Introduction

Daptomycin is an antibiotic of the class of lipopeptides. Its spectrum of activity includes most gram-positive bacteria. It is widely used in clinical practice and, being administrable once daily, also for long-term therapy in patients discharged from the hospital [1]. Despite its higher cost compared to other antibiotics, its preference is justified by favourable handling characteristics, low resistance development, and a strong safety profile [1,2,3]. Notably, daptomycin combined with rifampicin has demonstrated effectiveness in treating complex device-related infections due to its biofilm activity [1, 4].

Daptomycin is generally well-tolerated, yet like all medications, it carries potential adverse effects. One of the most severe is eosinophilic pneumonia, specifically daptomycin-induced eosinophilic pneumonia (DIEP), as documented in the literature. The exact mechanism behind daptomycin’s pulmonary toxicity remains unclear, but several theories have been proposed. Daptomycin is quickly inactivated by pulmonary surfactant. The drug is then absorbed by alveolar macrophages which present it as an antigen to T helper cells. These cells subsequently produce interleukin-5. Alongside eotaxin, which is released by endothelial cells, interleukin-5 leads to the recruitment of eosinophils into the alveolar spaces. This sequence of events triggers an inflammatory cascade that causes acute epithelial damage [5, 6]. Daptomycin can induce eosinophilic pneumonia causing dyspnea, hypoxemia and respiratory failure [7]. The literature provides limited and unclear information about this disease; however, given its severity and rapid progression, this type of pneumonia requires prompt diagnosis, as well as more specific diagnostic criteria and treatment approaches.

This study aims to present a series of cases collected between January 2019 and June 2020 in Santa Maria della Misericordia Hospital, Perugia, Italy. We also performed a systematic review of the available literature.

Case series

Case 1

P. U., a 69-year-old man, was admitted on February 16, 2019, diagnosed with Methicillin-Sensitive Staphylococcus aureus (MSSA) endocarditis affecting his native aortic valve, accompanied by multiple emboli and abscess formations. His medical history included two prior surgeries—a discectomy and a hip prosthesis implant—though no significant pathologies or known allergies were reported.

Upon admission, a comprehensive body CT scan revealed multiple findings: retroperitoneal abscess collections near the retrocaval area and within both psoas muscles, a cortical abscess in the right kidney, and several non-colliquated external iliac lymph nodes. In the upper and posterior mediastinum, additional abscess collections were identified, with the most significant adjacent to the thoracic esophagus, below the azygos vein, and behind the subclavian artery. Another abscess was located near the right scapular glenoid. Additionally, a septic embolus was noted in the dorsal segment of the right upper lobe.

Given these findings, the patient urgently underwent thoracoscopy for the placement of pleural and abscess drains. The procedure, however, was complicated by the development of a pneumothorax.The chosen treatment regimen began with oxacillin (3 gm IV q4h) and clindamycin (600 mg IV q8h). The treatment faced challenges, including delayed blood culture negativity extending beyond seven days and ongoing fever. As a result, on March 8, clindamycin was substituted with daptomycin (10 mg/kg per day) and rifampicin (600 mg IV q24h), while continuing oxacillin. Since the modification did not yield significant improvements, daptomycin was discontinued after a week. Despite the complex clinical situation, the patient did not experience respiratory failure during the initial three weeks. On March 28, daptomycin at a dosage of 10 mg/kg per day was reintroduced to his treatment. Eight days later, the patient developed symptoms including fever and respiratory failure characterized by hypoxemia and bronchospasm. Laboratory tests indicated elevated white blood cell count (12,560/mmc) and peripheral eosinophil count (690.8/mmc, 5.5%), along with a significant increase in C-Reactive Protein (CRP, 21.6 mg/dl, normal < 0.5 mg/dl). High-resolution computed tomography (HRCT) of the chest revealed bilateral patchy pneumonic infiltrates and bilateral pleural effusion. Eosinophilic pneumonia was suspected, leading to the immediate cessation of daptomycin and initiation of corticosteroid therapy. The patient’s respiratory function improved significantly within 24 h, and inflammatory markers gradually decreased. He was discharged on May 17, 2019, without the need for oxygen or ongoing corticosteroid treatment. A follow-up chest radiograph five weeks later confirmed the complete resolution of the radiological findings. This case was deemed probable according to ATS criteria and definite according to the Lyon algorithm, although it did not meet the Phillips Criteria [8,9,10].

