Clarithromycin

Limits on what this shows

• Null findings: No human measurement (muscle, blood cells, lactate, respirometry) after clarithromycin was found in the PubMed listing of 10 records [20]. The cell and organelle concentrations were not compared with the clinical plasma peak, which this review could not retrieve for the parent drug (the label read gave only the 14-OH metabolite peak of about 0.6 mcg/mL on 250 mg every 12 hours) [1].

Direction

Possible mitochondrial electron-transport inhibition (mechanistic plausibility). Not a nutrient deficiency. Clinical hepatotoxicity is a drug toxicity and is not filed here (Hard Rule 2).

Regulatory status

The label has a Hepatotoxicity warning ("Discontinue clarithromycin immediately if signs and symptoms of hepatitis occur"; hepatic failure with fatal outcome reported, usually with serious underlying disease or concomitant medications) and mentions lactic acidosis only in the verapamil interaction row; it does not describe mitochondrial effects [1].

Population evidence (Tier B, in vitro and organelle; abstracts and one full text)

Woodhead 2019 [16]: HepG2 cells, Seahorse assay; clarithromycin decreased basal oxygen consumption after 24 hours but not after 1 hour, fit as "saturable ETC inhibition at low concentrations and complete ETC inhibition at higher concentrations"; median intracellular-to-medium ratio 15.2; in the DILIsym liver-injury model (285 simulated people) the largest fall in simulated ALT elevations came when electron-transport inhibition was omitted, and the authors conclude clarithromycin toxicity is caused mainly by mitochondrial electron-transport inhibition (full text via a fetch tool; concentrations, IC50 values and ATP not in the text read). Salimi 2015 [17]: isolated rat heart mitochondria exposed to erythromycin, azithromycin or clarithromycin showed ROS formation, mitochondrial membrane permeabilization, swelling and cytochrome c release (reconstructed abstract; concentrations not retrieved). Yamamoto 2024 [18]: clarithromycin reduced oxygen consumption rate in NIPSNAP1/2 double-knockout cells in a dose-dependent manner and the authors conclude it suppresses IL-8 through mitochondrial quality control (reconstructed abstract; no numbers). Kuretu 2024 [19]: C2C12 muscle cells exposed for 24 hours (cytotoxicity 25 to 400 micromolar, functional assays 25 to 100 micromolar) to six drugs including clarithromycin; mitochondrial membrane potential increased significantly at all clarithromycin concentrations, GLUT4 and Akt expression fell at all concentrations, and tafazzin expression fell (significant only at 100 micromolar) (full text via a fetch tool; significance flags only, no numbers).

Supporting listing lines (not opened)

PubMed listing [20]: Qiao 2006 (N-demethyl-clarithromycin, a metabolite, triggers mitochondria- and caspase-linked apoptosis in HeLa cells, snippet only) and Mathis 2005 (clarithromycin reduced vesicle formation in the helminth Echinococcus multilocularis with mitochondrial alterations on electron microscopy, snippet only). Used as pointers, not as evidence.

Mechanism

Proposed: the mitochondrial large ribosomal subunit resembles the bacterial 50S ribosome, so a 50S-binding macrolide may impair mitochondrial translation; this is stated for macrolides generally and was not tested for clarithromycin in the sources opened. Direct electron-transport inhibition is the mechanism favored by Woodhead [16].

Counter-evidence held on record

Counter-evidence and confounding Cell-line and organelle exposures do not establish tissue exposure in people, and two of the studies used cancer, embryonic-kidney-like or immortalized lines. Directions differ between assays (see ATP production). Framing caution Do not state that clarithromycin damages mitochondria in people or causes fatigue through energy failure. This is a laboratory hypothesis, graded for mechanistic plausibility only.

Limits on what this shows

• Null findings: Plasma malondialdehyde unchanged [14]. In a PubMed listing of 43 records for "clarithromycin oxidative stress" the first 10 contained one human study (Larsen) and otherwise algae, cyanobacteria, mycobacteria and H. pylori work; 33 records were not read [15]. No human antioxidant-capacity or glutathione measurement with clarithromycin was found.
• At-risk subgroups: None established.

Direction

Excess oxidative modification of nucleic acids (pro-oxidant direction). It is not a measure of vitamin C, vitamin E, glutathione or any antioxidant nutrient, and no supplement is implied (Hard Rule 5).

Regulatory status

The label sections read contain no oxidative-stress statement; hepatotoxicity is a warning and is a drug toxicity [1].

