Accomplishments in 2007 in the Treatment of Metastatic Pancreatic Cancer

SUMMARY

1. Overview of the Disease

   1. Incidence
   2. Prognosis
2. Current Therapy Standards

   1. Chemotherapy
3. Accomplishments and Lack of Accomplishments During the Year

   1. Therapy
   2. Biomarkers
   3. Basic Science
4. What Needs To Be Done (Application of the Accomplishments)

   1. Controversies and Disagreements

      1. When To Transition to a Phase III Study?
      2. What is the Optimum Design for Pilot Studies?
5. Future Directions

   1. Comments on Research
   2. How To Capture Data From Preclinical Research?
   3. Obstacles to Progress

6. Conclusions

I. OVERVIEW OF THE DISEASE

I-A. Incidence

Pancreatic cancer remains one of the most difficult challenges in oncology. Worldwide, the disease claims approximately 227,000 lives per year, and is the eighth most common cause of cancer death in both sexes combined. Most cases and deaths (61%) occur in developed countries, where incidence and mortality are between 7 and 9 per 100,000 in men and 4.5 and 6 per 100,000 in women, with lower rates in developing countries. The incidence of pancreatic cancer is increasing at a rate of approximately 2% per year in developed countries. Approximately 50% of new cases are diagnosed at stage IVb (metastatic), but the majority of patients diagnosed with pancreatic cancer at an earlier stage will eventually develop metastatic disease.

I-B. Prognosis

Complete surgical resection of primary disease provides the only hope of cure for pancreatic cancer. Unfortunately, less than 20% of patients present with potentially resectable disease. More than 80% are diagnosed with locally advanced, unresectable, or metastatic (ie, advanced) pancreatic cancer, which is relatively unresponsive to conventional therapies. Chemotherapy is the primary therapeutic modality for patients with metastatic disease. The major objectives of treatment are to palliate symptoms, to improve quality of life, to delay disease progression, and to prolong survival. Median survival in patients with advanced disease is less than 6 months. Performance status (PS) is the major prognostic factor in patients with advanced pancreatic cancer; patients with good PS tend to derive the greatest benefit from treatment.

The majority of clinical trials in pancreatic cancer combined patients with metastatic and locally advanced disease instead of studying them separately. Thus, it is often difficult to extract and analyze data specifically on metastatic disease. Since current treatment modalities have largely failed to improve survival of patients with pancreatic cancer, interest has grown in identifying better diagnostic markers. Efforts are currently under way to identify biologic factors and markers that might help predict therapeutic efficacy in individual patients, but such efforts are often limited by the quality of specimens obtained at diagnosis.

II. CURRENT THERAPY STANDARDS

II-A. Chemotherapy

Symptomatic treatment (eg, ensuring adequate biliary drainage and pain management) is required in nearly all cases. Patients with an ECOG (Eastern Cooperative Oncology Group) PS of 0–2 should benefit from chemotherapy, which has proved superior to best supportive care alone in terms of survival and quality of life. Standard chemotherapy remains single-agent gemcitabine 1,000 mg/m² in a 30-minute infusion administered weekly.

Results of a recent phase III study in advanced pancreatic cancer showed that patients who were treated with erlotinib 100 mg/day in addition to standard gemcitabine had a modest but statistically significant improvement in overall survival vs. patients treated with gemcitabine alone.¹ Based on the results of this trial, erlotinib in combination with gemcitabine was approved in the United States for the treatment of patients with locally advanced, unresectable, and metastatic pancreatic carcinoma, and in Europe for metastatic pancreatic carcinoma only. The clinical use of this combination, however, is limited, particularly in Europe, due to toxicity, cost, and the relatively small benefit. Nevertheless, both single-agent gemcitabine and the gemcitabine/erlotinib combination are considered valuable standard therapies.

