Amlexanox

updated

Amlexanox was FDA-approved as a 5% oral paste (APHTHASOL, 1996) for aphthous ulcers, and both US products are now listed as discontinued. In the one controlled human metabolic trial, 42 obese adults with type 2 diabetes had a small HbA1c fall (registry P = 0.05) concentrated in responders, with no weight loss versus placebo. The metabolic-rate and fat-loss claims rest on mouse data.

TBK1 and IKK-epsilon inhibitor (mechanism shown in mice)

  • FDA approved APHTHASOL (amlexanox 5% oral paste) on 17 December 1996 and a 2 mg topical patch on 29 September 2004; both are listed as discontinued.
  • In the 2002 label's three controlled studies, median time to complete healing of aphthous ulcers was 0.7 days shorter than vehicle and 1.6 days shorter than no treatment.
  • In a 42-patient trial in obese adults with type 2 diabetes, mean HbA1c went from 7.68 to 7.44% on amlexanox and from 7.77 to 7.76% on placebo (registry P = 0.05); weight did not differ from placebo.
  • Improvements in insulin sensitivity and liver fat appeared only in a subset of responders in the trial abstract.
  • A PubMed search on 9 October 2026 returned 28 records on amlexanox with energy expenditure or weight terms in human contexts; the only human metabolic trial was the 42-patient study.
  • A Japanese approval was not located, and the paste's PK does not transfer to oral tablets or vendor powder.

What is amlexanox, and what was it approved for?

FDA approved APHTHASOL (amlexanox 5% oral paste) for aphthous ulcers on 17 December 1996, and a 2 mg topical patch on 29 September 2004. Both products are listed as discontinued.

FDA approved APHTHASOL (amlexanox 5% oral paste) under NDA 020511 on 17 December 1996 as a new molecular entity, and an amlexanox 2 mg topical patch under NDA 021727 on 29 September 2004. Both products are listed as discontinued, and DailyMed searches on 9 October 2026 returned no amlexanox or Aphthasol label.

The 2002 APHTHASOL label (labeling supplement approved 11 June 2002) indicates the paste for aphthous ulcers in people with normal immune systems.

No primary record of a Japanese approval was located. PMDA's prescription-drug database lists amlexanox by generic name under an oral anti-allergy code and an ophthalmic code, but both showed no products or package inserts on 9 October 2026. A 2013 mouse paper describes amlexanox as an approved drug in clinical use for aphthous ulcers and asthma; that is secondary support, and Japanese status is not established here.

Amlexanox is C16H14N2O4 (298.29 g/mol, CAS 68302-57-8). The US product was a 5% paste whose systemic exposure came mainly from swallowed drug, while the metabolic trials used oral Solfa tablets, so paste data do not transfer to oral tablets or vendor powder.

Drugs@FDA: APHTHASOL approved 17 December 1996, patch 29 September 2004, both discontinued

Regulatory record

U.S. Food and Drug Administration. Drugs@FDA application records for NDA 020511 (APHTHASOL, amlexanox 5% oral paste) and NDA 021727 (amlexanox 2 mg topical patch). Dates read from the openFDA Drugs@FDA records.

FDA approved APHTHASOL (amlexanox 5% oral paste) under NDA 020511 on 17 December 1996 as a new molecular entity, and an amlexanox 2 mg topical patch under NDA 021727 on 29 September 2004. Both products are listed as discontinued. DailyMed searches for amlexanox and Aphthasol on 9 October 2026 returned no label.

Participants / model
Not a clinical population
Treatment
5% oral paste; 2 mg topical patch
Follow-up
Approvals in 1996 and 2004
Study design
FDA application records
Funding
not stated in the source

These records give approval dates and product status. The indication comes from the 2002 label.

Read the original source
APHTHASOL label (2002): ulcer healing 0.7 days faster than vehicle, PK, adverse reactions

FDA prescribing information

U.S. Food and Drug Administration. APHTHASOL (amlexanox 5% oral paste) prescribing information, labeling supplement approved 11 June 2002 (NDA 020511).

The 2002 label indicates the 5% paste for aphthous ulcers in people with normal immune systems. In three controlled studies (464 amlexanox, 465 vehicle, 195 no treatment), amlexanox shortened median time to complete healing by 0.7 days versus vehicle and 1.6 days versus no treatment; at day 4, complete healing was 37% versus 27% on vehicle. The label says the mechanism is unknown; in vitro, amlexanox inhibits histamine and leukotriene release. After 100 mg of paste (5 mg amlexanox), serum levels peaked near 120 ng/mL at 2.4 hours, mostly through swallowed drug, with a half-life of 3.5 +/- 1.1 hours in healthy people. No related adverse reaction was reported by more than 5% of patients; transient pain, stinging or burning occurred in 1 to 2%.

Participants / model
People with aphthous ulcers and normal immune systems; healthy people for PK
Treatment
5% amlexanox oral paste applied to ulcers
Follow-up
Labeling supplement approved 11 June 2002
Study design
FDA prescribing information
Funding
not stated in the source

Paste data. The page gives no label dosing schedule, and paste PK does not transfer to oral tablets or vendor powder.

Read the original source
PMDA: no Japanese approval record located, listings empty

Regulatory record

Pharmaceuticals and Medical Devices Agency (Japan). Prescription drug search, generic-name listings for amlexanox under oral anti-allergy code 4490006 and ophthalmic code 1319725.

No primary record of a Japanese approval was located. PMDA's prescription-drug database has generic-name listings for amlexanox under the oral anti-allergy code 4490006 and the ophthalmic code 1319725, but both showed no products or package inserts on 9 October 2026. Support for Japanese asthma use here is secondary: the 2013 mouse paper describes amlexanox as an approved drug in clinical use for aphthous ulcers and asthma, and the Michigan trials used Solfa tablets.

Participants / model
Not a clinical population
Treatment
Oral tablets; eye drops
Follow-up
Database read on 9 October 2026
Study design
Regulator database listing
Funding
not stated in the source

Japanese approval and current marketing status are not established by this record.

Read the original source
Obese mice: energy expenditure up, weight loss, better insulin sensitivity, less fatty liver

Animal study

Reilly SM et al. Nat Med. 2013;19(3):313-21. PMID 23396211.

