Recognition and Diagnostic Workup of Suspected Skin Cancer: Screening, Clinical Assessment, Dermoscopy, and Biopsy — A Living Review

Scope, Audience, and Clinical Thesis

This review is written for dermatologists, dermatology residents, Mohs surgeons, primary-care clinicians, advanced practice clinicians, and medically sophisticated readers who want one coherent method for deciding whether a skin lesion should be reassured, photographed, monitored, treated, or biopsied.

The central thesis is simple: skin-cancer diagnosis is a calibrated pathway, not a single visual trick. Good decisions integrate the reason for the examination, the patient's baseline risk, the lesion's history and behavior, comparison with the rest of the skin, palpation, dermoscopy, and the consequences of waiting. Histopathology remains the reference standard when definitive tissue diagnosis is required, but indiscriminate biopsy is not the goal either.^1–3

This article focuses on suspected cutaneous melanoma, basal cell carcinoma, and cutaneous squamous cell carcinoma. It does not provide a population-wide mandate, a consumer-app recommendation, or a universal interval for skin examinations.

Screening and Diagnostic Evaluation Are Different Jobs

The 2023 US Preventive Services Task Force concluded that evidence is insufficient to determine the balance of benefits and harms of clinician visual skin-cancer screening in asymptomatic adolescents and adults without a history of premalignant or malignant skin lesions.¹ That statement is frequently overextended.

It does not apply to a person who presents with a changing, bleeding, painful, growing, or otherwise suspicious lesion. It does not apply to surveillance after skin cancer. It does not erase individualized follow-up for transplant recipients, people with many atypical nevi, strong family histories, or other established risk factors.

No randomized trial has shown that population-based visual screening reduces skin-cancer mortality, and no universal examination interval is supported for every adult.^1,2 The appropriate response is not nihilism. It is to match the intervention to the question:

Who Needs Risk-Stratified Surveillance

Professional surveillance is most defensible when the expected yield is higher or the cost of delay is greater. Factors that can justify a deliberate plan include prior melanoma or keratinocyte carcinoma, many or atypical nevi, strong family history, substantial cumulative ultraviolet exposure, Fitzpatrick I or II phenotype, chronic immune suppression, genetic susceptibility, and difficulty examining one's own skin.^1,3,4

The exact interval remains individualized. Tumor history, number and tempo of new lesions, transplant or hematologic context, medication changes, access to reliable self- or partner examination, and the availability of high-quality baseline images all matter. “Every three months forever” and “once a year for everyone” are both too blunt.

The role and limits of self-examination

Patients and partners often notice the first meaningful change. Self-examination is valuable because it supplies longitudinal information between visits and can trigger evaluation. It is not a diagnostic test.

An older case-control study associated skin self-examination with a large melanoma-mortality reduction, but 20-year follow-up did not confirm that finding.^5,6 The defensible claim is narrower: familiarity with one's own skin can improve awareness and timeliness, while no self-check can prove that a lesion is benign.

ABCDE Is a Prompt, Not a Rule-Out Test

ABCDE remains a useful starting language:

• A — Asymmetry: one half differs from the other.
• B — Border: an irregular, blurred, or notched edge.
• C — Color: multiple colors or uneven distribution.
• D — Diameter: size can matter, but a melanoma can be smaller than 6 mm.
• E — Evolution: change in size, shape, color, surface, symptoms, or behavior.

In lesion-selected studies, in-person visual inspection using melanoma checklists had pooled sensitivity near 83% and specificity near 88%.^2,7 Use ABCDE and the ugly-duckling comparison as reminders while preserving attention to atypical presentations; neither is a rule-out test. The studies were heterogeneous, usually involved selected lesions, and do not recreate the prevalence or uncertainty of a general population.

Evolution is often more informative than a fixed snapshot. The ugly-duckling comparison asks whether one lesion looks different from the patient's other lesions. Small image studies support the concept, but it is still a clinical prompt rather than a validated rule-out tool.^7,8

What ABCDE misses

Nodular melanoma may be symmetric, uniform in color, and smaller than expected while growing quickly. Amelanotic melanoma may be pink, red, or skin-colored. Acral and nail melanoma can present as a changing patch or band rather than a conventional mole. A rapidly growing, firm, elevated, symptomatic, or otherwise discordant lesion warrants evaluation even without the full mnemonic.^2,3

Recognizing Possible Basal and Squamous Cell Carcinoma

Melanoma education often dominates public teaching, but the examination must also recognize the variable faces of basal and squamous cell carcinoma.

