Spotlight commentary: New Frontiers in pneumonia treatment

Abstract

Community-acquired pneumonia (CAP) is a leading cause of hospitalization and mortality worldwide, necessitating continuous advancements in its management. As antimicrobial resistance rises and treatment complexities grow, researchers are exploring innovative therapeutic strategies to enhance patient outcomes. From evaluating the efficacy of adjunct therapies to investigating novel antibiotics and optimizing dosing regimens, recent studies have shed light on critical aspects of CAP treatment. The aim of this commentary is to highlight key emerging studies on adjunctive therapies, new antimicrobial agents, innovative dosing strategies and machine learning tools in order to demonstrate the expanding potential for a more individualized approach to the treatment of CAP. These developments collectively challenge the longstanding one-size-fits-all paradigm and invite a redefinition of how pharmacological care is delivered in this common yet heterogeneous condition.

For patients hospitalized with CAP without severe beta-lactam allergy and lacking risk factors for MRSA or Pseudomonas spp., empiric first line therapy typically consists of either a combination of an antipneumococcal beta-lactam (such as ceftriaxone, cefotaxime or ampicillin-sulbactam) with a macrolide or respiratory fluoroquinolone monotherapy. The 2023 ERS/ESICM/ESCMID/ALAT guidelines recommend β-lactam–macrolide therapy for severe CAP in ICU patients, while the 2019 ATS/IDSA guidelines support this combination in non-severe CAP when atypical pathogens are suspected.^{1, 2} However, much of this support is derived from observational studies and expert consensus rather than robust randomized evidence. The ACCESS trial,³ conducted between 2021 and 2023 across 18 Greek hospitals, provides timely randomized data. It enrolled 278 adults with more severe CAP, defined by a SOFA score ≥2 and procalcitonin (PCT) ≥0.25 ng/mL, who were randomized to receive either clarithromycin or placebo in addition to standard β-lactam therapy. The primary composite endpoint, defined as improvement in respiratory symptoms plus either a ≥30% reduction in SOFA score or favourable PCT kinetics, was achieved in 68% of patients in the clarithromycin group compared to 38% in the placebo group (p < 0.0001). A subgroup analysis suggested that patients with elevated inflammatory markers such as CRP and PCT derived the greatest benefit, supporting a biomarker-guided approach to macrolide use.⁴ In contrast, a recent meta-analysis by Prizão et al.,⁵ which included over 2600 patients across six randomized trials, found no significant mortality benefit from β-lactam and macrolide combination therapy at discharge, 30 or 90 days. Although treatment success was nominally higher in some cases, substantial heterogeneity limited the interpretability of pooled results and may have obscured benefits in select subgroups. Collectively, these studies underscore a shift away from routine empirical macrolide use towards more targeted therapy based on inflammatory profiles. While the ACCESS trial has limitations, including a modest sample size and single-country setting, its design aligns well with the principles of personalized medicine and antimicrobial stewardship. However, such approaches require validation in multicentre and pragmatic contexts before informing widespread changes in clinical practice.

In addition to reconsidering established drug combinations, clinicians must account for the persistent rise in antimicrobial resistance, which poses a significant threat to the effectiveness of existing therapies. This includes the growing resistance to fluoroquinolones, commonly employed in CAP treatment. In response, the development of novel agents has gained momentum. Nemonoxacin, a non-fluorinated quinolone, has emerged as a promising alternative, having completed several clinical trials and received Qualified Infectious Disease Product (QIDP) and fast-track designations from the US Food and Drug Administration. In mainland China, oral nemonoxacin has been approved since 2016 for adults with mild-to-moderate CAP. In a large Phase 3 double-blind randomized trial by Li et al.⁶ involving 525 hospitalized adults, intravenous nemonoxacin (500 mg once daily) demonstrated a clinical cure rate of 91.8% compared to 85.7% in the levofloxacin group. Although nemonoxacin led to a slightly higher incidence of adverse events, most were mild and resolved after discontinuation. To complement these clinical findings, the role of pharmacokinetics and pharmacodynamics (PK/PD) in dosing optimisation was further addressed by Li et al.⁷ in a follow-up study published in the British Journal of Clinical Pharmacology (BJCP), which assessed dose adjustments in patients with severe renal impairment using a population PK/PD approach. This modelling study demonstrated that a 0.5-g dose every 48 h maintained a ≥90% probability of target attainment at MIC ≤1 mg/L. These studies exemplify how combining clinical efficacy data with PK/PD modelling and individual patient factors such as renal function can enable more personalized antimicrobial therapy in CAP, with BJCP playing a central role in promoting the translation of pharmacokinetic modelling into tailored dosing strategies, particularly in settings where traditional trial infrastructure may be limited.

