Serotypes, virulence and antimicrobial susceptibility of Group B Streptococci isolated from infections in non-pregnant adults in Malaysia

Abstract

This is a cross-sectional study describing on Group B Streptococcus (GBS) serotypes, virulence gene distribution and antimicrobial susceptibility in non-pregnant adults from two major hospitals on East-coast Malaysia. The patients’ clinical information was reviewed, while serotyping and virulence gene detection for the GBS isolates (n = 75) were conducted by PCR. GBS serotypes III and Ia were the most common, followed by serotypes II and IV. The majority of GBS (n = 65; 86.7%) presented with virulence genes such as cylE, scpB, lmb, hylE and bca. Skin and soft tissue infection (SSTI) (n = 50; 66.7%) were the most common clinical diagnosis, with 90% of those diagnosed had underlying diabetes mellitus (DM). GBS serotype Ia and the virulence genes cylE and scpB were common in SSTI. Good sensitivity towards penicillin, erythromycin and clindamycin were observed.

KEYWORDS:

• GBS
• non-pregnant adult
• serotype
• virulence gene
• skin and soft tissue infection

1. Introduction

Group B Streptococcus (GBS) has emerged as significant causative agent of disease in non-pregnant adults, especially those of aged more than 65 years old and underlying medical conditions such as diabetes mellitus (DM) [1–3]. DM is the potential risk factor for GBS infections based on the studies from the United States (US), Europe and Southeast Asian countries, including Malaysia [4–7]. The incidence rate of GBS disease unrelated to pregnancy in Malaysia is still unknown. In the USA, the incidence rate rose from 8.1 cases per 100 000 population in 2008 to 10.9 in 2016, with the fatality rate of 6.5% [4]. Each year, there are 25 invasive GBS disease cases per 100,000 elderly population in the USA [8].

In non-pregnant adults, primary bacteremia without a clear source (39.3%) and SSTI (25.5%) were the most common clinical diagnoses seen in GBS disease [8–10]. Apart from that, pneumonia (12.5%), bone (9.4%) and joint infections (7.8%) were reported with GBS infections unrelated to pregnancy. Meanwhile, UTIs accounted for about 5–23% of invasive GBS disease in non-pregnant adults [11]. The least common clinical diagnoses were endocarditis (3.0%) and meningitis (1.6%) [12]. Disease, including septic arthritis and meningitis, was associated with GBS belonging to serotype III subtype 4 (serotype III-4) and multi-locus sequence type (MLST) 283 (ST283) [13]. The ST283 infection was different as it affected non-pregnant, younger adults with fewer co-morbidities suggesting greater virulence [14,15].

In several studies, SSTI was reported as the most common presentation of GBS infections in non-pregnant adults [5,16,17]. It is usually manifested as infected ulcers (11.3%). Cellulitis or erysipelas and wound infection accounted for 1.7% each, while abscess and necrotizing fasciitis was about 1.1% [7]. DM has been reported as the main risk factor for SSTI [17]. Loss of skin integrity in diabetic patients as a consequence of peripheral neuropathy promotes GBS to penetrate the skin and mucous membrane to cause infection. Early wound debridement or drainage with adequate coverage of antimicrobial therapy results in good outcome. In the USA, GBS invasive disease was typically seen in SSTI, rising from 27.2% in 2008 to 34.0% in 2016 [4].

Serotyping is one of the useful epidemiological tools for describing GBS. The primary virulence factor found in most of the GBS strains is capsular polysaccharide (CPS). This CPS defines each GBS serotype. Currently, there are 10 distinct GBS serotypes based on the CPS antigen: Ia, Ib, II, III, IV, V, VI, VII, VIII and IX [11,12]. One study done in Maryland USA reported that 29% of GBS disease in non-pregnant adults associated with serotype V [18]. In Malaysian population, a study reported that serotype Ia was the main CPS serotype identified (n = 27, 45%), followed by serotype III (n = 10, 16.7%), V (n = 9, 15%), VI (n = 8, 13.3%), VIII (n = 2, 3.3%) and VII (n = 1, 1.7%) [19]. In other parts of the world, including the US and some European countries, the common serotypes were serotypes Ia, II, III, and V and the less frequently identified CPS serotypes were VI to IX [20].

