Advances in the noninvasive diagnosis of melanoma—40 years beyond the ABCDs

Abstract

The early detection of cutaneous melanoma is critical to survival outcomes. Because less than one half of melanomas in the United States are diagnosed by dermatologists, the ABCD (asymmetry, border irregularity, color variation, diameter >6 mm) acronym, created 40 years ago with the later addition of “E” for evolution (ABCDE), was developed for nondermatologist health care professionals and the public to simplify and enhance the diagnosis of early melanoma. It continues to be the global, naked‐eye, nondevice‐assisted standard for initial triage of pigmented lesions. This clinical review discusses the changing clinical diagnostic landscape and examines the currently available first‐line and second‐line detection modalities for melanoma. It also provides updates to the first‐line triage approach and discusses the challenges of regulatory agency oversight for the safe and effective use of current and emerging skin cancer detection technologies. It is critical that health care professionals globally have knowledge of these technologies to enhance their diagnosis of melanoma.

INTRODUCTION

Melanoma is now the fifth most common US malignancy with an incidence that is rising faster than any other cancer. ¹ , ² The 5‐year survival rate for patients in the United States who have early localized disease is now >99% versus only 35% for those who have distant disease. ³ Therefore, early detection is essential for improving prognosis as well as minimizing health care expenses because >90% of melanoma‐related treatment costs are associated with patients who have advanced disease. ⁴ Visual assessment of pigmented skin lesions (PSLs) followed by histopathologic analysis continues to be the standard of care. ⁵

Over 200,000 melanomas will be diagnosed in the United States this year, but less than one half are initially diagnosed by dermatologists. ⁶ The development of the ABCD (asymmetry, border irregularity, color variation, diameter >6 mm) acronym 40 years ago, ⁷ with the later addition of E for evolution (ABCDE; Figure 1), was designed for nonspecialized health care professionals and the lay public to simplify and enhance the diagnosis of early melanoma, and ABCDE continues to be the global naked‐eye, nondevice‐assisted standard for the initial triage of pigmented lesions. ⁸

FIGURE 1.

ABCDEs of melanoma detection. (A) Asymmetry, (B), border irregularity, (C) color variation, (D) diameter >6 mm, (E) evolution.

Technologic innovations, including dermoscopy, laser‐based technologies, spectroscopy, and gene expression profiling (GEP) analysis, have expanded noninvasive device–assisted diagnostic techniques. Artificial intelligence (AI) has demonstrated melanoma diagnostic accuracy on par with and, in some studies, potentially superior to that of clinicians. ⁹ This review focuses on available first‐line and second‐line detection modalities for melanoma, an update of triage approaches, and the integration of technology into the clinical setting for optimizing early diagnosis and referral. Given the innovations described above, it is important for all clinicians who treat patients with skin conditions to be aware of the new technologies available for melanoma diagnosis and to facilitate their inclusion into the clinical setting. ¹⁰

ABCDE HISTORY AND FUNCTION AS THE FIRST‐LEVEL TRIAGE TOOL

ABCDE criteria were developed as an accessible way for physicians and the public to recognize features of early melanoma with a simple‐to‐learn, widely accessible, and low‐cost barrier method. The developers observed that early melanomas were asymmetric, had irregular borders, had variegated colors, and 95% of these lesions at that time were >6 mm in diameter. ⁷ This paradigm and its application have been validated in numerous subsequent studies. ¹¹ , ¹² , ¹³ , ¹⁴ , ¹⁵ The acronym was subsequently expanded to include the evolving (E) nature of melanoma with regard to size, color, shape, symptoms, and surface. These evolving features can capture melanomas that change and may not fit the ABCD criteria at one point in time. Although benign PSLs may have ABCDE features (e.g., seborrheic keratosis), this framework captures the majority of early melanomas to enhance early detection. The ABCDE method is regularly taught as a part of health care training and is often found on posters in the offices of health care professionals, patient education outreach materials, and the mainstream media. ¹⁶ , ¹⁷ , ¹⁸ , ¹⁹ , ²⁰

THE CHANGING DIAGNOSTIC LANDSCAPE AFTER FIRST‐LEVEL TRIAGE

With noninvasive, second‐level triage/diagnostic tools that add relevant visual, cytologic, and/or genetic insights after ABCDE evaluation, clinicians are able to further improve the accuracy of biopsy selection. High sensitivity and specificity, affordability, ease of use, and the ability to limit unnecessary triage and overdiagnosis are the targets for second‐level tool adoption and sustained use. Therefore, the use of dermoscopy has been widely adopted because it increases the likelihood of correctly selecting PSLs for biopsy compared with using ABCDE criteria alone.

