Tofacitinib

Limits on what this shows

• Key null findings: Membrane potential unchanged [17]. A search for tofacitinib mitochondrial toxicity returned no relevant tofacitinib record (the one cardiomyocyte toxicity-assay paper surfaced was not opened and its drug list is unknown).
• At-risk subgroups: None identified.

Direction

Reverse (protective or restorative in inflamed cells).

Regulatory status

No mitochondrial statement in the label sections retrieved or in the TGA product information read [1][3].

Population evidence

SW982 human synovial cells stimulated with TNF-alpha, tofacitinib 100 nM for 24 hours: proton leak decreased (p<0.0003) versus the stimulated state, while methotrexate increased proton leak (p<0.0001) and impaired oxidative phosphorylation [16]. Psoriatic arthritis fibroblast-like synoviocytes (n=18 patients), tofacitinib 1 uM for 24 hours: basal oxygen consumption rate rose (p=0.0313), maximal respiratory capacity rose (p=0.0156), no change in mitochondrial membrane potential, increased mitophagy and younger mitochondria (MitoTimer p=0.0244), higher PGC-1alpha and NRF2 [17].

Mechanism

Named in the authors' reading: JAK-STAT inhibition lowers inflammatory drive on mitochondria and enhances autophagy and mitochondrial quality control [17]. Evidence is a cell-line and ex vivo effect at single concentrations.

Counter-evidence held on record

Counter-evidence on record Single time points, cell lines or small ex vivo samples, supraphysiologic or unreported-relative-to-plasma concentrations (100 nM and 1 uM, with no comparison to plasma exposure in the retrieved summaries), and no human in vivo mitochondrial measurement. Absence of reported toxicity is not proof of safety. Framing caution This is not a mitochondrial deficiency or damage claim. The reverse-direction cell finding must not be turned into a supplement or energy claim.

Limits on what this shows

• Key null findings: No change in 8-iso-PGF2alpha with tofacitinib [16]. Two cross-sectional RA studies show higher oxidants and lower antioxidant capacity in untreated or mixed RA versus healthy controls (n=112 vs 52, n=60 vs 30) but did not include tofacitinib, so they describe the disease, not the drug [23][24].
• At-risk subgroups: None identified.

Direction

Reverse (antioxidant-like) or neutral. Not a depletion finding.

Regulatory status

None. No antioxidant or oxidative-stress statement in the retrieved label text [1][3][4].

Population evidence

PsA fibroblast-like synoviocytes (n=18): mitochondrial ROS fell from 24.384 +/- 3.583 to 14.606 +/- 1.159 arbitrary fluorescence units per cell (p=0.0156) with tofacitinib 1 uM [17]. SW982 cells: the lipid peroxidation marker 8-iso-PGF2alpha doubled with methotrexate and did not change significantly with tofacitinib [16]. TNF-alpha-stimulated Caco-2 intestinal cells (Exp Ther Med 2024, read via an aggregator summary): authors report that tofacitinib reduced oxidative stress enzyme activity (SOD1, catalase, NADPH), reduced apoptosis and raised claudin-1 [18]; the direction of the individual enzyme readouts was not verifiable from the summary.

Mechanism

Named: reduced inflammatory signaling lowers ROS generation. Direct antioxidant chemistry of tofacitinib was not found.

Counter-evidence held on record

Counter-evidence on record In vitro, single time point, and the readout overlaps the mitochondrial entry. The RA oxidative studies [23][24] are Tier D for the drug question (tofacitinib absent) and are cited only to show the disease baseline. Framing caution Do not market this as an antioxidant benefit of the drug, and do not extend it to a supplement recommendation.

Limits on what this shows

• Key null findings: The registry did not show better anemia correction with JAK inhibitors than with IL-6R or TNF inhibitors [9], which argues against a strong tofacitinib-driven hepcidin fall in practice. No drug-attributable iron deficit was found; none was measured.
• At-risk subgroups: People with active ulcerative colitis, where chronic bleeding and inflammation both produce anemia; patients with baseline anemia (see the oxygen delivery entry).

Direction

Neutral, with a possible reverse-direction mechanism (less hepcidin drive) that is unconfirmed in people. Not a depletion finding.

Regulatory status

The US label and EU SmPC do not mention iron. They direct hemoglobin monitoring only: baseline, 4 to 8 weeks, then every 3 months; avoid initiation below 9 g/dL; interrupt below 8 g/dL or a fall of more than 2 g/dL [1][4].

