Clinical research advances in pulmonary infections in 2025: a narrative review of multipathogen assessment, precision diagnosis, and integrated management

Abstract

Background and Objective

The diagnosis and treatment of pulmonary infections have become increasingly complex owing to antimicrobial resistance, the growing population of immunocompromised patients, and the ongoing circulation of respiratory viruses. In this narrative review, the major clinical advances reported in 2025 are summarised, and evidence-based support for multipathogen assessment, precision diagnosis, individualized treatment, and integrated management of pulmonary infections is provided.

Methods

PubMed, Web of Science, and the Chinese Medical Association Journal Database were searched for relevant English-language and Chinese-language publications. The search primarily focused on studies published in 2025 and was supplemented by the manual retrieval of relevant clinical studies and guidelines. Medical Subject Headings (MeSH) and free-text terms were used in PubMed, with corresponding free-text terms applied in the other databases. Clinically relevant publications were screened according to predefined criteria by two independent reviewers, and expert adjudication was used in the event of disagreements.

Key Content and Findings

Metagenomic next-generation sequencing (mNGS), integrated biomarker assessment, and artificial intelligence (AI) have exhibited increasing clinical value in pathogen identification, differential diagnosis, and prognostic assessment. Therapeutic advances have been reported for drug-resistant bacterial and fungal infections, as well as for coronavirus disease 2019 (COVID-19) and influenza, supporting more precise and individualized treatment. Studies on the gut–lung axis and immune modulation have further clarified the potential role of host-directed approaches, although clinical evidence for microbiota-targeted interventions remains limited. The risks associated with coinfection received increasing attention, while improved identification of risk factors and stratified management supported more individualized care for high-risk populations.

Conclusions

In 2025, the clinical management of pulmonary infections continued to transition towards precision-oriented and host-directed models. However, because a quantitative evidence synthesis was not included in this narrative review, pooled effect size estimates or formally graded clinical recommendations are not provided. The heterogeneity of the available studies also limits the overall strength and generalisability of the evidence. Future high-quality, multicentre studies are needed to further standardise clinical diagnosis and treatment strategies.

Introduction

Pulmonary infections are common in clinical practice. Their aetiology is complex, and their diagnostic and therapeutic challenges have intensified because of antimicrobial resistance, the growing number of immunocompromised patients, pathogen evolution, the persistent circulation of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), and the routine transmission of other respiratory viruses. In 2025, important advances were made in precision diagnosis, anti-infective agents, host immune regulation, and prognostic assessment systems. Together, these developments have supported a more integrated approach involving precise pathogen identification, targeted treatment, host modulation, and risk-based management. In this review, recent clinical advances in pulmonary infections caused by bacteria, fungi, and common respiratory viruses in 2025 are systematically summarised, and their implications for clinical diagnosis and treatment are discussed. This review focuses on three clinically relevant questions related to pulmonary infection research in 2025: recent advances in diagnosis, anti-infective and host-directed treatment, and risk-based integrated management. We present this article in accordance with the Narrative Review reporting checklist (available at https://jtd.amegroups.com/article/view/10.21037/jtd-2026-1375/rc).

Methods

The literature search for this review was conducted between 1 and 31 January 2026. To ensure comprehensive coverage, we searched PubMed, Web of Science, and the Chinese Medical Association Journal Database, supplementing this with manual searches of relevant clinical studies to minimize omissions inherent to database retrieval. In PubMed, Medical Subject Headings (MeSH) were combined with free-text terms; corresponding English free-text terms were used in Web of Science, and Chinese terms were used in the Chinese Medical Association Journal Database. The search query included terms related to pulmonary infections, pulmonary fungal infections, antimicrobial resistance, antifungal therapy, atypical bacterial pathogens, viral respiratory infections, and molecular diagnostic methods. Free-text terms included “pulmonary infection”, “pulmonary aspergillosis”, “invasive fungal disease”, “multidrug resistance”, “atypical pneumonia”, “Mycoplasma pneumoniae”, “Chlamydia pneumoniae”, “Chlamydophila pneumoniae”, “Legionella pneumophila”, “metagenomic next-generation sequencing”, and “molecular diagnostics”, among others. For the PubMed search, filters were applied for human studies and English-language publications. Chinese-language clinical studies, guidelines, and consensus documents were identified through the Chinese Medical Association Journal Database and via manual retrieval when directly relevant to clinical practice. The detailed PubMed search strategy is provided in Table S1.

