Keratoacanthoma sits in one of dermatology’s most stubborn gray zones. It looks and grows like a skin cancer, shares some of the same genetic mutations as squamous cell carcinoma, and under a microscope can be nearly impossible to tell apart from one. Yet unlike a true malignancy, it typically stops growing on its own and shrinks back to nothing within a few months. Most experts today treat keratoacanthoma as a low-grade or borderline tumor rather than assigning it neatly to either camp, and in practice it is almost always removed because leaving a potential squamous cell carcinoma untreated carries real risk.
Why the Classification Remains Unsettled
The debate over whether keratoacanthoma (KA) is benign or malignant has been running for decades, and it is not close to being resolved. One camp points to the tumor’s self-limiting behavior: KAs grow fast, plateau, and then regress without treatment, which no aggressive cancer does. Another camp argues that KAs are just well-differentiated squamous cell carcinomas (SCCs) that happen to regress, and that labeling them “benign” creates a dangerous false sense of security. A 2024 review in the dermatopathology literature described “considerable controversy” in differentiating KA from SCC, noting that while many regard KAs as benign neoplasms where intervention may be unnecessary, SCC carries significant risks of spreading and causing harm if left alone.1PubMed Central. Keratoacanthoma versus Squamous-Cell Carcinoma: Histopathological Features and Molecular Markers
One older but still-cited description calls KA “the prototype of cutaneous pseudomalignancies” and frames it as “an aborted malignancy that only rarely progresses into an invasive squamous cell carcinoma.”2Journal of the American Academy of Dermatology. Keratoacanthoma That phrase, “aborted malignancy,” captures the tension well. The tumor starts down the road toward cancer, using some of the same molecular machinery, but then something internal hits the brakes. Whether you call that process benign or malignant depends largely on which part of the journey you emphasize.
How a Keratoacanthoma Grows and Regresses
A keratoacanthoma follows a remarkably predictable life cycle compared to other skin tumors. It passes through three recognizable stages: a proliferative phase, a fully developed or mature phase, and an involuting phase where it shrinks and eventually disappears. Within weeks, a KA goes from a rapidly growing, firm, smooth nodule into a dome-shaped tumor with a central plug of hard keratin, and finally degenerates into a shrinking keratinous mass that flattens out and leaves a scar.3PubMed. The keratoacanthoma: a review The whole cycle from first appearance to resolution typically takes two to six months, though some lesions hang around longer.
This spontaneous regression is the single feature that most strongly separates KA from conventional squamous cell carcinoma. SCCs do not shrink on their own. They grow, persist, and if untreated can invade deeper tissue or spread to lymph nodes. KAs, by contrast, are described as “well-differentiated squamoproliferative skin lesions that grow rapidly and regress spontaneously.”4PubMed. Differentiating keratoacanthoma from squamous cell carcinoma by the use of apoptotic and cell adhesion markers The catch is that you usually cannot be certain a lesion is a KA until it starts regressing, and by then you have already spent weeks or months watching what could have been a cancer.
What Makes a Keratoacanthoma Regress
Two mechanisms appear to drive self-resolution, and they work through different pathways. The first is an internal developmental switch. Research using mouse models and human tissue has shown that regressing KAs shift from a growth program to a differentiation program, essentially telling their cells to mature and stop dividing. This process hijacks signaling pathways normally used in hair follicle regeneration. During growth, Wnt signaling sustains the tumor. During regression, retinoic acid signaling ramps up and suppresses Wnt, tipping the balance toward differentiation and tumor collapse.5PubMed Central. Spontaneous tumour regression in keratoacanthomas is driven by Wnt/retinoic acid signalling cross-talk
The second driver is the immune system. When researchers compared the immune environment inside KAs and SCCs, they found that KAs had significantly higher numbers of activated T cells infiltrating the tumor. These T cells expressed markers of active immune engagement at higher rates than those found in SCCs. Adhesion molecules that help immune cells stick to and attack tumor cells were also more abundant in KAs.6PubMed. Evidence that regression in keratoacanthoma is immunologically mediated: a comparison with squamous cell carcinoma The implication is that the immune system recognizes KA cells more effectively than SCC cells and mounts a stronger targeted attack. This aligns with the clinical observation that immunosuppressed patients, such as organ transplant recipients, develop more KAs and may have trouble resolving them.