Population evidence

CorEvitas RA registry (n=2,772; 439 of 481 JAK inhibitor patients on tofacitinib): hemoglobin change at 6 months was +0.44 g/dL on IL-6R inhibitors, +0.12 on TNF inhibitors and -0.09 on JAK inhibitors; anemia normalized in 65%, 43% and 34% respectively. Hepcidin, ferritin and iron were not measured [9]. Single-center UC series (n=22, tofacitinib): hemoglobin unchanged at week 2 (P=0.90); iron and ferritin not measured [11]. Class context, tofacitinib not in cohort: in 93 RA patients tocilizumab lowered serum hepcidin-25 by 82% and TNF inhibitors by 56% at 2 weeks, with hemoglobin +1.4 and +0.7 g/dL [14]. Filgotinib UC trial: about 50% of patients anemic at baseline, no association of filgotinib with anemia versus placebo [22].

Mechanism

Named but indirect. In RA, IL-6 receptor blockade lowers serum hepcidin and improves anemia [14], which identifies the IL-6 to hepcidin axis as one route by which inflammation affects iron handling. In human macrophages tofacitinib blocks IL-10/IL-10R signaling and prevents induction of intracellular HAMP, the hepcidin gene [15]. No in vivo human confirmation was found.

Counter-evidence held on record

Counter-evidence on record Confounding by indication is presumed (Hard Rule 6): anemia in ulcerative colitis and rheumatoid arthritis tracks disease activity and bleeding, and it improves as inflammation settles. Anemia is not iron deficiency (Hard Rule 2). The IL-6 hepcidin data [14] are class-adjacent context for a different drug and cannot be carried onto tofacitinib above yellow. Framing caution Low hemoglobin on tofacitinib should not be read as caused by or as proof of iron deficiency, and it is not a prompt to start iron on its own. Interpret ferritin alongside CRP and bleeding status with a clinician.

Limits on what this shows

• Why no grade: Dual condition of Step 7 met on what could be accessed: no Tier B or C evidence for a yellow grade and no circulating consumer claim to rebut as white. This is not a finding of safety.

What was searched

WebSearch "tofacitinib serum copper OR zinc OR vitamin D levels patients treated JAK inhibitor" returned a clinical trial registry page, an IBD blog on tofacitinib safety and vitamin D, BMC and PMC pages unrelated to copper, and no tofacitinib copper record; the vitamin D blog was opened and does not link the two drugs. PubMed listing "tofacitinib copper" and "tofacitinib AND copper" were refused (HTTP 429) on three attempts spaced several minutes apart. OpenAlex title-and-abstract search for tofacitinib and copper returned 429. Consumer pages (drugs.com monograph and food-interaction page) mention no copper or mineral [26][27]. The label, EU SmPC and TGA text read contain no copper statement, though label sections 7 and 12 were not fully retrievable [1][3][4]. Mechanism screen: no chelation, no enzyme induction, no renal wasting, no copper transporter involvement is described in the pharmacology.

Gap type

Mostly tooling-limited. Search budget exhaustion and PubMed and OpenAlex rate limiting prevented a full literature sweep; a later revision should re-run PubMed and Europe PMC for tofacitinib with copper, ceruloplasmin and cuproptosis.

Limits on what this shows

• Key null findings: No composition or activity change in the ex vivo model [19]. No 16S study in ulcerative colitis patients was found.
• At-risk subgroups: None identified.

Direction

Neutral to reverse (toward normal in a disease cohort). Not a microbiome-harm finding.

Regulatory status

No microbiome statement in the retrieved label text [1][4]. Serious infections are covered by label warnings, not as microbiome effects.

Population evidence

Clin Rheumatol 2025: 10 ankylosing spondylitis patients and 10 matched healthy controls, 16S rRNA sequencing and plasma metabolomics before and after tofacitinib. Patients had lower alpha-diversity (ACE, Chao1); tofacitinib partially normalized it, and Barnesiella, Coprobacter, Lachnospira and Lactobacillus tended to return to control levels (84 LEfSe biomarker taxa), with plasma choline, glycerophospholipid and phenylalanine pathways linked to taxa [20]. Eur J Pharm Sci 2025: ProDigest short-term colon model with microbiota from 3 Crohn's disease donors, tofacitinib for 48 hours: no significant effect on microbiota composition or fermentative and metabolic activity; some donors showed lower CXCL10, MCP-1 and IL-6 [19]. A PubMed listing summary of a 2023 review indicates newer IBD agents including tofacitinib are less likely to cause severe C. difficile infection; that record was seen only as a listing summary and its text was not opened [25].

Mechanism

Absent as a direct microbial effect (the in vitro model showed none). Indirect via mucosal inflammation and bile acid or metabolite changes is speculative.

