Semin Respir Crit Care Med 2016; 37(06): 839-854
DOI: 10.1055/s-0036-1592075
Review Article
Thieme Medical Publishers 333 Seventh Avenue, New York, NY 10001, USA.

# Community-Acquired Pneumonia in the Asia-Pacific Region

› Author Affiliations
Further Information

### Address for correspondence

Jae-Hoon Song, MD, PhD
Division of Infectious Diseases, Department of Medicine, Samsung Medical Center
Sungkyunkwan University School of Medicine, 81 Ilwon-ro, Gangnam-gu, Seoul
Korea

### Publication History

Publication Date:
13 December 2016 (online)

### Abstract

Community-acquired pneumonia (CAP) is an important cause of mortality and morbidity worldwide. Aging population, dense urbanization, and poor access to health care make the Asia-Pacific region vulnerable to CAP. The high incidence of CAP poses a significant health and economic burden in this region. Common etiologic agents in other global regions including Streptococcus pneumoniae, Mycoplasma pneumoniae, Haemophilus influenzae, Chlamydophila pneumoniae, Staphylococcus aureus, and respiratory viruses are also the most prevalent pathogens in the Asia-Pacific region. But the higher incidence of Klebsiella pneumoniae and the presence of Burkholderia pseudomallei are unique to the region. The high prevalence of antimicrobial resistance in S. pneumoniae and M. pneumoniae has been raising the need for more prudent use of antibiotics. Emergence and spread of community-acquired methicillin-resistant S. aureus deserve attention, while the risk has not reached significant level yet in cases of CAP. Given a clinical and socioeconomic importance of CAP, further effort to better understand the epidemiology and impact of CAP is warranted in the Asia-Pacific region.

#

The Asia-Pacific region in this article includes East Asia, Southeast Asia, South Asia, and Oceania. Asia-Pacific is home to more than 4.4 billion people, which is “nearly 60 percent of the world's population.”[1] While collectively categorized as Asia-Pacific, considerable diversity exists. Seven of the world's 10 most populous countries are located in this region, as well as some of the world's smallest countries. Some have leading economies of the world, while some are struggling to meet the most basic needs of their people. Despite its large heterogeneity, the region shares some distinct characteristics. Population growth rate is declining (0.9% per year), infant mortality rate is still high (124 deaths per 100,000 live births), proportion of older adults are growing (12.1% of population are aged 60 and above), a large population is living in urban areas (48%), and some of the world's largest megacities are located in Asia-Pacific.[1] These characteristics are associated with the high burden of community-acquired pneumonia (CAP) in this region, probably taking considerable toll on its population, economy, and societies.[2] Information on the epidemiology of CAP in this region is limited by multiple hurdles: poor accessibility to health care,[3] [4] lower utilization of microbiologic diagnosis,[5] lack of surveillance systems, and considerable heterogeneity among different geographic areas. Aging population, high population density, and high use of antibiotics are likely to result in increased incidence of CAP, in particular by less susceptible pathogens. In this review, we will describe the epidemiology, etiology, antimicrobial resistance, preventive measures, and outcomes of CAP in the Asia-Pacific region.

### Epidemiology of CAP in the Asia-Pacific Region

Lack of surveillance in many countries and discrepancies in the surveillance methods make the accurate estimation of the burden of CAP in the region very difficult ([Table 1]).[6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] Overall incidences of pneumonia and pneumonia-attributed mortality rates were recently estimated from a multicenter prospective surveillance in Japan from 2011 to 2013.[9] The estimated annual incidence rates of adult community-onset pneumonia, hospitalization, and inhospital deaths were 1,690, 530, and 70 per 100,000 person-years, respectively. The overall estimated annual number of adult CAP cases in the entire Japanese population was 1,880,000; importantly, 69.4% were ≥65 years old. A prospective study in one Japanese city (Kochi) from May 2008 to April 2010 cited an incidence of 960 per 100,000 person years; 73.3% of cases were ≥65 years old.[10] More attention was given to the geriatric population in the latter study, in which annual incidence of CAP in older adults (≥75 years) was estimated to be 4,290 per 100,000. An Asian country that showed similar socioeconomic and ethnic characteristics to Japan would be South Korea. Although the exact nationwide overall incidence of CAP in Korea has not been reported, hospitalization rate was reported to be similar (520 per 100,000).[16] This study confirmed the significantly larger burden of CAP in the elderly population by estimating that hospitalization rate in people ≥ 75 years of age was 2,030 per 100,000 population. Several reports have been published on the disease burden of CAP in Thailand, including both urban and rural areas. The hospitalization rate due to CAP was reported to be 177 to 580 per 100,000 people in Thailand, which was lower than those of Japan and South Korea.[22] [24] Because these studies were limited by low utilization of chest X-rays and variable access to health care, the estimates might not reflect the whole picture of CAP in Thailand.[24] Reports on the Pacific Island countries are scarce, but two studies from New Zealand reported the overall incidence in adults and hospitalization rate. Estimated incidences of CAP in all adults and those ≥65 years of age in New Zealand were 859 and 1,882 per 100,000, respectively.[11] Another study estimated that pneumonia hospitalization rate was 92 per 100,000.[12] Several studies on the burden of CAP among children in Southeast Asia have also been published. A surveillance in rural villages on Lombok Island, Indonesia, in 1998 to 1999 reported the incidence, hospitalization rate, and mortality among young children (≤2 year) as 21,000, 5,300, and 3,300 per 100,000 child-years, respectively.[8] Another interview survey in the Philippines revealed similar incidence rates: 10,500 cases of pneumonia, 6,100 admissions, and 90 deaths per 100,000 children each year.[15] The burden of CAP was much smaller in Taiwan (1,240 pneumonia cases per 100,000)[20] and New Zealand (500 hospitalizations per 100,000).[13] Rudan et al conducted an estimation of the global incidence of childhood pneumonia, in which the annual incidences in Southeast Asia and Western Pacific regions were estimated to be 36,000 and 22,000 per 100,000 children, respectively.[18] When specific demographic groups were studied for CAP, a larger burden was almost always observed in older adults,[6] [7] [10] [16] [22] those residing in rural areas,[6] [7] [8] [15] and minority ethnicity.[6] [13] Most studies on the epidemiology of CAP in the Asia-Pacific region are from either nationwide mortality statistics or surveillance in geographically limited areas. Differences in case definitions and potential underreporting due to limited accessibility to health care undermine the effort to measure the burden of CAP. Further studies on the epidemiology of CAP are warranted in the Asia-Pacific region based on more coordinated plans and resources.

