VEGF and VEGFR family members are expressed by neoplastic cells of NF1-associated tumors and may play an oncogenic role in malignant peripheral nerve sheath tumor growth through an autocrine loop.

Neurofibromatosis type 1 (NF1) is a common autosomal dominant disorder. The role of angiogenesis and VEGF pathway in the pathogenesis of neurofibromas and malignant peripheral nerve sheath tumors (MPNSTs) remains poorly understood. We assessed the expression of VEGF and VEGFR family members in cohorts of plexiform neurofibromas (pNF), MPNSTs and MPNST cell lines at transcript [pNF, n = 49; MPNST, n = 34] and protein levels [pNF, n = 21; MPNST, n = 9]. VEGF and VEGFR members were variably expressed in cell lines. VEGFA (p = 3.10-5), VEGFR1 (p = 0.08), and VEGFR2 (p = 2.10-4) mRNAs were overexpressed in MPNSTs in comparison with pNFs. Both VEGFA and VEGFR1 proteins were expressed by spindle tumor cells of pNFs and MPNSTs. VEGFA was expressed more in MPNSTs than in pNFs (p = 9.10-6) and a trend for VEGFR1 overexpression was observed (p = 0.06). VEGFR2 was not found at the protein level. The microvascular density was significantly reduced in MPNSTs as compared to pNFs (p = 0.0025), with no differences regarding the expression of the activated phosphorylated forms of ERK (P-ERK [p = 0.63]) and AKT (P-AKT [p = 0.41]) in endothelial cells, suggesting that VEGF-dependant angiogenesis may not be critical for MPNST oncogenesis. Altogether, these results indicate that the VEGF-VEGFR pathway may play a role in the development of pNFs and MPNSTs, independently of angiogenesis. Whether or not it drives an oncogenic autocrine/paracrine loop in neoplastic cells, participating in an increased activation of signaling pathways downstream of tyrosine kinase receptors, including VEGFRs, is a tempting hypothesis. Nevertheless, the specific targeting of angiogenesis in MPNSTs may not be sufficient to slow down tumor growth.

Prevalence of T-cell antigen losses in mycosis fungoides and CD30-positive cutaneous T-cell lymphoproliferations in a series of 153 patients.

Mycosis fungoides (MF) and primary cutaneous CD30-positive T-cell lymphoproliferative disorders (CD30LPD) are the most frequent primary cutaneous T-cell lymphomas. Our objective was to study pan-T-cell antigens and PD-1 expression in a large cohort of MF and CD30LPD with a special interest in antigen losses as a diagnostic tool. We retrospectively reviewed 160 consecutive samples from 153 patients over a 3 year period, including 104 with MF and 49 with CD30LPD. As controls, benign inflammatory dermatoses (BID, n=19) were studied. A semi-quantitative evaluation of CD2, CD3, CD4, CD5, CD7, CD8 expression was performed. PD-1 and double stainings (CD3+CD7 and PD-1+CD7) were performed in a subset of MF cases. CD8+ MF was frequent (23%) and CD7 was the most frequently lost antigen in both MF (45%) and CD30LPD (86%), while no significant T-cell antigen loss was observed in BID. CD7 loss was less frequent in folliculotropic MF (p<0.001). PD-1 was variably expressed in MF with no differences with BID. The CD3+/CD7- and PD-1+/CD7- neoplastic lymphocytes were highlighted by the use of chromogenic double staining experiments in MF with a CD7 loss identified with single staining. Multiple pan T-cell antigen losses were mostly seen in CD30LPD with CD2 being the most frequently preserved marker (90%). While PD-1 does not discriminate between MF and BID, CD7 is frequently lost in MF infiltrates as well as other pan-T-cell antigens in CD30LPD, which can be used as routine markers for diagnosis. We recommend the use of CD7 in addition to CD3, CD4 and CD8 as a minimal immunohistochemical panel for MF assessment, and the use of double stainings for CD3 and CD7 in difficult cases.