Isoniazid

證據等級: L5

預測適應症: 1 個

Isoniazid: From Tuberculosis to Conjunctivitis

One-Sentence Summary

Isoniazid is a first-line antituberculosis agent, most commonly used alone or in isoniazid-rifapentine combination regimens (e.g. 3HP/1HP) for active and latent tuberculosis. The TxGNN model predicts it may be effective for Conjunctivitis, with 1 clinical trial and 20 publications currently associated with this direction — though, as detailed below, the supporting evidence is largely indirect.

Quick Overview

Item	Content
Original Indication	Tuberculosis (active and latent TB infection)
Predicted New Indication	Conjunctivitis
TxGNN Prediction Score	99.36%
Evidence Level	L4
Market Status (Taiwan)	未上市 (Not currently marketed)
Number of NDAs/Licenses	0
Recommended Decision	Hold

Why is This Prediction Reasonable?

Currently, detailed mechanism of action data is not available (data gap DG002). Based on known pharmacology, isoniazid inhibits mycolic acid synthesis in the mycobacterial cell wall, an action specific to Mycobacterium tuberculosis and related organisms. This mechanism has no established direct pharmacological action on the conjunctiva or on common causes of conjunctivitis (allergic, viral, or bacterial).

The literature association between isoniazid and “conjunctivitis” almost entirely reflects a narrow clinical subgroup: tuberculous or phlyctenular keratoconjunctivitis, an ocular manifestation of underlying tuberculosis infection. In these cases, isoniazid improves the conjunctivitis indirectly, by treating the causative TB infection — not by acting on conjunctivitis as a general disease category. Two publications (PMID 14253168, PMID 5103251) do describe isoniazid used specifically for phlyctenular keratoconjunctivitis, but both are historical papers pertaining to TB-associated disease, not general conjunctivitis.

Given this, the TxGNN model’s high score (99.36%) most likely reflects the knowledge graph picking up a strong co-occurrence between isoniazid, tuberculosis, and conjunctivitis (as a TB complication), rather than a genuine, generalizable pharmacological effect on conjunctivitis. The one associated clinical trial (NCT04094012) is a Phase 3 safety comparison of latent TB regimens and was not designed to evaluate conjunctivitis outcomes.

Clinical Trial Evidence

Trial Number	Phase	Status	Enrollment	Key Findings
NCT04094012	Phase 3	Completed	490	Compared systemic adverse drug reaction rates between 3HP (rifapentine+isoniazid weekly ×12) and 1HP regimens for latent TB infection. Not designed to evaluate conjunctivitis efficacy; ocular findings, if any, would only appear as incidental adverse events (relevance grade C — low relevance to this indication).

Literature Evidence

PMID	Year	Type	Journal	Key Findings
5103251	1971	Case Report	Annales d’oculistique	Describes local (topical) use of isoniazid in treatment of ocular tuberculosis.
14253168	1965	Case Report	Am Rev Respir Dis	Isoniazid prophylaxis for phlyctenular keratoconjunctivitis among Alaskan Eskimos — the most direct isoniazid–conjunctivitis link identified.
1363080	1992	Review	Optometry Clinics	Review of ocular side effects of systemic drugs; conjunctivitis noted as an adverse effect associated with several drug classes (not specific to isoniazid).
14089390	1964	Case Report	Archives of Ophthalmology	Primary tuberculosis of the conjunctiva.
33607832	2021	Case Report	Medicine	Pediatric phlyctenular keratoconjunctivitis associated with primary sinonasal tuberculosis.
26692731	2015	Case Report	Middle East Afr J Ophthalmol	Tuberculous conjunctivitis in an anophthalmic socket.
17133069	2006	Case Report	Cornea	Mycobacterium tuberculosis presenting as chronic red eye (conjunctival TB).
25433746	2014	Case Report	Can J Ophthalmol	Conjunctival phlyctenulosis as a presenting sign of impending clinical tuberculosis.
10641112	1999	Case Report	Oftalmologia	28 cases of tuberculous/phlyctenular keratoconjunctivitis, mostly in children with primary TB.
4233886	1968	Case Report	Arch Ophtalmol Rev Gen Ophtalmol	Tuberculosis of the bulbar conjunctiva.

All identified literature concerns tuberculous/phlyctenular conjunctivitis specifically, not conjunctivitis of other etiologies. No RCTs supporting isoniazid for general conjunctivitis were found.

US Market Information

No TFDA/FDA marketing authorization records are currently available — isoniazid is recorded as 未上市 (not marketed) in this dataset, with 0 licenses on file.

Safety Considerations

Please refer to the package insert for safety information. Detailed TFDA warnings, contraindications, and drug-drug interaction data were not available in this evidence pack (data gap DG001, Blocking severity) and must be obtained before any safety evaluation can proceed.

Conclusion and Next Steps

Decision: Hold

Rationale:

The evidence supports isoniazid’s role in treating tuberculous conjunctivitis (a TB complication), not conjunctivitis as a general disease category — the TxGNN prediction likely reflects a co-occurrence artifact in the knowledge graph rather than a broad pharmacological effect. No RCT-level evidence and no trial specifically designed to test isoniazid for conjunctivitis exists.
A Blocking-severity data gap (missing TFDA label/warnings, DG001) also prevents any Stage 1 safety evaluation.

To proceed, the following is needed:

TFDA package insert (warnings, contraindications) — DG001
Detailed mechanism of action data from DrugBank — DG002
Clarification of target population: should this be re-scoped specifically to tuberculous/phlyctenular keratoconjunctivitis rather than conjunctivitis broadly, given the evidence base
If re-scoped, prospective data comparing isoniazid-treated vs. untreated TB-conjunctivitis outcomes
Disclaimer

This content is for research purposes only and does not constitute medical advice. Clinical validation is required before any clinical application.