The Shared Mutations That Complicate Everything
If the regression story makes KA sound reassuringly benign, the genetic evidence complicates that picture considerably. When researchers sequenced growing KAs at the molecular level, they found the same cancer-driving mutations that fuel cutaneous squamous cell carcinoma. Growing KAs carried mutations in genes like TP53, PIK3CA, and NOTCH1, all well-known oncogenic drivers in skin cancer. These were not random background mutations; they were specific, functionally significant variants known to promote uncontrolled cell growth.7Journal of Investigative Dermatology. Keratoacanthoma Shares Driver Mutations with Cutaneous Squamous Cell Carcinoma
Interestingly, none of the regressing KAs in that same study carried these driver mutations. The researchers concluded that sporadic KAs show a lower overall mutation burden than SCCs and lack the large chromosomal deletions common in true carcinomas. But the fact that growing KAs can harbor identical oncogenic drivers to SCCs makes it genuinely difficult to tell them apart at the molecular level during the growth phase, which is exactly when a clinician needs to make a treatment decision. The study’s authors put it bluntly: identifying biomarkers to differentiate the two tumors “may be challenging” precisely because they share the same oncogenic machinery during active growth.7Journal of Investigative Dermatology. Keratoacanthoma Shares Driver Mutations with Cutaneous Squamous Cell Carcinoma
How Pathologists Try to Tell Them Apart
Given the shared genetics, much of the diagnostic burden falls on pathologists examining tissue under the microscope and using special staining techniques. Several protein markers have shown promise in distinguishing KA from SCC, though none is perfect on its own.
A 2023 study examining fully excised tumors found that a combination of staining patterns for three proteins, p16, p53, and Ki-67, could separate KA from SCC with high accuracy. KAs displayed a characteristic mosaic pattern for p16 and a graded peripheral pattern for p53 in about 92% of cases, a combination that appeared in none of the SCCs tested. Conversely, highly abnormal staining patterns for p16 and p53 were present in roughly 84 to 91% of SCCs but in zero percent of KAs. Abnormal distribution of Ki-67, a marker of cell division, beyond the outermost cell layers was rare in KAs (about 4%) but common in SCCs.8PubMed. Aberrant p16, p53 and Ki-67 immunohistochemistry staining patterns can distinguish solitary keratoacanthoma from cutaneous squamous cell carcinoma
Another study evaluated a different protein, cytokeratin 17 (CK17), and found that a central staining pattern for CK17 within tumor clusters was 92% sensitive and 70% specific for identifying KA. Meanwhile, a diffuse staining pattern for Ki-67 identified SCC with 81% sensitivity and 100% specificity.9PubMed Central. Cytokeratin 17 and Ki-67: Immunohistochemical markers for the differential diagnosis of keratoacanthoma and squamous cell carcinoma These are encouraging numbers, but they depend on having adequate tissue to examine.
And that leads to a practical problem. The architectural features that help a pathologist distinguish KA from SCC require seeing the full structure of the tumor, including its relationship to the surrounding skin. Partial or shave biopsies, the quick sampling methods most commonly used in outpatient dermatology, often cannot capture these features.10DermNet. Keratoacanthoma A superficial biopsy that clips the top of the lesion may show atypical squamous cells without the broader context needed to distinguish KA from SCC, leading the pathologist to default to a diagnosis of squamous cell carcinoma. This is one reason KAs are probably over-diagnosed as SCCs in routine clinical practice.
Differences in the Tumor’s Blood Supply
Beyond immune activity and protein markers, researchers have also examined how KAs and SCCs build their blood supply. Both tumors promote the growth of new blood vessels, but the relationship between blood vessel density and tumor cell proliferation differs between them. In KAs, the formation of new blood vessels was closely linked to how actively the tumor cells were dividing. In SCCs, that correlation did not hold, suggesting the two tumors rely on different pathways to sustain their growth.11PubMed Central. Keratoacanthoma and Keratoacanthoma-Like Squamous Cell Carcinoma: Similar Morphology but Different Pathogenesis The lower cell division rate in KAs, combined with their apparent dependence on new blood vessel formation, points to a fundamentally different growth strategy, one that may be easier for the body’s own defenses to shut down.
Why Most Doctors Remove Them Anyway
Given all the uncertainty, the standard clinical recommendation is to excise KAs surgically rather than wait for spontaneous regression. A clinical and biological review in the British Journal of Dermatology explained the rationale plainly: because the similarities between KA and well-differentiated SCC make definitive differentiation difficult or impossible in many cases, the ambiguity has led to a general recommendation for surgical excision to ensure a potentially malignant SCC is not left untreated.12British Journal of Dermatology. A clinical and biological review of keratoacanthoma
This approach is pragmatic rather than elegant. Most KAs would probably resolve on their own, but “probably” is not good enough when the alternative diagnosis is a cancer that can invade and metastasize. The cosmetic outcome of waiting for a KA to regress is often poor anyway, since the involuting tumor can leave a depressed, irregular scar. Excision removes the lesion, provides a complete tissue sample for the pathologist, and typically heals with a more predictable scar.
