Monotherapy versus combination therapy in severe Legionella pneumonia: a multicenter retrospective observational study

Abstract

Introduction

The incidence of Legionella pneumophila pneumonia is rising. Since 2012, French guidelines recommend fluoroquinolone monotherapy or combination therapy with a fluoroquinolone and macrolide for severe cases, without favoring a strategy. This study assessed the impact of antibiotic strategies on mortality in severe Legionella pneumophila pneumonia in ICU patients.

Materials and methods

We conducted a retrospective, observational multicenter cohort study across ICUs at the university hospitals of Rouen, Caen, Amiens, Lille, and Elbeuf General Hospital, covering the period from January 1, 2017, to December 31, 2022. All adult patients admitted to ICUs for Legionella pneumophila were included. Patients were divided into two groups: monotherapy and combination therapy. The combination therapy group included those who received exclusive combination therapy and those who transitioned from monotherapy to combination therapy (non-exclusive combination therapy). The primary endpoint was 28-day mortality after ICU admission. Secondary outcomes included 90-day mortality, ventilation duration, ICU stay length, and antibiotic-related adverse events.

Results

We included 93 patients: 13 in the monotherapy group and 80 in the combination therapy group, with 24 (30%) receiving non-exclusive and 56 (70%) exclusive combination therapy. In the monotherapy group, 12 patients (92.3%) received fluoroquinolones and 1 (7.7%) received a macrolide. Combination therapy consisted of fluoroquinolones and macrolides in all but one case, treated with fluoroquinolones and rifampicin. High-flow oxygen was required in 59 patients (69.4%), invasive ventilation in 65 patients (69.9%), and acute respiratory distress syndrome occurred in 64 patients (68.8%). The 28-day mortality was not significantly different between the two groups (15.4% vs. 12.5). No significant differences were observed in invasive ventilation duration (16.5 vs. 10 days), ICU stay length (12 vs. 11.5 days), or 90-day mortality (23.1% vs. 16.25%). Adverse events attributed to antibiotherapy were cutaneous rash (n = 7; 87.5%) and confusion (n = 1 ;12.5%) and were not significantly different between the two groups (0% vs. 10%).

Conclusion

Our study suggests that monotherapy, particularly with fluoroquinolones, may be an alternative to combination therapy in managing severe Legionella pneumophila pneumonia in the ICU. Prospective and randomized studies would be necessary to confirm these results and refine treatment recommendations.

Introduction

The global incidence of Legionella pneumophila pneumonia has recently risen, driven by increased risk factors – particularly immunosuppression – as well as advances in diagnostic methods [1–4]. Legionella pneumophila pneumonia often involves severe lung injury requiring respiratory and/or hemodynamic support, along with extrapulmonary manifestations [5–7]. It is currently among the three leading causes of community-acquired pneumonia requiring intensive care unit (ICU) admission. Mortality among ICU immunocompromised patients can reach 40% despite appropriate antimicrobial therapy [8, 9].

Early diagnosis and prompt antibiotic initiation are critical for disease management and improve outcomes [9–11]. Effective antibiotic therapy against Legionella spp depends on the ability of the antibiotic to concentrate in alveolar macrophages. The main antibiotic classes include fluoroquinolones, macrolides, tetracyclines, and rifampicin. In vitro data suggest that fluoroquinolones – particularly levofloxacin – are superior [12, 13].

Current therapy relies on macrolides or fluoroquinolones, or their combination, with prolonged courses required in severe cases [14, 15]. Meta-analyses have shown no significant differences in clinical response or mortality between fluoroquinolones and macrolides [15, 16]. Guidelines consistently recommend either a fluoroquinolone or a macrolide as first-line treatment [17–19]. The role of combination therapy remains controversial, as international guidelines vary by pneumonia severity and offer no clear consensus for critically ill patients [17, 19]. In France, the 2012 guidelines recommended fluoroquinolone monotherapy (preferably levofloxacin) or combination therapy with a fluoroquinolone and a macrolide for severe cases, though the benefit of combination therapy in severe pneumonia remains controversial [20].