Table 1

### Incidence of community-acquired pneumonia in the Asia-Pacific region (per 100,000 population unless noted otherwise)

Country

Year

Reference

Population

Incidence

Hospitalization

Mortality

Notes

Australia

1988–1993

Williams et al[6]

All

17 (nonaboriginal urban young adults) and 4,035 (aboriginal rural old adults)

China

1991–2000

He et al[7]

All

43.9

Mortality surveillance

Indonesia

1998–1999

Sutanto et al[8]

Children (<2 y)

21,000

5,300

3,300

Rural areas showed higher incidence and mortality

Japan

2011–2013

Morimoto et al[9]

1,690

530

70

Japan

2008–2010

Takaki et al[10]

960, 4,290 (≥ 75 y)

New Zealand

2000–2002

Scott et al[11]

859

New Zealand

1999–2000

Chambers et al[12]

92

New Zealand

1993–1996

Grant et al[13]

Children (<15 y)

500

Pacific Islanders (1,400) and Maori (670) have higher incidence compared with Europeans/other (270)

Pakistan

2002–2003

Nizami et al[14]

Children (<5 y)

8,210

Philippines

2011–2012

Kosai et al[15]

Children (<5 y)

10,500

6,100

90

South Korea

2002–2005

Kim et al[16]

All

520 (all), 2,030 (≥75 y)

Influenza included

South Korea

2012

Lim et al[17]

All

20.8

Southeast Asia and Western Pacific

Estimate

Rudan et al[18]

Children (<5 y)

30,000

Taiwan

1994 (estimate)

Leung et al[19]

All

3.71–6.39%

Taiwan

1997–2004

Wu et al[20]

Children (<18 y)

1,240

6.7 (<5 y)

Thailand

2010

Reechaipichitkul et al[21]

9.63%

Thailand

2004–2006

Prapasiri et al[22]

All

199–256

6.9

Thailand

2003–2009

Aungkulanon et al[23]

All

20–25

Thailand

2002–2003

Olsen et al[24]

All

177–580

Thailand

1999–2001

Kanlayanaphotporn et al[25]

All

211

Thailand

2010

Teeratakulpisarn et al[26]

Children (<5 y)

11.29

#

### Etiologic Pathogens of CAP in the Asia-Pacific Region

Distribution of etiologic agents of CAP is the most important information for the selection of appropriate antibiotics. It has been known that major identifiable pathogens of CAP include Streptococcus pneumoniae, Haemophilus influenzae, Mycoplasma pneumoniae, Chlamydophila pneumoniae, and Legionella spp.[27] [28] [29] The last three have been referred to as “atypical pathogens,” the importance of which has been the subject of considerable debate.[29] Despite the importance of this subject, the majority of the data on the etiology of CAP have been reported from the U.S. and European countries. But, the number of studies from the Asia-Pacific region is recently increasing; they are summarized in [Table 2]. The most comprehensive data were reported by the Asian Network for Surveillance of Resistant Pathogens (ANSORP) in 2007.[27] In this study, a total of 955 cases with CAP were collected from seven countries (South Korea, China and Hong Kong SAR, India, Singapore, Vietnam, Taiwan, and the Philippines). Streptococcus pneumoniae was the most common isolate, which comprised 29.2% of identified pathogens. Pneumococcus was followed by Klebsiella pneumoniae (15.4%), H. influenzae (15.1%), C. pneumoniae (13.4%), and M. pneumoniae (11.0%). The overall distribution of etiologic pathogens from this study was comparable to those from western countries.[30] [31] [32] [33] Streptococcus pneumoniae was the most frequent pathogen identified in other studies from Japan,[34] [35] [36] [37] [38] South Korea,[39] [40] Taiwan,[41] [42] Australia,[43] and New Zealand.[44] However, the proportion of pneumococcus showed considerable variability, from 10%[37] [40] [43] [44] [45] to 25%.[46] The broad range seen in these studies could be attributed to variable detection rate in addition to the actual difference in the pathogen distribution. With regard to atypical pathogens, some reports from China,[47] [48] [49] Taiwan,[42] [45] and Thailand[50] reported a relatively more important role of these pathogens in CAP.

Table 2

### Etiology of CAP in the Asia-Pacific region. Note: Detection rates of pathogens are shown as percentage (%).

Country

Year

Reference

Age

No.[a]

Methods

Virusb

Sp

Hi

Kp

Mp

Cp

Mc

Sa

Lp

Bp

Pa

Ab

Flu A

Flu B

RSV

PIV

HRV

BoV

MPV

Asiac

2001–2002

Ngeow et al[5]

All (≥2)

1,374

PCR and serology

No

12.2

4.7

6.6

Australia

2004–2006

Charles et al[43]

885

Culture, PCR, serology

Yes

13.9

5.1

8.8

1.7

0.8

1.2

3.4

1.6

7.7

1.9

5.2

Australia

2005–2007

Rémond et al[111]

293

Culture

No

13

18.2

3.1

4.2

2.6

1

1

Cambodia

2007–2009

Vong et al[127]

All (>5)

959

Culture and PCR

Yes

2.2

5.4

2.9

0.5

2.6

2

1.7

2.4

9.1

0.7

China

2001–2003

Huang et al[47]

All (≥2)

389

Culture, PCR, serology

No

3.1

20.6

3.9

10.8

4.4

0.3

1.5

0.5

China

2002–2004

Song et al[27]

225

Culture and serology

No

9

5

China

2003–2004

Liu et al[48]

610

Culture and serology

Yes (n = 184)

6.1

5.4

3.8

13.4

4.8

0.8

2.8

2.8

0.8

3.3

1.1

1.6

China

2004–2005

Liu et al[ 88]