Amlexanox inhibits the noncanonical IkappaB kinases IKK-epsilon and TBK1, which are induced in liver and fat of high-fat-fed mice. In obese mice it raised energy expenditure through thermogenesis, produced weight loss, improved insulin sensitivity and reduced fatty liver.

Participants / model
Obese mice
Treatment
Amlexanox; route and dose not stated in the abstract
Follow-up
not stated in the source
Study design
Original preclinical study
Funding
not stated in the source

Mouse findings. The abstract does not state the route or dose, so no comparison with human exposure is possible.

Read the original source
PubChem CID 2161: C16H14N2O4, 298.29 g/mol

Chemical identity record

National Library of Medicine. PubChem compound record CID 2161 (amlexanox).

Amlexanox: PubChem CID 2161, C16H14N2O4, 298.29 g/mol, CAS 68302-57-8 (label molecular weight 298.30). The US product was a 5% paste whose systemic exposure came mainly from swallowed drug, while the metabolic trials used oral Solfa tablets, so paste PK does not transfer to oral tablets or vendor powder.

Participants / model
Chemical identity, not a clinical population
Treatment
Defined molecule
Follow-up
Not applicable
Study design
NLM chemical record
Funding
not stated in the source

Identity record only; it says nothing about efficacy or safety.

Read the original source

How well did amlexanox work on mouth ulcers?

Modestly. In the 2002 label's controlled studies, median time to complete healing was 0.7 days shorter than vehicle and 1.6 days shorter than no treatment, and at day 4 complete healing was 37% versus 27% on vehicle.

In three controlled studies in the 2002 label (464 amlexanox, 465 vehicle, 195 no treatment), amlexanox shortened median time to complete healing by 0.7 days versus vehicle and 1.6 days versus no treatment; at day 4, complete healing was 37% versus 27% on vehicle.

Four vehicle-controlled, randomized, double-blind, multicenter studies in 1335 people with 1 to 3 aphthous ulcers under 48 hours old found that the 5% paste sped pain resolution and healing compared with both vehicle and no treatment.

These are topical outcomes in people with mouth ulcers. They say nothing about metabolism.

APHTHASOL label (2002): ulcer healing 0.7 days faster than vehicle, PK, adverse reactions

FDA prescribing information

U.S. Food and Drug Administration. APHTHASOL (amlexanox 5% oral paste) prescribing information, labeling supplement approved 11 June 2002 (NDA 020511).

The 2002 label indicates the 5% paste for aphthous ulcers in people with normal immune systems. In three controlled studies (464 amlexanox, 465 vehicle, 195 no treatment), amlexanox shortened median time to complete healing by 0.7 days versus vehicle and 1.6 days versus no treatment; at day 4, complete healing was 37% versus 27% on vehicle. The label says the mechanism is unknown; in vitro, amlexanox inhibits histamine and leukotriene release. After 100 mg of paste (5 mg amlexanox), serum levels peaked near 120 ng/mL at 2.4 hours, mostly through swallowed drug, with a half-life of 3.5 +/- 1.1 hours in healthy people. No related adverse reaction was reported by more than 5% of patients; transient pain, stinging or burning occurred in 1 to 2%.

Participants / model
People with aphthous ulcers and normal immune systems; healthy people for PK
Treatment
5% amlexanox oral paste applied to ulcers
Follow-up
Labeling supplement approved 11 June 2002
Study design
FDA prescribing information
Funding
not stated in the source

Paste data. The page gives no label dosing schedule, and paste PK does not transfer to oral tablets or vendor powder.

Read the original source
Four randomized studies: 1335 people with aphthous ulcers, faster pain relief and healing

Randomized controlled trial

Khandwala A et al. Oral Surg Oral Med Oral Pathol Oral Radiol Endod. 1997;83(2):222-30. PMID 9117754.

Four vehicle-controlled, randomized, double-blind, multicenter studies in 1335 people with 1 to 3 aphthous ulcers under 48 hours old found that the 5% paste sped pain resolution and healing compared with both vehicle and no treatment. These are topical outcomes and say nothing about metabolism.

Participants / model
1335 people with recurrent minor aphthous ulcers
Treatment
5% amlexanox oral paste applied to ulcers
Follow-up
not stated in the source
Study design
Four vehicle-controlled, randomized, double-blind, multicenter studies
Funding
not stated in the source

Endpoints were complete healing and complete pain resolution of mouth ulcers, not metabolic outcomes.

Read the original source

What did the type 2 diabetes studies find?

In 42 obese adults with type 2 diabetes, amlexanox lowered HbA1c slightly, and insulin sensitivity and liver fat improved only in a subset of responders. Weight did not differ from placebo, and the registry HbA1c result is borderline (P = 0.05).

A proof-of-concept randomized double-blind placebo-controlled trial in 42 obese adults with type 2 diabetes and nonalcoholic fatty liver disease reported a statistically significant fall in HbA1c and fructosamine with amlexanox. Improvements in insulin sensitivity and hepatic steatosis appeared only in a subset of responders, who had a distinct inflammatory gene-expression signature in subcutaneous fat at baseline. The abstract reports no weight or energy-expenditure measurement; any link to energy expenditure is inferred from fat gene expression.

The trial's registry record posts results for Solfa amlexanox tablets 25 mg three times daily for 2 weeks, then 50 mg three times daily for 10 weeks, versus placebo (21 per arm). Mean HbA1c went from 7.68 to 7.44% on amlexanox and from 7.77 to 7.76% on placebo (t-test P = 0.05). MRI liver fat, measured in 20 patients, went from 18 to 16.22% on amlexanox versus 12.82 to 12.88% on placebo (P = 0.38). Weight went from 100.95 to 100.81 kg on amlexanox and from 100.24 to 97.94 kg on placebo (P = 0.51).

An earlier open-label single-arm pilot of Solfa tablets for 12 weeks (7 enrolled, 6 completed; terminated for lack of funding) posts mean changes of 0 in HbA1c, -2.2 percentage points in MRI liver fat and -2.2 kg in weight. With no control group, these changes cannot be attributed to the drug, and the arm description gives no dose.

Company and patent interests are disclosed in the 2021 mouse paper: its senior author founded Elgia Therapeutics and is a named inventor on amlexanox-related patents, one of them on obesity treatment. They are noted here because the human trial studied the same drug.

