Diagnostic Algorithm for Secondary Extramammary Paget Disease from Institutional Cases and Literature Review

<b>Background/Objectives</b>: Although clinicopathologic correlation with integration of clinical and radiographic data is the gold standard in distinguishing primary extramammary Paget disease (EMPD) from secondary EMPD, immunoprofiling of EMPD tumors enables distinction between primary and secondary EMPD. <b>Methods</b>: We evaluated the immunoprofiles of previously published cases in the literature as well as 12 secondary EMPD cases from our archives in order to construct a diagnostic algorithm that enables the distinction between primary and secondary EMPD. <b>Results</b>: Immunoprofiles of 480 primary (published cases) and 132 secondary (120 published cases and 12 institutional cases) EMPD cases were compared. CK7, CK20, CDX2, GATA3, GCDFP15, TRPS1, and SATB2 expression was significantly different in primary EMPD versus colonic secondary EMPD (<i>p</i> < 0.001 for all except SATB2, <i>p</i> = 0.036). CK20, GCDFP15, TRPS1, p63 and uroplakin II/III expression was significantly different in primary EMPD versus urothelial secondary EMPD (<i>p</i> < 0.001). CK7, CDX2, SATB2, GATA3 and p63 expression was significantly different in colonic versus urothelial secondary EMPD. CK20, CDX2, and GCDFP15 expression was significantly different in colonic versus prostatic secondary EMPD. CK20 expression was significantly different in colonic versus prostatic secondary EMPD (<i>p</i> = 0.018). CK20, GCDFP15 and TRPS1 are helpful in the distinction of primary EMPD versus colonic and urothelial secondary EMPD (<i>p</i> < 0.001). <b>Conclusions</b>: We propose that the initial IHC panel should include TRPS1, CK7 and CK20. In TRPS1-negative cases, additional immunostains should be performed: CDX2 and SATB2 for colonic; p63, GATA3 and uroplakin II/III for urothelial; and PSA and NKX3.1 for prostatic secondary EMPD.