Advances in Pneumococcal Antibiotic Resistance

Jae-Hoon Song


Expert Rev Resp Med. 2013;7(5):491-498. 

In This Article

Current Epidemiology of Pneumococcal Resistance

Macrolide Resistance

Resistance to macrolides and azalides is the most prominent problem of in vitro resistance in pneumococci in many parts of the world. The SENTRY Antimicrobial Surveillance Program in the United States showed that the rate of macrolide resistance has increased from 17.8% in 1998 to 44.8% in 2011.[12] In a survey performed between 2004 and 2005 in 15 European countries, macrolide resistance rate among isolates from patients with community-acquired respiratory tract infections was 24.6%, ranging from 6.9% in Norway to 57.1% in Greece.[13] However, many Asian countries showed much higher rates of macrolide resistance in S. pneumoniae isolates than in the western part of the world.[14–17] According to the Asian Network for Surveillance of Resistant Pathogens (ANSORP) study, the overall rate of erythromycin resistance in Asian countries significantly increased from 46.1% in 1996–1997 to 72.7% in 2008–2009.[14,16,18] In particular, very high resistance rates to macrolide were found in China (96.4%), Taiwan (84.9%), Vietnam (80.7%) and Korea (77.7%) in 2008–2009.[14]

The most common mechanism of macrolide resistance is either methylation of the 23S ribosomal target site, encoded by the ermB gene, which confers high-level resistance to macrolides (erythromycin minimum inhibitory concentrations [MICs]: ≥64 mg/L) or an efflux pump modified by the mef genes (mefA and mefE) that is associated with low-level resistance (erythromycin MICs: 1–32 mg/L).[19,20] Typically, the mefA-mediated low-level resistance was the most prevalent type of macrolide resistance in the United States, while ermB-mediated high-level resistance is more common type in European countries, South Africa and Asian countries.[21] In the United States, ermB-mediated resistance is also increasing recently, resulting in approximately equal in prevalence of mefA- and ermB-mediated resistance.[22] In most Asian countries, ermB was found in >50% of pneumococcal isolates either alone or in combination with mefA.[14,15] Recently, prevalence of macrolide-resistant pneumococci expressing both ermB and mefA has increased worldwide.[14,15,22,23] According to the ANSORP study, macrolide-resistant isolates expressing both genes have increased in Asian countries, particularly in Hong Kong (from 8.9% in 2000–2001 to 26.4% in 2008–2009), Taiwan (from 0% in 2000–2001 to 21.4% in 2008–2009) and Korea (from 38.6% in 2000–2001 to 43.3% in 2008–2009).[14,15] Of note, most isolates expressing both ermB and mefA from 2008 to 2009 in Asian countries were serotype 19F (61.3%), followed by serotype 19A (16.4%) and serotype 6A (9.8%).[14] Pneumococcal isolates carrying both ermB and mefA genes show resistance to multiple antimicrobials in addition to a high level of resistance to macrolides.[14,23,24]

Increasing prevalence of pneumococcal isolates with macrolide resistance is due to increased consumption of antimicrobial agents and the clonal spread of resistant strains. According to the European Surveillance of Antimicrobial Consumption project in 1998–2004 and antimicrobial surveillance study in 15 European countries in 2004–2005, macrolide use is significantly associated with increased prevalence of erythromycin resistance and MDR in S. pneumoniae.[13] The Prospective Resistant Organism Tracking and Epidemiology for Ketolide Telithromycin (PROTEKT) US study revealed that increase in macrolide-resistant pneumococci carrying both genes in the United States is related to the spread of specific clones of S. pneumoniae carrying both genes such as clonal complex 271 (CC271; Taiwan19F-14 clone), CC81 (Spain23F-1 clone), and CC242 (Taiwan23F-15 clone).[23] In particular, serotype 19A ST320 clone which belongs to CC271 has become dominant in the United States among macrolide-resistant pneumococci expressing both genes.[25,26] In Asian countries, CC271, which has been described as CC236 in the study by Ko and Song, is predominant in pneumococcal isolates containing both ermB and mefA genes.[27]

The prevalence of macrolide resistance has been persistently high in many parts of the world despite of PCV7 or PCV13 vaccination.[22,28] A recent surveillance study by the Canadian Bacterial Surveillance Network performed during 2000–2011 showed that erythromycin resistance in pneumococcal isolates has been steadily increasing both in isolates with PCV13 and non-PCV13 serotypes, since the introduction of PCV7 in 2001 and PCV13 in 2010.[29] The SENTRY Antimicrobial Surveillance Program in the United States from 1998 to 2011 also showed steady increase in macrolide resistance after the introduction of PCV7 in 2000 and PCV13 in 2010,[12] suggesting little impact of PCV7 or PCV13 on macrolide resistance in S. pneumoniae at the moment. Persistently high prevalence of macrolide resistance despite the use of PCV or even decreased use of macrolide antibiotics is due to the clonal spread of macrolide-resistant pneumococcal strains. However, epidemiology of macrolide resistance may change after the introduction of PCV13. According to the study by the United States Pediatric Multicenter Pneumococcal Surveillance Group, serotype 19A pneumococcal isolates has decreased since the introduction of PCV13, although PCV19A is still the most prevalent serotype among IPD in children in the United States.[30] Data suggested that PCV13 can prevent the clonal spread of macrolide-resistant pneumococci 19A strains.