42 obese adults with type 2 diabetes and NAFLD: HbA1c fell, benefit mainly in responders

Randomized controlled trial

Oral EA et al. Cell Metab. 2017;26(1):157-170.e7. PMID 28683283.

Amlexanox produced a statistically significant fall in HbA1c and fructosamine. Improvements in insulin sensitivity and hepatic steatosis appeared only in a subset of responders, who had a distinct inflammatory gene-expression signature in subcutaneous fat at baseline. The abstract reports no weight or energy-expenditure measurement; any link to energy expenditure is inferred from fat gene expression.

Participants / model
42 obese adults with type 2 diabetes and nonalcoholic fatty liver disease
Treatment
Oral amlexanox or placebo (regimen in registry record NCT01975935)
Follow-up
12 weeks (registry record)
Study design
Proof-of-concept randomized double-blind placebo-controlled trial
Funding
not stated in the source

Endpoints were HbA1c, fructosamine, insulin sensitivity, hepatic fat and fat gene expression. Only the abstract was read, so the responder definition is not described here.

Read the original source
Registry results: HbA1c 7.68 to 7.44% versus 7.77 to 7.76% (P = 0.05); weight not different

Trial registry record

ClinicalTrials.gov record NCT01975935. Sponsor University of Michigan.

The registry record posts results for Solfa amlexanox tablets 25 mg three times daily for 2 weeks, then 50 mg three times daily for 10 weeks, versus placebo (21 per arm). Mean HbA1c went from 7.68 to 7.44% on amlexanox and from 7.77 to 7.76% on placebo (t-test P = 0.05). MRI liver fat, measured in 20 patients, went from 18 to 16.22% on amlexanox versus 12.82 to 12.88% on placebo (P = 0.38). Weight went from 100.95 to 100.81 kg on amlexanox and from 100.24 to 97.94 kg on placebo (P = 0.51). Serious adverse events occurred in 1 of 21 in each arm.

Participants / model
42 obese adults with type 2 diabetes
Treatment
Oral amlexanox 25 mg three times daily, then 50 mg three times daily, versus placebo
Follow-up
12 weeks
Study design
Registered randomized placebo-controlled trial with posted results
Funding
Sponsor listed: University of Michigan

The HbA1c P value of 0.05 is borderline, and baseline liver fat differed between arms.

Read the original source
Open-label pilot: 7 enrolled, 6 completed, no control group, terminated for lack of funding

Trial registry record

ClinicalTrials.gov record NCT01842282.

An earlier open-label single-arm pilot of Solfa tablets for 12 weeks (7 enrolled, 6 completed; terminated for lack of funding) posts mean changes of 0 in HbA1c (range -0.6 to 0.6), -2.2 percentage points in MRI liver fat (range -5.9 to 0.3) and -2.2 kg in weight (range -5.7 to -0.1). With no control group, these changes cannot be attributed to the drug, and the arm description gives no dose.

Participants / model
6 obese adults with type 2 diabetes, aged 18 to 59
Treatment
Oral amlexanox (Solfa tablets), open label
Follow-up
12 weeks
Study design
Registered open-label single-arm pilot with posted results
Funding
not stated in the source

No control group and a dose not given in the arm description, so the changes cannot be attributed to amlexanox.

Read the original source
Obese mice: weight loss from FGF21-dependent energy expenditure; author patent and company interests

Animal study

Reilly SM et al. J Clin Invest. 2021;131(10). PMID 33822771.

In obese mice, amlexanox caused a transient drop in food intake, while lasting weight loss came from higher energy expenditure through FGF21-dependent beiging of white fat; acute glucose lowering ran through adipocyte IL-6 and hepatic STAT3. The abstract describes earlier work as showing weight loss in obese animals and patients, but the controlled registry results from the human trial show no weight difference versus placebo.

Participants / model
Obese mice
Treatment
Amlexanox; route and dose not stated in the abstract
Follow-up
not stated in the source
Study design
Original preclinical study
Funding
Funding not given in the abstract; the senior author discloses founding Elgia Therapeutics and being a named inventor on amlexanox-related patents, one of them on obesity treatment

Mouse findings. The senior author's company and patent interests are noted wherever the human trial is cited.

Read the original source

Does amlexanox raise metabolic rate, cut fat or improve insulin sensitivity in people?

Insulin sensitization has partial support in a subset of people with type 2 diabetes. Higher metabolic rate and fat loss are mouse findings, and a PubMed search on 9 October 2026 found no human study that measured energy expenditure or showed fat loss.

Sellers describe amlexanox as an insulin sensitizer that raises metabolic rate and drives fat loss in obesity-related metabolic dysfunction.

A PubMed search on 9 October 2026 for amlexanox with energy expenditure, thermogenesis, metabolic rate, weight loss, body weight, fat mass or obesity terms in human or clinical contexts returned 28 records. By title, the only human metabolic trial was the 42-patient study; the rest were mouse, cell, chemistry or review records.

In that trial, HbA1c fell slightly more on amlexanox, but improvements in insulin sensitivity and liver fat appeared only in responders and weight did not differ from placebo. Insulin sensitization is therefore partly supported in a subset of people with type 2 diabetes. The metabolic-rate and fat-loss claims rest on mouse data in the records found.

PubMed search, 9 October 2026: 28 records, one human metabolic trial

PubMed search record

PubMed search run on 9 October 2026: amlexanox AND (energy expenditure OR thermogenesis OR metabolic rate OR weight loss OR body weight OR fat mass OR obesity) AND (patients OR humans OR human OR randomized OR volunteers OR clinical). 28 records returned.

The search returned 28 records. By title, the only human metabolic trial was the 42-patient study; the rest were mouse, cell, chemistry or review records. The metabolic-rate and fat-loss claims rest on mouse data in the records reviewed.

Participants / model
Records returned, not a population
Treatment
Not applicable
Follow-up
Search run on 9 October 2026
Study design
Literature search
Funding
not stated in the source

A search result shows what PubMed returned on that date. It is not proof that no unindexed study exists.

Read the original source
42 obese adults with type 2 diabetes and NAFLD: HbA1c fell, benefit mainly in responders

Randomized controlled trial

Oral EA et al. Cell Metab. 2017;26(1):157-170.e7. PMID 28683283.

