Safety and Tolerability of More than Six Days of Tedizolid Treatment

Tedizolid has demonstrated its efficacy and safety in clinical trials; however, data concerning its tolerability in long-term treatments are scarce. The aim of the study was to assess the indications and to describe the long-term safety profile of tedizolid. A multicentric retrospective study of patients who received tedizolid for more than 6 days was conducted. Adverse events (AEs) were identified from patients’ medical records and laboratory data. The World Health Organization causality categories were used to discern AEs that were probably associated with tedizolid.

ABSTRACT

Tedizolid has demonstrated its efficacy and safety in clinical trials; however, data concerning its tolerability in long-term treatments are scarce. The aim of the study was to assess the indications and to describe the long-term safety profile of tedizolid. A multicentric retrospective study of patients who received tedizolid for more than 6 days was conducted. Adverse events (AEs) were identified from patients’ medical records and laboratory data. The World Health Organization causality categories were used to discern AEs that were probably associated with tedizolid. Eighty-one patients, treated with tedizolid 200 mg once daily for a median (interquartile range [IQR]) duration of 28 (14 to 59) days, were included; 36 (44.4%) had previously received linezolid. The most common reasons for selecting tedizolid were to avoid linezolid potential toxicities or interactions (53.1%) or due to previous linezolid-related toxicities (27.2%). The most common indications were off-label, including prosthetic joint infections, osteomyelitis, and respiratory infections (77.8%). Overall, 9/81 patients (11.1%) experienced a probably associated AE. Two patients (2.5%) developed gastrointestinal disorders, 1 (1.2%) developed anemia, and 6 developed thrombocytopenia (7.4%) after a median (IQR) duration of treatment of 26.5 (17 to 58.5) days. Four (5%) patients discontinued tedizolid due to AEs. Among 23 patients with chronic renal failure (CRF), the rate of myelotoxicity was 17.4%, and only 8.7% had to stop tedizolid; 20 out of 22 with previous linezolid-associated toxicity had no AE. Long-term tedizolid treatments had good tolerance with rates of gastrointestinal AE and hematological toxicity lower than those reported with linezolid, particularly in patients with CRF and in those with a history of linezolid-associated toxicity.

INTRODUCTION

Tedizolid phosphate is an expanded-spectrum oxazolidinone with activity against Gram-positive bacteria, including methicillin-resistant Staphylococcus aureus (MRSA) and cfr-mediated linezolid-resistant S. aureus that is indicated for the treatment of acute bacterial skin and skin structure infections (ABSSSI) in adults. Two phase III randomized, double-blind clinical trials (CTs), Establish-1 and Establish-2, demonstrated the noninferiority of 6 days of tedizolid (200 mg per day) versus 10 days of linezolid (600 mg every 12 h) in patients with ABSSSI (1, 2). Gastrointestinal (GI) disorders (nausea, diarrhea, and vomiting) and myelotoxicity were less frequent in tedizolid than in the linezolid arm (3, 4). A recent meta-analysis of four randomized CTs involving 2,056 patients, comparing the efficacy of linezolid versus tedizolid for the treatment of ABSSSI, reconfirmed the previous results (5). The potent activity against Gram-positive cocci (6–8), high oral bioavailability, improved dosage profile (once daily), the expected lower risk of drug-drug interactions (9, 10), and a better safety profile at 6 days of treatment compared to linezolid makes tedizolid an attractive alternative for infections that require longer periods of antibiotic treatment, such as bone and joint infections or mycobacterial infections, among others. However, the information about the tolerability of tedizolid for treatment courses longer than 6 days is scarce. Recently, Kim et al. (11) evaluated the safety and tolerability of tedizolid in 25 patients with nontuberculous mycobacterial infections who received tedizolid for a median of 91 days. They suggest that long-term tedizolid treatment might have a safety profile comparable to linezolid.

Tedizolid was approved in Spain in 2015, and it has been used off-label for more than 6 days in different indications. The aim of our study was to assess the most common indications of prolonged treatments and to describe the long-term safety profile of tedizolid in a real-world scenario.

