Propylthiouracil

Limits on what this shows

• Key negative/null findings: None identified at lower or clinically typical human-equivalent exposure; the cited study used short-term, high-dose rat administration, so the magnitude at ordinary therapeutic human dosing is unknown, and this review found no second, independent confirmatory study.
• At-risk subgroups: None established from this single animal study; PTU hepatotoxicity in general is idiosyncratic and can occur at any dose or time, more often within the first 6 months of therapy per label pharmacovigilance experience, though the mitochondrial-structure finding itself comes only from this one 1981 animal study.
• Counter-evidence on record: No study was found showing preserved human liver mitochondrial ultrastructure on PTU; this is an absence of a negative human study, not evidence of safety, and the 1981 finding has not been independently replicated.

Regulatory status

Not addressed directly, beyond the label's general boxed hepatotoxicity warning (liver injury, failure, transplantation, or death, rare and unpredictable onset).

Population evidence

Aguas AP. Giant mitochondria in rat liver cells after short-term administration of propylthiouracil. J Submicrosc Cytol. 1981;13(4). PMID 7334542. Rat liver, short-term PTU administration; ultrastructural (electron microscopy) study describing giant mitochondria with decreased cristae number and intracristal filaments. Retrieved via a PubMed search-listing summary; the individual abstract page was not independently opened.

Mechanism

Direct structural mitochondrial injury in hepatocytes, proposed as part of the broader PTU hepatotoxicity picture underlying the drug's boxed warning.

Counter-evidence held on record

This is a single, decades-old, short-term high-dose animal study. It supports biological plausibility for PTU's already-labeled hepatotoxicity, not a claim that routine therapeutic dosing damages mitochondria in most patients, and it does not support mitochondrial-support supplementation.

Limits on what this shows

• Key negative/null findings: No study was located measuring oxidative-stress markers in PTU-treated patients who are fully biochemically euthyroid, versus untreated hyperthyroid patients, which is the comparison that would isolate a drug-specific effect from the disease effect.
• At-risk subgroups: Hyperthyroid patients not yet biochemically controlled on PTU; the Seven 1998 finding tracks disease control status, not drug exposure duration, as best this review could determine from the retrieved abstract.
• Counter-evidence on record: The Seven 1998 comparator group is healthy controls, not an untreated-versus-treated hyperthyroid comparison, so Hard Rule 6 (confounding by indication) is presumed, not excluded. Two rat studies that used PTU specifically as the method to induce experimental hypothyroidism and then tested a candidate protective compound (aminoguanidine: Rastegar-Moghaddam et al., Adv Biomed Res 2025, DOI 10.4103/abr.abr_538_24; quercetin: Lebda et al., Al-Azhar Med J 2018, DOI 10.12816/0052263) found oxidative stress in the resulting hypothyroid state, which reflects the hypothyroid state PTU produces in that model, not necessarily an independent chemical toxicity of PTU distinct from thyroid status (see Overlap map).

Regulatory status

Not addressed in the label.

Population evidence

Seven A, Taşan E, İnci F, Hatemi H, Burçak G. Biochemical evaluation of oxidative stress in propylthiouracil treated hyperthyroid patients. Effects of vitamin C supplementation. Clin Chem Lab Med. 1998;36(10):767-770 (DOI 10.1515/cclm.1998.136). Human study: hyperthyroid patients receiving PTU therapy versus healthy controls showed higher lipid peroxidation and altered antioxidant-enzyme activity; one month of vitamin C 1000 mg/day raised glutathione and glutathione peroxidase and lowered lipid-peroxidation markers in both groups. Sample size and confidence intervals were not available in the retrieved abstract metadata; the publisher page (De Gruyter) returned an HTTP 405 error on direct fetch, so only the Crossref-hosted abstract, not the full text, was reviewed.

Mechanism

No PTU-specific mechanism is named for the human finding beyond the general thyrotoxic state raising metabolic/oxidative load. Separately, PTU's thiourea/sulfhydryl structure has documented radical-scavenging chemistry: Bagnato G, Bitto A, et al. Propylthiouracil modulates aortic vasculopathy in the oxidative stress model of systemic sclerosis. Vascul Pharmacol. 2015 (DOI 10.1016/j.vph.2014.12.006) — an animal study using PTU experimentally (off-label, non-thyroid indication) that found it lowered oxidative markers in that model.