Case 2

R. C., a 67-year-old man, was admitted to the post-cardiac surgery intensive care unit on April 9, 2019, following the replacement of his aortic and mitral valves with bio-prosthetic valves. His medical background included chronic kidney disease, chronic obstructive pulmonary disease (COPD), hemosiderosis, non-Hodgkin lymphoma, and allergies to amoxicillin/clavulanate and ceftriaxone.

Following the valve replacement surgery, a sample from the mitral valve was analyzed in the microbiology lab and tested positive for Methicillin-resistant S. epidermidis (MRSE). Based on the sensitivity profile, he was started on daptomycin at a dosage of 10 mg/kg. After 22 days on daptomycin, he presented with a high fever and respiratory failure. Laboratory tests showed elevated white blood cell count (19,790/mmc), peripheral eosinophil count (1,504/mmc, 7.6%), and CRP (28 mg/dl, normal < 0.5 mg/dl). The HRCT of the chest revealed bilateral patchy reticular interstitial patterns with ground glass opacities and septal thickening, along with left pleural effusion. Bronchoalveolar lavage identified 20.5% eosinophils, while tests for bacterial and Pneumocystis jirovecii infections returned negative results.

Given the suspicion of eosinophilic pneumonia, daptomycin was discontinued and corticosteroid therapy initiated. The patient’s respiratory function improved significantly within 96 h, and inflammatory markers decreased steadily. He was discharged without the need for supplemental oxygen or continued corticosteroid treatment. A follow-up chest radiograph two months later confirmed the complete resolution of the radiological findings. This case was classified as probable according to ATS criteria and definite according to both the Lyon algorithm and Phillips Criteria [8,9,10].

Case 3

P. F., a 77-year-old man, was admitted to the hospital for osteomyelitis on March 27, 2019. His medical history included chronic kidney disease, COPD, and an allergy to iodinated contrast media. Upon admission, he was placed on empirical treatment with daptomycin (7 mg/kg per day), ceftazidime (2 gm IV q8h), and ciprofloxacin (400 mg IV q12h). Four days into the treatment, he developed high fever and respiratory failure. Laboratory tests showed an increase in white blood cell count (9,840/mmc), peripheral eosinophil count (1,043/mmc, 10.6%), and CRP (7.9 mg/dl, normal < 0.5 mg/dl). Corticosteroid therapy was initiated immediately. The HRCT of the chest revealed bilateral patchy reticular interstitial patterns with ground glass opacities, predominantly in the upper regions. Bronchoalveolar lavage performed two days post-corticosteroid therapy showed only 0.3% eosinophils, and all microbiological tests for bacteria, viruses, and fungi returned negative results. Given the suspicion of eosinophilic pneumonia, daptomycin was discontinued. The patient’s respiratory function improved significantly within 36 h, and his inflammatory markers decreased steadily. He was discharged without the need for supplemental oxygen but continued on low-dose corticosteroid therapy. Follow-up chest radiography two months later confirmed the complete resolution of the radiological findings. This case has been classified as probable according to ATS criteria and definite according to both the Lyon algorithm and Phillips Criteria [8,9,10].

Case 4

T. M., a 75-year-old outpatient, was receiving daptomycin (7.73 mg/kg/die) for an early prosthetic knee infection. His medical background included COPD, high blood pressure, benign prostatic hyperplasia (BPH), and gastroesophageal reflux disease, with no known kidney injury or history of allergies. After 34 days on daptomycin, he exhibited symptoms of high fever and respiratory failure and was admitted to the hospital on July 22, 2019. Laboratory tests showed a normal white blood cell count (8,500/mmc) but an elevated peripheral eosinophil count (1,122/mmc, 13.2%) and CRP (7.9 mg/dl, normal < 0.5 mg/dl). HRCT scans of the chest revealed a bilateral patchy reticular interstitial pattern with ground glass opacities and septal thickening. Bronchoalveolar lavage was not conducted. Given the suspicion of eosinophilic pneumonia, daptomycin was discontinued, and treatment with corticosteroids and a new antibiotic regimen began. The patient’s respiratory function improved significantly within 48 h, and his inflammatory markers gradually decreased. He was discharged without the need for oxygen or ongoing corticosteroid treatment. This case was classified as probable according to ATS criteria and definite according to the Lyon algorithm, though it did not meet the Phillips Criteria [8,9,10].

Case 5

C. F., a 67-year-old man, was admitted to the post-cardiac surgery intensive care unit on September 27, 2019, following an aortic-coronary bypass operation. His medical history included ischemic cardiomyopathy, type 2 diabetes, hypertension, chronic kidney disease, COPD, and dyslipidemia, but no known allergies. During his stay, he developed acute kidney failure and a MSSA surgical wound and bloodstream infection. Based on the antibiogram results, treatment with daptomycin (11.67 mg/kg) commenced on October 29, 2019.

