The recent COVID-19 pandemic has raised serious concerns regarding disease and human health. In terms of bacterial infection, antimicrobial resistance is a global issue that threatens to worsen infections. Antimicrobial agents used to treat bacterial infections have been evaluated with artificial media such as Mueller–Hinton medium. However, several reports show that data obtained from artificial media do not always correspond to reactions in vivo, suggesting that effective compounds may have been overlooked by conventional screening methods. If compounds are evaluated under conditions similar to those inside the human body, assessments of effective compounds may be made more accurate. Experimental animals have been employed as a method to provide in vivo information, but animal models may not translate across species and could, furthermore, create ethical issues around their use. 3D organotypic tissue models have attracted attention in this context. As these models are composed of human culture cell lines, problems surrounding species compatibility and ethics can be avoided. In addition, the state of the model is closer to that of the human body than either monolayer cell culture or artificial media. The 3D tissue model can thus be used to reproduce infectious diseases in vitro, as well as evaluate therapeutic drugs under environmental conditions similar to those found in the human body. This review summarizes the characteristics of the 3D tissue model and its construction method and discusses current infection models for respiratory tract infections.

Antimicrobial susceptibilities for anaerobic bacteria and microaerophilic oral streptococci isolated from respiratory tract or oral cavities at medical facilities in the Chubu region of Japan between January, 2017 and December, 2018 were examined. β-lactamase production test was also performed on anaerobic bacteria.

β-lactamase was detected in 54.9% of Prevotella species. The susceptibility rates of Prevotella species for penicillin/β-lactamase inhibitor combinations and carbapenems were 100%, respectively. The susceptibility rates for ceftriaxone (CTRX), moxifloxacin (MFLX) and clindamycin (CLDM) were 86.3%, 72.6% and 74.5%, respectively. There were strains with decreased susceptibility to clarithromycin (CAM) or azithromycin (AZM).

The susceptibility rates of Fusobacterium species for penicillin/β-lactamase inhibitor combinations, CTRX and carbapenems were 100%, respectively. The strains resistant to MFLX or CLDM were detected. The MIC90 values of CAM and AZM were >64 μg/mL and >16 μg/mL, respectively, suggesting decreased antimicrobial activity of CAM and AZM.

As with Fusobacterium species, the susceptibility rates of Parvimonas species for penicillin/β-lactamase inhibitor combinations, CTRX and carbapenems were 100%, respectively. The susceptibility rate for MFLX was 92.3%. The MIC90 values of CAM, AZM and CLDM were >64 μg/mL, >16 μg/mL and 16 μg/mL, respectively. Decrease of susceptibility to CAM, AZM and CLDM was observed.

Penicillin/β-lactamase inhibitor combinations, CTRX and carbapenems showed good antimicrobial activity against Streptococcus anginosus group. The susceptibility rates for CTRX and meropenem (MEPM) were 100%, respectively. The antimicrobial activities of quinolones were weaker than those of carbapenems. The susceptibility rates for CAM and AZM were 74.2%, respectively. CLDM also showed good antimicrobial activity, the MIC90 value was 0.0625 μg/mL. However, there was one strain showed resistance to CLDM.

In this study, anaerobic bacteria and oral streptococci isolated from respiratory tract or oral cavities showed high susceptibility to penicillin/β-lactamase inhibitor combinations and carbapenems. However, decrease of susceptibility to macrolides and CLDM was observed. This suggests the importance of continuous surveillance to assess the antimicrobial susceptibility.