homogeneous to achieve a true decrease in the incidence of the disease.

The reemergence of pertussis is observed vis-a-vis with the redefinition of cases with the diagnostic use of PCR, including atypical forms that are not diagnosed. In addition to the underreporting of such cases, there is contagion at home and in the community. On the other hand, having had the natural disease does not confer protection for life, so the vaccination and reinforcement program should be continued, even in people who have had whooping cough.

Another important measure is maternal vaccination, which protects children who have not started the vaccination schedule and reduces the incidence and fatality of pertussis. Maternal vaccination as a safe strategy in reducing the incidence of pertussis is beyond the scope of this review; however, there are recent reviews of the subject and the first impact data in Latin America^{4,33,34,40,41}.

Recommendations of the SLIPE and ALAPE group of experts on infant vaccination

The SLIPE and ALAPE Group of Experts met to present recommendations based on the revised data. These recommendations are summarized in Table 3 and are presented below.