Aphthous ulcers

Meet the patient

A 24-year-old teacher returns for the fourth time in a year with two painful round ulcers on the inside of her lower lip. Each time they came on over a day of burning, lasted about ten days, and healed completely without a scar. She is otherwise well, takes no regular medication, and has no genital, eye, skin or bowel symptoms.[1][5]

That is the textbook minor aphthous ulcer, and it is the exception that earns its easy diagnosis. The examiner's job — and yours — is to recognise when the pattern strays: ulcers that last weeks, scar, cluster in dozens, appear on the hard palate or attached gingiva, or travel with genital, ocular, skin or gut symptoms. Every section below is built to answer one question: when is a mouth ulcer not just a mouth ulcer?[3][5]

What RAS is — and the three things it is not

RAS is a clinicopathological pattern, not a single disease. A susceptible mucosal barrier develops episodic, T-cell-driven ulceration on movable non-keratinized mucosa — buccal and labial mucosa, floor of mouth, ventral tongue, soft palate. The diagnosis is deliberately narrow and you should be able to say it in one breath.[1][3]

It is not herpes. Aphthae are ulcers from the outset; there is no vesicular stage and they favour movable mucosa. Herpes simplex begins as grouped vesicles and, when it recurs intra-orally, favours keratinized mucosa — hard palate and attached gingiva. That single distinction is the highest-yield discriminator in the viva.[3][5]

It is not every painful mouth ulcer. A single traumatic ulcer after cheek-biting is not RAS. A grouped vesicular eruption on the hard palate is not RAS. A solitary ulcer persisting for weeks in a smoker is oral cancer until proven otherwise. The label is safe only after morphology, site, tempo and recurrence have been matched to aphthae and the red flags actively sought.[3][5]

Minor, major, herpetiform — the three faces

The three classical subtypes determine healing time, scarring risk, and your threshold to investigate. Minor is the everyday pattern; major and herpetiform are the patterns that make the examiner ask about HIV, Behçet, IBD, coeliac, neutropenia and haematinic deficiency.[1][2]

[1] [2]

The classic trap: the name herpetiform is the most misleading word in oral medicine. These ulcers are not caused by herpes simplex, do not start as vesicles, and are not improved by aciclovir. If there genuinely were vesicles, or the lesions sit on keratinized mucosa, send HSV PCR — do not trust the name.[1][3]

A practical fourth label, complex aphthosis, describes almost-constant oral aphthae, recurrent genital aphthae, or major/herpetiform patterns that behave severely but do not yet meet criteria for Behçet disease. It is a management and referral category: it warns you not to reassure too quickly, because repeated genital ulcers, ocular symptoms, skin lesions, gut symptoms or fever cycles may declare a systemic disorder over time.[2][6]

Etymology for viva gold: aphtha is ancient Greek for "I set on fire" — a word that has survived two and a half millennia because the burning pain still tells the story before the lesion does. Stutton disease (major aphthae) is named after the surgeon who first described the deep, scarring variant in the nineteenth century.[1]

Why the mouth ulcerates — the final common pathway

RAS is best understood as mucosal barrier vulnerability plus dysregulated immune injury. A local trigger appears to expose epithelial antigens, stress signals or microbial products to a primed immune system. Th1-skewed T cells recruit macrophages and neutrophils, epithelium necroses, fibrin deposits, and the familiar yellow-white pseudomembrane forms over the defect.[1][4]

The yellow-white centre is not pus. It is fibrinous exudate and necrotic debris over an epithelial defect; the red halo is adjacent inflammatory hyperaemia. This is why chlorhexidine may reduce secondary colonisation but does not treat the primary immune driver, why topical corticosteroids started early can shorten attacks, and why antivirals do nothing for true aphthae even when the herpetiform subtype looks like herpes.[1][3]