Amlexanox produced a statistically significant fall in HbA1c and fructosamine. Improvements in insulin sensitivity and hepatic steatosis appeared only in a subset of responders, who had a distinct inflammatory gene-expression signature in subcutaneous fat at baseline. The abstract reports no weight or energy-expenditure measurement; any link to energy expenditure is inferred from fat gene expression.

Participants / model
42 obese adults with type 2 diabetes and nonalcoholic fatty liver disease
Treatment
Oral amlexanox or placebo (regimen in registry record NCT01975935)
Follow-up
12 weeks (registry record)
Study design
Proof-of-concept randomized double-blind placebo-controlled trial
Funding
not stated in the source

Endpoints were HbA1c, fructosamine, insulin sensitivity, hepatic fat and fat gene expression. Only the abstract was read, so the responder definition is not described here.

Read the original source
Registry results: HbA1c 7.68 to 7.44% versus 7.77 to 7.76% (P = 0.05); weight not different

Trial registry record

ClinicalTrials.gov record NCT01975935. Sponsor University of Michigan.

The registry record posts results for Solfa amlexanox tablets 25 mg three times daily for 2 weeks, then 50 mg three times daily for 10 weeks, versus placebo (21 per arm). Mean HbA1c went from 7.68 to 7.44% on amlexanox and from 7.77 to 7.76% on placebo (t-test P = 0.05). MRI liver fat, measured in 20 patients, went from 18 to 16.22% on amlexanox versus 12.82 to 12.88% on placebo (P = 0.38). Weight went from 100.95 to 100.81 kg on amlexanox and from 100.24 to 97.94 kg on placebo (P = 0.51). Serious adverse events occurred in 1 of 21 in each arm.

Participants / model
42 obese adults with type 2 diabetes
Treatment
Oral amlexanox 25 mg three times daily, then 50 mg three times daily, versus placebo
Follow-up
12 weeks
Study design
Registered randomized placebo-controlled trial with posted results
Funding
Sponsor listed: University of Michigan

The HbA1c P value of 0.05 is borderline, and baseline liver fat differed between arms.

Read the original source

What did the mouse studies find?

In obese mice, amlexanox raised energy expenditure, caused weight loss, improved insulin sensitivity and reduced fatty liver. The abstracts do not state the route or dose.

Amlexanox inhibits the noncanonical IkappaB kinases IKK-epsilon and TBK1, which are induced in liver and fat of high-fat-fed mice. In obese mice it raised energy expenditure through thermogenesis, produced weight loss, improved insulin sensitivity and reduced fatty liver.

A 2021 mouse paper found that amlexanox caused a transient drop in food intake, while lasting weight loss came from higher energy expenditure through FGF21-dependent beiging of white fat. Acute glucose lowering ran through adipocyte IL-6 and hepatic STAT3. Its abstract describes earlier work as showing weight loss in obese animals and patients, but the controlled registry results from the human trial show no weight difference versus placebo.

Neither abstract states the route or dose, so no mouse-to-human comparison is drawn here.

Obese mice: energy expenditure up, weight loss, better insulin sensitivity, less fatty liver

Animal study

Reilly SM et al. Nat Med. 2013;19(3):313-21. PMID 23396211.

Amlexanox inhibits the noncanonical IkappaB kinases IKK-epsilon and TBK1, which are induced in liver and fat of high-fat-fed mice. In obese mice it raised energy expenditure through thermogenesis, produced weight loss, improved insulin sensitivity and reduced fatty liver.

Participants / model
Obese mice
Treatment
Amlexanox; route and dose not stated in the abstract
Follow-up
not stated in the source
Study design
Original preclinical study
Funding
not stated in the source

Mouse findings. The abstract does not state the route or dose, so no comparison with human exposure is possible.

Read the original source
Obese mice: weight loss from FGF21-dependent energy expenditure; author patent and company interests

Animal study

Reilly SM et al. J Clin Invest. 2021;131(10). PMID 33822771.

In obese mice, amlexanox caused a transient drop in food intake, while lasting weight loss came from higher energy expenditure through FGF21-dependent beiging of white fat; acute glucose lowering ran through adipocyte IL-6 and hepatic STAT3. The abstract describes earlier work as showing weight loss in obese animals and patients, but the controlled registry results from the human trial show no weight difference versus placebo.

Participants / model
Obese mice
Treatment
Amlexanox; route and dose not stated in the abstract
Follow-up
not stated in the source
Study design
Original preclinical study
Funding
Funding not given in the abstract; the senior author discloses founding Elgia Therapeutics and being a named inventor on amlexanox-related patents, one of them on obesity treatment

Mouse findings. The senior author's company and patent interests are noted wherever the human trial is cited.

Read the original source
Registry results: HbA1c 7.68 to 7.44% versus 7.77 to 7.76% (P = 0.05); weight not different

Trial registry record

ClinicalTrials.gov record NCT01975935. Sponsor University of Michigan.

The registry record posts results for Solfa amlexanox tablets 25 mg three times daily for 2 weeks, then 50 mg three times daily for 10 weeks, versus placebo (21 per arm). Mean HbA1c went from 7.68 to 7.44% on amlexanox and from 7.77 to 7.76% on placebo (t-test P = 0.05). MRI liver fat, measured in 20 patients, went from 18 to 16.22% on amlexanox versus 12.82 to 12.88% on placebo (P = 0.38). Weight went from 100.95 to 100.81 kg on amlexanox and from 100.24 to 97.94 kg on placebo (P = 0.51). Serious adverse events occurred in 1 of 21 in each arm.

Participants / model
42 obese adults with type 2 diabetes
Treatment
Oral amlexanox 25 mg three times daily, then 50 mg three times daily, versus placebo
Follow-up
12 weeks
Study design
Registered randomized placebo-controlled trial with posted results
Funding
Sponsor listed: University of Michigan

The HbA1c P value of 0.05 is borderline, and baseline liver fat differed between arms.

Read the original source

How does amlexanox work, and what does the body do with it?

In obese mice, amlexanox inhibits the noncanonical IkappaB kinases IKK-epsilon and TBK1. For mouth ulcers the label says the mechanism is unknown, and after a paste dose its half-life was 3.5 hours in healthy people.