Population evidence (Tier A)

Larsen 2017 [14]: data from two randomized, placebo-controlled trials in healthy adults. Participants received one week of clarithromycin, trimethoprim, phenoxymethylpenicillin or placebo; 24-hour urinary 8-oxo-7,8-dihydro-2'-deoxyguanosine (8-oxodG, DNA oxidation) and 8-oxo-7,8-dihydroguanosine (8-oxoGuo, RNA oxidation) and plasma malondialdehyde were measured before and after treatment. Clarithromycin increased 8-oxodG by 22.0% (95% CI 3.6 to 40.4%) and 8-oxoGuo by 14.9% (95% CI 3.7 to 26.1%); trimethoprim lowered 8-oxodG by 21.7% (5.8 to 37.6%) with no change in 8-oxoGuo; penicillin V changed neither; no drug changed plasma malondialdehyde. The same result was read on three channels (OpenAlex works/pmid record, a PubMed uid record and the journal-society page). Participant numbers per arm, the clarithromycin dose and the baseline values were NOT retrieved: PMC (PMC5510073) returned a bot-verification page and the Wiley page did not load.

Supporting Tier B

Woodhead 2019 [16]: in HepG2 cells all five macrolides tested, clarithromycin included, increased reactive oxygen and nitrogen species in a concentration-dependent manner after 24 hours but not after 1 hour (DHE probe; no numbers in the text read). Salimi 2015 [17]: clarithromycin, erythromycin and azithromycin induced ROS formation in isolated rat heart mitochondria (reconstructed abstract; concentrations not retrieved). Kuretu 2024 [19]: in C2C12 muscle cells (25 to 100 micromolar, 24 hours) clarithromycin raised malondialdehyde at all concentrations and raised ROS, with the 50 micromolar ROS result not significant (full text via a fetch tool; no numbers).

Context, not counted

Götz 1997 [31]: in 63 patients given omeprazole or ranitidine plus clarithromycin and metronidazole, Mn superoxide dismutase fell and Cu/Zn superoxide dismutase rose in gastric mucosa 8 weeks after successful eradication. This is the effect of clearing the infection, in regimens that contain clarithromycin in both arms, and says nothing about the drug alone.

Mechanism

Proposed, not shown in people: mitochondrial electron transport inhibition raising ROS in cell and organelle models [16][17][19]. Larsen state only that the finding may relate to clarithromycin adverse effects.

Why orange, not red

One human trial report supplies the direct measurement, with a bounded exposure (one week), a bounded population (healthy adults) and a functional marker (urinary oxidised nucleoside excretion, a rise of 15 to 22%) rather than a measured antioxidant deficit; no second independent human study of clarithromycin and an oxidative marker was found, and plasma malondialdehyde did not move.

Counter-evidence held on record

Counter-evidence and confounding Per-arm sample size, dose and absolute values were not retrieved, so the clinical weight of a 15 to 22% shift in a urinary marker cannot be judged here. The in vitro supporting data used concentrations that were not compared with plasma levels. Infection and inflammation themselves change these markers; the trials were in healthy volunteers, which avoids that confound but does not tell what happens during illness. Framing caution Do not say clarithromycin depletes antioxidants or damages DNA in patients. The finding is a short-term rise in urinary markers of nucleic-acid oxidation in healthy adults. No antioxidant supplement is implied.

Limits on what this shows

• Null findings: No study of clarithromycin alone on iron absorption, ferritin or hemoglobin was found; the PubMed listing "clarithromycin anemia hemoglobin" [25] returned H. pylori and iron-deficiency papers and one mouse report. The PubMed listing "clarithromycin copper" returned the Pharmazie record as its only iron-related item [29]. A PubMed listing "clarithromycin ferritin" (17 records, first 10 read) held only H. pylori eradication papers and unrelated infection reports, with none flagged for an effect of clarithromycin itself on iron status, absorption or hemoglobin; snippets included Tseng 2019 (persistent iron deficiency anemia 24.3% in treated vs 26.5% in untreated groups, P = 0.71) and Wallace 2004 (no ferritin difference between eradicated and persistently positive adults), read at listing-line level only [39].

Direction

Reverse or neutral for the drug: regimens that include clarithromycin are associated with improved iron stores because they treat the infection. Indication-linked deficiency, a context flag (Hard Rule 6); not depletion by clarithromycin and not merged with Copper (Hard Rule 7).

Regulatory status

No iron statement in the label sections read [1].