Only one in three patients in the United States with a skin lesion of concern is actually seen by a dermatologist because of issues with access to specialty care. Access is improved by the inclusion of advanced practice providers (APPs), such as physician assistants and nurse practitioners, who are estimated to now see nearly one half of all suspicious skin lesion spot checks in the United States. ²¹ , ²² In Australia, primary care providers are the predominant front‐line health care professionals diagnosing >75% of melanomas. ²³ Because early melanomas can present with a wide range of features and may be difficult to diagnose with the naked eye alone, it is crucial for the choice of a tool for the first‐level triage of PSLs to evolve such that each patient receives equitable, quality care regardless of the level of training of their licensed health care professional.

FIRST‐LEVEL PSL TRIAGE

Dermoscopy is a handheld instrument that provides ×10 magnification and minimizes light reflection off of the skin surface, thereby enabling the visualization of colors and structures located below the stratum corneum that are not visible to the naked eye (Figure 2). ²⁴ It is commonly used in clinical settings to enhance visual detection of melanoma and is now considered standard of care for triage and biopsy selection. ²⁵ A meta‐analysis reported that the diagnostic accuracy for melanoma was four times higher for dermoscopy compared with the unaided eye. ²⁶ Dermoscopic evaluation of PSLs after ABCD evaluation significantly improves sensitivity for melanoma diagnosis versus the unaided eye (odds ratio, 0.90 vs. 0.71; p = .002) without a decrease in specificity (odds ratio, 0.90 vs. 0.81; p = .18). ²⁶ Although effectiveness depends on the experience of the user, dermoscopy has been shown to decrease the number of biopsies of benign lesions compared with malignant lesions ²⁷ , ²⁸ (Table 1). ²⁹ , ³⁰ , ³¹ , ³²

FIGURE 2.

(A) Benign nevus with globular pattern (uniform globules). (B) Benign nevus with reticular pattern (uniform pigment network). (C–F) Malignant melanoma with features including atypical pigment network, asymmetric colors, irregular dots and globules, atypical streaks, blue‐white structures, and regression structures.

TABLE 1.

First‐level and second‐level noninvasive detection devices for triage of pigmented skin lesions.

Dermoscopy algorithms	Sensitivity, %	Specificity, %	Diagnostic accuracy, %
Pattern analysis (Annessi 2007 29 )	85.0	79.0	71.0
ABCD (Annessi 2007 29 )	84.0	75.0	76.0
Seven‐point checklist (Annessi 2007 29 )	78.0	65.0	58.0
CASH (Henning 2007 30 )	98.0	68.0	—
Menzies (Dolianitis 2005 31 )	85.0	85.0	81.0
Triage amalgamated dermoscopy algorithm (TADA; Rogers 2017 32 )	94.0	75.5	—
Three‐point checklist (Rogers 2017 32 )	88.6	78.7	—

Abbreviations: ABCD, asymmetry, border irregularity, color variation, diameter >6 mm; CASH, color, architecture, symmetry, homogeneity.

When analyzing dermoscopy patterns, clinicians can implement algorithms to further enhance and simplify diagnosis. The dermatoscopic ABCD rule is an easy‐to‐use method and works off of the clinical ABCDE for melanoma. ³³ It calculates a total numerical score based on asymmetry, border, color, dermoscopic structures, and morphologic change. ³³ The three‐point checklist is another simple algorithm using asymmetry, atypical pigment network, and the presence of blue‐white structures; whereas the seven‐point checklist is more extensive and is based on a list of major and minor criteria. ³³ Furthermore, the CASH (color, architecture, symmetry, homogeneity) algorithm and the Menzies method (a simplified dermoscopy method based on 11 features scored as present or absent) are straightforward and are designed for clinicians with limited dermoscopy experience. ³³ Sensitivity (78%–98%), specificity (65%–85%), and accuracy (58%–81%) vary between methods; and, because each aims to simplify diagnosis, which algorithm to implement in clinical practice is an individual choice (Table 2). ⁸ We more frequently use the Menzies method. Based on the available evidence, dermoscopy is recommended for the evaluation of PSLs by health care professionals who have received appropriate training in its use.

TABLE 2.

Comparison of dermoscopy algorithms.