The primary search period was from 1 January to 31 December 2025, with a focus on the latest developments in pulmonary infection research during that year. Selected authoritative guidelines published outside this period were additionally considered when necessary to provide clinical context. Eligible publications included clinical trials, observational studies, clinically relevant translational studies, reviews, guidelines, and consensus documents related to the diagnosis, treatment, immune modulation, and prognosis of pulmonary infections. The scope covered bacterial, fungal, SARS-CoV-2, and influenza virus-related pulmonary infections, with an emphasis on molecular diagnostics, anti-infective agents, immunological interventions, and prognostic assessment. No geographic restrictions were applied during study selection. Differences in geographic settings were considered when interpreting the applicability of the findings. The exclusion criteria were basic research without clear clinical applicability, methodologically weak case reports, letters, and expert opinions without original clinical data, studies with incomplete data or inaccessible full texts, and publications outside the scope of this review.

Study selection was performed independently by two researchers (X.L. and Z.L.) with professional backgrounds in respiratory medicine and clinical epidemiology. Irrelevant studies were first excluded on the basis of their titles and abstracts, and potentially eligible publications were then assessed via full-text review. Any disagreements were resolved through discussion with a senior professor in respiratory medicine until consensus was reached. The overall search strategy is summarised in Table 1.

Table 1. Summary of the search strategy.

Items	Specification
Date of search	1 January to 31 January, 2026
Databases and other sources searched	PubMed, Web of Science, and the Chinese Medical Association Journal Database, supplemented by manual searches of relevant clinical studies and guidelines
Search terms used	MeSH and free-text terms related to pulmonary infections, fungal infections, antimicrobial resistance, antifungal therapy, atypical bacterial pathogens, respiratory viral infections, and molecular diagnostics. Corresponding English or Chinese terms were used as appropriate. The detailed PubMed strategy is provided in Table S1
Timeframe	The primary search covered the period from 1 January to 31 December 2025. Selected authoritative guidelines published outside this period were included when necessary for clinical context
Inclusion and exclusion criteria	Clinical trials, observational studies, translational studies, reviews, guidelines, and consensus documents relevant to the diagnosis, treatment, immune modulation, or prognosis of pulmonary infections were included
	Publications without clear clinical relevance, methodologically weak case reports, letters, and expert opinions without original clinical data, studies with incomplete data or inaccessible full texts, and articles outside the scope of the review were excluded
Selection process	Two researchers (X.L. and Z.L.) independently screened titles, abstracts, and full texts. Disagreements were resolved through discussion with a senior reviewer
Additional considerations	English-language publications were identified through PubMed and Web of Science, and Chinese-language publications through the Chinese Medical Association Journal Database. No geographic restrictions were applied

MeSH, Medical Subject Headings.

Use of artificial intelligence (AI) tools

Most English translation and language editing were undertaken by the authors. From 1 to 24 July 2026, ChatGPT (OpenAI; GPT-5.6 Thinking) was used to assist with the translation of selected passages summarising cited studies and the language editing of selected English paragraphs. The relevant prompts focused on translating the core content of cited references and improving the wording of specific passages while preserving their intended scientific meaning. The tool was not used to generate research data, perform statistical analyses, determine study eligibility, or make final scientific judgments. All AI-assisted text was checked and revised by the authors, who take full responsibility for the accuracy and integrity of the final manuscript.

Advances and clinical translation of precision diagnostic technologies

Rapid and accurate pathogen identification is central to the management of pulmonary infections. Traditional culture-based diagnostic methods are time-consuming and may delay pathogen-directed treatment, particularly in critically ill patients. In 2025, advances in molecular diagnostics, optimisation of respiratory sampling approaches, broader application of biomarkers, and increasing use of AI-assisted models supported pathogen identification, differential diagnosis, risk stratification, and prognostic assessment in pulmonary infections.

Expanded applications and multidimensional advances in molecular and multimodal diagnostics

