Dear Editor,
Mycoplasma pneumoniae has become the leading cause of paediatric community-acquired pneumonia in countries where pneumococcal vaccination is included in the national immunisation programme, including Singapore.1 M. pneumoniae is intrinsically resistant to beta-lactams due to the absence of cell walls. Macrolides, tetracyclines, and fluoroquinolones are used to treat M. pneumoniae infections, and macrolides are recommended in children, due to the potential adverse effects of tetracyclines and fluoroquinolones. Macrolide-resistant M. pneumoniae (MRMP) isolates were first observed in 2001.2 The prevalence of MRMP worldwide varies from 0.2% in Europe to 90% in East Asia.3 In Singapore, the prevalence of MRMP in hospitalised children was 13% in 2017.4 Studies in Asia have reported increased disease severity in persons infected with MRMP,5 while others have not detected differences in the clinical course in persons with MRMP versus those with macrolide-susceptible M. pneumoniae (MSMP) infections.6 We compare differences between hospitalised children with MSMP and those with MRMP infections.
From July 2019 to February 2020, patients younger than 16 years in KK Women’s and Children’s Hospital, Singapore, who tested positive for M. pneumoniae by polymerase chain reaction on nasopharyngeal or throat swab samples, were screened for inclusion. Assessment of genotypic macrolide resistance was done as previously described.4 Sequencing was performed retrospectively; thus, the results of genotypic macrolide resistance were not known at the time of treatment. Electronic medical records of patients with M. pneumoniae infection were retrospectively reviewed after informed consent was obtained. Demographic and clinical data were collected and matched to the results of genotypic macrolide-resistance testing. Comparisons between binary groups were analysed using chi-square or Fisher’s Exact test for categorical variables, and Mann-Whitney U test for continuous variables. All P values were two-tailed and differences were considered statistically significant at <0.05.
Between July 2019 and February 2020, 170 patients with M. pneumoniae infection consented to participate in this study. Thirty-one patients (18.2%) had M. pneumoniae strains for which no readable sequence was obtained; they were not included in further analysis. Of the remaining 139 patients, 125 (90%) had M. pneumoniae strains that had wild-type sequence and were presumed to be macrolide-susceptible, and 14 (10%) patients had strains with mutations conferring macrolide resistance. The A2063G point mutation accounted for all 14 M. pneumoniae strains with macrolide resistance.
There were no differences in demographics, presenting symptoms, clinical examination or haematological parameters at presentation (Table 1). A higher median C-reactive protein was observed in children with MRMP infections, although this was not statistically significant. Pneumonia constituted the clinical diagnosis in 86 (68.8%) and 11 (78.6%) hospitalised children with MSMP and MRMP infections, respectively (P=0.09). Children younger than 5 years were significantly more likely to have M. pneumoniae with viral co-infections compared to children older than 5 years (odds ratio [OR] 3.3, confidence interval [CI] 1.5–7.1, P=0.002).
Table 1. Demographics, clinical characteristics, treatment and outcomes of children with macrolide-susceptible and macrolide-resistant Mycoplasma pneumoniae infections
| MSMP
No. (%) (n=125) |
MRMP
No. (%) (n=14) |
P value | |
| Sex | 0.78 | ||
| Male | 62 (49.6) | 8 (57.1) | |
| Female | 63 (50.4) | 6 (42.8) | |
| Ethnicity | 0.81 | ||
| Chinese | 82 (65.5) | 10 (71.4) | |
| Malay | 23 (18.4) | 2 (14.3) | |
| Indian | 11 (8.8) | 2 (14.3) | |