Epidemiological data and clonal distribution of GBS are important to determine colonization at the molecular level. Certain GBS serotypes possess a higher potential to cause invasive disease, while others can be only colonizing strains. Colonization and persistence in different host niches are dependent on the ability of GBS to adhere to host cells and tissues. Although all GBS serotypes are capable of causing invasive infections, most of the current studies and a recent worldwide review of GBS reported that the five predominant GBS serotypes, i.e. Ia, Ib, II, III and V, account for the majority of disease both in neonates and adults [21,22]. However, it is noted that the predominance of GBS serotypes changes over time and varies with geographical region and population [20,23]. Several studies have reported the epidemiological changes of GBS over the years. For example, a study in Korea reported the serotype changes from Ib, Ia and III in 1995 [12] to III, V and VI in 2008–2009 [8] and to V, VIII and III in 2017–2019 [24]. In another study conducted from 1975 to 2014 in Iceland, serotype III was found significantly decreased in the population, while serotypes IV and V showed increasing trends over the time [25]. In addition, around 8–14% of the clinical isolates are classified as non-typeable (NT) strains because they cannot be distinguished on the basis of CPS antigenicity [12,26]. The antigenicity and immunogenicity of capsular antigens differ between serotypes and are associated with invasiveness of GBS diseases [19]. A close monitoring of the epidemiological changes and diversity of GBS serotypes is indeed important for the disease management and vaccination strategies in future.

The virulency of GBS strains was determined by the presence of specific virulence determinants such as surface associated proteins, pore-forming toxin and/or hydrolytic enzymes. Among the virulence factors that has been reported are C5a peptidase (scpB), laminin-binding protein (lmb), α and β-subunits of C protein (bca and bac) and Rib protein (rib). β-hemolysin/cytolysin (cylE), hyaluronidase (hylE) and the CAMP factor (cfb) [27]. These virulence factors facilitate the entry of GBS into host cells and promote its intra-cellular survival. Almost all human isolates contained the scpB and lmb genes, whereas these genes were absent in a majority of strains of bovine origin [11,17,18,28]. There is a hypothesis that scpB-lmb region may be essential for the colonization or infection in humans [11,29]. An increasing diversity of GBS serotypes, as well as serotype switching, represent powerful immune evasion strategies that may severely impair vaccine efforts. Moreover, the current conjugate vaccine incorporates only a limited number of GBS serotypes. Close monitoring of changing serotype distribution in many different countries is therefore crucial in guiding GBS vaccine development [30].

Future vaccination program will be the most appealing strategy for GBS disease prevention. Research and development of vaccines against GBS infection, especially in neonate and pregnant mother, are currently progressing in global. The development of vaccine candidates against GBS has begun since mid-70s, in which the polysaccharide-based antigen using purified GBS type-specific polysaccharides were tested as immunogen and proven to be nontoxic, safe and immunogenic [31]. Since GBS has been addressed as an important cause of invasive infections in the elderly and in those patients with underlying disorders, these subjects could represent two natural target populations for a GBS vaccination program. However, only one clinical vaccination study targeting the elderly has been published to date [32].

Serotyping for GBS is not routinely performed as it has no effect on the antibiotic management. In addition, almost all GBS today are sensitive to penicillin and have no implication on the infection control measures in the hospital settings. GBS serotyping, on the other hand, is important for epidemiological surveillance and can serve as a baseline for vaccination purposes. Although more informative methods are available for GBS genotyping (such as MLST and genome sequencing), PCR methods are preferred in most healthcare settings due to their cost-effectiveness and rapidity. Most of healthcare settings in Malaysia do not routinely perform GBS serotyping for surveillance purposes, most likely due to cost and a lack of awareness about the importance of serotype with severity of GBS infection. As such, the aim of this study was to identify GBS serotypes; the presence of virulence gene(s); antibiotic susceptibility patterns; and their relationship with clinical diagnosis and infection outcome in infected non-pregnant adults, with a little focus on the significance of SSTI.

2. Methods

2.1 Bacterial isolates and confirmation of GBS

A total of 75 GBS isolates were collected from various clinical samples of non-pregnant adults (age >18 years) in two main hospitals in Kelantan, which is in the East-Coast of Malaysia. The two selected hospitals for this study were Hospital Raja Perempuan Zainab II (HRPZ II), the largest government hospital in Kelantan which has capacity of 920 beds, and Hospital Universiti Sains Malaysia (Hospital USM), a teaching hospital in Kelantan with a total of 750 beds. Clinical samples were collected from both hospitals from January 2018 until January 2020. Inclusion criteria includes isolates from non-pregnant adults (age >18 years). Whereas the exclusion criteria were GBS isolated from vaginal swab samples (as most of the GBS isolated from non-pregnant adults represent vaginal colonizers) and GBS isolated from the same patient during the same episode of infection.

