By Mihai Ciustea, Ph.D.
Pancreatic cancer is one of the most difficult malignancies to treat, and very few drugs have achieved success in large-scale clinical trials. Pancreatic cancer, nevertheless, remains and attractive disease for drug companies to address. Over 43,000 new cases are reported each year in the United States, over 60,000 in Europe, and over 20,000 in Japan. About 85% of these cases are unresectable, locally advanced (stage III) or metastatic (stage IV) at the time of diagnosis, which is one of the reasons mortality is so high and patients typically have a median survival time of only 6 months from time of diagnosis. Since 1997, the powerful chemotherapy gemcitabine has been the first-line standard of care despite low response rates and short improvements to overall survival [OS].
In this note, we build upon a previous note, providing an update on the space and preview upcoming catalysts. Before discussing the ongoing studies, it is important to remember that a range of cytotoxic and targeted chemotherapeutic agents have been added to gemcitabine but failed to produce statistically significant survival benefits in large Phase III clinical trials, even when improvements were seen in smaller Phase II studies.
For example, the oral fluorinated pyrimidine, S-1, combined with gemcitabine was studied in a 454-patient randomized trial and showed a significantly better PFS (5.7 mo vs. 4.1 mo, P < 0.0001), but no significant difference in OS (10.1 mo vs. 8.8 mo, P = 0.15) . For the combinations of gemcitabine with platinum derivatives, encouraging improvements in PFS and OS seen in phase 2 trials did not materialize into statistically significant OS benefits in follow-up phase 3 trials [2-4]. Furthermore, most combination regimens have higher toxicities than gemcitabine alone. Similar conclusions emerged from meta-analyses of the available phase 3 clinical trials: gemcitabine combinations with cytotoxic compounds have no significant OS benefit, even when they delay the onset of disease progression .
Combinations with targeted therapeutics have also been disappointing. In November 2005, the EGFR tyrosine kinase inhibitor erlotinib was approved for use in combination with gemcitabine in pancreatic cancer, having modest but statistically significant OS benefit (6.24 mo vs 5.91 mo, P = 0.038) accompanied by increased toxicity [6, 7]. The analysis was based on a total of 521 patients randomized to gemcitabine plus erlotinib (n = 261) or to gemcitabine plus placebo (n = 260). The primary endpoint was overall survival, defined as the time from randomization to death. The combination of 100 mg of erlotinib and standard doses of gemcitabine was significantly more toxic than gemcitabine alone. Interestingly, only patients with severe (grade 2 or higher) rash caused by erlotinib had increased survival. Despite the modest survival improvement and increased toxicity, the Oncologic Drugs Advisory Committee voted in favor of approval by 10 to 3.
The combination of gemcitabine with cetuximab, an anti-EGFR antibody, showed moderate activity in phase 2, but no survival benefits in the follow-up, 745-patient phase 3 trial.
Combinations of gemcitabine with agents that target tumor angiogenesis also failed to receive FDA approval. The combination with bevacizumab, an anti-angiogenesis antibody that targets VEGF, showed a promising 8.8 mo OS benefit in 52 patients with advanced pancreatic cancer, but the results were not confirmed in a bigger phase 3 trial, where OS was only 5.8 mo. The combination of gemcitabine with sorafenib, a kinase inhibitor with anti-angiogenic properties, showed activity in phase 1 and 2, but also failed in phase 3, with no significant increase in OS (9.2 mo vs 8.5 mo, P = 0.146).
Of the regimens that do not include gemcitabine, the most promising was FOLFIRINOX (oxaliplatin, irinotecan, leucovorin, and 5-fluorouracyl). When compared to gemcitabine in a phase 3 trial, the OS for FOLFIRINOX was 11.2 mo vs. 6.8 mo for gemcitabine (HR: 0.57; 95% CI: 0.45-0.73; P < 0.001). The FOLFIRINOX regimen was more toxic and almost half the patients suffered grade 3 or 4 toxicities. FOLFIRINOX is typically used in patients with good prognostic indicators at diagnosis who physicians feel can handle the increased toxicity of this regimen.