1,193

Culture and serology

No

8.5

5.2

6.5

4.6

0.1

China

2006

Zhang et al[49]

All

610

N/A

No

6.1

5.4

3.8

13.4

4.8

2.8

2.8

China

2009–2013

Wei et al[157]

Children (≤16)

3181

Culture and PCR

Yes

14.4

4.3

3.7

14.9

35

24.9

12.3

2010–2012

Liu et al[158]

Children (≤15)

39,756

IFA

Yes

19.1

0.1

0.4

0.2

4.7

2

1.4

4.8

China

2010–2012

Wu et al[159]

Children (≤16)

10,435

Serology

Yes

56.9

0.2

1.6

2

35.4

18.9

7.5

4.9

China

2011–2013

Chen et al[160]

Children (4–14)

1,204

Serology

Yes

40.8

0.3

0.9

0.08

7.06

3.32

4.82

1.08

India

2002–2004

Song et al[27]

104

Culture and serology

No

10

2

8

India

2013

Acharya et al[55]

100

Culture

No

31

5

13

8

8

15

Indonesia

2007–2009

Farida et al[58]

148

Culture, PCR, serology

Yes

3

3

1

3

1

3

8

3

1

3

4

1

1

Japan

1994–1997

Ishida et al[34]

326

Culture and serology

Yes

23

7.4

4.3

4.9

1.8

2.1

0.6

2.5

0.3

0.9

0.3

0.6

Japan

1998–2000

Kawai et al[87]

231

Culture and serology

No

9.1

11.6

5

6.6

1.7

1.7

10.4

3.3

Japan

1998–2003

Miyashita et al[35]

506

Culture, IFA, serology

No

23.3

11.3

1.6

13.0

7.7

3.2

2.8

1.2

1.6

Japan

1998–2003

Miyashita et al[36]

200

Culture and serology

No

20.5

11

2.5

9.5

7.5

3

5

1

2

Japan

1999–2000

Saito et al[46]

232

Culture, PCR, serology

Yes

24.6

18.5

1.3

5.2

6.5

2.2

3.4

3.9

0.4

13.4

0.4

0.9

1.2

Japan

2000–2002

Motomura et al[37]

124

Culture and serology

No

12.1

8.0

2.4

3.2

2.4

Japan

2001–2004

Ishida et al[38]

349

Culture and serology

No

23.8

6

1.4

11.2

3.4

1.7

1.4

1.4

1.1

Japan

2011–2013

Morimoto et al[9]

1,772

Culture and PCR

Yes

9

10

6

8

5

4

9

2

Malaysia

1997–1999

Liam et al[113]

Mixed (≥12)

127

Culture and serology

No

5.5

5.5

10.2

3.9

1.6

1.6

3.9

Malaysia

2002–2003

Loh et al[112]

Mixed (≥12)

80

Culture

No

17.8

2.7

4.1

Malaysia

2006?

Liam et al[53]

Mixed (≥12)

346

Culture and serology

No

4

3.5

10.7

9

4

4

5.8

0.6

2.9

0.9

Malaysia

2009–2010

Mustafa et al[52]

46

Culture and PCR

No

21.7

2.1

17.3

6.5

4.3

2.1

13

6.5

2.1

New Zealand

1999–2000

Laing et al[44]

474

Culture and serology

Yes

14

10

3

1

1

2

4

1

7

2

3

2

2

Philippines

2002–2004

Song et al[27]

55

Culture and serology

No

11

20

11

Singapore

2002–2004

Song et al[27]

96

Culture and serology

No

6

3

Singapore

2006?

Chiang et al[161]

Children (≤16)

1,702

Culture, PCR, serology

Yes

6.6

2.4

20.6

0.2

0.4

0.3

1.5 (A and B)

5.8

1.5

0.7

1.5

South Korea

2001–2002

Sohn et al[39]

126

Culture, PCR, serology

No

13.5

0.8

3.2

6.3

7.1

0.8

2.4

3.2

3.2

South Korea

2002–2004

Song et al[27]

338

Culture and serology

No

14

1

3

South Korea

2007–2008

Jeon et al[40]

Elderly (>60)

63

Culture and serology

No

12.0

4.0

7.4

1.1

5.1

2.3

Taiwan

2001–2002

Lauderdale et al[41]

168

Culture and serology

Yes

23.8

4.8

4.8

14.3

7.1

1.8

1.2

6.5

1.2

1.2

1.2

Taiwan

2001–2002

Yen et al[42]

100

Culture and serology

No

26

9

5

20

13

2

1

3

Taiwan

2002–2004

Song et al[27]

65

Culture and serology

No

14

2

14

Taiwan

2007

Wu et al[51]

All

933

Culture

No

5.9

7

24.7

9.7

10.2

5.2

Taiwan

2007–2008

Lee et al[45]

156

Culture and serology

No

11.2

11.2

13.1

11.9

6

Thailand

1998–2001

Wattanathum et al[50]

245

Culture and serology

No

18.8

2.0

5.7

15.9

24.5

0.0

2.0

6.5

0.8

0.4

0.8

Thailand

2001–2002

Reechaipichitkul et al[54]

254

Culture and serology

No

11.4

4.3

10.2

3.9

8.7

0.8

3.5

11

2.4

Thailand

2001–2002

Prapphal et al[89]

All (≥2)

292

PCR and serology

No

14

3.4

0.4

Thailand

2005–2010

Hasan et al[162]

Children (<5)

28,543

Culture, PCR, serology

Yes

6.2

2

19.5

9.1

18.7

3.5

12.8

Vietnam

2002–2004

Song et al[27]

72

Culture and serology

No

11

11

3

Vietnam

2009–2010

Takahashi et al[59]

154

Culture and PCR

Yes

23

27

2

2

4

3

6

3

1

5

2

Abbreviations: Ab, Acinetobacter baumannii; AdV, adenovirus; BoV, bocavirus; Bp, Burkholderia pseudomallei; CAP, community-acquired pneumonia; Cp, Chlamydophila pneumoniae; Hi, Haemophilus influenzae; HRV, human rhinovirus; IFA, immunofluorescence assay; Kp, Klebsiella pneumoniae; Lp, Legionella pneumophila; Mc, Moraxella catarrhalis; Mp, Mycoplasma pneumoniae; MPV, metapneumovirus; N/A, not available; Pa, Pseudomonas aeruginosa; PCR, polymerase chain reaction; PIV, parainfluenza virus; RSV, respiratory syncytial virus; Sa, Staphylococcus aureus; Sp, Streptococcus pneumoniae.

a Number of patients included. bInclusion of testing for respiratory virus. cMulticenter study including China, South Korea, Taiwan, Thailand, Indonesia, Malaysia, and Singapore. Only atypical pathogens were tested for. dTest for S. pneumoniae was not performed.