Penicillin and β-lactam Resistance

Penicillin resistance is the most classic example of pneumococcal resistance, which had been first reported in 1967. The Active Bacterial Core Surveillance Program in the United States showed that the rate of penicillin non-susceptibility (intermediate resistance and resistance, ≥0.12 mg/L) has increased from 21.6% in 1996 to 25.9% in 2000 and after the introduction of PCV7, it has decreased again to 21.6% in 2004.[31] Another study performed in the United States using the Surveillance Network (TSN) database from 1996 to 2008 also showed fluctuating prevalence of penicillin resistance in pneumococci: 15.6% (1996), 23.2% (2000), 15.4% (2003) and 16.9% (2008).[28,32] The ANSORP study performed in 2000–2001 showed that many Asian countries particularly showed very high rates of penicillin resistance: Vietnam (71.4%), Korea (57.8%), Hong Kong (43.2%) and Taiwan (38.6%).[16,18] Despite the increasing prevalence of penicillin-resistant S. pneumoniae isolates based on the breakpoints for penicillin resistance (intermediate 0.12–1, resistant ≥2 mg/L) in many parts of the world, there was a lack of clinical correlation of in vitro penicillin resistance in nonmeningeal pneumococcal infections.[33] Therefore, the Clinical and Laboratory Standards Institute (CLSI) has revised the MIC breakpoints for penicillin resistance in nonmeningeal pneumococcal isolates in 2008 (susceptible ≤2, intermediate 4, resistant ≥8 mg/L).[33,34] According to the revised CLSI breakpoints, penicillin nonsusceptibility (intermediate and resistant, ≥4mg/L) rates of nonmeningeal isolates have been reported to be very low.[32,35–37] In Asia, prevalence rate of penicillin nonsusceptibility was 4.6% among nonmeningeal pneumococcal isolates during 2008–2009.[14] However, the SENTRY Antimicrobial Surveillance Program in the United States showed that the rate of penicillin nonsusceptibility has increased from 3.2% in 1998 to 11.7% in 2011, according to the revised breakpoints for penicillin.[12] Also, according to the analysis of pneumococcal isolates from 1996 to 2008 in the United States, pneumococcal isolates with MICs ≥4 mg/L have begun to increase since 2006.[32] Therefore, changing trends of penicillin resistance in S. pneumoniae should be continuously monitored.

Alteration of the cell wall penicillin-binding proteins (PBPs), resulting in decreased affinity for penicillin, is the main mechanism of penicillin resistance.[38] Among six PBPs, PBP1a, PBP1b, PBP2x, PBP2a, PBP2b and PBP3, alterations in PBP1a, PBP2b and PBP2x and sometimes PBP2a are closely related to penicillin resistance in pneumococci.[39] Penicillin resistance is caused by mutations in pbp genes by intraspecies or interspecies gene transfer, particularly from commensal Streptococcus species such as S. mitis and S. oralis. Gene transfer results in the mosaic gene sequences of pbp in S. pneumoniae.[40] Mosaic pbp gene sequences of penicillin-resistant international clones such as Spain23F-1, Spain6B-2 are thought have originated from resistant commensal Streptococcus species.[41]

Resistance to cefotaxime and ceftriaxone in S. pneumoniae remains relatively infrequent worldwide, although resistance to cefuroxime has been reported to be much higher. The rates of nonsusceptibility to cefotaxime in S. pneumoniae isolates in adults during 2001–2003 in 8 European countries varied from 5.1 to 11.1%, while the rates of nonsusceptibility to cefuroxime varied from 17.7 to 43.9%.[42] According to the SENTRY Antimicrobial Surveillance Program in the United States, the rate of nonsusceptibility to ceftriaxone in pneumococci has increased from 3% in 1998 to 11.7% in 2011.[12] The PROTEKT US study revealed that the incidence of cefuroxime resistance slightly decreased from 28.8% in 2000 to 20.4% in 2004.[43] According to the ANSORP surveillance study, however, nonsusceptibility rates to cefuroxime were relatively high (57.7% in 2008–2009), particularly in Korea (73.7%), Vietnam (71.7%), Taiwan (65.8%) and China (65.1%).[14,16] Decreased susceptibility to cephalosporins is also caused by the development of altered forms of PBPs.[40,44] In particular, it has been reported that alteration of PBP1a and PBP2x is related to resistance of cephalosporins such as cefotaxime and cefuroxime, while alterations in PBP1a, PBP2x and PBP2b is related to penicillin resistance and some cephalosporins such as cefaclor and cefprozil.[45,46]

