Cefuroxime

Limits on what this shows

• Negative or null findings: No oral, intravenous or in vitro respirometry study with cefuroxime was found. The widely cited 2013 study of bactericidal antibiotics and mitochondrial dysfunction was not opened in this review, so no class result is borrowed.
• At-risk subgroups: Not defined. The finding applies to intracameral injection during eye surgery, an off-label route.
• Counter-evidence on record: None identified; the evidence is a single local animal study.

Direction

Excess exposure toxicity at a local site (drug toxicity, not a nutrient deficiency; Hard Rule 2).

Regulatory status

Not addressed in any label opened. [1][3][4]

Population evidence

None in humans. Tier B, listing-level: a 2020 rabbit study (Cutan Ocul Toxicol, PMID 32854557) of intracameral cefuroxime 1 mg in 0.1 mL found mitochondrial swelling in corneal endothelial cells, similar whether diluted in normal saline or balanced salt solution, with none in untreated control eyes; no other antibiotic was compared. A PubMed listing for cefuroxime with mitochondria or mitochondrial returned 2 records, the other (swine sperm silver nanoparticles) not involving cefuroxime effects. [25]

Mechanism

Not established; the listing says nothing on respiration or membrane potential.

Counter-evidence held on record

Boundary One rabbit eye study seen at listing level; the concentration in the anterior chamber far exceeds systemic exposure. The class literature on bactericidal beta-lactams and mitochondria was not opened in this review, so no class extrapolation is made. Framing caution Local animal toxicity. It does not describe human mitochondrial function after oral or intravenous cefuroxime, and no supplement is implied.

Limits on what this shows

• Negative or null findings: No human antioxidant capacity, MDA or glutathione study with cefuroxime appeared in the listing; the remaining records were analytical chemistry, bacterial or environmental. [24]
• At-risk subgroups: Not defined. Two of the three studies are ocular and use the intracameral route.
• Counter-evidence on record: Route and dose far from usual oral or intravenous exposure; abstract-level data only; 65 of 75 records were not visible.

Direction

Excess oxidant load in animal tissue at local or high-dose exposure (drug toxicity). Not an antioxidant nutrient deficiency.

Regulatory status

Not addressed in any label opened. [1][3]

Population evidence

None in humans. Tier B, listing-level summaries from a PubMed listing "cefuroxime AND oxidative" (75 results, 10 shown): (a) Wistar rats, 30 animals in three groups of 10, intracameral cefuroxime versus balanced salt solution versus sham (Arq Bras Oftalmol 2019, PMID 31116319): corneal total antioxidant status differed from both controls (p<0.001), oxidative stress index differed from BSS (p=0.001) and sham (p=0.026), total oxidant status differed from BSS (p=0.023); the authors concluded cefuroxime increased oxidative stress in the cornea; (b) rabbits, 10 per group, cefuroxime versus vancomycin versus BSS (Graefes Arch Clin Exp Ophthalmol 2010, PMID 20076966): corneal thiol decreased (p=0.001) and malondialdehyde increased (p<0.001) with cefuroxime, no significant oxidative change with vancomycin; (c) Wistar rats given 0, 30, 60 or 120 mg/kg cefuroxime intraperitoneally daily for 7 days (Reprod Toxicol 2019, PMID 30797828): decreased antioxidant enzyme activities and increased MDA in the male reproductive system, reversible after about 70 days (the enzymes were not named in the abstract text). [24]

Mechanism

Not established in the abstracts; oxidative injury to the exposed tissue is the interpretation the authors give.

Counter-evidence held on record

Boundary and overlap Animal tissue toxicity at local or high exposure. The corneal studies share the local ocular-toxicity concept with the mitochondrial entry but are different studies (see the overlap map). Framing caution No evidence that cefuroxime changes oxidative stress or antioxidant nutrient needs in people; no antioxidant supplement is supported.

Limits on what this shows

• Negative or null findings: No ceruloplasmin, serum copper or absorption study was returned by the copper listing. [27]
• At-risk subgroups: Not defined.
• Counter-evidence on record: None; the evidence is chemistry and one unread enzyme paper.

Direction

Metal affects the drug in vitro; enzyme inhibition is a drug effect. No evidence of the drug lowering or raising body copper.