In obese mice, amlexanox inhibits IKK-epsilon and TBK1, and lasting weight loss came from higher energy expenditure through FGF21-dependent beiging of white fat. Acute glucose lowering ran through adipocyte IL-6 and hepatic STAT3.

For the mouth-ulcer use, the label says the mechanism is unknown. In vitro, amlexanox inhibits histamine and leukotriene release.

After 100 mg of 5% paste (5 mg amlexanox), serum levels peaked near 120 ng/mL at 2.4 hours, mostly through swallowed drug, with a half-life of 3.5 +/- 1.1 hours in healthy people. The metabolic trials used oral Solfa tablets, so the paste figures do not transfer to oral tablets or vendor powder, and the sources reviewed give no oral half-life.

Obese mice: energy expenditure up, weight loss, better insulin sensitivity, less fatty liver

Animal study

Reilly SM et al. Nat Med. 2013;19(3):313-21. PMID 23396211.

Amlexanox inhibits the noncanonical IkappaB kinases IKK-epsilon and TBK1, which are induced in liver and fat of high-fat-fed mice. In obese mice it raised energy expenditure through thermogenesis, produced weight loss, improved insulin sensitivity and reduced fatty liver.

Participants / model
Obese mice
Treatment
Amlexanox; route and dose not stated in the abstract
Follow-up
not stated in the source
Study design
Original preclinical study
Funding
not stated in the source

Mouse findings. The abstract does not state the route or dose, so no comparison with human exposure is possible.

Read the original source
Obese mice: weight loss from FGF21-dependent energy expenditure; author patent and company interests

Animal study

Reilly SM et al. J Clin Invest. 2021;131(10). PMID 33822771.

In obese mice, amlexanox caused a transient drop in food intake, while lasting weight loss came from higher energy expenditure through FGF21-dependent beiging of white fat; acute glucose lowering ran through adipocyte IL-6 and hepatic STAT3. The abstract describes earlier work as showing weight loss in obese animals and patients, but the controlled registry results from the human trial show no weight difference versus placebo.

Participants / model
Obese mice
Treatment
Amlexanox; route and dose not stated in the abstract
Follow-up
not stated in the source
Study design
Original preclinical study
Funding
Funding not given in the abstract; the senior author discloses founding Elgia Therapeutics and being a named inventor on amlexanox-related patents, one of them on obesity treatment

Mouse findings. The senior author's company and patent interests are noted wherever the human trial is cited.

Read the original source
APHTHASOL label (2002): ulcer healing 0.7 days faster than vehicle, PK, adverse reactions

FDA prescribing information

U.S. Food and Drug Administration. APHTHASOL (amlexanox 5% oral paste) prescribing information, labeling supplement approved 11 June 2002 (NDA 020511).

The 2002 label indicates the 5% paste for aphthous ulcers in people with normal immune systems. In three controlled studies (464 amlexanox, 465 vehicle, 195 no treatment), amlexanox shortened median time to complete healing by 0.7 days versus vehicle and 1.6 days versus no treatment; at day 4, complete healing was 37% versus 27% on vehicle. The label says the mechanism is unknown; in vitro, amlexanox inhibits histamine and leukotriene release. After 100 mg of paste (5 mg amlexanox), serum levels peaked near 120 ng/mL at 2.4 hours, mostly through swallowed drug, with a half-life of 3.5 +/- 1.1 hours in healthy people. No related adverse reaction was reported by more than 5% of patients; transient pain, stinging or burning occurred in 1 to 2%.

Participants / model
People with aphthous ulcers and normal immune systems; healthy people for PK
Treatment
5% amlexanox oral paste applied to ulcers
Follow-up
Labeling supplement approved 11 June 2002
Study design
FDA prescribing information
Funding
not stated in the source

Paste data. The page gives no label dosing schedule, and paste PK does not transfer to oral tablets or vendor powder.

Read the original source
PubChem CID 2161: C16H14N2O4, 298.29 g/mol

Chemical identity record

National Library of Medicine. PubChem compound record CID 2161 (amlexanox).

Amlexanox: PubChem CID 2161, C16H14N2O4, 298.29 g/mol, CAS 68302-57-8 (label molecular weight 298.30). The US product was a 5% paste whose systemic exposure came mainly from swallowed drug, while the metabolic trials used oral Solfa tablets, so paste PK does not transfer to oral tablets or vendor powder.

Participants / model
Chemical identity, not a clinical population
Treatment
Defined molecule
Follow-up
Not applicable
Study design
NLM chemical record
Funding
not stated in the source

Identity record only; it says nothing about efficacy or safety.

Read the original source

What risks do the label and the diabetes trial report?

On the 2002 paste label, no related adverse reaction was reported by more than 5% of patients, and transient pain, stinging or burning occurred in 1 to 2%. In the 42-patient diabetes trial, serious adverse events occurred in 1 of 21 in each arm.

On the 2002 paste label, no related adverse reaction was reported by more than 5% of patients. Transient pain, stinging or burning occurred in 1 to 2%.

In the registry results of the 42-patient trial of oral tablets, serious adverse events occurred in 1 of 21 patients in each arm. The sources reviewed give no safety data for vendor powder.

APHTHASOL label (2002): ulcer healing 0.7 days faster than vehicle, PK, adverse reactions

FDA prescribing information

U.S. Food and Drug Administration. APHTHASOL (amlexanox 5% oral paste) prescribing information, labeling supplement approved 11 June 2002 (NDA 020511).

The 2002 label indicates the 5% paste for aphthous ulcers in people with normal immune systems. In three controlled studies (464 amlexanox, 465 vehicle, 195 no treatment), amlexanox shortened median time to complete healing by 0.7 days versus vehicle and 1.6 days versus no treatment; at day 4, complete healing was 37% versus 27% on vehicle. The label says the mechanism is unknown; in vitro, amlexanox inhibits histamine and leukotriene release. After 100 mg of paste (5 mg amlexanox), serum levels peaked near 120 ng/mL at 2.4 hours, mostly through swallowed drug, with a half-life of 3.5 +/- 1.1 hours in healthy people. No related adverse reaction was reported by more than 5% of patients; transient pain, stinging or burning occurred in 1 to 2%.