In this context, we conducted a retrospective multicentre study to evaluate the impact of monotherapy or combination therapy strategies on mortality in ICU patients with severe pneumonia due to Legionella spp.

Methods

We conducted a multicenter retrospective observational study across five hospitals in Northwestern France between January 1, 2017, and December 31, 2022. We included all adult patients (aged 18 years and older) admitted to intensive care units for severe Legionella pneumophila pneumonia. We considered as severe Legionella pneumophila pneumonia requiring ventilatory support (excluding conventional oxygen therapy delivered via nasal cannula or medium- to high-concentration face masks), hemodynamic support with catecholamines, or renal support with renal replacement therapy. Eligible patients were screened in each center by the local investigator using data extracted from the French hospital discharge database (PMSI). Screening was based on the diagnostic code A481 (Legionnaires’ disease – Legionellosis). Data were collected in the case report form (CRF) by the local investigator using information extracted from the patients’ medical records.

We collected the following data in medical charts: demographic information (age, sex, medical history), patient characteristics at admission (SAPS II score [21] and laboratory test results), diagnostic method for Legionella pneumophila (urinary antigen testing, PCR, or culture of respiratory samples), the antibiotic regimen specifically used to treat Legionella pneumophila infection (the monotherapy group included patients who received a single antibiotic at any time while the combination therapy group included patients who received the concurrent use of at least two different antibiotics at any time, including patients who switched from or to monotherapy during treatment) and organ support during intensive care stay (vasopressors, mechanical ventilation, VV-ECMO, renal replacement therapy). Patients who developed acute respiratory distress syndrome (ARDS) were classified according to the Berlin criteria [22]. We also collected the patient vital status at 28 and 90 days from two sources: the center’s medical records and the national Death Register, maintained by the French National Institute for statistics and Economic Studies (INSEE – https://deces.matchid.io/search).

The primary outcome was 28-day all-cause mortality following ICU admission. Secondary outcomes included ICU length of stay, duration of invasive mechanical ventilation, 90-day mortality and the incidence of adverse drug reaction to antibiotics used to treat the Legionella pneumophila infection during hospitalization.

To address the objective of our study, we compared the monotherapy and combination therapy groups. We expressed categorical variables as counts and percentages and compared using the Chi-square test or Fisher’s exact test, as appropriate. We expressed continuous variables as mean ± standard deviation (SD) or as median [interquartile range], depending on their distribution, and compared them using Student’s t-test or one-way ANOVA, respectively. We analyzed survival using Kaplan-Meier curves and compared with the log-rank test.

As part of the retrospective setting, a post-hoc power analysis was performed from the collected 28-mortality data. A sensitivity analysis was performed to assess the impact of modified therapy on outcomes. To enhance the comparability of the groups and so generalize our drawing, we conceive a propensity score in which dependent variable was the antibiotics regimen (monotherapy versus combination therapy) and independent variables the patient’s characteristics. The conception of this score included a two-step procedure, in which variables with a p value < 0.15 were selected and included in the second step. Only variables with a p value < 0.10 in the second step were retained for the score.

To determinate risk factors of 28-day mortality, a logistic regression was performed. The dependent variable was mortality and independent variables were demographic data and ICU management. Variables with a p value < 0.15 were selected and included in the multivariable analysis. Only variables with a p value < 0.10 in the multivariable analysis were retained as risk factors of mortality and presented with odd ratios and their respective 95% confident interval and p value.

Statistical significance was set at p < 0.05. Analyses were performed using SPSS version 29 (IBM Corp. Released 2023. IBM SPSS Statistics for Macintosh, Version 29.0.2.0 Armonk, NY: IBM Corp).

This study was approved by the Research Project Qualification Committee of the Rouen University Hospital, under registration number 1058.