All the GBS isolates were identified based on its phenotypic characteristics as follows: colonies morphology was Gram-positive cocci in chain, catalase-negative and presence of narrow beta-haemolysis colonies on blood agar. GBS isolates produce a cytolytic toxin known as the CAMP factor encoded by the cfb gene. All isolates positive for group B Lancefield grouping (PathoDx™ Strep Grouping Kit, Thermo Scientific™, USA) were further confirmed by the presence of cfb gene using polymerase chain reaction (PCR). The confirmed GBS isolates were further tested to determine its specific serotypes and the presence of virulence genes.

Antibiotic sensitivity testing was routinely done as part of diagnostic procedure using disc diffusion Kirby Bauer technique. In this method, an agar plate was uniformly inoculated with GBS organism at 0.5 McFarland. A 6-mm filter paper disk impregnated with a known concentration of selected antibiotics for GBS (ceftriaxone (CRO-30 mg), clindamycin (DA-2 mg), erythromycin (E-15 mg) and penicillin (P-10 mg)) was placed on Mueller-Hinton agar plate with 5% sheep blood (Thermo Scientific™, USA). After an overnight incubation at 37°C, the diameter of inhibition zone was measured using Biomic Analyser (Giles Scientific Inc., USA). The results were interpreted based on CLSI guidelines [33].

2.2 DNA lysate preparation

All confirmed GBS isolates were subjected to DNA lysate preparation by using boiling method. A loopful of colonies was mixed with 100 µl of distilled water. The inoculum was boiled for 10 min and subsequently centrifuged at 12,000 g for 5 min to remove the cell debris. After centrifugation, the clear supernatant was transferred into a new sterile tube while the cell debris was discarded. This lysate was stored at –20°C before proceeded with specific GBS PCR serotyping.

2.3 PCR analysis

PCR amplification of the 193 bp cfb gene was performed using the specific primers 5'-GTGGCTGGTGCATTGTTATTTTCACCAGCTGTATTAGAAGTA-3’ and 5’-CATTAACCGGTTTTTCATAATCTGTTCCCTGAACATTATCTTTGAT-3’, as described by Ke et al. [34]. Apart from GBS confirmation by cfb-PCR, another PCR reaction for bacterial 16S ribosomal RNA [35] was also performed to validate the quality of each DNA template and omitting the possibilities of false negative results in further PCR analyses. 16S rRNA sequencing provides taxonomic resolution at species specific based on polymorphisms within the gene. All PCR analyses for serotyping and virulence gene screening were performed in a total volume of 25 µl PCR reaction mixture, containing 15.75 µl distilled water, 5.0 µl of 5X MyTaq Red reaction buffer, 1.0 µl each of 20 µM forward and reverse primer, 0.25 µl of 5U/µl MyTaq Red DNA polymerase, and 2.0 µl of DNA template (Bioline, UK). PCR cycling conditions consisted of an initial denaturation at 95°C for 5 min, 30 repeating cycles of denaturation at 95°C for 30 s, primer annealing at specific temperature (between 45°C and 55°C, depending on the primer pair used) for 30 s, and elongation at 72°C for 30 s, followed by final elongation at 72°C for 5 min. Post-PCR amplification products (amplicons) were analysed via agarose gel electrophoresis, in which 1% of agarose gel was used to separate the amplicons in according to their sizes. The gel electrophoresis was run at a constant voltage (90 v) for one hour. The agarose gel image was visualized and photographed using G-Box image analyser (Syngene, USA).

2.4 Determination of GBS serotypes and its virulence genes

GBS serotypes (Ia, Ib, II, III, IV, V, VI, VII and VIII) for each isolate were determined based on various capsular polysaccharide gene using previously published primers [29,36,37]. Seven virulence genes (scpB, lmb, bca, bac, rib, cylE and hylE) encoding for C5a peptidase, laminin-binding protein, à-C protein, ß-C protein, Rib protein, ß-hemolysin/cytolysin and hyaluronidase were also determined by PCR. Each GBS isolate was subjected to PCR serotyping and virulence gene screening by using each pair of primers. Isolates that were constantly negative for any GBS serotypes after two or three repeated PCR were considered as non-typeable. Representative amplicons from each GBS serotype and virulence gene were subjected to DNA sequencing to definitively confirm the specificity of PCR amplification.

2.5 Confirmation of GBS and virulence genes by DNA sequencing

Successfully amplified PCR products from each GBS and virulence genes were further validated through DNA sequencing. Representative amplicons were sent to Apical Scientific Sdn. Bhd., Selangor, Malaysia, for DNA sequencing services. The sequence data obtained from DNA sequencing was analysed by using a biological sequence alignment editor (BioEdit 7.2 software, Copyright 1991-2007 Tom Hall) [38] and Nucleotide BLAST: Basic Local Alignment Search Tool (National Centre for Biotechnology Information, USA), an online program for searching nucleotide databases using a nucleotide query (available at http://www.ncbi.nih.gov). Similarity index between the nucleotide query and database sequences is confirmed when the BLAST analysis scores hit ≥99% for both query coverage and identity.

