Invasive Pneumococcal Diseases in Korean Adults After the Introduction of Pneumococcal Vaccine into the National Immunization Program

Streptococcus pneumoniae is a Gram-positive aerobic diplococcus that acts as both a colonizer of the nasopharynx of healthy carriers and a dreadful pathogen in susceptible groups. S. pneumoniae (pneumococcus) is strictly a human pathogen. Pneumococcus still remains a significant cause of both mild infections such as otitis media, sinusitis, and bronchitis and more severe manifestations such as bacteremia, pneumonia, and invasive pneumococcal diseases (IPDs). IPD is defined as the isolation of S. pneumoniae from sterile site such as blood, cerebrospinal fluid (CSF), or pleural fluid. Bacteremic pneumococcal pneumonia is, by far, the most common clinical presentation, followed by primary pneumococcal bacteremia (also termed as pneumococcal bacteremia of unknown source). Meningitis, septic arthritis, peritonitis, osteomyelitis, endocarditis, and soft tissue infection are other major forms of IPDs [1].

While IPD is relatively rare in the general Korean population, it is still a very severe infectious disease; its mortality rate is as high as above 30% in many previous studies, which is much higher in extremely old population [2, 3, 4, 5, 6, 7, 8]. In this paper I will review the major characteristics of epidemiology, risk factors, and antibiotic resistance of IPD in Korean adults after introduction of pneumococcal vaccine into a national immunization program.

The incidence rates of IPD vary depending on the age group, comorbidities, immunocompromised condition, socioeconomic state, geographic location, residency and, importantly, vaccination status of the population [9]. The epidemiology of pneumococcus has changed over the past 20 years with the development of the pneumococcal vaccines targeting common, specific, and invasive serotypes. Following the introduction of the 7-valent pneumococcal conjugate vaccine (PCV7) in Korea in 2003, the proportion of IPD caused by vaccine-type pneumococci has decreased while non-PCV7 serotypes, especially serotypes 19A and 6A, became predominant among childhood IPD isolates [10]. In 2013 and in 2014, Korea introduced 10-valent, and 13-valent PCV (PCV10/PCV13) into the childhood national immunization program (NIP) and a 23-valent pneumococcal polysaccharide vaccine (PPSV23) into the elderly (≥65 years) NIP, respectively. In Korea, PCV10/PCV13 coverage in children aged <24 months has reached 97% in 2022, while the PPSV23 cumulative coverage rate in elderly populations had been achieved above 55% at the same year [11, 12].

The estimated annual incidence of IPD ranged from 4.1 to 6.5 cases per 100,000 persons per year during periods from 2011 to 2014, while the case fatality rate was estimated to be 30.8% in Korea [3]. Korea has strong surveillance for IPD that allows for tracking trends over time (Fig. 1). Since IPD was designated as a Korean national notifiable infectious disease in 2014, all cases occurring are being reported to the Korea Disease Control and Prevention Agency (KDCA). There was a total of 3,734 cases with IPD reported from September 2014 to mid-November 2023, and 66.8% (2,496) of all cases were men. The incidence rate by reporting year showed the highest incidence rate of 1.29 cases per 100,000 people per year in 2018, then decreased, then tended to rise again to 0.52 in 2021, 0.66 in 2022, and 0.69 in 2023. The same increasing trend in incidence was observed when only adult data occurring in the same year were analyzed. The incidence rate per 100,000 population of Korean IPD cases by age was highest in those aged 65 or older (32.1, 54.8%), followed by those aged 60 to 64 (15.7, 11.2%), and those aged under 5 (13.5, 8.1%). (Fig. 2). Figure 3 shows the age distribution of cases by year. By 2022, the overall mortality rate for IPD cases is 14.1% (447/3,377). By age, the highest mortality rate was 21.9% (325/1,487) in those aged 70 or older, followed by 10.1% (72/714) and 10.8% (51/471) in those aged 60 to 69 and 50 to 59, respectively. The mortality rate in the aged 0 - 9 group was the lowest at 1.6% (5/309). Detailed information on clinical features, such as clinical syndrome classification, hospitalization period, and cause of death, was not confirmed in the KDCA registration system [13].

