Histological correlates of optical coherence tomography in non-melanoma skin cancer.

BACKGROUND: Non-melanoma skin cancer (NMSC) is rarely fatal but is now the most common malignancy occurring in white populations, accounting for 70% of the cost of managing skin cancer. Optical coherence tomography (OCT) has the potential to improve diagnostic accuracy and help delineate pre-surgical margins in NMSC. Its widespread clinical acceptance awaits the accumulation of evidence from studies of direct histological comparisons. METHOD: In this study, seventy-eight subjects presenting with skin lesions, including 28 NMSCs, were imaged using the VivoSight OCT scanner and a biopsy taken. Haemotoxylin and eosin stained histology sections were compared with the OCT images. RESULTS: The depth of superficial basal cell carcinoma (BCC) lesions (<1 mm) can be measured accurately using OCT. A low-strength OCT signal at the periphery of the cell nests seen in superficial and nodular BCC is identified as corresponding to cellular palisading. A weak inverse linear correlation (r(2) = 0.3) is found between the optical attenuation coefficient measured on OCT and the nuclear-cytoplasmic ratio (N/C) of cells determined from histology. CONCLUSIONS: OCT has clinical value in providing accurate dimensional measurement of superficial BCC and in identifying the presence of peripheral palisading in nodular BCC.

A genomic and expression study of AP-1 in primary cutaneous T-cell lymphoma: evidence for dysregulated expression of JUNB and JUND in MF and SS.

Activator protein 1 (AP-1) consists of a group of transcription factors including the JUN and FOS family proteins with diverse biological functions. This study assessed the genomic and expression status of the AP-1 transcription factors in primary cutaneous T-cell lymphoma (CTCL) by using immunohistochemistry (IHC), Affymetrix expression microarray, real-time reverse transcriptase-polymerase chain reaction (RT-PCR) and fluorescent in situ hybridization (FISH). IHC showed JUNB protein expression in tumor cells from 17 of 33 cases of Sezary syndrome (SS) and JUND protein expression in 16 of 23 mycosis fungoides cases. There was no correlation between JUNB and CD30 expression. However, 7 of 12 JUNB-positive SS cases expressed both phosphorylated and total extracellular signal-regulated kinase (ERK) 1/2 mitogen-activated protein kinase (MAPK) proteins. Expression microarray showed over threefold increased expression of JUNB in three of six SS patients and similar findings were also noted after re-analysis of previously published data. Real-time RT-PCR confirmed the overexpression of JUNB in four SS cases and of JUND in three of four cases. FISH showed increased JUNB copy number in four of seven SS cases. These findings suggest that deregulation of AP-1 expression in CTCL is the result of aberrant expression of JUNB and possible JUND resulting from genomic amplification and constitutive activation of ERK1/2 MAPK in this type of lymphoma.

Heterogeneous abnormalities of CCND1 and RB1 in primary cutaneous T-Cell lymphomas suggesting impaired cell cycle control in disease pathogenesis.

Upregulation of cyclin D1/B-cell leukemia/lymphoma 1 (CCND1/BCL1) is present in most mantle cell lymphomas with the t(11;14)(q13;q32) translocation. However, little is known about the abnormalities of CCND1 and its regulator RB1 in primary cutaneous T-cell lymphomas (CTCL). We analyzed CCND and RB status in CTCL using fluorescent in situ hybridization (FISH), immunohistochemistry (IHC), and Affymetrix expression microarray. FISH revealed loss of CCND1/BCL1 in five of nine Sézary syndrome (SS) cases but gain in two cases, and RB1 loss in four of seven SS cases. IHC showed absent CCND1/BCL1 expression in 18 of 30 SS, 10 of 23 mycosis fungoides (MF), and three of 10 primary cutaneous CD30+ anaplastic large-cell lymphoma (C-ALCL). Increased CCND1/BCL1 expression was seen in nine MF, seven C-ALCL, and six SS cases. Absent RB1 expression was detected in 8 of 12 MF and 7 of 9 SS cases, and raised RB1 expression in 7 of 8 C-ALCL. Affymetrix revealed increased gene expression of CCND2 in four of eight CTCL cases, CCND3 in three cases, and CDKN2C in two cases with a normal expression of CCND1 and RB1. These findings suggest heterogeneous abnormalities of CCND and RB in CTCL, in which dysregulated CCND and RB1 may lead to impaired cell cycle control.

