Potentially pathogenic microorganisms from removable prosthesis biofilms have been suggested to be associated with significant systemic health problems including gastrointestinal infections, bacterial endocarditis, and respiratory infections [1]. The association between opportunistic respiratory pathogens and the oral cavity has been gaining popularity in the dental and medical fields because of the increased risk of pneumonia [2], [3], [4], [5], [6], [7], [8]. Two studies have shown that half of the removable prostheses examined were colonized by opportunistic respiratory bacteria which may lead to life-threatening respiratory tract infections [9,10]. The colonization of the respiratory pathogens was mainly associated with compromised oral immunity, reduced oral commensal bacteria, and maturation of oral plaque [2]. Notably, a high correlation was found between the microbiome residing on removable prostheses and the pharynx [11], suggesting the possibility of pathogenic bacteria on prosthesis surfaces facilitated the process of colonisation or infection of the pharynx. The finding was consistent with Venkataraman et al. study [12], which reported that microorganisms residing intraorally may be the primary source of the lung microbiome.

Pneumonia for older adults is mainly associated with underlying comorbidities, compromised mucociliary functions, and a decreased host immune system. The predominant respiratory pathogens of community-acquired pneumonia are Streptococcus pneumoniae and gram-negative enteric bacilli including Klebsiella pneumoniae and Escherichia coli. However, for nosocomial pneumonia, Staphylococcus aureus and gram-negative enteric bacilli are the main respiratory pathogens [13]. In addition, the major source of pneumonia may also be caused by micro- or macro-aspiration of pathogenic pharyngeal bacteria [14], which has been strongly correlated with the bacterial species on removable prostheses [11]. Furthermore, the fungi and bacteria on prosthesis surfaces were also reported to trigger secondary coinfections and aggravate existing pulmonary infections, resulting in longer hospitalization times and increased risk of death [15]. Unlike young adult dentate patients, the oral and prosthesis hygiene care of frail elders with comorbidities is difficult to be maintained due to their poor manual dexterity, reduced cognitive function, and or compromised systemic health. Therefore, this group of the population is potentially exposed to these unexpected respiratory infections.

Pathogenic bacteria screening using the Microbial Index of Pathogenic bacteria (MIP) [16], a type of microbiome-based index allows the risk assessment of oral and systemic health diseases, for various targeted human organs or body sites [17]. In addition, the MIP can potentially help to control the transmission of pathogens and reduce the risk of oral and systemic diseases. Various microbiome-based indices have been developed for skin health, gut microbiome, dental caries, and periodontal diseases [18], [19], [20], [21]. However, the MIP for clean and unclean removable prostheses, particularly MIP of the respiratory system is unknown. Removable prosthesis-wearing was found to have a 7-fold higher risk associated with community-acquired pneumonia compared with the control group [22]. Prosthesis-wearing at night also increased the risk of pneumonia [23]. Kusama et al. [24] reported that infrequent denture cleaning was significantly associated with the self-reported incidence of pneumonia. Thus, the association of removable prosthesis hygiene care and wearing habits is of critical importance, not just for oral health, but also for life-threatening opportunistic respiratory infections. A microbiome-based index for respiratory pathogens residing on the removable prosthesis surfaces can be calculated as the sum of the relative abundance of opportunistic respiratory pathogens in a microbial community according to 300 published categories of opportunistic pathogenic bacteria by the Chinese Center for Disease Control and Prevention. This newly developed MIP can be named as Opportunistic Respiratory Pathogenic Index (ORPI), which allows the screening of pathogenic respiratory bacteria, via high-throughput metagenomic sequencing (Type IIB Restriction-site Associated DNA Sequencing for Microbiome, 2bRAD-M) [16,25].

Scientific evidence of MIP is lacking, and there is scant evidence to support this newly developed ORPI. Therefore, this study aimed to address this knowledge gap by determining and comparing the ORPI and the prevalence of respiratory pathogens between clean and unclean removable prostheses. The removable prosthesis cleanliness was determined by the percentage plaque area coverage (PPC) after quantification using a semi-automated planimetric assessment [26]. The null hypothesis was that no significant difference in ORPI and prevalence of respiratory pathogens was observed in removable prostheses with different levels of cleanliness.

The primary outcome of this study was that there would be at least a 30 % difference in the abundance of respiratory pathogens colonising prostheses that were classified as clean (PPC < 25 %) compared to unclean removable prostheses (PPC ≥ 25 %). Removable prostheses were considered as a reservoir for respiratory pathogens with the most prevalent respiratory pathogen (Staphylococcus aureus) detected at 1.3 × 105 colony-forming equivalent (CFE) per prosthesis [9]. Based on the hypothesis that ‘be at least a 30 % difference in the abundance of this common respiratory pathogen’ [27], S. aureus was estimated to be detected at 1.0 × 105 CFE among ‘clean prostheses’ allowing for a standard deviation of approximately half the mean (i.e. 0.5), then a sample size of 44 (per group) would have an 80 % statistical power to identify significant differences in the abundance of common respiratory pathogens colonising removable prostheses between ‘unclean’ and ‘clean’ groups. To allow for the potential for errors is prudent to sample 10 % more, thus a total of 96 removable prostheses was required.