Figure 1
Invasive pneumococcal disease (IPD) case count and incidence rate by year from September 2014 to November 2023.
Source: Korea Disease Control and Prevention Agency. https://npt.kdca.go.kr/.

• Click for larger image

Figure 2
Invasive pneumococcal disease (IPD) incidence rate by age group, from September 2014 to November 2023.
Source: Korea Disease Control and Prevention Agency. https://npt.kdca.go.kr/.

• Click for larger image

Figure 3
The age distribution of invasive pneumococcal disease case by year group from September 2014 to November 2023.
Source: Korea Disease Control and Prevention Agency. https://npt.kdca.go.kr/.

• Click for larger image

There are a few studies in Korea that classified and analyzed IPD as a clinical syndrome. Bacteremic pneumonia was the most common type of IPD among Korean adults overall (overall, 64.3%; range: 54.3 - 72.0%); primary bacteremia (13.5%) was the second most common type of IPD and meningitis (9.4%) was the third most common type of IPD in collected prospectively from 20 hospitals through the nationwide surveillance program from 2013 to 2015 [4]. In a prospective multicenter IPD surveillance study conducted in adults aged 19 years or older from 2019 to 2021, most samples were collected from blood (90.6%), while the remainder were collected from pleural fluid (5.6%) and CSF (3.1%), respectively [5].

Increases in the rate of non-vaccine serotype (VT) carriage or disease after implementation of a vaccine, a phenomenon known as serotype replacement, is of significant concern with any NIP, including pneumococcal vaccines. Although the proportion of VT-IPD and overall IPD in the unvaccinated population, especially in the adult ≥65 years old (indirect effect) and the vaccinated population (direct effect), decreased after PCV was administered in NIP for children, serotype replacement was observed with an increase in unvaccinated serotype. Indirect effects after childhood PCV NIP, a decrease in VT-IPD and overall IPD, and an increase in non-VT IPD may be influenced by the pre-PCV incidence of IPD and the duration after PCV introduction [14]. Prior to the introduction of PCV13, serotypes included in PCV10, PCV13, and PPV23 accounted for 39.8%, 67.3%, and 73.4%, respectively, of pneumococcal infections in Korean adults [15]. Studies on surveillance data after the implementation of PCV revealed an increase in disease caused by serotypes found in PPV23. PCV13, PPSV23, and non-vaccine strains accounted for 31.9%, 58.6%, and 41.4%, respectively [5].

After PCV13 introduction, the percentage of non-VT in 2014 - 2017 as the post-PCV13 era on Jeju Island was increased compared to in the 2009 - 2010 as the pre-PCV13 era (61.1% vs. 29.4%, P <0.01) [16]. As a result of analyzing changes in IPD serotype distribution in adults after the introduction of PPSV23 vaccination for seniors over 65 years of age in 2013 and PCV10/PCV13 vaccination for children in 2014, PCV13 serotype was found to have decreased. PCV13 coverage for IPD decreased from 36.9% (2013 - 2015) to 31.9% in the 2019 - 2021 study after the introduction of PCV10/PCV13. In adults, serotypes 3 and 19A remain the most common serotypes causing IPD [4, 5]. On the other hand, among the serotypes preventable by PCV13 in Korean children under 5 years of age, the most frequently occurring serotype was 19A, and 75% of cases occurred in infants under 1 year of age [9]. Serotype 3, which occurs most commonly in adults and is especially prominent in those over 65 years of age, has rarely been observed in infants and young children. It can be assumed that the difference in serotype distribution by age is due to the effect of PCV13 in infants and young children. PCV13 serotypes, especially 3A and 19A, remain prevalent in adult IPD, suggesting the need for additional PCV13 vaccination in older adults and patients with chronic diseases.

