Accuracy of In-vivo Confocal Imaging for Pigmented Lesion Diagnosis

DESCRIPTION (provided by applicant): The effort under this proposal is to prepare for and perform a large clinical study using in-vivo confocal microscopy to validate the results of a smaller, recently published study of the efficacy of this technology for diagnosis of melanoma. The main goal of this Type 2 Phase II Competing Continuation SBIR Proposal is to obtain data for submission to the FDA as a step in the Federal regulatory approval process. The clinical research plan is intended to demonstrate the sensitivity, specificity, positive predictive value, negative predictive value and the test efficiency of in-vivo confocal microscopy for detection of the most deadly skin cancer - melanoma - in a population of approximately 600 patients presenting with a suspicious pigmented lesion on clinical examination. The American Cancer Society estimates that in 2005 approximately 59,580 new cases of invasive melanoma will be diagnosed in the United States, and that 7,770 deaths will occur due to the disease. If melanoma is diagnosed and removed while the tumor is localized, the five-year survival rate is 97.6%; however if the lesion has metastized to either a regional or distant site, the five-year survival rate drops to 60.3% or 16.2%, respectively [Cancer Facts and Figures, 2005]. Consequently, the need for early detection is crucial to a melanoma patient's long-term survival of this disease. This proposal represents one of the final steps in the technological innovation needed to bring a revolutionary medical device to the market. Confocal microscopy can non-invasively image skin in-vivo with cellular resolution. This potentially permits pathologic diagnosis without the need to excise tissue. Just as MRI and CT scans have largely eliminated the need for routine "exploratory surgery", in-vivo reflectance confocal microscopy may one day eliminate the need for routine invasive skin biopsy. The proposed study has a high probability of a positive outcome. A recently published pilot study of the efficacy of confocal imaging for the diagnosis of melanoma demonstrated high sensitivity and specificity for images read by a panel of diagnostic readers with relatively little training (30 minutes). Consequently, the proposed study, a comparison of confocal imaging to the "gold standard" of pathologic examination of excised tissues, conducted at leading U.S. clinical sites and utilizing well-trained diagnostic readers, should produce a compelling clinical data for use in the Federal regulatory approval process. In addition, the project will produce and submit for publication an extensive atlas of confocal images and a glossary of terms for the description of confocal images of pigmented lesions.