Linezolid

Limits on what this shows

• Key null findings: In the in vitro study, mitochondrial membrane potential did not change at Cmin or Cmax and basal oxygen consumption was not affected in HL-60 cells [4]. In one case with lactic acidosis, thiamine was normal [7], so the lactic acidosis was not explained by a measured thiamine deficit.
• At-risk subgroups: Treatment beyond 28 days, higher trough concentration (above about 2 ug/mL in one trial [3]), renal impairment (higher exposure of metabolites and higher hematologic toxicity [1][11]). A single case suggests the mitochondrial J1 haplogroup as a susceptibility factor; this is one patient and unconfirmed [6].

Direction

Drug toxicity (excess/toxic direction). This is not a nutrient deficiency, and no nutrient was measured in these sources.

Regulatory status

Label section 5.7 warns of lactic acidosis and says patients with recurrent nausea or vomiting, unexplained acidosis or a low bicarbonate need immediate evaluation; postmarketing reports are primarily in patients treated beyond the maximum labeled 28 days; no incidence figure is given in the sections retrieved. Section 13 reports that in male rats the reversible fertility effect involved spermatids with "abnormally formed and oriented mitochondria" [1].

Population evidence

Song 2015 [3]: prospective randomized trial in chronic extensively drug-resistant tuberculosis; blood-cell mitochondrial function measured serially as the cytochrome c oxidase to citrate synthase activity ratio; trough concentrations in 38 participants at 600 mg and 300 mg. Participants on 600 mg had a higher risk of mitochondrial toxicity-related adverse events than on 300 mg (HR 3.10, 95% CI 1.23 to 7.86). Mean mitochondrial function was higher before linezolid than on 300 mg (P=0.004) or 600 mg (P<0.0001). Higher trough concentration was associated with lower mitochondrial function (Spearman rho -0.48, P=0.005), and all participants with troughs above 2 ug/mL developed a mitochondrial toxicity-related event at either dose. Read as a reconstructed OpenAlex abstract; the full text was not opened. A congress poster [5] (peripheral blood mononuclear cells from patients, Seahorse respirometry) reports progressive decline with longer treatment, maximal respiratory capacity most affected, and complex I and IV proteins affected; patient numbers and effect sizes are not stated, so it is not counted as a second Tier A source. Liu 2021 [25] (second, independent Tier A cohort; lactate is a downstream outcome marker, not a mitochondrial measurement): retrospective propensity-matched cohort of patients older than 85, 98 per group after matching (199 teicoplanin and 108 linezolid before matching); lactic acidosis (pH below 7.35 and arterial lactate at least 3.5 mmol/L) in 35 of 98 (35.7%) on linezolid versus 0 of 98 (0%) on teicoplanin (P<0.0001); in multivariable analysis linezolid duration of 9 days or more had OR 3.54 (95% CI 1.16 to 10.79, P=0.026), arterial glucose 8 mmol/L or more OR 4.55 (1.51 to 13.73) and SOFA score OR 1.43 (1.21 to 1.69) per point.

Mechanism

Named and partly demonstrated. In HL-60 and THP-1 human cells, linezolid at therapeutic concentrations (clinical Cmin 2.5 and Cmax 15 mg/L) almost completely inhibited expression of CYTox I (a mitochondrially encoded subunit) at Cmax over 48 to 72 hours, with an IC50 for expression of 4.81 mg/L and for cytochrome c oxidase activity of 2.41 mg/L; effects were fully reversible within 48 to 72 hours of withdrawal [4]. A letter and author reply state that McKee et al. showed inhibition of mitochondrial protein synthesis in rat tissues (primary paper not opened) [23]. A single case report showed cytochrome c oxidase deficiency and depressed respiratory chain complexes (all except II) in muscle after linezolid-associated lactic acidosis [6]. Bacterial-ribosome similarity to the mitochondrial ribosome is the proposed basis [6]. A rat and cell-line study [24] (Tier B) also found lower serum fatty acids and trimethyl phosphate in linezolid-dosed rats and lower ATP content in cultured hepatoma cells, but ran no mitochondrial function assay.