Population evidence (Tier C, regimen level; clarithromycin not separable)

Hudak 2017 [27]: systematic review and meta-analysis; H. pylori infection vs none, 14 observational studies for iron deficiency anemia pooled OR 1.72 (95% CI 1.23 to 2.42), 30 studies for iron deficiency OR 1.33 (1.15 to 1.54), 23 for anemia OR 1.15 (1.00 to 1.32); across 7 randomized trials eradication therapy increased ferritin (standardized mean difference 0.53, 95% CI 0.21 to 0.85) but did not significantly change hemoglobin (0.36, -0.07 to 0.78); significant heterogeneity and publication bias (OpenAlex reconstructed abstract). Which regimens the 7 trials used was not retrieved. Sarker 2008 [28]: randomized trial in Bangladeshi children whose indexing keywords name clarithromycin, amoxicillin and omeprazole with ferrous sulfate and hemoglobin, ferritin and soluble transferrin receptor outcomes (metadata only; abstract and results not retrieved).

CLASS EXTRAPOLATION / regimen level

Cohort test: clarithromycin is named in the Sarker trial keywords only; Hudak's trial regimens were not retrieved, so for this drug the iron data are regimen-level (Tier C), capped at yellow, and they run toward better iron status.

Mechanism

The mechanism of the infection effect was not retrieved in the sources read, and none is established for the drug. In vitro, a Pharmazie 2003 study of roxithromycin, clarithromycin and azithromycin with Cu(II), Zn(II) and Fe(II) exists, but its record has no abstract, so whether it found iron binding is unknown [30].

Counter-evidence held on record

Counter-evidence and confounding Confounding by indication is presumed: the same patients have a gastric infection that independently lowers iron. Eradication regimens differ in co-drugs, and the pooled hemoglobin effect was not significant. Framing caution Do not state that clarithromycin lowers iron or causes iron deficiency, and do not suggest iron supplementation because of clarithromycin. Iron status in H. pylori infection is a clinician question.

Limits on what this shows

• Why no grade: Dual condition met on the evidence actually read: no Tier B or Tier C result linking clarithromycin to a copper change was retrieved (the one in vitro record has no readable result), and no circulating consumer claim about clarithromycin and copper was found. Tooling-limited rather than a proven genuine gap: if the Hamdan full text shows clarithromycin binding Cu(II), this should move to yellow. Not a finding of safety.

What was searched

DailyMed label [1] and a second label-derived page [2] (no copper statement); PubMed listing "clarithromycin copper" (10 records, one page) [29]: items were environmental or synthetic, one H. pylori study on gastric superoxide dismutases (Götz 1997 [31]), a 2002 Pakistani complexation paper whose abstract describes erythromycin, not clarithromycin (Sultana 2002 [32]), and Hamdan 2003 [30]; PubMed listing "clarithromycin vitamin deficiency" (13 records, first 10 read; no mineral or vitamin lowered) [33]; consumer pages [34][35].

Channel outcomes

Hamdan 2003 ("Comparative in vitro investigations of the interaction between some macrolides and Cu(II), Zn(II) and Fe(II)", Pharmazie 58:223-224) indexes clarithromycin, roxithromycin and azithromycin with copper, zinc and iron by HPLC, NMR and UV methods, but its PubMed record has no abstract, so whether it found binding is unknown; the OpenAlex record had only metadata. Götz 1997 measured gastric Cu/Zn superoxide dismutase protein after H. pylori eradication, which is an infection effect, not copper status. No full text of the Pharmazie note was located with the tools available. PMC and Europe PMC returned bot-verification or 429 responses, and Crossref and OpenAlex search endpoints returned errors or unrelated results.

Limits on what this shows

• Null and contrary findings: Yap 2016 [10]: 17 H. pylori-positive Malaysian adults on amoxicillin 1,000 mg, clarithromycin 500 mg and pantoprazole 40 mg twice daily for 7 days; no significant alpha-diversity difference between baseline and 6 months after eradication (P > 0.05); Actinobacteria 6.68% at baseline, 4.86% at 6 months and 8.14% at 12 months. Sampling began at 6 months, so it does not test the acute window. In the Elvers volunteer study (d) anaerobes did not change [6]. Liou T14 recovered fully by week 8 [7]. No clarithromycin-specific analysis of short-chain fatty acids or other metabolites was found. PubMed listings "clarithromycin intestinal microbiota diversity" (18 records) and "clarithromycin microbiome randomized" (32 records) surfaced the cohorts above and nothing else with a clarithromycin-only sequencing design in the first 10 records each [12][13].
• At-risk subgroups: People with C. difficile risk factors (label) [1]; children given repeated macrolide courses (Korpela, asthma odds ratio 6.11, 95% CI 1.53 to 26.58 for more than 2 courses in the first 2 years, N = 32 exposed vs 116, an association not a nutrient outcome) [5]; people receiving clarithromycin with metronidazole as concomitant or quadruple therapy, in whom recovery was slower [7].