A wide range of clinical symptoms are associated with bacterial virulence and host factors [17]. Pneumococcal virulence is largely dependent on the polysaccharide capsule, of which more than 100 serotypes have been identified. Overall, serotypes 1, 7F, and 8 were associated with decreased serotype-specific mortality risk ratios, and serotypes 3, 6A, 6B, 9N, and 19F were associated with increased risk ratios. Serotypes with higher risk ratios had higher carriage rates, were less invasive, and were more encapsulated in vitro [18]. The more invasive serotypes are more likely to cause infection upon acquisition and are less likely to be found in the nose [19]. Less invasive serotypes, despite the low virulence of the pathogen, are more likely to colonize in the nose of vulnerable host who may develop clinically significant infectious diseases [20].

Healthy adults develop antibodies against the capsular polysaccharide within a few weeks of being colonized with pneumococci [21]. Persons with a reduced ability to produce antibodies to capsular polysaccharide remain susceptible while colonized. The level and quality of antibodies to capsular polysaccharides decrease with aging. As vaccinated people age, IgG levels decrease and antibodies become less effective as opsonizing agents [22, 23].

Several host-related and social factors influence the severity of pneumococcal infection. In adults, IPD incidence and mortality increase with age. Many factors that cause IPD tend to increase severity and mortality. All non-immunocompromised (IC) chronic diseases, including chronic obstructive pulmonary disease, smoking, diabetes, chronic heart failure, liver disease, and chronic kidney disease, are associated with a higher incidence of IPD [24]. The risk of IPD in specific chronic conditions was compared with the risk in healthy adults using 1999 and 2000 data from the Active Bacterial Core (ABC) Surveillance and the National Health Interview Survey, controlling for age, race, and other chronic conditions [24]. The overall incidence rate per 100,000 was 8.8 in healthy adults, 51.4 in adults with diabetes, 62.9 in adults with chronic lung disease, 93.7 in adults with chronic heart disease, and 100.4 in adults with alcohol abuse. Among the evaluated high-risk groups, the highest risks were for adult solid cancer (300.4), human immunodeficiency virus (HIV)/Acquired Immune Deficiency Syndrome (AIDS) (422.9), and blood cancer (503.1) [21].

A combination of risk factors can further increase the risk of IPD. In adults with chronic lung disease, diabetes, and solid tumors, the incidence increased with age [25]. IPD risk can be estimated based on a person's underlying medical conditions and total number of medical conditions. The underlying conditions with the highest adjusted relative risk (RR) for IPD were chronic liver disease (RR = 2.1, 95% confidence interval [CI]: 1.5 - 2.8) and chronic obstructive pulmonary disease (COPD; RR = 2.1, 95% CI: 1.8 - 2.5). IPD risk increased with increasing number of medical conditions: adjusted RR, 2.2 (95% CI: 1.9 - 2.5) for one condition, 2.9 (2.5 - 3.5) for two conditions, and 5.2 (4.4 - 6.1) for three conditions. For a single non-IC condition, the risk of IPD was twice that of the general population, and people with multiple chronic non-IC conditions had an increased risk of IPD with each additional condition [26]. Alternatively, either active smoking or passive inhalation may have a dose-response relationship with the development of IPD, but the risk may return to normal when exposure is eliminated [27].

Immunosuppression due to drugs, HIV, or asplenia is an important risk factor for IPD. Of the 36 million people with private health insurance in the United States, 17% had risk factors for IPD, and 36% of these had immunosuppression (defined as the presence of cancer, organ transplant, asplenia, HIV, or chronic kidney disease) [21]. The incidence of IPD in these high-risk individuals varied depending on the underlying disease. Patients who develop IPD while immunosuppressed are not only more likely to develop bacteremia without focus and septic shock, but are also nearly three times more likely to die compared to non-immunocompromised patients [28]. In epidemiological data on the pneumococcal infection in patients with autoimmune inflammatory rheumatoid disease (AIRD), patients with anti-tumor necrosis factor (TNF) drugs were reported to have a fivefold (5.97 vs. 1.07 - 1.2) higher incidence of pneumonia than healthy people [29]. The mortality rate in patients with rheumatoid arthritis due to pneumonia has increased by two to fivefold, and the hospitalization rate is twofold higher than that in the general population [30]. As the proportion of those with pneumococcal infection and complications among patients with AIRD increases, all patients with AIRD are recommended to be vaccinated against S. pneumoniae [31].