Threshold's TH-302. In March 2012, Threshold (THLD) released Phase II data where impressive progression free survival [PFS], the primary endpoint of the trial, was featured. The stock price responded positively and soared from around $3 to above $8. At the end of September, however, the company presented additional Phase II data at the European Society for Medical Oncology [ESMO], which included an overall survival analysis. Some investors had been hoping for a statistically significant result and the possibility of filing for regulatory approval based on this Phase II study, and when this did not occur the stock sold off to its currently level of approximately ~$4.00.
Going forward, marketing approval of TH-302 will depend on the results of a Phase III randomized trial that is expected to be initiated by Threshold and partner Merck KGaA at the end of 2012. The trial will compare TH-302 in combination with gemcitabine against gemcitabine alone in patients with metastatic or locally advanced unresectable pancreatic cancer. The primary endpoint will be OS and the trial will not contain a crossover option.
Merck KGaA will be responsible for conducting the trial, which is being conducted pursuant to a Special Protocol Assessment [SPA]. According to the SPA, FDA agrees that the design and planned analysis of the study adequately address the objectives necessary to support a regulatory submission. In other words, the FDA agrees that the study is sufficient to support registration, assuming statistically significant OS benefit, and no new safety issues. This is important because the two other drugs, Celgene's Abraxane and Clovis' CO-101, that are currently in Phase III studies for the same indication might have positive trial read-outs in 4Q12. According to the SPA, however, positive outcomes for these agents will not prevent the approval of TH-302, should its Phase III trial be positive. All three trials contain a gemcitabine control arm, so should all the trials be positive, investors should be able to identify which product has the greatest market potential based on the relative best benefit/risk ratio of these agents. Furthermore, there is no reason why TH-302 could not be combined with either CO-101 or abraxane/gemzar, should these chemotherapies gain the indication.
Review of the TH-302 Phase II Data. With respect to the new data presented at ESMO, analysis of the secondary endpoints of the Phase 2b trial indicated an improvement in OS in the TH-302 + gemcitabine arms relative to gemcitabine alone (see Table below). This OS benefit was not statistically significant, but the Phase 2b trial was not designed to have the statistical power for OS analysis. Furthermore, crossover of 26 patients from gemcitabine alone arm to the TH-302 + gemcitabine arms confounds the OS analysis.
We take a closer look at the survival data for TH-302. Patients treated with gemcitabine + TH-302 (340 mg/m2) demonstrated an OS of 9.2 months (N=74; HR=0.955 (95% CI: 0.67-1.37); Log-rank test: P=0.800), while patients treated with gemcitabine + TH-302 (240 mg/m2) had a median OS of 8.7 months (N=71; HR=0.960 (95% CI: 0.67-1.38); Log-rank test: P=0.827). Our view is that this data is about as good as one could reasonably expect in this indication for a trial of this size.
Gemcitabine + TH-302 (240 mg/m2) (N=71)
Gemcitabine + TH-302 (340 mg/m2) (N=74)
Median OS (MO)
(95% CI: 0.67-1.38); P=0.827
(95% CI: 0.67-1.37); P=0.800
If one excludes the 26 crossover patients from the survival analysis, patients treated with gemcitabine alone (N=43) demonstrated OS of 6.3 months (below). We note that this is in line with the median OS typically observed for patients treated with gemcitabine (i.e. ~6 months).
Gemcitabine + TH-302 (240 mg/m2) (N=71)
Gemcitabine + TH-302 (340 mg/m2) (N=74)
(95% CI: 0.50-1.14); P=0.183
(95% CI: 0.51-1.16); P=0.207
The OS of 6.3 months for the group excluding the crossover patients is only slightly less than the 6.9 months observed when they are included. When the crossover patients are excluded, the statistical tests for the TH-302 arms are improved: 340 mg/m2 TH-302 (N=74), HR=0.77 (95% CI: 0.51-1.16); Log-rank test: P=0.207, and 240 mg/m2 TH-302 (N=71), HR=0.756 (95% CI: 0.50-1.14); Log-rank test: P=0.183. These hazard ratios are inline with what one would expect for adding an active agent onto another chemotherapy and a treatment naïve oncology indication.
