At a glance
A pregnancy tumor is a benign pyogenic granuloma that can appear on the gums, the nose or a finger. What the studies show about who gets them, which ones shrink after delivery, and why excision during pregnancy carries a higher recurrence rate.
A nodule on a finger, and four possible answers
In the November–December 2010 issue of Annals of Saudi Medicine, a dermatology resident at McGill published a diagnostic quiz built from a single patient. A 27-year-old woman, 28 weeks pregnant, had come in with a rapidly growing nodule on her left index finger that had been bleeding on its own for four weeks. The photograph showed a pedunculated vascular growth. Readers were offered four answers: malignant melanoma, pyogenic granuloma, basal cell carcinoma, orf.
The answer was pyogenic granuloma, the lesion most clinicians call a pregnancy tumor. It is worth pausing on why the question needed asking at all. A beefy red nodule that appears within weeks and bleeds when nothing has touched it is, in an adult, melanoma until somebody proves otherwise. That the correct answer turned out to be a benign reactive overgrowth is reassuring. That the two can look alike is the entire clinical problem.
The case carries a second lesson that patient-facing writing on this subject usually drops. Cutaneous pyogenic granulomas, the author wrote, do not involute spontaneously. Whatever a pregnant woman is told about the lump on her gum shrinking after delivery, that promise does not transfer to a lump on her hand.
Three names for one lesion, and none of them is accurate
Here the word "tumor" means swelling, not cancer. The same growth is called a pregnancy epulis, a granuloma gravidarum, or — the term pathologists tend to prefer — a lobular capillary haemangioma. The oldest name, pyogenic granuloma, was introduced by Hartzell in 1904 and is wrong on both counts: the lesion contains no pus and is not a granuloma. What sits under the microscope is a proliferation of small blood vessels set in an oedematous stroma, infiltrated by chronic inflammatory cells, arising where long-standing minor irritation meets a changed hormonal environment.
Hullihen described the first case in 1844. It has taken the intervening years to settle that the lesion is reactive rather than infective — a conclusion with practical consequences, because it means the treatment is not an antibiotic.
The irritant is mundane. Calculus, plaque at the gingival margin, a rough filling, a badly fitting crown, a fingernail, a nose-picking habit, the friction of a denture. Take a pregnant woman with a modest amount of plaque and add gestational hormones, and the tissue can respond with granulation tissue out of all proportion to the insult. Two Iranian investigators found defective restorations or crowns adjacent to nine of 72 lesions in one series.
The prevalence figures disagree, and the disagreement is the interesting part
The number quoted most often is 5%: pyogenic granuloma of the gingiva is said to develop in up to 5% of pregnancies. Where investigators examine every pregnant woman in a clinic and count what they find, the figure moves considerably. A Tehran study of 1,600 pregnant women found 72 lesions, 4.5%. A Tripoli multicentre study of 402 women found four, 1.0%. A Hungarian series that examined 2,260 women found 12, or 0.48%.
One study explains the spread. Iranian investigators who required histological confirmation of every lesion before counting it reported a prevalence well below the clinical surveys, and pointed at the reason: studies that diagnose by eye alone overcount, because a peripheral giant cell granuloma or a peripheral ossifying fibroma looks much like a pregnancy tumor until somebody puts it under a microscope. That biopsy-confirmed survey is a useful corrective to a figure everyone repeats. It is not that 5% is wrong; it is that 5% answers a vaguer question than readers assume.
Two findings recur across populations. Lesions favour the maxillary gingiva — 66 of the 72 Tehran cases — particularly the interdental papilla between the front teeth. And they cluster late: 69.4% of the Tehran cases presented in the third trimester, and all four Libyan cases did. Poor oral hygiene was recorded in 80.6% of the Tehran group. Whether hygiene causes the lesion or merely permits it is a distinction the surveys were not designed to make.
Photo by Juan Encalada on Unsplash
Hormones are the obvious explanation, which is not the same as the correct one
The standard account runs as follows. Oestrogen and progesterone rise; gingival tissue becomes more permeable and more reactive to plaque; progesterone raises signalling through vascular endothelial growth factor A and its receptor VEGFR2, the main driver of new blood vessel formation, which produces the endothelial proliferation visible on a slide. Angiostatin, an inhibitor of vessel growth, is expressed at lower levels in these lesions than in healthy gingiva. When hormone levels fall after delivery, the stimulus falls with them and the vessels regress. Lesions provoked by oral contraceptives or retinoid therapy behave the same way when the drug is stopped.
That account explains a great deal and proves rather less than it appears to. When researchers counted oestrogen and progesterone receptors in pyogenic granuloma tissue taken from men, from pregnant women and from non-pregnant women, the counts came out similar — an awkward result for a hormone-driven theory. The likely reconciliation is that receptor expression spikes early in the lesion's life and settles later, which would make the date of the biopsy determine the answer. Nobody has run the study that would settle it, and these lesions are common enough that somebody could.
One molecular detail matters clinically, because it predicts behaviour. If the growth is fed by a signal that pregnancy supplies, removing the signal should starve it. Broadly, that is what happens, and it is the basis for every conservative management plan in the literature.