Counter-evidence held on record

Do not read this as "PTU depletes antioxidants." No validated tissue-level or functional marker has been shown to move specifically because of PTU independent of thyroid status. The vitamin C dose used in the cited research, 1000 mg/day, is a study dose, not a recommendation, and is below the NIH Office of Dietary Supplements tolerable upper intake level of 2,000 mg/day for adults; no dosing recommendation is made here because this entry is graded yellow.

Limits on what this shows

• Key negative/null findings: Apoptosis rates were unchanged across all three copper diets. No study was located measuring serum or hepatic copper concentration in PTU-treated animals or humans maintained on a normal, unmanipulated diet, so there is no evidence PTU itself changes copper status under ordinary dietary conditions.
• At-risk subgroups: None identified in humans. This is a rodent carcinogenesis-bioassay finding using an initiator-promoter protocol and PTU dosing/duration designed to maximize thyroid tumor yield, not ordinary therapeutic human exposure.
• Counter-evidence on record: No source was found contradicting the direction shown, but per Hard Rule 4's direction check, the documented relationship is copper intake affecting PTU's carcinogenic action, not PTU affecting the animal's copper economy; this entry is retained under Copper because it is the only genuine PTU-specific literature located connecting the drug to copper-dependent biology at any tier, not because it shows a depletion or excess mechanism.

Regulatory status

Not addressed in the label.

Population evidence

Shima T, Nishimura J, Dewa Y, Saegusa Y, Matsumoto S, Kawai M, Harada T, Mitsumori K, Shibutani M. Modification of dietary copper levels on the early stage of tumor-promotion with propylthiouracil in a rat two-stage thyroid carcinogenesis model. Chem Biol Interact. 2009;180(2):262-270 (DOI 10.1016/j.cbi.2009.02.002). Male F344 rats initiated with N-bis(2-hydroxypropyl)nitrosamine and then promoted with PTU while fed copper-deficient (0.6 ppm), normal (6 ppm), or copper-supplemented (60 ppm) diets. Both the deficient and supplemented diets reduced preneoplastic focal follicular-cell hyperplasia multiplicity and Ki-67-positive proliferating cells versus the normal-copper diet, with corresponding downregulation of cell-cycle genes and decreased ceruloplasmin and metallothionein-1/2 expression at both copper extremes. Retrieved via the publisher's Elsevier ScienceDirect abstract page.

Mechanism

The study authors propose an altered cellular redox-balance mechanism operating through copper-dependent proteins (ceruloplasmin, metallothionein), not a change in systemic thyroid hormone levels. The causal direction is dietary copper intake modulating PTU's tumor-promoting potency, the reverse of a drug-depletes-nutrient story.

Counter-evidence held on record

This is not evidence that PTU changes a patient's copper status, and it does not support copper supplementation or restriction during PTU therapy. It is an animal cancer-bioassay finding about how dietary copper modulates PTU's own tumor-promoting potency through copper-dependent proteins, not a nutrient-depletion signal, and it should not be described to a reader as "PTU affects copper levels."

Limits on what this shows

• Key negative/null findings: Plasma iron itself did not change significantly (P=.143). Inflammatory markers were unchanged, which argues against a generic acute-phase-reaction confound even though the iron-marker pattern resembles one. The ferritin fall with treatment is the reversal of an elevation already present in untreated hyperthyroidism, not a treatment-induced deficiency signal.
• At-risk subgroups: None specific to PTU could be identified; the 2 PTU-treated patients were not reported as a separate subgroup, so this entry cannot bound the effect to PTU as opposed to carbimazole (methimazole's prodrug, an imidazole-class antithyroid drug chemically distinct from PTU's thiouracil ring).
• Counter-evidence on record: Because 29 of 31 patients received carbimazole and no PTU-only subgroup analysis was reported, this study cannot establish that PTU itself, through any thiouracil-specific chemistry, drives these iron-marker changes rather than the thyroid-hormone-normalization pathway shared by any effective antithyroid treatment (Hard Rule 6, confounding by indication/treatment-class, is not excluded for PTU specifically).

Regulatory status

Not addressed in the label.

Population evidence

Fischli S, von Wyl V, Trummler M, Konrad D, Wueest S, Ruefer A, Heering K, Streuli R, Steuer C, Bernasconi L, Recher M, Henzen C. Iron metabolism in patients with Graves' hyperthyroidism. Clin Endocrinol (Oxf). 2017 (PMID 28833367). Prospective cohort of 31 patients with untreated Graves' hyperthyroidism (22 female, 9 male, mean age 48), treated with carbimazole (n=29) or propylthiouracil (n=2) until normal TSH, free T3, and free T4 were sustained for at least 4 weeks. Median ferritin fell from 156 to 80 µg/L (P=.015); transferrin rose from 2.3 to 2.6 g/L (P=.002); hepcidin (mass spectrometry) fell from 9.0 to 5.9 nmol/L (P=.011, significant in men only); plasma iron rose from 16.8 to 19.1 µmol/L (not significant, P=.143). In a companion cell-culture experiment in the same paper, T3 directly induced hepcidin production. Full text retrieved via an open-access archived copy of the manuscript (Zurich Open Repository, ZORA).