Participants / model
People with aphthous ulcers and normal immune systems; healthy people for PK
Treatment
5% amlexanox oral paste applied to ulcers
Follow-up
Labeling supplement approved 11 June 2002
Study design
FDA prescribing information
Funding
not stated in the source

Paste data. The page gives no label dosing schedule, and paste PK does not transfer to oral tablets or vendor powder.

Read the original source
Registry results: HbA1c 7.68 to 7.44% versus 7.77 to 7.76% (P = 0.05); weight not different

Trial registry record

ClinicalTrials.gov record NCT01975935. Sponsor University of Michigan.

The registry record posts results for Solfa amlexanox tablets 25 mg three times daily for 2 weeks, then 50 mg three times daily for 10 weeks, versus placebo (21 per arm). Mean HbA1c went from 7.68 to 7.44% on amlexanox and from 7.77 to 7.76% on placebo (t-test P = 0.05). MRI liver fat, measured in 20 patients, went from 18 to 16.22% on amlexanox versus 12.82 to 12.88% on placebo (P = 0.38). Weight went from 100.95 to 100.81 kg on amlexanox and from 100.24 to 97.94 kg on placebo (P = 0.51). Serious adverse events occurred in 1 of 21 in each arm.

Participants / model
42 obese adults with type 2 diabetes
Treatment
Oral amlexanox 25 mg three times daily, then 50 mg three times daily, versus placebo
Follow-up
12 weeks
Study design
Registered randomized placebo-controlled trial with posted results
Funding
Sponsor listed: University of Michigan

The HbA1c P value of 0.05 is borderline, and baseline liver fat differed between arms.

Read the original source

Why does amlexanox sit in B tier?

A past FDA approval and vehicle-controlled ulcer studies support the B for the approved paste, and the mouse mechanism is well supported. In the one controlled human metabolic trial, the HbA1c fall was small and concentrated in responders, with no weight difference versus placebo.

The B reflects a mechanism that is well supported in obese mice, partial benefit in a subset of people in one 42-patient trial, and a discontinued approved product for a different use. On the approved ulcer use, median healing was 0.7 days faster than vehicle on the label.

The supplier's metabolic claims are separate. Insulin sensitization has partial support in a subset of people with type 2 diabetes, while higher metabolic rate and fat loss are mouse findings.

Drugs@FDA: APHTHASOL approved 17 December 1996, patch 29 September 2004, both discontinued

Regulatory record

U.S. Food and Drug Administration. Drugs@FDA application records for NDA 020511 (APHTHASOL, amlexanox 5% oral paste) and NDA 021727 (amlexanox 2 mg topical patch). Dates read from the openFDA Drugs@FDA records.

FDA approved APHTHASOL (amlexanox 5% oral paste) under NDA 020511 on 17 December 1996 as a new molecular entity, and an amlexanox 2 mg topical patch under NDA 021727 on 29 September 2004. Both products are listed as discontinued. DailyMed searches for amlexanox and Aphthasol on 9 October 2026 returned no label.

Participants / model
Not a clinical population
Treatment
5% oral paste; 2 mg topical patch
Follow-up
Approvals in 1996 and 2004
Study design
FDA application records
Funding
not stated in the source

These records give approval dates and product status. The indication comes from the 2002 label.

Read the original source
APHTHASOL label (2002): ulcer healing 0.7 days faster than vehicle, PK, adverse reactions

FDA prescribing information

U.S. Food and Drug Administration. APHTHASOL (amlexanox 5% oral paste) prescribing information, labeling supplement approved 11 June 2002 (NDA 020511).

The 2002 label indicates the 5% paste for aphthous ulcers in people with normal immune systems. In three controlled studies (464 amlexanox, 465 vehicle, 195 no treatment), amlexanox shortened median time to complete healing by 0.7 days versus vehicle and 1.6 days versus no treatment; at day 4, complete healing was 37% versus 27% on vehicle. The label says the mechanism is unknown; in vitro, amlexanox inhibits histamine and leukotriene release. After 100 mg of paste (5 mg amlexanox), serum levels peaked near 120 ng/mL at 2.4 hours, mostly through swallowed drug, with a half-life of 3.5 +/- 1.1 hours in healthy people. No related adverse reaction was reported by more than 5% of patients; transient pain, stinging or burning occurred in 1 to 2%.

Participants / model
People with aphthous ulcers and normal immune systems; healthy people for PK
Treatment
5% amlexanox oral paste applied to ulcers
Follow-up
Labeling supplement approved 11 June 2002
Study design
FDA prescribing information
Funding
not stated in the source

Paste data. The page gives no label dosing schedule, and paste PK does not transfer to oral tablets or vendor powder.

Read the original source
Four randomized studies: 1335 people with aphthous ulcers, faster pain relief and healing

Randomized controlled trial

Khandwala A et al. Oral Surg Oral Med Oral Pathol Oral Radiol Endod. 1997;83(2):222-30. PMID 9117754.

Four vehicle-controlled, randomized, double-blind, multicenter studies in 1335 people with 1 to 3 aphthous ulcers under 48 hours old found that the 5% paste sped pain resolution and healing compared with both vehicle and no treatment. These are topical outcomes and say nothing about metabolism.

Participants / model
1335 people with recurrent minor aphthous ulcers
Treatment
5% amlexanox oral paste applied to ulcers
Follow-up
not stated in the source
Study design
Four vehicle-controlled, randomized, double-blind, multicenter studies
Funding
not stated in the source

Endpoints were complete healing and complete pain resolution of mouth ulcers, not metabolic outcomes.

Read the original source
Obese mice: energy expenditure up, weight loss, better insulin sensitivity, less fatty liver

Animal study

Reilly SM et al. Nat Med. 2013;19(3):313-21. PMID 23396211.

Amlexanox inhibits the noncanonical IkappaB kinases IKK-epsilon and TBK1, which are induced in liver and fat of high-fat-fed mice. In obese mice it raised energy expenditure through thermogenesis, produced weight loss, improved insulin sensitivity and reduced fatty liver.

Participants / model
Obese mice
Treatment
Amlexanox; route and dose not stated in the abstract
Follow-up
not stated in the source
Study design
Original preclinical study
Funding
not stated in the source

Mouse findings. The abstract does not state the route or dose, so no comparison with human exposure is possible.

