Predictive biomarkers and tumor microenvironment in female genital melanomas: a multi-institutional study of 55 cases.

Female genital melanomas are rare. At diagnosis, most affected patients have advanced disease. Surgery remains the primary treatment, and adjuvant therapy is largely ineffective. Recently, immune checkpoints and the mitogen-activated protein kinase pathway have been explored as treatment targets. However, evaluation of these biomarkers in genital melanomas is limited. We evaluated the clinicopathological features of 20 vulvar, 32 vaginal, and three cervical melanomas and assessed programmed cell death ligand 1 (PD-L1) expression, CD8 tumor-infiltrating lymphocyte density, mismatch repair proteins, VE1 immunohistochemistry, and KIT and BRAF mutations. The median age of the patients was 66 years, and median tumor sizes were 25, 30, and 20 mm for vulvar, vaginal, and cervical tumors, respectively. Mean mitotic figures were 18, 19, and 30 per mm2. Thirty-seven patients (67%) had operable tumors. After a median follow-up of 15 months, only nine patients (16%) were alive. Eight of the nine survivors did not have lymph node metastasis. Using 5% as the threshold, PD-L1 expression was observed in 55%, 50%, and 33% of vulvar, vaginal, and cervical tumors, respectively, when the Roche SP263 antibody was used and 20%, 53%, and 0%, respectively, when the Dako 28-8 antibody was used. The median CD8 tumor-infiltrating lymphocyte density was significantly higher in vulvar/vaginal than cervical melanomas and correlated with PD-L1 expression. No cases exhibited loss of mismatch repair proteins. Five cases harbored KIT mutations, three of which were hotspots. BRAF V600E mutation was not detected. Univariable analysis showed that tumor size greater than or equal to 33 mm, mitotic figures of greater than or equal to 10 per mm2, lymph node metastasis, and low CD8+ tumor-infiltrating lymphocyte density were adverse prognostic factors. Thus, patients with genital melanomas have a poor prognosis, and evaluation of multiple biomarkers is necessary to identify patients who may benefit from immunotherapy or targeted therapy.

Mapping high-fidelity volume rendering for medical imaging to CPU, GPU and many-core architectures.

Medical volumetric imaging requires high fidelity, high performance rendering algorithms. We motivate and analyze new volumetric rendering algorithms that are suited to modern parallel processing architectures. First, we describe the three major categories of volume rendering algorithms and confirm through an imaging scientist-guided evaluation that ray-casting is the most acceptable. We describe a thread- and data-parallel implementation of ray-casting that makes it amenable to key architectural trends of three modern commodity parallel architectures: multi-core, GPU, and an upcoming many-core Intel architecture code-named Larrabee. We achieve more than an order of magnitude performance improvement on a number of large 3D medical datasets. We further describe a data compression scheme that significantly reduces data-transfer overhead. This allows our approach to scale well to large numbers of Larrabee cores.