Technology	Common features	Advantages	Limitations
First‐level triage
Dermoscopy	Low cost Easy to use Learned in dermatology training	Improves SE and SP	User skill–dependent False positives
Digital dermoscopy imaging		Medical‐grade objective image documentation	Dermoscopy better when evaluating patient vs. image
Second‐level triage
Reflectance confocal microscopy	Laser‐based technologies Improves SE and SP	Higher SE than dermoscopy alone	Requires training Cost of equipment Limited depth of imaging Takes longer to administer
Optical coherence tomography		Examines skin to a depth of 1 mm	Requires training Cost of equipment Small field of view Takes longer to administer
Multiphoton laser scanning combined with reflectance confocal microscopy		Scans deep into tissue without causing damage	Requires training Cost of equipment Small field of view Takes longer to administer
Electrical impedance spectroscopy	Spectroscopy‐based technologies High SE and NPV More commonly used in clinical practice	Enhanced accuracy with dermoscopy Less expensive vs. others Instant results Low false‐negative rate	May need to take multiple measures of the same lesion because the device's electrodes may not cover the entire lesion
Elastic scattering spectroscopy		FDA indication for primary care physicians	Not intended for stand‐alone for biopsy selection Only indicated for patients older than 40 years
Two‐gene expression profiling	—	High SE, SP, and NPV Reduces surgery and morbidity	Need to return for biopsy if test indicates not benign Potential genetic analysis failure
Ultrasound	—	Deeper penetration and improved resolution Potentially assesses depth	Requires additional research for integration in clinical settings

Abbreviations: FDA, US Food and Drug Administration; NPV, negative predictive value; PSL, pigmented skin lesion; SE, sensitivity; SP, specificity.

Digital dermoscopy imaging (DDI), through attachment of the dermatoscope to a camera, has enabled improved medical‐grade objective image documentation of PSL findings in patient electronic records. ³⁴ , ³⁵ DDI provides context concerning where dermoscopic features are located within the PSL and has the potential to complement clinician ABCDE evaluation by increasing pretest probability.

A previously available noninvasive, handheld imaging device, MelaFind (Electro‐Optical Sciences) was the first optical diagnostic device for melanoma detection approved by the US Food and Drug Administration (FDA) in 2011. ³⁶ It compared images from nine visible and infrared bands to distinguish early melanoma from benign PSLs, ³⁷ with 98.4% sensitivity and 9.9% specificity. ³⁸ , ³⁹ However, its clinical availability was ended in 2017 because of low specificity, challenges integrating it into the clinical workflow, and reimbursement difficulties. ⁴⁰ Compared with the ease of clinical utility of dermoscopy, incorporating this device into practice was found to be difficult.

SECOND‐LEVEL PSL TRIAGE

Laser‐based techniques

Reflectance confocal microscopy (RCM) uses near‐infrared, low‐power laser energy to provide noninvasive, in‐vivo imaging of PSLs down to the upper dermis. ⁴¹ Prospective interventional studies and meta‐analyses demonstrated that RCM has sensitivity equal to that of dermoscopy but increases specificity, enabling a reduction in the number of unnecessary benign PSL biopsies. ⁴² , ⁴³ A multicenter, randomized clinical trial of >3000 equivocal PSLs resulted in a reduced number needed to excise a melanoma from 3.7 to 1.8 in the RCM arm and excisions of benign PSLs were reduced by 43.4%. ⁴⁴ RCM limitations include the limited depth of imaging and difficulty in differentiating pagetoid melanocytes from Langerhans cells. ⁴¹ Pellacani et al. developed an algorithm for RCM features to identify melanoma that achieved 96.3% sensitivity and 52.1% specificity for a score ≥2.⁴¹ The algorithm calculated a total score with an algorithm based on various features identified through RCM, including nonedged papillae, mild‐to‐marked cell atypia, roundish pagetoid cells, widespread pagetoid infiltration, cerebriform nests, and nucleated cells in the dermis. ⁴⁵

Optical coherence tomography (OCT) is an imaging modality that uses infrared light to derive two‐dimensional and three‐dimensional images of tissue and can examine skin to a depth of 1 mm.⁴² Conventional OCT has had produced mixed results, with several studies reporting findings correlated to histopathology, whereas others did not. ⁴⁶ Four OCT imaging systems have been developed, including high‐definition OCT), speckle‐variance OCT, conventional OCT, and line‐field OCT. High‐definition OCT has been correlated with histopathology for melanoma with 74.1% sensitivity and 92.4% specificity. ⁴⁷ Speckle‐variance OCT provides visualization of vascular patterns. ⁴⁸ Line‐field OCT has high resolution and deeper image penetration but has a limited field of view. ⁴⁹