The clinical value of metagenomic next-generation sequencing (mNGS) in the diagnosis of pulmonary infections was validated, and it exhibited particularly prominent advantages for severe infections. Studies published in 2025 expanded the boundaries of this technology by enabling one-step precision differentiation between pulmonary infections and lung cancer, thereby offering a new diagnostic approach for difficult cases with overlapping clinical manifestations. In a study of 223 patients with suspected pulmonary infections or tumours, Li et al. (1) reported that the Onco-mNGS assay, an integrated mNGS approach combining pathogen identification with host chromosomal copy number variation analysis for tumour detection, achieved a sensitivity of 81.82% and an accuracy of 77.58% for pulmonary infections and showed higher sensitivity and accuracy than conventional testing. When bronchoalveolar lavage fluid (BALF) was used, the sensitivity and specificity for lung cancer diagnosis reached 87.5% and 100%, respectively. The method also yielded a favourable turnaround time: 92.36% of the reports were available within 48 hours, and 58.30% were completed within 24 hours (1). Thus, while addressing the frequent diagnostic confusion between infection and malignancy, this approach maintained high diagnostic efficiency and provided a diagnostic approach for patients with suspected pulmonary infections or malignancy. Wu et al. (2) evaluated the impact of mNGS on clinical outcomes in patients with severe community-acquired pneumonia (CAP). Their results showed that mNGS facilitated early pathogen identification and informed antimicrobial adjustment. Compared with the conventional microbiological testing group, the mNGS group had a significantly shorter time to clinical improvement, indicating that the median time to clinical improvement was reduced from 13 to 10 days in critically ill patients. In invasive pulmonary aspergillosis (IPA), BALF mNGS also showed favourable diagnostic performance. Zhao et al. (3) reported that BALF mNGS achieved a sensitivity of 94.03% and a specificity of 96.20% and showed higher overall diagnostic performance than culture, galactomannan (GM) testing, and polymerase chain reaction (PCR), helping address delays in fungal diagnosis.

The clinical use of targeted molecular panels has become more refined, providing efficient diagnostic solutions for infections in defined clinical settings. In intensive care units (ICUs), the FilmArray Pneumonia Plus panel exhibited excellent diagnostic performance for the 15 typical bacterial targets included in the panel; compared with conventional microbiological culture, it achieved a sensitivity of 94% and a specificity of 98%, enabling rapid identification of bacterial targets and supporting early microbiological assessment in hospital-acquired pneumonia (HAP) and ventilator-associated pneumonia (VAP) (4). In a multicentre study on patients with suspected VAP and HAP, Enne et al. (5) examined the effect of molecular diagnostic-guided therapy. Rapid syndromic PCR improved antibiotic stewardship at 24 hours, although non-inferiority was not demonstrated for clinical cure at 14 days. These findings indicate that rapid molecular testing should complement, rather than replace, conventional culture and clinical assessment.

The increasing availability of point-of-care and non-invasive diagnostic approaches has provided practical options in resource-limited settings and for patients in whom invasive testing is difficult. Aspergillus-specific lateral flow devices (LFDs) can detect serum antigen within approximately 15 minutes. A diagnostic meta-analysis showed that LFDs had a pooled sensitivity of 0.67 and specificity of 0.90 for invasive aspergillosis. BALF-based testing was more sensitive than serum-based testing (0.72 versus 0.49), indicating that specimen source should be considered when interpreting the results (6). Among selected patients with chronic lung disease and positive sputum cultures for Aspergillus, the combination of Aspergillus-specific immunoglobulin G and high-resolution computed tomography (HRCT) scores achieved an area under the receiver operating characteristic curve (AUC) of 0.90 for distinguishing clinically diagnosed Aspergillus infection from colonization. This approach may help identify patients who require further invasive evaluation and reduce unnecessary antifungal treatment (7).

Optimisation of pulmonary infection sampling and less invasive alternatives

Bronchoalveolar lavage (BAL) is widely used to obtain lower respiratory tract samples for the diagnosis of IPA, but deep BAL may be difficult to perform in critically ill patients. In a prospective exploratory pilot study of 40 critically ill patients, Schroeder et al. (8) compared non-directed bronchial lavage (BL) with BAL for Aspergillus testing. GM enzyme immunoassay results showed a strong correlation between the two sample types (Pearson’s r=0.78). At optical density index cutoffs of 0.8 for BAL and 1.2 for BL, sensitivity was 94% and specificity was 67%. Aspergillus PCR also showed good agreement between the two sample types (κ=0.75). These findings suggest that BL may be a less invasive alternative to BAL in selected critically ill patients, although larger studies are needed.