Direction

Disturbance risk (excess-risk framing). This is not the depletion of a nutrient, and no probiotic or supplement is implied.

Regulatory status

The label states that Clostridium difficile associated diarrhea "may range in severity from mild diarrhea to fatal colitis" and has been reported to occur more than two months after antibacterial use; it also has a "Development of Drug Resistant Bacteria" warning [1]. No label text addresses the microbiome itself.

Population evidence: single-drug healthy volunteers (Tier A, systematic review stratified to clarithromycin)

Elvers 2020 [6]: systematic review of 31 articles of antibiotics commonly used in UK primary care; its abstract states clarithromycin "decreased the populations of Enterobacteria" and reduced Bifidobacterium and Lactobacillus in numbers and diversity for up to 5 weeks. The text describes four clarithromycin studies, all culture-based healthy-volunteer studies. (a) 10 volunteers, 500 mg twice daily for 7 days: fewer Enterobacteria and streptococci, lactobacilli, bifidobacteria and Bacteroides suppressed, C. albicans isolated, normal within 28 days. (b) 12 men: E. coli significantly reduced during treatment, 6 subjects colonised by resistant Gram-negative rods, bifidobacteria and Bacteroides significantly reduced, whole flora normal in all by 35 days. (c) 10 volunteers, 250 mg twice daily for 10 days: E. coli significantly reduced, overgrowth of Klebsiella, Citrobacter or Enterobacter in 6, lactobacilli and bifidobacteria significantly reduced for the study duration, normal within 28 days. (d) 6 volunteers: no change in Candida, enterococci or anaerobes, Enterobacteriaceae slightly reduced, baseline by day 21. The review page showed reference numbers only, so the mapping of its four studies to named primary reports was not confirmed; two primary reports that fit the description were found and their abstracts read (next row). This row is a second-hand summary read through a fetch tool, and the review itself reports heterogeneity and small samples.

Population evidence: single-drug primary reports (Tier A, abstracts read)

Edlund 2000 [36]: 12 healthy men aged 24 to 40 in a crossover with a 6-week washout, moxifloxacin 400 mg daily and clarithromycin 500 mg twice daily, each for 7 days, in two 35-day periods. During clarithromycin E. coli decreased significantly; enterococci, Enterobacter, Citrobacter, Klebsiella and Pseudomonas increased markedly; bifidobacteria, lactobacilli and clostridia were suppressed; peptostreptococci, Veillonella, Bacteroides and fusobacteria were unchanged; the flora returned to normal after 35 days in all volunteers (reconstructed abstract; counts and P-values not retrieved). Brismar 1991 [37]: 20 healthy volunteers, 10 per group, clarithromycin 250 mg every 12 hours or erythromycin ethylsuccinate 1,000 mg every 12 hours for 7 days, stool before, during and after; with clarithromycin streptococci and Enterobacteriaceae decreased and the anaerobic flora was also affected, with more disruption in the erythromycin group (reconstructed abstract; counts and recovery times not retrieved). A third, double-blind placebo-controlled volunteer study of azithromycin and clarithromycin on oropharyngeal and bowel flora (Matute 2002 [38]) exists, but its record has no abstract and no result was read. None of these has a placebo arm that was read.

Population evidence: randomized trial of a clarithromycin triple regimen (Tier A, drug in cohort with co-drugs)

Liou 2019 [7]: multicentre open-label trial in Taiwan, 1,620 adults randomized 1:1:1; 1,214 (75%) analysed at 1 year. Triple therapy T14 was lansoprazole 30 mg, amoxicillin 1 g and clarithromycin 500 mg, each twice daily for 14 days; concomitant therapy C10 was the same three drugs plus metronidazole 500 mg for 10 days; the third arm (bismuth quadruple) contained no clarithromycin. Compared with baseline, alpha diversity was significantly reduced 2 weeks after T14 (P = 0.0002) and after C10 (P < 0.0001). After T14 alpha and beta diversity were restored at week 8 (P = 0.14 and 0.918) and at 1 year; after C10 they were not fully recovered at week 8 (P = 0.0001 and 0.013) or at 1 year (P = 0.019 and 0.064). Faecal E. coli resistance to ampicillin-sulbactam rose from 12% to 66% after T14 and from 7% to 64% after C10 at week 2 and returned to baseline by week 8. Read from a PubMed record summarized by the fetch tool; the OpenAlex record had no abstract and the Lancet page did not open. No numeric diversity indices in the abstract.

