Introduction

Traumatic brain injury (TBI) is damage to the brain caused by external trauma. Patients with TBI often suffer from post-traumatic disturbances of consciousness and require endotracheal intubation and mechanical ventilation, which increases the risk of hospital-acquired pneumonia (HAP)1. Staphylococcus aureus, in particular methicillin-resistant Staphylococcus aureus (MRSA), is the most common pathogen causing pneumonia in TBI patients, accounting for 29–35%2,3. Linezolid has excellent antibacterial activity against MRSA and has shown superior penetration into the lung epithelial lining fluid and lung macrophages4,5. It is recommended by the Infectious Diseases Society of America (IDSA) as a first-line treatment for MRSA pneumonia6.

The recommended dose of linezolid in the instructions is 600 mg q12h, and the amount of exposure in vivo exposure is strongly correlated with its clinical efficacy and adverse reactions. Currently, the recommended steady-state trough concentrations (Cmin) of linezolid is in the range of 2–7 mg/L7,8. Previous studies have shown that factors such as renal impairment or dysfunction, hepatic impairment, and low or high body weight significantly contribute to its pharmacokinetic (PK) variability9,10,11,12. This variability may result in either increased or decreased exposure to linezolid, potentially leading to adverse effects such as thrombocytopenia or clinical treatment failure7. We have found that TBI patients have lower plasma concentrations of linezolid than other patients, which may increase the risk of treatment failure. TBI-induced cellular and biochemical cascades can lead to secondary tissue damage, such as disruption of renal autoregulation and alterations in hepatic drug metabolism13. In addition, in our patient population, there exist such factors as hyperdynamic states, cerebrospinal fluid (CSF) drainage or the influence of concurrent medications (e.g. vasopressors, diuretics)14,15. These pathophysiological changes may increase the volume of distribution and clearance of antimicrobial agents, resulting in lower exposure, as reported in previous studies treated with vancomycin16 and meropenem17. Luque et al.15 found a significant increase in the volume of distribution and clearance rate of linezolid in neurosurgical patients, which was differently with other populations.

The aim of this study was to investigate the effect of TBI on the plasma trough concentrations and microbiological efficacy of linezolid at the normal dose in patients with HAP, and to explore the possible influencing factors leading to the pharmacokinetic variability of linezolid in the patients with TBI.

Methods

Study design and population

This two-center study was conducted in the intensive care unit (ICU) and emergency intensive care unit (EICU) of Affiliated Suzhou Hospital of Nanjing Medical University and Suzhou Science and Technology Town Hospital. Critically ill patients treated with linezolid between July 2020 and June 2022 were retrospectively collected. The inclusion criteria were as follows: (1) age ≥ 18 years; (2) the clinical diagnosis was hospital acquired pneumonia6; (3) receiving linezolid therapy for at least 3 days, and the first course of linezolid treatment was included if there was more than one course of linezolid therapy; (4) the Cmin of linezolid was determined; (5) patients with TBI admitted to the EICU with a Glasgow Score of 5 to 8 and diagnosis was confirmed by imaging tests such as cranial computed tomography (CT) and magnetic resonance imaging (MRI). Exclusion criteria were as follows: (1) baseline platelet count < 75 × 109/L; (2) hemodialysis, continuous renal replacement therapy (CRRT), or extracorporeal membrane oxygenation during linezolid treatment; (3) concomitant disseminated intravascular coagulation or other bleeding disorders; (4) concomitant use of heparin or anti-platelet agents; (5) severe hepatic impairment; (6) linezolid Cmin outside the limits of detection or incorrect sampling timing; (7) incomplete or missing demographic and clinical information.

Data collection

Demographic and clinical data were collected for each patient by searching the hospital information system and the following information was recorded: (1) demographic characteristics include sex, age, weight, underlying diseases, and hospitalization time; (2) diagnosis, comorbidities, complications and fluid balance; (3) the type of infection, microbiological isolates and minimum inhibitory concentration (MIC); (4) medication regimen including dose, intervals, duration of linezolid administration, Steady-state trough concentration, concomitant antibiotic therapy and the use of dehydrating agents (e.g., mannitol, glycerin fructose); (5) laboratory indexes: the complete blood count (CBC) includes white blood cell (WBC) counts, neutrophil percentage (N%), Platelet (PLT) counts; serum inflammatory biomarkers include C-reactive protein (CRP), procalcitonin (PCT); biochemical indexes include hemoglobin (HB), total bilirubin (TBIL), albumin (ALB), alanine aminotransferase (ALT), creatinine (Cr); estimated glomerular filtration rate (eGFR) was measured within 1–3 days of linezolid administration and within one week of discontinuation of the drug. Laboratory indexes were tested every 3–5 days; (6) linezolid-related adverse reactions and treatment outcomes (28-day survival rate). An EXCEL database was set up for two-person entry and verification. The eGFR was calculated using the 2021 CKD-EPI equation18,19.

