Vigabatrin

Limits on what this shows

• Key negative/null findings: A combined Crossref works-API search ("vigabatrin mitochondria astrocyte vacuolation energy metabolism") returned no additional vigabatrin-specific hit beyond these two studies. No study measuring mitochondrial structure, number, or electron-transport-chain activity in vigabatrin-treated human tissue was located.
• At-risk subgroups: Not established in humans; both supporting studies are animal/invertebrate models (mouse, Drosophila), not stratified by age, dose, or duration in a way that maps onto the pediatric infantile-spasms or adult partial-seizure populations vigabatrin actually treats.
• Counter-evidence on record: Torin 1's partial normalization of the mitochondrial increase in Vogel et al. indicates the effect is pharmacologically modifiable rather than fixed, but this was tested only as a mechanistic probe, not as a proposed clinical intervention, and no vigabatrin-specific study found the opposite (no change in mitochondrial number or structure).

Regulatory status

Not named in the FDA label. The label's own disclosed, vigabatrin-specific structural-imaging finding (reversible MRI T2-hyperintensity/intramyelinic edema; see Overlap map and Monitoring panel) is a distinct, already-described radiological phenomenon, not a measured mitochondrial-structure finding, and is not used as evidence for this entry.

Population evidence

Animal, Tier B, vigabatrin-specific: Vogel et al. (2015, Ann Clin Transl Neurol, DOI 10.1002/acn3.200) found vigabatrin exposure produced significantly enhanced mitochondrial numbers in mouse tissue, a mitochondrial-biogenesis effect linked to mTOR-pathway activation and increased apoptotic/phosphorylation markers; the mTOR inhibitor Torin 1 selectively, partially normalized the mitochondrial increase. Separately, Shamapari and Nagaraj (2024, Toxicol Res (Camb), DOI 10.1093/toxres/tfae180) exposed Drosophila larvae to vigabatrin (VGB) and carbamazepine (CLB) and found both drugs upregulated four genes by proteomic/mRNA analysis, including ND75, the fly ortholog of the mitochondrial complex I (NADH dehydrogenase) subunit NDUFS1; both drugs also produced DNA damage (genotoxicity) in this model.

Mechanism

Named and drug-specific in both studies, but distinct from each other and from vigabatrin's own intended GABA-transaminase inhibition (Hard Rule 2): Vogel et al. propose an mTOR-mediated mitochondrial-biogenesis/mitophagy imbalance; Shamapari and Nagaraj report transcriptional upregulation of a complex I subunit gene alongside broader genotoxic stress-response signaling. Neither mechanism operates at the GABA-transaminase active site itself.

Framing caution

This is not a nutrient-deficiency finding and nothing here is reachable by supplementation. Both supporting studies are animal/invertebrate models; magnitude and clinical relevance in vigabatrin-treated humans is unknown, and no human monitoring marker for this specific effect exists. See Overlap map: this entry shares its one Drosophila source with the Oxidative stress and ATP production entries below.

Counter-evidence held on record

No study was located finding vigabatrin has no effect on mitochondrial number or structure; the one modifying factor found (Torin 1) was a mechanistic probe, not evidence against the underlying finding.

Limits on what this shows

• Key negative/null findings: Two Crossref works-API searches ("vigabatrin oxidative stress antioxidant"; "vigabatrin oxidative stress apoptosis") returned no additional vigabatrin-specific hit beyond these two studies. No human study measuring an oxidative-stress or antioxidant-capacity marker in vigabatrin-treated patients was located.
• At-risk subgroups: Not established in humans. The protective finding is from a rat gynecologic ischemia-reperfusion model, not an epilepsy model or population; the contrasting finding is from Drosophila larvae. Neither model population resembles vigabatrin's actual pediatric infantile-spasms or adult partial-seizure patients.
• Counter-evidence on record: The two sources are themselves mutual counter-evidence: a tissue-protective, antioxidant-raising direction in rat ovarian tissue versus a pro-oxidant-leaning, contrasting direction in Drosophila. This is disclosed as a genuine, unresolved, model-dependent conflict rather than resolved in either direction.

Regulatory status

Not named in the FDA label.

