Background:Desmoplastic melanoma (DMM) is an uncommon melanoma variant with a distinct morphology, including a prominent spindle cell component with fibrosis , as well as a distinct immunohistochemical profile. Histologically, the spindle cell component of DMM can be confused with sclerotic/desmoplastic nevi, non-pi gmented blue nevi, scar, and neural tumors. The histological distinction between sclerotic/desmoplastic/blue nevi and DMM using standard light microscopic techn iques can be exceedingly subtle. Therefore, we investigated whether immunohistoc hemical staining for Melan-A and Ki-67 expression can be used to discriminate these lesions, distinguishing between epithelioid and spindle cell compartments of the lesions. Design:Fifty cases of DMM and 13 cases of sclerotic/desmoplasti c/blue nevi were identified. Standard immunohistochemical techniques were used w ith antibodies towards HMB-45, Melan-A (Al03), and Ki-67; 43 of 50 DMM cases were available for staining with Melan-A, 42 of 50 for HMB-45, and 31 of 50 ca ses were stained with Ki-67. All 13 nevi were stained for Melan-A and 8 for Ki -67. Immunoreactivity to Ki-67 antibody was scored as O to 5%, 6 to 10%, 11 to 30%, or greater than 30%positive tumor cells. Results:Only 3 of 43 and 3 o f 42 of spindle cell compartments of DMMs were positive for Melan-A and HMB-45 , respectively. Focal staining of epithelioid cells in the junctional component or superficial demis was observed in 33%(14/43). In contrast, 100%of the 13 ne vi were strongly positive for Melan-A (P <.0.001). Seventeen melanomas (55%) w ere O to 5%positive for Ki-67, five (16%) fell into the 6 to 10%category, th ree (10%)were between 11 and 30%, and six (19%) were at least focally greater than 30%positive. All 8 nevi (100%) had less than 5%positive cells for Ki-6 7 (P=0.02), with only 2 cases having more than 2%positive cells. Conclusion:Th e sclerotic/desmoplastic and hypopigmented blue nevi were uniformly positive for Melan-A, while the vast majority of DMM were negative in their spindle cell co mpartments. Melan-A is very useful in distinguishing between DMM and sclerotic nevi. Ki-67 appears to be an inconsistent marker for DMM. However, a high label ing index (over 5%) may be used as a clue in diagnosing DMM. Background: Desmoplastic melanoma (DMM) is an uncommon melanoma variant with a distinct morphology, including a prominent spindle cell component with fibrosis, as well as a distinct immunohistochemical profile. Histologically, the spindle cell component of DMM can be confused with sclerotic / desmoplastic nevi , non-pi gmented blue nevi, scar, and neural tumors. The histological distinction between sclerotic / desmoplastic / blue nevi and DMM using standard light microscopic techniques can be exceedingly subtle. Thus, we investigated whether immunohistoc hemical staining for Melan-A and Ki-67 expression can be used to discriminate these lesions, distinguishing between epithelioid and spindle cell compartments of the lesions. Design: Fifty cases of DMM and 13 cases of sclerotic / desmoplasti c / blue nevi were identified. Standard immunohistochemical techniques were used w ith Antibodies to HMB-45, Melan-A (Al03), and Ki-67; 43 of 50 DMM cases were available for staining with Melan-A, 42 of 50 for HMB-45, and 31 of 50 ca ses were stained with Ki-67. All 13 nevi were stained for Melan-A and 8 for Ki-67. Immunoreactivity to Ki-67 antibody was scored as O to 5% 6 to 10%, 11 to 30%, or greater than 30% positive tumor cells. Results: Only 3 of 43 and 3 of 42 of spindle cell compartments of DMM were positive for Melan-A and HMB-45, respectively. Focal staining of the epithelioid cells in the junctional component or superficial demis was observed in 33% (14/43). In contrast, 100% of the 13 ne vi were strongly positive for Melan-A (P <.0.001). Seventeen melanomas (55% (16%) fell into the 6 to 10% category, th ree (10%) were between 11 and 30%, and six (19%) were at least focally All 8 nevi (100%) had less than 5% positive cells for Ki-6 7 (P = 0.02), with only 2 cases having more than 2% positive cells. Conclusion: Th e sclerotic / desmoplastic and hypopigmented blue nevi were uniformly positive for Melan-A, while the vastMost of DMM were negative in their spindle cell co mpartments. Melan-A is very useful in distinguishing between DMM and sclerotic nevi. However, a high label ing index (over 5%) may be used as a clue in diagnosed DMM.