Broader biomarker use and machine learning-assisted multidimensional diagnosis and prognostic assessment

In addition to traditional biomarkers such as C-reactive protein (CRP) and procalcitonin (PCT), studies in 2025 have further expanded biomarker applications to the assessment of immune dysfunction, persistent inflammation, and prognosis. Wang et al. (9) characterised anti-interferon-gamma autoantibodies (AIGAs) as polyclonal antibodies, with immunoglobulin G subclass 3 (IgG3) as the predominant subtype, and demonstrated their neutralising activity. Affinity chromatography enabled absolute quantification of AIGAs and showed higher diagnostic accuracy than enzyme-linked immunosorbent assay testing. Piqueras et al. (10) reported that persistently elevated serum interleukin-6 (IL-6) levels measured 30 days after discharge following hospitalisation for CAP were independently associated with one-year mortality. In pulmonary fungal infections, Luo et al. (11) found that lower total iron-binding capacity was independently associated with pulmonary mucormycosis, whereas lower serum iron independently predicted mortality in patients with pulmonary mucormycosis.

Integrated multibiomarker models further improved prognostic assessment. Voza et al. (12) reported that pentraxin 3 (PTX3) has independent prognostic value in CAP. A machine learning model combining PTX3 and mid-regional pro-adrenomedullin measured at baseline and 72 hours with the oxygen saturation to fraction of inspired oxygen ratio measured at 72 hours achieved an AUC of 0.950. Wang et al. (13) developed an in-hospital mortality risk model for patients with acquired immunodeficiency syndrome (AIDS) and Pneumocystis pneumonia (PCP) based on serum albumin levels, arterial oxygen partial pressure, total bilirubin levels, lactate dehydrogenase levels, and cluster of differentiation 4-positive (CD4⁺) T-lymphocyte counts. The AUC for model discrimination was 0.994 in the training cohort and 0.947 in a temporally separated testing cohort from the same institution. However, because the model has not been evaluated in an independent external cohort, its applicability to other populations and healthcare settings remains uncertain.

Machine learning has also been applied to differential diagnosis and imaging analysis. In human immunodeficiency virus (HIV)-negative individuals, Qiu et al. (14) developed a logistic regression model to distinguish talaromycosis from pulmonary tuberculosis. The model achieved AUCs of 0.918 and 0.900 in the training and validation sets, respectively, and provided probability estimates for the two diseases. Li et al. (15) developed disease-specific three-dimensional U-Net models for computed tomography (CT) lesion segmentation. For pulmonary mucormycosis, the Dice coefficient was 93.38% in internal validation and 77.53% in external validation.

Innovation and optimised application of anti-infective agents

Antimicrobial resistance and the emergence of novel pathogens remain central challenges in the treatment of pulmonary infections. In 2025, key advances were made in the development of new agents targeting multidrug-resistant bacteria, drug-resistant fungi, and respiratory viruses. In parallel, the optimisation of treatment duration and dosing strategies, alongside the precise use of passive immunotherapy, further improved anti-infective treatment systems and supported the goals of precision dosing, resistance prevention, and enhanced therapeutic efficacy.

Antibacterial agents: breaking resistance barriers and promoting rational use

Clinical data for new agents targeting multidrug-resistant Gram-negative bacteria (MDR-GNB) continued to mature, providing new options for refractory infections. A phase 3 clinical trial published in The Lancet Infectious Diseases in 2025 (16) compared aztreonam-avibactam (ATM-AVI) with meropenem for HAP and VAP caused by Gram-negative bacteria. This study supported ATM-AVI as a potential treatment option for HAP and VAP caused by Gram-negative bacteria. Real-world evidence further provided comparative clinical data for newer established agents. One study (17) suggested that, compared with ceftazidime-avibactam, ceftolozane-tazobactam was associated with higher clinical success in invasive infections caused by multidrug-resistant Pseudomonas aeruginosa, with the difference mainly observed among patients with pneumonia.

The risk-stratified management of antimicrobial resistance has become increasingly refined and provides a stronger basis for rational drug use. A systematic review published in 2025 summarised the risk factors for drug-resistant pathogens in CAP (18), including tracheostomy, severe respiratory failure requiring early mechanical ventilation, prior antibiotic use, chronic lung disease, and wound care. A comprehensive review also summarised current and emerging treatment options for serious infections caused by carbapenem-resistant Pseudomonas aeruginosa and considered monotherapy with ceftazidime-avibactam, imipenem-cilastatin-relebactam, cefiderocol, or high-dose ceftolozane-tazobactam to be generally acceptable (19).

