Travis Vandergriff1, Kaveh A Nezafati1, Joseph Susa1, Laszlo Karai1, Amy Sanguinetti2, Linda S Hynan3, Josephine M Ambruzs4, Dwight H Oliver4, Amit G Pandya1. 1. Department of Dermatology, University of Texas Southwestern Medical Center, Dallas, TX, USA. 2. Veterans Affairs Hospital, Sierra Nevada Health Care System, Reno, NV, USA. 3. Department of Clinical Science, University of Texas Southwestern Medical Center, Dallas, TX, USA. 4. Department of Pathology, University of Texas Southwestern Medical Center, Dallas, TX, USA.
Abstract
BACKGROUND: Mycosis fungoides (MF) is the most common subtype of cutaneous T-cell lymphoma and is often difficult to diagnose. Early-stage disease is particularly challenging and requires clinical and histopathologic correlation to make an accurate diagnosis. In order to facilitate the diagnosis of early MF, an algorithm has been proposed by the International Society for Cutaneous Lymphomas (ISCL) whereby clinical and histopathologic characteristics as well as immunohistochemistry and T-cell receptor gene rearrangement studies may be applied to suspected cases of MF. The diagnostic utility of this algorithm has not yet been validated. We sought to determine the validity of the proposed algorithm via an investigator-blinded, retrospective, case-control study. METHODS: A total of 34 cases were randomly selected from the database of a clinic for cutaneous T-cell lymphomas and included patients with MF and patients with clinicopathologic mimics. The proposed diagnostic algorithm was systematically applied to the entire cohort. Each case was assigned a composite score based on the parameters in the proposed algorithm. RESULTS: Among the 24 cases of MF, 21 cases achieved four or more points through application of the algorithm. Among the 10 cases of MF mimics, only four achieved four or more points. This difference was significant (Fisher's exact test, p = 0.009). The sensitivity of the 4-point threshold for a diagnosis of MF was 87.5% and the specificity was 60%. CONCLUSIONS: The diagnostic algorithm proposed by the ISCL is a statistically valid method for defining cases of early MF and distinguishing these cases from other benign dermatoses. However, the clinical utility of the algorithm may be limited by its low specificity. Further refinement of the algorithm may improve its accuracy.
BACKGROUND:Mycosis fungoides (MF) is the most common subtype of cutaneous T-cell lymphoma and is often difficult to diagnose. Early-stage disease is particularly challenging and requires clinical and histopathologic correlation to make an accurate diagnosis. In order to facilitate the diagnosis of early MF, an algorithm has been proposed by the International Society for Cutaneous Lymphomas (ISCL) whereby clinical and histopathologic characteristics as well as immunohistochemistry and T-cell receptor gene rearrangement studies may be applied to suspected cases of MF. The diagnostic utility of this algorithm has not yet been validated. We sought to determine the validity of the proposed algorithm via an investigator-blinded, retrospective, case-control study. METHODS: A total of 34 cases were randomly selected from the database of a clinic for cutaneous T-cell lymphomas and included patients with MF and patients with clinicopathologic mimics. The proposed diagnostic algorithm was systematically applied to the entire cohort. Each case was assigned a composite score based on the parameters in the proposed algorithm. RESULTS: Among the 24 cases of MF, 21 cases achieved four or more points through application of the algorithm. Among the 10 cases of MF mimics, only four achieved four or more points. This difference was significant (Fisher's exact test, p = 0.009). The sensitivity of the 4-point threshold for a diagnosis of MF was 87.5% and the specificity was 60%. CONCLUSIONS: The diagnostic algorithm proposed by the ISCL is a statistically valid method for defining cases of early MF and distinguishing these cases from other benign dermatoses. However, the clinical utility of the algorithm may be limited by its low specificity. Further refinement of the algorithm may improve its accuracy.