Results

Over the six-year study period, 95 patients with severe Legionnaire’s disease were admitted to the five participating ICUs. We excluded two patients, one refused participation, and one was a false positive (positive urinary antigen test but negative PCR on respiratory sample). Finally, we included 93 in the analysis (Fig. 1). Legionella pneumophila pneumonia was diagnosed by urinary antigen testing in 86% of cases, respiratory PCR in 43%, and culture in 16%. Thirteen patients (14%) received monotherapy and 80 (86%) received a combination therapy.

Fig. 1.

Flow chart

Post hoc power analysis

By using an ɑ risk of 0.05 with a b risk of 0.20 (i.e. power of 0.8), with a 28-day mortality of 15.4% (monotherapy group) and 12.5% (combined therapy group), the calculated power was 5.8%.

Baseline characteristics and medical history of the study population

Baseline characteristics are shown in Table 1. Patients were predominantly male (male/female sex ratio 2.4), with a mean age of 60.5 years (SD +/- 11,5 years). Over half were chronic smokers (55%) and nearly a quarter reported regular alcohol consumption (24%) and/or were immunosuppressed (24%). An attempt for a propensity score calculation was done to balance the groups but no characteristic was significantly different between groups (Table 2).

Table 1.

Baseline characteristics of the study population

Characteristics	Missing data n = 93	Study population n = 93	Monotherapy n = 13		Combination therapy n = 80	Non exclusive combination therapy n = 25
Sex, n (%)
Male	0	66 (71)	7 (69.2)	57 (71.25)		17 (68)
Female	0	27 (29)	4 (30.8)	23 (28.75)		8 (32)
Age, mean +/- SD	0	60 ± 11.5	61.5 ± 10.7	54 ± 13.8		62.5 ± 11.1
Smoking history > 10 pack-years, n (%)	0	51 (55)	6 (46.1)	45 (56.3)		14 (56)
Alcohol consumption > 2 units/day, n (%)	0	21 (22.5)	2 (11.4)	19 (23.7)		7 (28)
Heart failure, n (%)	0	8 (9)	2 (15.4)	6 (7.5)		1 (4)
Chronic respiratory failure on long-term oxygen, n (%)	0	0 (0)	0 (0)	0 (0)		0 (0)
Chronic obstructive pulmonary disease (COPD), n (%)	0	6 (6.5)	1 (7.7)	5 (6.3)		1 (4)
Cirrhosis CHILD B or C, n (%)	0	0 (0)	0 (0)	0 (0)		0 (0)
Chronic kidney disease, n (%)	0	7 (7.5)	0 (0)	7 (8.75)		2 (8)
Diabetes mellitus, n (%)	0	15 (16)	2 (15.4)	13 (16.25)		5 (20)
Hematologic malignancy, n (%)	0	13 (14)	0 (0)	13 (16.2)		0 (0)
Immunosuppression, n (%)	0	22 (24)	5 (38.5)	17 (21.3)		13 (52)
IGSII score, median [IQR]	2/93	43 [34–54]	41 [19–50]	44.5 [35–57.5]		27 [11–40]

Table 2.

Conception of a propensity-score

Variables	Univariable analysis			Multivariable analysis
	Odd Ratio (95% confident interval)	P value	Odd Ratio (95% confident interval)		P value
Age	1.06 (0.90–1.12)	0.360	-		-
Male sex	0.91 (0.25–3.24)	0.882	-		-
Smoking history > 10 pack-years	1.50 (0.46–4.86)	0.490	-		-
Alcohol consumption > 2 units/day	1.71 (0.35–8.42)	0.512	-		-
Heart failure	0.44 (0.08–2.49)	0.368	-		-
Chronic obstructive pulmonary disease (COPD)	0.80 (0.09–7.46)	0.843	-		-
Chronic kidney disease	2.65 (0.83–7.63)	0.981	-		-
Diabetes mellitus	1.07 (0.21–5.39)	0.948	-		-
Hematologic malignancy	3.00 (0.16–7.57)	0.971	-		-
Immunosuppression	0.43 (0.13–1.49)	0.285	-		-

Characteristics of the study population at ICU admission

At ICU admission (Table 3), over 80% of patients required respiratory support, most commonly high-flow nasal oxygen (41%). There were no significant differences between treatment groups regarding the proportion of different respiratory support and laboratory tests.