In addition to MDR-GNB, atypical bacterial pathogens remain important causes of CAP (20). Mycoplasma pneumoniae, Chlamydia pneumoniae (formerly Chlamydophila pneumoniae), and Legionella pneumophila differ from typical bacterial pathogens in terms of both diagnostic approaches and antimicrobial treatment. Routine clinical assessment and standard microbiological tests may not support the prompt identification of these pathogens, and beta-lactam monotherapy is ineffective against Mycoplasma pneumoniae because of its lack of a cell wall (20,21). In recent years, Mycoplasma pneumoniae infections have resurged in some regions, and increasing macrolide resistance has complicated empirical treatment decisions (21,22). In patients with suspected atypical bacterial pneumonia, particularly those with relevant epidemiological exposure, atypical clinical features, severe disease, or a poor response to empirical beta-lactam therapy, early pathogen identification using nucleic acid testing or other targeted microbiological methods, including urinary antigen testing for Legionella, may be considered (20,23). Early recognition of these pathogens may facilitate more targeted antimicrobial selection and the optimisation of antibiotic use (23).

Antifungal agents: new mechanisms and responses to resistance

The global spread of triazole-resistant Aspergillus fumigatus (TRAF) has challenged conventional antifungal strategies, and the development of new antifungal agents is beginning to reshape the therapeutic landscape. Epidemiological data from Asia indicate that TRAF occurs in both clinical and environmental isolates, with pooled proportions of 4% and 14%, respectively, posing a challenge to triazole-based treatment (24). A meta-analysis from Iran reported pooled resistance rates of 17.61% to itraconazole and 14.68% to voriconazole in A. fumigatus, with the TR34/L98H resistance mutation commonly identified among resistant isolates, supporting the need for regional antifungal resistance surveillance (25). Moreover, a systematic review and individual patient data meta-analysis showed substantial mortality in chronic pulmonary aspergillosis (CPA). Overall mortality was 35% in cohorts predominantly involving chronic obstructive pulmonary disease and 25% in those predominantly involving pulmonary tuberculosis (26).

The 2025 Chinese guidelines for invasive pulmonary fungal disease provide recommendations tailored to the clinical context of China, including the management of resistant and refractory infections. For patients with severe, refractory, or treatment-failure IPA, these guidelines recommend combination therapy with triazoles plus echinocandins and/or polyenes, or echinocandins plus polyenes (27). Similarly, the 2025 American Thoracic Society clinical practice guideline addresses initial antifungal strategies for proven or probable IPA and conditionally recommends either mould-active triazole monotherapy or combination therapy with a mould-active triazole plus an echinocandin on the basis of low-quality evidence (28). Taken together, these recommendations support the use of individualized antifungal strategies for severe or difficult-to-treat IPA, with treatment selection guided by disease severity, local epidemiology, antifungal availability, and resistance considerations.

Maertens et al. (29) reported phase 2b clinical data for olorofim, the first orotomide antifungal developed for invasive fungal infections. Olorofim selectively inhibits fungal dihydroorotate dehydrogenase (DHODH), thereby blocking de novo pyrimidine biosynthesis and leading to fungal cell death. Olorofim exhibited activity against difficult-to-treat fungi, including azole-resistant Aspergillus spp. and Scedosporium spp., and was generally well tolerated in patients with few or no therapeutic options. Because this was a single-arm, open-label study, its efficacy could not be directly compared with that of existing salvage therapies (29). This highlights the expansion of antifungal drug development beyond the traditional polyene, azole, and echinocandin classes and supports further evaluation of olorofim for drug-resistant fungal infections.

Antiviral agents: new drug targets and refined dosing strategies

The development of new nucleoside analogues has provided additional therapeutic options for coronavirus disease 2019 (COVID-19) treatment. Obeldesivir is an oral nucleoside analogue prodrug that inhibits SARS-CoV-2 replication by targeting viral ribonucleic acid (RNA)-dependent RNA polymerase (30). In a phase 3 randomised, double-blind, placebo-controlled trial involving low-risk nonhospitalised patients, Ogbuagu et al. (30) reported that compared with a placebo, obeldesivir was generally safe and well tolerated and reduced SARS-CoV-2 viral RNA copy numbers, although it did not significantly shorten the time to symptom alleviation. Its oral administration and favourable safety profile suggest its potential utility in future outpatient antiviral strategies, although additional studies are needed to clarify its clinical benefits in specific populations. The effect of treatment duration on viral rebound was also evaluated. In a randomised phase 2 trial among immunocompromised patients, Weinstein et al. (31) reported no difference among the 5-, 10-, and 15-day regimens in the primary endpoint. Viral rebound occurred in 17.3%, 2.1%, and 2.0% of participants, respectively. Extending treatment beyond 5 days was associated with a nominal reduction in viral rebound, although its clinical benefit remains uncertain.