Read the original source
42 obese adults with type 2 diabetes and NAFLD: HbA1c fell, benefit mainly in responders

Randomized controlled trial

Oral EA et al. Cell Metab. 2017;26(1):157-170.e7. PMID 28683283.

Amlexanox produced a statistically significant fall in HbA1c and fructosamine. Improvements in insulin sensitivity and hepatic steatosis appeared only in a subset of responders, who had a distinct inflammatory gene-expression signature in subcutaneous fat at baseline. The abstract reports no weight or energy-expenditure measurement; any link to energy expenditure is inferred from fat gene expression.

Participants / model
42 obese adults with type 2 diabetes and nonalcoholic fatty liver disease
Treatment
Oral amlexanox or placebo (regimen in registry record NCT01975935)
Follow-up
12 weeks (registry record)
Study design
Proof-of-concept randomized double-blind placebo-controlled trial
Funding
not stated in the source

Endpoints were HbA1c, fructosamine, insulin sensitivity, hepatic fat and fat gene expression. Only the abstract was read, so the responder definition is not described here.

Read the original source
Registry results: HbA1c 7.68 to 7.44% versus 7.77 to 7.76% (P = 0.05); weight not different

Trial registry record

ClinicalTrials.gov record NCT01975935. Sponsor University of Michigan.

The registry record posts results for Solfa amlexanox tablets 25 mg three times daily for 2 weeks, then 50 mg three times daily for 10 weeks, versus placebo (21 per arm). Mean HbA1c went from 7.68 to 7.44% on amlexanox and from 7.77 to 7.76% on placebo (t-test P = 0.05). MRI liver fat, measured in 20 patients, went from 18 to 16.22% on amlexanox versus 12.82 to 12.88% on placebo (P = 0.38). Weight went from 100.95 to 100.81 kg on amlexanox and from 100.24 to 97.94 kg on placebo (P = 0.51). Serious adverse events occurred in 1 of 21 in each arm.

Participants / model
42 obese adults with type 2 diabetes
Treatment
Oral amlexanox 25 mg three times daily, then 50 mg three times daily, versus placebo
Follow-up
12 weeks
Study design
Registered randomized placebo-controlled trial with posted results
Funding
Sponsor listed: University of Michigan

The HbA1c P value of 0.05 is borderline, and baseline liver fat differed between arms.

Read the original source

Studies and sources

Drugs@FDA: APHTHASOL approved 17 December 1996, patch 29 September 2004, both discontinued

Regulatory record

U.S. Food and Drug Administration. Drugs@FDA application records for NDA 020511 (APHTHASOL, amlexanox 5% oral paste) and NDA 021727 (amlexanox 2 mg topical patch). Dates read from the openFDA Drugs@FDA records.

FDA approved APHTHASOL (amlexanox 5% oral paste) under NDA 020511 on 17 December 1996 as a new molecular entity, and an amlexanox 2 mg topical patch under NDA 021727 on 29 September 2004. Both products are listed as discontinued. DailyMed searches for amlexanox and Aphthasol on 9 October 2026 returned no label.

Participants / model
Not a clinical population
Treatment
5% oral paste; 2 mg topical patch
Follow-up
Approvals in 1996 and 2004
Study design
FDA application records
Funding
not stated in the source

These records give approval dates and product status. The indication comes from the 2002 label.

Read the original source
APHTHASOL label (2002): ulcer healing 0.7 days faster than vehicle, PK, adverse reactions

FDA prescribing information

U.S. Food and Drug Administration. APHTHASOL (amlexanox 5% oral paste) prescribing information, labeling supplement approved 11 June 2002 (NDA 020511).

The 2002 label indicates the 5% paste for aphthous ulcers in people with normal immune systems. In three controlled studies (464 amlexanox, 465 vehicle, 195 no treatment), amlexanox shortened median time to complete healing by 0.7 days versus vehicle and 1.6 days versus no treatment; at day 4, complete healing was 37% versus 27% on vehicle. The label says the mechanism is unknown; in vitro, amlexanox inhibits histamine and leukotriene release. After 100 mg of paste (5 mg amlexanox), serum levels peaked near 120 ng/mL at 2.4 hours, mostly through swallowed drug, with a half-life of 3.5 +/- 1.1 hours in healthy people. No related adverse reaction was reported by more than 5% of patients; transient pain, stinging or burning occurred in 1 to 2%.

Participants / model
People with aphthous ulcers and normal immune systems; healthy people for PK
Treatment
5% amlexanox oral paste applied to ulcers
Follow-up
Labeling supplement approved 11 June 2002
Study design
FDA prescribing information
Funding
not stated in the source

Paste data. The page gives no label dosing schedule, and paste PK does not transfer to oral tablets or vendor powder.

Read the original source
Four randomized studies: 1335 people with aphthous ulcers, faster pain relief and healing

Randomized controlled trial

Khandwala A et al. Oral Surg Oral Med Oral Pathol Oral Radiol Endod. 1997;83(2):222-30. PMID 9117754.

Four vehicle-controlled, randomized, double-blind, multicenter studies in 1335 people with 1 to 3 aphthous ulcers under 48 hours old found that the 5% paste sped pain resolution and healing compared with both vehicle and no treatment. These are topical outcomes and say nothing about metabolism.

Participants / model
1335 people with recurrent minor aphthous ulcers
Treatment
5% amlexanox oral paste applied to ulcers
Follow-up
not stated in the source
Study design
Four vehicle-controlled, randomized, double-blind, multicenter studies
Funding
not stated in the source

Endpoints were complete healing and complete pain resolution of mouth ulcers, not metabolic outcomes.

Read the original source
PMDA: no Japanese approval record located, listings empty

Regulatory record

Pharmaceuticals and Medical Devices Agency (Japan). Prescription drug search, generic-name listings for amlexanox under oral anti-allergy code 4490006 and ophthalmic code 1319725.

No primary record of a Japanese approval was located. PMDA's prescription-drug database has generic-name listings for amlexanox under the oral anti-allergy code 4490006 and the ophthalmic code 1319725, but both showed no products or package inserts on 9 October 2026. Support for Japanese asthma use here is secondary: the 2013 mouse paper describes amlexanox as an approved drug in clinical use for aphthous ulcers and asthma, and the Michigan trials used Solfa tablets.