Population evidence: other eradication cohorts (Tier A, drug in cohort with co-drugs)

Zhou 2020 [8]: 22 penicillin-allergic children with H. pylori and 23 controls; omeprazole, clarithromycin, metronidazole and bismuth for 14 days; stool 16S at weeks 0, 2, 6 and 52. Alpha diversity fell at week 2 (P < 0.05, except Chao1 P = 0.16), beta diversity changed (P = 0.001 and 0.002), some changes persisted at week 6 and the community had returned almost to baseline at 1 year; Bifidobacterium, Faecalibacterium, Blautia and Bacteroides fell and Escherichia-Shigella, Klebsiella, Enterococcus and Streptococcus rose (reconstructed abstract). Wu 2019 [9]: duodenal ulcer patients randomized to PPI, clarithromycin and amoxicillin with or without a probiotic; the abstract states anti-H. pylori treatment "significantly reduces intestinal microbiota diversity" and fungi decreased, with sample size and numeric results not retrieved (reconstructed abstract only). Jakobsson 2010 [11]: 6 adults, 3 treated with metronidazole 400 mg, clarithromycin 250 mg and omeprazole 20 mg twice daily for 7 days; fecal Actinobacteria (Bifidobacterium, Collinsella) fell sharply and Shannon diversity fell; erm(B) rose 3 to 5 orders of magnitude, was still more than 1,000-fold higher at 1 year and detectable at 4 years; the authors state the sample was too small to test significance.

Population evidence: children, macrolides pooled (Tier A for the exposure, clarithromycin not separable)

Korpela 2016 [5]: 142 Finnish children with fecal samples (257 samples) and purchase records. Cohort test, run on the full text: "azithromycin and clarithromycin" were grouped as macrolides, with no counts or analysis by drug. Macrolide use was followed by lower Actinobacteria (Bifidobacterium 4 to 10-fold lower in the first 6 months, "in many cases to below detectability"), about 2-fold more Bacteroides and Proteobacteria, lower bile-salt hydrolase, and macrolide-resistance genes that fell with time since the last course (r = -0.31, P = 0.005); richness "did not reach the level of the control samples even 12-24 months after the course" and Collinsella, Lactobacillus and Anaerostipes stayed reduced up to 2 years. Observational; clarithromycin is in the cohort but its share is unknown, so it supports the direction without isolating the drug.

CLASS EXTRAPOLATION

None used for the grade. The azithromycin and erythromycin trials covered in sibling dossiers (mass-treatment cohorts and the 4-week erythromycin trial) did not include clarithromycin and were not opened for this entry; they contribute nothing here.

Mechanism

Present and named: binding of the bacterial 50S ribosomal subunit (23S rRNA) blocks protein synthesis [4], and commensals with susceptible ribosomes are exposed to a drug that distributes readily into tissues and fluids [1]. The taxa-to-function link is inferred from co-movement in the cohorts, not proven.

Why orange, not red

Clarithromycin clears the independent-source threshold on direction and mechanism, but red is withheld for four reasons. (a) In the single-drug volunteer studies and in the clarithromycin triple arm the effect recovered within about 28 to 56 days [6][7], which is time-bounded in the series sense. (b) In every large cohort clarithromycin was co-administered with amoxicillin or metronidazole and a PPI, so attribution to clarithromycin alone rests on small culture-based volunteer studies (10 to 12 people each) without a placebo arm that could be read [6][36][37]; the one placebo-controlled volunteer study found had no readable result [38]. (c) The persistence evidence (erm(B) at 4 years in 3 people, richness gaps at 12 to 24 months in children) comes from cohorts where clarithromycin is mixed with other drugs [11][5]. (d) One cohort found no alpha-diversity change at 6 months [10], and no cohort linked a change to a measured nutrient or clinical outcome.