When you look at the 26 crossover patients who were randomized to receive gemcitabine + 340 mg/m2 TH-302 (N=12) or gemcitabine + 240 mg/m2 TH-302 (N=14), the median OS values are 13.4 months and 2.6 months from the time of crossover, respectively. Although the 340 mg/m2 TH-302 crossover arm shows improved OS with statistical power (HR=0.448, p=0.010), this is a small sample size and results need viewed with caution. Nevertheless, the fact that there appears to be a dose response is encouraging as is the totality of the TH-302 Phase II data set.
Gemcitabine + TH-302 (240 mg/m2)
Gemcitabine + TH-302 (340 mg/m2)
Median OS from time of crossover
(95% CI: 1.9-4.3)
(95% CI: 4.1-15.0)
Threshold and Merck KGaA intend to start the Phase III by year-end 2012, however, there are two ongoing Phase III clinical trials now on investor's radar screens that should read-out by the end of 2012 and help shape investor's opinion of the space.
The first study is the LEAP Study from Clovis Oncology (CLVS). While the LEAP study enrolled a total of 360 patients, only 64% or 230 patients who are confirmed low hENT1 expressing patients will be evaluated in the primary endpoint analysis, which is overall survival. In LEAP study, the gemcitabine-based drug, CO-101, is being compared to gemcitabine. Study details can be found here.
CO-101. CO-101 is a lipid conjugate of gemcitabine designed to enter cells independently of the mechanism though by some responsible for the transport of gemcitabine across cell membrane. Researchers in the field believe that the most important transporter for gemcitabine is the human Equilibrative Nucleoside Transporter 1 (hENT1). Tumor cells in vitro with low hENT1 expression are resistant to gemcitabine therapy and retrospective analyses of clinical studies have shown a correlation between hENT1 tumor expression and overall survival of pancreatic cancer patients treated with gemcitabine. Clovis would argue that the patients with low hENT1 expression have poorer outcomes because gemcitabine fails to enter these cells and exert its anti-proliferative activity.
An alternative viewpoint is that hENT1 is just a prognostic indicator. In other words, sicker patients with more advanced disease have decreased hENT1 expression. As a simple prognostic indicator, like ECOG status, for example, mortality would then correlate to hENT1 expression. In this regard, hENT1 would not have an active role in modulating response to gemcitabine.
The LEAP (Low hENT1 and Adenocarcinoma of the Pancreas) clinical trial is testing the hypothesis of whether or not hENT1 expression plays a role in gemcitabine response as measure by overall survival. The LEAP trial has enrolled 360 metastatic pancreatic cancer patients that were randomized 1:1 to receive either CO-101 (1250 mg/m2, which equates to 625 mg/m2 of gemcitabine) or gemcitabine (1000 mg/m2). The primary endpoint is overall survival in the low hENT1 expressing patients, which, per the company, is 64% of the patients, or 230 patient and results are expected in the fourth quarter of 2012. The trial is 98% powered to show a near doubling of overall survival. We note that this is a relatively high bar in this indication, but success would unequivocally establish the benefit of the agent.
Abraxane. Abraxane is an albumin-bound nanoparticle formulation of paclitaxel. The MPACT trial (NCT00844649) is a randomized, open-label, phase 3 study that evaluates the efficacy of the combination of Abraxane and gemcitabine versus gemcitabine alone in improving overall survival in patients with metastatic adenocarcinoma of the pancreas. With an enrollment of 842 patients, the study has completed accrual and final data collection for primary outcome measure is expected in December 2012. This is a large trial and the size indicates that it is powered to detect even a small increase in survival benefit.
In summary, historically, Phase III trials in pancreatic cancer have a very low probability of success and investors should be very cautious owning these stocks into these binary events. For many drug and drug combinations studied in pancreatic cancer, increases in PFS and OS seen in small early trials did not translate into statistically significant OS benefits in larger, Phase III trials. Some of these past Phase III trials included adding very active and well-established chemotherapies like irinotecan or cisplatin to gemcitabine, and yet these doublets failed versus gemcitabine alone. Since the approval of gemcitabine in 1997, only erlotinib + gemcitabine and FOLFIRINOX have had positive Phase III trials against gemcitabine alone. It appears to us that TH-302 may be the most promising of the agents discussed in this note given the Phase 2b signs of activity. Obviously, the known chemotherapy, Abraxane, is being evaluated in over-powered trial, yet it remains to be seen what the magnitude of any benefit looks like. Our view is that LEAP trial evaluating CO-1010 is a relative long-shot given the speculative nature of hENT1 and gemcitabine response and the high-bar of establishing success in a 230 patients Phase III study.