Some pathogens are worth attention due to their unique importance in the Asia-Pacific region. Klebsiella pneumoniae, which is relatively uncommon in other regions, contributes to larger cases of CAP in Southeast Asia. This is especially evident in studies from Taiwan,[27] [45] [51] Malaysia,[52] [53] Thailand,[54] India,[55] and the Philippines[27] (all >10% of total CAP cases), which showed a stark contrast to East Asian countries (usually ≤5%). Another important pathogen in this region is Burkholderia pseudomallei, which is endemic in Southeast Asia and often results in severe infections. It was detected in 13% of hospitalized CAP patients in Malaysia[52] and in 11% of patients with severe CAP in Thailand.[56] In another study from Singapore, which also focused on the patients who required intensive care unit admission, B. pseudomallei was isolated from 10% of the cases.[57]

There have been only small number of surveillance studies in which the burden of viral infections in CAP were reported. We could find 16 studies for our review, which are summarized in [Table 2]. The prevalence of respiratory virus varied from 1.8 to 21%. Most of the studies used polymerase chain reaction (PCR) for the detection of virus, but serologic tests were applied in some reports. Methods used for the detection seems to result in variable results, as different studies from same countries often showed vastly different detection rates. Proportion of viral pathogens among CAP was only 1.8% in an earlier report from Japan that used serologic test,[34] but virus was identified in 20% among 1,772 patients in a recent study from Japan using PCR.[9] In adult population, influenza A and B viruses seem to predominate, comprising 5 to 15% of pathogens detected including bacteria.[9] [41] [43] [44] [46] [58] [59] Rhinovirus, which is increasingly identified as etiologic agents of CAP in adults, was the second most commonly detected virus (4–9%).[60]

Multiple limitations hinder the effort to elucidate the etiologic agents of CAP, including suboptimal quality of respiratory specimens, difficulty to culture certain species of bacteria, interpretation of commensal bacteria detected in patients with CAP, ambiguous results of serologic tests, and methods for detection of virus. PCR, often performed in multiplex, enabled sensitive and accurate detection of respiratory pathogens, and recent studies using this technique are broadening our understanding of the pathogens causing CAP.[31]

#

### Specific Pathogens of Community-Acquired Pneumonia

There are a couple of specific pathogens of CAP that are unique or notable in the Asia-Pacific region with regard to the incidence, antimicrobial resistance, clinical features, or clinical outcomes.

#### Streptococcus pneumoniae

The importance of S. pneumoniae as a major pathogen causing CAP remains unchallenged in the Asia-Pacific region, as discussed previously. Furthermore, the high prevalence rate of antimicrobial resistance in pneumococci in this region is a very serious threat to public health. Important data on the antimicrobial resistance of S. pneumoniae in the Asia-Pacific are summarized in [Table 3]. The most prominent resistance issue is macrolide resistance.[61] Two surveillance studies on pneumococcus conducted by ANSORP in the early 2000s revealed that about half of the isolates were resistant to erythromycin.[27] [62] Considerable variability does exist between different countries, and resistance rates in China, Korea, Taiwan, and Vietnam exceed 70% with MIC90 (minimal inhibitory concentration, MIC, for 90% of the isolates) of >128 mg/L.[62] High resistance rates in East Asian countries were confirmed by Prospective Resistant Organism Tracking and Epidemiology for the Ketolide Telithromycin (PROTEKT) study.[63] In this international surveillance study in 40 countries, macrolide-resistant S. pneumoniae was most prevalent in Far East countries (China, Japan, South Korea, and Taiwan) with resistance rates ≥ 80%. While Southeast Asian countries were not included in this study, the resistance rate in Australia was comparable to countries in Northern Europe and America at <30%. Some Asian countries are showing the increasing resistance trend over time. A report from Sri Lanka, in which the resistance rate had been reported to be 16.7% in ANSORP studies in the early 2000s, showed that the resistance rate increased to 60.9% in the late 2000s.[64] Notable exceptions are India and the Philippines, where only <20% of the organisms were reported to be resistant to macrolides in ANSORP studies, although current status should be investigated. Since the macrolide MIC level in pneumococci from some Asian countries is too high to be achieved by increased dose of macrolides, single empiric therapy with macrolides for the treatment of CAP is generally not recommended.[65] [66]

Table 3

### Antimicrobial resistance or nonsusceptibility rates (%) of Streptococcus pneumoniae in the Asia-Pacific region

Country

Year

Reference

MIC breakpoint for penicillin (mg/L)

Antibiotic class

Penicillin

Amox/clav[a]

Ceftriaxone

Erythromycin

Levofloxacin

ANSORPb

2002–2004

Song et al[27]

2

35.1

3.5

7

56.1

0

ANSORPb

2008–2009

Kim et al[72]

8/4

0.7/4.6

3.7

China

1980–2008

Chen et al[163]

2

15.6

3.3

5

81.7

Chinab

2000–2001

Song et al[62]

2

23.4

7.3

1.8

73.9

China

2001–2003

Huang et al[47]

2

0

50

0

China

2003–2004

Liu et al[48]

2

3.2

1.6

79.4

0

Chinab

2008–2009

Kim et al[72]

8/4

2.2/13.2

8

Hong Kongb

2000–2001

Song et al[62]