RESULTS

A total of 81 patients with a median (interquartile range [IQR]) age of 66 (56 to 77) years were included in the study, and the characteristics are depicted in Table 1. The proportion of males was 59%; 35% had diabetes mellitus; 28% had chronic renal failure, which included 3 patients undergoing hemodialysis; and 40.7% had immunosuppression (HIV, solid organ transplantation, bone marrow transplantation, autoimmune diseases, cancer, or history of cancer). Thirty-six (44.4%) patients had previously received linezolid treatment. The most common reasons for starting tedizolid were to avoid linezolid toxicity (38.3%), due to previous linezolid toxicity (27.2%), to avoid linezolid-drug interactions (9.9%), or to avoid both linezolid toxicity and interactions (4.9%).

TABLE 1.

General characteristics of patients treated with tedizolid for more than 6 days^a

Patient and treatment characteristic(s) (n = 81)
Age (yrs) (median [IQR])	66 (56–77)
Male (no. [%])	48 (59.3)
Median (IQR) body mass index (kg/m2)b	25.8 (22–29.6)
Comorbidity (no. [%])
Diabetes mellitus	28 (34.6)
Chronic renal failure	23 (28.4)
Cancer or history of cancer	17 (20.9)
Autoimmune diseases	7 (8.6)
Solid organ or bone marrow transplantation	7 (8.6)
Cirrhosis (Child B)	2 (2.4)
HIV	2 (2.4)
Source of infection (no. [%])
Prosthetic joint infection	20 (24.7)
ABSSSI	18 (22.2)
Osteomyelitis-septic arthritis	18 (22.2)
Respiratory focus	7 (8.6)
Endovascular infections	4 (4.9)
Biliary focus	3 (3.7)
Pulmonary mycobacteria infection	3 (3.7)
Disseminated mycobacteria infection	2 (2.5)
Intra-abdominal focus	2 (2.5)
Other	4 (4.9)
Microorganism (no. [%])
MRSA	15 (18.5)
CoNS-MR	12 (14.8)
MSSA	8 (9.9)
CoNS-MS	5 (6.2)
Enterococcus spp.	8 (9.9)
Corynebacterium striatum	7 (8.6)
Nontuberculous mycobacteria	5 (6.2)
Not recovered	14 (17.3)
Other	7 (8.6)
Previous linezolid therapy (no. [%])	36 (44.4)
Reasons for tedizolid prescription (no. [%])
To avoid linezolid potential toxicity	31 (38.3)
To avoid linezolid potential interactions	8 (9.9)
To avoid linezolid potential toxicities and interactions	4 (4.9)
Previous GI toxicity with linezolid	9 (11.1)
Previous GI and hematologic toxicities with linezolid	2 (2.5)
Previous hematologic toxicity with linezolid	11 (13.6)
Suspected therapeutic failure with linezolid	1 (1.2)
Linezolid-resistant strain	1 (1.2)
Other	14 (17.3)
Route of tedizolid administration (no. [%])
Oral	65 (80.2)
Intravenous	14 (17.3)
Intravenous and oral	2 (2.5)
Site of treatment (no. [%])
Ambulatory care	56 (69.1)
Hospital	24 (29.6)
Not available	1 (1.2)
Duration of tedizolid treatment (days) (median [IQR])	28 (14–59)
Tedizolid duration of treatment (days) (no. [%])
>6 to ≤15	23 (28.4)
16 to ≤30	21 (25.9)
31 to ≤60	19 (23.5)
61 to ≤100	11 (13.6)
>100	7 (8.6)
Concomitant antibiotics (no. [%])
Fluoroquinolones	17 (21)
Carbapenems	14 (17.3)
Rifampicin	3 (3.7)
Cephalosporin/penicillin	3 (3.7)
Other	3 (3.7)
Multiple antibiotics	14 (17.3)
Final outcome (no. [%])
Favorable clinical outcome	61 (75.3)
Clinical/microbiological failure	16 (19.8)
Nonrelated exitus	2 (2.5)
Not available/no evaluable	2 (2.5)

^aAbbreviations: GI, gastrointestinal; ABSSSI, acute bacterial skin and soft structures infections; MRSA, methicillin-resistant S. aureus; MSSA, methicillin-susceptible S. aureus; CoNS-MR, methicillin-resistant coagulase-negative staphylococci; CoNS-MS, methicillin-susceptible coagulase-negative staphylococci.

^bAvailable for 46 patients.

