No abstract

We describe here the first in vivo targeting of tumors with a single-chain antigen-binding protein. The molecule, which was constructed and expressed in Escherichia coli, is a novel recombinant protein composed of a variable light-chain (VL), amino acid sequence of an immunoglobulin tethered to a variable heavy-chain (VH) sequence by a designed peptide. We show that this protein, derived from the DNA sequence of the variable regions of the antitumor monoclonal antibody B6.2, has the same in vitro antigen-binding properties as the B6.2 Fab' fragment. Comparative pharmacokinetic studies in athymic mice demonstrate much more rapid alpha and beta phases of plasma clearance for the single-chain antigen-binding protein than for the Fab' fragment, as well as an extremely rapid whole-body clearance. Half-life values for alpha and beta phases of single-chain antigen-binding protein clearance were 2.4 minutes and 2.8 hours, respectively, versus 14.8 minutes and 7.5 hours for Fab'. Furthermore, the single-chain antigen-binding protein molecule did not show accumulation in the kidney as did the Fab' molecule or, as previously shown, the F(ab')2 molecule. Despite its rapid clearance, the single-chain antigen-binding protein showed uptake in a human tumor xenograft comparable to that of the Fab' fragment, resulting in tumor to normal tissue ratios comparable to or greater than those obtained with the Fab' fragment. These studies thus demonstrate the in vivo stability of recombinant single-chain antigen-binding proteins and their potential in some diagnostic and therapeutic clinical applications in cancer and other diseases.

Combination of chemotherapy with cancer vaccines is currently regarded as a potentially valuable therapeutic approach for the treatment of some metastatic tumors, but optimal modalities remain unknown. We designed a phase I/II pilot study for evaluating the effects of dacarbazine (DTIC) on the immune response in HLA-A2 1 disease-free melanoma patients who received anticancer vaccination 1 day following chemotherapy (800 mg/mq i.v.). The vaccine, consisting of a combination of HLA-A2 restricted melanoma antigen A (Melan-A/MART-1) and gp100 analog peptides (250 lg each, i.d.), was administered in combination or not with DTIC to 2 patient groups. The combined treatment is nontoxic. The comparative immune monitoring demonstrates that patients receiving DTIC 1 day before the vaccination have a significantly improved long-lasting memory CD8 1 T cell response. Of relevance, these CD8 1 T cells recognize and lyse HLA-A2 1 /Melan-A 1 tumor cell lines. Global transcriptional analysis of peripheral blood mononuclear cells (PBMC) revealed a DTIC-induced activation of genes involved in cytokine production, leukocyte activation, immune response and cell motility that can favorably condition tumor antigen-specific CD8 1 T cell responses. This study represents a proof in humans of a chemotherapyinduced enhancement of CD8 1 memory T cell response to cancer vaccines, which opens new opportunities to design novel effective combined therapies improving cancer vaccination effectiveness. '

Although the monocentric validation of the presented predictors might represent a limitation, these results demonstrate that a model based on MKL and RO may represent a novel methodological approach to derive VTE risk classifiers. Moreover, this study highlights the advantages of optimizing the relative importance of groups of clinical attributes in the selection of VTE risk predictors.

Machine learning (ML) has been recently introduced to develop prognostic classification models that can be used to predict outcomes in individual cancer patients. Here, we report the significance of an ML-based decision support system (DSS), combined with random optimization (RO), to extract prognostic information from routinely collected demographic, clinical and biochemical data of breast cancer (BC) patients. A DSS model was developed in a training set (n = 318), whose performance analysis in the testing set (n = 136) resulted in a C-index for progression-free survival of 0.84, with an accuracy of 86%. Furthermore, the model was capable of stratifying the testing set into two groups of patients with low- or high-risk of progression with a hazard ratio (HR) of 10.9 (p < 0.0001). Validation in multicenter prospective studies and appropriate management of privacy issues in relation to digital electronic health records (EHR) data are presently needed. Nonetheless, we may conclude that the implementation of ML algorithms and RO models into EHR data might help to achieve prognostic information, and has the potential to revolutionize the practice of personalized medicine.