Prevalence of Resistance to β-Lactam Antibiotics and bla Genes Among Commensal Haemophilus parainfluenzae Isolates from Respiratory Microbiota in Poland

(1) Background: Beta-lactams are the most frequently used antimicrobials, and are the first-line drugs in many infectious diseases, e.g., pneumonia, otitis media. Due to this fact, various bacteria have developed resistance to this group of drugs. (2) Methods: Eighty-seven Haemophilus parainfluenzae isolates were obtained from adults 18–70 years old in eastern Poland. The presence of 10 bla genes and 2 substitutions in ftsI reported as the most frequent in H. parainfluenzae were analyzed. (3) Results: Among 57 beta-lactam-resistant isolates, 63.2% encoded bla genes; blaTEM-1 predominated (54.4%), followed by blaOXA (19.3%), blaDHA (12.3%), blaSHV (10.5%), blaGES (7.0%), blaCMY (5.3%), blaVEB (1.8%) and blaROB-1 (1.8%). Lys-526 was the most common substitution in ftsI gene. The resistance genotypes were as follows: gBLNAS (17.5%), low-gBLNAR I (1.8%), low-gBLNAR II (1.8%), gBLNAR II (15.8%), gBLPAS (15.8%), gBLPAR (19.3%), gBLPBS I (8.8%) and gBLPBS II (1.8%); (4) Conclusions: This has been the first study to report on the high diversity of bla genes in H. parainfluenzae isolates in Poland. High sensitivity and specificity of benzylpenicillin test, as well as PCR of bla genes were shown, indicating that these methods may be useful as tools for the rapid screening of beta-lactamase prevalence and resistance to beta-lactams among H. parainfluenzae isolated from respiratory microbiota.


Introduction
Beta-lactam antibiotics and especially penicillins, are the most frequently used group of antimicrobials in the European Union. In Poland, their consumption in the community for systemic use was about 10.0 defined daily doses (DDD) per 1000 inhabitants per day in 2017, while in Germany, Sweden, and Austria it was in the range of 5.0-7.2 DDD per 1000 inhabitants per day [1]. Beta-lactams (e.g., amoxicillin, second-or third-generation cephalosporins) are of great importance in the treatment of many infections, including those caused by haemophilic rods (Haemophilus spp., Pasteurellaceae family). They are the first-line drugs in the outpatient treatment of pneumonia, with suspicion of aspiration pneumonia in adults, and in treatment of acute bacterial rhinosinusitis, acute otitis media, and bacterial meningitis [2]. Amoxicillin is recommended for the treatment of non-beta-lactamase-producing

Culture and Identification
Isolates were stored as a frozen stock in trypticasein soy broth (TSB, Biocorp, Warsaw, Poland) supplemented with Haemophilus test medium supplement (HTMS, Oxoid, Hampshire, Great Britain) with addition of 30% (v/v) glycerol at −70 ± 2 • C until its use. Bacteria were then re-cultured by applying the frozen stock to a chocolate agar (BioMérieux, Craponne, France) and incubated for 24 h at 35 • C in microaerophilic (5-10% CO 2 , 80-90% N 2 , 5-10% O 2 , Generbag microaer, BioMérieux, Craponne, France) conditions. Haemophilus spp. isolates were then identified by colony morphology, Gram-staining, and identified to the species level by API NH microtests (BioMérieux, Craponne, France) and by the Ultraflextreme Matrix Assisted Laser Desorption Ionization Time of Flight Mass Spectrometry (Bruker Daltonics, Bremen, Germany) (MALDI-TOF MS) with MALDI-Biotyper 3.0 software (Bruker Daltonics, Bremen, Germany) according to the procedure described earlier [30]. The correctness and reliability of the abovementioned identifications were expressed in the form of a point indicator, as presented previously [31]. Only H. parainfluenzae isolates with identification scores >1.999 on the basis of protein profile were taken for further analysis.