Results of several phase III trials and/or meta-analysis indicated that gemcitabine in a fixed-dose rate infusion,² gemcitabine plus cisplatin or oxaliplatin,³ and gemcitabine plus capecitabine³ are all acceptable therapeutic options, particularly in patients with a good performance status.

III. ACCOMPLISHMENTS AND LACK OF ACCOMPLISHMENTS DURING THE YEAR

III-A. Therapy

Unfortunately, two large phase III trials reported at the 2007 annual meeting of the American Society of Clinical Oncology (ASCO), as well as final data from a Swiss phase III study, did not yield positive results.

• Cancer and Leukemia Group B (CALGB) 80303 randomized 602 patients (85% with metastatic disease) to receive either single-agent gemcitabine or gemcitabine combined with bevacizumab. ⁴ Survival was not statistically different (median 6.0 and 5.7 months, respectively) between the two arms. Thus, it was concluded that adding bevacizumab to gemcitabine does not improve survival in advanced pancreatic cancer.

• In the Southwest Oncology Group (SWOG) S0205 study, 766 patients (78.5% with metastatic disease) were randomized to receive either single-agent gemcitabine or gemcitabine combined with cetuximab.⁵ Again, no statistically significant difference in survival was observed (median 6.0 and 6.5 months, respectively). The study failed to demonstrate a significant advantage in adding cetuximab to gemcitabine in advanced pancreatic cancer.

• In addition to the studies reported at ASCO,^4,5 the final results of a Swiss phase III trial were published in 2007 confirming previously reported data, where gemcitabine/capecitabine did not significantly improve overall survival compared with single-agent gemcitabine.⁶

Several studies reported positive results in 2007, and these data have led to extended study and suggest potentially promising avenues for additional research.

• Results of a randomized phase II trial comparing single-agent gemcitabine to the non-gemcitabine–containing regimen FOLFOXIRI (modulated 5-fluorouracil [5-FU]/oxaliplatin/irinotecan) were reported at ASCO.⁷ Based on the encouraging results of the FOLFOXIRI combination in terms of response rate (> 30%), this study was extended into a phase III trial, with overall survival as the primary end point.

• Several meta-analyses published during 2007 indicated that gemcitabine-based doublets are more active than single-agent gemcitabine, despite a lack of statistical significance in the individual trials.^3,8,9 This seems to be particularly true for patients with good performance status, and for gemcitabine/platinum and gemcitabine/fluoropyrimidine combinations.

• Second-line therapy after gemcitabine failure in metastatic pancreatic cancer is feasible, but probably in less than half of patients.¹⁰ FOLFOX4 (5-FU/leucovorin/oxaliplatin) in this setting achieved interesting results, with a 57% disease control rate in a small phase II study.¹¹

III-B. Biomarkers

One of the major challenges in the treatment of metastatic pancreatic cancer is to define subpopulations of patients that will benefit more from new therapies, based on biomarkers. Moore et al reported a possible relationship of K-ras mutations and epidermal growth factor receptor (EGFR) gene copy number to outcomes in patients treated with erlotinib.¹² However, much additional data in this promising field of study are clearly needed.

III-C. Basic Science

The other major challenge is to increase our knowledge in basic science about pancreatic cancer. Data are still limited and the advances that have been made have yet to translate into significant clinical benefit. Researchers have now identified pancreatic cancer stem cells that may not only explain biology and response to therapy but also represent a target of therapy in the future. This work is still too early to be applied to clinical drug development in pancreatic cancer.¹³

IV. WHAT NEEDS TO BE DONE (APPLICATION OF THE ACCOMPLISHMENTS)

More than a dozen phase III clinical trials were performed over the past decade with the hope of improving treatment outcomes in patients with advanced pancreatic cancer. Most such studies assessed gemcitabine-based doublets. The lack of benefit seen with gemcitabine-based combinations, with the exception of erlotinib and possibly capecitabine, raises the important question of how to choose and test new therapies to minimize failures in phase III settings. What is the optimum cytotoxic platform?

