Pancreatic carcinoma is the fourth leading cause of cancer death in the United States. Among the approximately 28,000 cases diagnosed each year, fewer than 5% live 5 years or longer. The American Joint Commission on Cancer (AJCC) has divided the progression of pancreatic cancer into four stages, but clinical decision-making is based on three levels of disease: local and resectable, locally advanced and unresectable, and metastatic disease. This article will discuss the role of chemotherapy at each stage of pancreatic cancer.
Although a number of agents have been evaluated as treatment for pancreatic cancer, efficacy has been limited. Agents such as mitomycin(Drug information on mitomycin) (Mutamycin), doxorubicin(Drug information on doxorubicin), and epirubicin(Drug information on epirubicin) (Ellence) have produced some of the highest response rates, ranging from 13% to 27%. A comparison of phase II trials is limited in that most have been small, with varying proportions of patients with locally advanced (stages II and III) and metastatic (stage IV) disease. In addition, the use of various approaches to radiologic assessment and response criteria have resulted in widely disparate response rates, often for the same therapies.
The antimetabolite fluorouracil(Drug information on fluorouracil) (5-FU) was among the earliest chemotherapy agents studied for the treatment of this disease. Until recently, it represented the standard against which other agents or combination chemotherapy regimens were tested. Early trials reported response rates > 20% for single-agent 5-FU. However, trials in the early 1990s reported response rates < 15%. In fact, a recent randomized trial showed no response in 57 patients with measurable disease treated with single-agent 5-FU. This decline in the response rate may have been caused by a number of factors. Changes in radiographic technology, which, in turn, led to an improvement in the ability to accurately measure disease, may be the largest contributing factor to this change.
Coadministration of other agents appears to modulate the intracellular effects or metabolism of 5-FU. Modulation of 5-FU with agents such as interferon-alpha, N-phosphonacetyl-l-aspartate (PALA), and leucovorin does not appear to have improved the efficacy of 5-FU.[1,3,4] Altering the mode of 5-FU administration from bolus to infusion clearly changes the side-effect profile of 5-FU but does not affect efficacy. Phase II trials of infusional 5-FU have produced results similar to those obtained in trials of bolus 5-FU. Infusional 5-FU given to 24 patients in conjunction with interferon-alpha resulted in an 8% response rate and a median survival of only 4.6 months. Among 35 patients treated with infusional 5-FU in conjunction with PALA, the response rate was 14%, and the median survival was 5.1 months.
More recently, oral formulations of 5-FU have been developed that mimic infusional 5-FU pharmacokinetically without the inconvenience of an infusion port. One such agent is capecitabine(Drug information on capecitabine) (Xeloda), a prodrug converted intracellularly to 5-FU. Another agent (UFT) combines a prodrug, tegafur(Drug information on tegafur), with uracil; this compound competes with 5-FU for catabolism by dihydropyrimidine dehydrogenase and thereby reduces 5-FU breakdown. UFT and capecitabine have both been investigated for the treatment of pancreatic cancer and have produced response rates of 0% and 7.3%, respectively, and median survivals of 3.5 and 6 months.[5,6]
Investigators have tried to improve the efficacy of single-agent 5-FU through the use of combination regimens. Table 1 lists a sampling of randomized chemotherapy trials in pancreatic cancer.[2,7-11] Although combinations of 5-FU with doxorubicin (Adriamycin) and mitomycin (FAM), and 5-FU with a platinum, have demonstrated promising results in phase II trials, neither has proved superior to single-agent 5-FU.[7,8]
Other regimens have met similar fates. The 10.2-month median survival reported in a small, early trial of an intensive multiagent schedule entitled the Mallinson regimen could not be duplicated in a larger randomized trial, which showed the strategy to be no more effective than 5-FU.
Finally, although combination regimens have not improved upon the efficacy of single-agent 5-FU, multiagent regimens such as FAM and ELF (etoposide, leucovorin, 5-FU) have produced longer survivals than best supportive care in small randomized trials (Table 1).[2,7-11] Nevertheless, single-agent 5-FU remained the standard of care for patients with advanced pancreatic carcinoma until recently.
Gemcitabine (Gemzar) has become the drug of choice for first-line treatment of pancreatic cancer. A pyrimidine nucleoside analog, the agent is activated to its triphosphate form by a series of enzymatic steps, starting with the actions of a rate-limiting enzyme, deoxycytidine kinase. The antitumor activity of gemcitabine(Drug information on gemcitabine) correlates with DNA incorporation of gemcitabine triphosphate, and may be caused partially by its inhibition of ribonucleotide reductase (a key enzyme in the production of deoxynucleotides), leading to the depletion of normal deoxynucleotide pools.
Phase II TrialsPhase II trials of gemcitabine in pancreatic cancer have resulted in only 11% and 6.3% response rates.[12,13] Despite these low response rates, gemcitabine produced median survivals of 5.6 and 6.3 months, and investigators noted that some patients who did not achieve objective tumor responses did experience relief of tumor-related symptoms. To better study this observation, subsequent trials were designed to assess the effects of gemcitabine on specific disease-related symptoms, using a new measure"clinical benefit response." Clinical benefit response encompasses an evaluation of pain, Karnofsky performance status (KPS), and weight loss and was first reported in a phase II trial in patients with pancreatic cancer refractory to 5-FU. Although only 10.5% of patients with measurable disease had objective partial responses, 27% demonstrated clinical benefit responses.