Antimicrobial Susceptibility Testing
The isolates grown overnight on chocolate agar (HAEM, BioMérieux, Craponne, France) were resuspended in TSB + HTMS medium. Cell concentration in broth medium was determined using a densitometer (BioMérieux, Craponne, France). Bacterial suspensions with a density of 0.5 (1.5 × 10 8 colony forming units CFU/mL) were used according to the McFarland standard and then incubated overnight at 35 • C in microaerophilic atmosphere. Each isolate was tested for beta-lactamase activity by the Pen reaction on API NH strip, and tested with a benzylpenicillin disk (1 U, Becton Dickinson, Franklin Lakes, NJ, USA) and the cefinase disk assay (Becton Dickinson, Franklin Lakes, NJ, USA). According to the benzylpenicillin disk screening test [32], isolates with zone diameter ≥ 12 mm were reported as susceptible to any beta-lactam agents for which clinical breakpoints were available (all beta-lactam resistance mechanisms excluded). Isolates with zone diameter < 12 mm were suspected to produce beta-lactamases and/or to have PBP3 mutations, and needed further analysis. As for ampicillin and amoxicillin without beta-lactamase inhibitors, if beta-lactamase was positive, the isolate was reported as resistant, and if beta-lactamase was negative, then as susceptible according to the clinical breakpoints. As for other beta-lactams, except cefepime, cefpodoxime, and imipenem, the isolate was reported according to the clinical breakpoints for the agent in question. Additionally, for the mentioned agents, if resistant according to both screen and agent disk diffusion test, the isolate was reported as resistant. If isolate was resistant according to the screen test and susceptible according to the agent disk diffusion test, the minimal inhibitory concentration (MIC) of the agent was determined and interpreted according to latest breakpoints. As regards the cefinase test, all H. parainfluenzae isolates were classified according to the color change of the nitrocefin disc after 2-3 min as either beta-lactamase-positive isolates (cef + , change of the yellow color to pink) and beta-lactamase-negative (cef − , no change of color). In a subsequent step, the susceptibility to the following beta-lactam antibiotics: ampicillin (2 µg), amoxicillin-clavulanate (2/1 µg), ampicillin-sulbactam (10/10 µg), cefuroxime (30 µg), cefotaxime (5 µg), imipenem (10 µg), and meropenem (10 µg) was determined using the Kirby-Bauer disk-diffusion method [33], using Mueller-Hinton agar medium with 5% horse blood with the addition of 20 mg/L NAD (MHF, BioMérieux, Craponne, France). Moreover, MIC alues of ampicillin (MIC Am ) for ampicillin-resistant isolates were determined by the E-test method using E-test strips (BioMérieux, Craponne, France) at a concentration gradient of 0.016-256 mg/L. Diameters of bacterial growth inhibition zones were measured using the Interscience Scan ® 1200 version 8.0.3.0 (Interscience, St Nom la Bretèche, France); the results were interpreted according to the latest European Committee on Antimicrobial Susceptibility Testing (EUCAST) 2019 criteria [32].

DNA Extraction
Genomic DNA was extracted using a modified boiling method [34]. The culture was transferred into sterile 1.5 mL Eppendorf tubes, followed by centrifugation for 10 min at 14,000 rpm. The obtained supernatant was decanted from the precipitate. The pellet was resuspended in 200 µL of sterile deionized water and centrifuged again (14,000 rpm for 5 min). Next, the obtained supernatant was recovered from the pellet and the pellets were once more suspended in 100 µL of sterile deionized water. The contents of the tubes were thoroughly mixed by vortexing for 10 s at 13,000 rpm and then incubated for 10 min on a thermoblock (Biometra, Göttingen, Germany) at 100 • C. After cooling, samples were subjected into centrifugation for 10 min at 14,000 rpm. The supernatant, containing isolated DNA, was transferred to new sterile Eppendorf tubes, which were reheated for 10 min at 100 • C. Extracted DNA was frozen to −70 • C for further analysis.

Amplification Experiments and Gene Detection
Determination of the presence of ten beta-lactamase bla genes: bla TEM-1 , bla GES , bla OXA , bla VEB , bla CTX-M-1 , bla SHV , bla CMY , bla DHA , bla PER , and bla ROB-1 (also called further: TEM-1, GES, OXA, VEB, CTX-M-1, SHV, CMY, DHA, PER, and ROB-1, respectively) was carried out for all H. parainfluenzae isolates. PCR amplification from genomic DNA was performed for parts of the abovementioned genes. Table 1 presents previously published sequences of the oligonucleotides (Novazym, Poznan, Poland). The PCR cycling conditions were 30 cycles of the following: 95 • C for 30 s, 46 • C for 60 s, and 72 • C for 30 s. All reactions were carried out using the AmpliMIX HiFi (Novazym, Poznan, Poland) in a total volume of 25 µL containing: 0.1 µL of 0.5 U HiFi Taq polymerase DNA, 2.5 µL of reaction buffer (10×) pH 8.6, 0.75 µL of 2 mM dNTPs mix, 1 µL of each 0.6 µM primer, followed by electrophoresis in 1.5% agarose gel (Sigma-Aldrich, Saint Louis, MO, USA). Each reaction included 2 µL of DNA templates from individual H. parainfluenzae isolates. Determination of presence of the ftsI (pbp3) gene encoding the transpeptidase domain of PBP3 protein was performed by PCR reaction according to Touati et al. [18]. Additionally, the amplifications with primers complementary to the part of ftsI gene within which amino acid substitutions (Table 1) resulting in resistance to beta-lactam antibiotics have been commonly reported used pbp3-BLN (Thr-385 and Lys-526 substitutions) and pbp3-INT (Lys-526 substitution) [19].
Based on the results obtained by the beta-lactams susceptibility testing, drug resistance phenotypes and genotypes of drug resistance to beta-lactam antibiotics were classified as follows ( Table 2).

