Vitamin A

Evidence Level: L5

Predicted Indications: 10

Vitamin A

Pharmacist Assessment Report

Vitamin A: Congenital prothrombin deficiency (rank 1); best-supported alternative: perinatal disease (congenital prothrombin deficiency)

One-Sentence Summary

Vitamin A is a fat-soluble vitamin originally used to treat vitamin deficiency disorders, including vitamin A deficiency. The TxGNN model’s top-ranked prediction is congenital prothrombin deficiency, but that prediction has no clinical trials and no publications behind it. Its score comes from the TxGNN model (deep learning) and may reflect similarities the model learned between related diseases, not biological evidence. Among the ten predictions, the best-supported is perinatal disease (specifically bronchopulmonary dysplasia in very-low-birth-weight infants), with 1 clinical trial and multiple Cochrane systematic reviews.

Quick Overview

Item	Content
Original Indication	Not recorded in the local data
Predicted New Indication	Congenital prothrombin deficiency (rank 1); best-supported alternative: perinatal disease (congenital prothrombin deficiency)
TxGNN Prediction Score	0.9997
Evidence Level	L5 (congenital prothrombin deficiency); L1 (perinatal disease)
India Market Status	Listed in a compiled Indian medicine price dataset (8 of 8) – CDSCO approval status is not included in this data
Number of Listed Products	8 (valid: 8)
Recommended Decision	Hold (congenital prothrombin deficiency); Proceed with Guardrails (perinatal disease, limited to BPD prevention in very-low-birth-weight infants)

Why is This Prediction Reasonable?

Currently, detailed mechanism of action data is not available. Based on known information, vitamin A (retinol) acts through its metabolite retinoic acid, which binds nuclear retinoic acid receptors to regulate cell differentiation, proliferation, and epithelial integrity. Its efficacy in correcting vitamin A deficiency is established.

Top-ranked prediction (congenital prothrombin deficiency). This is a coagulation factor II defect. Vitamin A has no established link to it; vitamin K is the vitamin relevant to prothrombin synthesis. The very high score comes from the TxGNN model (deep learning) and may reflect similarities the model learned between related diseases, not biological evidence. It should not be acted on without supporting studies.

Best-supported prediction (perinatal disease). Vitamin A is needed for lung epithelial integrity and repair. Preterm infants are born with low vitamin A stores, and this is associated with bronchopulmonary dysplasia (BPD). Cochrane reviews of randomised trials indicate that vitamin A supplementation in very-low-birth-weight infants modestly reduces death or oxygen requirement at 36 weeks’ postmenstrual age. The supporting evidence covers BPD in preterm infants, a subset of “perinatal disease”, not the whole category.

Other predictions were mostly weak or not actionable:

Biotin metabolic disease, non-syndromic esophageal malformation, florid cemento-osseous dysplasia, segmental odontomaxillary dysplasia: no direct evidence; the scores come from the TxGNN model (deep learning) and may reflect similarities the model learned between related diseases, not biological evidence. Retinoic acid signalling in foregut development is plausible for the esophageal malformation, but untested and not relevant to postnatal supplementation.
Disease by subcellular system affected: an ontology classification node, not a clinical indication.
Injury (L2): wound-healing reviews and two neonatal trials, but only indirectly relevant.
Cell proliferation disorder and radiation or chemically induced disorder (L4): mechanistic and review literature only; categories too broad to define an indication.

Clinical Trial Evidence

No clinical trials were found for the top-ranked prediction, congenital prothrombin deficiency. Trials for the best-supported prediction (perinatal disease) and for other predictions with trial data are shown below.

Trial Number	Phase	Status	Enrollment	Key Findings
NCT01115478	Not applicable	Completed	2500	Zinc and/or vitamin A supplementation to reduce placental malaria and adverse pregnancy outcomes (perinatal disease; partly relevant)
NCT01203488	Phase 1/2	Completed	807	Higher-dose vitamin A vs control in extremely-low-birth-weight infants, assessing survival without BPD and sepsis risk (injury; neonatal lung injury setting)
NCT02102711	Not applicable	Completed	62	Oral vitamin A in neonates under 1500 g to prevent BPD and retinopathy of prematurity (injury; indirectly relevant)
NCT04000789	Not applicable	Unknown	96	Vitamin A + vitamin E + sodium hyaluronate eye drops vs sodium hyaluronate alone in diabetic retinopathy dry eye (cell proliferation disorder; loosely related)

Literature Evidence

No publications were found for congenital prothrombin deficiency. Key publications for the best-supported prediction (perinatal disease) and for injury are listed below.