Three cases of lymphomatoid papulosis with a CD56+ immunophenotype.

We report 3 cases of lymphomatoid papulosis (LyP) with a CD56+, cytotoxic immunophenotype. All 3 patients presented with clinical histories typical of LyP, with one patient having associated mycosis fungoides. Histologically, two cases were type A LyP and one was type B. All 3 cases demonstrated a T-cell receptor clone in lesional skin without evidence of blood involvement. The atypical lymphocytes in each of the 3 cases expressed cytotoxic granules (T-cell intracellular antigen-1+ and granzyme B+) and were CD8+ and CD56+. Expression of CD56 is associated with a poor prognosis in subcutaneous panniculitis-like T-cell lymphoma and blastic natural killer cell lymphoma. However, the two cases of CD56+ LyP previously reported and the 3 cases in this series all appear to be pursuing an indolent course with no evidence of systemic disease.

Mycosis fungoides with a CD56+ immunophenotype.

We report 3 cases of mycosis fungoides (MF) with a CD56+ cytotoxic immunophenotype. Each patient presented with a different clinical phenotype: one exhibited limited poikilodermatous patches (skin stage T1); one, widespread hypopigmented lesions (skin stage T2); and one, poikiloderma with a single cutaneous tumor (skin stage T3). MF was confirmed both histologically and by the presence of a T-cell receptor clone in lesional skin in all cases. CD56 and T-cell intracellular antigen-1 were expressed by the malignant lymphocytes in all patients and two expressed CD8. No sample demonstrated loss of the pan T-cell markers CD2 or CD3. None of the 3 developed systemic disease and T-cell receptor gene analysis of peripheral blood was polyclonal in all cases. Only 3 cases of CD56+ MF have been reported previously, none of which exhibited tumor-stage disease. Currently, the disease in our patients appears to be behaving in a manner similar to that predicted for MF with a normal immunophenotype but the prognosis has to be guarded in view of the rarity of this subtype.

Microsatellite instability is associated with hypermethylation of the hMLH1 gene and reduced gene expression in mycosis fungoides.

Fifty-one mycosis fungoides samples were analyzed for microsatellite instability (MSI) using the panel of markers recommended for hereditary nonpolyposis colorectal cancer kindred and a panel we designed for cutaneous T cell lymphoma in order to compare detection rates and determine if MSI is a genome-wide phenomenon. Samples demonstrating MSI were analyzed for abnormalities of the hMLH1 gene including loss of heterozygosity, mutations, and promoter hypermethylation. MSI was detected in 16% using the hereditary nonpolyposis colorectal cancer panel and 22% with the cutaneous T cell lymphoma panel. Overall, 27% demonstrated MSI and 73% had a stable phenotype. hMLH1 gene studies did not detect loss of heterozygosity or reveal any mutations. Promoter hypermethylation was detected in nine of 14 patients with MSI, however (64%). In addition hMLH1 and hMSH2 protein expression was studied using immunohistochemical techniques. Five of nine patients with MSI and hMLH1 promoter methylation showed abnormal hMLH1 protein expression with normal hMSH2 gene expression. All other patients tested demonstrated normal hMLH1 and hMSH2 protein expression. MSI was found to be more prevalent in tumor stage mycosis fungoides (47%) than early stage disease (20%) and was associated with an older age of onset of mycosis fungoides. MSI may be a consequence of hMLH1 promoter hypermethylation in mycosis fungoides patients and may prevent transcription in a subset of patients. This suggests that the development of a mutator phenotype may contribute to disease progression in mycosis fungoides.

Inactivation of tumor suppressor genes p15(INK4b) and p16(INK4a) in primary cutaneous B cell lymphoma.