Possible basal cell carcinoma

Classic nodular BCC may appear as a pearly or translucent papule with arborizing telangiectasia, a rolled border, or ulceration. Superficial BCC can resemble a persistent scaly dermatitis, and these classic clues remain nonspecific rather than diagnostic. Morpheaform BCC can look like a pale, indurated scar with poorly defined borders.^9,10

Pearly quality, recurrent bleeding, and a nonhealing surface are useful clues, not diagnoses. The absence of those clues does not exclude BCC, and no validated four-week threshold converts a persistent lesion into cancer.

Possible cutaneous squamous cell carcinoma

cSCC may be keratotic, crusted, indurated, tender, ulcerated, bleeding, or rapidly enlarging. Well-differentiated tumors may produce visible keratin, while poorly differentiated tumors can be smooth, eroded, or deceptively nonspecific.^9,11,12

Pain and tenderness occur more often in cSCC than in several common comparison lesions in observational cohorts, but they remain nonspecific.^11,12 Palpation matters because firmness, fixation, and depth can be invisible in a photograph.

The shared safety boundary

Persistent change, rapid growth, spontaneous bleeding, ulceration, pain, induration, or clinical discordance should lower the threshold for assessment. None is an automatic diagnosis, and none should be isolated from the full examination.

Sites and Skin Tones That Checklists Miss

Skin cancer occurs in every skin tone. A complete examination should not stop at the sun-exposed face and forearms. The scalp, ears, interdigital spaces, palms, soles, nails, and mucosal-adjacent skin deserve attention when the history or risk profile supports it.^3,4

Acral melanoma is proportionally more common among Black, Asian, and Hispanic patients than among White patients, although absolute melanoma incidence is lower in darker skin tones. Delayed recognition, site-specific biology, access, and diagnostic assumptions all contribute to disparities. A changing nail band, palm or sole lesion, nonhealing acral wound, or unusual pigmented or amelanotic spot deserves the same disciplined assessment as a conventional mole.³

Scalp lesions can remain hidden by hair. Partner assistance, mirrors, photographs, and professional examination may improve visibility, but none substitutes for attention to change.

The Integrated Clinical Examination

The most defensible diagnostic pathway combines several imperfect but complementary inputs:^3,9

A strong assessment combines history, change over time, comparison with the patient's other lesions, close inspection, palpation and regional-node examination when indicated, dermoscopy, and the threshold for tissue diagnosis.^3,9

1. Reason for the encounter: routine risk assessment, a new lesion, change in an existing lesion, symptoms, or surveillance after cancer.
2. Risk context: prior cancers, immune status, nevus phenotype, UV history, family history, and patient reliability.
3. History of the lesion: onset, evolution, growth rate, bleeding, pain, treatment attempts, and prior photographs.
4. Comparison: whether the lesion fits or violates the patient's background pattern.
5. Palpation: firmness, scale, mobility, depth, fixation, tenderness, and regional nodes when indicated.
6. Dermoscopy: subsurface structures not visible to the naked eye.
7. Decision threshold: reassure, treat clinically, photograph, monitor at a defined interval, or obtain tissue.

The pathway is intentionally redundant. When one input is weak—poor image quality, uncertain duration, an atypical skin tone, or a lesion in a hidden site—other inputs become more important.

Dermoscopy Improves the Examination

Dermoscopy reduces surface reflection and magnifies subsurface pigment and vascular structures. In the Cochrane melanoma review, trained dermoscopy increased estimated sensitivity from about 76% to 92% at a fixed specificity of 80%.^13,14

That number has three important boundaries:

• Most evidence came from specialist settings.
• Training and experience materially affect performance.
• Dermoscopy narrows probabilities; it does not create histology.

An experienced clinician uses dermoscopy within the total pattern: a network, streaks, dots and globules, blue-white structures, shiny white lines, ulceration, vascular architecture, and site-specific clues are interpreted in relation to history and morphology. The device is not a magic magnifying glass. It is a learned diagnostic language.

Photography Is a Memory Aid, Not a Diagnosis

Total-body photography can be helpful for selected patients with many nevi or difficulty detecting change. It improves the memory of the examination: a new lesion, a changing lesion, and a stable lesion can be compared with a reproducible baseline.

More imaging is not automatically better. In a 2025 randomized trial, three-dimensional total-body photography plus sequential dermoscopy increased biopsies and excisions without increasing melanoma detection per person; a mortality benefit has not been established.¹⁵ Photography should answer a defined surveillance question rather than create an endless search for microscopic change.

When short-interval dermoscopic monitoring is reasonable

A carefully selected, mildly atypical, flat lesion without high-concern features may be photographed and reassessed by an experienced clinician. In a prospective monitoring cohort, lack of change at three months had 99.2% reliability for benignity within that narrow population.¹⁶

That estimate should never be generalized to a nodular, rapidly enlarging, symptomatic, ulcerated, bleeding, or clinically discordant lesion. A monitoring plan must include a specific interval, reproducible image, defined trigger for biopsy, and reliable follow-up.