At this time, researchers continue to use gemcitabine as the cytotoxic agent of choice for developing multidrug regimens, including targeted agents. However, the efficacy of gemcitabine is at best modest in patients with advanced pancreatic cancer, though it does have a very good safety profile. Other cytotoxic agents are currently being studied, including ixabepilone, which was recently approved by the US Food and Drug Administration (FDA) for advanced breast cancer, S-1, capecitabine, nab-paclitaxel, and glucofosfamide. Some of these have demonstrated single-agent activity that may be comparable to that associated with gemcitabine. Whether combinations using any of these agents would provide a better platform for targeted agents remains to be seen. For example, an ongoing multi-institutional trial sponsored by Bristol-Myers Squibb in the United States is studying the combination of cetuximab and ixabepilone in patients with advanced pancreatic cancer who are not previously treated. Table 1 lists randomized phase III trials of new systemic therapies in advanced pancreatic cancer.

Table 1.

Randomized phase III trials of new systemic therapies in advanced pancreatic cancer.

Organization/Sponsor	Experimental Arm	Control Arm	Target(s)	Availability of Pilot Data on Activity of Agent*
Untreated Patients
Pfizer	Gemcitabine + axitinib	Gemcitabine	VEGFR	Yes
sanofi-aventis	Gemcitabine + aflibercept	Gemcitabine	VEGF	None
Marseille Institute of Cancer	Gemcitabine + sorafenib	Gemcitabine	VEGFR, Raf	Yes
Royal Liverpool University Hospital	Gemcitabine, capecitabine, GV101	Gemcitabine, capecitabine	Vaccine	None
Taiho	Gemcitabine + S-1	Gemcitabine	Fluoropyrimidine	Yes
Abraxis	Gemcitabine + Abraxane	Gemcitabine	Tubulin	Yes
Tel-Aviv Sourasky Medical Center	Gemcitabine, curcumin, celecoxib	Gemcitabine	COX-2 and multiple others	None
Gemcitabine-Refractory Patients
sanofi-aventis	Loretaxel	5-FU	Tubulin	None
sanofi-aventis	S-1	5-FU	Fluoropyrimidine	Yes

^*Availability of pilot data does not necessarily mean positive results with respect to efficacy.

Another question relates to testing new drugs without combining them with gemcitabine or any other cytotoxic drug. At this time, the only application of this concept is in the setting of gemcitabine-refractory disease. The advantage of this approach is the ability to detect single-agent activity of a new drug. In the future, it is likely that combinations of targeted agents alone will be tested in the frontline setting in gemcitabine-naïve patients.

IV-A. Controversies and Disagreements

IV-A. 1. When To Transition to a Phase III Study?

Interestingly, the phase III study of gemcitabine with erlotinib was not preceded by a phase II trial. This approach is considered inappropriate at this time because of the risk of wasting resources on a large study that may be unnecessary and exposing a lot of patients to potentially toxic regimens. A major concern is that reliable assessment of benefit in a single phase II study is still inadequate because of selection biases. This is largely due to the heterogeneity of patient populations but is also inherent in flawed trial designs that may lead to a phase III study. With targeted agents, this is compounded by our lack of knowledge on patient selection with respect to the target. The magnitude of the desired difference in pilot studies (vs. a control arm or historical series) might be set relatively high. However, the difference should also take into account whether it is realistic, based on the biology of this disease and the expected benefit of therapy, and ensure a magnitude that is also clinically acceptable and worthwhile.

IV-A. 2. What is the Optimum Design for Pilot Studies?

A summary of pilot studies of drugs under evaluation in patients with advanced pancreatic cancer is presented in Table 2. Pressure is increasing to adopt randomized pilot study designs to account for the heterogeneity and patient selection factors inherent in such trials. Gemcitabine or the combination of gemcitabine and erlotinib may be used as the control (standard) arm. However, it is not uniformly accepted that inclusion of erlotinib is necessary in this setting, especially in light of the modest benefit it adds to gemcitabine. It is worth noting that randomization using a control arm may still be associated with a significant risk for false positive study outcomes. There are several potential designs for randomized studies that include the more popular screening design and a design that may not include a control arm but has two or more regimens tested in what is called a selection design to pick the “winner.”