Counter-evidence held on record

Counter-evidence on record n=10 per group, no abundance data reported in the summary, no stated limitations, and an ankylosing spondylitis population rather than ulcerative colitis [20]. The in vitro model used Crohn's donors, short exposure and no host [19]. Neither study is powered to rule out a clinically relevant shift. Framing caution This entry is not a claim of dysbiosis or of probiotic need.

Limits on what this shows

• Key null findings: Placebo-controlled induction showed fewer anemia events on tofacitinib [7][8]; no dose dependence [7][8]; hemoglobin stable in a small UC cohort [11].
• At-risk subgroups: Patients with baseline hemoglobin below 9 g/dL (label: avoid initiation), active GI bleeding, or concurrent myelosuppressive drugs.

Direction

Mixed. Net rise with disease control in trials; rare fall. Outcome entry, not a nutrient deficiency (Hard Rule 2).

Regulatory status

Hemoglobin monitoring and action thresholds as in the Iron entry [1]. The EU SmPC states tofacitinib "has been associated with decreases in haemoglobin levels" and recommends the same schedule [4].

Population evidence

OCTAVE UC program (conference abstracts, Gastroenterology 2019 and J Crohn's Colitis 2019, same analysis): induction anemia adverse events 22 (2.4%) tofacitinib vs 9 (3.8%) placebo; maintenance incidence 2.91 (placebo), 5.51 (5 mg) and 2.55 (10 mg) per 100 patient-years, no dose dependence; hemoglobin rose through week 52; 5 (0.6%) vs 1 (0.4%) discontinued for hemoglobin decline in induction [7][8]. Pooled RA phase 3 (n=4,271) and extension (n=4,858): modest hemoglobin changes that stabilized over time; notable falls (3 g/dL or more from a baseline of 7 g/dL or lower) in under 1% [6]. RA baseline anemia (hemoglobin below 10 g/dL, 152 of 4,736 tofacitinib patients, 3.2%): mean hemoglobin +1.25 g/dL by month 6 vs +0.15 in non-anemic patients; the placebo arm converged to a similar anemia proportion by month 6 (about 29 to 31%) [10]. Registry JAK inhibitor change -0.09 g/dL [9]. UC single center n=22: no change [11]. Case report: transfusion-dependent autoimmune hemolytic anemia after starting tofacitinib in UC, resolved with drug stopping, corticosteroids and rituximab [12]. A real-world RA database study (n=3,126 tofacitinib starters of 153,788) is retrievable only as an abstract focused on tocilizumab; the tofacitinib hemoglobin figure was not reported there [13].

Mechanism

Not established in the retrieved sources: the label and SmPC text retrieved give no mechanism for the hemoglobin decreases. The hemolysis in the case report was diagnosed as autoimmune [12]. The anemia-of-inflammation pathway is discussed in the Iron entry. Directly measured oxygen delivery (arterial saturation, tissue oxygenation) was not found.

Counter-evidence held on record

Counter-evidence on record Maintenance incidence at 5 mg (5.51 per 100 patient-years) was about twice placebo (2.91) but not at 10 mg (2.55), which does not fit a dose-driven drug effect; maintenance populations were responders, adding selection bias [7][8]. The two OCTAVE sources are conference abstracts read through an aggregator and are one analysis, not two. The Padula registry is a completer analysis with channeling bias [9]. Framing caution Hemoglobin findings here are clinical outcomes to be monitored, not evidence that tofacitinib reduces oxygen delivery or that any nutrient is deficient.

Limits on what this shows

• Key null findings: No study found lower ATP with tofacitinib. No whole-body or clinical energy endpoint was found; fatigue and vitality scores did not track hemoglobin changes in the RA analysis [6].
• At-risk subgroups: None identified.

Direction

Reverse (higher ATP in inflamed cells).

Regulatory status

No ATP or energy-metabolism statement in the retrieved label text [1][3][4].

Population evidence

SW982 cells, tofacitinib 100 nM: ATP levels increased (p<0.0001) while methotrexate reduced ATP (p<0.0096) [16]. PsA fibroblast-like synoviocytes (n=18), 1 uM: ATP production increased (p=0.0313) [17].

Mechanism

Same improved oxidative phosphorylation reading as the mitochondrial entry. ATP output was a direct assay in both studies.

Counter-evidence held on record

Counter-evidence on record Rests on the same two papers as the mitochondrial entry (see Overlap map). Fatigue in these diseases is driven by inflammation, anemia and sleep, not by an ATP deficit, so no fatigue-energy link is drawn. Framing caution Not an energy or fatigue claim. Cell-culture ATP rises do not imply any change in a patient's energy level.