Quantitative histopathology identifies patients with thin melanomas who are at risk for metastases.
AuthorGlazer, Evan S
Bartels, Peter H
Kha, Stephanie T
Morgan, Sherif S
da Silva, Vinicius D
Yozwiak, Michael L
Bartels, Hubert G
Cranmer, Lee D
de Oliveira, Jefferson K
Alberts, David S
Warneke, James A
Krouse, Robert S
AffiliationUniversity of Arizona College of Medicine
University of Arizona Cancer Center
University of Arizona College of Science
Southern Arizona Veterans Affairs Health Care System
MetadataShow full item record
PublisherLIPPINCOTT WILLIAMS & WILKINS
CitationQuantitative histopathology identifies patients with thin melanomas who are at risk for metastases. 2016, 26 (3):261-6 Melanoma Res.
RightsCopyright © 2016 Wolters Kluwer Health, Inc. All rights reserved.
Collection InformationThis item from the UA Faculty Publications collection is made available by the University of Arizona with support from the University of Arizona Libraries. If you have questions, please contact us at firstname.lastname@example.org.
AbstractThis small exploratory study was designed to test the hypothesis that thin melanoma lesions contain nuclei of two similar phenotypes, in different proportions. In lesions likely to progress to metastatic disease, one of these phenotypes predominates. Histopathological sections from 18 cases of thin melanomas which did not progress to metastasis, and from 10 cases which did progress were imaged and digitized at high resolution, with a total of 2084 and 1148 nuclei, respectively, recorded. Five karyometric features were used to discriminate between nuclei from indolent and from potentially metastatic lesions. For each case, the percentage of nuclei classified by the discriminant function as having come from a potentially metastatic lesion was determined and termed as case classification criterion. Standard histopathological criteria, such as ulceration and high mitotic index, indicated in this material the need for intensive therapy for only one of the 10 participants, as compared with 7/10 identified correctly by the karyometric measure. Using a case classification criterion threshold of 40%, the overall accuracy was 86% in the test set. The proportion of nuclei of an aggressive phenotype may lend itself as an effective prognostic clue for thin melanoma lesions. The algorithm developed in this training set appears to identify those patients at high risk for metastatic disease, and demonstrates a basis for a further study to assess the utility of prognostic clues for thin melanomas.
Note12 month embargo.
VersionFinal accepted manuscript
SponsorsThis research was supported in part by grants from the National Cancer Institute (P01 CA27502-33, and the Arizona Cancer Center Support Grant CA023074), Bethesda, MD. The authors also gratefully acknowledge support from The Jim Click Family Foundation, Tucson, Arizona, and the J. Russell Skelton Family, Phoenix, Arizona.