2

43.2

3.6

0

76.8

Hong Kongb

2008–2009

Kim et al[72]

8/4

0/1.5

6.6

India

1993–2008

Thomas et al[164]

4

2.7

< 20

Indiab

2000–2001

Song et al[62]

2

0

0

0

1.3

Indiab

2008–2009

Kim et al[72]

8/4

0/0

0

Japan

1999–2004

Inoue et al[73]

2

30.9–44.5

0

77.2–81.9

1.0–1.3

Japan

2001–2003

Qin et al[74]

2

22.8

0

80.7

1.8

Japan

2003–2004

Ishida et al[75]

8/4

0/0

0

0.7

83.7

3.5

Japan

2003–2005

2

21.7

Japanb

2008–2009

Kim et al[72]

8/4

0/0

0

South Koreab

2000–2001

Song et al[62]

2

54.8

9.7

3.2

80.6

South Koreab

2008–2009

Kim et al[72]

8/4

0.3/2.2

1.9

Malaysia

1999–2007

Le et al[166]

2

21.2

Malaysiab

2000–2001

Song et al[62]

2

29.5

0

2.3

34.1

Malaysiab

2008–2009

Kim et al[72]

8/4

0/0

0.7

Philippines

1994–2000

Sombrero et al[167]

2

0

0.2

Philippinesb

2000–2001

Song et al[62]

2

0

0

0

18.2

Philippinesb

2008–2009

Kim et al[72]

8/4

0/0

0.9

Saudi Arabiab

2000–2001

Song et al[62]

2

10.3

0

0

10.3

Singaporeb

2000–2001

Song et al[62]

2

17.1

0

0

40

Sri Lankab

2000–2001

Song et al[62]

2

14.3

0

0

16.7

Sri Lankab

2008–2009

Kim et al[72]

8/4

0/0

0

Taiwanb

2000–2001

Song et al[62]

2

38.6

1.8

0

86

Taiwan

2000–2001

Lee et al[168]

2

41.9–45.5

Taiwan

2001–2006

Hsieh et al[76]

1.2–2.5

Taiwan

2004–2006

Hsieh et al[169]

8/4

1.7/10.2

7.8

4.7

Taiwan

2007

Hsieh et al[76]

4.2

Taiwanb

2008–2009

Kim et al[72]

8/4

0/0.4

1.3

Taiwan

2009–2012

Lee et al[77]

2

39.4

13.8

90.8

1

Thailand

1998–2001

Sangthawan et al[170]

2

4.3

4.3

34.8

Thailandb

2000–2001

Song et al[62]

2

26.9

0

0

36.5

Thailandb

2008–2009

Kim et al[72]

8/4

0/0.5

0

Vietnamb

2000–2001

Song et al[62]

2

71.4

22.2

3.2

92.1

Vietnam

2007

Hoa et al[171]

8/4

4/36

4/36

70

Vietnamb

2008–2009

Kim et al[72]

8/4

0/0.9

1.8

Abbreviations: ANSORP, Asian Network for Surveillance of Resistant Pathogens; MIC, minimal inhibitory concentration.

a Amoxicillin/clavulanic acid. bMultinational surveillance study conducted by ANSORP, including South Korea, China, Taiwan, India, Singapore, Vietnam, and the Philippines.

High resistance rates of pneumococcus to penicillin had raised concerns in the Asia-Pacific region.[62] [67] But after subsequent reports showing comparable clinical outcomes in infections caused by S. pneumoniae with MIC ≤ 0.06 mg/L and those with 0.06 to 2 mg/L,[68] [69] [70] the penicillin MIC breakpoint for resistance in pneumococcus was revised from 2 to 8 mg/L in nonmeningeal isolates by the Clinical and Laboratory Standards Institute in 2008.[71] Since the MIC breakpoints used for the determination of penicillin resistance varied by studies, we selected the data using penicillin MIC breakpoint of 2 mg/L as resistance in [Table 3] to compare the temporal trend of penicillin resistance. Aforementioned ANSORP studies reported that 30 to 35% of the pneumococcus isolates in this region were resistant to penicillin.[27] [62] Like macrolide resistance, higher penicillin-resistance rates were observed in East Asian countries, while those in Southeast and South Asian countries were considerably lower. The most comprehensive multinational surveillance study based on revised criteria had been conducted by ANSORP in 2008 to 2009, which revealed that the resistance rate according to the revised MIC breakpoint (8 mg/L) was only 0.7%.[72] It indicates that the resistance of pneumococci to penicillin is not a serious threat, at least in nonmeningeal infections treated with intravenous antibiotics. Fluoroquinolone resistance has been reported to be <5% in most countries.[27] [47] [48] [73] [74] [75] [76] [77] The PROTEKT international surveillance study also showed the low resistance rates (<3%) in all countries except Hong Kong (14.3%),[78] which was likely to due to the dissemination of fluoroquinolone-resistant variant of the Spain23F-1 clone.[79] Spread of unrelated resistant clone was also reported from Taiwan.[76]

Widespread vaccination against pneumococci has significantly affected the incidence of pneumococcal diseases, serotype distribution, and antimicrobial resistance. Introduction of the 7- and 13-valent pneumococcal conjugate vaccine (PCV-7 and PCV-13) led to dramatic reduction of pneumococcal infections in the United States. As serotypes included in PCV-7 are often associated with penicillin and multidrug resistance, the incidence of antibiotic-resistant invasive pneumococcal infections has also declined.[80] However, there have been reports of the emergence of pneumococcal infections by nonvaccine types, especially 19A, which is often multidrug-resistant.[81] [82] [83] Two recent studies on adults from Japan during 2010 to 2013 reported that PCV-7 serotypes, especially 6B, decreased from 43.3 to 23.8%.[84] [85] Some nonvaccine serotypes emerged, but genotypic penicillin resistance rate declined.[84] Nonetheless, the notable emergence of invasive pneumococcal infections caused by 19A was reported in Taiwan, which was associated with reduced susceptibility to β-lactams.[77] Serotype 19A was also the most prominent non-PCV-7 serotype in the latest ANSORP study, comprising 8.2% of isolates, while 52.5% showed PCV-7 serotypes. The majority of serotype 19A isolates were erythromycin-resistant (86.0%) and multidrug-resistant (79.8%).[72] Most prevalent clone among serotype 19A was ST320 (51.1%), which was found in Hong Kong, India, South Korea, Malaysia, Saudi Arabia, and Taiwan.[86] High prevalence of serotype 19A with multidrug resistance, even in countries with low vaccination rate, needs to be carefully evaluated.