Counter-evidence held on record

Counter-evidence and confounding Culture-based volunteer studies miss most taxa; the sequencing cohorts treat H. pylori with other antibiotics and a PPI, and PPIs alone alter the gut community. The pooled-macrolide children cohort has confounding by infection and age. Effect sizes in the sequencing cohorts were mostly P-values in abstracts, with no confidence intervals retrieved. Framing caution This is a disturbance risk during and after a course, not a depleted nutrient and not a reason to take a probiotic or supplement. Do not present it as permanent damage; most cohorts show recovery. Persistent or severe diarrhea is a clinician question because of C. difficile [1].

Limits on what this shows

• Null findings: No human or animal ATP, lactate or respirometry measurement after clarithromycin was found [20].

Direction

Unclear, possibly reduced or increased cellular ATP output in vitro (mechanistic plausibility only). Not a nutrient deficiency and not independent of the Mitochondrial burden entry.

Regulatory status

Not addressed in the label sections read [1].

Population evidence (Tier B, in vitro; one full text, one abstract)

Kuretu 2024 [19]: in C2C12 cells the article states that "rifampicin, simvastatin and clarithromycin significantly increased the concentration of ATP"; the clarithromycin increase was not significant at 25 and 50 micromolar; no percentage change in the text (the abstract, by contrast, says the mitotoxicant drugs reduced mitochondrial ATP synthesis, which conflicts with the full text for clarithromycin). Woodhead 2019 [16]: lower basal oxygen consumption in HepG2 cells at 24 hours, a respiration surrogate; ATP was not reported in the text read.

Mechanism

Proposed: electron-transport inhibition lowering respiration [16]; the opposite ATP direction in C2C12 cells has no named mechanism in the text read.

Counter-evidence held on record

Counter-evidence and confounding Rests on two in vitro papers that disagree in direction; shares its sources with Mitochondrial burden (see the Overlap map). Effects in cultured cells do not transfer to people. Framing caution Do not attribute low energy, fatigue or reduced ATP to clarithromycin from this entry.

Limits on what this shows

• Null findings: PubMed listing "clarithromycin hemolytic anemia" (19 records, first 10 read): none identified clarithromycin as the cause of hemolysis; clarithromycin appeared as treatment or comparator [24]. PubMed listing "clarithromycin anemia hemoglobin" (23 records, first 10 read): the only record attributing a hematologic effect to clarithromycin was the mouse report; the rest were H. pylori and iron-deficiency papers [25]. Pages 2 and 3 of those listings, methemoglobinemia and arterial oxygenation were not searched separately.

Direction

Drug-drug pharmacodynamic enhancement of another drug's hematologic toxicity (mouse). Not an iron deficiency and not an oxygen-carrying nutrient deficiency; anemia is not iron deficiency (Hard Rule 2).

Regulatory status

The label sections read list white-cell and liver-enzyme laboratory abnormalities in the prophylaxis trial (WBC below 1 x 10^9/L in 4% vs 0% placebo) and postmarketing thrombocytopenia; no hemoglobin or anemia statement appeared [1][2]. The label advises separating clarithromycin and zidovudine by at least two hours because simultaneous oral dosing of the immediate-release tablet may lower zidovudine concentrations [1].

Population evidence (Tier B, animal; PubMed record abstract through a fetch tool)

Rao 2003 [26]: NIEHS technical report in Swiss CD-1 mice dosed by gavage (clarithromycin 250 to 1,000 mg/kg, about 1 to 4 times the human dose; zidovudine 200 or 400 mg/kg). Zidovudine alone caused slight hematopoietic toxicity with mild declines in red cell count, hemoglobin and hematocrit; clarithromycin alone at 250 or 500 mg/kg caused no significant toxicity, and at 1,000 mg/kg multi-organ toxicity in a few animals and mild leukocytosis; the combination produced anemia "far more severe" than zidovudine alone, with bone-marrow depletion, reticulocytopenia and macrocytic (microcytic in the highest-dose group) anemia. No numeric blood counts in the abstract.

Mechanism

Not stated in the abstract. The label's zidovudine statement is a pharmacokinetic timing interaction, not an explanation for the mouse finding.

Counter-evidence held on record

Counter-evidence and confounding The mouse doses were 1 to 4 times the human dose by body weight, zidovudine was the anemia-causing agent, and the abstract gives no counts. Relevance to people taking clarithromycin alone is not shown. Framing caution Do not state that clarithromycin lowers hemoglobin or oxygen delivery. The one hematologic signal is an interaction with zidovudine in mice. Anemia in H. pylori patients is covered under Iron.