With respect to influenza therapy, important advances have been made in the use of several new antiviral agents. The 2025 Chinese influenza guidelines provide evidence-based recommendations for the treatment and prophylaxis of influenza in the Chinese clinical setting. Oseltamivir is recommended as an important antiviral option for severe influenza, along with early initiation of antiviral therapy for high-risk patients (preferably within 48 hours of symptom onset), avoidance of routine corticosteroid use, and the use of antiviral prophylaxis as a supplementary strategy for selected individuals when vaccination alone is insufficient or not feasible (32). These recommendations are generally in agreement with the 2024 World Health Organization clinical practice guidelines for influenza, which conditionally recommend oseltamivir for patients with severe influenza, advise against routine corticosteroid use, and support antiviral prophylaxis for selected asymptomatic exposed individuals at extremely high risk of hospitalisation (33). Together, these guidelines emphasize early antiviral treatment and individualized prevention strategies according to disease severity, host risk factors, and exposure status.

Beyond the recommendations in these guidelines, Yang et al. (34) evaluated onradivir (ZSP1273), a potent inhibitor of the polymerase basic protein 2 (PB2) subunit of the influenza A virus polymerase, in a phase 3 clinical trial. A 600 mg once-daily oral regimen for adults with acute uncomplicated influenza A exhibited clinical efficacy comparable to that of oseltamivir. Compared with a placebo, onradivir shortened the time to symptom alleviation by 24.52 hours. It also shortened the duration of detectable viral activity compared with oseltamivir (34). Suraxavir marboxil (GP681) is an oral antiviral agent that targets the polymerase acidic (PA) subunit of the influenza virus RNA polymerase. In a phase 3 randomised trial involving otherwise healthy outpatients aged 5–65 years with uncomplicated influenza, Wang et al. (35) reported that compared with a placebo, a single 40 mg oral dose significantly shortened the median time to symptom alleviation (42.0 vs. 63.0 hours), accelerated viral load decline, and limited the emergence of resistance-associated variants. In phase 2 and phase 3 randomised, double-blind, multicentre, placebo-controlled trials, Zhao et al. (36) evaluated ADC189, a novel small-molecule anti-influenza agent whose active metabolites bind divalent metal ions within the PA subunit, thereby inhibiting cap-dependent endonuclease activity and viral RNA transcription. A single oral dose of ADC189 was associated with clinical efficacy and reduced viral loads and exhibited a favourable safety profile in adults and adolescents with uncomplicated influenza. Compared with a placebo, ADC189 shortened the median time to symptom alleviation (50.0 vs. 68.1 hours), accelerated viral RNA decline, and reduced the duration of detectable viral RNA, with most adverse events being mild to moderate (36).

Innovation in drug delivery has expanded antiviral treatment strategies. Khaitov et al. (37) evaluated MIR19, an inhaled small interfering RNA (siRNA)-based antiviral therapy targeting a highly conserved region of the SARS-CoV-2 RNA-dependent RNA polymerase (RdRp) gene through RNA interference. In a multicentre, randomised, controlled phase 2b–3 trial, compared with standard therapy, MIR19 reduced the proportion of patients with mild COVID-19 who progressed to moderate or severe disease (2.85% versus 6.91%) and was associated with fewer adverse events (15.65% vs. 20.33%) (37). This RNA interference-based strategy represents an alternative antiviral approach with a mechanism distinct from that of conventional direct-acting antiviral agents.

Optimising passive immunity and antibody-based therapy for COVID-19

The optimal timing of neutralising antibody therapy relative to vaccination has become clearer. Isa et al. (38) reported that casirivimab and imdevimab administered before or at the time of mRNA-1273 vaccination reduced vaccine-induced neutralising antibody titres, whereas minimal reduction was observed when vaccination preceded monoclonal antibody administration. The clinical significance of this reduction remains unclear. The clinical value of high-titre convalescent plasma was also redefined. Hoffmann et al. (39) reported that the infusion of high-titre convalescent plasma early in the course of COVID-19 may bridge the temporal gap before endogenous antibody production, helping prevent hospitalisation or death in immunocompromised patients with mild COVID-19 and offering a complementary intervention for immunodeficient populations.

New dimensions in host immune regulation and microecological intervention

An important trend in pulmonary infection research in 2025 was the shift from the focus on pathogens alone to an emphasis on host-pathogen interactions. Deeper investigations of gut-lung axis immune regulation, mechanisms of host immune dysregulation, and the optimized precision use of adjunctive immunotherapy have provided new targets for comprehensive pulmonary infection care. These advances have promoted a transition from single anti-infective therapy to integrated strategies that combine pathogen-directed treatment with immune balance regulation.