Patients who develop acquired immunodeficiency due to immunosuppressant treatment are at higher risk of developing IPD due to suppressed response to pneumococcal vaccine [32]. Pneumococcal vaccines, as inactivated vaccines, can be administered regardless of the state of immunity, and do not affect the activity of AIRD. In general, the effects of S. pneumoniae vaccines on patients with AIRD are nearly identical to those on healthy persons [31]. As for the effects of medications administered to patients with AIRD on the immunogenicity of vaccines, methotrexate, rituximab, and abatacept were reported to decrease immunogenicity. The immunogenicity of pneumococcal vaccines differed depending on the type of anti-TNF drugs in early studies. In patients who are expected to administer immunosuppressants for a long period of time, if vaccination is necessary after confirming their pneumococcal vaccination history, it is recommended to administer vaccination prior to administration of immunosuppressants.

The pneumococcal vaccination rates in HIV-infected patients between 2006 and 2017 were estimated using the Korean HIV/AIDS cohort database. In HIV-infected Korean patients, pneumococcal vaccination rates (3.0% in 2015, 7.6% in 2016, and 9.6% in 2017) increased annually, remained low despite the recommendations for high-priority immunization. Considering the low pneumococcal vaccination rates, there was a discordance between experts' opinions and real clinical practice in the medical field [33].

A systematic review reported an incidence of IPD of 2.9% in patients with severe post-splenectomy infections [34]. In a prospective cohort study of 173 German intensive care units, asplenia was the only predictor independently associated with pneumococcal sepsis (adjusted relative risk, 2.53 [95% CI: 1.06 - 6.08]) [35]. The incidence and relative risk (RR) of IPD in asplenic/hypoplenotic patients were investigated through the National Health Insurance Service of Korea. The 8-year cumulative incidence of IPD was 0.5%. 45.6% of infections occurred within 2 years of diagnosis. The age-standardized incidence rate was 104.5 per 100,000 person-years (95% CI: 103.6 - 105.4). Patients younger than 5 years had a 15.1 times higher risk of IPD than patients older than 60 years (95% CI: 5.8 - 39.5, P <0.0001). The RR of IPD was 32.0 times higher (95% CI: 21.7 - 47.0) in patients with asplenia/hyposplenism compared to the general population. After age standardization, the RR of IPD was 17.9 times higher in the asplenia/hyposplenism group than in the general population (standardized incidence rate = 17.9, 95% CI: 11.8 - 26.0) [36].

Viral infections usually occur in association with pneumococcal infections. The temporal correlation between IPD and the incidence of viral infections of both influenza and respiratory syncytial virus is reproducible from year to year [37].

In a meta-analysis study, the overall mortality rate of IPD was reported to be 20.8% (95% CI: 17.5 - 24%). Factors associated with mortality included age (odd ratio [OR] = 3.04, 95% CI: 2.5 - 3.68), nursing home (OR = 1.62, 95% CI: 1.13 - 2.32), nosocomial infection (OR = 2.10, 95% CI: 1.52 - 2.89), and septic shock. (OR = 13.35, 95% CI: 4.54 - 39.31), underlying chronic disease (OR = 2.34, 95% CI: 1.78 - 3.09), solid organ tumor (OR = 5.34, 95% CI: 2.07 - 13.74), immunosuppression (OR = 1.67, 95% CI: 1.31 - 2.14), and alcohol abuse (OR = 3.14, 95% CI: 2.13 - 4.64) [38]. In Korean adults, the elderly population has the highest rate of IPD incidence at 59.9%, but the mortality rate is quite high at 51.0%, and for those aged 75 or older, the fatality rate reaches 34.2%. Older age (≥65 years), higher Pitt bacteremia score (≥4, 62.7% vs. 10.2%, P <0.001), and bacteremic pneumonia (82.4% vs. 61.2%, P = 0.019) were independently associated with IPD mortality. In all age groups, an additional risk factor for mortality was identified for immunocompromised status (OR = 7.26, 95% CI: 1.57 – 33.46, P = 0.011) for patients aged 50 - 64 years [4]. Mortality from IPD remained high, and these findings may help clinicians provide appropriate initial treatment for IPD.