Seventeen days into treatment, he presented with high fever and Acute Respiratory Distress Syndrome (ARDS). Laboratory results indicated a normal white blood cell count (6,330/mmc) but an increased peripheral eosinophil count (367/mmc, 5.8%) and CRP (7.3 mg/dl, normal < 0.5 mg/dl). HRCT scans displayed a bilateral patchy reticular interstitial pattern with ground glass opacities and septal thickening, along with bilateral pleural effusion. Bronchoalveolar lavage revealed an eosinophil percentage of 11%, and all microbiological tests for bacteria, fungi, and Pneumocystis jirovecii returned negative results. Suspecting eosinophilic pneumonia, daptomycin was replaced with ceftobiprole, which showed limited effectiveness. Corticosteroid therapy commenced ten days later, leading to rapid improvement in respiratory function and a gradual reduction in inflammatory markers. Corticosteroids and oxygen therapy were eventually discontinued. A chest radiograph taken five weeks later confirmed the complete resolution of the radiological abnormalities. This case was classified as probable according to ATS criteria and definite according to the Lyon algorithm, though it did not fulfill the Phillips Criteria [8,9,10].

Case 6

On June 1st, 2019, B. S., a 43-year-old woman with metabolic syndrome, uncontrolled type 2 diabetes, and a history of osteomyelitis, was admitted due to severe pain in her right leg and foot. Physical examination showed crepitation in her leg and a fever. Radiography revealed pulmonary blurring, furthermore, X-rays identified air within the superficial cutaneous layers of her leg and discontinuity in the cortical bone. Early signs of ischemia in her leg led to the disarticulation of the fifth toe and multiple surgical incisions that drained purulent material. She initiated antibiotic treatment with meropenem and teicoplanin. She was diagnosed with sepsis and necrotizing fasciitis after blood, skin biopsy, and swab cultures isolated several types of gram-positive cocci, including Streptococci and Bacteroides. On June 3rd, her antibiotic regimen was adjusted, replacing teicoplanin with ceftaroline and daptomycin due to the rapid spread of infection on healthy skin and emerging respiratory distress, prompting urgent surgical intervention. Her right leg was amputated. Echocardiography ruled out endocarditis.

As healing was poor, tigecycline was added to target a multi-resistant Acinetobacter baumannii identified in a rectal swab. Despite these measures, a higher-level amputation was necessary. On June 23rd, daptomycin was discontinued, lightening the antibiotic load. The following day, she experienced a rapid decline in respiratory function and became dyspnoic; CT scans showed patchy areas, and peripheral eosinophilia (16%) raised suspicions of DIEP, although bronchoalveolar lavage was not performed due to her severe respiratory distress.

Her respiratory condition gradually improved after all antibiotic treatments were halted and replaced with amoxicillin/clavulanate. She was discharged on July 10th in stable condition, with a clean wound and low-dose oxygen therapy. This case meets the criteria for a probable diagnosis according to ATS guidelines, and definite according to both the Lyon algorithm and Phillips Criteria [8,9,10].

Systematic review of literature

Materials and methods

We conducted a systematic review of literature following Preferred Reporting Items for Systematic Reviews and MetaAnalyses (PRISMA) Statement protocol [11].

A systematic search on PubMed, Google Scholar, Directory of Open Access Journal and Cochrane Database of Systematic Reviews (DOAJ) was performed from inception to the end of September 2023. We used the following search strategy: every search included the terms “eosinophilic”, “pneumonia” and “daptomycin”. Each article was analyzed to establish if reached eligibility criteria. The inclusion criteria were as follows: English, Italian or Spanish language, pertinence on the issue, and articles that reported sufficient patient information such as age, sex, clinical and radiological presentation, treatment and outcome.

The exclusion criteria were as follows: languages different from those mentioned above; abstract and posters from conference proceedings because they did not pass under literature review. All cases have been evaluated according to American Thoracic Society (ATS) criteria, Phillips criteria and Lyon algorithm [8,9,10].

Data were summarized using descriptive statistical analysis where possible. Descriptive characteristics were summarized using numbers and percentages for categorical variables, and either means with standard deviations (SD) or medians with interquartile ranges (IQR) for numerical variables, as appropriate. The Kolmogorov-Smirnov test was employed to determine the distribution of continuous variables.

Additionally, the Chi-square test was used to assess whether different treatments were associated with recovery or improvement outcomes. A p-value of < 0.05 was considered statistically significant.