1. Ozaka, M., Y. Matsumura, H. Ishii, Y. Omuro, T. Itoi, H. Mouri, K. Hanada, Y. Kimura, I. Maetani, Y. Okabe, M. Tani, T. Ikeda, S. Hijioka, R. Watanabe, S. Ohoka, Y. Hirose, M. Suyama, N. Egawa, A. Sofuni, T. Ikari, and T. Nakajima, Randomized phase II study of gemcitabine and S-1 combination versus gemcitabine alone in the treatment of unresectable advanced pancreatic cancer (Japan Clinical Cancer Research Organization PC-01 study). Cancer Chemother Pharmacol., 2012. 69(5): p. 1197-204. Epub 2012 Jan 17.
2. Colucci, G., R. Labianca, F. Di Costanzo, V. Gebbia, G. Carteni, B. Massidda, E. Dapretto, L. Manzione, E. Piazza, M. Sannicolo, M. Ciaparrone, L. Cavanna, F. Giuliani, E. Maiello, A. Testa, P. Pederzoli, M. Falconi, C. Gallo, M. Di Maio, and F. Perrone, Randomized Phase III Trial of Gemcitabine Plus Cisplatin Compared With Single-Agent Gemcitabine As First-Line Treatment of Patients With Advanced Pancreatic Cancer: The GIP-1 Study. Journal of Clinical Oncology, 2010. 28(10): p. 1645-1651.
3. Heinemann, V., D. Quietzsch, F. Gieseler, M. Gonnermann, H. Schonekaes, A. Rost, H. Neuhaus, C. Haag, M. Clemens, B. Heinrich, U. Vehling-Kaiser, M. Fuchs, D. Fleckenstein, W. Gesierich, D. Uthgenannt, H. Einsele, A. Holstege, A. Hinke, A. Schalhorn, and R. Wilkowski, Randomized phase III trial of gemcitabine plus cisplatin compared with gemcitabine alone in advanced pancreatic cancer. Journal of Clinical Oncology, 2006. 24(24): p. 3946-3952.
4. Louvet, C., R. Labianca, P. Hammel, G. Lledo, M.G. Zampino, T. Andre, A. Zaniboni, M. Ducreux, E. Aitini, J. Taieb, R. Faroux, C. Lepere, and A. de Gramont, Gemcitabine in combination with oxaliplatin compared with gemcitabine alone in locally advanced or metastatic pancreatic cancer: Results of a GERCOR and GISCAD phase III trial. Journal of Clinical Oncology, 2005. 23(15): p. 3509-3516.
5. Bria, E., M. Milella, A. Gelibter, F. Cuppone, M.S. Pino, E.M. Ruggeri, P. Carlini, C. Nistico, E. Terzoli, F. Cognetti, and D. Giannarelli, Gemcitabine-based combinations for inoperable pancreatic cancer: Have we made real progress? A meta-analysis of 20 phase 3 trials. Cancer, 2007. 110(3): p. 525-533.
6. Moore, M.J., D. Goldstein, J. Hamm, A. Figer, J.R. Hecht, S. Gallinger, H.J. Au, P. Murawa, D. Walde, R.A. Wolff, D. Campos, R. Lim, K. Ding, G. Clark, T. Voskoglou-Nomikos, M. Ptasynski, and W. Parulekar, Erlotinib plus gemcitabine compared with gemcitabine alone in patients with advanced pancreatic cancer: A phase III trial of the National Cancer Institute of Canada clinical trials group. Journal of Clinical Oncology, 2007. 25(15): p. 1960-1966.
7. Senderowicz, A.M., J.R. Johnson, R. Sridhara, P. Zimmerman, R. Justice, and R. Pazdur, Erlotinib/gemcitabine for first-line treatment of locally advanced or metastatic adenocarcinoma of the pancreas. Oncology (Williston Park). 2007. 21(14): p. 1696-706; discussion 1706-9, 1712, 1715.