Azathioprine

Limits on what this shows

• Null findings: No human study of mitochondrial function on azathioprine was found. Concentrations used in vitro were not checked against clinical exposure, so clinical relevance is unknown.
• At-risk subgroups: Not defined. Hepatic impairment raises 6-MP exposure [3].

Regulatory status

No mitochondrial statement in the US label or SmPC. The US label warns of hepatic veno-occlusive disease in transplant patients [2].

Population evidence

Tier B only. Isolated rat hepatocytes and mitochondria (abstract read): azathioprine decreased viability, depleted reduced GSH, caused necrosis via the permeability transition pore, swelling and increased oxygen consumption of intact mitochondria, activated JNK and p38, all prevented by N-acetylcysteine; 6-mercaptopurine showed no effects on GSH content and metabolic activity [27]. HCT116 and HT29 colorectal cancer cells: azathioprine, 6-MP and 6-thioguanine generated ROS, caused loss of mitochondrial transmembrane potential, and N-acetylcysteine reduced apoptosis and autophagy [28]. Hepatoma cells: azathioprine plus buthionine sulfoximine caused mitochondrial dysfunction and necroptosis (abstract) [29]. Rats: azathioprine lowered MnSOD, GPx and GSH with metabolic enzyme loss, reversed by quercetin (abstract via a bibliographic record) [30]. Tier C: 6-thioguanine, an active metabolite, is incorporated into and oxidatively damages mitochondrial DNA in human cells [31].

Mechanism

Present, preclinical. The azathioprine-specific GSH depletion seen with the parent but not with 6-MP [27] points to the imidazole leaving group and GSH conjugation rather than purine analogue action; thioguanine nucleotide incorporation into mtDNA is a separate route [31]. Direction: excess toxic exposure.

Class cohort test

Azathioprine was in the cohort for [27][28][29][30]. [31] used 6-thioguanine, not azathioprine, and is Tier C for this drug (metabolite extrapolation, yellow cap).

Counter-evidence held on record

Counter-evidence and confounding Cell-line and rat-hepatocyte systems at undefined concentrations. No clinical biomarker (lactate, mtDNA copy number, respiratory chain activity) was found moving in patients. Framing caution Not a deficiency and not a supplement indication. N-acetylcysteine and quercetin protection in these studies is preclinical and is not a recommendation.

Limits on what this shows

• Null findings: [26] found markers minimized by adaptive response; the azathioprine group showed higher, not lower, glutathione peroxidase.
• At-risk subgroups: Not defined.

Regulatory status

No antioxidant or oxidative stress statement in the US label or SmPC [1][2][3].

Population evidence

Preclinical as in Mitochondrial burden: GSH depletion in hepatocytes [27], ROS generation reversed by N-acetylcysteine [28], lowered MnSOD, GPx and GSH in rats [30]. Clinical: 50 Crohn's patients on azathioprine monotherapy or anti-TNF plus azathioprine versus 25 controls (cross-sectional): serum and saliva MDA raised and vitamin E lowered in the monotherapy group versus the combination group and controls [33]. 40 renal transplant recipients on cyclosporine or azathioprine versus 19 controls: lower superoxide dismutase in both groups, higher glutathione peroxidase and higher zinc in the azathioprine group, with the authors describing an adaptive response and no meaningful difference between regimens [26].

Mechanism

Present, preclinical: GSH conjugation consumption and ROS [27][28].

Class cohort test

Azathioprine was in the cohort for all sources above. No class extrapolation.

Counter-evidence held on record

Counter-evidence and confounding In [33] the comparison groups differ in disease severity and co-therapy, and there is no untreated Crohn's arm; the disease itself raises oxidative markers. [26] confounds transplantation and cyclosporine. The two clinical studies point in different directions. Framing caution Antioxidant supplementation is not supported by this entry. Direction is mixed and the evidence is not attributable to the drug.

Limits on what this shows

• Null findings: Anemia prevalence fell from 33.8% to 16.7% across two IBD registers as immunosuppressant use more than doubled [15].
• At-risk subgroups: Active IBD, menstruating women, extensive small-bowel disease, people with GI bleeding.

Regulatory status

No iron mention and no iron interaction in the US label or SmPC [1][2][3].

Population evidence

1,036 IBD patients in 21 French centers: iron deficiency 23.7% (95% CI 19.8 to 28.1); risk factors were moderate or severe disease activity and concurrent anemia (both OR 3.66); thiopurine analysis not reported in the abstract [22]. 291 Romanian IBD patients: anemia 31.3%, iron deficiency 53.3%, higher anemia in Crohn's disease, whose patients more often received azathioprine, with no causal link drawn [23]. 32 IBD patients in remission on mesalazine and azathioprine versus 30 controls had lower red cell counts, hemoglobin and iron (cross-sectional, no drug-free IBD arm) [21]. In 103 IBD patients on azathioprine, sideropenia had no effect on MCV dynamics [8].