#

#### Atypical Pathogens: Mycoplasma pneumoniae, Chlamydophila pneumoniae, and Legionella species

The most comprehensive study on the role of atypical pathogens in the Asia-Pacific region was reported by Ngeow et al in a multicenter surveillance study on the prevalence of atypical pneumonia in the early 2000s.[5] They used serology and PCR to detect M. pneumoniae, C. pneumoniae, and L. pneumophila from 1,374 patients in 8 countries (Malaysia, Thailand, China, the Philippines, Taiwan, South Korea, Singapore, and Indonesia). These three atypical pathogens were associated with 23.5% of CAP cases in this study, with M. pneumoniae, C. pneumoniae, and L. pneumophila detected in 12.2, 4.7, and 6.6% of patients, respectively. The ANSORP study also showed that atypical pathogens account for 25.5% of the cases in which serologic tests were performed.[27] But C. pneumoniae (13.4%) and M. pneumoniae (11.0%) were detected in the comparable proportion of CAP patients in this study. Studies from individual countries reveal considerable differences in the burden of atypical pathogens within the region. In general, studies from Japan[34] [36] [37] [46] [87] and South Korea[39] have reported lower proportion of atypical pathogens in CAP, accounting for less than 10% in CAP cases ([Table 2]). However, studies from China,[47] [48] [49] Taiwan,[42] [45] and Thailand[50] showed the larger role of atypical pathogens. In a multicenter prospective study conducted at 12 centers in seven Chinese cities, atypical pathogens accounted for 31.3% of the cases with CAP when fourfold increase in titers of antibodies were defined as serologic evidence of the infection: M. pneumoniae (13.4%) was the single most prevalent pathogen, followed by S. pneumoniae (6.1%), H. influenzae (5.4%), and C. pneumoniae (4.8%).[48] Legionella pneumophila was detected in only 2.8% of the patients, which is in accordance with other studies in Asia.[5] [27] Other studies from China, including one which used PCR, have reported the similar proportion of atypical pathogens (M. pneumoniae: 10%; C. pneumoniae: 4–5%).[27] [49] Another report from Hong Kong on 1,193 adult patients with CAP requiring hospitalization showed 6.5% for M. pneumoniae, but still two atypical pathogens accounted for 28% of the patients in whom the etiologic agents were identified.[88] Two studies from Taiwan, which included adult CAP patients who were hospitalized, reported that the serologic evidence of M. pneumoniae and C. pneumoniae infection was found in 13.1 to 14.3% and in 7.1 to 11.9%, respectively.[41] [45] Similar distribution and higher prevalence in mild CAP was also shown in a multicenter, prospective study from Thailand in 1998 to 2001.[50] A more recent study at seven centers in Bangkok (2001–2002) using both PCR and serology reported comparable results.[89] Despite the existence of many studies on atypical pneumonia in this region, our understanding of the exact role of these pathogens is still inadequate. The diagnosis of CAP caused by atypical pathogens still mostly relies on the serologic tests, which requires serial testing and often yields equivocal results.[49] Further studies using molecular techniques can improve the correct understanding about the epidemiology of atypical pathogens.[90]

The most notable issue concerning M. pnuemoniae is the emergence of resistance to macrolides. As M. pnuemoniae harbors intrinsic resistance to β-lactams, most guidelines for the treatment of CAP recommend the inclusion of macrolides as empirical treatment regimen when the coverage for atypical pathogens are required.[28] [65] [66] [91] [92] But increased use of macrolides resulted in the emergence of erythromycin-resistant M. pneumoniae in the Asia-Pacific region. After the first report of macrolide-resistant M. pneumoniae in Japan from patients in 2000,[93] Matsuoka et al reported the isolation of 13 erythromycin-resistant strains (17%) among 76 M. pneumoniae strains isolated in Japan during 2000 to 2003.[94] All but one isolate harbored a point mutation (A2063G/C) in domain V of 23S rRNA gene, a binding site for macrolides. Another surveillance in Japan during 1976 to 2006 revealed that there were no resistant strains prior to 2000, yet the resistance rates were 14.6 and 21.6% in years 2000 to 2004 and 2005 to 2006, respectively.[95] The resistance rate in Japan increased further to approximately 45% in 2007 to 2008.[96] Macrolide-resistance strains were subsequently reported in China,[97] [98] [99] South Korea,[100] and Taiwan.[101] Reports from Beijing[98] and Shanghai[97] revealed remarkable resistance rates of 92% (46/50) and 83% (44/53), respectively. In the latter study, all strains isolated in 2007 and 2008 were resistant to macrolides. A Korean study with 378 isolates during 2000 to 2011 showed a similar picture; there were no resistant strains in 2000, but the resistance rate surged from 2.9% in 2003 to 62.9% in 2011.[100] A recent survey from Hong Kong reported the resistance rate of 47.1% (24/51) in 2014 and showed that the macrolide resistance was associated with increasing resistance in multilocus variable-number tandem-repeat analysis type 4-5-7-2.[102] A report from northern Taiwan showed that 12.3% of M. pneumoniae isolates were resistant to macrolides.[101] In contrast, only one strain was resistant among 30 specimens from Sydney, Australia.[103] Previous reports on the macrolide-resistant M. pneumoniae have been concentrated in three East Asian countries: Japan, China, and South Korea. The vast majority of macrolide-resistant M. pneumoniae strains found in Asia harbor point mutations on A2063 or A2064 in 23S rRNA gene. Mutations on A2063 or A2064 result in a high level of resistance to various macrolides, but do not affect the susceptibility to doxycycline or fluoroquinolones.[96] Information on the current status in other countries within the Asia-Pacific region is not available. Furthermore, the presence of a considerable regional difference in resistance rates within a single country has been reported.[104] Additional studies and enhanced surveillance are urgently warranted to clarify this issue. The clinical course of macrolide-resistant M. pneumoniae was reported to be prolonged; the duration of fever was 2 to 2.5 days longer and cough persisted for more than 4 days longer compared to patients with macrolide-susceptible M. pneumoniae infection.[101] [105] [106] [107] Efficacy of macrolide was reduced to 22.7% in cases with resistant strains compared with 91.5% in cases caused by susceptible strains.[106] Treatment with broad-spectrum tetracyclines (minocycline and doxycycline) or fluoroquinolones has been suggested, and two small-scale studies reported that minocycline or doxycycline was superior to fluoroquinolone in terms of the duration of fever after the initiation of treatment.[108] [109] Both classes of antibiotics have safety concerns in children (tooth discoloration and joint/cartilage toxicity, respectively), however, in whom M. pneumoniae infections are most prevalent. Use of fluoroquinolones is further complicated by its tendency to accelerate the emergence of antimicrobial resistance and the relatively high prevalence of tuberculosis in the region. As M. pneumoniae infections are often mild and self-limited, the conservative use of alternative agents other than macrolides only in severe or persistent cases was suggested.[110]