Table 2.

Pilot studies of drugs being tested in patients with advanced pancreatic cancer (combinations incorporating radiotherapy are excluded).

Drug	Major Known Mechanism(s) of Action/Target(s)	Line of Therapy
PTK787/ZK222584	VEGFR	Plus gemcitabine
AZD0530	Abl/src	Plus gemcitabine
Imatinib mesylate	PDGFR, c-Kit	Plus gemcitabine
Enzastaurin	Protein kinase C-beta	Plus gemcitabine*
Sirolimus	mTOR inhibitor	Unspecified
RAD001	mTOR	Gemcitabine refractory
Everolimus	mTOR	Plus gemcitabine
CS1008	Death receptor 5	Plus gemcitabine
RP101	STAT3, APEX	Plus gemcitabine
RTA 402	NFkB	Plus gemcitabine
Cetuximab	EGFR	Plus ixabepilone in gemcitabine-naïve patients
AZD6244	MEK	Gemcitabine refractory*
WX-671	uPA	Plus gemcitabine*
Dasatinib	Src	Gemcitabine refractory Plus gemcitabine*
Lapatinib	EGFR/HER-2	Plus gemcitabine
Actimid	Immune modulation	Plus gemcitabine
AVN944	Inosine monophosphate dehydrogenase	Plus gemcitabine
Nardroparin	Low molecular weight heparin	Not specified
Dalteparin	Low molecular weight heparin	Plus gemcitabine*
Tanespimycin	HSP90	Plus gemcitabine
Curcumin	Multiple	Plus gemcitabine
Ipilimumab	CTLA-4	Unspecified
MORAb-009	GP-9 cell surface antigen	Plus gemcitabine
MKC-1	Cell cycle	Plus gemcitabine
PHY906	Herbal	Plus capecitabine
E7389	Cytotoxic	Gemcitabine refractory
Romidepsin	HDAC	Plus gemcitabine
PX-12	Thioredoxin-1	Gemcitabine refractory
ARQ 197	c-Met	Plus gemcitabine
Rubitecan	Cytotoxic	Plus gemcitabine*
Mitomycin C	Cytotoxic	In patients with BRCA2 mutations
Etoposide	Cytotoxic	Plus gemcitabine
32P BioSilicon	Cytotoxic	Plus gemcitabine
TPI 287	Tubulin	Gemcitabine refractory
Ixabepilone	Tubulin	Gemcitabine naive

^*Randomized study

One aspect of clinical trials in pancreatic cancer is the rapid decline of patients’ general health and performance status. This has resulted in the inclusion of patients in clinical trials who have been erroneously labeled as having a favorable performance status only to progress and die within weeks of study entry. Unfortunately, some patients do have end-stage pancreatic cancer and their inclusion in clinical trials is inappropriate. Stricter inclusion guidelines are therefore necessary to limit accrual as much as possible to patients who do have a more favorable performance status. Investigators have also recognized the need to separate the study of locally advanced unresectable pancreas cancer from that of metastatic disease.

V. FUTURE DIRECTIONS

V-A. Comments on Research

Most new drug development at this time involves targeted agents. Several of those drugs have gone through clinical trials without producing any important sign of activity. These studies were in populations that were not selected on the basis of the target of interest. Recent examples include EGFR and vascular endothelial growth factor (VEGF) targeting monoclonal antibodies. The two major considerations for a successful application of new drugs are the identification of patients who will benefit (or not benefit) from a treatment and addressing the multigenic molecular alterations in pancreatic cancer. The latter dictates the need to test multitargeted approaches to reduce treatment failures resulting from the redundancies and cross talks between signaling pathways.