Grouping and dosage regimen

Patients enrolled were classified into TBI and non-TBI groups. TBI is a type of brain injury caused by an external force. TBIs can include penetrating injuries (in which an object breaches the skull and dura, with direct damage to the brain parenchyma) and closed-head injuries (in which the skull and dura remain intact). All patients with TBI were confirmed by CT or MRI to have skull fractures, brain bruising, bleeding, or swelling. The initial dose of linezolid in all patients was 600 mg q12h, after which the dose was adjusted based on Cmin and efficacy assessment. Only the first determination of linezolid Cmin was included in our study for statistical analysis.

Determination of linezolid Cmin

Blood samples (1 mL) were obtained and placed in yellow vacuum sampling vessel (containing coagulant and separation glue) within 30 min before the fifth dose20, centrifuged within 2 h, and then 0.5 mL supernatant was collected and placed in the refrigerator at − 80℃ for freezing and storage. The serum isolated above was used to determine the linezolid Cmin. The Concentration of linezolid was determined using a high-performance liquid chromatography-tandem mass spectrometry assay. Quantifying of linezolid was validated over the 0.5–50 µg/mL concentration range with satisfactory accuracies (− 0.59–5.14%), intra-day precisions (≤ 3.45%) and inter-day precisions (≤ 6.99%). The recommended target range of Cmin for linezolid is 2–7 mg/L7,8.

Primary outcome

The primary outcome of this study was to compare the Cmin distribution and microbiological efficacy between the two groups. Microbiological efficacy was evaluated only for patients with confirmed Gram-positive pathogenic microorganisms. The evaluation was based on the most recent microorganism clearance result at the end of the treatment course, and the assessment was recorded as documented eradication, presumed eradication, documented persistence, and presumed persistence. Eradication of Gram-positive bacteria is defined as documented eradication and presumed eradication21.

Secondary outcomes

Secondary outcomes included clinical efficacy, 28-day survival, and adverse events. Clinical efficacy was evaluated only in patients with an identified Gram-positive pathogen. The clinical efficacy was evaluated after 7 to 14 days of linezolid administration, and the results were recorded as clinical success or failure21. Clinical success was defined as a composite clinical improvement, including: (1) Clinical symptoms: resolution of fever and reduced sputum volume; (2) Inflammatory biomarkers: normalization of white blood cell count (WBC), neutrophil percentage (N%), C-reactive protein (CRP), and procalcitonin (PCT) levels; (3) Respiratory support: successful liberation from mechanical ventilation, transition to conventional oxygen therapy, or cessation of high-flow nasal oxygen; and (4) Radiographic improvement: reduction of inflammatory exudates on chest imaging. 28-day survival and adverse reactions were assessed for all patients. Adverse reactions include LIT, severe LIT and other adverse reactions. According to the causality assessment criteria for adverse reaction, cases which judged to be possible, probable, or certain were recognized as related to linezolid treatment22. LIT was assessed during linezolid treatment and within 1 week after discontinuation of the drug. LIT was defined as a platelet count below 112.5 × 109/L (75% lower limit of normal) at any time during therapy in patients with platelet counts above the lower limit of normal (150 × 109/L) at baseline, and severe LIT was defined as a platelet count below 75 × 109/L. For patients with platelet counts below the lower limit of normal range at baseline (75 × 109/L to 149 × 109/L), LIT was defined as a reduction of at least 25% from baseline, and severe LIT was defined as a platelet count below 50 × 109/L12. Other adverse reactions refer to other events associated with linezolid treatment15, such as headache, diarrhea, nausea, etc.

Statistical analysis

Continuous variables of normal distribution were expressed as mean ± SD, and comparison between two groups was tested using the independent t-test. Continuous variables of non-normal distribution were described as median with inter-quartile range, and Mann–Whitney test was used for comparison between groups. Categorical variables were expressed as frequencies and proportions (%) and Chi-square test or Fisher’s exact test was used for comparison between groups. Linear correlation analysis or rank correlation analysis was used between the two quantitative variables, and the correlation coefficient was used to assess the correlation between the two variables. Receiver operation characteristic (ROC) curve was plotted to analyze the predictive effect of linezolid Cmin on gram-positive bacteria eradication. Statistical analysis was performed using SPSS 22.0 (SPSS, Inc., USA) statistical software. Both ROC curve and statistical plotting were performed using GraphPad Prism 9.0 (GraphPad, Inc., USA) software. All p values were two-sided and were considered significant at P < 0.05.