Increasing antibiotic resistance in pneumococci has been a global problem. Since the first penicillin-resistant pneumococcal infection was reported in Australia in 1967 [39]. Pneumococcal resistance to penicillin and other antibiotics has been increasing worldwide, especially in Asia [15, 40, 41, 42]. In the Asian Network for Surveillance of Resistant Pathogens Study, which included patients with pneumococcal pneumonia in 11 Asian countries in 2008 - 2012, 59.3% of cases had multidrug-resistant (MDR) pneumococci. Korea has the third highest incidence of antibiotic resistance at 63.9%, following China and Vietnam [15]. A prospective surveillance study on S. pneumoniae collected from adult patients (≥50 years old) with IPD or community-acquired pneumonia was performed at 6 Asian countries (Korea, China, Malaysia, Singapore, the Philippines, and Thailand) in 2012 - 2017 [40]. The proportions of isolates with serotypes covered by PCV13 were 37.0% in Korea, 53.4% in China, 77.2% in Malaysia, 35.9% in the Philippines, 68.7% in Singapore, and 60.2% in Thailand. Major serotypes were 19F (10.4%), 19A (10.1%), and 3 (8.5%) in 2012 - 2017, with different serotype distributions in each country. The prevalence rates of penicillin non-susceptible pneumococci significantly increased from 4.9% in 2008 - 2009 to 9.0% in 2012 - 2017 (P <0.05) with the revised breakpoints (intermediate [I], 4 mg/L; resistant [R], ≥8 mg/L). Macrolide resistance in pneumococci was high (66.8%) and prevalence of MDR to three or more types of antibiotics also remained high (50.8%). MDR non-PCV13 serotypes such as 11A, 15A, 35B, and 23A have emerged in Asian countries [40].

The recent penicillin non-susceptibility (I or R) rate in IPD was 26.6% compared to the 2011 IPD study (10.5%), reflecting the continuously increasing trend of penicillin non-susceptibility in IPD in Korea [2, 4]. From 2017 to 2019, the overall rate of non-susceptibility to penicillin and ceftriaxone was 37.2% and 29.7%, respectively, across all age IPD groups, and the rate of penicillin non-susceptibility was higher in children under 5 years of age than in adults age 65 or older (41.9% vs. 35.4%) [8].

In addition to penicillin, resistance to macrolides, lincosamides, tetracyclines, trimethoprim-sulfamethoxazole (TMP/SMX), and fluoroquinolones has also been reported and is of growing concern [20, 43, 44]. From 2017 to 2019, the resistance rates to erythromycin, clindamycin, and tetracycline in Korea were high at 80.3%, 66.7%, and 75.9%, respectively. Resistance rates to TMP/SMX and levofloxacin were 27.7% and 4.1%, respectively [8]. Resistance to macrolide drugs has been recorded as high as 85% in Hong Kong [43], while it is as low as 25% in Canada [44] and as low as 7% in the UK [45]. In Korea, the rate of MDR was high at 57.5% [4].

In previous studies, risk factors for acquiring resistance included age, previous antibiotic exposure, residence in a long-term care facility, and comorbidities [46, 47, 48, 49]. Antibiotic resistance to penicillin in S. pneumoniae has been considered a significant problem, but the use of penicillin-based drugs has not led to clinical failure in practical therapeutic applications. The clinical relevance of the in vitro susceptibility of pneumococcal isolates remains controversial. While some studies have reported that antibiotic resistance in S. pneumoniae is not clinically relevant [50, 51], some studies have shown that patients infected with antibiotic-resistant S. pneumoniae have a higher mortality rate [52, 53, 54]. The relationship between antibiotic resistance and mortality in pneumococcal disease remains unclear because there are few controlled studies.