All patients received 200 mg once daily and were mainly treated as an outpatient (69.1%). Median (IQR) duration of tedizolid treatment was 28 (14 to 59) days, and the longest treatment was of 357 days. Tedizolid was administered in monotherapy in 27 patients (33.3%), and the most frequent concomitant antibiotics were fluoroquinolones (21%) and carbapenems (17.3%). Regarding the indications of tedizolid, 22.2% were for ABSSSI and 77.8% for off-label entities (prosthetic joint infections [PJI], osteomyelitis, and respiratory tract infections). Five patients with nontuberculous mycobacteria infection (3 with pulmonary focus and 2 disseminated) were treated with tedizolid during a median of 65 days (62 to 68 days). Patients with osteomyelitis, PJI, and ABSSSI received tedizolid for a median of 53 (27.8 to 68.5), 44 (29.3 to 69.3), and 14.5 (12 to 34) days, respectively. The most frequent microorganisms were methicillin-resistant S. aureus and coagulase-negative staphylococci (CoNS-MR) (33.3%), followed by methicillin-susceptible S. aureus (9.9%) and Enterococcus spp. (9.9%).

Of the 9 patients who reported previous linezolid-associated GI adverse events, 8 (89%) presented good tolerance with tedizolid without AEs and completed the prespecified duration of treatment. On the other hand, of the 13 patients (2 also had GI adverse events) who reported previous hematologic toxicity under linezolid treatment, 12 patients (92.3%) did not develop alterations in laboratory data related to tedizolid, and they finished the prespecified duration of treatment.

Overall, 9 out of 81 (11.1%) patients experienced a total of 11 adverse events probably associated with tedizolid (Table 2). Two patients (2.5%) developed GI disorders; one had severe diarrhea (8 bowel movements a day) that led to treatment discontinuation after 25 days of monotherapy with tedizolid. The patient improved after stopping tedizolid, and Clostridium difficile infection was discarded as a potential diagnosis. The other patient, after 30 days of tedizolid in monotherapy, experienced nausea and vomiting that disappeared after changing the administration route (from oral to intravenous). Regarding laboratory data abnormalities, one patient (1.2%), who had previously suffered GI disorders and anemia with linezolid, presented anemia that led to treatment discontinuation after 52 days of tedizolid concomitantly with azithromycin and ethambutol for a Mycobacterium kansasi and Mycobacterium avium-intracellulare infection. Hemoglobin recovery was complete 1 week after stopping tedizolid. Fifty-five patients (68%) had a baseline platelet count of ≥150,000 cells/μl, and 26 (32%) patients had a baseline count <150,000 cells/μl. The rate of thrombocytopenia was 7.4% (6/81 patients) after a median (IQR) of 26.5 (17 to 58.5) days of treatment. In five patients, thrombocytopenia was mild, and in one patient, it was severe with a final platelet count of 37,000 cells/μl after 68 days of therapy, so it was stopped. One patient with mild thrombocytopenia also had mild reductions of white blood cell (WBC) and lymphocyte counts after 15 days of tedizolid with the consequent treatment suspension. No differences in terms of demographics, comorbidity, or baseline parameters were found between patients who did and did not develop tedizolid-induced thrombocytopenia during tedizolid treatment.

TABLE 2.

Patients with adverse events likely related to tedizolid from medical records and laboratory data^a

Patient no.	Age, sex	CKD	Tedizolid indication	Previous linezolid use and toxicity	Tedizolid treatment duration (days)	Adverse event	Baseline value	Final value (day no.)	Stopped due to AE
1	60, M	N	Osteomyelitis	Y (GI disorder)	25	GI disorder (severe diarrhea)	NA	NA	Y
2	62, M	Yb	Prosthetic joint infection	N	270	GI disorder (nausea and vomiting)	NA	NA	N
3	76, M	Y	Pulmonary mycobacteria infection	Y (GI and anemia)	52	Anemia	Hb, 116 g/liter	Hb, 96 g/liter (52)	Y
4	47, M	Yb	Osteomyelitis	NK/NA	30	Thrombocytopenia	PLT, 253,000 cells/μl	PLT,123,000 cells/μl (30)	N
5c	57, M	N	Osteomyelitis	Y (None)	98	Thrombocytopenia	PLT, 250,000 cells/μl	PLT, 142,000 cells/μl (95)	N
6d	67, M	Y	ABSSSI	Y (GI disorder)	14	Thrombocytopenia	PLT, 212,000 cells/μl	PLT, 118,000 cells/μl (15)	N
7	64, F	N	Prosthetic joint infection	N	23	Thrombocytopenia	PLT, 347,000 cells/μl	PLT, 124,000 cells/μl (25)	N
8e	73, M	Y	Disseminated mycobacteria infection	N	68	Thrombocytopenia	PLT, 150,000 cells/μl	PLT, 37,000 cells/μl (68)	Y
9	53, M	N	Osteomyelitis	Y (GI disorder)	15	Thrombocytopenia	PLT, 412,000 cells/μl	PLT, 125,000 cells/μl (15)	Y
						Leukopenia	LKC, 6,490 cells/μl	LKC, 3,700 cells/μl (15)
						Lymphopenia	ALC, 1,100 cells/μl	ALC, 800 cells/μl (15)