Confocal Microscopy, Teledermatology, and Consumer Apps

Reflectance confocal microscopy

Reflectance confocal microscopy can improve specificity for selected equivocal lesions at expert centers and may prevent some unnecessary excisions. Its limits include specialized equipment, training, limited penetration, anatomic constraints, and interpretive uncertainty. It is an adjunct, not a universal substitute for pathology.^22,23

Teledermatology

Teledermatology can improve access and triage. A high-quality image plus clinical history may support urgency decisions, but image-only review cannot reproduce palpation, total-body comparison, or regional-node examination and remains sensitive to framing, lighting, focus, and scale.¹⁷

Consumer applications

In the Cochrane review of smartphone melanoma-triage applications, sensitivity ranged from 7% to 73%; in the evaluated datasets, 27% to 93% of melanomas were classified as low risk.¹⁸ The evidence was small and at high risk of bias, so the result should not be used to rank current products. It establishes a safety boundary: an app cannot safely clear a concerning lesion.

When Tissue Diagnosis Is Needed

Histopathology is the reference standard when a definitive tissue diagnosis is required. The threshold depends on pretest concern, the consequence of delay, the feasibility of complete sampling, and whether monitoring can be performed safely.^3,9

Not every seborrheic keratosis, angioma, dermatofibroma, or stable nevus needs biopsy. Conversely, a clinically familiar label should not override evolution, symptoms, palpation, dermoscopic discordance, or a sampling concern.

Choosing a biopsy for possible melanoma

When melanoma is suspected, complete sampling with narrow clinical margins and sufficient depth is preferred when feasible because Breslow thickness and ulceration guide staging. An appropriately deep saucerization can be a complete biopsy; “shave” is not one uniform depth or technique.^3,19

Partial sampling can be reasonable for a large facial, acral, nail, or otherwise anatomically difficult lesion when complete removal would be impractical or harmful. The clinician should target the most suspicious area, document the remaining lesion, and anticipate that final depth may be greater than the sample shows.

What the shave-biopsy numbers actually mean

In a meta-analysis of 14 studies of melanoma diagnosed by shave biopsy, the deep margin was positive in 42.9%, T stage changed after definitive excision in 7.7%, and additional treatment changed in 2.3%.¹⁹ The studies were heterogeneous and did not establish that shave biopsy worsens survival.

The teaching point is neither “shaves are unsafe” nor “depth never matters.” It is to match technique to the staging question and obtain enough depth to avoid preventable uncertainty.

Sampling basal and squamous cell carcinoma

BCC subtype concordance between the initial biopsy and definitive specimen is imperfect. One retrospective series found 18% overall subtype disagreement; another found that 28% of tumors called nonaggressive on biopsy revealed an aggressive component during Mohs surgery.^20,21 A superficial sample may identify BCC while understating the pattern that determines margin strategy.

For suspected invasive cSCC, the specimen should be deep and representative enough to assess invasion and adverse pathology. A shallow surface sample can miss the deepest, most consequential component.²⁴

Benign Mimics and Malignant Impostors

Seborrheic keratosis, dermatofibroma, angioma, nevus, scar, inflammatory dermatitis, actinic keratosis, and benign adnexal tumors can mimic cancer. Melanoma, BCC, and cSCC can imitate each of them.

The correct response is not to abandon pattern recognition. It is to use pattern recognition honestly:

• A stuck-on surface favors seborrheic keratosis but does not explain every evolving pigmented lesion.
• A firm dimple sign favors dermatofibroma but does not excuse rapid enlargement or ulceration.
• Bright-red lacunae favor angioma but do not explain every bleeding red papule.
• Symmetry favors a benign nevus but does not overrule evolution or an atypical dermoscopic pattern.
• Scale favors inflammation or keratinocytic disease but does not distinguish superficial BCC, actinic keratosis, SCC in situ, and dermatitis by itself.

The lesion earns reassurance when the history, examination, dermoscopy, and expected behavior agree—not because it resembles one internet photograph.

The Monitor-or-Biopsy Boundary

Monitoring is a clinical intervention, not indecision. A safe monitoring plan requires:

• a lesion appropriate for observation;
• a reproducible baseline image;
• a defined interval;
• a specific change threshold;
• a patient who can return; and
• a backup plan if uncertainty increases.

Biopsy is safer when the lesion is nodular, rapidly enlarging, spontaneously bleeding, ulcerated, symptomatic, palpably deep or fixed, clinically discordant, or sufficiently concerning that a delay would change staging or treatment.^3,9,16

A Practical Decision Framework

Related Patient Guides

The following guides turn these principles into focused answers to common patient questions.