#

#### Klebsiella pneumoniae

Klebsiella pneumoniae accounts for approximately 6% of CAP cases in the ANSORP study,[27] while it is infrequently found in the Europe and Americas.[30] [31] [32] [33] Even within the Asian-Pacific regions, Australia, Vietnam, and East Asian countries report smaller incidence at ≤3%,[35] [36] [38] [39] [46] [47] [48] [49] [59] [111] while a recent report from South Korea showed a higher frequency of K. pneumoniae in elderly patients.[40] High burden of CAP caused by K. pneumoniae has been mostly seen in Taiwan (14–25%),[45] [51] Thailand (10.2%),[54] India (13%),[55] the Philippines (11%),[27] and Malaysia (10.2–17.8%).[52] [53] [112] [113] In a worldwide study on K. pneumoniae bacteremia, only 6% of community-acquired K. pneumoniae bacteremia were caused by CAP in the Europe and Americas.[114] In contrast, CAP was responsible for 29% of K. pneumoniae bacteremia in Taiwan, which made CAP the leading cause of bloodstream infection by this pathogen. According to a clinical study from Taiwan,[115] evaluating clinical outcome of bacteremic CAP caused by K. pneumoniae (49 patients) and S. pneumoniae (44 patients), mortality rate was significantly higher in patients with K. pneumoniae pneumonia (55.1 vs. 27.3%). High mortality rate was also reported in another study from Cambodia (37.5%).[116] Among 36 strains of K. pneumoniae tested for antimicrobial susceptibility in the 2008 ANSORP study, all but one were susceptible to ceftriaxone.[27] More recent data from Taiwan and Japan also suggest low resistance rate of community-acquired K. pneumoniae in this region, but further surveillance is warranted.[115] [117]

#

#### Burkholderia pseudomallei

Melioidosis, which is caused by B. pseudomallei, is an endemic infectious disease in Southeast Asia, Northern Australia, Southern China, and India.[118] Humans are infected by exposure to contaminated soil or surface water.[119] Incubation period is usually 3 to 14 days, but latency for decades has been reported.[120] Clinical manifestations have a broad spectrum, from asymptomatic infections to fulminant illness leading to death.[121] Approximately half of the patients with melioidosis present with pneumonia, which makes pulmonary infection the most common clinical presentation.[119] Among aforementioned studies, six reported the incidence of pneumonia caused by B. pseudomallei ([Table 2]). Malaysian study that used multiplex PCR for pathogen detection from 46 adult patients reported that B. pseudomallei accounted for 13% of CAP.[52] In this study, 83% were positive by PCR alone and only 17% were culture positive. Among 145 patients with CAP from Northern Thailand, B. pseudomallei was identified in 11%, which is slightly less frequent than S. pneumoniae (11.4%) but more frequent than K. pneumoniae (10.3%).[54] The annual incidence of bacteremic melioidosis was reported to be 4.6 and 14.4 cases per 100,000 persons in two Thailand provinces.[122] A study on the etiology of severe CAP in Singapore between 1989 and 1993 revealed that B. pseudomallei was identified in 10 cases among 48 patients.[57] The presence of endemic melioidosis has been also reported from Northern Australia,[123] [124] [125] [126] Cambodia,[127] [128] Hong Kong,[129] India,[130] Taiwan,[131] and Southern China.[132] Melioidosis has been associated with poor outcomes in multiple studies. A retrospective review from Royal Darwin Hospital in Australia reported that its mortality rate in 1989 to 1997 was 92%, although it was reduced to 26% in 1998 to 2013.[133] Mortality rate of 20% was reported from the aforementioned Northern Thailand hospital between 1996 and 2002.[134] Also, in a case series of 11 patients with imported melioidosis from South Korea, overall mortality rate was 36.4%.[135] Ceftazidime, sometimes in combination with cotrimoxazole, has been the treatment of choice during the initial intensive phase.[119] Burkholderia pseudomallei is highly susceptible to carbapenems in vitro, and imipenem or meropenem showed comparative outcomes to ceftazidime.[136] [137] After 2 to 4 weeks of initial intensive therapy, subsequent antimicrobial therapy for eradication of the bacteria should be followed using the combination of cotrimoxazole, doxycycline, and chloramphenicol for longer than 3 months.[138]