Table 3.

Ventilatory status at admission, laboratory data and diagnostic methods

Characteristics	Missing data n = 93	Study population n = 93	Monotherapy n = 13	Combination therapy n = 80	p	Non exclusive combination therapy n = 25
Ventilation mode at ICU admission, n (%)	0/93				0.63
Spontaneous ventilation		18 (19.3)	5 (38.4)	13 (16.25)	0.08	2 (8)
High-flow nasal oxygen		38 (40.9)	4 (30.8)	34 (42.5)	0.62	14 (56)
Non-invasive ventilation		5 (5.4)	0 (0)	5 (6.25)	0.79	1 (4)
Invasive mechanical ventilation		30 (32.3)	4 (30.8)	26 (32.5)	0.98	7 (28)
Invasive mechanical ventilation + ECMO		2 (2.1)	0 (0)	2 (2.5)	0.97	1 (4)
Neutrophils, mean +/- SD	18/93	11.9 ± 7.1	11 ± 7.5	12.9 ± 7.7	0.48	11.2 ± 7.0
C-reactive protein, mean +/- SD,	17/93	366 ± 142	363 ± 154	371 ± 136	0.83	367 ± 156
Procalcitonin, median [IQR]	33/93	6 [2–68]	4 [2.3–22]	12 [3.8–63]	0.10	6.5 [4.1–58]
Positive urinary antigen test, n (%)	1/93	80 (86)	12 (92.3)	68 (85)	0.68	16 (64)
PCR from lower respiratory tract sample, n (%)		40 (43)	7 (53.8)	33 (41.3)	0.54	12 (48)
Positive, n (%)	1/47	29/40 (72.5)	6/7 (85.7)	23/33 (69.7)	0.35	7/12 (58.3)
From BAL	0/47	12/40 (30)	2/7 (28.6)	10/33 (30.3)	0.99	4/12 (33.3)
From PDBS	0/47	17/40 (42.5)	4/7 (57.1)	13/33 (39.4)	0.38	4/12 (33.3)
From endotracheal aspiration	0/47	11/40 (27.5)	1/7 (14.3)	10/33 (30.3)	0.97	5/12 (41.7)
PCR from nasopharyngeal swab, n (%)	0/93	7 (7.5)	0 (0)	7 (8.75)	0.59	6 (24)
Culture from lower respiratory tract sample, n (%)	0/93	15 (16.1)	3 (23)	12 (15)	0.43	4 (16)
Positive culture n (%)	0/8	7/15 (46.7)	1/3 (33)	6/12 (50)	1	2 (50)

Organ support during ICU stay

During the ICU stay (Table 4), 70% of patients required invasive mechanical ventilation. ARDS developed in 64 patients (69%), with 10 of them (15%) requiring veno-veinous extracorporeal membrane oxygenation (VV-ECMO). There were no significant differences between treatment groups regarding organ support (mechanical ventilation, vasopressors, or renal replacement therapy). However, ARDS severity within the first 48 h was significantly greater in the combination therapy group, as indicated by a higher use of prone positioning (p = 0.002). The post-intubation P/F ratios and ARDS severity grades did not differ between groups and are provided in (Table 4).

Table 4.