Counter-evidence held on record

Counter-evidence and limits Only one peer-reviewed cohort measured mitochondrial function directly (38 participants, a tuberculosis trial population with other drugs and disease effects). The second cohort [25] supplies the lactic acidosis incidence but is a single-center retrospective study in very elderly, severely ill patients (shock, renal replacement therapy and other lactate-raising conditions excluded), linezolid concentrations were not monitored, and the authors state the mechanism is unclear; lactic acidosis is a drug outcome under Hard Rule 2. Human case-series biopsy papers exist but their pages were not opened, so they are not cited. Why orange and not red A second independent Tier A source with linezolid in the cohort now exists, but it measures lactate, not mitochondrial function, and both signals are bounded (prolonged exposure or high trough, and in [25] age over 85). The directly measured mitochondrial endpoint still rests on one cohort [3]. Framing caution This is a drug toxicity entry bounded to prolonged or high exposure, not a deficiency claim. It must not be read as a reason to take mitochondrial or energy supplements. Any acidosis symptoms during treatment are a prescriber question.

Limits on what this shows

• Null findings: Thiamine alone gave no significant protection (P>0.05) [19]. No human oxidative-stress marker study of linezolid was located in the channels that could be reached.

Regulatory status

No oxidative-stress warning in the label text retrieved [1].

Population evidence

None in humans. Isik 2025 [19]: 24 male Wistar rats, 6 per group; linezolid 125 mg/kg orally every 12 hours for 28 days (a far higher dose than clinical). Liver MDA 3.6 to 6.9 nmol/mg protein (about 92% rise), total GSH 8.3 to 4.1 nmol/mg protein (about 51% fall), SOD 7.7 to 3.8 U/mg protein, catalase 5.6 to 3.1 U/mg protein; ALT 52.3 to 139.4 U/L and AST 76.8 to 274.6 U/L. Thiamine pyrophosphate restored the markers (MDA 4.7, GSH 6.5, SOD 7.3, catalase 5.1) while thiamine alone did not (MDA 6.8, GSH 4.3). Tier B (drug-specific animal study).

Mechanism

Presumed secondary to mitochondrial dysfunction; the authors' proposed pathway (impaired thiamine conversion) was not tested directly [19].

Counter-evidence held on record

Counter-evidence and limits No thiamine-only or TPP-only control groups, males only, one time point, six animals per group, semiquantitative histology [19]. The same experiment supplies the thiamine entry, so the two are not independent. Framing caution Not an antioxidant-depletion claim for patients and not a reason to take antioxidants or glutathione precursors.

Limits on what this shows

• Null findings: No iron indices, transferrin saturation or ferritin were reported in any opened linezolid source.

Direction

Misutilization or mitochondrial iron accumulation (a retention-type disturbance). It is not iron depletion, and iron supplementation would be the hazard, not the remedy (Hard Rule 5).

Regulatory status

Sideroblastic anemia is listed among postmarketing myelosuppression reports; no iron statement or interaction in the label text retrieved [1].

Population evidence

Bernstein 2003 [17]: case report, 78-year-old woman with endocarditis; marrow showed adequate megakaryocytes, ringed sideroblasts and vacuolated pronormoblasts; anemia was present from admission and did not change; the report gives no hemoglobin, ferritin or iron values. Tier B. Hij 2021 [26] (Tier B, abstract only): a 70-year-old man treated for one month with several antibiotic lines for a skin infection presented with cytopenias, vacuolated erythroid precursors and ring sideroblasts; linezolid was proposed as the cause and he improved after it was stopped; the abstract gives no hemoglobin, ferritin or iron values and does not itself name linezolid as the drug received. The same report's other patient (a 15-year-old girl with ring sideroblasts and low serum copper and ceruloplasmin) was not exposed to linezolid and is not evidence for this drug.

Mechanism

Plausible: heme synthesis finishes in mitochondria, which linezolid impairs. Not demonstrated by measured iron indices.

Counter-evidence held on record

Counter-evidence and limits One case with pre-existing anemia and several confounders (endocarditis, multiple prior antibiotics). The case authors also attribute the thrombocytopenia to an immune mechanism [17]. Framing caution Anemia is not iron deficiency. Do not read this entry as a reason to take iron, and keep it separate from Copper (Hard Rule 7).

Limits on what this shows

• Why no grade: Dual condition met after the repair pass: no Tier B or C source was found to support a yellow grade (the only copper-deficiency patient in the linezolid-related hits was not on linezolid), and no circulating claim about linezolid and copper was encountered in any opened source (a consumer-page audit was not possible after the search budget ran out). This is a disclosed gap, not a finding of safety. Independent of Iron (Hard Rule 7).

What was searched

Original pass: web search "linezolid copper deficiency OR serum copper anemia neutropenia" returned general copper-deficiency papers with no linezolid title; none was opened. Repair pass: PubMed listing "linezolid copper" (14 records; none reports copper status in a linezolid-treated patient; hits were copper alloy surfaces, nanoparticle delivery, resistance genes and similar) and PubMed listing "linezolid copper anemia" (1 record, Hij 2021 [26]). In [26] the only copper-deficient patient (a 15-year-old girl) had no linezolid exposure, and the linezolid-treated patient had ring sideroblasts but no copper measurement in the abstract. The label sections read contain no copper statement [1].