^aAbbreviations: CKD, chronic kidney disease; AE, adverse event; M, male; N, no; GI, gastrointestinal; Y, yes; Hb, hemoglobin; ABSSSI, acute bacterial skin and soft structures infections; NK/NA, not known/not answered; PLT, platelet count; NR, nonrecovery; LKC, leukocyte count; ALC, absolute lymphocyte count; NA, not applicable.

^bPatients undergoing hemodialysis.

^cPatient 5 had a risk factor for thrombocytopenia; he used low-molecular-weight heparin for recurrent deep-vein thrombosis due to congenital thrombophilia (factor V Leiden).

^dFor patient 6, all factors (transplant patient with immunosuppressors, renal failure, antibiotics, etc.) could influence thrombocytopenia, but the main factor was tedizolid itself.

^eMultifactorial thrombocytopenia with worsening in probable relationship with tedizolid in patient 8. Although there was no evidence of platelet count recovery after tedizolid withdrawal (no laboratory tests were available beyond 1 week after withdrawal, and the patient was exitus), the patient was included because, according to clinical judgment, the AE was probably due to tedizolid.

In total, 4 (5%) patients discontinued tedizolid therapy due to adverse events, and the characteristics are depicted in Table 2.

No increase in transaminases, reduction in absolute neutrophil counts, lactic acidosis, peripheral or optical neuropathy, or hypersensitivity reactions were described during tedizolid use. Seven patients were concomitantly treated with antidepressants such as selective serotonin reuptake inhibitors (SSRIs), tricyclic antidepressants, or other serotoninergic drugs. Median (IQR) duration of tedizolid therapy in these patients was 40 days (23 to 52.5 days), and no serotoninergic syndrome was observed.

A favorable clinical outcome was documented in 61 (75.3%) patients, 16 (19.8%) had clinical or microbiological failure, 2 patients died due to unrelated causes (2.5%), and 2 were lost at follow-up.

DISCUSSION

We present the largest case series published to date of patients receiving more than 6 days of tedizolid. There is just one approved indication (ABSSSI) by regulating agencies, but our study shows that there is a high percentage (77.8%) of patients treated for unapproved indications, mostly long-term treatments for mycobacterial infections, osteomyelitis, and PJI. In line with this, our study included a heterogeneous population of patients, very different from that evaluated in the pivotal clinical trials of tedizolid. The main reason for selecting tedizolid instead of linezolid was toxicity (4, 5, 12). Interestingly, 22 patients (27.2%) had previous linezolid-associated toxicities that motivated the tedizolid prescription, and 20 did not report AEs.

The incidence of GI disorders was 2.5% (2/81 patients) and, in one patient, motivated the interruption of tedizolid. It is a low proportion compared with the rate observed in CTs (16%) and in the recent meta-analysis conducted by Lan et al. where they reported a 6.8% incidence of nausea, 3.3% of diarrhea, and 2.6% of vomiting. (4, 5) Kim et al., in their study of 25 patients using tedizolid for a median of 91 days, observed GI intolerance in five patients (20%) (11). A possible explanation for this difference could be the retrospective design of our study and the possible lack of registration of these AEs in medical histories; however, we can confirm that these AE were not severe enough to stop the treatment. This prevalence is also lower than the one reported with linezolid in CTs or in a real-life scenario (≥23%) (4, 13).