After ethical clearance (IRB Reference Number: UW22-256) was obtained from the Institutional Review Board of the University of Hong Kong/Hospital Authority Hong Kong West Cluster, participants (97 removable prosthesis wearers) who complied with the following inclusion criteria were enrolled in the study using a convenience sampling method from October 2022 to February 2023. Participants were recruited from those attending the Prince Philip Dental Hospital. The inclusion criteria were a participant must be aged 18 years and above and wearing a removable prosthesis for at least 3 months. Exclusion criteria included participants who had received antimicrobial or antifungal treatment in the past month, taken antibiotics within a month, taken steroids in the past 6 months, undergone radiotherapy or chemotherapy, diagnosed with upper or lower respiratory tract infections within a month, and rinsed with mouthwashes prior to sample collection. Written informed consent was provided by all participants and they agreed to comply with the study protocol. Microbiological laboratory technician and data analyst at Qingdao OE Biotech Co., Ltd. (Qingdao, China) and Faculty of Dentistry (The University of Hong Kong) were blinded to the group classification.

After plaque area coverage images of the removable prostheses were obtained and quantified using a semi-automated planimetric assessment method, prostheses were grouped as clean (PPC < 25 %) or unclean (PPC ≥ 25 %) [26]. Prosthesis plaque sample collection [28] was performed by placing the prosthesis in a sterile bag containing 50 mL sterilized phosphate-buffered saline and then immersing it in an ultrasonic bath for 15 mins at 45 kHz to remove the plaque from the surface. In the microbiology laboratory, the sonicate was centrifuged at 9880 rpm (14,000 g) for 10 mins and the plaque pellet was resuspended in 180 μL of 20 mM Tris–HCl; 2 mM EDTA; 1.2 % Triton with 20 μL of 20 mg/mL lysozyme and incubated overnight at 37 °C. 20 μL proteinase K extraction buffer was added and mixed by vortexing. These samples were incubated at 56 °C for 2 h followed by 95 °C for 15 mins. Ethanol (200 μL) was added to the sample and mixed by pulse-vortexing for 15 s. A QIAmp Mini DNA Extraction Kit (Qiagen GmbH, Germany) was used to extract the DNA according to the manufacturer's instructions. DNA concentration was quantified using the Qubit 2.0 Fluorometer (Life Technologies, Carlsbad, California, US) prior to storing at −70 °C.

The sequencing was performed by the principle of 2bRAD-M [25] at Qingdao OE Biotech Co., Ltd. In brief, the extracted DNA was digested using a Type IIB restriction enzyme (BcgI), followed by fragments enrichment and random amplification for DNA sequencing. With 2bRAD-M sequencing data (32-bp long reads), the species-resolved compositional profile for each prosthesis plaque sample was obtained from the bioinformatic pipeline (https://github.com/shihuang047/2bRAD-M). For each microbiota, microbial species were identified according to a prebuilt 2bRAD species-specific marker database. For each species, the abundance was estimated based on the sequencing coverage of its species-specific markers.

The relative abundance in this study was calculated as follows: the average read coverage of 2bRAD markers for each species was confirmed and the ratio was determined, by dividing this value by the total number of detected microorganisms in a sample. The detailed formula and analysis have been previously published [16]. Species-level abundance profiles were obtained to determine ORPI scores for each prosthesis microbiota. ORPI was calculated as the sum of the relative abundance of opportunistic pathogenic respiratory pathogens for each sample according to 300 published categories of opportunistic pathogenic bacteria by the Chinese Center for Disease Control and Prevention [17,29], using Eq. (1). The index ranged from 0 to 1. The R scripts for this study and the list of opportunistic respiratory pathogenic bacteria are presented at https://github.com/yfz-96/ORPI. The variations in the prevalence and ORPI between the two groups were determined using the Wilcoxon signed-rank test. The significance level was determined at 0.05 and the analysis was performed using R software (version 4.2.1).

For the ORPI calculation using Eq. (1), M denotes the number of respiratory pathogens in a sample, while N denotes the number of all microbes identified in a sample. m denotes the relative abundance of a microbe in a sample.ORPI=∑i=1Mpathogensi∑j=1Nmj

The prevalence of opportunistic respiratory pathogens in the clean and unclean prostheses was calculated using Eq. (2), in which N denotes the number of samples, and w denotes the presence of opportunistic respiratory pathogen in the sample.Prevalence=∑i=1NwiN