Therapeutic interventions during ICU stay and ARDS severity

Characteristics	Missing data n = 93	Study population n=93	Monotherapy n = 13	Combination therapy n = 80	p	Non exclusive combination therapy n = 25
High-flow nasal oxygen, n (%)	0/93	59 (63.4)	7 (53.8)	52 (65)	0.54	17 (68)
Non-invasive ventilation, n (%)	0/93	8 (8.6)	0 (0)	8 (10)	0.59	2 (8)
Invasive mechanical ventilation, n (%)	0/93	65 (69.9)	7 (53.8)	58 (72.5)	0.21	18 (72)
ARDS, n (%)	0/93	64 (68.8)	8 (61.5)	58 (72.5)	0.10	19 (76)
Prone positioning, n (%)	0/93	44 (47.3)	1 (7.7)	43 (53.7)	0.002	15 (60)
Vasopressor support, n (%)	0/93	54 (58)	6 (46.1)	48 (60)	0.37	13 (52)
Neuromuscular blockade (beyond induction), n (%)	0/93	55 (59.1)	4 (30.8)	51 (63.75)	0.035	15 (60)
Use of renal replacement therapy, n (%)	0/93	17 (18.3)	1 (7.7)	16 (20)	0.45	5 (20)
Use of veno-venous ECMO, n (%)	0/93	10 (10.75)	0 (0)	10 (12.5)	0.35	2 (8)
P/F ratio at 0–24h post intubation, mean +/- SD	0/93	110.2 ± 62	119.8 ± 58	105.3 ± 59	0.54	131.1 ± 83
P/F ratio at 24 h post intubation, mean +/- SD	0/93	150.1 ± 61	152.4 ± 53	148.5 ± 60	0.86	162.6 ± 75
Weaning tracheostomy, n (%)	0/93	10 (10,6)	3 (23.1)	7 (8.75)	0.15	1 (4)
ARDS, n (%)	0/93	64 (68.8)	8 (61.5)	58 (72.5)	0.10	19 (76)
Maximum ARDS severity during first 48 h, n (%)	0/93				0.20
Mild		6 (6.5)	2 (25)	4 (6.9)		4 (16)
Moderate		22 (23.7)	3 (37.5)	19 (32.7)		5 (20)
Severe		38 (40.9)	3 (37.5)	35 (60.4)		10 (40)

Characteristics of the antibiotic treatment

All patients received appropriate antibiotic therapy within 24 h of ICU admission. Eighty patients (86%) received a combination therapy. Fluoroquinolones were administered to 90 patients (97%), primarily levofloxacin (n = 79; 88%), followed by ofloxacin (n = 6; 6.5%) and ciprofloxacin (n = 5; 5.5%). Macrolides were prescribed in 81 patients (87%), predominantly spiramycin (n = 79; 98%). The median duration of antibiotic treatment in the ICU was 11 days [IQR 6–18].

Most combination therapies consisted of a fluoroquinolone–macrolide regimen, with one exception involving fluoroquinolone-rifampicin. Monotherapy was used exclusively in three centers, while two centers did not use monotherapy at all. Among the 13 monotherapy cases, 10 (77%) were fluoroquinolone-based and 3 (23%) involved macrolides.

Primary outcome

28-day mortality did not differ significantly between groups (15.4% (95% confident interval 14.9–15.9%) in monotherapy vs. 12.5% (95% confident interval 11.9–13.1%) in combination therapy; p = 0.59; Fig. 2). No risk factor for 28-day mortality was found (Table 5).

Fig. 2.

Kaplan-Meier survival curve showing 28-day survival from ICU admission

Table 5.

Risk factors of 28-day mortality

Variables	Univariable analysis				Multivariable analysis
	Odd Ratio (95% confident interval)		P value		Odd Ratio (95% confident interval)		P value
Age	1.05 (0.99–1.12)		0.078		1.03 (0.94–1.14)		0.51
Monotherapy	0.78 (0.15–4.07)		0.774		-		-
Male sex	0.79 (0.19–3.18)		0.742		-		-
Smoking history > 10 pack-years	0.36 (0.10–1.29)		0.119		0.38 (0.05–2.95)		0.36
Alcohol consumption > 2 units/day	0.65 (0.13–3.24)		0.602		-		-
Heart failure	18.57 (3.65–94.49)		< 0.001		8.89 (0.15–32.10)		0.24
Chronic obstructive pulmonary disease (COPD)	1.38 (0.15–12.96)		0.778		-		-
Chronic kidney disease	6.41 (1.23–33.36)		0.027		1.37 (0.03–72.31)		0.88
Diabetes mellitus	1.04 (0.21–5.34)		0.957		-		-
Hematologic malignancy	6.25 (1.45–26.92)		0.014		2.36 (0.22–25.52)		0.48
Immunosuppression	2.69 (0.76–9.54)		0.126		4.83 (0.71–33.06)		0.11
IGSII score	1.06 (1.02–1.11)		0.007		1.04 (0.97–1.13)		0.20
ARDS	0.99 (0.97–1.03)		0.950		-		-
Prone positioning	1.13 (0.34–3.81)		0.841		-		-
Vasopressor support	9.72 (1.19–78.84)		0.033		7.68 (0.10–58.51)		0.36
Neuromuscular blockade (beyond induction)	4.01 (0.82–19.42)		0.086		1.31 (0.96–4.96)		0.99
Use of renal replacement therapy	1.59 (0.38–6.65)		0.524		-		-
Use of veno-venous ECMO	0.717 (0.08–6.23)		0.769		-		-