Results

Patient characteristics

A total of 177 patients who met the criteria were enrolled in the study, including 57 patients with TBI and 120 without TBI. Among all patients, the differences in age, weight, hemoglobin (HB), total bilirubin (TBIL), albumin (ALB) and alanine aminotransferase (ALT) between the two groups in the total study population were statistically significant (P < 0.05). Propensity score matching (PSM) was performed between the two groups of patients, after 1:1 PSM, 46 patients with well-balanced baseline levels were enrolled in each of TBI group and non-TBI group (Fig. 1). and there were no significant differences in demographic and clinical characteristics between the two groups (Table 1).

Fig. 1
figure 1

Study flow chart. TBI, traumatic brain injury; PSM, Propensity score matching.

Full size image
Table 1 Demographics and clinical characteristics of the study population before and after PSM. PSM, propensity score matching; TBI, traumatic brain injury; IQR, interquartile range; SD, standard deviation; APACHE II, acute physiology and chronic health evaluation II; SOFA, sequential organ failure assessment; CRRT, continuous renal replacement therapy; MODS, multiple organ dysfunction syndrome; WBC, white blood cell; N, neutrophil; hb, hemoglobin; PLT, platelet; CRP, C-reactive protein; PCT, procalcitonin; TBIL, total bilirubin; ALB, Albumin; ALT, Alanine aminotransferase; cr, creatinine; eGFR, estimated glomerular filtration rate.
Full size table

Site of infection and pathogenesis

Pulmonary infection was diagnosed in all patients. Nine patients in TBI group had combined intracranial infections, there were no statistically significant differences between the two groups in other sites of infections except for intracranial infections. 29 patients in TBI group and 33 patients in non-TBI group were treated with gram-positive bacteria, and there were no statistical differences in the distribution of bacteria, fungal detection rate, and distribution of the MIC of gram-positive bacteria between the two groups, as shown in Table 2.

Table 2 Site of infection and pathogenesis of patients in both groups. TBI, traumatic brain injury; MRSA, methicillin-resistant Staphylococcus aureus; MRCNS, methicillin resistant coagulase negative staphylococci; MIC, minimum inhibitory concentration.
Full size table

Clinical outcomes of patients with gram-positive bacterial infections

There was no statistically significant difference in the dosage or duration of linezolid administration between the two groups. The eradication rate of gram-positive bacteria in TBI group was lower than that in non-TBI group, and the difference was statistically significant (69.0% vs. 90.9%, P = 0.029). Although the clinical success rate was lower in TBI group compared with non-TBI group, the difference was not statistically significant (72.4% vs. 78.8%, P = 0.559). The entire list of clinical efficacy evaluations can be found in Supplementary Table S1 online. LIT and severe LIT occurred in fewer patients in TBI group than in non-TBI group (4.3% vs. 28.3%, P = 0.005; 2.2% vs. 17.4%, P = 0.035). One patient in TBI group developed diarrhea, while three patients in non-TBI group developed 5-HT3 syndrome, intermittent fever, and diarrhea separately. The clinical outcomes of the two groups were shown in Table 3.

Table 3 Clinical efficacy and safety evaluation of the two groups. TBI, traumatic brain injury; IQR, interquartile range; LIT, linezolid-induced thrombocytopenia; adrs, adverse drug reactions.
Full size table

Analysis of the predictive effect of linezolid Cmin on the eradication of gram-positive bacteria

As described above, 29 patients in TBI group and 33 patients in non-TBI group were targeted therapy with Gram-positive bacteria, 20 and 30 patients were eradicated with Gram-positive bacteria at the end of treatment in the two groups respectively. Moreover, these patients were divided into the Gram-positive bacteria eradication group and the non-eradication group and analyzed for the linezolid Cmin in both groups, the results showed that the linezolid Cmin was higher in the Gram-positive bacteria eradication group than in the non-eradication group [5.8 (2.4–11.0) mg/L vs. 2.6 (0.6–9.3) mg/L, P = 0.024]. The ROC curve of linezolid Cmin for predicting the eradication of Gram-positive bacteria indicated that: the area under the curve was 0.76 (95% CI 0.63–0.90); the Youden index was 0.51; the calculated cut-off value of linezolid Cmin was 1.62 mg/L, with the sensitivity 0.84 and the specificity 0.67. The ROC curve was shown in the Fig. 2.