Mechanism

Thyroid hormone (T3) directly induces hepatic hepcidin production, shown in vitro in the same paper; as hyperthyroidism resolves with treatment, falling T3 lowers hepcidin, shifting iron-handling proteins in a pattern the authors describe as resembling reversal of an acute-phase-like state, not classical iron deficiency or iron overload. This is a hormonal/disease-state mechanism common to correcting hyperthyroidism by any effective antithyroid drug, not a PTU-chemistry-specific mechanism.

Counter-evidence held on record

Do not present this as "PTU depletes iron" or use it to justify iron supplementation. Ferritin falling with successful treatment is not iron deficiency here; ferritin also behaves as an acute-phase reactant that is elevated by untreated hyperthyroidism itself and falls as thyroid function normalizes, and plasma iron itself did not significantly change. This is a hormonally mediated effect of treating the underlying disease, common to antithyroid therapy in general, and this small, unstratified PTU subgroup (2 of 31 patients) cannot itself confirm the pattern applies to PTU specifically as opposed to carbimazole.

Limits on what this shows

• Key negative/null findings: The liver-injury correlation analysis was reported across the combined 40-patient treated cohort rather than separately powered within the PTU-only n=20 subgroup, so exact correlation coefficients specific to PTU alone were not independently reported. PTU's dysbiosis index and Firmicutes/Bacteroidetes shift were smaller than MMI's, a quantified difference in magnitude between the two drugs, not a uniform class effect.
• At-risk subgroups: Patients in the first month of PTU initiation, the study's own sampling window. Whether dysbiosis persists, resolves, or changes beyond four weeks of treatment was not addressed by this study.
• Counter-evidence on record: PTU's smaller dysbiosis index and smaller Firmicutes/Bacteroidetes shift relative to methimazole in the same cohort is itself evidence bounding the size of the PTU-specific effect; this is not a severe or class-uniform finding, and Hard Rule 6 confounding by indication is only partly excluded, since both drug arms were compared to each other and to healthy controls but not to an untreated-hyperthyroid comparator.

Regulatory status

Not addressed in the PTU label; the label's hepatotoxicity boxed warning does not name a microbiome mechanism.

Population evidence

Sun J, Zhao F, Lin B, Feng J, Wu X, Liu Y, Zhao L, Zhu B, Wei Y. Gut microbiota participates in antithyroid drug induced liver injury through the lipopolysaccharide related signaling pathway. Front Pharmacol. 2020;11:598170 (DOI 10.3389/fphar.2020.598170). Human arm: 40 Graves' disease patients randomized to methimazole (MMI, n=20) or propylthiouracil (PTU, n=20), plus 50 healthy controls; stool and serum sampled at baseline and biweekly through the first treatment month. PTU patients showed a distinct taxonomic shift from MMI (higher Bacteroidetes proportion versus MMI's higher Firmicutes; genus-level enrichment of Bacteroides, Lachnospiraceae_NK4A136_group, Lachnoclostridium, Eubacterium_rectale, and Dorea), with alpha diversity rising by week 2 then declining by week 4. Taxa enriched with antithyroid treatment correlated positively with ALT, AST, and AKP across the combined 40-patient treated cohort. PTU patients had a lower microbial dysbiosis index and a smaller Firmicutes/Bacteroidetes shift than MMI patients. Animal arm: 45 female Sprague-Dawley rats in six groups including PTU 7.5 mg/kg/day and 12.5 mg/kg/day versus MMI and vehicle groups; PTU rat groups showed enrichment of Spirochaetae, Clostridium_sensu_stricto_1, Blautia, and Treponema_2, intestinal barrier damage, and higher fecal/serum LPS, with ALT rising more in PTU rat groups than MMI rat groups (p=0.033), consistent with a hepatocellular injury pattern.

Mechanism

Intestinal endotoxemia: antithyroid-drug-associated dysbiosis increases LPS-producing taxa (Klebsiella, Escherichia-Shigella, enriched more with MMI than PTU in this study), gut barrier integrity is disrupted (shown directly in the animal arm), fecal and serum LPS rise, and LPS-related signaling correlates with hepatocellular injury markers. This is named as a proposed causal chain by the study authors, not confirmed by a knockout or blocking experiment.