2.6 Record tracing and data collection

The medical record of the patients was retrieved and reviewed. Patients’ data, including demographic details, underlying medical illness, clinical diagnosis and clinical outcomes, were extracted from the medical record for analysis. Antibiotic sensitivity testing results were collected from the Biomic V3 Microbiology System (Giles Scientific, USA) from both hospitals.

2.7 Statistical analysis

Statistical analysis was performed using IBM SPSS Statistics version 27. Descriptive statistics were used to summarize the socio-demographic of the patient, underlying medical conditions, clinical diagnosis of the infected patient, distribution of GBS serotypes and its virulence genes. Categorical data were presented as frequency (percentage). The association between GBS serotypes and the clinical diagnosis was determined using the Chi-square test.

3. Result

A total of 1703 GBS isolates excluding samples from urine were identified during the study period. Of that number, 4.4% (n = 75) were isolated from invasive samples from non-pregnant adults and were confirmed with the presence of cfb gene. Majority of patients were female (60%, n = 45) and Malay ethnicity (96%, n = 72). Detail clinical demographic data of all patients was summarized in Table 1. The age of the patient with GBS infection range from 18 to 89 with a mean of 54.04 years. Higher rate of infection was detected in the adult age group of 41–60 years (66.7%, n = 50). Majority of the isolates were obtained from sterile samples (53.3%, n = 40), including from blood (17.3%, n = 13), tissue (24%, n = 18), bone (6.7%, n = 5), and body fluid (5.3%, n = 4). The remaining 35 (46.7%) were isolated from non-sterile samples such as from pus swab 30.7% (n = 23), urine 13.3% (n = 10) and sputum 2.7% (n = 2). Clinically, most of the patients infected with GBS in this study (88%, n = 66) had at least one underlying medical illness. Half of them (50.7%, n = 38) had more than one risk factors, while majority of them had underlying DM (76%, n = 57) (see Supplementary Table 1).

Table 1. Socio-demographic data of non-pregnant adult with GBS infection and its serotype (n = 75).

		Serotype
Characteristic	n (%)	Ia	Ib	II	III	IV	V	VI	VII	VIII	IX	*NT
Age, years
18–40	8 (10.7)	2	0	2	1	2	0	0	1	0	0	0
41–60	50 (66.7)	11	0	9	12	8	1	0	4	0	0	5
61–90	17 (22.7)	1	0	1	4	0	3	3	1	2	0	2
Sex
Male	30 (40.0)	8	0	4	6	5	2	1	2	1		1
Female	45 (60.0)	6	0	8	11	5	2	2	4	1	0	6
Race
Malay	72 (96.0)	14	0	11	17	10	4	3	4	2	0	7
Chinese	1 (1.3)	0	0	0	0	0	0	0	1	0	0	0
Other: Siamese	2 (2.7)	0	0	1	0	0	0	0	1	0	0	0
Type of specimen
Blood	13 (17.3)	0	0	3	2	3	3	0	1	0	0	1
Tissue	18 (24.0)	6	0	1	5	0	1	2	1	1	0	1
Bone	5 (6.7)	0	0	3	1	0	0	1	0	0	0	0
Urine	10 (13.3)	2	0	0	3	3	0	0	0	0	0	2
Pus	23 (30.7)	6	0	1	6	3	0	0	4	0	0	3
Sputum	2 (2.7)	0	0	0	0	1	0	0	0	1	0	0
Body fluid	4 (5.3)	0	0	4	0	0	0	0	0	0	0	0
Types of infection*
*SSTI	50 (66.7)	11	0	7	10	5	2	3	6	1	0	5
*RTI	7 (9.3)	0	0	3	0	1	2	0	0	1	0	0
*BSI	3 (4.0)	0	0	0	3	0	0	0	0	0	0	0
*BJI	1 (1.3)	1	0	0	0	0	0	0	0	0	0	0
Genitourinary infection	12 (16)	2	0	0	4	4	0	0	0	0	0	2
Hepatobiliary infection	1 (1.3)	0	0	1	0	0	0	0	0	0	0	0
Peritonitis	1 (1.3)	0	0	1	0	0	0	0	0	0	0	0
Underlying conditions
Cardiovascular disease	5 (6.7)	0	0	1	1	2	0	1	0	0	0	0
Diabetes mellitus	57 (76.0)	12	0	9	14	3	4	3	5	2	0	5
Hypertension	35 (46.7)	6	0	5	7	3	3	3	3	1	0	4
Hyperlipemia	14 (18.7)	1	0	2	3	2	2	0	2	1	0	1
Liver disease	1 (1.3)	0	0	1	0	0	0	0	0	0	0	0
Renal disease	10 (13.3)	1	0	2	2	1	2	1	0	0	0	1
Autoimmune disease	1 (1.3)	1	0	0	0	0	0	0	0	0	0	0
Malignancy	4 (5.3)	1	0	0	1	2	0	0	0	0	0	0
Respiratory	0 (0.0)	0	0	0	0	0	0	0	0	0	0	0
Outcome
Alive	73 (97.3)	14	0	12	15	10	4	3	6	2	0	7
Died	2 (2.7)	0	0	0	2	0	0	0	0	0	0	0