In bold, p-values less than 0.05

Secondary outcome

Secondary outcomes (Table 6) showed no significant differences between groups in median ICU length of stay, duration of mechanical ventilation, high-flow oxygen therapy, or 90-day mortality. An additional analysis was conducted by subdividing the combination therapy group into exclusive and non-exclusive combination therapy. Kaplan–Meier survival analysis confirmed the absence of a significant difference in 30-day survival from ICU admission according to antibiotic strategy (log-rank p = 0.132; Fig. 3). No adverse events were reported in the monotherapy group, compared to 10% in the combination therapy group (p = 0.59), primarily mild cutaneous reactions (n = 7) and one case of confusion (n = 1).

Table 6.

Primary and secondary endpoints

Characteristics	Monotherapy N = 13	Combination therapy n = 80	P	Non exclusive combination therapy n = 25
28-day mortality from ICU admission, n (%)	2 (15.4)	10 (12.5)	0.59	1 (4)
90-day mortality from ICU admission, n (%)	3 (23.1)	13 (16.25)	0.69	1 (4)
ICU length of stay, median (days) [IQR]	12 [3–22]	11.5 [6–21]	0.44	12 [7–24]
Duration of HFNO, median (days) [IQR]	3 [1–4]	2 [1–4]	0.68	2 [1–4]
Duration of invasive mechanical ventilation, median (days) [IQR]	16.5 [8–20.5.5]	10 [6.5–14]	0.52	14 [7–19]
Antibiotic-related adverse events, n (%)	0 (0)	8 (10)	0.59	2 (8)

Fig. 3.

Kaplan–Meier survival curves showing 30-day survival from ICU admission according to antibiotic strategy (monotherapy, exclusive combination therapy, and non-exclusive combination therapy)

Discussion

In this retrospective multicenter cohort of 93 ICU patients with severe Legionella pneumophila pneumonia, 28-day mortality did not differ between monotherapy and combination therapy in accordance with one retrospective study [23] and meta-analyses [15, 16]. The 28-day ICU mortality rate was 13% which aligns with reported ranges (9.1%–41.7%) [6, 9, 23–26]. Early diagnosis and administration of appropriate antibiotics within 24 h likely contributed to this outcome and may have reduced differences between the two groups, highlighting the critical role of prompt therapy in severe Legionella pneumophila pneumonia [5, 9–11, 27, 28].

The patients were severely respiratory ill, with over two-thirds requiring mechanical ventilation and two-thirds developing ARDS, consistent with previously reported rates (mechanical ventilation: 20–70%; ARDS: 50–80%) [7, 9, 23]. ARDS severity within the first 48 h was higher in the combination therapy group, with more frequent use of prone positioning and neuromuscular blockade, suggesting that antibiotic strategy was guided by perceived disease severity and potentially explaining the observed trend toward higher mortality in this group.

Antibiotic-related toxicity, though infrequent and non-severe, occurred in 10% of the combination therapy group and none in the monotherapy group. Both fluoroquinolones and macrolides carry notable adverse effects: macrolides can prolong the QT interval, cause gastrointestinal disturbances, and interact via CYP3A4 [29], while fluoroquinolones are linked to neurotoxicity [30], dysglycemia [31], and increased cardiovascular risks, including aortic aneurysm and dissection [32]. These risks are particularly relevant in older, polymedicated ICU patients, making monotherapy probably preferable to minimize cumulative toxicity and drug interactions [5].