Statistical Analysis
The data processing and analysis were performed using GraphPad InStat 3.00 (GraphPad Software, San Diego, CA, USA). Fisher's exact test was used to calculate the 95% confidence interval ranges (95% CI) and the relative risk (RR). A p value of <0.05 was considered statistically significant. The following indexes were calculated for the conventional phenotypic methods used to detect beta-lactamases in relation to PCR samples: sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV). Sensitivity (defined as the percentage of true positives) was calculated as the proportion of H. parainfluenzae isolates that tested positive among all isolates tested that actually produced beta-lactamase, whereas specificity (defined as the percentage of true negatives) was calculated as the proportion of isolates that tested negative among all isolates that actually did not produce beta-lactamase. PPV and NPV allowed a clinical perspective to be obtained of how likely production of beta-lactamase was in comparison to PCR results among H. parainfluenzae isolates tested. Positive predictive value was the probability that, following a positive test result, an individual bacterial isolate would truly produce that specific beta-lactamase, while negative predictive value was the probability that following a negative test result, that individual bacterial isolate would truly not produce that specific enzyme.
Within all beta-lactam-resistant isolates, 36.8% (21/57) were beta-lactamase-negative, classified into gBLNAS, low-gBLNAR, and gBLNAR genotypes. The gBLNAS isolates were resistant to one (Ctx/Ipm) or two (Cxm Ctx/Ctx Mem/Cxm Mem) antibiotics simultaneously, and the resistance to cefotaxime found in 33.3% (7/21) of isolates was predominant. Two low-gBLNAR isolates were resistant only to ampicillin. One of them-the W1HC isolate-had a MIC Am value of 0.8 mg/L (susceptible according to EUCAST) and the Lys-526 substitution in ftsI. The second one-the IM18GB isolate-had a MIC Am value of 2.0 mg/L susceptible, increased exposure (formerly intermediate according to EUCAST), did not harbor the ftsI gene or any amino acid substitutions, and was classified into the low-gBLNAR I genotype.   All 14.0% (8/57) tested gBLNAR isolates were 100% resistant to ampicillin and cefuroxime, 88.9% resistant to cefotaxime, seven of them had an Am Cxm Ctx resistance pattern, and one had an Am Sam Cxm Ctx pattern. The MIC Am values ranged from 2.0 to 8.0 mg/L (two had MIC > 4.0 mg/L), and one 39CU isolate was cefinase-positive (BLPAR phenotype). All gBLNAR isolates were classified into group II with an amino acid substitution at the Lys-526 position in the ftsI gene.
Of the 36 phenotypically beta-lactamase-positive isolates, 41.7% (15/36) were susceptible to ampicillin and were classified into the BLPAS phenotype. Among them, resistance to cefuroxime was predominant, observed in 73.3% (11/15) of isolates, followed by cefotaxime, found in 53.3% (8/15) of cases. Four of them were resistant to two beta-lactam antibiotics (Cxm Ctx or Cxm Ipm), one was resistant to three antimicrobials (Cxm Ctx Mem) and one was resistant to four antibiotics, presenting a Cxm Ctx Ipm Mem resistance pattern. In 40% (6/15) of BLPAS isolates, amino acid substitutions at the Lys-526 position of the ftsI gene were found, in one isolate, the Thr-385 substitution was observed. Due to that, we classified these isolates into the gBLPBS genotype, divided further into two groups: I with the Lys-526 substitution and II with the Thr-385 substitution. Eleven (30.5%) of the 36 beta-lactamase-positive isolates were BLPAR, among which two were cefinase-positive. All BLPAR isolates were ampicillin-resistant, one had MIC Am = 0.75 mg/L (sensitive according to EUCAST), six had MIC Am values from 1.0 to 3.0 mg/L (susceptible, increased exposure according to EUCAST), and four had MIC Am = 6.0 mg/L (resistant according to EUCAST). According to PCR results, the assigned phenotype of one isolate was not covered by the genotype due to the presence of an amino acid substitution at the Lys-526 position in the ftsI gene, thus, the W5HD isolate was classified into the gBLPACR I genotype. Similarly, two cefinase-positive BLPBR isolates, one resistant to four (Am AmC Cxm Ctx), and one resistant to six (Am AmC Cxm Ctx Ipm Mem) beta-lactam antibiotics, with Lys-526 substitutions in the ftsI gene, were classified into the gBLPACR I genotype.