The microbial story is nuanced. No single bacterium or virus causes idiopathic RAS, but the oral microbiome, epithelial barrier and mucosal immune system communicate continuously; dysbiosis may lower the threshold for inflammation. The safe viva answer is therefore "non-infectious and immune-mediated" rather than "sterile" — and routine bacterial swabs are unhelpful unless secondary infection, candidiasis or immunocompromise is suspected.[4]

Risk factors and triggers — the modifiable contributors

Elicit triggers as modifiable contributors, not as assumed causes. Local trauma (cheek biting, a sharp tooth, orthodontic appliance, vigorous brushing, recent dental work) can initiate an ulcer in susceptible mucosa. Emotional stress, sleep deprivation and intercurrent illness are common patient-reported triggers.[1][2]

Sodium lauryl sulfate (SLS) toothpaste worsens ulcer frequency in some patients, probably by irritating or altering the barrier; an SLS-free trial is a cheap, harmless intervention. Acidic and spicy foods, nuts, chocolate, cheese, tomatoes and citrus may aggravate pain and sometimes act as reproducible triggers — but blanket dietary restriction is less useful than a symptom diary.[1][2]

[1] [5]

The haematinic hunt — common enough to matter, treatable enough to miss

Iron, B12, folate (and sometimes zinc) deficiency is the examiner's favourite because it is common and correctable. Look for angular cheilitis, glossitis, fatigue, menorrhagia, gastrointestinal symptoms, a restricted or vegan diet, bariatric surgery, or malabsorption — any of these should lower the threshold to test. Correcting a true deficiency is disease-modifying; supplementing a normal patient is not the same thing.[2][8]

The systemic associations are not rare zebras in examinations. Behçet disease is the classic vasculitis association; inflammatory bowel disease can produce aphthous-like oral ulcers; coeliac disease may present with oral aphthae and iron deficiency; HIV and other immunodeficiency states produce large, persistent, treatment-resistant ulcers; and cyclic neutropenia produces periodic fever, infections and oral ulcers. Medication-related ulceration — especially nicorandil, NSAIDs, bisphosphonates or cytotoxic drugs — should be considered when the temporal history fits.[2][6][9]

The differential — the mouth has few ways to show many diseases

The differential is the section that separates a safe clinician from a candidate who memorised one line. Infection, trauma, autoimmune blistering disease, inflammatory dermatoses, cancer, nutritional deficiency and systemic inflammation can all produce painful erosions. The discriminators are site, onset, vesicles, number, recurrence, systemic symptoms, induration, healing time and the background mucosa.[3][5]

[1] [2] [3] [5]

The discriminator line: vesicles first and keratinized mucosa means HSV; ulcers from the outset on movable mucosa means aphthae. When the patient is immunocompromised, when lesions are atypical, or when the distinction changes treatment, swab early for HSV PCR rather than guessing.[3][5]

The do-not-miss mimic is oral squamous cell carcinoma. Pain is not reassuring — cancers hurt once they ulcerate or infect. A non-healing ulcer beyond two to three weeks, with induration, rolled border, spontaneous bleeding, erythroplakia, leukoplakia, fixation, dysphagia, referred otalgia, a neck mass, or tobacco, alcohol or betel exposure, needs urgent specialist assessment. Reassuring the patient that "it is probably aphthous" without safety-net review is the classic pitfall.[3][5]

The bedside script — say it out loud

Bedside assessment begins before any investigation. Inspect the entire oral cavity with good light, gloves and a tongue depressor — lips, labial and buccal mucosa, gingiva, floor of mouth, ventral and dorsal tongue, lateral tongue, hard and soft palate, tonsillar pillars and oropharynx. Note size, depth, border, base, number, site, symmetry and surrounding mucosa, and whether lesions are ulcers, erosions, vesicles, plaques or bullae.[3][5]