*NT: Non-typeable; SSTI: Skin and soft tissue infections; RTI: respiratory tract infections; BSI: bloodstream infections, BJI: bone & joint infections. One patient can present with more than one underlying condition.

All the GBS isolates were tested positive by 16S ribosomal RNA PCR, confirming that the GBS DNA lysate was successfully amplified (Figure 1). This is important for the validation of true negative results in the further PCR serotyping and resistance gene screening, as the presence of PCR inhibitors might prevent amplification of target genes. Serotype distribution (Figure 2) shows that serotype III (22.7%, 95% CI [13.8–33.8], n = 17), Ia (18.7%, 95% CI [10.6–29.3], n = 14), II (16%, 95% CI [8.6–26.3], n = 12), and IV (13.3%, 95% CI [6.6–23.2], n = 10) were the most common serotypes isolated in this study. Other identified serotypes were VII (8%, 95% CI [3.0–16.6], n = 6), V (5.3%, 95% CI [1.5–13.1], n = 4), VI (4%, 95% CI [0.8–11.3], n = 3) and VIII (2.7%, 95% CI [0.3–9.3], n = 2). None of the serotypes Ib and IX were detected in this study. This study also showed that 9.3%, 95% CI [3.8–18.3] (n = 7) of the isolates were PCR-negative for all tested serotypes and hence were classified as non-typeable (NT).

Figure 1. Representative image of agarose gel electrophoresis for 16S ribosomal RNA gene. The presence of target amplicon (805 bp) in all isolates validated the quality of DNA template for further PCR analyses of this study. Lane M: 100 bp DNA ladder, lane N: no template control, lanes 1–18: representative GBS isolates.

Figure 2. Serotypes distribution of GBS isolated from various clinical samples (n = 75).

Representative images of agarose gel electrophoresis for the determination of GBS serotype and the virulence gene sequences are shown in Figure 3(a) and (b), respectively. The predominant virulence factors detected among the isolates were β-hemolysin/cytolysin (cylE) (86.7%, n = 65), C5a peptidase (scpB) (76%, n = 57), laminin-binding protein (lmb) (70.7%, n = 53), hyaluronidase (hylE) (66.7%, n = 50), and α subunits of C protein (bca) (61.3%, n = 46) (Figure 4). The rib and bac were detected in 10.7% (n = 8) and 2.7% (n = 2) of the total isolates. DNA sequencing for each GBS and virulence gene detected was carried out for nucleotide confirmation showed that all the specific serotypes have 100% similarities to the respective gene sequences of the GBS serotypes available in the NCBI database. Similar finding was also obtained for all the virulence gene sequences submitted for DNA sequencing. Hence, the results supported that all the GBS serotypes and virulence genes detected in this study were correctly identified.

Figure 3. (a) A representative gel electrophoresis of PCR amplification using primer target on cpsIII gene for serotype III (389 bp). PCR amplicons of cpsIII were shown in lanes 4, 7, 13 and 16. Lanes with no band indicate that the cpsIII gene was not detected. Lane M: 100 bp DNA ladder, lane N: no template control, lanes 1–17: representative GBS isolates. (b) A representative gel electrophoresis of PCR amplification to target virulence gene of scpB (255 bp). PCR amplicons of scpB were shown in lanes 3–6 and 8–10. Lanes with no band indicate that the gene was not detected. Lane M: 100 bp DNA ladder, lane N: no template control, lanes 1–10: representative GBS isolates.

Figure 4. Percentage of virulence genes detected in GBS isolated from non-pregnant adults (n = 75).

