Current genomic and gene expression analyses provide versatile tools to improve cancer chemotherapy. However, it is still difficult to predict whether each patient responds to a particular regimen or not. To predict chemosensitivity in each patient with colorectal cancer, we developed an evaluation method using the primary tumor-initiating cells (TIC, aka cancer stem cells) xenografted in nude mice subcutaneously (patient-derived spheroid xenografts; PDSX). Simultaneously, we also prepared the conventional patient-derived xenografts (PDX) from the same patients' tumors and compared the dosing results with those of PDSXs. We further compared the chemosensitivities of PDSXs with those of 7 patients who had been given regimens such as FOLFOX and FOLFIRI to treat their metastatic lesions. As per the results, the PDSX method provided much more precise and predictable tumor growth with less variance than conventional PDX, although both retained the epithelial characteristics of the primary tumors. Likewise, drug-dosing tests showed essentially the same results in PDXs and PDSXs, with stronger statistical power in PDSXs. Notably, the cancer chemosensitivity in each patient was precisely reflected in that of the PDSX mice along the clinical course until the resistance emerged at the terminal stage. This "paraclinical" xenograft trials using PDSXs may help selection of chemotherapy regimens efficacious for each patient, and, more importantly, avoiding inefficient ones by which the patient can lose precious time and QOL. Furthermore, the PDSX method may be employed for evaluations of off-label uses of cancer chemotherapeutics and compassionate uses of yet-unapproved new drugs in personalized therapies. Mol Cancer Ther; 17(10); 2187-96. ©2018 AACR.
©2018 American Association for Cancer Research.