Biodistribution and tumor imaging of an anti-CEA single-chain antibody-albumin fusion protein.

Albumin fusion proteins have demonstrated the ability to prolong the in vivo half-life of small therapeutic proteins/peptides in the circulation and thereby potentially increase their therapeutic efficacy. To evaluate if this format can be employed for antibody-based imaging, an anticarcinoembryonic antigen (CEA) single-chain antibody(scFv)-albumin fusion protein was designed, expressed and radiolabeled for biodistribution and imaging studies in athymic mice bearing human colorectal carcinoma LS-174T xenografts. The [125 I]-T84.66 fusion protein demonstrated rapid tumor uptake of 12.3% injected dose per gram (ID/g) at 4 h that reached a plateau of 22.7% ID/g by 18 h. This was a dramatic increase in tumor uptake compared to 4.9% ID/g for the scFv alone. The radiometal [111 In]-labeled version resulted in higher tumor uptake, 37.2% ID/g at 18 h, which persisted at the tumor site with tumor: blood ratios reaching 18:1 and with normal tissues showing limited uptake. Based on these favorable imaging properties, a pilot [64 Cu]-positron emission tomography imaging study was performed with promising results. The anti-CEA T84.66 scFv-albumin fusion protein demonstrates highly specific tumor uptake that is comparable to cognate recombinant antibody fragments. The radiometal-labeled version, which shows lower normal tissue accumulation than these recombinant antibodies, provides a promising and novel platform for antibody-based imaging agents.

Neural stem cells as a novel platform for tumor-specific delivery of therapeutic antibodies.

BACKGROUND: Recombinant monoclonal antibodies have emerged as important tools for cancer therapy. Despite the promise shown by antibody-based therapies, the large molecular size of antibodies limits their ability to efficiently penetrate solid tumors and precludes efficient crossing of the blood-brain-barrier into the central nervous system (CNS). Consequently, poorly vascularized solid tumors and CNS metastases cannot be effectively treated by intravenously-injected antibodies. The inherent tumor-tropic properties of human neural stem cells (NSCs) can potentially be harnessed to overcome these obstacles and significantly improve cancer immunotherapy. Intravenously-delivered NSCs preferentially migrate to primary and metastatic tumor sites within and outside the CNS. Therefore, we hypothesized that NSCs could serve as an ideal cellular delivery platform for targeting antibodies to malignant tumors. METHODS AND FINDINGS: As proof-of-concept, we selected Herceptin (trastuzumab), a monoclonal antibody widely used to treat HER2-overexpressing breast cancer. HER2 overexpression in breast cancer is highly correlated with CNS metastases, which are inaccessible to trastuzumab therapy. Therefore, NSC-mediated delivery of trastuzumab may improve its therapeutic efficacy. Here we report, for the first time, that human NSCs can be genetically modified to secrete anti-HER2 immunoglobulin molecules. These NSC-secreted antibodies assemble properly, possess tumor cell-binding affinity and specificity, and can effectively inhibit the proliferation of HER2-overexpressing breast cancer cells in vitro. We also demonstrate that immunoglobulin-secreting NSCs exhibit preferential tropism to tumor cells in vivo, and can deliver antibodies to human breast cancer xenografts in mice. CONCLUSIONS: Taken together, these results suggest that NSCs modified to secrete HER2-targeting antibodies constitute a promising novel platform for targeted cancer immunotherapy. Specifically, this NSC-mediated antibody delivery system has the potential to significantly improve clinical outcome for patients with HER2-overexpressing breast cancer.