2. Pertussis Diagnostics
Pertussis diagnostics can be divided into three main approaches: (1) culture, (2) nucleic acid detection (PCR) and (3) serology. Patient age, vaccination history and onset of the symptoms should be considered when choosing the correct diagnostic method
[8]. Culture can be performed up to 2 weeks after the symptoms have appeared, before the bacteria is cleared by the immune defence. Specimen from freshly obtained nasopharyngeal (NP) samples should be cultured on Regan-Lowe (RL, charcoal) or Bordet-Gengou (BG, blood) agar containing cephalexin. Suspected
B. pertussis specific colonies are further cultured on RL/BG agar (without cephalexin), and identified with e.g., slide agglutination test with specific anti-
B. pertussis and anti-
B. parapertussis sera or MALDI-TOF
[8,9,10][8][9][10]. Specific nucleic acid identification (targeting IS481/
ptxp) with PCR requires only a small amount of DNA for detection and identification of the bacterium and is therefore far more sensitive than culture. Furthermore, it can be used even three to four weeks after the onset of symptoms. Therefore, PCR-based approaches are more widely used than culture, especially with infants and small children. For school children and adults, serology is commonly used as there is less interference in antibodies induced from previous vaccinations and the only symptom may have been a prolonged cough (>3–4 weeks, culture nor PCR can be used). Serological diagnosis should be made based on the measurement of serum IgG antibodies against pertussis toxin
[11]. Furthermore, laboratory confirmation of
B. pertussis from clinical samples is needed before antimicrobial susceptibility testing (AST) is performed.
3. Epidemiology
The first macrolide resistant
B. pertussis strain was identified in a 2-month-old infant from Yuma, Arizona, US in 1994
[7]. The isolate was highly resistant to erythromycin with a minimum inhibitory concentration (MIC) > 64 µg/mL. However, the origin of this isolate was not known. Breakpoints to detect antimicrobial resistance of clinical
B. pertussis isolates were not standardized but the reported resistant strains had MICs of >256 µg/mL with erythromycin (ERY) and clarithromycin (CHL) by Etest method suggesting macrolide resistance. Concurrently, seven additional
B. pertussis isolates from the same area were tested, but macrolide resistance was not detected in these cases. In a review of 47
B. pertussis isolates from children in Utah, US, in 1985–1997, one isolate from January 1997 was resistant against erythromycin
[12].
Since the first appearance of macrolide-resistant
B. pertussis, macrolide susceptibility has been tested in thousands of cultured isolates all over the world (
Table 1,
Figure 1). In a study of 1030 isolates collected from various parts of the the US, five (0.5%) isolates were erythromycin resistant. Four out of five isolates were from Arizona (1994–1995) and one from Georgia (1995). All isolates initially showed the growth inhibition of
B. pertussis by disc diffusion method, but after 5–7 days of incubation, novel bacterial colonies appeared on the plate inside the growth inhibition area, demonstrating heterogeneous phenotype
[13]. In a review of 38
B. pertussis isolates from France in 2003, none of them were resistant to erythromycin
[14]. However, nine years later in 2012, the first patient in Europe with macrolide-resistant
B. pertussis was diagnosed in Lyon, France
[15]. A three-week-old neonate with severe pertussis was treated repeatedly with macrolides before the detection of the resistant isolate. Of the three serial isolates from the patient, the first two were sensitive, but the third one turned to be resistant, suggesting that the
B. pertussis isolate acquired the mutation leading to macrolide resistance during the macrolide treatment. Sporadic cases of macrolide-resistant
B. pertussis isolates were also reported from Iran in 2009
[16].
Figure 1.
Countries where
B. pertussis
antimicrobial susceptibility studies have been performed (created with MapChart).
In Asia, studies from Cambodia, Japan, Taiwan and Vietnam have found some macrolide-resistant
B. pertussis isolates that seem to be related to resistant strains in mainland China
[17,18,19][17][18][19]. In northern Vietnam, of NP swab samples from 184 patients with pertussis diagnosed during 2016–2020, 24 (13.0%) were found to be resistant. In Japan, the first isolation of a macrolide-resistant strain was from a 2-month-old baby in 2018. The MICs of the isolate showed >256 µg/mL for ERY and CHL and >32 µg/mL for AZT. The complete genome sequence of the macrolide resistant
B. pertussis strain from Japan has been published
[20]. It confirms that the isolate has a homogeneous A2047G mutation in each of the three copies of its 23S rRNA gene and that it belongs to the genotype that is common in Chinese macrolide resistant
B. pertussis isolates. The issue of macrolide-resistant
B. pertussis is greater and reported in more detail in China than in anywhere else in the world. The first macrolide-resistant isolates from Shandong Province in China were reported in 2011 in two asymptomatic pupils
[21]. No macrolide resistance has been detected in historical isolates in China from 2008 or earlier
[22,23][22][23]. More recent reports show very high prevalence of macrolide resistance among
B. pertussis isolates in different parts of China (
Table 1).
Table 1.
Global frequencies of macrolide-resistant
Bordetella pertussis
.