Mechanism

Absent for azathioprine. Disease pathways (chronic GI blood loss, inflammation-driven hepcidin, malabsorption) are outside the drug.

Class cohort test

No class extrapolation. Azathioprine was in the cohort only in [21] (all patients), and that study has no azathioprine-free IBD comparator.

Counter-evidence held on record

Counter-evidence and confounding Confounding by indication is the whole signal: disease activity was the main risk factor [22]. [21] cannot separate drug from disease. Framing caution Iron deficiency is not anemia, and anemia is not iron deficiency (Hard Rule 2). Iron status needs ferritin and transferrin saturation read alongside CRP in IBD. This entry does not recommend iron supplementation.

Limits on what this shows

• Null findings: No copper difference in [21].
• At-risk subgroups: Post-resection Crohn's disease, chronic diarrhea, alcohol use, high-dose zinc exposure.

Regulatory status

No copper mention in the US label or SmPC [1][2][3].

Population evidence

32 IBD patients on mesalazine and azathioprine had serum copper 1.03 versus 0.97 mg/L in controls (P=0.32, no difference) [21]. A children's IBD abstract states there are conflicting reports of copper status in adults with IBD (excerpt only) [25]. A 55-year-old with Crohn's disease, prior hemicolectomy and alcohol use presented with pancytopenia (hemoglobin 8.8 g/dL, MCV 135.1 fL) from copper deficiency and was not on immunosuppressants [24].

Mechanism

Absent for azathioprine. Disease and surgery pathway: malabsorption and chronic diarrhea [24].

Class cohort test

No class extrapolation. Azathioprine was in the cohort only in [21]. [24] and [25] did not involve azathioprine.

Counter-evidence held on record

Counter-evidence and confounding [21] is a 32-patient cross-sectional study without an azathioprine-free IBD arm. [24] is one case in a patient not on immunosuppressants, so it shows the disease and surgery pathway only. Graded yellow only as a context flag, kept separate from Iron (Hard Rule 7). Framing caution Not a depletion claim for azathioprine. Copper testing is a differential for unexplained macrocytic cytopenia in malabsorptive IBD, not a drug-based recommendation.

Limits on what this shows

• Null findings: A UC cohort with only one patient on azathioprine drew no thiopurine-microbiota conclusion [36].
• At-risk subgroups: Not defined.

Regulatory status

No microbiome statement in the US label or SmPC [1][2][3].

Population evidence

Effenberger 2021: 65 IBD patients on azathioprine or anti-TNF followed at baseline, 12 and 30 weeks; 1.7-fold higher microbial butyrate production capacity in remission than non-remission; in responders Proteobacteria fell and Bacteroidetes rose; the effect was most pronounced in azathioprine-treated Crohn's disease (p=0.008) [34]. Azathioprine was in the cohort, but the design compares response groups with no drug-free comparator. Wills 2014: 19 IBD patients (10 Crohn's, 9 UC), 16S pyrosequencing, thiopurine use correlated with lower richness and diversity [35]; the abstract did not stratify azathioprine from other thiopurines, so this is CLASS EXTRAPOLATION from thiopurines, azathioprine not confirmed in the cohort by name in what was read. Reverse direction: E. coli, Enterococcus faecalis and Bacteroides thetaiotaomicron convert azathioprine and 6-MP to thioguanine nucleotides (mini review) [37].

Mechanism

Not established for drug-to-microbiome effects. Thiopurines are antiproliferative, but no direct effect on bacterial communities was demonstrated in the sources read.

Class cohort test

[35] is CLASS EXTRAPOLATION from thiopurines, azathioprine not confirmed in the cohort; capped at yellow. [34] included azathioprine but cannot separate treatment from response and disease state.

Counter-evidence held on record

Counter-evidence and confounding Confounding by indication is presumed: IBD itself changes the microbiome, remission status drives the composition differences in [34], and [35] has 19 patients with individualized shifts. Framing caution Not a dysbiosis claim for azathioprine. The microbiome-to-drug direction [37] is a drug-exposure variable, not a nutrient effect.