Resistance to Beta-Lactams among H. parainfluenzae
Increasing resistance to antimicrobial agents is a global phenomenon nowadays, commonly occurring among haemophilic bacteria, including the rare etiological factor of such infections as caused by H. parainfluenzae [13][14][15]21]. This may have a huge impact on therapeutic treatment for infections caused by these bacteria. The importance of this issue has been evidenced by an increasing number of reports on beta-lactamase-producing strains and the weakening activity of beta-lactam antibiotics against H. parainfluenzae [7,13,14,16,21]. Since 2013 only a few publications have been published in the PubMed database about H. parainfluenzae resistance to beta-lactam antibiotics and its molecular mechanisms [7,13,16,21], despite the fact that it belongs to the fastidious Gram-negative bacterial group of the Haemophilus-Aggregatibacter-Cardiobacterium-Eikenella-Kingella genus (HACEK). That means H. parainfluenzae is an etiological factor of many documented chronic or recurrent infections also occurring within the respiratory system [2,18]. This was the first study to report the high diversity of beta-lactamase genes in H. parainfluenzae isolates from respiratory tract microbiota in eastern Poland.

The Susceptibility to Beta-Lactam Antibiotics among H. parainfluenzae Isolates
The susceptibility of H. influenzae, and similarly H. parainfluenzae, to beta-lactam antibiotics is mainly determined by ampicillin susceptibility results. The resistance phenotypes for these bacteria are divided into many groups and subgroups [9][10][11][12]19,24,42]. In H. influenzae clinical strains, resistance to ampicillin resulting from the bacterial ability to synthesize beta-lactamases is usually detected on the basis of MIC values for ampicillin and amoxicillin with clavulanic acid above the recommended values as the cut-off point for resistant strains. According to the actual EUCAST recommendations for amoxicillin-clavulanate, the area of technical uncertainty (ATU) is relevant only if the benzylpenicillin 1 U disk screen is positive [32].
In our study, among all tested isolates, 34.5% were sensitive and 65.5% were resistant to tested beta-lactams, with resistance to ampicillin, cefuroxime, and cefotaxime predominating and found in 36.8%, 37.9%, and 35.6% of isolates, respectively. Other researchers have also reported a high extent of resistance among H. parainfluenzae isolates [7,14,15,21,23]. Tinguely et al. [7] described an extensively drug-resistant (XDR) H. parainfluenzae isolate resistant to almost all tested beta-lactams, comprising: ampicillin, amoxicillin, amoxicillin-clavulanate, cefuroxime, ceftriaxone, cefotaxime, and cefepime. Abotsi et al. [14], in studies on H. parainfluenzae isolated from sputum of patients with pneumonia, also demonstrated its resistance to fluoroquinolones and telithromycin. In turn, the phenotypic resistance to beta-lactam antibiotics, azithromycin, and trimethoprim-sulfamethoxazole was demonstrated in H. parainfluenzae by Kosikowska et al. [15]. The research of these authors was devoted to microorganisms isolated from the airways of patients suffering from lung cancer resistant simultaneously to five, six, or eight drugs from different therapeutic groups, including beta-lactams.

