Streptococcus pneumoniae is a gram-positive bacterium with more than 95 identified capsular serotypes1. Infection with this pathogen is a major cause of morbidity and mortality in both adults and children worldwide, with a wide clinical spectrum that ranges from asymptomatic colonization to mucosal disease, and invasive infection (involving previously sterile sites)2,3. Before the introduction of the pneumococcal conjugate vaccine, S. pneumoniae caused between 8% and 12% of all deaths in children under 5 years, amounting to an estimated 1 million deaths per year worldwide. In 2008, the mortality was estimated at 500 000 deaths per year worldwide4–6.

Invasive pneumococcal disease (IPD) is most prevalent in the very young and the elderly, especially in children under 5 years. The most frequent presentations are pneumonia, meningitis and bacteraemia. The introduction of pneumococcal conjugate vaccines has had a significant impact on IPD in terms of disease incidence and the serotype distribution. After the introduction of the heptavalent pneumococcal conjugate vaccine (PCV7) in the United States, there was a dramatic decline in the incidence of IPD. However, in our region there was a significant increase in the incidence of IPD caused by non-PCV7 serotypes, a slight reduction in the rate of IPD caused by PCV7 serotypes and emergence of previously stable virulent clones of non-PCV7 serotypes2,7–9. With the aim of improving vaccine coverage against non-PCV7 serotypes, the 13-valent pneumococcal conjugate vaccine (PCV13) was approved in 2010 for active immunization of children aged 6 weeks to 5 years. Although the Advisory Committee on Vaccines of the Asociación Española de Pediatría (Spanish Association of Pediatrics, AEP) has recommended routine administration of conjugate pneumococcal vaccines (PCV7 in the 2003–2010 period and PCV13 once it had been authorised/introduced), at the time of this writing these vaccines were not funded by the public health system of Catalonia except for children with selected risk factors, and therefore their administration depended on the judgment of the paediatrician and the willingness of families. The pneumococcal vaccine coverage in Catalonia in children under 60 months has been estimated at 50% in the pre-PCV13 period and 63% in the PCV13 period, based on previous studies conducted by our group9–11.

The aim of the present study was to compare the incidence and describe epidemiological variables, the clinical presentation, current trends and serotype distribution for IPD among children before and after the introduction of the PCV13 in Catalonia, Spain, a region where this vaccine was not included in the routine vaccination schedule.

During the study period, 493 cases of IPD were detected, 319 in the pre-PCV13 period and 174 in the PCV13 period. A serotype had been documented in 395 of the total cases (245 in the pre-PCV13 period and 150 in the PCV13 period). Table 1 presents the analysis of the baseline characteristics. We found differences in terms of breastfeeding, attendance to childcare centre or school, previous history of respiratory infection, risk factors for pneumococcal disease, microbiological diagnosis and the sites from which positive samples were obtained.

Although the vaccine coverage in our region is not high, the incidence of invasive pneumococcal disease in children under 60 months has decreased significantly after the introduction of PCV13. The general characteristics of the population in both periods were similar.

Pneumonia, meningitis and occult bacteraemia accounted for more than 90% of cases of IPD in both groups. We observed a significant decrease in the cases of pneumonia complicated with empyema associated with a shift in the distribution of serotypes that caused IPD in the PCV13 period (decline of serotype 1), which was consistent with the literature24–26, while cases of necrotising pneumonia increased substantially from 0.84 to 3.3 cases per 100 000 inhabitants. Although length of stay and mortality did not change with the introduction of the PCV13, we found a relative increase in severity reflected in an increase in the proportion of patients requiring admission to the ICU, requiring ventilation or presenting with septic shock. It is worth noting that there was also a significant increase in the proportion of children with risk factors. We assume that the changes in clinical presentation were due to the impact of vaccination with the PCV13, which was associated with an evident shift in the serotypes causing IPD. In the PCV13 period, opportunistic serotypes, or serotypes with low invasive disease potential, have caused disease with a significantly greater frequency in patients with a higher rate of comorbidities, which may in turn be associated with an increased risk of complications and mortality27,28. We believe that the increase in the proportion of cases caused by serotype 3 could have an important impact on the incidence of necrotising pneumonia, we observed an increase in both the absolute and relative frequency of this form of disease between the 2 periods (from 3 [0.8%] to 19 [10.9%]) despite a decrease in overall incidence, which suggests that there are other factors at play in the increase in necrotising pneumonia, such as an increased proportion of patients with comorbidities in the PCV13 period.