Population evidence

Animal, Tier B, vigabatrin-specific: Hizal et al. (2025, J Ovarian Res, DOI 10.1186/s13048-025-01794-0), 30 female Wistar rats across 5 groups (sham, vigabatrin-only, ischemia-reperfusion, sham+vigabatrin, ischemia-reperfusion+vigabatrin), oral vigabatrin 50 mg/kg: in the ischemia-reperfusion+vigabatrin group versus ischemia-reperfusion alone, vigabatrin significantly decreased malondialdehyde (MDA, p<0.001) and increased total glutathione, superoxide dismutase, and catalase, alongside reduced TNF-alpha/IL-1beta/IL-6 (p<0.05) and improved ovarian histopathology. Separately, Shamapari and Nagaraj (2024, Toxicol Res (Camb), DOI 10.1093/toxres/tfae180, same Drosophila study cited under Mitochondrial burden) reported that carbamazepine reduced reactive oxygen species and lipid peroxidation with increased antioxidant enzymes, while vigabatrin "showed contrasting results," and that both drugs nonetheless produced DNA damage.

Mechanism

Not unified between the two studies. Hizal et al. propose a direct antioxidant/anti-inflammatory action of vigabatrin protecting ovarian tissue from ischemia-reperfusion injury. Shamapari and Nagaraj's contrasting result in Drosophila is reported without a named mechanism in the metadata available this session.

Framing caution

Neither direction supports any supplementation or antioxidant-avoidance recommendation. Neither model is vigabatrin's clinical population (ovarian ischemia-reperfusion surgery; Drosophila toxicology), and the conflict between them means this cannot be read as settled evidence of either protection or depletion in humans. See Overlap map: the Drosophila source is shared with Mitochondrial burden and ATP production.

Counter-evidence held on record

The two vigabatrin-specific sources located directly contradict each other on direction; both are reported here rather than one being selectively used to support a single narrative.

Limits on what this shows

• Why no grade: No present mechanism was identified at Step 2 connecting vigabatrin's pharmacology (minimal metabolism, no hepatic conjugation, renal elimination unchanged, no documented chelation) to copper handling. A Crossref search surfaced condition-level literature on serum copper in epileptic/drug-resistant-epilepsy populations generally (Wojciak et al. 2013, Epilepsy Research; Kour et al. 2026 systematic review/meta-analysis, Epilepsy Research; Su et al. 2024 preprint), none of which stratifies by drug or confirms vigabatrin in the studied cohort; per Hard Rule 6, confounding by the seizure disorder itself is presumed until excluded, and these sources are Tier C at best for vigabatrin and are not used to support any grade here. No circulating consumer claim specifically linking vigabatrin to copper status was located. The dual condition for ungraded (Step 7) is met.

Channels tried

Crossref works API, "vigabatrin copper ceruloplasmin trace element": attempted; returned 429 before results could be reviewed on that exact phrasing. Crossref works API, "vigabatrin serum copper zinc iron micronutrient children epilepsy": completed, surfaced the condition-level (not vigabatrin-specific) sources named above. Europe PMC, PubMed search-listing, Semantic Scholar, OpenAlex bulk search: 429 on attempts made this session for adjacent queries. General WebSearch: unavailable (budget exhausted). Second pass: PubMed search-listing, "vigabatrin copper OR vigabatrin ceruloplasmin OR vigabatrin zinc OR vigabatrin trace element": completed successfully, 10 results reviewed by title; none isolated vigabatrin and copper (results concerned P-glycoprotein substrate status, GABA-aminotransferase structural biology, drosophila genotoxicity, and GABA/ischemia energy metabolism, reviewed and used elsewhere in this dossier for other categories, not for copper). OpenAlex single-record lookup was used successfully elsewhere this pass but had no vigabatrin-copper record to look up. Europe PMC and Semantic Scholar remained 429 on this pass.

Assessment

Most likely a genuine literature gap for vigabatrin specifically, consistent with the absence of any copper-relevant mechanism in its pharmacokinetic profile, and now reinforced by a second, independently-phrased, cleanly completed PubMed search that also found nothing vigabatrin-and-copper-specific. The condition-level epilepsy/copper literature found is a real signal worth flagging as a confound (seizure disorders generally, not any specific drug, may associate with altered trace-element status) but not a drug-caused finding.