Fig. 2
figure 2

ROC curve for linezolid Cmin predicting eradication of gram-positive bacteria.

Full size image

Effect of traumatic brain injury on linezolid Cmin

Patients from two groups (46 in each group) were included to analyze the impact of traumatic brain injury on the trough concentration of linezolid. The results showed that patients in TBI group had lower Cmin than those in non-TBI group [2.06 (0.89, 6.89) mg/L vs. 6.70 (3.09, 13.48) mg/L, P < 0.001]. There were more patients with Cmin < 2 mg/L in TBI group and more patients with Cmin >7 mg/L in non-TBI group, and the differences were statistically significant (50.0% vs. 15.2%, P < 0.05; 23.9% vs. 47.8%, P < 0.05). As shown in Fig. 3a,b. There were more patients in TBI group with a trough concentration of < 1.62 mg/L (41.3% vs. 10.9%, P = 0.001), similar to the results of patients with linezolid Cmin < 2 mg/L.

Fig. 3
figure 3

(a) Distribution of linezolid Cmin in the two groups. (b) More patients with Cmin < 2 mg/L in the TBI group and more patients with Cmin >7 mg/L in the non-TBI group. *Represents a statistically significant difference.

Full size image

Analysis of the reasons for lower linezolid Cmin in TBI group

The correlation between linezolid Cmin and eGFR was analyzed in both groups, and the results showed that Cmin was negatively corelated with eGFR in non-TBI group (spearman’s r = -0.378, P = 0.010), however, there was no correlation between Cmin and eGFR in TBI group (spearman’s r = -0.287, P = 0.059), shown in Fig. 4a,b.

Fig. 4
figure 4

(a) Correlation between eGFR and linezolid Cmin in the TBI group. (b) Correlation between eGFR and linezolid Cmin in the non-TBI group. Statistical significance was assessed by Spearman correlation.

Full size image

We statistically analyzed the factors that might affect the Cmin, such as fluid intake and output, diuretic use, etc., and found that more patients were using mannitol and glycerol fructose in TBI group, which resulted in a higher 24 h intake and output before the determination of the concentration, as shown in Table 4.

Table 4 Analysis of factors associated with substandard Cmin in the TBI group. TBI, traumatic brain injury; IQR, interquartile range.
Full size table

Discussion

This two-centre retrospective study found that patients with TBI was at a low exposure, had reduced microbial clearance and clinical treatment success. To our knowledge, this is the first study to demonstrate that hyperosmotic agents indirectly lead to increased urine volume in patients with TBI and is responsible for the difference pathophysiological changes when compared with other populations.

Luque et al.15 also reported that the neurosurgical patients had low exposure to linezolid, with plasma Cmin ranging from < 0.2 mg/L to 2 mg/L, which was lower than TBI group in our study, mainly due to differences in body weight and renal function. The median dose of linezolid in the study by Luque et al.15 was 15 mg/kg, which was lower than the 18.5 mg/kg in TBI group in our study. In addition, all patients in the study by Luque et al.15 had an eGFR >80 ml/min, with a median eGFR of 158.5 ml/min. In our study, 8 patients in TBI group had eGFR < 80 ml/min, of which 4 patients had eGFR < 60 ml/min, and the median eGFR was 97.5 ml/min, which was significantly lower than in the study by Luque et al.15. In another two studies, the plasma Cmin of linezolid in the neurosurgical patients was 5.35 ± 6.25 mg/L and 5.6 ± 5.0 mg/L, respectively23,24, both of which were approximately double that of TBI group in this study. The creatinine clearance rate (97.87 ± 38.55 ml/min and 81.3 ± 39.6 ml/min, respectively) and body weight (58.5 ± 15.6 kg and 67.6 ± 11.8 kg, respectively) in these two studies23,24 were not significantly different compared with TBI group in our study, indicating that there were other contributing factors.

Linezolid exposure was strongly correlated with efficacy and adverse reactions in both groups. In our study, the cut-off value of linezolid plasma Cmin for eradication of gram-positive bacteria was 1.67 mg/L using ROC analysis. All patients in both groups had pulmonary infections, and most of the gram-positive bacterial samples came from sputum. The concentration of linezolid in pulmonary lining fluid is 8.35 times higher than in plasma5, and the effective concentration is more readily achieved in the lung, so the plasma Cmin cut-off value for microbiological eradication in this study is below 2 mg/L. The proportion of patients with Cmin >1.67 mg/L in TBI group and non-TBI groups was 63% (29/46) and 89.13% (41/46), respectively, which was closer to the clearance rate of gram-positive bacteria in these two groups. We also observed a significant difference in the incidence of thrombocytopenia between the TBI and non-TBI groups (4.3% vs. 28.3%, P = 0.005), which correlated with linezolid exposure. The linezolid Cmin associated with the development of thrombocytopenia has been reported to be in the range of approximately from 6 to 11 mg/L7,25. The incidence of thrombocytopenia was significantly lower in our study compared to those associated with higher linezolid exposure to, as the risk factors of thrombocytopenia are also affected by the duration of linezolid treatment25. The occurrence of thrombocytopenia was reported in the linezolid treatment for duration of 10 to 14 days26, whereas the median duration of linezolid treatment was 11 days in both study groups, indicating a relatively short duration.