The gut-lung axis and immune regulation in pulmonary infection

Accumulating evidence suggests that the gut microbiota may influence pulmonary infection outcomes through microbiota-derived metabolites. Using a translational design that integrated a large general-population cohort, a cohort of critically ill patients with severe CAP, a randomised human intervention trial of antibiotic-mediated microbiota modulation, and mouse models, Kullberg et al. (40) demonstrated that gut microbiota-derived tryptophan metabolites, particularly indole-3-acetic acid (IAA), were associated with pneumonia risk and mortality. Experimental studies further suggested that IAA may aggravate pulmonary injury through aryl hydrocarbon receptor signalling and increased reactive oxygen species release from neutrophils (40). However, the clinical implications of targeting gut microbiota-derived metabolites in pulmonary infection remain uncertain.

Precision use of immunomodulators and ongoing controversies

The controversy surrounding the use of glucocorticoids in the treatment of pulmonary infections has become more clearly defined, with population heterogeneity emerging as a key determinant of benefit. In a pragmatic randomised trial conducted in a low-resource setting, Lucinde et al. (41) reported that adjunctive low-dose glucocorticoid therapy reduced 30-day mortality among adults hospitalised with CAP. These findings suggest that the clinical value of steroids varies across patient backgrounds and that individualized use is necessary. A meta-analysis by Pitre et al. (42) revealed that corticosteroids probably reduced short-term mortality and reduced the need for invasive mechanical ventilation among adults hospitalised with non-viral CAP. The role of adjunctive corticosteroids in PCP was also evaluated. In HIV-negative patients with severe PCP, Lemiale et al. (43) reported no significant reduction in 28-day mortality. However, 90-day survival was higher in the corticosteroid group than in the placebo group (72.0% versus 56.8%), and among patients who were not intubated at randomization, subsequent intubation was less frequent (10.1% versus 26.1%). These secondary outcomes suggest a possible benefit, although the primary endpoint was not met.

In patients with severe COVID-19, immunomodulator use requires a careful balance between efficacy and safety. Amstutz et al. (44) found that Janus kinase (JAK) inhibitors reduced 28-day mortality and the need for new mechanical ventilation or other respiratory support among adults hospitalised with COVID-19. JAK inhibitors were also associated with fewer severe or serious adverse events than no JAK inhibitor treatment. A subsequent sub-study investigated a post hoc safety signal among vaccinated patients treated with baricitinib but did not identify a statistically significant biomarker-based explanation for an interaction between vaccination status and serious adverse events. The findings require cautious interpretation because of the moderate sample size (45). A randomised clinical trial (46) reported that colchicine did not improve functional capacity, respiratory function, or inflammatory markers in adults with long COVID. These findings do not support colchicine for this indication.

Refinement of prognostic assessment and clinical strategies

Clinical outcomes in patients with pulmonary infections vary substantially among individuals. Even among patients infected with the same pathogen, differences in immune status, underlying diseases, and host factors may contribute to heterogeneous clinical trajectories. Identifying factors associated with poor outcomes and tailoring management according to patient characteristics are important for improving risk assessment and clinical decision-making.

Prognostic risk factors and risk stratification in pulmonary infection

With respect to aspiration-related CAP, Almirall et al. (47) reported that aspiration-related factors and frailty contribute substantially to outcome variability among older adults. Dysphagia, impaired consciousness, and reduced airway protective mechanisms are important contributors to aspiration risk and may partly explain the poor outcomes observed in this population. With respect to fungal infection, risk assessment in nonneutropenic patients has become an increasingly important clinical issue. Recent studies indicate that neutropenia alone is insufficient for identifying patients at risk for IPA, as invasive aspergillosis may also occur in individuals with other forms of immune impairment or immunosuppressive conditions (48). Patients receiving Bruton tyrosine kinase inhibitors, particularly those with lymphoid malignancies, may develop impaired antifungal immune surveillance despite the absence of profound marrow suppression, increasing their susceptibility to opportunistic fungal infections, including aspergillosis and PCP (49). Patients with diabetes admitted to the ICU also represent a population with an increased risk of invasive aspergillosis (50). These findings highlight the need for individualized risk assessment rather than reliance on traditional single-factor models.

The clinical impact of respiratory viral coinfection has also been recognized. Recent evidence indicates that the clinical consequences of viral coinfection may vary according to the specific pathogen combination and clinical setting. For example, SARS-CoV-2 coinfection with different respiratory viruses may have distinct associations with disease severity (51).