#

#### Staphylococcus aureus

Staphylococcus aureus is not a common etiologic agent of CAP, as it accounts for less than 5% of cases.[30] [32] [33] In the Asia-Pacific region, S. aureus has also been found in similar proportion, although there was a report of higher incidence of S. aureus in 10.4% of CAP cases in Japan ([Table 2]).[87] One of the remarkable issues regarding S. aureus is the emergence of community-associated methicillin-resistant S. aureus (CA-MRSA) with varying clinical syndromes and different strains during the last decade.[139] Most common presentation of CA-MRSA infections is skin and soft-tissue infection, but CA-MRSA can also cause severe CAP presenting as necrotizing pneumonia.[140] [141] [142] Since the emergence of CA-MRSA in Western Australia in the early 1990s,[143] numerous reports on small number of cases have been published from countries in the Asia-Pacific region.[144] [145] [146] [147] [148] But data on the prevalence of CA-MRSA causing CAP are lacking. A retrospective study from South Korea reported that S. aureus was isolated from 11.1% of cases with pathogens identified, and among them, two-thirds (6/9) were MRSA.[40] MRSA accounted for 4.3% of the cases with CAP in a Taiwanese multicenter study.[51] However, both studies did not examine the genotypic and phenotypic characteristics of MRSA isolates. In a report from South Korea that studied the community-onset sequence type 72 MRSA-SCCmec type IV infection, the predominant CA-MRSA clone in the country, showed that pneumonia was the focus of infection in 19% of the cases.[149] A multinational study conducted by ANSORP in 2004 to 2006 provided the most comprehensive information on the epidemiology of CA-MRSA in the Asia-Pacific region.[150] This multinational study collected 1,463 S. aureus isolates from various community-acquired infections, of which 373 (25.5%) were MRSA. Respiratory infection was the second most common type of infection (8.3%), following skin and soft-tissue infection (66.7%). Albeit more susceptible than hospital-acquired MRSA (HA-MRSA), CA-MRSA isolates from Asian countries also showed considerable resistance to gentamicin (61.2%), ciprofloxacin (52.5%), clindamycin (91.6%), tetracycline (69.3%), and trimethoprim/sulfamethoxazole (31.3%). There have been insufficient data on the exact incidence of CAP caused by CA-MRSA in this region. CA-MRSA pneumonia is often associated with poor clinical outcome, which emphasizes the importance of early appropriate treatment.[151] Therefore, further study on the epidemiology of CA-MRSA pneumonia in the Asia-Pacific region is of critical importance.

#
#

### Clinical Outcomes and Socioeconomic Burden of CAP in the Asia-Pacific Region

Studies published since 2000 on the mortality caused by CAP in the Asia-Pacific region are summarized in [Table 4]. Reported mortality rates varied between 1.1 and 30%, depending on the country, study population, and hospitalization. But with some exceptions, mortality rates were between 5 and 15%, while a recent study showed moderately improved outcomes compared with a previous review.[61] Although it is difficult to draw a conclusion from these limited data, countries with more advanced economy seem to show better outcomes with regard to pneumonia-specific mortality. Older age,[152] [153] [154] comorbidities,[27] [154] nursing home residence,[27] and poor performance status[152] [154] were associated with worse outcome, as in other regions of the world.

Table 4

### Mortality rates of adult patients with CAP in the Asia-Pacific region

Country

Year

Reference

No. of cases

Mortality rate (%)

ANSORP

2002–2004

Song et al[27]

955

7.3

Australia

2004–2006

Charles et al[43]

885

5.6

Australia

2005–2007

Rémond et al[111]

293

1.1

Indonesia

2007–2009

Farida et al[58]

148

30

Japan

1999–2002

Fujiki et al[154]

227

11.3

Japan

2012

Morimoto et al[9]

1,772

8

Malaysia

2002–2003

Loh et al[112]

108

12

New Zealand

1999–2000

Chambers et al[12]

474

6.1

South Korea

2007–2008

Jeon et al[40]

175

5.7

South Korea

2008–2010

Lee et al[153]

693

4.4 (age ≥ 65);

0.5 (age 50–65)

South Korea

2009–2011

Kim et al[172]

883

4.5

Taiwan

2001–2002

Lauderdale et al[41]

168

8.3

Taiwan

2007–2008

Lee et al[45]

208

13.9

Thailand

1998–2001

Wattanathum et al[50]

245

17.5

Thailand

2001–2002

Reechaipichitkul et al[54]

254

5.9

Thailand

2002–2003

Olsen et al[24]

777

9

Thailand

2004–2006

Prapasiri et al[22]

4,993

3

Vietnam

2009–2010

Takahashi et al[59]

174

9.8

Abbreviations: ANSORP, Asian Network for Surveillance of Resistant Pathogens; CAP, community-acquired pneumonia.

A relatively small number of studies on the economic burden of CAP have been performed in the Asia-Pacific region. A multicenter study from Korea over a decade estimated that the mean direct medical cost was US$7,452 per case, with no difference among age and risk groups.[152] In New Zealand, the direct medical cost was estimated at US$636 per episode, which would translate into the national cost of US$16.8 million.[11] The total annual cost, which includes direct and indirect medical cost and loss of productivity, was US$36.6 million. Chen et al conducted a study to evaluate the cost benefits of pneumococcal vaccination and, in the process, estimated the national cost of CAP in the elderly to be US$30 million each year.[155] Another study on the cost of CAP in China reported the median cost for hospitalization to be US$556.50.[156] Because direct and indirect costs caused by the medical condition are determined by multiple socioeconomic factors, direct comparison of the cost between countries is not appropriate. But the published data invariably revealed that the economic burden of CAP is quite significant, especially in countries with limited resources.

#

### Summary

The Asia-Pacific region shows its own landscape of CAP with regard to the incidence, etiologic pathogens, antimicrobial resistance, clinical outcomes, and socioeconomic burden of the disease, reflecting the diversity of the region. Since data on major issues of CAP need to be further collected in many countries in the region, appropriate and continuous surveillance of CAP is strongly warranted given the clinical and socioeconomic importance of the disease.

#
#

### Conflict of Interest

All authors have nothing to declare.

None.

### Funding

This study was partly supported by a grant from the Korean Health Technology R&D Project, Ministry of Health & Welfare, Republic of Korea (Grant No. HI12C0756), and partly by the Asia Pacific Foundation for Infectious Diseases.

### Address for correspondence

Jae-Hoon Song, MD, PhD
Division of Infectious Diseases, Department of Medicine, Samsung Medical Center
Sungkyunkwan University School of Medicine, 81 Ilwon-ro, Gangnam-gu, Seoul
Korea