Non-Surgical Options When Surgery Is Difficult
Not every KA is in a convenient spot for surgery. Lesions on the nose, ears, lips, or near the eyes can be challenging to excise without cosmetic or functional compromise, especially if the tumor is large. For these cases, several non-surgical approaches have shown good results.
Injecting medication directly into the tumor is the best-studied alternative. A comparative trial of intralesional methotrexate versus intralesional 5-fluorouracil found complete clearance in 70% of patients receiving methotrexate and 80% receiving 5-fluorouracil, with the 5-fluorouracil group needing fewer injections (a median of two sessions versus three).13PubMed Central. Intralesional methotrexate versus 5-flurouracil in the treatment of keratoacanthoma A broader review found that both topical and intralesional treatments achieved KA eradication rates between 92 and 100%, with intralesional 5-fluorouracil leading to faster healing (about 3.7 weeks) compared to methotrexate (about 4.6 weeks).14PubMed. Relative Efficacy of Nonoperative Treatment of Keratoacanthomas These numbers are reassuring, though clinical decisions still depend on confirming the diagnosis and ruling out true SCC.
What Triggers Keratoacanthomas
Several factors have been linked to KA development. Ultraviolet light exposure is the most consistent risk factor, which explains why KAs tend to appear on sun-exposed skin, particularly the face, forearms, and hands. Genetic susceptibility, immunosuppression, certain viruses, and physical trauma have all been implicated as potential triggers as well.15Journal of the American Academy of Dermatology. Keratoacanthoma developing in sites of previous trauma: A report of two cases and review of the literature The trauma connection is interesting because KAs have been documented arising at sites of previous surgery, burns, or even tattoos, suggesting that skin disruption can set the stage for the tumor to develop.
A more recently recognized trigger involves certain cancer medications. BRAF inhibitors, drugs used to treat melanoma by blocking a specific growth-promoting protein, have been shown to induce KAs and squamous cell carcinomas as a side effect. Vemurafenib, one of the most widely used BRAF inhibitors, can cause these lesions to erupt on previously normal skin during treatment.16PubMed Central. Diverse cutaneous side effects associated with BRAF inhibitor therapy: a clinicopathologic study The paradox of a cancer drug triggering skin tumors has been well documented and is thought to involve activation of an alternative growth pathway in skin cells when the drug’s primary target is blocked.
When Multiple Keratoacanthomas Appear at Once
A solitary KA on sun-damaged skin is the most common presentation, but some people develop multiple KAs, and these cases often signal an underlying genetic syndrome. Several named syndromes involve multiple KA-like lesions, including Ferguson-Smith syndrome (also called multiple self-healing squamous epithelioma), eruptive keratoacanthoma of Grzybowski, multiple familial keratoacanthoma of Witten and Zak, and Muir-Torre syndrome.17PubMed Central. Multiple Keratoacanthoma-like Syndromes: Case Report and Literature Review
These syndromes differ in important ways. Ferguson-Smith syndrome is inherited in an autosomal dominant pattern and involves tumors that grow and regress repeatedly over a lifetime. Muir-Torre syndrome carries the added concern of visceral cancers, particularly of the colon and genitourinary tract, because it involves defects in DNA mismatch repair genes. For anyone presenting with multiple KAs, especially at a young age or in unusual locations, genetic evaluation is worth pursuing because the skin lesions can be the first visible sign of a systemic cancer predisposition.
Keratoacanthomas in Animals
Keratoacanthomas are not exclusive to humans. Dogs, horses, and other animals develop histologically similar tumors, sometimes called infundibular keratinizing acanthomas in veterinary pathology. In dogs, these growths behave in much the same way as their human counterparts: they form a keratin-filled nodule that is typically benign and can be cured with surgical removal. Veterinary pathologists use the same diagnostic tools, including Ki-67 and p53 staining, to confirm the diagnosis and distinguish these tumors from more aggressive skin cancers.18PubMed Central. Canine nail bed keratoacanthoma diagnosed by immunohistochemical analysis The existence of KA-like tumors across species suggests that the underlying biology, a squamous proliferation that initiates like a cancer but stalls and regresses, is a conserved pattern rather than a quirk of human skin.