Channel outcomes

PMC and NCBI article pages: reCAPTCHA interstitial. A PubMed listing with ceruloplasmin, hypocupremia and serum copper terms, a Crossref bibliographic search for linezolid copper deficiency, and the web search tool (budget exhausted) were refused or unavailable (HTTP 429 or budget), and the refused pages were not re-fetched. Two dedicated PubMed copper queries were completed and returned no usable record. Likely a genuine gap in the indexed literature, but not fully confirmed: the ceruloplasmin and hypocupremia phrasings could not be run. Copper deficiency can itself cause anemia and neutropenia, so a future search should specifically exclude a copper contribution in prolonged-course cytopenia reports.

Limits on what this shows

• Key null findings: No effect on anaerobes other than those named; full normalization by day 35 [18]. No 16S or functional sequencing study of linezolid was opened.
• At-risk subgroups: Not defined by the sources read. Prolonged courses and co-prescribed antibacterials were not studied here.

Regulatory status

Label 5.5: Clostridioides difficile-associated diarrhea (CDAD) is a labeled warning for linezolid, ranging from mild diarrhea to fatal colitis, and has been reported to occur over 2 months after antibacterial administration [1].

Population evidence

Lode 2001 [18]: crossover study in 12 healthy adults (6 women, 6 men, 18 to 40 years), linezolid 600 mg twice daily for 7 days, then a 4-week washout and amoxicillin/clavulanic acid. Stool culture on days -2, -1, 4, 8, 14, 21 and 35. On linezolid, enterococci were markedly suppressed and Klebsiella increased; bifidobacteria, lactobacilli, clostridia and Bacteroides decreased markedly; other anaerobes were unaffected; the flora normalized in all volunteers by day 35. Effect sizes were not given in the abstract-level record read (reconstructed OpenAlex abstract).

Mechanism

Present: an antibacterial with Gram-positive activity acting on the commensal Gram-positive and some anaerobic flora. The effect on Bacteroides was observed only as a stool culture count change.

Counter-evidence held on record

Counter-evidence and limits Culture-based, 12 healthy volunteers, short course; it does not show a lasting change, loss of diversity measured by sequencing, or any downstream effect on vitamin synthesis or absorption. The study's comparator drug produced overlapping reductions in bifidobacteria, lactobacilli and clostridia. Framing caution This is a transient composition finding, not a claim of dysbiosis harm and not a reason to take probiotics or B vitamins. Persistent diarrhea during or after treatment is a clinical question for the prescriber.

Limits on what this shows

• Key null findings: Pediatric label data show no excess anemia versus vancomycin [1]. The 600 mg arms in ZeNix had no participant with hemoglobin more than 25% below baseline [8]. Pyridoxine did not change anemia risk in two comparative cohorts [13][15].
• At-risk subgroups: Courses beyond 2 weeks, 1200 mg/day, renal impairment (OR 2.31 for anemia [11]; and OR 4.15 for thrombocytopenia in a separate 24-study meta-analysis, seen as a summary only [12]), pre-existing myelosuppression, moderate to severe hepatic impairment [1].

Direction

Decrease in hemoglobin from drug myelosuppression. Under Hard Rule 2, anemia is an outcome and is not filed as iron, B12, B6 or folate deficiency; none of these was measured in the cohorts below.

Regulatory status

Label 5.1: myelosuppression including anemia, leukopenia, pancytopenia and thrombocytopenia; weekly complete blood counts, with particular attention beyond 2 weeks, in renal or hepatic impairment and pre-existing myelosuppression; consider stopping if it worsens. Anemia in adults was 2.1% on 600 mg every 12 hours in non-skin indications versus 1.4% in all comparators, and 0.4% versus 0% for the 400 mg skin-infection regimen; a substantially abnormal hemoglobin value was seen in 7.1% of adults on the 600 mg regimen (comparator rate not captured). Pediatric anemia was 5.6% versus 7.1% on vancomycin [1]. Sideroblastic anemia is listed among postmarketing myelosuppression reports [1].