Most of them shrink; some of them don't
In the Hungarian series of 12 pregnancy epulides, nine resolved spontaneously between one and four months after delivery. A Chinese report describes multiple epulides that did not respond to a single supragingival scaling during pregnancy and then vanished five months postpartum. A Nigerian survey of 15 lesions found five that regressed and ten that required excision. The pattern holds across all of them: regression is common, not universal, and slow.
Against that sits a recurrence figure worth knowing. Excision of a pyogenic granuloma during pregnancy is followed by recurrence in up to 20% of cases, higher than the rates reported when the same operation is done after delivery, and higher than the 3% recurrence reported for surgical excision of pyogenic granulomas in general. Incomplete removal and persistent local irritation account for most of the difference.
| Question | Watch and wait through pregnancy | Excise during pregnancy |
|---|---|---|
| What it involves | Treat the irritant: scaling, plaque control, a soft brush, review each trimester | Complete removal, usually under local anaesthetic, with the specimen sent to histopathology |
| Supporting evidence | Case series in which most lesions regressed one to four months after delivery | Case reports and small series; no randomised comparison exists |
| Recurrence | Not applicable, though lesions can return in a later pregnancy | Up to 20%, with incomplete excision and persisting irritation the usual causes |
| Preferred when | The lesion is small, painless and does not interfere with eating or speech | Bleeding is heavy or uncontrolled, pain is significant, function is impaired, or the diagnosis is in doubt |
| What it cannot tell you | Whether this particular lesion would have shrunk | Whether the operation was necessary |
What a patient should actually do
Show it to somebody. That is the whole of the advice, and it is less trivial than it sounds, because surveys keep finding lesions that have been present for months without examination — often on the assumption that dental treatment during pregnancy is unsafe. It isn't. The American Dental Association's guidance is that preventive, diagnostic, restorative and surgical dental treatment is safe throughout pregnancy, a position the American College of Obstetricians and Gynecologists shares and has reaffirmed.
A few things follow from that, and none of them is exotic.
- Get a professional cleaning. Plaque and calculus are the half of the cause that can be removed in a chair.
- Brush twice daily with a soft-bristled brush and fluoride toothpaste. Scrubbing an inflamed papilla with a hard brush achieves nothing except more bleeding.
- Ask for a review appointment rather than a decision. Many clinicians will measure the lesion each trimester and operate only if it bleeds heavily, hurts, or obstructs eating.
- If anything is removed, insist that the specimen goes to histopathology. The differential includes peripheral giant cell granuloma, peripheral ossifying fibroma and, in rare instances, malignancy.
The single case published in Annals of Saudi Medicine — available in full — makes the last point in the plainest way. The author's advice was that excision with cauterisation of the base is the preferred approach when the clinician is in any doubt about the diagnosis. In doubt is doing a lot of work in that sentence, and it should.
Photo by Anastasiia Chepinska on Unsplash
The tail of the distribution
Most pregnancy-associated pyogenic granulomas are small, gingival and forgettable. A minority are not, and the case reports describing them are worth reading for what they say about how far this lesion can go.
A 2025 report describes a woman with recurrent left-sided nosebleeds and a mass that eventually blocked her left nostril entirely. The lesion arose from the anterior margin of the inferior nasal turbinate, measured 1.7 by 2.0 by 1.4 cm on MRI, and was managed initially with topical decongestants and then silver nitrate cautery. Because of her deteriorating sleep and the anxiety the bleeding caused, she was delivered by Caesarean section at 37 weeks. The mass was gone by twelve months postpartum, and a second pregnancy produced no recurrence. Only about 7% to 10% of pregnancy-associated pyogenic granulomas involve the nasal cavity, which is one reason the diagnosis is often late.
A 2015 report from Taipei describes a 33-year-old at 34 weeks with a lesion on the mandibular gingiva and haemorrhage described as life-threatening. Rather than operate on a highly vascular mass in a late-stage pregnancy, the team embolised the feeding vessels through a microcatheter. The bleeding stopped, the lesion regressed within a month, no excision was needed, and there was no recurrence at one year.
These are the extremes, and the existence of a treatment as specialised as transarterial embolisation should not recalibrate anyone's expectation of the ordinary case. It does, though, mark the boundary of what "benign" covers. A benign lesion can still exsanguinate a patient, obstruct an airway, or frighten a woman badly enough to affect her sleep in the third trimester.
The number nobody has collected
Read the literature end to end and a gap appears that no case report fills. Regression rates after delivery come from small series; recurrence rates after excision during pregnancy come from other small series; the two are almost never measured in the same population with the same definitions. The Tehran figure of 4.5% counts lesions found on examination. The Hungarian figure of 0.48% counts something similar in a different country with different diagnostic habits. Neither tells a woman at 28 weeks what will happen to the specific lump on her gum.
What follows is an honest question rather than a finding. Given that the natural history usually ends in regression, and that excision before delivery carries a higher recurrence rate than excision afterwards, how many of these operations are performed on lesions that would have resolved unaided? Nobody has collected that number, and the case reports that make up most of the evidence base were never designed to produce it. The lesion is common, the operation is minor, and the question has been sitting open for decades. That is usually a sign that the answer would be inconvenient.
- dermatology
- pregnancy-tumor
- pyogenic-granuloma
- lobular-capillary-haemangioma
- oral-health-in-pregnancy