Additionally, to date, there is a lack of data on the clinical impact of ceftriaxone resistance. Resistance to cephalosporins in S. pneumoniae is associated with mutations in penicillin binding proteins (PBPs), particularly PBP1a, PBP1b and PBP2x [55]. In particular, during the 2019 - 2021 study period, 32.5% of invasive pneumococcal isolates were not susceptible to ceftriaxone, and during the 2013 - 2015 period, 18.7% of IPD isolates were not susceptible to ceftriaxone, but in most countries reported at a low rate. <10%: Japan (8.4%, 2011 - 2019), South Africa (8.0%, 2003 - 2010), United States (8.7%, 2010 - 2011) [56, 57]. Penicillin resistance does not appear to affect survival, but there may be worse outcomes in patients with cephalosporin resistance. In previous studies, mortality was significantly associated with host risk factors, such as age (≥80 years) and presence of underlying liver diseases, but was not associated with bacterial factors such as serotype or antibiotic resistance [5, 58].

Antibiotic susceptibility results were different depending on the serotype. Resistance rates were higher for several specific serotypes. In a recent Korean prospective study, serogroups 3 (13.8%), 19A (9.5%), 23A (6.9%), 10A (4.3%), 15B (4.3%), and serogroups 11 (6.9%) were common causative agents of IPD [5]. While antibiotic resistance has been reported in some serotypes (6A, 6B, 9V, 14, 15A, 19F, 19A, 23F, etc.), in Korea, serogroup 11, serotypes 19A, 19F, 15B, and 23A are not susceptible to ceftriaxone [5]. Although serotypes 3 and 19A were still prevalent in Korea, their susceptibility to ceftriaxone was clearly different. Most serotype 3 isolates were susceptible to ceftriaxone (93.8%, 15/16), whereas more than two-thirds (81.8%, 9/11) of serotype 19A isolates were not susceptible to ceftriaxone. Among serogroup 11, serotypes 11A and 11E and serotype 19A pneumococci were more likely to cause ceftriaxone non-susceptible infections [4, 5]. Serotype 10A was the most common serotype in children (≤5 years old, 32.6%), and most of these isolates were collected from children ≤1 year old (85.7%) in prospective IPD study in 16 hospitals in Korea between 2017 and 2019 [8]. Serotype 3 is most prevalent in adults. Serotype 35B was common in those ≥65 years old in Korea, and was closely related to levofloxacin resistance, as previously reported, whereas this serotype is highly prevalent in children in the United States [59]. An increase in serotype 3 was observed in patients aged >50 years, especially in those ≥65 years old, whereas it was hardly ever found in children. This change may be resulted from PCV13 use as the NIPs were provided only for children, and may be associated with lower immunogenicity of the PPSV for serotype 3.

Despite increasing resistance rates, ceftriaxone and cefotaxime are still recommended as first-line treatment regimens for pneumococcal disease [60]. The emergence of high-level resistance to cephalosporins may limit treatment options and raise concerns about clinical outcomes, necessitating further research on clinical outcomes with a focus on extended-spectrum cephalosporins.

In summary, I discussed the major changes of epidemiology, risk factors, and antibiotic resistance of IPD in Korean adults after introduction of PCV10/PCV13 and PPSV23 into the NIP. One of the deadliest infectious diseases known to humans, smallpox is the only human infection that has been eradicated through effective vaccination. IPD is also a representative disease in which a significant number of infectious diseases can be prevented through vaccination, but it is difficult to eradicate IPD with vaccines alone. The incidence of IPD have been decreasing since the introduction of PCV, but like the balloon effect, the incidence of IPD due to serotypes not included in the vaccine is gradually increasing after the expansion of vaccination. National surveillance data show that the incidence of IPD in Korea has been increasing in recent years. Additionally, there has been a notable trend of increasing resistance to extended-spectrum cephalosporins (e.g., ceftriaxone/cefotaxime) in Korea and other regions. In many previous studies, mortality was significantly associated with host risk factors such as extreme age (>80 years) and presence of underlying diseases, but not with bacterial factors such as serotype or antibiotic resistance. Although there has not yet been research on the impact of high-level resistance to ceftriaxone or cefotaxime on mortality, the number of strains with non-susceptibility to cephalosporins is increasing. Proper management of IPD and application of new pneumococcal vaccines to diverse vulnerable hosts will require continuous monitoring of incidence, serotypes, and antibiotic resistance.