Limits on what this shows

• Why no grade: No present mechanism connecting vigabatrin's pharmacology to iron absorption, transport, or storage was identified. The label's anemia/hemoglobin finding is graded separately under Oxygen delivery/utilization (orange) specifically because no source measured iron status, ferritin, or transferrin saturation in vigabatrin-treated patients; per Hard Rule 2, anemia is not graded as iron deficiency without the nutrient itself being measured, and per Hard Rule 7, Iron and Oxygen delivery/utilization are kept as two independent entries even though they both originate from the same adverse-event tabulation (see Overlap map). The same condition-level epilepsy/trace-element sources found under Copper (Wojciak 2013, Kour 2026) also report serum iron but fail the same cohort test and are not used as support here. No circulating consumer claim specifically linking vigabatrin to iron status was located. The dual condition for ungraded (Step 7) is met.

Channels tried

Crossref works API, "vigabatrin serum copper zinc iron micronutrient children epilepsy": completed (same search as Copper, above), surfaced only condition-level, non-drug-specific sources. Europe PMC, PubMed search-listing, Semantic Scholar, OpenAlex bulk search: 429 on attempts made this session. General WebSearch: unavailable (budget exhausted). Second pass: covered by the same combined PubMed search described under Copper, above (completed cleanly, no vigabatrin-iron-specific hit).

Assessment

Most likely a genuine gap for a drug-caused iron effect specifically, now reinforced by a second cleanly completed PubMed search; the label's anemia signal (graded separately above) is the one real hematologic finding, and no source distinguishes it as iron-mediated.

Limits on what this shows

• Why no grade: No present mechanism connecting vigabatrin's pharmacology (not hepatically metabolized, no bile-acid or broad-spectrum antimicrobial activity described, renally cleared) to gut microbiome composition, diversity, or function was identified at Step 2. A second-pass PubMed search located Mu et al. (2022, JCI Insight, PMID 35730569), "Targeted gut microbiota manipulation attenuates seizures in a model of infantile spasms syndrome": this rodent infantile-spasms model found a targeted probiotic combination (Streptococcus thermophilus and Lactococcus lactis) reduced seizure frequency and improved developmental outcomes, comparable to a ketogenic diet. Vigabatrin, a first-line infantile-spasms treatment, is not identified as a tested agent or comparator in this study's available record; it is a microbiota-to-seizure intervention study in the disease model vigabatrin treats, not a vigabatrin-to-microbiota study, so it fails the Step 4 cohort test for vigabatrin specifically and is disclosed here only as condition-level context (per Hard Rule 6, confounding by the seizure disorder/model itself, not the drug) rather than used as support. No circulating consumer claim specifically linking vigabatrin to microbiome disruption was located. The dual condition for ungraded (Step 7) is met on the evidence actually confirmed.

Channels tried

First pass: Crossref works API, "vigabatrin gut microbiome microbiota": attempted after a 230-second wait following a prior rate-limit response; returned 429 again. Europe PMC, PubMed search-listing, Semantic Scholar, OpenAlex bulk search: 429 on attempts made in the first pass. Second pass: PubMed search-listing, "vigabatrin gut microbiome OR vigabatrin microbiota": completed successfully, 2 results reviewed (Scantlebury et al. 2025, a ketogenic-diet-for-infantile-spasms review, not vigabatrin-and-microbiome-specific; Mu et al. 2022, described above). General WebSearch: unavailable (budget exhausted, both passes).

Assessment

Now a completed, clean search rather than a tooling-limited one: PubMed returned reviewable results on this exact term, and neither result is a vigabatrin-specific microbiome finding. This is most likely a genuine literature gap for vigabatrin's own effect on the microbiome, though the Mu et al. 2022 finding (gut microbiota manipulation as a seizure treatment in the same disease model) is worth flagging as adjacent context: it shows active research interest in the gut-brain axis for infantile spasms, just not yet pointed at vigabatrin itself.

Limits on what this shows

• Key negative/null findings: None located specifically testing whether the hemoglobin decline reverses with dose reduction, discontinuation, or iron/B12 repletion. A second corroborating peer-reviewed source was sought (Crossref, "vigabatrin anemia hemoglobin infantile spasms clinical trial") but the attempt returned a 429 before any result could be reviewed; this gap is why the entry stays orange rather than advancing toward red.
• At-risk subgroups: Not stratified by age or dose in the retrievable label text; infantile-spasms (pediatric) trials and adult complex-partial-seizure trials are reported together in the cited adverse-event tabulation.
• Counter-evidence on record: None located distinguishing this from a trial-population artifact (e.g., confounding by the seizure disorder itself, blood draws, or concomitant AEDs with their own hematologic effects); Hard Rule 6's confounding-by-indication caution applies and is not excluded by any source opened this session.

Regulatory status

Labeled adverse reaction in the current FDA-approved Sabril prescribing information (Adverse Reactions section).