Comparatively, in‐vivo RCM creates high‐resolution digital pathology images that are read by a trained dermatology or pathology clinician. In‐vivo RCM uses RCM in addition to a multiphoton microscope to scan deep into tissue without causing damage. This device received FDA clearance for use in 2023. ⁵⁰ , ⁵¹ , ⁵² In addition, multiphoton microscopy (MPM) is a noninvasive imaging technique that is being developed to differentiate benign nevi from melanoma. ⁵³ , ⁵⁴ MPM induces epidermal fluorescence and a collagen‐specific signal in the dermis, after which sequential sections are captured. ⁵³ , ⁵⁴ Each lesion is assigned a multiphoton microscopy score, calculated as the weighted sum of malignant features. ⁵³ , ⁵⁴ There is limited literature on the utility of MPM for clinical melanoma diagnosis, although several studies with few patients produced promising results for differentiating malignant from benign nevi. ⁵⁵ , ⁵⁶ , ⁵⁷ However, OCT, in‐vivo RCM, and MPM require additional research and applicability for clinicians in clinical practice.

Spectroscopy

Perhaps the most integrated technology for the analysis of PSLs is spectroscopy‐based imaging, specifically, electrical impedance spectroscopy (EIS) and elastic scattering spectroscopy (ESS).

EIS is a noninvasive imaging modality that detects variations in electrical resistance between benign and malignant cells by applying a matrix of low‐voltage electrodes placed directly onto the skin. ⁵⁸ , ⁵⁹ , ⁶⁰ An EIS device using a program developed with AI is the only currently FDA‐approved (class 3) device for melanoma detection. ⁶¹ In a study of >2000 lesions, EIS achieved 96.6% sensitivity, 34.4% specificity, and a 98.2% negative predictive value. ⁶⁰ Another analysis reported similar sensitivity findings of 98.1% ⁶² and a false‐negative rate of 3.4% for melanoma. ⁶³ It also has been demonstrated that EIS alters dermatology trainees' biopsy decisions by 24.3%, improves their melanoma diagnosis sensitivity from 80.7% to 95.2%, ⁶⁴ and improves the biopsy decisions of APPs and dermatologists. ⁶⁵ EIS had higher sensitivity for melanoma or severely dysplastic nevi versus dermoscopy (66% vs. 75%). ⁶⁶ In addition, integration of EIS by dermatology professionals after dermoscopic evaluation significantly improved the rate of correct biopsy choices for melanoma (91.1% vs. 85.2%) and decreased incorrect biopsy decisions (8.9% vs. 14.8%), ⁶⁷ elucidating the ability of EIS to positively assist dermatology clinicians in evaluating indeterminate lesions beyond dermoscopy alone. ⁶² A contemporary expert consensus panel unanimously recommended using EIS for clinically indeterminate lesions because it can significantly enhance diagnostic assessment and clinical decision making for PSLs, including cutaneous melanoma (SORT [Strength of Recommendation Taxonomy] grade A). ⁶⁸

ESS is a form of reflectance optical spectroscopy that can detect melanoma by differentiating microscale and nanoscale structure of skin tissue. ⁶⁹ An ESS in‐vivo, handheld device classifies lesions as high‐risk or low‐risk and received FDA clearance (class 2) in 2024 for use by primary care physicians (PCPs) to triage PSLs (melanoma, basal cell carcinoma, squamous cell carcinoma, severely atypical melanocytic nevi) to a dermatologist for further evaluation in patients older than 40 years. ⁷⁰ A multicenter, prospective, investigator‐blinded study indicated that ESS melanoma sensitivity, when used by dermatologists, was 95.5%, specificity was 32.5%, and the negative predictive value was 98.1%. ⁷¹ In another study evaluating ESS use by PCPs, diagnostic sensitivity for melanoma increased from 67% to 88%, but there was no difference in specificity. ⁷² Currently, ESS is not intended as stand‐alone clinical support for biopsy selection but is recommended for use by PCPs to evaluate which lesions should be referred to dermatologists.