Our study further investigated the factors that influence the plasma concentration of linezolid in TBI patients. Low linezolid exposure has been associated with augmented renal function, overweight or obesity, and younger age in previous studies10,12,27,28, however, there were no statistical differences in eGFR, body weight, and age between the two groups in this study. We found a correlation between linezolid plasma concentration and eGFR in non-TBI group, which was not shown in TBI group, suggesting that there are other factors affecting TBI group. Our study found significant differences in hypertonic drug use and urine output between these two groups. Intracranial hypertension and cerebral edema are common complications of TBI27, and appropriate hypertonic agents (such as mannitol and hypertonic saline) are often required to reduce brain tissue injury and improve cerebral perfusion28,29. In our study, 24 (52.2%) of 46 patients in TBI group were treated with mannitol or glycerol fructose, in comparison only 3 (6.5%) in non-TBI group. Hypertonic agents have a diuretic effect and significantly increase the volume of urine. The median output of urine volume in TBI group 24 h before monitoring was 2540mL, which was significantly higher than that 1550 mL in non-TBI group (P = 0.005), suggesting that this may increase the clearance rate of linezolid in TBI group, resulting in low exposure. In addition, in this study, some patients in TBI group had external ventricular drainage (EVD) or lumbar drainage to reduce intracranial pressure and resolve post-traumatic hydrocephalus, and CSF drainage may increase the clearance rate of some drugs, such as vancomycin and meropenem16,17. However, the degree of impact on drug clearance varies based on drug permeability. Alexia Chauzy et al.29 showed that EVD drainage had a negligible effect on the overall clearance rate of metronidazole, as there is an exchange between CSF and extracellular cerebral fluid, and metronidazole is highly permeable and easily reabsorbed into the cerebrovascular system. Currently, there is a lack of research on the effect of CSF drainage on the clearance rate of linezolid. In the study of Li SC et al.23, no effect of CSF drainage on the clearance rate of linezolid was observed, which maybe due to an insignificant difference in the drainage flow of CSF in enrolled patients. Because linezolid and metronidazole have similar hyperpermeability, such as being widely distributed in the central nervous system and only controlled by passive diffusion control15,20,29,30, we speculate that external CSF drainage also has little effect on linezolid. In addition, in our study, nine patients in TBI group had central nervous system infection. Due to the destruction of the blood-brain barrier and the blood-cerebrospinal fluid barrier, patients with brain injury and central nervous system infection have increased permeability and may also increase drug distribution as a result20,30,31, but drugs with high permeability have been little affected29. Yogev et al.32 concluded that there was no difference in the permeability of linezolid in children with and without inflammatory meninges.

In this study, the effects of brain injury on plasma linezolid concentration were investigated, and the underlying reasons for this result were also analysed. Limitations of this study include (1) The 24-h clearance of linezolid in urine and CSF drainage have not been studied to explore their effect on linezolid clearance rates. (2) Gram-positive bacteria in this study included MRSA, methicillin resistant coagulase negative staphylococci (MRCNS), enterococci, and positive rods, each with different Clinical and Laboratory Standards Institute (CLSI) sensitivity breakpoints for linezolid (e.g., Staphylococcus spp. MIC ≤ 4, Enterococcus spp. MIC ≤ 2), but further stratified analysis of microbial clearance rates and linezolid plasma concentration breakpoints for different Gram-positive bacteria was not performed. (3) Due to limitations of sample size, the results of this study require further validation.

Consequently, TBI patients are extremely likely to have low exposure to standard doses of linezolid, and patients with comorbid pneumonia have a lower trough concentration but are still at risk of treatment failure. This study did not find an association between low exposure and renal function in TBI patients, but demonstrated a correlation between low exposure and increased urine output due to the use of hyperosmolar agents. It is suggested that this subgroup of patients requires therapeutic drug monitoring of linezolid concentrations and the therapeutic dose can be adjusted by establishing of population pharmacokinetic models.