Abstract: Objective To establish and characterize human primary esophageal xenografts by subcutaneous implantation of fresh resected esophageal cancer (EC) specimens in immunodeficient mice with unchanged tumor histological features and validate their application value. Methods Thirty-six fresh resected human esophageal cancer specimens were subcutaneously implanted into SCID mice. After tumor developed, they were serially passaged with nude mice using the same methods. For each case, when the xenograft tumors reached size 500-1000mm³, tumors were excised and devided into four parts: the first part was implanted into nude mice for the growth of next generation; the second part of tumor was viably saved in liquid nitrogen for future model recovery; the third part of tumor was snap frozen for DNA/RNA extraction; and the last part of tumor was fixed in formalin and embedded into paraffin (FFPE) for H&E staining and pathology analysis. Using four cases of human primary EC xenografts passaged in nude mice, chemosensitivity study in vitro and efficacy study in vivo were performed to evaluate the application value of these models. Results Twenty-three in 36 cases by fresh EC specimens implantation were successfully developed into tumor in immunodeficient mice, the implantation success rate was 63.89%. These xenografts were grown up stably and could be continuous passaged to next generation. Histopathology examination of transplanted tumors showed that these xenografts were stable with similar histological features compared with the original donors. Twenty-one in 23 models were successfully established in SCID mice when implanted with recovering cryopreserved xenograft tissues, the recovery success rate was 91.30%. In in vitro chemo-sensitivity assay, the selected four cases were sensitive to Sorafenib, 5-FU, Doxorubicin, Tarceva and Irinotecan; however, there were individual differences in sensitivity to Docetaxel, even more, ESX003 and ESX004 were showed no response to Docetaxel. At dosage of 100 mg/kg, Irinotecan could significantly inhibit the growth of these four human primary EC xenografts in nude mice with relative tumor proliferation rate (T/C) of 22.02%, 39.63%, 21.69% and 39.70%, respectively. Conclusions Human primary EC xenograft models could be established with holding original donor’s histological features. The cryopreserved xenografts tissues could be recovered successfully in SCID mice, indicating that these xenograft models can be used repeatedly. The results in chemosensitivity study in vitro and efficacy study in vivo with four selected EC xenograft models show that the human primary EC xenograft models are valuable and very helpful for anti-cancer drugs screening and translational research of human esophageal cancer.

Key words: Human primary esophageal cancer, Xenograft model, Resected specimen, Immunodeficient mice, Anti-cancer drugs