Phase III TrialA phase III trial comparing single-agent gemcitabine to 5-FU as first-line therapy was also conducted with a primary end point of clinical benefit response. Survival was the secondary end point. Of 160 patients enrolled, 34 were not randomized, mostly due to rapid clinical deterioration. For the remaining 126 patients, clinical benefit response in the gemcitabine arm was significantly better than in the 5-FU arm (23.8% vs 4.8%, respectively, P = .0022). Survival analysis also favored the gemcitabine arm with a median and 1-year survival of 5.65 months and 18%, respectively, compared to 4.41 months and 2%, respectively, for the 5-FU arm.
The treatment IND (investigational new drug application) supported the randomized trial data with a similar median survival of 5.1 months for chemonaive patients treated with gemcitabine. Food and Drug Administration approval of gemcitabine was based on the randomized data. Gemcitabine was replaced by 5-FU as the control arm and current standard in more recent randomized trials.
Significant effort has been focused on improving results with gemcitabine. First, laboratory data suggested that an alternative gemcitabine administration schedule might be advantageous. The rate-limiting step in activating gemcitabine to its triphosphate form is deoxycytidine kinase. At an infusion rate higher than 10 mg/m²/min, deoxycytidine kinase is saturated. Thus, the standard schedule of 1,000 mg/m² over 30 minutes may not allow full conversion and activation of gemcitabine to gemcitabine triphosphate.
After a phase I study established 1,500 mg/m² administered at a fixed-rate infusion over 150 minutes as the recommended phase II dose, that dose was evaluated against 2,200 mg/m² over 30 minutes in a randomized phase II trial in patients with advanced pancreatic cancer.[16,17] As postulated, the 10 mg/m² infusion rate produced significantly higher intracellular levels of gemcitabine triphosphate in peripheral blood mononuclear cells. Patients treated with the longer infusion schedule also showed a trend toward improved survival, but patient numbers were small. These results justify further testing of the fixed-rate infusion in phase III trials.
Gemcitabine-Based CombinationsAnother important development was the evaluation of gemcitabine-based combinations. One logical partner was 5-FU. Laboratory studies demonstrated additive effects against HT-29 cells when gemcitabine was administered either prior to or following 5-FU.[18,19] Examples of combination schedules that have been investigated are listed in Table 2.[20-24] As noted earlier, phase II trials have enrolled varying percentages of patients with locally advanced disease and other characteristics that may have had an effect on survival.
Three trials using identical regimens of gemcitabine, 1,000 mg/m² IV over 30 minutes, followed by a weekly bolus of 5-FU at 600 mg/m² for 3 of 4 weeks, illustrated the potential effects on survival in locally advanced patients.[20-22] The Eastern Cooperative Oncology Group (ECOG) trial only enrolled patients with metastatic disease and found a median survival of 4.4 months. A larger study from the Italian Group for the Study of Digestive Tract Cancer (GISCAD) enrolled 54 patients, 46% of whom had locally advanced disease, and reported a median survival of 7 months. The Gruppo de Tumores Gastrointestinales (GETG) study, with 61% of patients having locally advanced disease, showed an 11-month median survival. However, patients in the latter two studies also received leucovorin, 25 mg/m².[21,22] Because leucovorin and other factors may have contributed to the variability in survival rates, interpretation of these comparisons is limited.
Other institutions have approached the combination strategy using infusional 5-FU schedules.[23,24] One of these trials, in which 34% of patients had locally advanced disease, reported a promising median survival > 10 months. The ECOG has completed accrual of over 320 patients with locally advanced and metastatic pancreatic cancer to a phase III trial comparing gemcitabine plus bolus 5-FU with gemcitabine alone. Survival was the primary end point. Results were presented at the 2001 annual meeting of the American Society of Clinical Oncology (ASCO). Although the gemcitabine plus 5-FU arm had a survival of 6.7 months compared to 5.4 months for gemcitabine alone, this did not achieve significance (P = .09). Both arms of the trial had low response rates and similar toxicity profiles.[24a]
Gemcitabine Plus Cisplatin(Drug information on cisplatin)Another agent demonstrating evidence of synergy in the laboratory when administered with gemcitabine is cisplatin (Platinol). Limited trials of cisplatin have been conducted in pancreatic cancer. While these have yielded mixed responses, early results from combination trials of cisplatin with gemcitabine have reported response rates ranging from 11.5% to 36.4%.[25,26] The preliminary results of a randomized trial of cisplatin plus gemcitabine vs single-agent gemcitabine in 63 evaluable patients reported a higher response rate in the combination arm vs the gemcitabine arm (31% vs 10%, respectively). However, this trial also measured clinical benefit response, which was fairly similar for the combination and gemcitabine arms (38% and 45%, respectively). Other platinum agents, such as carboplatin(Drug information on carboplatin) (Paraplatin) and oxaliplatin(Drug information on oxaliplatin), are currently under investigation.