The systemic questions earn their keep. Behçet screening covers genital ulcers, eye pain, photophobia, blurred vision, floaters, erythema nodosum, acneiform lesions, pathergy, arthralgia, neurological and vascular symptoms. Gastrointestinal screening covers chronic diarrhoea, abdominal pain, rectal bleeding, weight loss and perianal disease. Haematinic screening covers fatigue, pallor, pica, heavy menses and dietary risk. Neutropenia screening covers recurrent fevers, bacterial infections and striking periodicity.[2][6][8]

The pathergy test supports Behçet disease when a sterile needle prick yields a papule or pustule at 24-48 hours — but sensitivity varies by ethnicity and technique, and a negative test does not exclude it. Ocular symptoms are time-critical: uveitis and retinal vasculitis can threaten vision, so same-day ophthalmology referral is the correct reflex.[6]

When to investigate — calibrate to risk

Simple, infrequent minor RAS in a well patient with a long stable pattern needs no panel. Investigate when ulcers are frequent, severe, major, herpetiform, adult-onset, persistent, treatment-resistant, associated with systemic symptoms, or occurring in an immunocompromised patient. A scattergun panel is less impressive than a clear rationale.[1][2]

[1] [2] [3] [5] [8]

Histopathology is not required for classic minor RAS, and when biopsied an aphthous ulcer is non-specific — ulceration with fibrinopurulent exudate, mixed inflammation and granulation tissue. Biopsy is performed not to "prove RAS" but to exclude malignancy, immunobullous disease, lichen planus, infection or granulomatous disease. If pemphigus or pemphigoid is suspected, send perilesional mucosa for direct immunofluorescence, not only the necrotic ulcer bed.[3][5]

Management — the stepwise ladder

Management has five aims: reduce pain, shorten duration, reduce recurrence, identify treatable triggers, and not miss systemic disease or cancer. Most patients need explanation and topical therapy, not systemic immunosuppression. Tell them RAS is common, non-contagious, not poor hygiene and not HSV — and give them clear reasons to return.[1][2]

For mild infrequent disease: a soft toothbrush, correction of dental trauma, avoidance of reproducible food triggers, an SLS-free toothpaste trial, and topical pain control. Chlorhexidine mouthwash can reduce secondary colonisation and severity, but it stains teeth and alters taste, so use it as a short course. Benzydamine or lidocaine gel helps eating — but anaesthetised mucosa can be bitten, and children need careful dosing.[1][2][3]

Topical corticosteroids are first-line pharmacological therapy for frequent, moderate or painful RAS. The principle is early local anti-inflammatory treatment, ideally during the prodrome or at onset.[1][2]

[1]

Tetracycline mouthwash (doxycycline or tetracycline capsule contents dispersed in water, used as a rinse-and-spit for several minutes) is sometimes used for recurrent or herpetiform disease through anti-inflammatory and antimicrobial effects. It is not for young children or pregnancy, can irritate, and should not become a reflex prescription before topical steroids and trigger correction have been optimised.[2]

Systemic corticosteroids are rescue therapy, not chronic management. A short oral prednisolone course may rescue severe major aphthae causing major pain or poor intake, but infection and malignancy must be excluded when the presentation is atypical. Repeated steroid bursts without a diagnosis are a warning that the patient needs oral medicine, dermatology or rheumatology review.[2][5]

Refractory major aphthae, complex aphthosis or Behçet-associated disease may need steroid-sparing therapy. Colchicine is often first for mucocutaneous Behçet-like disease (watch for GI intolerance and renal/hepatic dosing). Dapsone helps neutrophil-mediated ulceration but requires G6PD testing and FBC/liver monitoring — haemolysis, methaemoglobinaemia and agranulocytosis are serious risks. Thalidomide is highly effective in HIV-associated severe aphthae but teratogenicity, neuropathy, sedation and thrombosis make it specialist-only with strict pregnancy prevention.[2][7]

Treat deficiencies and associated disease at the source. Iron, B12, folate or zinc replacement when deficiency is demonstrated — and seek the cause. Coeliac disease needs a gluten-free diet and gastroenterology follow-up, not oral gels. HIV-associated ulcers need antiretroviral optimisation and exclusion of HSV, CMV and fungal disease. Crohn oral ulcers improve with control of the gut. Behçet needs organ-based management — especially urgent eye care when uveitis is suspected.[2][6][8][9]