Results

A total of 115 results were found on Google Scholar, 71 on Pubmed and 17 on DOAJ, among these 44 papers matched the eligibility criteria. There were 32 case reports, 7 case series, 3 case reports plus review and 2 case series plus review, with a total of 68 cases. We analyzed data also including our five presented cases, with a total of 74. Among these, applying clinical criteria from ATS [8], we identified 15 definite cases (20.3%), 54 probable cases (73.0%), 5 possible cases (6.8%), and no unlikely cases. Patients’ characteristics are summarized in Table 1.

Table 1 A systematic review of literature: clinical characteristics of cases with suspected daptomycin induced eosinophilic pneumonia (DIEP).
Full size table

Of note, six cases did not match the criteria for definite cases because the authors did not report if the patient had or not fever. Applying criteria proposed by Phillips and colleagues [9], 43 (58.2%) cases are definite for DIEP. According to the Lyon algorithm, 64/67 (95.5%) cases are classified as definite. Of note, about 7 cases data were not enough to apply the Lyon algorithm [10].

The median age was 67.0 (interquartile range, IQR, 60.0–76.0, range 26–92) and 58 (79.4%) cases were male. The median daptomycin dose was 6.0 mg/kg/die (IQR 6-7.9, range 4–11.7) and the eosinophilic pneumonia symptoms appeared after a median of 19.0 days (IQR 12–24) of treatment. The most common initial symptoms were fever (49 cases, 70%) and respiratory symptoms (67 cases, 90.5%) such as dyspnea. For 52 cases, the value of peripheral eosinophils has been reported (in absolute number or percentage), among these, eosinophilia was diagnosed in 45 patients (86.5%). The bronchoalveolar lavage (BAL) was performed in 46 cases but eosinophils count was available for 43 cases with a median of eosinophils of 20.5% (IQR 10.5–20.5, range 0–98.0). Of note, the only patient with 0% eosinophils on BAL and peripheral blood count was a leukopenic woman with hematologic disease [12]. A lung biopsy was performed in eight cases and in six of them eosinophilic pneumonia was confirmed.

Radiological findings (available for 72 cases) were various, but some findings recurred such as bilateral opacities (49 cases, 68.1%), ground-glass opacities (30 cases, 41.7%), patchy infiltrates (22 cases, 30.6%) and peripheral predominance (14 cases, 19.4%).

In all cases, when eosinophilic pneumonia was suspected, daptomycin was withdrawn and in 20 cases (27.0%) no other therapy was added. In 38 cases (51.4%), the daptomycin withdrawal was associated only with corticosteroid therapy and in 12 cases (16.2%) with also antibiotic therapy. In 4 cases (5.4%), corticosteroid was not necessary but antibiotic therapy was added. As shown in Table 1, for each kind of treatment we found a good percentage of recovery (≥ 73.7%). Among the different treatments, none were significantly associated with a higher recovery rate (p > 0.05). The outcome was favorable for most of the patients with a recovery of 78.4% and improvement of 18.9%. Two patients (2.7%) died. Furthermore, a rapid improvement after daptomycin withdrawal has been described in most of the reports. The time of improvement was available for 46 cases and most of them (40 cases, 87.0%) significantly improved after 96 h. The improvement has been especially described as oxygen support reduction/suspension.

Specific data from every case included in the literature are resumed in Table 2.

Table 2 Summary of 74 cases of presumed daptomycin-induced eosinophilic pneumonia
Full size table

Discussion

Daptomycin-induced eosinophilic pneumonia belongs to a diverse group of interstitial lung diseases, which are marked by peripheral blood eosinophilia, elevated eosinophils in BAL fluid, or eosinophilic infiltration of the lung parenchyma as confirmed by lung biopsy [52]. Various drugs, including antibiotics, non-steroidal anti-inflammatory drugs (NSAIDs), and cardiac anti-arrhythmic medications like amiodarone, have been linked to eosinophilic pneumonia [53]. Typically, patients with drug-induced eosinophilic pneumonia show a favourable response to steroid therapy and discontinuation of the offending drug [52].

Recent studies indicate a prevalence of DIEP between 4.8 and 15% among patients treated with daptomycin, increasing with age, particularly in those over 70 [54,55,56]. Factors such as male gender, longer treatment durations (over 21 days), higher dosages, and haemodialysis also increase risk [56, 57]. According to our literature review, the median age of patients was 67 years, with a predominance of males. The median duration of therapy was consistent with previous studies at 19 days (IQR 12–24). However, therapy dosages varied, with 71% of the population receiving a daptomycin dose of less than 8 mg/kg/day.

