From the 2024 HVPA National Conference
Stephen Merry MD (Mayo Clinic), Ivana Croghan PhD, Nathan Falk MD, Harold Rabinovitz MD, Elizabeth Seiverling MD, Todd Thames MD, Kiran Chatha MD MPH, David Leffell MD
Background:
Skin cancer, which is prevalent in the U.S. and is steadily increasing in incidence, often poses a diagnostic challenge. Primary care physicians (PCPs), who may have limited dermatologist access, often lack confidence deciding which skin lesions need further evaluation based on clinical assessment. Various technologies have been developed to aid these providers in their referral decisions regarding suspicious skin lesions. Elastic scattering spectroscopy (ESS) distinguishes between normal and abnormal tissue in vivo based on subcellular differences and can be used to assess equivocal skin lesions rapidly in clinic without the need for a biopsy.
Objective:
The objective of this study was to evaluate how PCPs could benefit from a novel point-of-care ESS device in evaluating challenging skin lesions.
Methods:
This blinded, prospective, multi-center study was performed at 22 primary care sites across the U.S. and Australia. A total of 1,005 subjects with 1,579 lesions suggestive of skin cancer were enrolled. Lesions were clinically evaluated by PCP investigators and then scanned with the ESS device. Patients and providers were blinded to device results. PCPs recorded their prediction of whether the lesion was malignant or benign and their confidence level in that clinical assessment. All lesions were biopsied, and final diagnosis from dermatopathology was used as the reference standard to evaluate device performance. Statistical analyses included sensitivity, specificity, and AUROC.
Results:
Among 1,579 enrolled lesions, PCPs reported “low confidence” for 648 lesions (41.0%) on 456 patients. Dermatopathology confirmed 73 (11.3%) malignant lesions (54 NMSC, 19 melanomas (including severe atypia)) in this “low confidence” subgroup. The average subject age was 59.6 years, with 51.3% female subjects. Subjects had a median of 2 risk factors for skin cancer recorded. The ESS device had an overall sensitivity of 95.5% with a corresponding 20.7% specificity when compared to dermatopathology. Device sensitivity was 97.3% for PCP-rated low-confidence lesions (with melanoma and NMSC sensitivity of 94.7% and 98.1%, respectively). Depending on the assumptions used for AUROC calculations, the range of improvement for high-confidence lesions when aided by the device reading was 3.4-11.3%, while the improvement for low-confidence lesions ranged from 11.8-20.2%. No device-related adverse events were reported.
Conclusions:
The ESS device demonstrated high sensitivity in detecting skin cancer, especially for lesions where PCPs reported having low confidence in their clinical assessment. The addition of the handheld ESS device to PCP’s clinical assessment can improve physician confidence and performance in managing equivocal skin lesions. Addressing the well-recognized gap in PCP skin cancer assessments may enhance earlier detection and improve overall patient outcomes, potentially reducing the burden of later-stage skin cancer care on hospital systems.
Clinical Implications:
Incorporation of the ESS device into clinical care can improve physician confidence and performance in appropriately managing skin lesions, leading to greater cancer detection by PCPs and an ability to triage high-risk patients to dermatologists sooner, potentially enhancing earlier treatment.