The mTOR is a validated target in many human cancers, and rapamycin analogs (rapalogs) have shown promising results in various tumor types, including non-Hodgkin's lymphoma, sarcoma, glioblastoma and endometrial cancer. The PI3K/Akt/mTOR pathway is activated in pancreatic cancer by overexpression or activation of EGFR and insulin-like growth factor (IgF1R), by PTEN loss or secondary to k-ras mutation and activation of the Ras/Raf/MEK pathway. The investigation of rapalogs for pancreatic cancer treatment is therefore based on sound rationale.
We observed dose-limiting toxicity from standard doses of temsirolimus in Study A and a lack of antitumor activity in both studies (A and B). The reason for the toxicity noted with standard doses of temsirolimus or everolimus is not clear. Patients with gemcitabine-refractory pancreatic cancer can be frail and are often not the best candidates for aggressive systemic therapy. However, our institutional experience using capecitabine and oxaliplatin in this patient group reveals that patients with gemcitabine-refractory pancreatic cancer and good performance status can tolerate and derive benefit from systemic chemotherapy . Therefore, the high toxicity particularly observed with temsirolimus was surprising, despite the use of an approved dose. The concern was raised that the "toxicity" noted resulted at least in part from disease progression. While this degree of toxicity did not occur in Study B, the study was discontinued after enrolling 15 patients, because of disease progression in all patients.
The weekly scheduling of everolimus was based on the dosing schedule identified in our phase I study. Taberbero et al. suggest that daily dosing of everolimus may result in a more sustained inhibition of S6 kinase than intermittent dosing . However, Wolpin et al, recently reported no clinically relevant anti-tumor effect in a recent phase II trial of daily everolimus for patients with advanced, gemcitabine-refractory pancreatic cancer . Therefore, we do not believe that the schedule of everolimus used in Study B resulted in the absence of antitumor effect.
Our recent studies revealed that mTOR inhibitors such as rapamycin and temsirolimus increase Akt phosphorylation/activation and cyclin D1 expression levels in pancreatic cancer cells . Similar results were reported by Wan et al., who suggested that mTOR inhibitors might eventually promote cellular proliferation and survival if the Akt activation goes unchecked . In support of this possibility, pancreatic cancer cells were observed by Wan et al to become more resistant to the mTOR inhibitors beyond 72-96 hours of continuous treatment in vitro. Li et al, also observed that the phosphorylation of IRS-1 on Ser612, a modification that likely inhibits IRS-PI3K interactions, was abolished by rapamycin and CCI-779. It is therefore conceivable that IRS-Ser612 is a mTOR-regulated site that is part of a negative feedback loop, and that mTOR inhibitors enhance pAkt levels in pancreatic cancer cells by blocking its phosphorylation . Why this feedback loop is more relevant in pancreatic cancer than in other cancer types is unclear at this time. Whether the paradoxical Akt activation secondary to mTOR inhibition led to the increased toxicity or rapid tumor progression in Study A cannot be stated with any degree of certainty as no post-treatment biopsies were performed.
The sensitivity of cancer cell lines to erlotinib depends at least partially on the inhibition of the PI3K/Akt/mTOR pathway. The erlotinib-rapamycin combination was found to have a synergistic cytocidal effect and erlotinib alone could inhibit rapamycin-induced Akt activation in non-small cell lung cancer cell lines . These data supported the investigation of the erlotinib-everolimus combination in Study B. However, in Study B, we observed neither clinically significant antitumor activity nor disease stability. One possible explanation may be that most pancreatic cancers are k-ras mutated, which may induce baseline Akt activation in pancreatic cancer. Indeed our pretreatment biopsies revealed an elevated pAkt/Akt ratio in pancreatic cancer cells pre-treatment. It must be noted however, that both studies did not include post-treatment biopsies; therefore, we cannot definitively rule out an erlotinib effect on Akt activation. One can hypothesize, based on our results, that Akt inhibition as a therapeutic strategy for pancreatic cancer should include the simultaneous inhibition at several points along the PI3K pathway. Ongoing studies are investigating mTOR inhibitors in combination with IgF1-R or direct Akt antagonists.
Standard second-line chemotherapy regimens for advanced pancreatic cancer at the present time include oxaliplatin with fluoropyrimidines or capecitabine . Clinical trials that are based on molecular profiling, or which include one of the above standard regimens should be encouraged.