Population evidence

ZeNix trial [8]: 181 randomized to four linezolid regimens. Hemoglobin more than 25% below baseline in 13 participants (7%): 9 at 1200 mg for 26 weeks, 4 at 1200 mg for 9 weeks, none on either 600 mg arm; one participant had hemoglobin below 8 g/dL. Laboratory-confirmed myelosuppression was 22% (10/45), 15% (7/46), 2% (1/45) and 7% (3/45) in the four arms respectively. Pooled drug-resistant tuberculosis series: anemia 38.1% of 107 patients with adverse event data (12 studies) [9] and 25% (95% CI 15 to 34) of 367 patients (15 studies) [10]. Meta-analysis of 25 studies, 3,831 patients: anemia OR 2.31 (95% CI 1.27 to 4.21) in decreased versus preserved renal function [11]. In a 52-patient cohort of courses of at least 3 weeks, linezolid was stopped for severe anemia in 3/24 (12.5%) without and 1/28 (3.5%) with pyridoxine (P=0.24) [13]. Abstract-level: anemia 6/24 (25%) in 24 bone-infection patients [14].

Mechanism

Proposed, not proven in humans: suppression of erythropoiesis through inhibition of mitochondrial protein synthesis. Label section 13 reports dose- and time-dependent bone marrow hypocellularity and decreased hematopoiesis in animals [1]. One case showed ringed sideroblasts and vacuolated pronormoblasts, which the authors read as a chloramphenicol-like suppression; the anemia there predated linezolid [17].

Counter-evidence held on record

Why orange and not red Several independent Tier A sources measure hemoglobin directly in linezolid cohorts with consistent direction and a proposed mechanism, so the evidence is strong. The effect size is bounded by exposure window and dose (2.1% in labeled courses versus 25 to 38% in prolonged tuberculosis regimens), which is the orange boundary condition; the nutrient-style severity of red is not met. The tuberculosis series are uncontrolled and confounded by other drugs and disease. Framing caution Anemia on linezolid is a drug effect and not evidence of an iron, B12 or B6 deficiency. Iron, B12 or B6 supplements are not recommended by this entry. Platelet fall is a related but separate hematologic toxicity and is not an oxygen-delivery finding.

Limits on what this shows

• Null findings: Mitochondrial membrane potential (TMRM) was unchanged at Cmin and Cmax; basal OCR unchanged in HL-60; no assay measured ATP content [4]. In the rat study serum ATP was below the detection limit and no mitochondrial function, electron transport or ATP synthase assay was run [24].

Regulatory status

No ATP or energy-output statement in the label text retrieved [1].

Population evidence

None for ATP in humans. Milosevic 2018 [4]: HL-60 and THP-1 cells; spare respiratory capacity decreased at Cmax but not Cmin (exact percentages given only in figures); cytochrome c oxidase activity 10 to 30% residual at Cmax at 72 hours (value reported for both oxazolidinones combined); basal oxygen consumption unaffected in HL-60; ATP-linked respiration and glycolysis not reported. A poster [5] on patient PBMCs reports a decline in maximal respiratory capacity with treatment length and proteomic changes in 28 ATP-synthesis-related proteins, with no ATP values. Ye 2018 [24] (Tier B, rat plus cell line): 24 male Sprague-Dawley rats (control 7, low dose 8, high dose 9) given linezolid 60 or 120 mg/kg by oral gavage for 7 days; serum ATP was below the assay detection limit (0.3 ug/mL), so ATP was not measured in vivo. In HepG2-C3A hepatoma cells exposed for 24 hours, ATP by HPLC was 2.26 plus or minus 0.50 ug/mL versus 5.13 plus or minus 0.16 ug/mL in untreated cells (P<0.05; group size not stated). Trimethyl phosphate, which the authors used as an indirect proxy for ATP, was lower at the high dose (1.7 plus or minus 0.9 versus 5 plus or minus 2.3 in controls, P=0.007) and several serum fatty acids were lower.

Mechanism

Plausible through impaired mitochondrially encoded respiratory chain subunits. Whether this lowers cellular ATP in vivo is not shown.

Counter-evidence held on record

Counter-evidence The in vitro null findings in [4] argue against a net ATP deficit at therapeutic concentrations. The paper titled "Linezolid Inhibited Synthesis of ATP in Mitochondria" [24] was opened this repair pass through a publisher PDF: it is small (all male, one 7-day time point), the effective cell concentration is unclear ("1mg/mL 50uL"), ATP was measured only in a hepatoma line and not in rat tissue or isolated mitochondria, no multiple-comparison correction was applied, and the authors themselves call the ATP-synthesis mechanism a hypothesis. It supports a yellow grade, not a higher one. Framing caution A respiratory-capacity signal in cell lines is not an energy-deficit claim. Kept separate from Mitochondrial burden and Oxygen delivery/utilization (Hard Rule 7); shared evidence is disclosed in the Overlap map.