| Caucasian | 2 (1.6) | 0 | |
| Others | 7 | 0 | |
| Age, median (range), year | 4.4 (0.1–14.6) | 4.4 (1.1–11.5) | 0.88 |
| Overseas travel in preceding 4 weeks | 21 (16.8) | 4 (28.6) | 0.28 |
| Sick household contacts | 64 (51.2) | 8 (57.1) | 0.78 |
| Underlying conditions | |||
| Asthma | 8 (6.4) | 3 (21.4) | 0.08 |
| Global developmental delay | 6 (4.8) | 0 | 1.00 |
| Prior treatment with antibiotic | 46 (37.0) | 2 (14.0) | 0.14 |
| Beta-lactam | 41 (89.1) | 2 (100) | |
| Macrolide | 2 (4.3) | 0 | |
| Unknown | 3 (6.5) | 0 | |
| Presenting symptoms | |||
| Fever | 117 (93.6) | 14 (100) | 1.00 |
| Maximum temperature, median (range), oC | 39.0 (38.0–41.8) | 39.8 (38.0–40.6) | 0.11 |
| Cough | 124 (99.2) | 14 (100) | 1.00 |
| Rhinorrhoea | 30 (24.0) | 1 (7.1) | 0.19 |
| Shortness of breath | 54 (43.2) | 7 (50.0) | 0.78 |
| Sore throat | 122 (97.6) | 14 (100) | 1.00 |
| Rash | 6 (4.8) | 0 | 1.00 |
| Respiratory examination | |||
| Tachypnoea | 67 (53.6) | 7 (50.0) | 1.00 |
| Pulse oximetry reading <95% | 28 (22.4) | 3 (21.4) | 1.00 |
| Crepitations | 61 (48.8) | 8 (57.1) | 0.59 |
| Rhonchi | 32 (25.6) | 2 (14.3) | 0.52 |
| Investigations | |||
| Chest radiograph performed | 115 (92) | 13 (93) | |
| Focal consolidation | 77 (67.0) | 10 (77.0) | 0.37 |
| Multifocal consolidation | 17 (14.8) | 3 (23.1) | |
| Peribronchial thickening | 15 (13.0) | 0 | |
| CRP, mg/l, median (range) | 25.4 (1.0–163.8) | 69.2 (4.3–282.0) | 0.15 |
| Viral co-infection | 120 (96) | 14 (100) | |
| MP with 1 viral co-infection | 32 (26.7) | 7 (50.0) | 0.21 |
| MP with 2 viral co-infections | 7 (5.8) | 1 (7.1) | |
| Clinical diagnosis | |||
| Pneumonia | 86 (68.8) | 11 (78.6) | 0.09 |
| Bronchitis | 27 (21.6) | 0 | |
| Upper respiratory tract infection | 12 (9.6) | 3 (21.4) | |
| Total duration of fever, median (range), days | 7.0 (0–16) | 9.0 (4–28) | 0.047 |
| Antibiotic switched from macrolide to fluoroquinolone/tetracycline | 4 (3.2) | 3 (21.4) | 0.02 |
| Pleural effusion | 23 (18.4) | 7 (50) | 0.01 |
| Chest drain inserted | 0 | 2 (28.5) | 0.048 |
| Oxygen supplementation | 44 (35.2) | 6 (42.9) | 0.57 |
| Non-invasive ventilation | 2 (1.6) | 1 (7.1) | 0.28 |
| Length of hospitalisation, median (range), days | 2.0 (1–15.0) | 2.5 (1–22.0) | 0.57 |
CRP: C-reactive protein; MRMP: macrolide-resistant Mycoplasma pneumoniae; MSMP: macrolide-susceptible Mycoplasma pneumoniae
All 139 children with MSMP or MRMP infections received treatment with clarithromycin. The median total duration of fever in children with MRMP infection was significantly longer compared to those with MSMP infection (9.0 vs 7.0 days, P=0.047). Despite the fact that clinicians were unaware of the presence of macrolide resistance, the odds of antibiotics switched from macrolide to fluoroquinolone or tetracycline in children with MRMP infections was 8.2 that of children with MSMP infections (CI 1.6–41.7, P=0.02). Children with MRMP infection who had a switch of antibiotic therapy had a significantly longer median duration of fever during hospitalisation (median 12.0 days, range 6.0–21.0) compared to children with MRMP infection who remained on macrolide therapy (median 1.0 day, range 0–5.0) (P=0.005). Eleven (78.5%) children with MRMP infections remained on macrolide therapy, and all defervesced and were discharged.
Pleural effusions were significantly more frequent in children with MRMP infections compared to those with MSMP infections (OR 4.4, CI 1.4–13.9, P=0.01), and required more frequent chest drain insertions (P=0.048). There were no significant differences between children with MSMP and MRMP infections in terms of the need for oxygen supplementation, non-invasive ventilatory support or length of hospitalisation.