In the pivotal CTs, the incidence of tedizolid-induced thrombocytopenia was 6.4% at the end of treatment visit (4), and Kim et al. reported 1 patient (4%) developing thrombocytopenia out of 25 patients being treated with tedizolid during a median of 91 days (11). Here, we described thrombocytopenia in 7.4% of patients after a median of 26.5 days of tedizolid, which is in line with previous results and suggests a lower incidence than the one reported with linezolid in real-life studies (13–25). Since anemia at baseline was common in our patients, we decided to evaluate the proportion of patients with a reduction of ≥2 g/liter of hemoglobin at any time during treatment with tedizolid. Only 1 patient developed anemia that led to stopping medication after 52 days. Legout et al. (26), using the same definition, described a rate of anemia of 47% among 51 patients taking linezolid without rifampin for ≥4 weeks. If we consider only the 37 patients receiving tedizolid for ≥4 weeks, the rate of anemia was 2.7%, still significantly lower than the one reported with linezolid. In addition, the rate of myelotoxicity among patients under treatment with linezolid and with chronic renal failure is >30% (15), and it has been associated with a high serum concentration of linezolid in this population (27). However, in our cohort, the rate was 17.4% (4 out of 23), and only 2 patients had to stop tedizolid (8.7%). A potential explanation is that the pharmacokinetics of tedizolid in patients with severe renal impairment remain unchanged (28).

No episodes of C. difficile infection were documented that could be attributed to tedizolid use, a fact that should be highlighted because of the ecological protection this may provide for intestinal microbiota, as described by Tanaka et al. (29).

No cases of lactic acidosis or peripheral or optical neuropathy were described during tedizolid use in our study. In fact, in the study conducted by Fang et al. in volunteers receiving different doses of tedizolid for up to 21 days, they found no evidence of neurologic or visual changes (30). Kim et al., in their study of patients receiving tedizolid during a median of 91 days, reported peripheral neuropathy in five patients (20%). However, one of them had a previous neuropathy during linezolid treatment, and the other 4 were receiving ethambutol concomitantly (11).

The better safety profile of tedizolid compared to linezolid after long-term exposure in our cohort could be explained by several reasons. Oxazolidinones’ main toxicity (lactic acidosis, myelosuppression, and neuropathy) is probably due to its mitochondrial protein synthesis inhibition (31–33). Although the tedizolid concentration that inhibits 50% of the mitochondrial protein synthesis (IC_50mp) is lower than the linezolid IC_50mp, the total daily dose of tedizolid is 200 mg, and the trough serum levels are below the IC_50mp, while the linezolid daily dose is 1,200 mg, and the trough levels are frequently over the IC_50mp. In addition, the analysis of the subcellular location of tedizolid suggested that the cytosol is the main compartment, and it was not associated in a stable fashion with mitochondria or other organelles (34). On the other hand, prolonged treatment with linezolid has been associated with drug accumulation, while data from tedizolid suggests that no accumulation occurs (35).

The main limitation of this study is its retrospective design and therefore the risk of an incomplete description of potential AEs in the medical records. Furthermore, an accurate description of all concomitant medications was not collected, and therefore, we could not evaluate all potential drug-drug interactions. Additionally, we have not recorded Gram-positive strains that would have become tedizolid resistant during the treatment, nor superinfections by Gram-negative bacilli (enterobacteria, Pseudomonas spp.) or by Candida spp. In this sense, the lack of detection of cases of superinfection by Gram-negative bacilli or yeasts would be a limitation of the study. However, this study is the largest to present real-world data about the safety and tolerability of tedizolid, and in all cases, it was possible to identify the reason to stop tedizolid, given the real information about the tolerability of long-term treatment with this antibiotic.

In conclusion, tedizolid has been increasingly used in recent years for off-label indications that require prolonged treatments, mainly to avoid bone marrow and GI toxicities associated with linezolid. Tedizolid seems to have a good tolerance in clinical practice with rates of GI adverse events and hematological toxicity lower than those reported with linezolid, particularly in patients at risk, such as those with chronic renal failure. In addition, tedizolid demonstrated a good safety profile in patients with a history of linezolid-associated toxicity. Nevertheless, it would be advisable to obtain a complete blood count every 2 to 4 weeks to detect laboratory abnormalities. Prospective and controlled studies would be necessary to know exactly the incidence of hematological toxicities related to the long-term use of tedizolid.