The systemic hunt — the patterns that pay off

Behçet disease is suspected when recurrent oral ulcers travel with genital ulcers, ocular inflammation, erythema nodosum, acneiform lesions, pathergy, arthritis, neurological symptoms or vascular disease. Oral ulcers are usually the earliest and most frequent manifestation, so the dermatologist often sees the patient before the systemic diagnosis is obvious. Eye symptoms are time-critical.[6]

IBD and coeliac disease belong on the list when aphthae accompany chronic diarrhoea, abdominal pain, rectal bleeding, weight loss, iron deficiency, growth faltering, perianal symptoms or a family history. Aphthous-like ulcers can precede the gastrointestinal diagnosis. Do not start a gluten-free diet before coeliac testing if you intend to pursue the diagnosis.[2][8][9]

HIV-associated aphthous ulcers are often large, deep, persistent and painful, especially with advanced immunosuppression, and opportunistic infections mimic aphthae — so HSV, CMV, candidiasis and malignancy must be considered. Thalidomide showed efficacy for oral aphthae in HIV in a randomised trial, but modern use is constrained by toxicity and sits firmly in specialist care alongside antiretroviral optimisation.[2][7]

Cyclic neutropenia presents with recurrent fever, malaise, bacterial infections, gingivitis or oral ulcers at roughly three-week intervals. A single normal neutrophil count does not exclude it if the timing is wrong — serial FBC with differential across several weeks may be needed. This is a favourite viva corner because the mouth ulcer is the visible clue to a haematological rhythm.[2][5]

Special populations

Children commonly present with mouth ulcers, but RAS is only one possibility. Primary HSV gingivostomatitis causes fever, diffuse gingivitis, vesicles and ulcers and may dehydrate. Hand-foot-mouth disease causes oral lesions plus acral or buttock rash in outbreaks. PFAPA causes periodic fever, aphthae, pharyngitis and cervical adenitis. Always assess hydration, fever, systemic toxicity and the ability to drink.[3][5]

Pregnancy demands conservative treatment. Trigger avoidance, chlorhexidine, pregnancy-compatible analgesia and carefully selected topical corticosteroids are preferred. Tetracycline mouthwash is generally avoided. Thalidomide is absolutely contraindicated. Never list systemic agents without first checking pregnancy status.[2][7]

Older adults deserve a medication and malignancy lens. New-onset oral ulceration later in life is less typical for idiopathic RAS — review nicorandil, NSAIDs, bisphosphonates, cytotoxics and appliances, examine for denture trauma and xerostomia, and keep a low threshold for biopsy of persistent or atypical ulcers, especially with tobacco, alcohol or betel exposure.[3][5]

Immunocompromised patients are high risk for both severe aphthous-like disease and mimics. HSV, CMV, deep fungal infection, candidiasis-associated erosions, drug ulcers and malignancy all enter the differential. If topical steroids are used, follow-up is closer; if systemic steroids are contemplated, exclude infection and involve the specialist.[2][5][7]

Preventable harm — the mistakes that cost patients

[1]

A subtle pitfall is over-investigation of straightforward disease and under-investigation of atypical disease. The patient with a ten-year history of two minor ulcers a year does not need repeated autoimmune panels. The patient with new adult-onset major ulcers, weight loss and anaemia does. Calibrate investigation to risk.[2][5]

Evidence and regional differences

The evidence base is a mixture of clinical reviews, small randomised trials, specialist experience and extrapolation from Behçet and HIV-associated aphthae trials. Topical corticosteroids are widely accepted first-line for frequent or painful disease because they target the inflammatory phase with low systemic exposure; chlorhexidine and topical anaesthetics are symptomatic adjuncts; systemic agents carry more toxicity and less generalisable evidence, so they stay specialist-only.[1][2][5]