Counter-evidence held on record

This does not establish that gut dysbiosis on PTU causes the drug's own labeled hepatotoxicity risk, which the label describes as rare, idiosyncratic, and unpredictable in onset; no interventional study was located testing whether a probiotic, prebiotic, or dietary change reduces that risk on PTU, and no such product or dose is recommended here. This is a correlational, first-month, observational finding, not a demonstrated causal driver of clinical liver injury in an individual patient.

Limits on what this shows

• Key negative/null findings: The normocytic, normochromic anemia pattern described in the one disease-level source located is not the microcytic, hypochromic pattern typical of iron-deficiency anemia, which argues against an iron-depletion mechanism specifically, though this source did not measure iron indices directly (see Iron entry and Overlap map).
• At-risk subgroups: Patients over-titrated into iatrogenic hypothyroidism; patients with coexisting autoimmune conditions (e.g., pernicious anemia co-occurring with Graves' disease) where anemia has a separate, unrelated cause.
• Counter-evidence on record: Untreated hyperthyroidism is also associated with anemia in the same disease-level source (via diluted plasma volume and impaired iron incorporation despite normal or increased erythropoiesis), so being on antithyroid treatment does not predict a single direction for hemoglobin; direction depends on whether the patient ends up over-, under-, or adequately treated, which this review could not quantify for PTU specifically.

Regulatory status

Not addressed in the PTU label as a hematologic/oxygen-delivery mechanism; the label's only hematologic warning is the separately mechanistic, immune-mediated agranulocytosis risk (see Monitoring panel), which is a white-cell, not a red-cell, effect.

Population evidence

No propylthiouracil-specific interventional or observational study measuring hemoglobin, hematocrit, or red-cell mass was located despite multiple search phrasings. Disease-level (not drug-specific): a retrospective study of 145 patients with thyroid dysfunction (Blatt, Clin Pathol Res J, 2022) found normocytic, normochromic anemia in both directions of thyroid disease, more often with hypothyroidism (76.3% of anemic patients) than hyperthyroidism (29.7%); this study did not stratify by antithyroid drug and propylthiouracil was not confirmed to be in its cohort.

Mechanism

Thyroid hormone stimulates erythropoietin production and erythroid progenitor proliferation; hypothyroidism, whether disease-driven or from antithyroid-drug overtreatment, lowers erythropoietin drive and tissue oxygen demand, reducing red-cell mass. This is the same mechanism class named in Hard Rule 2's beta-blocker example: an intended or foreseeable pharmacologic effect of correcting a hypermetabolic state, not a nutrient deficiency.

Counter-evidence held on record

This is not a classical nutrient-depletion signal, and iron or B12 supplementation is not the fix if anemia occurs on PTU. The correction is dose titration of the antithyroid drug itself, guided by TSH and free T4, not a nutrient intervention. Do not present this as "PTU depletes iron" or "PTU depletes B12."

Limits on what this shows

• Key negative/null findings: Whether IF1 induction lowers net cellular ATP output in vivo, as opposed to only regulating ATP synthase reversal under stress, was not directly measured in the retrieved summary; no human ATP-output data for PTU were located.
• At-risk subgroups: None established in humans; this is an animal-only finding.
• Counter-evidence on record: The same mechanism was associated with a protective effect against a separate hepatotoxin in this study, which cuts against a simple "PTU starves cells of ATP" narrative.

Regulatory status

Not addressed in the label.

Population evidence

Banerjee A, et al. Induction of an ATPase inhibitor protein by propylthiouracil and protection against paracetamol (acetaminophen) hepatotoxicity in the rat. Br J Pharmacol. 1998. PMID 9720771. Rat model; PTU pretreatment induced the mitochondrial ATPase-inhibitor protein and was associated with protection against paracetamol-induced hepatotoxicity. Retrieved via a PubMed search-listing summary; the individual abstract page was not independently opened.

Mechanism

IF1 (ATPase inhibitor protein) regulates the reversal of ATP synthase activity under low mitochondrial membrane potential, preventing futile ATP hydrolysis; its induction by PTU is described as a regulatory response, not necessarily a destructive mitochondrial effect.

Counter-evidence held on record

Do not present this as evidence that PTU causes fatigue or "low energy" through ATP depletion. No human ATP-output or fatigue-outcome study tied to this mechanism was found, and the same study found a protective, not purely harmful, downstream effect.