Population evidence

North American controlled trials: anemia reported as an adverse event in 6% of patients (16/280) on SABRIL versus 2% (3/188) on placebo. Mean hemoglobin decrease approximately 3% on SABRIL versus 0% on placebo. No confidence interval or p-value is published in the retrievable label text; this is the label's own descriptive adverse-event tabulation, not a dedicated hematology substudy.

Mechanism

Not named in the label. No vigabatrin-specific study was located this session proposing a mechanism (e.g., marrow suppression, hemodilution, or a nutrient-cofactor pathway); it is reported here as a measured hematologic finding of unestablished mechanism, not attributed to iron, B12, copper, or any other specific nutrient.

Framing caution

This is a measured hematologic adverse effect, not a diagnosed iron-, B12-, or copper-deficiency anemia. See the separate Iron entry (ungraded) for why this finding is not extended to a nutrient-specific claim.

Counter-evidence held on record

No dedicated null-result study was located; the label itself is the only source reviewed, and no second corroborating or contradicting peer-reviewed source could be opened this session (see Key negative/null findings).

Limits on what this shows

• Key negative/null findings: No study was located measuring ATP production as reduced or impaired specifically by vigabatrin; the one gene-expression finding (sesB upregulation) does not by itself establish a direction for net ATP output and is disclosed as such rather than forced into either a depletion or protection framing.
• At-risk subgroups: Not established in humans. Both directional findings are from acute ischemia models (isolated mitochondria; gerbil bilateral ischemia), not from chronic anticonvulsant dosing in the pediatric infantile-spasms or adult partial-seizure populations vigabatrin actually treats.
• Counter-evidence on record: None located contradicting the protective direction in the two studies that measure it; the gene-expression-only finding (Shamapari/Nagaraj) neither confirms nor contradicts it.

Regulatory status

Not named in the FDA label.

Population evidence

Animal/in-vitro, Tier B, vigabatrin-specific: (1) Ravasz et al. (2017, Neurochem Int, DOI 10.1016/j.neuint.2017.03.008), isolated mouse brain and liver mitochondria, used vigabatrin specifically as the pharmacological GABA-transaminase inhibitor and found that catabolism of GABA, succinic semialdehyde, or gamma-hydroxybutyrate through the GABA shunt, the pathway vigabatrin blocks, impairs mitochondrial ATP generation via substrate-level phosphorylation; the study's own design used vigabatrin to establish this causal role of the shunt rather than to characterize vigabatrin's own net ATP effect, so this is reported here as supporting the shunt-blockade/sparing direction rather than a depletion claim. (2) Abel and McCandless (1992, J Neurochem, DOI 10.1111/j.1471-4159.1992.tb09780.x), Mongolian gerbils, found that vigabatrin (gamma-vinyl-GABA) pretreatment to elevate brain GABA attenuated the ischemia-induced depletion of ATP, phosphocreatine, glucose, and glycogen compared to untreated ischemic controls, a directly measured, protective, drug-specific energy-metabolite finding. (3) Shamapari and Nagaraj (2024, Toxicol Res (Camb), DOI 10.1093/toxres/tfae180, same Drosophila study cited above) found vigabatrin upregulated sesB, the Drosophila gene encoding the mitochondrial ADP/ATP translocase, alongside genotoxic stress-response signaling; direction of net ATP effect from this one gene-expression change is not established in that study.

Mechanism

Named and drug-specific, converging on vigabatrin's own intended pharmacology acting one step upstream (Hard Rule 2 distinction preserved: the category graded here is ATP output, a downstream metabolic consequence, not the GABA-transaminase inhibition itself): by blocking GABA-transaminase, vigabatrin prevents GABA from entering the shunt pathway that Ravasz et al. show impairs substrate-level phosphorylation, consistent with the ATP-sparing effect Abel and McCandless measured directly during ischemia.

Framing caution

This is not a deficiency and nothing here supports a supplementation claim. Both ischemia-model findings are acute, animal, and not demonstrated in vigabatrin's actual treated population or at anticonvulsant-range chronic dosing; the Drosophila gene finding is directionally uninterpreted. See Overlap map: the Drosophila source is shared with Mitochondrial burden and Oxidative stress.

Counter-evidence held on record

No vigabatrin-specific source found this session reports vigabatrin impairing ATP production; both sources that establish a direction point the opposite way, under acute ischemic stress specifically.