Regulatory status

Copper is not mentioned in the labels except as copper reduction tests for urine glucose (a false-positive reaction) [1][2][3].

Population evidence

None in humans or animals for copper status. Tier B in vitro, listing-level, from a PubMed listing "cefuroxime AND copper" (9 results, all shown) [27]: voltammetry indicated 1:2 metal-to-ligand cefuroxime complexes of Cu(II), Cd(II) and Zn(II) (Bioelectrochemistry 2005, PMID 15713559); cefuroxime axetil was complexed with CuCl2 and the complex was about 5 times more active than the free ligand in a beta-glucosidase inhibition assay (J Biomol Struct Dyn 2025, PMID 40495555); the other records were nanocomposite degradation, analytical chemistry or environmental isolates. Separately, a 1985 paper (Sattler et al., Agents Actions, PMID 3925736) was summarized in that listing as showing that cefuroxime markedly inhibited human diamine oxidase; its abstract was not retrievable (OpenAlex returned metadata only) [28], so the magnitude, concentration and mechanism are unknown and copper chelation is not shown.

Mechanism

Chelation chemistry in vitro; physiological relevance not studied. Whether diamine oxidase inhibition involved copper binding is unknown.

Counter-evidence held on record

Boundary and overlap Tier B chemistry kept distinct from iron (Hard Rule 7). The in vitro metal findings are the same type of evidence as any cefuroxime-metal timing concern and are cross-listed in the overlap map. Framing caution No evidence that cefuroxime changes copper status. Do not infer a need for copper supplementation, and the diamine oxidase line is not a copper finding.

Limits on what this shows

• Why no grade: Dual condition met: (1) no Tier B or Tier C evidence that cefuroxime changes iron status or absorption was found, and (2) no circulating claim about cefuroxime and iron was found on the consumer pages and labels read. The mineral effect that does exist in the beta-lactam literature is cefdinir with iron salts, which runs from iron to a different drug and does not support an iron-status entry for cefuroxime [16]. The hemoglobin and hematocrit decrease on the injection label is graded under oxygen delivery and is not iron evidence.

Direction

Not determined.

Channels tried

PubMed listing "cefuroxime AND iron" (19 results, 10 shown: a moxifloxacin review, the Wiesner beta-lactam review, rat sepsis, plasmid, electro-Fenton degradation, bacterial isolates, a dental resin colour-change study; none measured iron in people or animals given cefuroxime) [29]; the Wiesner abstract and Table 1 (the named iron effect is cefdinir; no cefuroxime antacid or mineral rows visible) [16]; the tablet labels (no iron text) and injection labels (none) [1][2][3][4]; Drugs.com monograph (no iron claim) [32]; PubMed listings for oxidative stress, ATP and copper (no iron-status record) [24][26][27]. PubMed listing "cefuroxime AND ferritin" (1 result: a 1992 thalassaemia major partial splenic embolization versus splenectomy follow-up in 16 patients, indexed under cefuroxime therapeutic use and blood ferritin, whose abstract mentions neither cefuroxime nor ferritin results; cefuroxime there is a perioperative drug and not an iron exposure) [44].

Genuine gap or tooling

Mostly a genuine gap; partly tooling-limited: 9 of 19 PubMed records (page 2) of the broader iron listing were not seen and WebSearch was exhausted before consumer pages could be searched for claims. The ferritin-specific phrasing was run in the repair and returned one unrelated record.

Counter-evidence held on record

Re-check Re-check page 2 of the "cefuroxime AND iron" listing and a "cefuroxime anaemia iron deficiency" or serum iron phrasing when access returns. Framing caution This is a disclosed gap, not a finding of no effect.