Garreau et al. identified a daptomycin AUC24h > 939 mg/h/L as a risk factor for DIEP and suggested as safe an AUC24h target range of 666 to 939 mg/h/L [58]. However, it is necessary to specify that therapeutic drug monitoring (TDM) data are not available in most clinical centres. The TDM was not used in most of the cases included in this review.

Our systematic review found no significant association was noted with past allergies, chronic lung disease, or renal impairment. The initial symptoms often included fever and respiratory issues of varying severity.

Diagnosing DIEP is challenging due to the restrictive nature of ATS criteria. Although increased eosinophils on BAL are essential for a definitive diagnosis, BAL was only performed in 62.2% of cases, with a significant count in just under half. This data demonstrated that it is not always possible to perform a BAL and, when it is, the result is diriment in less than a half of cases. The timing of BAL could affect results, as daptomycin is typically withdrawn quickly once DIEP is suspected, potentially reducing eosinophil counts by the time of the procedure.

The ATS criteria exclude patients who have less than 25% Eosinophils on BAL or who have not undergone the test. In these cases, it is possible to define the case as “possible”.

Regarding the criteria proposed by Phillips, patients who do not present dyspnoea cannot be defined as cases of DIEP. Furthermore, it does not consider patients who cannot undergo BAL or lung biopsy.

In contrast, the Lyon algorithm, which classified 95.5% of reviewed cases as DIEP, offers a more sensitive approach. The advantage of this algorithm is the possibility of identifying such a serious and rapidly evolving pathology more easily and promptly. However, a diagnosis, perhaps incorrect, requires the exclusion of an effective and easy-to-handle drug such as daptomycin. Therefore, it is essential to exclude an alternative diagnosis.

Special attention is necessary for immunocompromised patients, particularly those with haematological conditions, as they may not exhibit typical signs like peripheral eosinophilia or increased eosinophils on BAL [12].

Laboratory findings for the disease are nonspecific, including leukocytosis, eosinophilia, and the detection of Il-5 in bronchoalveolar lavage fluid, which are all associated with acute eosinophilic pneumonia [59].

West et al. observed in a cohort of 330 patients in therapy with daptomycin that elevated peripheral eosinophilia ≥ 5% was a risk factor for daptomycin pulmonary eosinophilia [55]. Peripheral eosinophilia is particularly important for raising suspicion of DIEP, especially since bronchoalveolar lavage (BAL) cannot always be performed. In our data, peripheral eosinophilia was present in 86.5% of reported cases.

Based on the data available, we found that a particular importance should be placed on the radiological findings that, together with the exposure to daptomycin, should represent a red flag for physicians. According to our findings, bilateral and ground-glass opacities, patchy infiltrates, and peripheral predominance are the most frequent radiological characteristics described. The significance of imaging, particularly HRCT, is well-established for diagnosing various forms of acute eosinophilic pneumonia. Common imaging findings include ground glass opacities, interlobular septal thickening, pleural effusion, thickened bronchovascular bundles, air space consolidations, and centrilobular nodules [60].

Regarding therapy, it is obviously essential to quickly discontinue daptomycin. If respiratory function does not improve rapidly or the onset is particularly severe (e.g. need for orotracheal intubation) it may be considered to introduce steroid therapy. Currently, there is no specific dosage or specific criteria for introducing therapy. Patients generally exhibit a rapid response to corticosteroids, with clinical improvement often observed within 48 h and resolution of infiltrates typically occurring within a month. Although the optimal corticosteroid regimen for acute eosinophilic pneumonia has not been definitively established, severe respiratory failure often [61] necessitates high intravenous doses (up to 500 mg/day of methylprednisolone), which can be transitioned to oral prednisone (40–60 mg/day) upon improvement. Oral prednisone is usually maintained for 2–4 weeks, followed by a gradual taper over the subsequent weeks [61].

In several cases, since bacterial pneumonia could not be ruled out, antibiotic therapy was still administered.

The overall percentage of recovery is 78.8% and rapid improvement is a predominant characteristic of the disease. More than half of patients improved after 48 h from daptomycin withdrawal and 87% after 96 h.

The study’s limitations include variable study quality, insufficient data, and a small number of definite cases per ATS criteria.

In conclusion, DIEP is a rare but rapidly progressive adverse event. Early diagnosis and treatment are crucial to avoiding serious consequences. Diagnosis relies on clinical evaluation supported by lab tests and imaging. Daptomycin cessation is fundamental, and corticosteroids may be necessary. Continued research is essential to refine diagnostic criteria for this serious adverse event.