Gene expression profiling

Another second‐level triage technology that can be feasibly applied in the clinical setting is GEP testing. The two‐gene expression profile (2‐GEP) test uses an adhesive patch to collect samples of epidermal RNA that are evaluated with a GEP assay. ⁷³ The assay evaluates the tissue for the expression of two genes (LINC00518 and PRAME), ⁷³ and the initial validation study, which evaluated 398 PSLs, reported that it is able to differentiate benign lesions and melanoma with 91% sensitivity and 69% specificity. ⁷³ Another study of 381 PSLs that were clinically suspicious for melanoma reported 95% sensitivity, 91% specificity, and a negative predictive value >99% at 6‐month follow‐up. ⁷⁴ When patients were followed up to 12 months, 99.9% of 2‐GEP–negative lesions were clinically monitored and appropriately avoided a surgical procedure, and 96.5% of 2‐GEP–positive lesions were appropriately biopsied. ⁷⁵

For PSLs that are clinically indeterminate, the 2‐GEP test may be used to aid in diagnosis and potentially reduce surgical intervention and morbidity. ⁷⁶ This technology was particularly useful during the coronavirus disease 2019 pandemic because suspicious PSLs could be tested by patients at home, obviating the need for an office visit if the results were negative. The 2‐GEP test was intended for use in lighter skin types but now has been shown to be effective for all skin phenotypes. ⁷⁷ An expert consensus panel unanimously recommended that 2‐GEP testing can significantly enhance diagnostic assessment and clinical decision making for PSLs (SORT [Strength of Recommendation Taxonomy] grade B). ⁶⁸ The drawback to the 2‐GEP test is that it can take ≥5 days to receive results, so patients need to return to the clinic for biopsy if the test indicates that a PSL is not benign. Future research in this area may lead to the development of tests using similar technology that integrate more genes into the discriminating algorithms to enhance the efficacy of this approach.

Ultrasound

The application of ultrasound to the diagnosis of melanoma has been explored. ⁷⁸ By using a higher frequency wavelength (high‐frequency ultrasound [HFUS]), skin lesions can be visualized with deeper penetration and improved resolution. ⁷⁹ Melanoma is often depicted as a poorly echoic, well defined lesion on HFUS, which may be useful in determining excision margins for superficial melanomas and subclinical, metastatic foci. ⁷⁹ , ⁸⁰ Melanomas may also be hypervascular, although studies report that using vascularity to differentiate benign nevi from melanoma may be difficult in thin lesions. ⁸¹ One systematic review indicated that sensitivity based on qualitative HFUS characteristics (e.g., lesion appearance) ranged from 83% to 100% for melanoma and specificity ranged from 33% to 73%. ⁷⁸ The review concluded that that studies were generally poorly reported and had varied methodologies, limiting conclusions. ⁷⁸ Furthermore, a study evaluating ultrasound identification of locoregional metastases reported 89.2% sensitivity and 97% specificity for finding occult metastases compared with clinical examination alone. ⁸² A meta‐analysis demonstrated that HFUS may serve as a supplementary tool for preoperative melanoma assessment because there are strong correlations to histopathology (ranging from 0.417 to 0.997; standard deviation, 0.13). ⁸³ Similar to laser‐based techniques, ultrasound requires additional research for integration in clinical settings.

Mobile applications

In addition to dermatologist evaluation, multiple publicly available mobile applications (apps; e.g., SkinVision, Skinive [Stichting Skinive Foundation], and canofyMD SCAI [LifeSemantics]) offer a unique approach to evaluating PSLs. However, there is insufficient evidence supporting their accuracy and reliability. ⁸⁴ Studies have indicated that these apps typically have low sensitivity and limited agreement with dermatologist PSL evaluation. ⁸⁵ , ⁸⁶ In addition, there is no currently established regulatory oversight process to uphold standards of melanoma screening with apps. ⁸⁴ The development of accurate diagnostic mobile apps could make a major impact on broad‐based early diagnosis by the public.

Integrating AI into PSL diagnostic devices

Research into the integration of AI into PSL diagnostic devices is becoming increasingly prevalent. One of the major targets of these initiatives is AI‐image–based algorithms that can augment the analysis of digital dermoscopy images to identify dermoscopic patterns and features of melanoma with high accuracy. ²⁶ , ⁸⁷ , ⁸⁸ , ⁸⁹ , ⁹⁰ The potential to augment first‐level triage is under active exploration. ³⁵