Limits on what this shows

• Negative or null findings: Vacarean-Trandafir: the authors conclude cefuroxime "does not promote short-term dysbiosis" in surgical patients without additional perioperative procedures, with only 3.71% of variation explained [14]; this was one prophylactic intravenous dose with no unexposed arm, and the study still saw lower diversity and a significant subgroup PERMANOVA. A randomized trial of 42 mothers (J Pediatr 2019, listing-level summary) found that the timing of maternal cefuroxime prophylaxis did not clearly change infant gut composition at 10 days; at 9 months infants whose mothers received it after cord clamping had more observed species [15]. A 2020 mouse study with five animals per antibiotic found few overall changes with cefuroxime, whereas dicloxacillin and clindamycin caused large shifts [15]. Leigh found no C. difficile in any volunteer [13]. A pediatric Kaiser Permanente abstract listed 1 community CDI case in 852 cefuroxime prescriptions (117 per 100,000) without an effect size [12]. The Ontario case-cohort study of adults over 65 [11] did not report cefuroxime at all, so it neither supports nor contradicts the estimate. A 2000 J Antimicrob Chemother opinion piece, "Ecological antibiotic policy" (PMID 11051626, listing-level summary, not opened), is described as favoring prodrugs such as cefuroxime axetil for lower impact on normal flora; it reports no trial data and is not counted [42].
• At-risk subgroups: Adults aged 65 or older with a recent hospitalization or Charlson index of 3 or more: about 6.6 times the overall cefuroxime absolute risk (42.65 versus 6.48 per 10,000 prescriptions) [8]. Hospitalized adults, with the strongest odds for oral cefuroxime in the prior 60 days [10].
• Counter-evidence on record: The reference in Zhang is another antibiotic (doxycycline, a low-risk drug), so the OR compares antibiotics and not antibiotic versus none [8]; in Baxter every comparator also received an antibiotic, which is why its estimate is the smallest (2.16) [39]. All estimates are observational and confounded by indication and healthcare exposure; the Intermountain authors note that the lower odds seen for daptomycin and doxycycline may reflect confounding by indication [10]. Effect sizes vary with design and window (4.57 at 90 days against doxycycline, 9.59 at 30 days in matched case-control, 6.4 at 60 days in hospitalized adults, 2.16 against other antibiotic users), and cephalexin was lower in the same two MarketScan analyses (1.97 and 2.88) [8][9]. The Miller analysis adjusted for matching variables, Elixhauser comorbidity indicators, PPI receipt and visit counts; the indication for each prescription was not among the adjustments listed [9].

Direction

Disturbance and infection risk (an outcome entry with a direct composition arm). Not a nutrient deficiency, and no probiotic or supplement is implied. The named mechanism is loss of colonization resistance from antibiotic exposure of the gut flora.

Regulatory status

Tablet labels, section 5.2: "Clostridioides difficile-associated diarrhea (CDAD) has been reported with use of nearly all antibacterial agents, including cefuroxime axetil"; superinfection with fungal or bacterial pathogens "should be considered"; diarrhea and nausea/vomiting are the most common adverse reactions (3% or more); in multiple-dose trials (n=912) diarrhea 4% and nausea/vomiting 3%; in the 20-day Lyme disease regimen diarrhea 10.6%. The injection label says CDAD must be considered in any patient with diarrhea after antibiotic use and can occur more than two months after the last dose. UK SmPC: overgrowth of Candida and C. difficile, pseudomembranous colitis ("not known" frequency), and that cefuroxime may affect the gut flora, lowering oestrogen reabsorption (an oral contraceptive statement, not a nutrient finding). [1][2][3][4][5]