GBS isolates mainly discovered from patients presented with SSTI (66.7%, n = 50) and UTI (16%, n = 12). Among the SSTI cases, 90% (n = 45) had underlying DM. The most common serotypes associated with SSTI were Ia (22%, n = 11) and III (13.3%, n = 10). Apart from SSTI, 4.0% (n = 3) of the patients presented with bacteraemia without primary source and were associated with serotype III (Figure 5). Other manifestations seen in this studied population were peritonitis (serotype II), septic arthritis (serotype 1a) and ascending cholangitis in chronic pancreatitis patients (serotype II). None of the adults in this study presented with meningitis and endocarditis. In SSTI cases, cylE was the commonest virulence gene detected (80%, n = 40), followed by scpB (70%, n = 35). Others were lmb (66%, n = 33), hylE (66%, n = 33), rib (12%, n = 6) and lastly bca and bac (2%, n = 1) (see Supplementary Table 2). Among all these five GBS isolated from blood samples, virulence gene detected varies with no specific pattern of distribution.

Figure 5. Distribution of GBS serotypes isolated from patient with skin and soft tissue infections (n = 50).

Of all GBS isolated from SSTI patients, majority showed good sensitivity towards penicillin (98%, n = 49), erythromycin (88%, n = 44) and clindamycin (86%, n = 43). However, sensitivity towards tetracycline was low (24%, n = 12) (see Supplementary Table 2). Majority of the patients infected with GBS survived (97.3%, n = 73). Two mortality cases (2.7%) were recorded, in which the infected GBS isolates belong to serotype III. Both cases presented with septic shock secondary to severe SSTI and both succumbed to death within 24 hours of admission. Virulence genes detected for both cases were scpB and cyIE (first case) and scpB, cylE, hylB and lmb (second case). However, diabetic status for both of these patients was unknown since the death was recorded within 24 hours of admission. Nonetheless, regarding antibiotic sensitivity, both isolates show good sensitivity to all antibiotic tested (penicillin, erythromycin, clindamycin and tetracycline) (see Supplementary Table 2).

4. Discussion

Our study on GBS among non-pregnant adults demonstrated that the most predominant serotypes were serotype III (22.7%, 95% CI [13.8–33.8]) and serotype Ia (18.7%, 95% CI [10.6–29.3]). GBS serotypes III and Ia are among the most common serotypes associated with infections as reported elsewhere. For instance, several GBS studies on non-pregnant adults reported that serotype III was the commonest isolated serotype in global, including Thailand [7], the United States [11], Belgium [39], China [40] and Taiwan [41]. Whereas, a study conducted in Malaysia in 2003 reported that serotype Ia was the most common strain isolated among non-pregnancy related GBS infections [5]. Another study done in 2010 also showed that serotype Ia (45%) was the most common, followed by serotype III (16.7%) and V (15%). However, this study included placenta and high vaginal swab specimens [19]. A slightly higher proportion of serotype III than serotype Ia found in this study could be due to several reasons. Apart from the differences in study population and type of sample, other factors that may contribute to serotype dissimilarity could be the small sample size and geographical location. In contrast to this study, those two studies were conducted on the west region of peninsular Malaysia. Additionally, there might be some changes in term of serotype distribution of GBS after 10–15 years of time.

GBS serotype distribution is also known to vary with age groups, across regions and population. In this study population, the majority of GBS infections were more common in the age group of 41–60 years old as compared to the global incidence, which is more likely to occur in the elderly of above 65 years old. This finding is consistent with the previous study in Malaysia in 2009 [5]. Previous study from a neighbouring country, Thailand, reported that the highest incidence was among the age group of 20–60 years old compared to the elderly over 65 years [7]. The shorter life expectancy of Asian people than that of Caucasian may contribute to this disparity [5,11]. Additionally, Malaysia is a multi-ethnic country. Malays (95.7%) are the predominant ethnic group in Kelantan, with Chinese (3.4%), Indian (0.3%), Siamese and others as minorities. This is proportionate with our finding that 96% of GBS diseases were accumulated among Malays.

In the present study, virulence genes cylE, lmb, scpB, hylB and bca were detected in more than 60% of GBS isolates, as illustrated in Figure 4. The findings were different from a previous study done in France in 2010 that reported only 2 out of 111 strains lacked the scpB and lmb genes [42]. However, in contrast to our sample population, their study has included the samples from paediatric populations with neonatal meningitis infection. The bca gene was found at a higher prevalence (61.3%) as compared to a previous report in 2015 [16]. The rib gene was detected in 10.7%, and only two isolates were positive for bac gene. The cylE, lmb, scpB, hylB and bca genes were mainly found in serotype III. Another study identified several virulence genes, i.e. cylE, lmb, scpB and hylB, which were present in more than 94% of the isolates, whereas bac (9.7%), bca (14.6%), and rib (29.1%) genes were detected at a lower rate. The authors reported a significant association between the specific serotype and the virulence genes, but the virulence contribution could not be confirmed [16]. The prevalence of predominant virulence genes such as cylE, lmb and scpB positive isolates in this study is lower compared to other studies [16,43]. Further investigations by using phenotypic characterization on culture media, alternative target primers, or whole genome sequencing might be useful to understand this finding.

