SDPA Presents 20th Annual Fall Dermatology Conference–Abstracts and Posters
November 16-20, 2022, Miami, FL
7. Validation of a Handheld Elastic-scattering Spectroscopy Device on Lesions Suggestive of Melanoma
Rebecca I Hartman, MD, MPH1; Kelly Tepedino, MD3; Maxwell A Fung, MD4; Jennifer M McNiff, MD5; Genevieve Patrick, BS2; Vivien Nguyen, PharmD7; Kiran Chatha, MD7; Jane Grant-Kels, MD6
1Brigham and Women’s Hospital, USA; 2Florida State University College of Medicine, USA; 3North Florida Dermatology, USA; 4University of California Davis School of Medicine, USA; 5Yale University School of Medicine, USA; 6University of Connecticut, USA; 7DermaSensor, Inc., USA
Introduction: Skin cancer is the most common cancer worldwide; leveraging technology for skin cancer evaluation may aid in reducing associated mortality and morbidity, avoiding unnecessary biopsies, and decreasing healthcare costs. Elastic-scattering spectroscopy (ESS) is an optical tissue sampling technique that can distinguish between normal and abnormal tissue in vivo by recording photons scattered back from chromophores. A handheld, non-invasive ESS device utilizing machine learning to evaluate subcellular changes may promote earlier identification of melanomas. Our study aimed to investigate the sensitivity and specificity of the ESS device in skin lesions suspicious for melanoma.
Methods: A blinded, prospective study was performed at 10 dermatology centers in the U.S. and Australia. Patients with lesions clinically suspicious for melanoma were enrolled for evaluation by the device. The ESS device classified the lesions as either high risk or low risk for malignancy by providing a result of “Investigate Further” or “Monitor”, respectively. Additionally, “Investigate Further” results included malignancy spectral similarity scores between 1-10. Biopsy was performed with dermatopathology consensus review. Statistical analyses included sensitivity, specificity, Negative Predictive Value (NPV) and Positive Predictive Value (PPV).
Results: A total of 328 patients with 440 lesions were enrolled. Most patients were male (54%), white (98%), non-Hispanic (96%), and Fitzpatrick skin type II (53%) or III (20%). Lesions were most commonly described as flat (84%), pigmented (96%) and with a median length of 5mm. Biopsy revealed 114 high-risk lesions (44 melanomas, 44 severely atypical nevi [SAN], 26 non-melanoma skin cancers [NMSC]). Diagnostic sensitivity of the device was 96% for melanoma, 91% for melanoma + SAN, and 93% for all high-risk lesions. Overall device specificity was 33%, and 43% for mild + moderate atypia. Dermatologist’s sensitivity was 91% for melanoma, 72% for melanoma + SAN, and 84% for all high-risk lesions. The device NPV for a “Monitor” result was 98.1%for melanoma. The overall PPV for melanoma for an “Investigate Further” result was 16% (i.e. an NNB of 6) with a PPV for melanoma of 47% for spectral scores between 8-10 (i.e. an NNB of 2). The modeled area under the curve (AUC) of the ESS device for melanoma detection was comparable to that of the dermatologists, 0.76 and 0.75, respectively.
Conclusions: The high sensitivity of the ESS device can aid in detection of melanoma in clinically equivocal lesions. Furthermore, the high NPV of this device may reduce unnecessary biopsies and aid dermatologists’ assessment of lesions suspicious for melanoma.