Population evidence

Zhang 2022 [8]: retrospective cohort, IBM MarketScan commercial and Medicare supplemental databases, 2008 to 2020, 36.6 million adults given an outpatient antibiotic, 11,607 community-acquired CDI cases (ICD code within 90 days plus CDI treatment). Cefuroxime: 459,679 prescriptions, 298 cases (the 2.6% printed beside it is the share of all 11,607 cases, not a rate), 6.48 cases per 10,000 prescriptions, unadjusted OR 5.65 (95% CI 4.79 to 6.66), adjusted OR 4.57 (3.87 to 5.39) versus doxycycline, attributable risk 5.33 per 10,000, number needed to harm 1,876. In the high-risk stratum (age 65 or older with a hospitalization in the prior 90 days or Charlson index 3 or more): 10,551 prescriptions, 45 cases, 42.65 per 10,000, number needed to harm 303 (the stratum OR sits in a supplement that was not readable). Same table, for scale: clindamycin 8.81, cefdinir 5.86, fluoroquinolones 4.05, cephalexin 1.97, azithromycin 0.77. Miller 2023 [9]: matched case-control (5 controls per case), MarketScan, 159,404 community-associated CDI cases and 797,020 controls, 2001 to 2021, exposure in the 30 days before the index date; cefuroxime exposed in 2,460 cases versus 1,190 controls, unadjusted OR 10.48, adjusted OR 9.59 (8.79 to 10.45); cephalexin 2.88, doxycycline 0.96. Zhang and Miller draw on the same database family and are counted as one source. Webb 2020 [10]: retrospective cohort of 506,068 consecutive adult admissions to 21 Intermountain Healthcare hospitals, 2006 to 2012, 2,356 hospital-associated CDI cases (hospital-onset, or community-onset within 30 days after discharge; positive assay plus more than three stools per day or toxic megacolon); multivariable logistic regression adjusted for age, Charlson score, acid suppression, recent hospitalization and ICU stay, GI procedure, antibiotic days of therapy and others; cefuroxime in the prior 60 days adjusted OR 6.4 (95% CI 5.1 to 7.9, FDR-adjusted P=0.009) and during the index admission 1.6 (1.3 to 1.9, P=0.033); the text names oral cefuroxime among the highest-risk agents, and the tables give one cefuroxime estimate per window with no route split. Agent-level exposure counts are in supplementary tables that were not read. This source is a different institution, years, case definition and covariate set from MarketScan. Baxter 2007 [39]: retrospective case-control study in a nonprofit integrated health system in Northern California, 1999 to 2005, 1,142 patients with hospital-acquired CDAD and 3,351 matched controls (matched on facility, calendar quarter, diagnosis-related group and length of stay), all of whom had received an antibiotic in the prior 60 days, so the comparison is against other antibiotic users. Cefuroxime adjusted OR 2.16, listed among agents with a significantly increased risk (imipenem-cilastatin 2.77, clindamycin 2.31, moxifloxacin 1.88, ceftazidime 1.82, cefpodoxime 1.58, ceftizoxime 1.57, ceftriaxone 1.49; metronidazole 0.67 and doxycycline 0.41 lower). Two independent abstract-level records agreed; the 95% interval and exposed counts were not in either. Leigh 1990 [13]: 10 healthy volunteers, cefuroxime axetil 250 mg twice daily for 4.5 days; mean blood concentration 3.6 mg/L two hours after the ninth dose; four volunteers had high faecal cefuroxime on day 4 or 5. Total anaerobes fell, most where faecal cefuroxime was high; Enterobacteriaceae decreased significantly or were eliminated; streptococci increased; Candida counts rose in six volunteers; two had mild diarrhoea and one Candida vaginitis; C. difficile was not isolated and toxin was absent in the two with diarrhoea. Abstract only (the publisher page showed the abstract, read through a summarizing fetch tool); no log counts were in it. Novelli 1995 [40]: comparison in 24 patients with acute exacerbation of chronic bronchitis given a 10-day course of cefetamet-pivoxil 500 mg twice daily, cefixime 400 mg daily or cefuroxime axetil 250 mg twice daily (about 8 per arm; per-arm counts not in the abstract), stool cultures before treatment, at day 10 and at day 24. Cefetamet changed the flora slightly and not significantly; "CFX and CA significantly affect Enterobacteriaceae and clostridia"; Salmonella spp. and C. difficile newly appeared in 4 and 2 cases respectively across the cefixime and cefuroxime treatment. Abstract only (reconstructed OpenAlex record); the sentence pairing the case counts with each drug is ambiguous, so no per-drug C. difficile count is relied on. Thomakos 1998 [41]: prospective study of 28 adult surgical patients given one of cefamandole, cefuroxime or cefoxitin for 10 days, quantitative stool yeast cultures before, at the end of and 1 week after treatment; all three increased fecal yeast concentrations (cefoxitin the most, about 2.5 log10 CFU/g); the authors call the increases minor; no cefuroxime-specific figure or group size in the abstract. Vacarean-Trandafir 2023 [14]: 128 surgical oncology patients (120 paired samples analyzed), single 1.5 g intravenous cefuroxime before incision (maximum 2 doses in 12 hours), 16S V3-V4 sequencing before and 7 days after, no unexposed control group. Post-treatment samples were less diverse; Firmicutes fell from 52.3% to 45.3% and Proteobacteria rose from 6.2% to 11%; antibiotic explained 3.71% of compositional variation versus 6.13% for the bowel preparation method; in the 38 patients without mechanical bowel preparation the pairwise PERMANOVA was significant (P=0.0039) although the abstract states composition was not significantly changed. Read through a summarizing fetch tool.