In terms of risk factors, more than half of the GBS patients in this study were diagnosed with DM. In addition, DM is more prevalent in Malays than in other ethnicities. In the latest 2019 record, 56.2% of 1.6 million diabetes patients in Malaysia were from the Malay ethnic [44]. These data eventually support the higher proportion of GBS infections among Malays and among those with DM. Besides, DM is also the leading risk factor causing GBS disease in the USA, Europe and Thailand populations [4,7,12]. A previous study in Argentina reported that GBS infections in adults were mainly associated with DM, and Ib was the most common serotype [45]. In contrast, our study population lacked of serotype 1b. One of the possible reasons could be due to different epidemiological region, as it is known that different locations encompass different serotype distributions. In our study, GBS was mainly discovered in patients diagnosed as SSTI (66.6%, n = 50). Among them, 90% (n = 45) were diabetics. A similar study done in Malaysia also reported the similar finding, where the most common clinical presentation was SSTI (71.4%, n = 35) and DM was the most common underlying condition (71.4%, n = 35) [5]. Diabetic patients with SSTI have a high tendency to be infected with GBS. DM is potentially associated with abnormalities in phagocyte function. The hyperglycaemic state and the presence of glycation end products reduce the neutrophil response, which leads to a defect in bacterial engulfment by the neutrophil [10]. Patients with underlying diabetes may have peripheral neuropathy or peripheral vascular disease as a complication of uncontrolled glucose. Injury to the lower extremities may alter the skin and mucous membrane integrity, which allows the colonizing GBS to penetrate the skin and mucous membrane, causing SSTI [10].

Among SSTI cases, the majority of the serotypes detected were types Ia (22%, n = 11), III (13.3%, n = 10), II (9.3%, n = 7) and VII (8.0%, n = 6). In contrast to our findings, a study in Taiwan by Lee in 2005 showed that type V was the most frequent serotype (32.5%), followed by types III (20%), Ia (15%), Ib (7.5%) and IV (2.5%) [28]. Limited data is available to specifically analyse the serotypes among SSTI patients. There is no statistically significant association between the most common serotypes and SSTI (p = 0.376). This possibly related to the small sample size involved in our study. In contrast to the study done in Thailand, which engaged a larger sample size, there was a significant association between the commonest serotype and clinical presentation [7]. They reported that serotype III has significant associations with meningitis, sepsis and septic arthritis, while serotype V was significantly associated with UTI. Meanwhile, a study in Japan also reported that capsular serotype had no significant correlation with clinical diagnosis, even though this study involved more than 400 cases [46].

The mortality rate for GBS infections in this study was low (2.7%) in comparison to other studies. Chaiwarith et al. reported that the mortality rate of GBS infections in non-pregnant adults ranged from 3% to 30% in northern Thailand [11]. This variation depends on the severity of the infection, whether it is associated with bloodstream infection, hypotension, concurrent infection, virulence strain and older age. Early diagnosis and proper management like administering appropriate antibiotics and early surgical intervention can reduce the mortality rate [9]. Both of the fatal cases in this current study belong to serotype III and were found to be associated with more than one virulence gene. One of the patients was diagnosed with septic shock secondary to leg necrotizing fasciitis. The GBS isolated from this patient carried scpB, lmb, bca, cylE and hylE virulence genes. In another fatal case, the patient had severe sepsis and was isolated with GBS that carried scpB, lmb, bca, bac, cylE and hylE virulence genes.

The majority of GBS isolated from SSTI patients in this study showed good sensitivity towards penicillin. The finding was supported by several reported findings from other studies [47–49]. Erythromycin and tetracycline, which were prescribed to those who were allergic to penicillin, showed a sensitivity of 88% and 24%, respectively. This study also discovered that one isolate of serotype IV had reduced susceptibility to penicillin. This isolate was also found resistant to clindamycin and erythromycin. Kimura and colleagues were the first to account for penicillin reduced susceptibility GBS (PRGBS) in 1995 [50]. According to a study conducted in Japan, the penicillin resistance rate of GBS increased dramatically from 2.3% to 14.7% between 2005 and 2013 [51]. GBS with reduced susceptibility to penicillin has also been reported in Canada [52], Korea [53], Germany [54] and Thailand [55].

