Gemcitabine resistance in pancreatic cancer: picking the key players.

The diagnosis of pancreatic cancer confers a grim prognosis that has changed little over the last 30 years. Due to the lack of any efficacious screening or biomarkers for pancreatic cancer, presentation with advanced local disease is typical (classically, painless jaundice) and morbidity aside, 4% of patients are alive 5 years after the time of diagnosis. The nucleoside analogue gemcitabine, which today remains the cornerstone of neoadjuvant and adjuvant chemotherapy in pancreatic cancer, has only a 5.4% partial response rate (1) and imparts a progressionfree survival interval ranging from 0.9 to 4.2 months (2). However, many forms of pancreatic cancer show initial sensitivity to gemcitabine therapy followed by the rapid development of resistance—a feature that essentially characterizes this fatal disease. Although the rapid acquisition of resistance to gemcitabine treatment inevitably translates into poor patient outcomes, the tumor’s initial vulnerability and subsequent resistance strongly suggest either the preexistence of resistant cell subpopulation(s) or the rapid development of resistant cells from the tumor itself or from tumor/stromal alterations. Thus, a better understanding of the origins of gemcitabine resistance is critical to the development of superior combination therapies or the replacement of gemcitabine as the gold standard in pancreatic cancer. Several genetic and/or epigenetic alterations have been associated with gemcitabine resistance. Not surprisingly, these include gene products associated with gemcitabine transport and metabolism. Alterations in the nucleoside transporter-1 (hENT1), an important element in gemcitabine uptake, as well as various gemcitabine metabolism gene products, including deoxycytidine kinase and ribonucleoside reductases M1 and M2, have been shown in gemcitabine resistance (3–6). Aberrant expression of genes associated with cellular survival and apoptosis have been implicated, such as the S100 family member S100A4, the expression of which may increase resistance in part by regulation of the hypoxia-induced proapoptotic gene BNIP3 (7, 8). The phosphatidylinositol 3-kinase/Akt survival pathway has also been implicated in gemcitabine resistance (9–12) along with integrin-linked kinase (13). Increased expression and activation of the non–receptor protein tyrosine kinases focal adhesion kinase (14) and c-Src (15, 16) have both been associated with gemcitabine resistance. Finally, c-Met activation has also been implicated in gemcitabine resistance (17). In this issue of Clinical Cancer Research , Liau and Whang extend their previous work (18) demonstrating that expression of HMGA1, a transcriptional ‘‘enhanceosome,’’ plays a role in sensitivity/resistance of pancreatic tumor cells to gemcitabine both in vitro and in mouse xenograft models (19). Overexpression of HMGA1 has been implicated in a number of human cancers and may correlate with poor prognosis (20, 21). HMGA1 complexes on chromatin regulate transcription of numerous genes downstream of the Ras/extracellular signalregulated kinase signaling pathway. HMGA1 itself is a target gene of c-Myc, c-Jun, and activator protein-1 (22–24). Thus, this protein is associated with numerous functions that could be important in drug resistance. Building upon their previous in vitro study that implicated HMGA1 in the sensitization of the pancreatic cancer cell lines BxPC3 and MiaPcCa2 (18), in this work the authors show that decreased HMGA1 expression by RNA interference techniques increases apoptosis and sensitizes cells to gemcitabine in vitro, and also results in decreased tumor burden upon gemcitabine treatment in subcutaneous xenografts in nude mice. Similar sensitization occurred with a dominant negative Akt construct, suggesting that HMGA1 signals through the prosurvival Akt pathway. These studies once again reaffirm the importance of activation of the Akt pathway in drug resistance. The most important finding in the work by Liau andWhang is the effect of gemcitabine treatment on HMGA1 knockdown cells in vivo. Although less characterized than MiaPaCa2 cells, gemcitabine administration in mice bearing subcutaneous xenografts of BxPC3 cells, in which HMGA1 was down-regulated, led to tumor regression, whereas parental cells continued to grow. As with in vitro studies, increased apoptosis was associated with tumor regression in vivo, although involvement of the Akt pathway was not determined. Although further investigation is needed to validate HMGA1 as therapeutic target, Liau and Whang’s results are exciting because they suggest that it may be feasible to identify ‘‘gemcitabine-sensitizing agents’’—drugs that attenuate inherent and acquired resistance to gemcitabine. As of yet, we have little idea what transcriptional program(s) are regulated by HMGA1 or how HMGA1 overexpression leads to Akt activation. Consequently, studies in these areas will be required to determine whether strategies to suppress HMGA1 activity will translate to clinical application. An important consideration in the study of gemcitabine resistance in pancreatic cancer is the apparent phenotypic and molecular variability among gemcitabine-resistant models isolated to date, as well as the large number of gene products Editorial