MATERIALS AND METHODS

Data collection.

We conducted a multicentric and retrospective study of all consecutive patients who received tedizolid for more than 6 days from 2015 to April 2019 in seven Spanish hospitals (Hospital Germans Trias i Pujol, Hospital Clínico Universitario de València, Clínica Universidad de Navarra, Hospital Universitario Fundación Alcorcón, Hospital Universitario Virgen de las Nieves de Granada, Hospital Universitario y Politécnico La Fe, and Hospital Clínic de Barcelona).

Patient demographic data (age, gender, weight, and height), comorbidities (diabetes mellitus, chronic renal failure, need for renal replacement therapies, liver cirrhosis and Child-Pugh score, cancer or history of cancer, autoimmune diseases, solid organ or bone marrow transplantation, and HIV), source of infection, isolated microorganisms, prior administration of linezolid, and indication for tedizolid (failure or toxicity of previous antibiotic regimen or to avoid linezolid toxicity or drug interactions) were recorded. In regard to tedizolid, dose, route of administration, treatment duration, concomitant antibiotics, potential adverse events related to tedizolid, and outcome of the infection were registered.

Definitions.

Adverse events (AEs) were identified from patients’ medical records and laboratory data. AEs from patients’ medical records were classified into gastrointestinal disorders (nausea, vomiting, diarrhea, and/or abdominal pain), neurological disorders (peripheral and/or optical neuropathy), hypersensitivity (pruritus and/or exanthema), and other AEs. Laboratory data were recorded at baseline, periodically (every 5 to 7 days), and at the end of treatment. Hemoglobin (g/liter), leukocyte count (cells/μl), absolute lymphocyte count (cells/μl), absolute neutrophil count (cells/μl), platelet count (cells/μl), alanine aminotransferase (ALT), and aspartate aminotransferase (AST) were registered. AEs from laboratory data were defined as (i) anemia, decrease in hemoglobin ≥2 g/liter from baseline after tedizolid initiation; (ii) leukopenia, WBC count <4,000/μl after tedizolid initiation and classified as mild (less than low limit normal to 3,000/μl), moderate (2,000 to <3,000/μl), and severe (<2,000/μl); (iii) lymphocytopenia, absolute lymphocyte count of <1,000 μl after tedizolid initiation and classified as mild (less than low limit normal to 800/μl), moderate (500 to <800/μl), and severe (<500/μl); (iv) neutropenia, absolute neutrophil count of <1,500 cells/μl after tedizolid initiation, (v) thrombocytopenia, platelet count of <150,000/μl after tedizolid initiation and further classified as mild (75,000 to 150,000 cells/μl), moderate (50,000 to <75,000 cells/μl), or severe (<50,000 cells/μl). For patients with a baseline platelet count of <150,000 cells/μl, thrombocytopenia was defined as a reduction of 25% from the baseline; and (vi) elevated AST or ALT 3 times above the upper limit of normal. The classification of adverse events severity was made on the basis of the National Cancer Institute Common Terminology Criteria for Adverse Events (CTCAE) version 5.0 (https://ctep.cancer.gov/protocolDevelopment/electronic_applications/docs/CTCAE_v5_Quick_Reference_8.5x11.pdf). Tedizolid treatment duration at the time of AEs and whether they caused treatment discontinuation were recorded. AEs were considered “probable” if there was a reasonable time relationship with the drug intake, it was unlikely to be caused by the disease or other drugs, and there was a clinically reasonable response after tedizolid withdrawal, as defined by the World Health Organization (https://www.who.int/medicines/areas/quality_safety/safety_efficacy/WHOcausality_assessment.pdf). Only AEs probably related to tedizolid intake, based on investigators’ judgment and/or medical records, were included.

Statistical analysis.

Categoric variables were described as proportions and continuous variables using the mean and standard deviation (SD) or the median and interquartile range (IQR). Statistical analysis was performed with STATA version 15.1 statistical software package (StataCorp., TX).

This study was approved by the Clinical Research Ethics Committee of Hospital Clinic Barcelona, and the informed consent from patients was waived due to the retrospective nature of the study by the Ethics Committee.