Adenine

證據等級: L5

預測適應症: 1 個

Adenine: From Purine Nucleobase to Drug-Induced Osteoporosis

One-Sentence Summary

Adenine (DB00173) is a naturally occurring purine nucleobase and fundamental structural component of nucleotides (AMP/ADP/ATP), with no established therapeutic indications and no US market authorisation on record. The TxGNN model predicts it may be effective for Drug-Induced Osteoporosis with a score of 99.16%; however, this prediction is very likely a false positive — all retrieved evidence describes adenine analogues (Adefovir, Tenofovir) as causative agents of drug-induced bone loss, not as treatments. Currently only 1 registry study (Grade C, unrelated) and 4 indirect publications (all Tier 3) are available, yielding an evidence level of L5.

Quick Overview

Item	Content
Original Indication	None established
Predicted New Indication	Drug-Induced Osteoporosis
TxGNN Prediction Score	99.16%
Evidence Level	L5
US Market Status	Not Marketed
Number of NDAs	0
Recommended Decision	Hold

Why is This Prediction Reasonable?

This prediction is not supported by mechanistic evidence — the available data points decisively in the opposite direction, and the high TxGNN score most likely reflects a model artefact rather than a genuine therapeutic signal.

Adenine (DB00173) is a purine nucleobase: one of the four bases in DNA and RNA, and the structural core of energy-carrying molecules such as AMP, ADP, and ATP. It is not an approved drug in the US, and no mechanism of action data is available for clinical therapeutic use. There is no established pharmacological pathway by which adenine itself would reverse or prevent drug-induced osteoporosis.

The high TxGNN score (0.9916) almost certainly arises from semantic confusion in the knowledge graph between adenine and adenine nucleotide analogues — specifically Adefovir dipivoxil and Tenofovir, which are structurally derived from adenine. These antiretroviral/antiviral drugs are well-documented causes of drug-induced osteoporosis: they inhibit mitochondrial DNA polymerase and impair renal tubular phosphate reabsorption, triggering Fanconi syndrome, hypophosphataemia, and ultimately hypophosphatemic osteomalacia or osteoporosis. Every piece of literature retrieved for this indication describes this toxicity pathway. The model appears to have linked “adenine → bone disease” co-occurrences in the literature without distinguishing therapeutic agent from causative agent.

Clinical Trial Evidence

Trial Number	Phase	Status	Enrollment	Key Findings
NCT06065852	N/A	Recruiting	35,000	National Registry of Rare Kidney Diseases (RaDaR) — an observational data-collection registry for rare renal conditions; entirely unrelated to Adenine as a treatment for drug-induced osteoporosis. Inclusion is likely a retrieval artefact from the known adenine-nephropathy rat model.

Literature Evidence

PMID	Year	Type	Journal	Key Findings
41924521	2026	Case Report	Frontiers in Endocrinology	Adefovir dipivoxil (an adenine analogue) induced Fanconi syndrome → hypophosphatemic osteomalacia in a 67-year-old woman. Describes the adenine analogue as the cause of drug-induced bone disease, not a treatment.
22943210	2012	PK/PD Review	Expert Opinion on Drug Metabolism & Toxicology	PK/PD review of emtricitabine/tenofovir for HIV; explains how nucleotide analogues act as chain terminators of viral reverse transcriptase. Not relevant to adenine as a therapeutic agent.
20026012	2010	In Vitro Mechanistic Study	Biochemical and Biophysical Research Communications	Tenofovir downregulates Gnas, Got2, and Snord32a in primary osteoclasts, providing a mechanistic basis for tenofovir-induced bone density loss — again describing an adenine analogue as the causative agent.
31026554	2019	Animal Study	Journal of Ethnopharmacology	Xian-Ling-Gu-Bao herbal formula hepatotoxicity in rats; used clinically to treat osteoporosis. No connection to adenine or drug-induced osteoporosis as a target indication.

Safety Considerations

Please refer to the package insert for safety information.

Conclusion and Next Steps

Decision: Hold

Rationale: The TxGNN prediction for Adenine → Drug-Induced Osteoporosis is assessed as a false positive driven by semantic conflation: adenine nucleotide analogues (Adefovir, Tenofovir) co-occur with drug-induced osteoporosis in the literature precisely because they cause the condition, and the model has not distinguished this from a therapeutic relationship. There is no mechanistic hypothesis, no clinical trial evidence, and no published literature supporting adenine itself as a treatment for drug-induced osteoporosis.

To proceed, the following would be needed:

A credible mechanistic hypothesis explaining how free adenine (not its phosphonate analogues) could therapeutically reverse or prevent drug-induced osteoporosis
Preclinical evidence (cell-based or animal model) specifically testing adenine — not adefovir or tenofovir — in a bone loss model
Knowledge graph audit to flag and correct the adenine ↔ adenine-analogue semantic conflation in TxGNN, to prevent similar false-positive propagation to other candidates
If no mechanistic rationale can be established after literature review, this candidate should be formally deprioritized and recorded as a confirmed model artefact
Disclaimer

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