[7]

Ward-round test

[1]

1. A 28-year-old has dozens of 2 mm painful ulcers on the buccal mucosa, no vesicles, recurring monthly. What are they, and what must you NOT prescribe?[1][3]

Answer

Herpetiform aphthae — clustered, non-vesicular, on movable mucosa. Do not prescribe aciclovir: the name mimics herpes, the cause does not. Consider HSV PCR only if there were vesicles or keratinized-mucosa involvement.[1][3]

2. A 55-year-old smoker has a solitary ulcer on the lateral tongue for five weeks, firm base, no pain relief. What is the single next step?[3][5]

Answer

Urgent oral medicine, ENT or maxillofacial referral for biopsy to exclude oral squamous cell carcinoma. Pain is not reassuring; induration, a rolled edge, fixation and a neck node make cancer the working diagnosis until proven otherwise. Do not extend a trial of topical steroids.[3][5]

3. Recurrent oral aphthae plus genital ulcers, blurred vision and erythema nodosum. Name the syndrome and the urgent referral.[6]

Answer

Behçet disease. The urgent referral is same-day ophthalmology — uveitis and retinal vasculitis can threaten vision. Mucocutaneous disease may be treated with topical steroids, colchicine or apremilast; organ-threatening disease needs specialist immunosuppression.[6]

4. Frequent major aphthae, ferritin 12, menorrhagia. First-line drug therapy and the disease-modifying step?[2][8]

Answer

First-line drug therapy is a topical corticosteroid started early (triamcinolone 0.1% paste or a betamethasone 0.5 mg rinse). The disease-modifying step is iron replacement for the deficiency plus investigation of the menorrhagia — correcting a true deficiency is disease-modifying; supplementing a normal patient is not.[2][8]

5. A child has mouth ulcers with fever every three weeks and recurrent bacterial infections. Which test, and why serial rather than once?[2][5]

Answer

Serial FBC with differential across several weeks to catch the neutrophil nadir of cyclic neutropenia. A single normal count at the wrong point in the cycle does not exclude it — the mouth ulcer is the visible clue to a haematological rhythm.[2][5]

References

1. [1]Gasmi Benahmed A, Noor S, Menzel A, et al. Oral Aphthous: Pathophysiology, Clinical Aspects and Medical Treatment Arch Razi Inst, 2021.PMID 35355774
2. [2]Saikaly SK, Saikaly TS, Saikaly LE. Recurrent aphthous ulceration: a review of potential causes and novel treatments J Dermatolog Treat, 2018.PMID 29278022
3. [3]Hargitai IA. Painful Oral Lesions Dent Clin North Am, 2018.PMID 30189985
4. [4]Lin D, Yang L, Wen L, et al. Crosstalk between the oral microbiota, mucosal immunity, and the epithelial barrier regulates oral mucosal disease pathogenesis Mucosal Immunol, 2021.PMID 34040155
5. [5]Stoopler ET, Villa A, Bindakhil M, et al. Common Oral Conditions: A Review JAMA, 2024.PMID 38530258
6. [6]Alibaz-Oner F, Direskeneli H. Update on the Diagnosis of Behçet's Disease Diagnostics (Basel), 2022.PMID 36611332
7. [7]Jacobson JM, Greenspan JS, Spritzler J, et al. Thalidomide for the treatment of oral aphthous ulcers in patients with human immunodeficiency virus infection. National Institute of Allergy and Infectious Diseases AIDS Clinical Trials Group N Engl J Med, 1997.PMID 9154767
8. [8]Koparal M, Ege B, Dogan EI, et al. Evaluation of biochemical variables in patients with recurrent aphthous stomatitis J Stomatol Oral Maxillofac Surg, 2023.PMID 36162803
9. [9]Sahin Y. Celiac disease in children: A review of the literature World J Clin Pediatr, 2021.PMID 34316439