One study proposed that patients receiving long-term penicillin or beta-lactam treatment are at risk of harbouring PRGBS [56]. In our case, a patient who received long-term penicillin presented with lower urinary tract symptoms and was diagnosed with a prostate tumour (see Supplementary Tables 1 and 2: GBS isolate number 66). However, the presumptive finding of PRGBS in this patient needs further confirmation, for instance, by molecular testing to detect amino acid substitution at PBPs. Clindamycin and erythromycin resistance is increasing globally [12]. Erythromycin resistance in GBS is caused by an efflux pump mechanism encoded by mefA genes and ribosomal modification expressed by erm genes (ermB, erm A/TR). Whereas, apart from erm genes, clindamycin resistance may also be due to ribosomal translocation mediated by linB genes. One study from Taiwan reported that GBS isolates have a varied susceptibility to clindamycin, azithromycin, erythromycin and chloramphenicol, ranging from 53.5% to 71.4%, and most of their isolates (95%) were resistant to tetracycline [28]. In Thailand, the prevalence of resistance measured for tetracycline, clindamycin, azithromycin, and erythromycin were 41.3%, 22.0%, 22.0% and 22.0%, respectively. Strains that were resistant to all four of those drugs were significantly associated with non-serotype III (p < 0.001) [55]. Clindamycin resistance is clinically significant and raises concern among clinicians, as clindamycin is used as the primary therapy for SSTI caused by the toxic producing organism [4]. However, the prevalence of resistance in Malaysia was low in the previous study. This study also reported a low prevalence of resistant to clindamycin (14%) and erythromycin (12%). Of seven isolates with clindamycin resistance, two were discovered from serotype IV, while one case was reported from each of the serotypes Ia, II, V, VII and IV. Both isolates of serotype IV that were resistant to clindamycin were also resistant to erythromycin.

One of the limitations of this study includes population coverage, which involves only one state in Malaysia. The short study period, budget constraints and pandemic COVID-19 are among the reasons for the small sample size involved in this study. Another limitation of this study is the exclusion of non-haemolytic GBS that could lead to the low prevalence of GBS isolates in non-pregnant adults in this study. Thus further studies with a larger sample size and inclusion of non-haemolytic GBS are needed to confirm the significance of the findings. However, the selection of Gram-positive cocci in chain, catalase-negative, and narrow beta-haemolysis colonies on blood agar, in addition to the presence of cfb gene by PCR, confirmed that all the 75 isolates were GBS. Further analyses on GBS subtyping and screening of PRGBS, erythromycin and clindamycin resistant isolates were not carried out as it was beyond the scope of our study. Additionally, screening of macrolide-lincosamide resistant GBS phenotypes was not routinely done in our diagnostic lab. Although capsular typing by PCR offers better accuracy and sensitivity than conventional methods, the sensitivity relies on the use of primer sequences, types of Taq polymerase enzyme, optimized concentration of PCR reagents, cycling parameters and quality of DNA template. Previous studies used latex agglutination kits and had successfully serotyped more than 95% of the GBS isolates [57,58]. Meanwhile, some other studies used both latex agglutination kits and PCR to maximize the assignment of the GBS strains [59]. Therefore, we suggest that a combination of PCR and conventional serotyping methods should be used in future studies to accurately assign the GBS strains into their specific serotypes.

5. Conclusion

This study reported that GBS serotypes III and Ia were the predominant strains in non-pregnant adults, in consistent with the global findings. The common virulence genes detected were cylE, scpB, lmb, hylE and bca. Invasive GBS among non-pregnant adult presented with SSTI is common. DM is the additional risk of acquiring GBS disease and can result in increased morbidity and mortality. GBS isolates showed good sensitivity towards penicillin and macrolides. Although the macrolide resistance GBS phenotype was low, routine screening for macrolide-lincosamide (ML) phenotypes is recommended. The detection of serotypes and virulence genes may help in GBS vaccine development. GBS vaccine for adults should most likely be considered as part of a prevention strategy to reduce the incidence and mortality rate of invasive GBS infection among non-pregnant adult.

Ethical approval

Not required.

Acknowledgements

We express our gratitude to all staffs in Department of Medical Microbiology USM who involved in this study especially Laboratory technologist Amanina binti Aminudin for helping in the collection of the isolates and the director of the Hospital Universiti Sains Malaysia (USM), Kubang Kerian, Kelantan for granting the investigators permission to use patients’ medical record, space and assets belong to the hospital during the process of conducting the research.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This study was funded by USM Research Incentive Grant No 1001/PPSP/AUP5001 and USM RUI Grant no 1001/PPSP/8012258.