Mechanism

Named: suppression of susceptible anaerobes, clostridia and Enterobacteriaceae with overgrowth of Candida and, in vulnerable hosts, C. difficile. The tablet label gives absolute bioavailability of 37% fasting to 52% after food [1], so a large share of an oral dose is not absorbed; that this unabsorbed or biliary drug reaches the colon is an inference from the label and from the high faecal levels in four of 10 volunteers [13], not a measured finding here.

Why red

Red was re-evaluated against its five conditions. (1) Two independent Tier A sources with cefuroxime in the cohort: the MarketScan family [8][9] (counted once) and the Intermountain hospital cohort [10] are different databases, years, settings and case definitions, and a third, the Northern California case-control study [39], is a further separate health system; this is the same logic applied to cefpodoxime in this batch, where a claims case-control (Miller) and a Northern California case-control (Baxter) were counted as two independent datasets. (2) The category is directly measured: stool flora was cultured or sequenced in adults given cefuroxime in four studies [13][14][40][41], and C. difficile infection is the category's own measured clinical endpoint, so the entry is an outcome-plus-composition entry labeled as such and not an outcome posing as a nutrient. (3) Direction is consistent across every adult dataset (odds ratios 2.16 to 9.59, all lower bounds above 1 where intervals are given; flora shifts toward fewer Enterobacteriaceae, clostridia and anaerobes and more yeasts). (4) Size is clinically meaningful (a labeled adverse reaction with a high-risk stratum of 42.65 cases per 10,000 prescriptions) and the mechanism is named. (5) No substantial contradicting Tier A evidence: the one adult study reading no short-term dysbiosis [14] is a single prophylactic intravenous dose without a control arm that nevertheless showed lower diversity, and the infant trial and mouse study are different populations or species [15]. The grade attributes a disturbance and infection risk to the drug; it is not a deficiency claim.

Counter-evidence held on record

Boundary and next step Applies to adults and to the infection and flora-disturbance findings in the datasets named (30 to 90 day claims windows; 60 days in the hospital cohort and the Northern California study; 4.5 to 10 day courses in the flora studies). It does not cover infants exposed through maternal prophylaxis and is not a nutrient shortfall. Advance to only after an editor opens the full texts to confirm the cefuroxime rows, the Baxter interval, the Webb supplementary counts and the Novelli per-arm counts. customerVisible is never assigned here. The labels direct that diarrhea after antibiotic use be evaluated for C. difficile. Framing caution This is an infection and disturbance risk, not a nutrient deficiency. No probiotic or supplement recommendation is made or implied.

Limits on what this shows

• Negative or null findings: No study of cefuroxime measuring arterial oxygenation, red-cell mass or tissue oxygen delivery was found. A PubMed listing for cefuroxime with oxidative terms and the listings for iron and ATP returned no hemoglobin outcome study [24][26][29].
• At-risk subgroups: Not defined by the evidence beyond the case report (a child) and hospitalized patients receiving the parenteral product.
• Counter-evidence on record: The label figure has no comparator and may reflect the underlying illness and treatment setting rather than the drug; the tablet labels do not repeat it. Several other cephalosporin hemolysis case reports opened during the search concerned ceftizoxime, cefoperazone/sulbactam, cefixime and cefazolin, not cefuroxime, and were not used (cefuroxime was not the suspect drug in those reports).

Direction

Adverse drug reaction and laboratory change (red-cell destruction in the immune cases; cause of the common hemoglobin drop not stated). It is not a nutrient deficiency, not iron deficiency and not a drug-intended effect.