Conventional Methods for Detection of Beta-Lactamase in Haemophili
Out of the three methods available for the detection of beta-lactamase in haemophili, the cephalosporin (nitrocefin) method is the most reliable and recommended in the case of haemophilic bacteria [32,43]. According to a manufacturer of the API NH strips, a positive penicillinase test indicates the presence of a penicillinase, which prohibits the use of penicillins (penicillin G, amino-, carboxy-, and ureidopenicillins). Additionally, a susceptibility test is required for the other beta-lactams. However, according to EUCAST, the benzylpenicillin 1 U disk screen test can be used to exclude beta-lactam resistance mechanisms [32,42]. When screening is negative, all beta-lactam agents for which clinical breakpoints are available can be reported sensitive without further testing. This result excludes both beta-lactamase production and other beta-lactam resistance mechanisms. All positive screen results indicate a possibility of both resistance mechanisms-beta-lactamase production and/or PBP3 mutations, as was explained in the Materials and Methods section based on the EUCAST flow chart [32]. At the same time, additional results should be taken into account, e.g., ampicillin and amoxicillin susceptibility [32,42]. However, phenotypic methods are sometimes insufficient to visually detect these enzymes, or the data obtained are ambiguous and difficult to interpret [20,43].
On the basis of the three phenotypic methods used in our study, 13.8% of isolates were able to phenotypically synthesize beta-lactamases, and in 4.6% of isolates, at least one of the tests used (PEN in API NH strip) gave a less precise and difficult to interpret (weak color change) result, which was statistically significant (p < 0.0001). Moreover, statistical diagnostic values of these methods among H. parainfluenzae isolates in comparison to PCR amplification results (bla gene identification) showed that the test of susceptibility to amoxicillin-clavulanate was the most sensitive, while the benzylpenicillin screen test was the most specific for H. parainfluenzae clinical isolates.

PCR Amplification as Rapid Detection Method of Beta-Lactamase Production
Many researchers have questioned the utility of conventional phenotypic methods in microbial diagnostics, requiring at least 2-3 days to fully detect and determine drug susceptibility [20,36,43,44]. They indicate treatment failure, poor outcome, and development of beta-lactam resistance among Haemophilus spp. strains as consequences of negative identification. It has been suggested that more specific and sensitive methods should be used, including PCR amplification or real-time PCR for quantification of Haemophilus spp. strains and simpler, more rapid and reliable detection of beta-lactamase production [7,13,14,[17][18][19]21]. For example, [19] revealed 92.9% and 91.8% sensitivity and specificity, respectively, of the PCR amplification method used to detect ampicillin-resistant intermediate H. influenzae strains, and 100% sensitivity and specificity of primers used to identify TEM-1 beta-lactamase compared to the conventional phenotypic methods.

Prevalence of Beta-Lactamase Type TEM-1 and ROB-1
In this study, among beta-lactamase-positive beta-lactam-resistant H. parainfluenzae, type TEM beta-lactamase was the most dominant, found in 86.1% isolates; the ROB-1 gene was identified in 2.8%. Our findings are in general agreement with the first four years of the global PROTEKT study (duration 1999-2003) comprising 137 centers of 38 countries (including Poland) [22]. The overall prevalences of TEM-1 and ROB-1 positive H. influenzae isolates in this study were 93.7% and 4.6%, respectively. In Poland, authors found 95.0% TEM-1 positive isolates and none with ROB-1 beta-lactamase [22]. Similar results were obtained by Nakamura et al. [19] and Touati et al. [18], who did not isolate any H. influenzae clinical strain producing ROB-1 type beta-lactamase either from sputum or from nasopharynx, respectively. According to Garcia-Cobos et al. [21], of 40 clinical (e.g., from genital mucosae, urine, respiratory secretions, peritoneal fluid, blood) H. parainfluenzae isolates, 84.6% of bla + isolates were identified as possessing TEM-1 beta-lactamases, and 7.7% possessed TEM-34 and TEM-182. According to the literature, strains positive for both TEM-1 and ROB-1 are rare [5,22], and conferring resistance to second generation cephalosporins [22] was also confirmed by this study. We also found that 30.6%, 19.4%, 16.7%, 11.1%, 8.3%, and 2.8% of H. parainfluenzae isolates from respiratory microbiota harbored the other following beta-lactamase genes: bla OXA , bla DHA , bla SHV , bla GES , bla CMY , and bla VEB , respectively. None of the tested isolates expressed bla PER or bla CTX-M-1 genes.
Many difficulties with the cefinase test results have been reported, warning against its weak detection level of some beta-lactamases. There have been some conflicting reports about the sensitivity of nitrocefin hydrolysis, especially in detecting the ROB-1 beta-lactamase [5,22]. There have also been some cases where nitrocefin hydrolysis was positive and both TEM-1 or ROB-1 genes were negative in the PCR reaction, accounting for 1% of tested isolates, which was supposedly due to unrecognized beta-lactamase in that study [22]. According to Tinguely et al. [7], who reported in Switzerland a case of XDR H. parainfluenzae isolate which carried the bla TEM-1 gene, the phenotypically produced beta-lactamase was not expressed according to the nitrocefin test (for both the cefinase paper disc and the hydrolytic activity against nitrocefin).