A major ongoing issue is to justify and prioritize research in EGFR and VEGF/VEGFR-related pathways, given the outcomes of phase III trials with these approaches. Future possibilities include the use of alternative strategies to block these pathways and the inclusion of drugs already tested in combination studies of multitargeted agents (Table 3). An example is the upcoming Southwest Oncology Group trial on combining the anti-EGFR tyrosine kinase inhibitor erlotinib and gemcitabine with a monoclonal antibody targeting insulin-like growth factor 1 receptor (IGF1R).

Table 3.

Pilot studies of combination targeted therapies in advanced pancreatic cancer.

Drugs	Major Known Mechanism(s) of Action/Target(s)	Organization/Sponsor
Gemcitabine/bevacizumab with either cetuximab or erlotinib*	VEGF, EGFR	University of Chicago Phase II Consortium
Gemcitabine, anti-IGF-1R antibody, and erlotinib*	EGFR, IGF-1R	Southwest Oncology Group
Gemcitabine, genistein, and erlotinib	EGFR, NFkB, AKT	Karmanos Cancer Institute/M. D. Anderson Cancer Center
Bevacizumab, erlotinib	VEGF, EGFR	University of California at San Francisco
Erlotinib, gemcitabine, ± panitumumab*	EGFR axis	North Central Cancer Treatment Group

^*Randomized phase II study

V-B. How To Capture Data From Preclinical Research?

This is an evolving area in pancreatic cancer where there has been relatively little funding, and even interest, compared with that in other cancer types, such as breast and prostate cancers. Cell-based preclinical in vitro models that involve in vitro drug treatments may help us understand better the mechanisms of action of drugs and combinations. However, their ability to predict clinical success, or lack of it, is very low. In addition, in vitro systems do not account for the influence of stromal and stem cells on treatment outcome. Animal tumor models may be more appropriate, but they also suffer from a limited ability to predict clinical outcomes and a narrow range of molecular abnormalities that mirror human disease. Despite these limitations, every effort should be made to use models for rational new drug development and to minimize empiricism. Genetically engineered models have attracted lots of interest in this respect and continue to be developed. Primary tumor models are relatively early in development at this time and require validation. In addition to hypothesis-driven testing of new therapies, preclinical models offer the potential for identifying and clinically validating biomarkers that may predict sensitivity or resistance to drugs.

V-C. Obstacles to Progress

A major challenge in developing new agents and combinations is the presence, or lack, of a scientific basis for a research hypothesis. Empiricism still dominates the choice of drugs and is largely justified by the paucity of available preclinical models to help make decisions on how to combine drugs. Lack of good models also limits the availability of biomarkers that may predict treatment outcome when a certain pathway(s) is targeted and, hence, hampers patient selection. Other challenges include the potential interactions between constituent drugs at the pharmacokinetic and pharmacodynamic levels that may either antagonize another’s activity or increase toxicity. Another hurdle is the increasing cost of regimens that use relatively expensive drugs and the need to balance both cost and toxicity against the expected benefit of such a regimen. Finally, logistics and political challenges must also be considered when combining agents from multiple sponsors.

VI. CONCLUSIONS

Clinical and translational research to develop more effective treatment strategies in pancreatic cancer has been disappointing. Investigators recognize the limitations of research questions and study designs and have identified multiple pitfalls. A new generation of trials appears to be focusing more on performing adequately controlled pilot studies. Recognition of the requirement of a more stringent phase II to phase III transition will result in fewer phase III trials in the coming years. However, the availability of an increasing number of targeted agents and improvements in our understanding of the molecular biology of pancreatic cancer will certainly create very exciting opportunities for those interested in research into this disease. Emphasis in the future should be placed on continued development and use of preclinical models, adoption of multitargeted therapy approaches based on rational design, and improvements in patient selection based on biomarkers.