Participants / model
Not a clinical population
Treatment
Oral tablets; eye drops
Follow-up
Database read on 9 October 2026
Study design
Regulator database listing
Funding
not stated in the source

Japanese approval and current marketing status are not established by this record.

Read the original source
Obese mice: energy expenditure up, weight loss, better insulin sensitivity, less fatty liver

Animal study

Reilly SM et al. Nat Med. 2013;19(3):313-21. PMID 23396211.

Amlexanox inhibits the noncanonical IkappaB kinases IKK-epsilon and TBK1, which are induced in liver and fat of high-fat-fed mice. In obese mice it raised energy expenditure through thermogenesis, produced weight loss, improved insulin sensitivity and reduced fatty liver.

Participants / model
Obese mice
Treatment
Amlexanox; route and dose not stated in the abstract
Follow-up
not stated in the source
Study design
Original preclinical study
Funding
not stated in the source

Mouse findings. The abstract does not state the route or dose, so no comparison with human exposure is possible.

Read the original source
Obese mice: weight loss from FGF21-dependent energy expenditure; author patent and company interests

Animal study

Reilly SM et al. J Clin Invest. 2021;131(10). PMID 33822771.

In obese mice, amlexanox caused a transient drop in food intake, while lasting weight loss came from higher energy expenditure through FGF21-dependent beiging of white fat; acute glucose lowering ran through adipocyte IL-6 and hepatic STAT3. The abstract describes earlier work as showing weight loss in obese animals and patients, but the controlled registry results from the human trial show no weight difference versus placebo.

Participants / model
Obese mice
Treatment
Amlexanox; route and dose not stated in the abstract
Follow-up
not stated in the source
Study design
Original preclinical study
Funding
Funding not given in the abstract; the senior author discloses founding Elgia Therapeutics and being a named inventor on amlexanox-related patents, one of them on obesity treatment

Mouse findings. The senior author's company and patent interests are noted wherever the human trial is cited.

Read the original source
42 obese adults with type 2 diabetes and NAFLD: HbA1c fell, benefit mainly in responders

Randomized controlled trial

Oral EA et al. Cell Metab. 2017;26(1):157-170.e7. PMID 28683283.

Amlexanox produced a statistically significant fall in HbA1c and fructosamine. Improvements in insulin sensitivity and hepatic steatosis appeared only in a subset of responders, who had a distinct inflammatory gene-expression signature in subcutaneous fat at baseline. The abstract reports no weight or energy-expenditure measurement; any link to energy expenditure is inferred from fat gene expression.

Participants / model
42 obese adults with type 2 diabetes and nonalcoholic fatty liver disease
Treatment
Oral amlexanox or placebo (regimen in registry record NCT01975935)
Follow-up
12 weeks (registry record)
Study design
Proof-of-concept randomized double-blind placebo-controlled trial
Funding
not stated in the source

Endpoints were HbA1c, fructosamine, insulin sensitivity, hepatic fat and fat gene expression. Only the abstract was read, so the responder definition is not described here.

Read the original source
Registry results: HbA1c 7.68 to 7.44% versus 7.77 to 7.76% (P = 0.05); weight not different

Trial registry record

ClinicalTrials.gov record NCT01975935. Sponsor University of Michigan.

The registry record posts results for Solfa amlexanox tablets 25 mg three times daily for 2 weeks, then 50 mg three times daily for 10 weeks, versus placebo (21 per arm). Mean HbA1c went from 7.68 to 7.44% on amlexanox and from 7.77 to 7.76% on placebo (t-test P = 0.05). MRI liver fat, measured in 20 patients, went from 18 to 16.22% on amlexanox versus 12.82 to 12.88% on placebo (P = 0.38). Weight went from 100.95 to 100.81 kg on amlexanox and from 100.24 to 97.94 kg on placebo (P = 0.51). Serious adverse events occurred in 1 of 21 in each arm.

Participants / model
42 obese adults with type 2 diabetes
Treatment
Oral amlexanox 25 mg three times daily, then 50 mg three times daily, versus placebo
Follow-up
12 weeks
Study design
Registered randomized placebo-controlled trial with posted results
Funding
Sponsor listed: University of Michigan

The HbA1c P value of 0.05 is borderline, and baseline liver fat differed between arms.

Read the original source
Open-label pilot: 7 enrolled, 6 completed, no control group, terminated for lack of funding

Trial registry record

ClinicalTrials.gov record NCT01842282.

An earlier open-label single-arm pilot of Solfa tablets for 12 weeks (7 enrolled, 6 completed; terminated for lack of funding) posts mean changes of 0 in HbA1c (range -0.6 to 0.6), -2.2 percentage points in MRI liver fat (range -5.9 to 0.3) and -2.2 kg in weight (range -5.7 to -0.1). With no control group, these changes cannot be attributed to the drug, and the arm description gives no dose.

Participants / model
6 obese adults with type 2 diabetes, aged 18 to 59
Treatment
Oral amlexanox (Solfa tablets), open label
Follow-up
12 weeks
Study design
Registered open-label single-arm pilot with posted results
Funding
not stated in the source

No control group and a dose not given in the arm description, so the changes cannot be attributed to amlexanox.

Read the original source
PubChem CID 2161: C16H14N2O4, 298.29 g/mol

Chemical identity record

National Library of Medicine. PubChem compound record CID 2161 (amlexanox).

Amlexanox: PubChem CID 2161, C16H14N2O4, 298.29 g/mol, CAS 68302-57-8 (label molecular weight 298.30). The US product was a 5% paste whose systemic exposure came mainly from swallowed drug, while the metabolic trials used oral Solfa tablets, so paste PK does not transfer to oral tablets or vendor powder.

Participants / model
Chemical identity, not a clinical population
Treatment
Defined molecule
Follow-up
Not applicable
Study design
NLM chemical record
Funding
not stated in the source

Identity record only; it says nothing about efficacy or safety.

Read the original source

Why is Amlexanox in B tier?

B fits a drug with a past FDA approval and vehicle-controlled ulcer studies, plus a mouse mechanism that is well supported. In the one controlled human metabolic trial, 42 obese adults with type 2 diabetes had a small HbA1c fall (registry P = 0.05) concentrated in responders and no weight loss versus placebo. The grade covers the approved paste and the partial diabetes signal, not the supplier's metabolic-rate and fat-loss claims.