Discrepancies in Phenotype and Genotype Classification for H. parainfluenzae
Furthermore, many authors have pointed to the imperfection of the EUCAST recommendations and the diagnostic scheme of haemophilic bacteria in the context of detection and phenotypic activity of beta-lactamases [44,45]. It is often underlined that phenotypic methods (e.g., in vitro sensitivity to ampicillin, amoxicillin, or amoxicillin with clavulanic acid) could lead to faulty results. According to Garcia-Cobos et al. [21], phenotype and genotype classification for H. parainfluenzae are the same as for H. influenzae. For this reason, it has been suggested that the scheme given by EUCAST [32,42] does not lead to the extraction of H. influenzae and/or H. parainfluenzae strains belonging to particular phenotypes of resistance to beta-lactams, including BLNAS, BLNAR, BLPAR, or BLPACR [44].

Polymorphism of the ftsI Resistance Gene
At the same time, the need to analyze the polymorphisms of resistance genes (e.g., the ftsI gene) is strongly indicated. It also follows that not all changes in the amino acid structure of PBP3 proteins are related to the determined phenotype of in vitro resistance [44]. This fact has been particularly emphasized in the case of strains with the genotype gBLNAR, which phenotypically show sensitivity to ampicillin, amoxicillin, and amoxicillin with clavulanic acid [44]. Similar observations have been made concerning strains belonging to the BLNAS phenotype, which are often positive in the cefinase assay, but not found to contain beta-lactamase genes [45].
In this study, 28.1% of all commensal H. parainfluenzae isolates were beta-lactam-resistant ftsI + bla + , indicating the existence of two simultaneous mechanisms of beta-lactam resistance, defined as gBLPACR I/II or gBLPBS I/II. This type of isolates is still rare, and their detection is difficult [5,18]. Touati et al. [18] observed only 6% of these isolates among tested nasopharyngeal H. influenzae strains. The same authors indicated that for these strains, resistance to amoxicillin-clavulanate is doubtful.

Beta-Lactamase-Negative ftsI-Positive Isolates
Furthermore, we found that 52.4% of gBLNAR or low-gBLNAR bla − isolates had an ftsI gene with a substitution at the position Lys-526. This was in agreement with Touati et al. [18], who found that H. influenzae nasopharyngeal isolates had the bla TEM-1 gene but did not phenotypically produce any beta-lactamases (one isolate), or were BLNAR ftsI − and bla TEM -negative. Some researchers have suggested the possibility of horizontal transfer of genes [14], including the ftsI gene, among H. influenzae and H. parainfluenzae clinical isolates. This has been supported, for example, by a transfer of Asn-526-Lys amino acid substitution specific to H. influenzae and now also reported in H. parainfluenzae isolates [21]. This confirmed that H. parainfluenzae might be a huge reservoir of multiple beta-lactamase-carrying plasmids for other bacterial species [14]. This is an important issue because exact determination of the mechanism of resistance (production of beta-lactamases and/or mutations in the ftsI gene) to beta-lactam antibiotics in haemophilic bacteria will enable the implementation of effective therapeutic options to limit the spread of this phenomenon.
The authors are aware of the limitations associated with this study, especially related to the low number of samples. The authors also note that official breakpoints recommended by EUCAST were not available for H. parainfluenzae, only for H. influenzae in general, which may not be sufficient nowadays. Furthermore, mapping of bla TEM gene, P3 promoter variants, and detection of other amino acid substitutions in the ftsI gene need to be investigated to analyze the degree of beta-lactamase expression among H. parainfluenzae isolates.
In conclusion, this was the first study to highlight the high diversity of beta-lactamase genes among commensal H. parainfluenzae isolates from respiratory microbiota in eastern Poland. This study showed the high sensitivity and specificity of the benzylpenicillin screen test, as well as PCR detection of bla genes for H. parainfluenzae. These methods may be a useful tool for the rapid screening of beta-lactamase prevalence among H. parainfluenzae clinical isolates resistant to beta-lactam antibiotics.