BJCP has also contributed to the expanding evidence base supporting tailored antimicrobial strategies in CAP by highlighting research examining how patient-specific and physiological factors influence drug concentrations at the site of infection. For extracellular pathogens such as Streptococcus pneumoniae and Haemophilus influenzae, drug concentration in the pulmonary epithelial lining fluid (ELF), the principal site of pathogen interaction in the lungs, is a key determinant of therapeutic success. Dong et al.⁸ measured ELF ceftriaxone levels in 22 children with CAP and found that ELF concentrations were, on average, more than 12 times higher than plasma levels. These findings suggest that therapeutic concentrations in the lung may be achieved even when systemic levels appear subtherapeutic. ELF penetration is particularly important in critically ill patients, where inflammation and large volume shifts lead to variable ELF drug concentrations. Palmer et al.⁹ compared nebulised to intravenous vancomycin in 10 mechanically ventilated patients with suspected MRSA pneumonia, reporting that nebulised vancomycin produced sputum concentrations ~20-fold higher than intravenous administration. Interestingly, only patients with severe hypoalbuminaemia attained adequate sputum levels via intravenous dosing alone, suggesting that albumin status may influence alveolar drug penetration. While the certainty of evidence is limited by small sample size and mechanistic study design, these findings demonstrate that even established antibiotics may benefit from revised administration routes or dosing regimens when guided by pathophysiological markers or compartmental PK data. A structured overview of the studies discussed in this section and those that follow, including their design, population, principal outcomes and level of evidence, is provided in Table 1.

The trajectory towards more individualized pharmacotherapy in CAP is further advanced by computational approaches, with a study by Tang et al.¹⁰ serving as a prominent example. Machine learning techniques were employed to refine azithromycin dosing in paediatric CAP using simulated pharmacokinetic profiles derived from a validated population model. The authors developed predictive tools capable of estimating individual drug exposure with high accuracy, achieving R² values exceeding 0.98 for both a priori and a posteriori models. This framework allows routinely collected clinical data to inform precise dosing recommendations, moving beyond standard weight-based regimens and enabling more consistent therapeutic exposure. The study also explored the integration of this framework into electronic health record systems, indicating the potential for real-time clinical decision support aligned with personalized care. By publishing this work, BJCP has spotlighted one of the most forward-looking applications of artificial intelligence in infectious disease pharmacotherapy and has reinforced its role as a platform for innovative and clinically relevant research. Successful implementation in routine care, however, will require interoperability with healthcare IT systems and close alignment with regulatory frameworks.

Beyond antimicrobials, corticosteroid therapy in CAP has become a prime example of data-driven, criteria-based treatment. The US-based Society of Critical Care Medicine (SCCM) recently updated its guidance based on 18 randomized trials, including the pivotal CAPE COD trial by Dequin et al., and now recommends corticosteroids for adults with severe bacterial CAP requiring intensive care under a broader range of indications,¹¹ in contrast to the more restrictive recommendations of the ERS, ESCMID and ATS/IDSA guidelines.^{1, 2} The landmark CAPE COD trial¹² showed that among ICU-admitted patients with severe CAP, early adjunctive hydrocortisone significantly improved survival (28-day mortality 6.2% vs. 11.9% with placebo). This nearly 50% relative reduction in mortality underscores that steroids can be lifesaving when used in the right patient subset, reinforcing the importance of severity-based selection rather than blanket use. Likewise, Cangemi et al.¹³ found that corticosteroid use attenuated inflammation-mediated cardiac injury in CAP: Patients with elevated troponin who received steroids had less troponin rise and fewer inhospital major cardiac events than those not treated, an effect not seen in patients without initial troponin elevation. Such data suggest that integrating biomarkers (e.g., high-sensitivity troponin) and validated risk criteria into clinical decisions can pinpoint patients most likely to benefit from anti-inflammatory adjuncts. Collectively, these findings highlight the need to move beyond generalized recommendations and adopt more nuanced, criteria-based patient selection. Identifying patient subgroups most likely to benefit from corticosteroids will be crucial for evidence-based integration into clinical practice.

The future of CAP management calls for more than incremental therapeutic refinements. It requires a fundamental reconceptualization of pharmacological care. From biomarker-guided antibiotics and novel agents to tailored administration methods and predictive modelling, recent research shows that individualized therapy in CAP is no longer aspirational, it is rapidly becoming essential. In parallel, BJCP has served as a platform for transformative work that advances patient-centred approaches. The next step is the translation of these insights into clinical reality. Achieving this will depend on coordinated efforts to embed validated tools into electronic systems, support clinician education and harmonize innovation with regulatory and practical frameworks. Only through such alignment can the treatment of CAP consistently deliver the right therapy to the right patient at the right time.

All coauthors took part in writing of the manuscript.

None to disclose.