Primary cutaneous B cell lymphomas represent a distinct group of lymphoproliferative disorders that can be distinguished from systemic lymphoma by their good response to local treatment and favorable prognosis. In systemic B cell lymphoma, inactivation of p15(INK4b) and p16(INK4a) is frequently observed and may be associated with a poor prognosis. There have been no comprehensive studies in primary cutaneous B cell lymphomas, however. Mechanisms of p15/p16 inactivation include loss of heterozygosity, homozygous deletion, promotor region hypermethylation, and point mutation. We analyzed DNA from 36 cases of primary cutaneous B cell lymphomas, four systemic B cell lymphomas, and six benign B cell lymphoproliferative infiltrates for abnormalities of p15 and p16 using microsatellite markers for 9p21, methylation specific polymerase chain reaction, and polymerase chain reaction/single stranded conformational polymorphism analysis with exon specific primers. Expression of both p15 and p16 protein was assessed by immunohistochemistry. Loss of heterozygosity at 9p21 was identified in 2 out of 36 primary cutaneous B cell lymphomas. Hypermethylation of p15 and p16 promotor regions was identified in 8 of 35 (23%) and 15 of 35 (43%) cases, respectively. In two cases p16 hypermethylation was identified in recurrent disease but not in the initial tumor. No point mutations were identified. In seven patients, however, a polymorphism was observed in exon 3 of the p16 gene. In primary cutaneous B cell lymphomas with allelic loss or promotor hypermethylation of either p15 or p16, loss of expression in tumor cells was identified in 5 of 8 and 9 of 10 cases, respectively. Our findings suggest that p15(INK4b) and p16(INK4a) biallelic gene abnormalities are common in primary cutaneous B cell lymphomas, most frequently as a result of promotor hypermethylation. The presence of abnormalities in recurrent disease in some cases suggests that inactivation of p15 and p16 may be involved in disease progression.

Reduced expression of insulin-like growth factor-binding protein-3 (IGFBP-3) in Squamous cell carcinoma complicating recessive dystrophic epidermolysis bullosa.

Squamous cell carcinoma (SCC) is a common complication in individuals with recessive dystrophic epidermolysis bullosa (RDEB). For the severe Hallopeau-Siemens subtype, the mortality rate from SCC is over 55% by the age of 40 y. Currently, little is known about the molecular pathology or cell biology of SCC in RDEB. In this study, we compared gene expression in RDEB SCC (n=3) and non-EB SCC (n=3) with corresponding RDEB and non-EB peri-tumoral skin, with microarray analysis using DermArray membranes as well as semi-quantitative and real-time RT-PCR. Both tumor sets showed downregulation of epidermal differentiation markers (e.g., profilaggrin, keratins 1 and 10) as well as certain pro-apoptotic genes (e.g., death-associated kinase-3 or ZIP kinase). Likewise, in both groups there was upregulation of matrix metalloproteinase 1 and laminin 5 in the tumors. But we found that the expression of insulin-like growth factor-binding protein-3 (IGFBP-3) was lower (mean of 5.8-fold) in RDEB SCC compared with non-EB SCC. These data were verified by immunohistochemistry. IGFBP-3 has an important role in cancer cell apoptosis mediated via the nuclear retinoid X receptor alpha (RXRalpha). Reduced expression of IGFBP-3 in RDEB SCC may provide a partial explanation for the aggressive behavior and poor prognosis of these tumors in this genodermatosis.

Frequent abnormalities of the p15 and p16 genes in mycosis fungoides and sezary syndrome.

There are few data on the molecular pathogenesis of cutaneous T cell lymphomas. A recent allelotyping study by our group identified frequent allelic loss on 9p, 10q, and 17p including losses on 9p21 in 16% of patients with mycosis fungoides and 46% with Sezary syndrome. The P15 and P16 genes are intricately linked on 9p21 and can be inactivated in melanoma and non-Hodgkin's lymphoma. We have therefore studied 76 patients with either mycosis fungoides or Sezary syndrome for abnormalities of these genes. DNA samples were analyzed for loss of heterozygosity, homozygous deletion, intragenic mutations, and promoter methylation. In addition P15 and P16 protein expression was assessed. Microsatellite analysis was informative in 73 of 76 cases: allelic loss on 9p21 was identified in 18 patients (25%), including 12 of 57 with mycosis fungoides (21%) and six of 16 with Sezary syndrome (37%). Single strand conformation polymorphism analysis of the entire coding regions of both genes did not identify any mutations, although two polymorphisms were identified including C613A, which has not previously been described. P15 and P16 gene promoter methylation was found in 45% and 29% of patients, respectively. Furthermore aberrant P15 protein expression was detected in 85% of patients analyzed with P15 gene abnormalities and abnormal P16 expression in 59% with P16 gene abnormalities. These abnormalities were not dependent on cutaneous stage of disease. This study suggests that abnormalities of the P15 and P16 genes are common in both early and advanced stages of mycosis fungoides and Sezary syndrome and that these genes may be inactivated by allelic loss and aberrant promoter methylation.