Overall, the use of pneumococcal DNA detection by PCR increased the frequency of cases of IPD with microbiological confirmation, as previously reported29. However, in the PCV13 period there was a lesser increase of cases with detection by PCR only. This was probably due to the decline in cases manifesting with empyema, as pneumococcal PCR is significantly more sensitive in samples of pleural fluid (where a high bacterial load is expected) compared to blood, where the bacterial load is lower.

There was an overall decrease in the incidence of IPD caused by PCV13 serotypes, with a decline in serotype 1 from being the most frequent (20.9%) in the pre-PCV13 period to accounting for only 11.2% of IPD cases in the PCV13 period. Although we observed a significant decrease in the incidence of IPD caused by serotype 3, this serotype became the most prevalent in the PCV13 period, causing 20.9% of IPD cases. This proportional increase can be explained by the failure rate of the PCV13 vaccine for serotype 3 in our study, a phenomenon previously reported by our group and other authors12,30,31. The causes of serotype 3 vaccine failure are not well understood, but are probably related to the low immunogenicity of the vaccine for this particular serotype12.

Serotype 3 is the most frequent serotype, especially in the most severe cases, such as those requiring ICU admission, mechanical ventilation or prolonged hospitalization and those with complications during the hospital stay or sequelae after discharge. This combination of high incidence and high proportion of severe cases was among the salient results of this study.

One of the limitations of the study was that the molecular techniques used to detect pneumococcal DNA and in serotyping changed during the study. In the first period, we used amplification of the ply gene, which has a high sensitivity, and a multiplex PCR assay for detection of 24 serotypes, while in the second period we used amplification of LytA gene for pneumococcal detection, which offers a better specificity, and a different molecular assay to detect 40 serotypes. This change in methodology could have played a role in the significant decrease of patients with IPD caused by serotypes not included in the assay.

In conclusion, in the PCV13 period there was a decrease in the incidence of IPD without serotype replacement. We observed changes in epidemiological variables and clinical manifestations, with a proportional increase in cases in children with risk factors and an increase in cases of necrotising pneumonia, with required admission to the ICU and mechanical ventilation in a greater proportion of patients.

We now need to see what happens in upcoming years following the inclusion of the PCV13 in the routine immunisation schedule of Catalonia in July 2016.

Dr de Sevilla MF received grants from the Instituto de Salud Carlos III and fees from Pfizer Inc. while the study was underway. Dr Moraga-Llop has received fees from Pfizer Inc. outside the study period. Dr Campins received fees from Pfizer while the study was underway. Dr Muñoz-Almagro has received grants from Pfizer, bioMérieux, Stat DX and the Instituto de Salud Carlos III and fees from Roche and GSK outside the study period. Dr García-García received grants from Instituto de Salud Carlos III and fees from Pfizer Inc. while the study was underway. The remaining authors had no conflicts of interest to disclose.

This work was supported by the National Plan of Research, Development and Innovation of Spain (PI11/02081, PI11/02345 and PI13/01729 projects) through the Insituto de Salud Carlos III (General Vice Directorate of Health Research Evaluation and Promotion) and the European Regional Development Fund (ERDF).

Johanna (J) Martínez-Osorio, MD*: Design and methodology, data collection, writing and data analysis. Mariona (M) F de Sevilla, PhD*: Design and methodology, data collection, writing and data analysis. Fernando (F) Moraga-Llop, MD: Expertise, feedback and writing. Alvaro (A) Díaz, PhD: Expertise, feedback and writing. Sergi (S) Hernández MSc: Data analysis and writing. Anna (A) Solé-Ribalta, MD: Data collection and writing. Sebastià (S) González-Peris, MD: Data collection and writing. Conchita (C) Izquierdo, PhD: Expertise, feedback and writing. Cristina (C) Esteva, PhD: Laboratory sample processing and writing. Gemma (G) Codina, PhD: Laboratory sample processing and writing. Ana María (AM) Planes, PhD: Laboratory sample processing and writing. Sonia (S) Uriona, MD: Laboratory sample processing and writing. Magda (M) Campins, PhD: Expertise, feedback and writing. Pilar (P) Ciruela, PhD: Expertise, feedback and writing. Luis (L) Salleras, PhD: Expertise, feedback and writing. Ángela (Á) Domínguez, PhD: Expertise, feedback, writing and fundraising. Carmen (C) Muñoz-Almagro, PhD: Design and methodology, Expertise, feedback, laboratory sample processing, writing of the manuscript and fundraising. Juan José (JJ) García-García, PhD: Design and methodology, Expertise, feedback, writing and fundraising.