These findings add to a growing literature that reports a prolonged clinical course and more complications in MRMP infections.5,7 Delayed effective antimicrobial treatment for M. pneumoniae has been postulated to be related to an immune response, which may lead to prolonged or extrapulmonary disease.8 It is crucial to note that the majority of children with MRMP infections (78.6%) eventually became afebrile even when receiving macrolide therapy. Clinical response to macrolides in MRMP infections has been described, although the duration of illness tends to be longer.9 The usefulness of macrolide treatment in children with MRMP infection is unclear, but there may be anti-inflammatory effects due to the inhibition of cytokine production, including interleukin-8.7
The prevalence of MRMP isolates sampled between July 2019 and February 2020 was 10% in our study, which did not differ significantly from the prevalence of 13% in an earlier cohort in Singapore studied between 2013 and 2014.4 With relatively low prevalence rates, macrolides remain a suitable first-line antimicrobial for community-acquired pneumonia in children with suspected M. pneumoniae infection.
The lack of distinguishing clinical features among children with infections by MRMP and MSMP has also been described in other comparative studies.10 Until the time when molecular tests for macrolide resistance in M. pneumoniae become available on a real-time basis, clinicians will have to rely on the clinical course of the patient and a setting’s prevalence of MRMP to guide decisions on switching antimicrobial therapy for children with M. pneumoniae infections. Persistent fever despite macrolide therapy and the presence of pleural effusions in children with M. pneumoniae infections are factors that may alert clinicians to the presence of macrolide resistance.
Our study involved hospitalised children in a single centre, and macrolide resistance rates may not reflect those of the wider community. The number of children with MRMP infection was also small, limiting the power of the study.
In conclusion, a longer duration of fever and higher rates of pleural effusions were found in children with MRMP infections compared to those with MSMP infections. The prevalence of MRMP isolates remains low in children in Singapore despite the high rates reported in East Asia, and macrolides remain a suitable first-line antibiotic for children with suspected M. pneumoniae infection.
Funding
This work was supported by the SingHealth Duke-NUS Paediatrics Academic Clinical Programme Tan Cheng Lim research and education fund grant (grant number PAEDSACP-TCL/2019/RES/004).
REFERENCES
- Chiang WC, Teoh OH, Chong CY, et al. Epidemiology, clinical characteristics and antimicrobial resistance patterns of community-acquired pneumonia in 1702 hospitalized children in Singapore. Respirology 2007;12:254-61.
- Okazaki N, Narita M, Yamada S, et al. Characteristics of macrolide-resistant Mycoplasma pneumoniae strains isolated from patients and induced with erythromycin in vitro. Microbiol Immunol 2001;45:617-20.
- Liu Y, Ye X, Zhang H, et al. Characterization of macrolide resistance in Mycoplasma pneumoniae isolated from children in Shanghai, China. Diagn Microbiol Infect Dis 2010;67:355-8.
- Loo LH, Soong HY, Maiwald M, et al. Assessment of Genotypic Macrolide Resistance among Mycoplasma pneumoniae Infections in Children in Singapore. Ann Acad Med Singap 2017;46:290-2.
- Chen YC, Hsu WY, Chang TH. Macrolide-Resistant Mycoplasma pneumoniae Infections in Pediatric Community-Acquired Pneumonia. Emerg Infect Dis 2020;26:1382-91.
- Waites KB, Ratliff A, Crabb DM, et al. Macrolide-Resistant Mycoplasma pneumoniae in the United States as Determined from a National Surveillance Program. J Clin Microbiol 2019;57:e00968-1019.
- Morozumi M, Takahashi T, Ubukata K. Macrolide-resistant Mycoplasma pneumoniae: characteristics of isolates and clinical aspects of community-acquired pneumonia. J Infect Chemother 2010;16:78-86.
- Yang TI, Chang TH, Lu CY, et al. Mycoplasma pneumoniae in pediatric patients: Do macrolide-resistance and/or delayed treatment matter? J Microbiol Immunol Infect 2019;52:329-35.
- Kawai Y, Miyashita N, Yamaguchi T, et al. Clinical efficacy of macrolide antibiotics against genetically determined macrolide-resistant Mycoplasma pneumoniae in paediatric patients. Respirology 2012;17:354-62.
- Kawakami N, Namkoong H, Saito F, et al. Epidemiology of macrolide-resistant Mycoplasma pneumoniae by age distribution in Japan. J Infect Chemother 2021;27:45-8.