Regulatory status

Zinacef US label (2013 copy, read twice through a summarizing fetch tool) and a generic cefuroxime for injection label (revised May 2021): "A decrease in hemoglobin and hematocrit has been observed in 1 in 10 patients"; transient eosinophilia in 1 in 14; positive Coombs' test "each occurred in fewer than 1 in 250 patients" under hypersensitivity reactions; hemolytic anemia appears in the cephalosporin-class adverse reaction list without an incidence. The label does not give the trial population, route or comparator for these figures. Tablet labels: "Positive Coombs' test" among adverse reactions (less than 1%), and hemolytic anemia, leukopenia, pancytopenia and thrombocytopenia in postmarketing reports; hemoglobin and hematocrit are not listed. UK SmPC: positive Coombs test uncommon, haemolytic anaemia "not known" in the table and "very rarely" in the text, with a note that a positive Coombs test can interfere with cross-matching. [1][2][3][4][5]

Population evidence

No cohort or trial quantifying hemoglobin change or hemolysis incidence with cefuroxime against a comparator was found. Tier B case report: Almeida 2025 [19], a 10-year-old girl with immune hemolytic anemia after cefuroxime and ceftriaxone for a parotid lymphangioma infection; hemoglobin fell from 12.9 to 2.6 g/dL by day 4 with hypovolemic shock and transient renal dysfunction, the direct antiglobulin test was positive for IgG and C3d, her serum reacted with cefuroxime-treated group O red cells by the indirect antiglobulin test (drug-dependent antibody), and hemoglobin was 9.1 g/dL within 48 hours of stopping both drugs; abstract level, read through a summarizing fetch tool. The authors held both drugs responsible, so the contribution of cefuroxime alone is not separable.

Mechanism

Named in outline: drug-dependent IgG and complement antibody binding to cefuroxime-coated red cells. The mechanism of the common 1-in-10 hemoglobin decrease is not given in the label text read; competing explanations are illness, hemodilution from intravenous fluids, surgery and blood sampling.

Counter-evidence held on record

Boundary and why not higher Orange would need the hemoglobin decrease quantified against a comparator, or an incidence for immune hemolysis with cefuroxime; the label figure of 1 in 10 is a Tier A statement about this drug but carries no denominator, magnitude or comparator, and anemia is an outcome here and never filed as a nutrient finding. Oxygen delivery is kept independent of ATP production and of iron (Hard Rule 7). Framing caution A laboratory and allergic-type adverse reaction flag for clinicians. It says nothing about iron, B12 or intake, and no supplement is indicated.

Limits on what this shows

• Why no grade: Dual condition met: no Tier B or Tier C source measuring ATP, respiration or energy output with cefuroxime was opened, and no circulating claim about cefuroxime and energy was found. The mitochondrial swelling in rabbit corneal endothelium [25] is a structural local finding and is not ATP evidence; it is graded under mitochondrial burden and kept separate (Hard Rule 7).

Direction

Not determined.

Channels tried

PubMed listing "cefuroxime AND ATP" (16 results, 10 shown): bacterial genomics (ABC transporter ATP-binding proteins), an extracellular-ATP signaling study of shock-wave drug delivery, a macrolide-digoxin cohort using cefuroxime as a comparator, ATP bioluminescence used as a bacterial viability assay, and a Streptomyces enzyme study; none measured ATP or respiration in cells exposed to cefuroxime [26]. PubMed listing "cefuroxime AND mitochondria" (2 results; see the mitochondrial entry) [25]. Repair round: PubMed "cefuroxime oxidative phosphorylation" returned 0 results (processed as cefuroxime AND oxidative AND phosphorylation) [45]; a "cefuroxime mitochondrial respiration ATP" and a "cefuroxime mitochondrial membrane potential" listing were refused with HTTP 429 and are not counted. Labels (no energy or mitochondrial text) [1][2][3][4]; Drugs.com monograph and Wikipedia page (no energy claim) [32][7].

Genuine gap or tooling

Probable genuine gap with some tooling limitation: 6 of 16 PubMed records of the first ATP listing were not seen; the oxidative phosphorylation phrasing returned nothing; the membrane potential phrasing was refused; the broader 2013 bactericidal-antibiotic mitochondria study could not be opened, so no class evidence was read; WebSearch was exhausted.

Counter-evidence held on record

Re-check Re-check page 2 of the ATP listing and a membrane potential or bioenergetics phrasing, and open the class mitochondrial literature to see whether cefuroxime was among the beta-lactams tested. Framing caution This is a disclosed gap, not a finding of no effect. The claim that antibiotics lower energy has no cefuroxime-specific support.