Development of gemcitabine-resistant patient-derived xenograft models of pancreatic ductal adenocarcinoma

Academic Article


  • Aim: Gemcitabine is a frontline agent for locally-advanced and metastatic pancreatic ductal adenocarcinoma (PDAC), but neither gemcitabine alone nor in combination produces durable remissions of this tumor type. We developed three PDAC patient-derived xenograft (PDX) models with gemcitabine resistance (gemR) acquired in vivo , with which to identify mechanisms of resistance relevant to drug exposure in vivo and to evaluate novel therapies. Methods: Mice bearing independently-derived PDXs received 100 mg/kg gemcitabine once or twice weekly. Tumors initially responded, but regrew on treatment and were designated gemR. We used immunohistochemistry to compare expression of proteins previously associated with gemcitabine resistance [ribonucleotide reductase subunit M1 (RRM1), RRM2, human concentrative nucleoside transporter 1 (hCNT1), human equilibrative nucleoside transporter 1 (hENT1), cytidine deaminase (CDA), and deoxycytidine kinase (dCK)] in gemR and respective gemcitabine-na├»ve parental tumors. Results: Parental and gemR tumors did not differ in tumor cell morphology, amount of tumor-associated stroma, or expression of stem cell markers. No consistent pattern of expression of the six gemR marker proteins was observed among the models. Increases in RRM1 and CDA were consistent with in vitro -derived gemR models. However, rather than the expected decreases of hCNT1, hENT1, and dCK, gemR tumors expressed no change in or higher levels of these gemR marker proteins than parental tumors. Conclusion: These models are the first PDAC PDX models with gemcitabine resistance acquired in vivo . The data indicate that mechanisms identified in models with resistance acquired in vitro are unlikely to be the predominant mechanisms when resistance is acquired in vivo . Ongoing work focuses on characterizing unidentified mechanisms of gemR and on identifying agents with anti-tumor efficacy in these gemR models.
  • Digital Object Identifier (doi)

    Author List

  • Miller AL; Garcia PL; Gamblin TL; Vance RB; Yoon KJ
  • Start Page

  • 572
  • End Page

  • 585
  • Volume

  • 3
  • Issue

  • 3