Recommendations are consistent for non-severe Legionella pneumophila pneumonia, for which combination therapy has not shown benefit over monotherapy [15]. However, in severe cases, guidance varies widely. The 2019 IDSA guidelines recommend a fluoroquinolone-macrolide combination [19], while the 2019 BTS guidelines advise a FQ alone, with the option of short macrolide or rifampicin course [17]. In Japan, FQ monotherapy is standard even in severe [33]. French guidelines from 2012 allowed either monotherapy (preferably levofloxacin) or combination therapy [18], whereas more recent French recommendations favor monotherapy in critically ill patients [20], though this stance relies on limited evidence.

The heterogeneity of available in vitro and in vivo data likely explains the variability in international guidelines for severe Legionella pneumophila. In vitro studies suggest inconsistent synergy between fluoroquinolones and macrolides [12, 34, 35]. Fluoroquinolones are bactericidal, with low minimum inhibitory concentrations for levofloxacin and moxifloxacin [36, 37], whereas macrolides are mostly bacteriostatic [38], with variable partial bactericidal activity at high concentrations [36, 37, 39]. In vivo findings are similarly conflicting : two large retrospective studies (French ICU cohort and Japanese nationwide cohort) found no mortality benefit with combination therapy [23, 40], while a small Spanish prospective study suggested a benefit in septic shock [26]. Overall, recommendations remain heterogeneous due to limited and predominantly retrospective evidence across populations of differing severity.

Beyond mortality and tolerance, several practical factors influence the choice of therapy for.

Legionella pneumophila. A large Japanese cohort of 3,560 patients found no major difference in hospitalization costs across FQ monotherapy, macrolide monotherapy or combination therapy [40], though cost-effectiveness varies by healthcare system. FQ – particularly levofloxacin – offer advantages due to their excellent oral bioavailability and pharmacokinetic equivalence to IV formulations [41], enabling early oral switch and potentially shortening ICU and hospital stays [42]. From an ecological standpoint, monotherapy generally limits antibiotic exposure, reducing selective pressure and the risk of multidrug resistance [43, 44]. In line with this, recent French expert recommendations emphasize reducing antibiotic use, shortening treatment durations, and optimizing administration routes in the ICU [45].

This study has several limitations. The small size of the monotherapy group (n = 13) limits statistical power, particularly in the context of multiple comparisons. Therefore, non-significant findings should be interpreted with caution and should not be considered evidence of equivalence between groups. As a retrospective analysis, treatment was not randomized, which may have biased results. Missing data were not imputed, as analyses were restricted to variables with complete data. Grouping all combination therapy patients aimed to reduce bias. Intercentre variability reflects real-world practice. Pharmacokinetics, pharmacodynamics, and co-infections were not assessed.

Conclusion

This multicenter retrospective study suggests that early fluoroquinolone monotherapy may be as effective as combination therapy for severe Legionella pneumophila pneumonia in ICU patients. Randomized clinical trials are needed to compare the safety and efficacy of different antibiotic regimens.

Abbreviations

ARDS

Acute respiratory distress syndrome

BTS

British thoracic society

ICU

Intensive care unit

IDSA

Infectious diseases society of America

PCR

Polymerase chain reaction

SAPS

Simplified acute physiology score

VV-ECMO

Veno-veinous extracorporeal membrane oxygenation

Author contributions

MB and GJ conceived the study, were in charge of the study management, data analysis and manuscript preparation. MB, GJ and CC analysed the results. MB and GJ drafted the manuscript. All authors participated in the study design and data collection; and read and approved the final manuscript.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Data availability

Data and materials supporting the findings of this study can be entirely shared if asking.

Declarations

Ethics approval and consent to participate

The ethics committee for non-interventional research of Rouen University Hospital gave approval for this study (N°1058).

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.