Stratified management of special populations

The management of pulmonary infections in special populations should extend beyond pathogen-directed therapy and incorporate assessments of host vulnerability, immune status, and modifiable risk factors. In older or frail adults, clinical management should include the early assessment of aspiration-related risk factors, including swallowing dysfunction, impaired consciousness, and functional decline. When aspiration risk is identified, preventive interventions and supportive measures should be considered to reduce complications (47). In immunocompromised hosts, clinical management should account for the type and severity of immune impairment, underlying diseases, and immunosuppressive therapies. Early and targeted microbiological evaluation, including antigen detection and molecular testing when appropriate, may facilitate timely diagnosis and treatment (52).

In patients with haematologic malignancies, particularly those receiving Bruton tyrosine kinase inhibitors, clinicians should maintain awareness of opportunistic fungal infections despite the absence of severe neutropenia. The early evaluation of new pulmonary abnormalities or persistent respiratory symptoms should be considered (49). Proven or probable IPA was more frequent in patients with Philadelphia chromosome-negative acute lymphoblastic leukemia than in those with acute myeloid leukemia. Prolonged neutropenia and the absence of antimold prophylaxis were associated with infection, supporting consideration of antimold prophylaxis in this population (53). For critically ill patients with diabetes, clinicians should maintain a high index of suspicion for invasive aspergillosis, particularly in patients with COVID-19, in whom the association between diabetes and invasive aspergillosis was strongest (50). When invasive aspergillosis is suspected, appropriate fungal diagnostic testing should be considered (52).

Strengths and limitations

This review focuses on recent developments in pulmonary infection research published in 2025. The use of multiple databases combined with manual supplementation and independent screening by two reviewers helped improve the comprehensiveness and consistency of the literature selection process. The review covers key clinical dimensions of pulmonary infection management, including multipathogen assessment, precision diagnosis, anti-infective drug development, host immune regulation, and prognostic risk stratification. It also addresses antimicrobial resistance, difficult-to-treat infections, and the management of special populations.

Several limitations should be acknowledged. First, although both English-language and Chinese-language publications were included, the evidence was derived from studies conducted across different geographic and healthcare settings. These contextual differences may limit the applicability of the findings to local clinical practice. Second, as a narrative review, a formal heterogeneity assessment or quantitative meta-analysis was not performed; therefore, pooled effect size estimates and formally graded recommendations were not generated, which limits the overall strength of the evidence. Third, the review focuses mainly on clinically oriented evidence, with relatively limited discussion of basic mechanistic studies and early-stage technologies. Fourth, evidence on coinfections and patients with complex comorbidities remains fragmented. Fifth, some novel diagnostic technologies and anti-infective regimens still lack long-term follow-up data, and their long-term efficacy and safety require validation. Finally, current risk stratification and management strategies are largely qualitative and lack widely accepted quantitative thresholds, limiting their immediate clinical implementation.

Conclusions

Research on pulmonary infections in 2025 reflected a continued transition from empirical approaches towards more precision-oriented, individualized, and host-directed management. At the diagnostic level, advances in molecular technologies such as mNGS, optimisation of sampling methods, and integration with novel biomarkers and AI-assisted models have exhibited increasing value in pathogen identification, differential diagnosis, and prognostic assessment, particularly in severe or diagnostically challenging infections. At the therapeutic level, important progress has been made in the development of new agents targeting multidrug-resistant bacteria, drug-resistant fungi, and respiratory viruses. More individualized selection and duration of anti-infective therapy may help balance treatment efficacy with antimicrobial stewardship. Studies of the gut-lung axis and immunomodulatory therapy have also highlighted the potential importance of host factors, although clinical evidence supporting microbiota-targeted interventions remains limited. At the prognostic management level, the improved identification of infection-related risk factors and more specific management strategies for high-risk populations may support earlier assessment and better-informed clinical decision-making.

Overall, studies published in 2025 reported important advances in the diagnosis, treatment, and integrated management of pulmonary infections. Nevertheless, this narrative review did not include a quantitative synthesis of the available evidence and therefore did not provide pooled effect size estimates or formally graded clinical recommendations. The interpretation and clinical application of these findings should also account for differences in study design, population characteristics, geographic setting, and evidence quality. Further high-quality clinical studies and independent multicentre validation are needed to confirm the generalisability of these findings and support the development of more consistent diagnostic and treatment strategies.

Supplementary

The article’s supplementary files as

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.