Total-Body Photography in Skin Cancer Screening: The Clinical Utility of Standardized Imaging

Skin cancer is an important public health issue in the United States, as 1 in 5 Americans are projected to develop a cutaneous malignancy during their lifetime. Currently, 75% of all skin cancer–related deaths are due to malignant melanomas (MMs), though melanomas account for less than 5% of all skin cancers.¹ Early detection of MM is essential, as prognosis depends on tumor stage, particularly the depth of the melanoma.^2-4 In general, patients with thin, early-stage melanomas have a more than 96% survival rate, which drops to 14% in late-stage disease.^5,6 Five percent to 30% of all melanomas are identified incidentally on total-body skin examinations (TBSEs) performed by a trained provider and thus would not have been caught with only a focused skin examination or patient self-examination.^7,8 Nonetheless, the clinical utility of skin cancer screening with TBSEs remains controversial, largely due to the poor quality of data available to establish a notable mortality benefit from skin cancer screening. As a result, obtaining endorsement from the larger medical community, federal government, and health insurance industry to include routine TBSEs as part of a preventive care health care strategy has not occurred. The absence of definitive clinical care guidelines mandating routine TBSEs is one of the greatest barriers preventing access to appropriate dermatologic screening along with the paucity of trained providers; however, standardized total-body photography (TBP) promises to provide a way forward by lowering the costs of dermatologic screening while simultaneously leveraging technology to increase availability.

Impact on Biopsy Efficiency

Current US Preventive Services Task Force (USPSTF) guidelines state that evidence is insufficient to assess the balance of benefits and harms of visual skin examination by a clinician to screen for skin cancer in adults. The USPSTF noted that “[d]irect evidence on the effectiveness of screening in reducing melanoma morbidity and mortality is limited to a single fair-quality ecologic study with important methodological limitations” (ie, the Skin Cancer Research to Provide Evidence for Effectiveness of Screening in Northern Germany [SCREEN] study), and although information on harm is similarly sparse, “[t]he potential for harm clearly exists, including a high rate of unnecessary biopsies, possibly resulting in cosmetic or, more rarely, functional adverse effects, and the risk of overdiagnosis and overtreatment.”⁹ The majority of suspicious skin lesions excised during screenings are not cancerous. For example, the SCREEN study found that 20 to 55 excisions were performed to detect 1 case of melanoma.¹⁰ At that rate, the USPSTF also noted that approximately 4000 excisions would be required to prevent a single death from melanoma.⁹ Following the lead of the USPSTF, the Patient Protection and Affordable Care Act did not mandate that skin examinations be included as essential preventive coverage in its requirements for insurance coverage of primary care prevention. As such, dermatologists face financial pressure to avoid performing time-consuming TBSEs, regardless of their perceived utility.¹¹

As the USPSTF points out, the value of TBSEs relies on the examiner’s ability to correctly identify malignant lesions and minimize biopsies of benign lesions, a concept known as biopsy efficiency.⁹ Secondarily, a TBSE is time consuming, and the time required for a dermatologist to complete a TBSE given the high rate of benign findings may not be financially viable. We argue that the routine use of total-body skin imaging may offer a way forward in addressing these issues. Total-body photography involves a photographic system that can allow dermatologists to acquire standardized images that can be used for primary diagnosis and to track individual lesions over time. Nonmedical personnel and medical assistants can be easily trained to use standardized photography devices to quickly obtain high-quality clinical images, thereby greatly reducing the time and cost of obtaining these images. Studies have found that the use of photographic monitoring may improve biopsy efficiency.^12-15 A recent study by Truong et al¹⁶ found that TBP used to monitor all existing melanocytic lesions on patients substantially reduced the number of biopsies that patients required. These results reflect that most nevi, including clinically atypical nevi, are usually stable and unlikely to exhibit suspicious changes over time.^17,18 For this reason, the use of TBP could minimize unnecessary biopsies because clinically suspicious but stable nevi can be objectively documented and followed over time.

Standardized TBP also offers the ability for dermatologists to work synergistically with modern computer technology involving algorithms capable of analyzing high-quality images to autodiagnose or flag concerning lesions that may require biopsy. Esteva et al¹⁹ described their development of a deep learning algorithm that relies on a convolutional neural network (CNN). This CNN was trained to identify melanomas using a large data set of clinical dermatologic images and subsequently was able to distinguish MMs from benign nevi at a rate on par with a board-certified dermatologist.¹⁹ Widespread use of total-body imaging would create an enormous database of high-resolution images that would be ideally suited to the development of such computerized algorithms, which could then be applied to future images by way of artificial intelligence. Convolutional neural networks that use a single patient’s imaging over time could be developed to assess the change in number or size of benign nevi and identify lesions that are concerning for MM while simultaneously comparing them to a population-based data set.

On a large scale, such a capability would minimize the inefficiency and subjectivity of TBSEs as a tool for identifying malignancy. Currently, dermatologists are only able to track and document a few concerning lesions on a patient’s body, rendering the choice of which lesions require biopsy more subjective. Total-body photography, particularly if used with an algorithm capable of quickly analyzing all the nevi on a person’s body, largely eliminates such subjectivity by creating a standardized set of images that can be tracked over time and flagging concerning lesions prior to the dermatologist examining the patient. In this way, the specialty of dermatology could achieve the same model of objective evaluation of standardized clinical images as those employed in radiology, cardiology, and other clinical disciplines. The additional benefit of such a system would be lower costs, as the images could be acquired by nonmedical personnel and then undergo initial assessment by an algorithm, which would flag concerning lesions, similar to a modern electrocardiogram machine, allowing the dermatologist to use his/her time more efficiently by only focusing on concerning lesions with the confidence that the patient’s entire body has already been rigorously screened.

By using TBP to improve biopsy efficiency and the objectivity of the TBSE as a tool to detect skin cancer, we propose that the benefit-to-harm ratio of the TBSE would remarkably improve. Ultimately, this type of screening would meet the strict requirements to be included in preventive health care strategies and thereby improve access to dermatologic care.