Most DDI‐AI algorithms have been trained on the publicly available International Skin Imaging Collaboration database, and those with the highest balanced accuracy (sensitivity vs. specificity) have been further retrained using private DDI data sets that have been annotated by dermoscopy experts. ⁹¹ In the United States, the FDA requires a threshold performance of at least 95% sensitivity and 30% specificity and suggests that DDI‐AI outputs displayed to the user are either suspicious, unremarkable, or error (requiring another attempt up to a maximum of three). If these criteria are met, DDI‐AI will be a significant development to assist health care professionals who have limited dermoscopy experience with interactive education and a possible stand‐alone bedside clinical support tool. Of note, studies report that most AI studies for skin cancer detection include relatively few images of darker skin phenotypes. ⁹² , ⁹³ Therefore, diverse populations need to be included in future studies and machine‐learning data sets so that patients with the full spectrum of skin tones may benefit from these technologies.

In addition, the use of AI permits the creation of a heatmap that may give context to clinicians where the positive dermoscopy findings of concern are present. It can also inform the dermatopathologist why the clinician chose to complete a biopsy, which may increase diagnostic precision. ⁹⁴ , ⁹⁵ Guided by DDI‐AI, in‐vivo placement of a tissue‐marking ink on the PSL before biopsy may further enable visualization of the ink above the stratum corneum in hematoxylin‐and‐eosin–stained sections. This method may be a means to improve diagnostic precision, although further studies are needed to evaluate its application in clinical practice.

CONTEMPORARY RECOMMENDATIONS

Throughout the past 40 years, there have been numerous technologies that have enhanced the diagnostic capabilities of clinicians evaluating PSLs. Several of these have demonstrated high diagnostic accuracy and sensitivity for detecting melanoma, and more development is being done to evaluate the advantages and disadvantages of these techniques. Currently, the major devices readily available to the PCP are dermoscopy, which is effective and inexpensive; and ESS, which has FDA clearance (class 2) specifically for use by PCPs. For dermatologists, dermoscopy, EIS, and 2‐GEP testing can aid in timely detection and augment the differentiation of benign lesions and melanoma. Other techniques, such as RCM, OCT, and in‐vivo RCM, may be available only in institutional settings. Each one of these builds upon the naked‐eye assessment that the ABCDEs provide, and knowledge is still required to be able to accurately select lesions for evaluation and to use the technologies available to enhance decision making.

Conclusion

The ABCD paradigm for the clinical recognition of early melanoma was established 40 years ago and has remained a vital tool for early detection of this cancer. Since then, various innovative technologies have been developed to augment the noninvasive diagnosis of melanoma and to enhance the selection of lesions for biopsy, many of which are already available for use in the clinical office setting and several that do not require specialty training in dermatology. Because early detection remains the most effective strategy for enhancing melanoma prognosis and the majority of these cancers are diagnosed by nondermatologists, it is imperative that all types of health care professionals involved in melanoma diagnosis should have the knowledge to integrate these tools into clinical practice to benefit patients.

Regulatory challenges—in the United States and globally—may exist that could affect the development of certain diagnostic approaches. Therefore, it is also essential for clinicians to understand the role of regulatory agency oversight for the safe and effective use of evolving technologies.

Despite all of the advances that have occurred over the 40 years since the development of the ABCDs, early melanoma diagnosis still faces challenges because one American still dies every hour from this cancer. ⁹⁶ Further innovation and studies will continue to enhance diagnostic technologies that will have the potential to improve clinicians' ability to better recognize and identify appropriate lesions for biopsy, with AI playing an increasingly important role. With the integration of these advances into clinical practice, both clinicians and the public have the potential to benefit from the enhanced ability to detect early melanoma, leading to improved patient survival.

CONFLICT OF INTEREST STATEMENT

Alexander Witkowski is a co‐founder and fiduciary officer of Sklip Inc. and reports support for professional activities from Castle Biosciences Inc. outside the submitted work. Joanna Ludzik is a co‐founder of Sklip Inc. and owns stock options in the company. Giovanni Pellacani is a member of the Sklip Inc. advisory board. Darrell Rigel reports personal/consulting fees from Almirall LLC, Castle Biosciences Inc., Kenvue, Pfizer, SciBase, Skincure, Sun Pharmaceuticals, and Takeda Pharmaceuticals outside the submitted work and owns stock in VYNE Pharmaceuticals Inc. The remaining authors disclosed no conflicts of interest.

Burshtein J, Witkowski A, Zakria D, et al. Advances in the noninvasive diagnosis of melanoma—40 years beyond the ABCDs. CA Cancer J Clin. 2026;e70065. doi: 10.3322/caac.70065