Fluoroquinolone Resistance

Despite the widespread use of fluoroquinolones in the clinical practice, particularly for the treatment of respiratory infections, the prevalence of fluoroquinolone resistance in pneumococci still remains low.[47–49] According to the SENTRY Antimicrobial Surveillance Program in the United States, the rate of nonsusceptibility to levofloxacin was reported to be 0.2% in 1998 and 1.2% in 2011.[12] The nonsusceptibility rate of levofloxacin was 0.5% with MIC90 of 1 mg/L in pneumococci collected across 24 European countries between 2004 and 2007.[50] In the ANSORP surveillance study, the overall rates of nonsusceptibility to levofloxacin in Asian countries have been persistently low; 1.7% in 2000–2001 and 2.4% in 2008–2009,[14,16] although relatively high prevalence of levofloxacin resistance was found in Taiwan (9.1%) and Korea (5.2%) in 2008–2009.[14]

Pneumococcal resistance to fluoroquinolones is usually mediated by spontaneous point mutations in the quinolone resistance determinant region of gyrA and/or parC.[38] The preferential target sites in S. pneumoniae are topoisomerase IV (subunit parC) for ciprofloxacin and levofloxacin and DNA gyrase (subunit gyrA) for moxifloxacin.[51] Isolates with low-grade resistance to ciprofloxacin typically develop from mutations altering parC and remain susceptible to the newer fluoroquinolones such as levofloxacin, gatifloxacin, moxifloxacin and gemifloxacin, while high-level fluoroquinolone-resistant strains typically have dual mutations affecting both parC and gyrA.[52,53]

Resistance to Other Antimicrobials

Another concern is inducible clindamycin resistance in pneumococci which is mediated by ermB gene that modifies the binding site for both macrolides and lincosamides.[54,55] Methylation of the 23S ribosomal target site encoded by the ermB gene leads to cross-resistance to macrolides, lincosamides and streptogramins B (MSLB), displaying constitutive or inducible MSLB phenotype.[38,55] In particular, inducible MLSB resistance can occur following exposure to macrolides. Clindamycin resistance may be induced during treatment in patients infected with S. pneumoniae which is resistant to erythromycin and susceptible to clindamycin, resulting in treatment failure.[55] Since clindamycin has been frequently used in treating various pneumococcal infections, such as acute otitis media or sinusitis, inducible clindamycin resistance, which is not detected by standard disk or MIC testing could be a concern in clinical practice.

Tetracycline is recommended as the first choice for treatment of lower respiratory tract infections by European guidelines.[56] Doxycycline is also recommended by the Infectious Diseases Society of America (IDSA)/American Thoracic Society (ATS) as a first-line empirical treatment for low-risk persons with community-acquired pneumonia (CAP) caused by S. pneumoniae due to the emergence of resistance in pneumococci to penicillin, cephalosporins and macrolides.[57] However, according to the SENTRY Antimicrobial Surveillance Program, susceptibility to tetracycline (≤2 mg/L) among S. pneumoniae has decreased from 88.8% in 1998 to 78.2% in 2011 in the United States.[12] In a survey performed in 15 European countries during 2004–2005, tetracycline nonsusceptibility (intermediate and resistant) rate was 19.8%. Tetracycline resistance in S. pneumoniae is mediated by ribosomal protection protein encoded by tetM.[58] The tetM gene is frequently carried on transposons Tn916 derivatives (Tn3872, Tn6002, Tn6003 and Tn1545) containing ermB in pneumococci, resulting in high incidence of tetracycline resistance among macrolide-resistant pneumococcal strains.

Multidrug Resistance

S. pneumoniae with MDR, which is defined as resistance to ≥3 antimicrobial classes, has increased worldwide and in particular, the emergence of serotype 19A MDR clones is an increasing concern.[59] MDR in S. pneumoniae usually involves resistance to β-lactams, macrolides, tetracyclines and sulfonamides, while fluoroquinolones are rarely involved in MDR in pneumococci.[60] It is noted that more than 30% of pneumococci are MDR worldwide[59] and data from the PROTEKT study, which collected S. pneumoniae from 38 countries during 2003–2004, indicated that about 40% of pneumococci displayed MDR phenotypes and 89.2% of erythromycin-resistant isolate showed MDR.[61] In a survey of 15 countries in Europe in 2004–2005, 15.8% of pneumococcal isolates were MDR (40.8% in France and 42.9% in Greece).[13] MDR is very prevalent in Asian countries compared with other parts of the world.[1,14,16,62] According to the ANSORP surveillance study, the overall rate of MDR was 26.8% in 2000–2001 and 59.3% in 2008–2009 with the highest rate of 83% in China, followed by Vietnam (75.5%), Korea (63.9%), Hong Kong (62.2%) and Taiwan (59.7%). Prominent increase in MDR in Asian countries was due to the spread of a few clones, including Taiwan19F and Spain23F.[14,16,27]