A fast and efficient model for real-time anti-tumor drug efficacy test in the clinical setting based on patient-derived tumor xenograft

Abstract

Drug response rates in cancer therapies are frequently unsatisfactory (first-line 30-40%, subsequent-line ~10% or less). The mouse model, patient-derived tumor xenograft (PDTX), can help promote drug effective rate to over 80% by precisely electing efficacious agents. However, the low engraftment rates, extended testing cycle, and high cost limit its utilization in the clinical setting. A novel method, FastPDTX, was developed, in which a substitutive engrafting procedure was applied, and the labor-intensive and time-consuming propagation of tumorgrafts was circumvented by pathological analysis. A comprehensive evaluation involving tumor biology and clinical performance was performed for this new model. The clinical evaluation engaged 431 PDTX cases, 1050 FastPDTX cases, 531 therapeutic entities, 6535 regimens, and a broad spectrum of solid tumors. Typical cases were also studied, covering various refractory malignant tumors. FastPDTX kept biological architectures and genetic characteristics of the primary tumor and displayed an equivalent positive predictive value as PDTX for the drug efficacy test. The testing cycle was reduced from 3-6 months to 3 weeks, and the cost was decreased by 85%. More importantly, the obstacle of low engraftment rate was circumvented, resulting in the percentage of appliable tests nearly 5-fold higher (94.8% vs. 16.7%) than PDTX. FastPDTX is particularly suitable in circumstances where standard care is not available and the therapeutic window is short, and for cases where tumors have indolent growth in mice. FastPDTX preserved the advantages of classical PDTX, overcame its major limitations, and showed great potential for real-time clinic drug selection in cancer therapy.