Limits on what this shows

• Null findings: Infants of 19 mothers on thiopurines had hemoglobin comparable to controls (median 177 vs 181 g/L; total n=106); this concerns infants, not the treated adult [38]. Anemia prevalence in one gastroenterologist's IBD register fell from 33.8% (1993) to 16.7% (2003) while immunosuppressant use rose from 13.8% to 33.3%, mostly azathioprine; separate cohorts, so this is ecological, but it argues against a large population-level anemia signal [15].
• At-risk subgroups: TPMT or NUDT15 variant carriers, patients also taking allopurinol, febuxostat, aminosalicylates, ACE inhibitors, trimethoprim-sulfamethoxazole or ribavirin, and people with hepatic impairment (6-MP exposure 1.6-fold in non-cirrhotic and 6-fold in cirrhotic impairment per the SmPC) [1][3].

Regulatory status

US label: "Severe leukopenia, thrombocytopenia, anemias including macrocytic anemia, and/or pancytopenia may occur"; patients with TPMT or NUDT15 deficiency "may be at increased risk of severe myelotoxicity"; ACE inhibitors with azathioprine "reported to induce anemia and severe leukopenia" [1][2]. UK SmPC: bone marrow depression very common, anaemia uncommon, megaloblastic anaemia and erythroid hypoplasia rare [3]. Both labels were read through summarizing fetches.

Population evidence

No primary cohort quantifying hemoglobin change from baseline on azathioprine was opened, so no effect size or confidence interval is available. Case reports with azathioprine as the drug: a 66-year-old on azathioprine plus allopurinol for 3 months had hemoglobin 56 g/L, MCV 107.8 fL, WBC 3.2, and normal B12 and folate [10]; a 26-year-old with TPMT*2/*3A and mesalazine co-therapy had hemoglobin 8.2 g/dL, neutrophils 0.06 and platelets 37 after dose escalation to 150 mg/day [12]; two Crohn's patients developed pure red cell aplasia that recovered after stopping azathioprine [13]; one renal transplant recipient developed megaloblastic pancytopenia after 22 years [11]. Genotype context: in 142 Japanese IBD patients with thiopurine exposure, NUDT15 R139C carried an odds ratio of 28.4 for early leukopenia (azathioprine in the cohort) [6]; CPIC gives genotype-based thiopurine dosing for TPMT and NUDT15 [5].

Mechanism

Present. Thioguanine nucleotide incorporation and inhibition of purine synthesis act on proliferating cells (general mechanism statement from a Tier D reference [39]); the labels attribute the cytopenias to the drug [1][3]. Allopurinol and febuxostat (xanthine oxidase inhibition) and aminosalicylates (in vitro TPMT inhibition) raise myelotoxicity risk through higher active-metabolite exposure [1][3][10][12]. Direction: reduced red-cell production (drug toxicity).

Class cohort test

No class extrapolation. Every case report and genotype study listed involved azathioprine by name.

Counter-evidence held on record

Counter-evidence and confounding Baseline anemia in IBD is common and mostly iron deficiency or inflammation (see Iron). Case reports cannot give incidence. Leukopenia, not anemia, is the usual early NUDT15 phenotype [6]. Why not orange: the label confirms the hazard qualitatively but no Tier A primary cohort with hemoglobin measured against baseline was opened. Framing caution Anemia is an outcome, not iron or B12 deficiency (Hard Rule 2). The signal is drug toxicity managed by blood-count monitoring and genotype-guided dosing, not by supplements.

Limits on what this shows

• Null findings: No human ATP, phosphocreatine or energy-expenditure measurement on azathioprine was found.
• At-risk subgroups: Not defined.

Regulatory status

No ATP or energy-metabolism statement in the US label or SmPC. The US label states azathioprine is metabolized to 6-mercaptopurine [1][2].

Population evidence

Fernandez-Ramos 2017: Jurkat T cells exposed to 6-MP showed rapid ATP reduction with AMPK activation, lowered ADP and AMP at 48 hours, raised ADP/ATP and AMP/ATP ratios, and reduced glycolytic and glutaminolytic flux (abstract-level read via a summarizing fetch) [32]. Azathioprine-specific indirect: increased oxygen consumption in intact isolated rat liver mitochondria with azathioprine (ATP not measured) [27]; the colorectal cell study described mitochondrial depolarization but not direct ATP depletion [28].

Mechanism

Present for 6-MP: inhibition of de novo purine synthesis in cells that depend on it [32]. For azathioprine itself, only the mitochondrial findings above.

Class cohort test

CLASS (METABOLITE) EXTRAPOLATION: azathioprine was not in the cohort of [32]; 6-MP is its active metabolite per the label [1]. Capped at yellow. 6-MP is a purine analogue in a leukemia cell line, so relevance to treated patients is unproven.

Counter-evidence held on record

Counter-evidence and confounding Leukemic cell line, in vitro, concentrations not compared with patient exposure. Kept independent of Mitochondrial burden and Oxygen delivery/utilization (Hard Rule 7). Framing caution Fatigue on azathioprine is not attributed to ATP loss by this entry. No supplement is implied.