PMID	Year	Type	Journal	Key Findings
27552058	2016	Systematic review (Cochrane)	Cochrane Database Syst Rev	Vitamin A supplementation to prevent mortality and morbidity in very-low-birth-weight infants
21975731	2011	Systematic review (Cochrane)	Cochrane Database Syst Rev	Earlier version of the same review
17943744	2007	Systematic review (Cochrane)	Cochrane Database Syst Rev	Earlier version of the same review
12519545	2002	Systematic review (Cochrane)	Cochrane Database Syst Rev	Earlier version of the same review
10796372	2000	Systematic review (Cochrane)	Cochrane Database Syst Rev	First version of the review; vitamin A for lung growth and epithelial integrity in preterm infants
22742601	2012	Systematic review / meta-analysis	Paediatr Perinat Epidemiol	Vitamin A and carotenoids in pregnancy and maternal, neonatal, and infant outcomes
25222155	2014	Review	JAMA Pediatrics	Vitamin A shortage and risk of bronchopulmonary dysplasia
34181622	2021	Review	Plast Reconstr Surg	Systematic summary of vitamin and mineral interventions in wound care; little evidence for official recommendations
34105132	2021	Meta-analysis	J Spec Oper Med	Vitamin A intake and fracture risk (safety signal at high intake)
40901583	2025	Review	World J Hepatol	Vitamin A toxicity and liver pathology

India Market Information

Authorization Number	Product Name	Dosage Form	Approved Indication
133468	M.V.I. Injection	Ampoule of 10 ml injection	Not stated
122794	Lycobold Capsule	Strip of 10 capsules	Not stated
218506	Tilvit Injection	Vial of 10 ml injection	Not stated
233452	Vitamin A 100000IU Syrup	Bottle of 50 ml syrup	Not stated
144556	Multiband Injection	Vial of 10 ml injection	Not stated

Eight valid registrations exist in total (injection, capsule, syrup, and chewable tablet forms).

Review correction (2026-11-06): Originally stated: “Eight valid single-ingredient registrations exist in total (injection, capsule, syrup, and chewable tablet forms).” Removed: it described the products as single-ingredient or combination products in a way that did not match the Indian medicine data (the Indian Medicine Dataset as of 2026-11-06). Source: Indian Medicine Dataset.

Safety Considerations

Contraindications suggested by the knowledge graph (not label text): kidney disease, liver disease, gallbladder disease, peptic ulcer disease, sarcoidosis, hypercalcemia, nephrocalcinosis, nephrolithiasis, hypervitaminosis, inflammatory bowel disease (including ulcerative colitis), gastroenteritis, sickle cell anemia, hemolytic anemia, hereditary hemochromatosis, hemosiderosis, Wilson disease, primary hyperoxaluria, and several other metabolic conditions.
Drug Interactions (68 recorded in total; main ones listed):
- Major: isotretinoin, acitretin, tretinoin (additive retinoid toxicity), and tetracycline-class antibiotics (doxycycline, tetracycline, minocycline, omadacycline, demeclocycline, eravacycline, oxytetracycline, sarecycline).
- Moderate: bexarotene, bile-acid sequestrants (cholestyramine, colesevelam, colestipol), sevelamer, selpercatinib.
- Minor: orlistat.
Dose-related concerns from the literature: excess vitamin A is associated with liver toxicity and increased fracture risk, so dosing limits are essential, particularly in neonates and pregnancy.

Conclusion and Next Steps

Decision: Hold for the top-ranked prediction (congenital prothrombin deficiency). Decision: Proceed with Guardrails for perinatal disease, limited to BPD prevention in very-low-birth-weight infants.

Rationale:

Congenital prothrombin deficiency has no trials or literature and no plausible mechanism, so its score from the TxGNN model (deep learning) likely reflects similarities the model learned between related diseases, not biological evidence. Perinatal disease is supported by repeated Cochrane reviews and large neonatal trials (L1), though the benefit is modest and the evidence applies to a subset of the category.
Injury (L2) is only indirectly supported, and the remaining predictions are L4–L5 and not actionable.

To proceed, the following is needed:

Package insert warnings and contraindications for the marketed products (currently unavailable and a blocking gap for safety screening)
Detailed mechanism of action data
For perinatal disease: a precisely defined indication (BPD prevention in very-low-birth-weight infants), dose, route, and toxicity limits, and an assessment of how the injection burden fits the available Indian formulations
Route compatibility assessment against the marketed forms (oral, injectable, syrup)
Safety monitoring plan covering hepatic toxicity, bone effects, and interactions with retinoids and tetracyclines

Review Log

Below is the review log for this page: statements checked against official Indian medicine data (the Indian Medicine Dataset used by this site). Corrections are limited to basic factual statements; model predictions, evidence levels, and conclusions were not rewritten.

Review date	Item	Action	Source
2026-11-06	Removed: “single-ingredient”	Corrected	Indian Medicine Dataset

Disclaimer

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