[494] Evaluation of the Impact of International Society for Cutaneous Lymphoma (ISCL) Scoring Algorithm on Diagnosis of Early Mycosis Fungoides

H Vakil, M Raoufi, P Sochacki, N Khoury, W Pfeifer, E Levi. Wayne State University, Detroit, MI; Henry Ford Hospital, Detroit, MI; VA Medical Center, Detroit, MI; Boston University, Boston, MA

Background: Recently, ISCL proposed a scoring algorithm for diagnosis of early Mycosis Fungoides (MF), incorporating clinical, histopathologic, molecular and immunopathologic criteria. However, specifying the histopathologic features by which the early MF may be differentied from inflammatory diseases is still challenging, controversial and yet crucial issue. This study is designed to evaluate how the algorithm proposed by ISCL for scoring histopathologic criteria may affect the interpertation of early lesions of MF.
Design: We selected 20 patients from our archives (1992-2008) with known MF, and total number of 30 biopsies (including the ones which were not originally diagnosed as MF) were retrieved. We also added 10 cases of benign dermatoses from patients with no history of MF. Two pathologists were asked to blindly score all 40 biopsies based on the ISCL histopathologic criteria (epidermotropism without spongiosis, and lymphoid atypia). After the blinded scoring, the clinical and laboratory data were incorporated, and the initial diagnoses were reassessed based on ISCL scoring criteria. We investigated the inter-rater agreement for each of these two criteria, as well as for the final diagnoses.
Results: The inter-rater agreement (kappa coefficinet) for epidermotropism without spongiosis was 0.31, for lymphoid atypia 0.41, and for diagnosis of MF 0.45. After comparing the diagnoses made by two pathologists with original ones, 13% and 43% of cases were upstaged from early benign dignosis to MF. Interstingly, only one out of ten cases of benign dermatoses was diagnosed as MF by one of the pathologists. The most common diagnoses in prior biopsies were subacute eczematous dermatitis (10%), poikiloderma (10%), and fixed drug erruption (6.7%).
Conclusions: This study shows that application of algorithm proposed by ISCL for scoring histopathologic criteria contribute to improve the diagnostic yield for early lesions of MF, and upstage the diagnoses in a significant number of cases which were primarily diagnosed as benign lesions. However, it is worth to note that utility of histopathologic criteria is preserved by the interdependence on the other diagnostic criteria proposed by ISCL, i.e. clinical, molecular, and immunopathologic to establish the diagnosis of early MF.
Category: Dermatopathology

Tuesday, March 10, 2009 1:00 PM

Poster Session IV # 59, Tuesday Afternoon