Papillary thyroid carcinoma and its variants can be difficult to distinguish from cellular adenomatous nodules. Prior studies have advocated various antibodies to aid in the differential diagnosis, but there is little agreement on their utility. We undertook this study to evaluate immunohistochemical markers in the diagnosis and differential diagnosis of papillary thyroid carcinoma. Ten cases of papillary thyroid carcinoma were initially stained for HBME1, CK19, fibronectin1, Ki-67, Calretinin, p16, SFTPB and CITED1. Additionally, two previously untested antibodies to molecules that have been found to be upregulated in papillary thyroid carcinoma (CST6 and EPS8) were also evaluated. Of these, only HBME1, CK19 and fibronectin1 showed diagnostic utility. These three markers were then further evaluated in 51 papillary thyroid carcinomas and 57 benign thyroids. HBME1 was the most sensitive and specific marker, staining 49/51 papillary thyroid carcinomas and only 4/57 benign thyroids. CK19 was equally sensitive staining all 51 papillary thyroid carcinomas, but it was nonspecific staining 39 of 57 benign thyroids. A negative result, however, was helpful in excluding papillary thyroid carcinoma. Fibronectin1 was positive in 35/51 papillary thyroid carcinomas (69%) and 4/57 (7%) benign thyroids, but its utility was hampered by high background staining. These findings suggest that the combination of HBME1 and CK19 has the greatest diagnostic utility in the differentiation of papillary thyroid carcinoma from its benign mimics.

Differentiating malignant melanoma from benign melanocytic lesions can be challenging. We undertook this study to evaluate the use of the immunohistochemical mitosis marker phospho-Histone H3 (pHH3) and the proliferation markers Ki-67 and survivin in separating malignant melanoma from benign nevi. Sixty-six melanocytic lesions (18 malignant melanomas, 8 Spitz nevi, 20 dysplastic nevi, and 20 compound nevi) were stained with antibodies to pHH3, Ki-67, and survivin. No pHH3 expression was detected in the dermis of compound and dysplastic nevi. Rare mitoses were observed in the superficial dermis in 3 of 8 Spitz nevi (37%). Staining for pHH3 was higher in malignant melanomas [average 25 per 10 high-power field (HPF), range 2-75 per 10 HPF] than in Spitz nevi (average 0.5 per 10 HPF, range 0-2 per 10 HPF) and was heterogeneously distributed in the malignant melanomas compared with a superficial dermal location in Spitz nevi. There was no cytoplasmic staining for survivin in any of the 66 melanocytic lesions and no nuclear staining in any of the benign ones. Survivin nuclear staining was present in 12 of 18 cases of malignant melanoma (67%) with an average index of 7% (range 0%-15%). In benign melanocytic lesions, the Ki-67 index was less than 5% (range 0%-4%) and staining was present close to the dermo-epidermal junction compared with an average index of 27% in melanomas (range 5%-50%) and a generally heterogeneous pattern of staining throughout the dermis. pHH3 and Ki-67 can be useful adjuncts to histopathology to separate malignant melanoma from benign nevi. pHH3 is especially useful to highlight mitoses and to rapidly assess the mitotic activity in melanocytic lesions.

The response rate at relapse to rituximab in prior responders B-cell non-Hodgkin's lymphoma (NHL) patients is below 50%. Loss of CD20 expression after rituximab therapy may explain this secondary resistance. However, the frequency of CD20 negative relapses cannot be assessed since most patients that relapsed after rituximab therapy have not been re-biopsied. Here, we present two patients with CD20 positive low grade B-cell NHL that lost the cell surface and cytoplasmic expression at relapse after rituximab therapy. Our findings suggest that confirmation of CD20 expression on the malignant B cells is required whenever rituximab therapy is considered.

Primary CNS diffuse large B-cell lymphoma (PCNS-DLBCL) and systemic DLBCL harbor mutations in MYD88 and CD79B. DNA methyltransferase (MGMT) is methylated in some DLBCL. Our goal was to investigate the frequencies of these events, which have not been previously reported within the same series of patients with PCNS-DLBCL. Fifty-four cases of PCNS-DLBCL from two institutions were analyzed by Sanger sequencing for MYD88 and CD79B, and pyrosequencing for MGMT. MYD88 mutations were identified in 68.8% (35 of 51 cases), with L265P being the most frequent mutation. Mutations other than L265P were identified in 21.6% of cases, of which eight novel MYD88 mutations were identified. Of mutated cases, 17.6% had homozygous/hemizygous MYD88 mutations, which has not been previously reported in PCNS-DLBCL. CD79B mutations were found in six of 19 cases (31.6%), all in the Y196 mutation hotspot. MGMT methylation was observed in 37% (20 of 54 cases). There was no significant difference in median overall survival (OS) between the wild type and mutated MYD88 cases, or between methylated and unmethylated MGMT cases. However, a significant difference (P = 0.028) was noted in median OS between the wild type and mutated CD79B cases.