Clinical Experience With Gemcitabine in Pancreatic Carcinoma
Clinical Experience With Gemcitabine in Pancreatic Carcinoma
Pancreatic adenocarcinoma is the number five cause of cancer mortality in North America, with an estimated 27,600 new cases in 1997. Only 20% of patients present with localized, and thus potentially resectable, disease. Surgical resection is associated with considerable morbidity and mortality and, even when complete, yields a 5-year survival rate of less than 25%.
There appears to be no clear consensus concerning the standard chemotherapy for patients with metastatic or locally advanced disease. Until recently, no systemic therapy has been shown to improve quantity and quality of life. Combination therapy has not provided additional benefit over single agents.[2-4]
Fluorouracil (5-FU) is the most extensively evaluated single agent. Carter et al reviewed small single-institutional studies from the 1960s and 70s. Rates of response ranged from 0% to 60%, reflecting the small numbers of patients studied, patient selection, and variable response criteria.
A collation of results by Moertel, incorporating the findings from his own institution, four other studies from larger institutions, and the previous report of Carter et al, reduced the response rate to 26%. Recent prospective randomized trials using a 5-FU bolus as the control arm have further reduced this rate to less than15%.[7,8] In addition, the optimal dose and administration schedule of 5-FU have not been assessed in a randomized fashion.
Attempts at biochemically modulating 5-FU with folinic acid have yielded response rates of less than10%. Similarly, the combination of 5-FU and folinic acid with interferon-alfa (Intron A, Roferon-A) has been less than impressive and has been associated with significant toxicity.
The list of other single agents studied in pancreatic cancer is extensive; these agents have been well summarized in the literature. The numbers of patients assessed have been small, and response rates have generally been less than 15%. Moreover, no agent has demonstrated any impact on overall survival.[3,4,11]
Unfortunately, the results of combination therapy are no better.[3,4] Most of the regimens tested include 5-FU plus one or two other active agents, such as an anthracycline, mitomycin (Mutamycin), streptozocin (Zanosar), cisplatin (Platinol), and/or a nitrosourea. Despite initial enthusiasm for certain regimens in phase II studies, randomized trials involving a total of more than 50 patients have shown response rates of less than 20% in patients with measurable disease (Table 1).[3,7,8,11-17]
The only exception is the SMF (streptozocin, 5-FU, and mitomycin) arm of one study, which showed a 34% response rate in 56 patients with measurable disease. This was not confirmed in a randomized Cancer and Leukemia Group B (CALBG) study comparing SMF to FAM (5-FU, Adriamycin, and mitomycin) in over 180 patients, which reported a 4% response rate for the SMF regimen.
Randomized studies have found no significant survival advantage from combination therapy; reported median survival has ranged from 2.7 to 10 months (Table 1).[3,4,11] One large study randomized over 150 patients to receive either no treatment or 5-FU/lomustine (CCNU). Patients assigned to the treatment arm showed no significant survival advantage.
The possibility of achieving therapeutic benefit is not enhanced by the poor physiologic status of patients with advanced pancreatic cancer. These patients are generally older (median age, 70 years), of poor performance status, and anorexic, often with malignant cachexia. Concurrent medical problems often include pain, nausea, jaundice, and a thrombotic propensity.[2,11] The combination of these factors increases toxicity, and hence, intolerance of any systemic therapy. Thus, the distinction between disease-related symptoms, disease progression, and side effects of therapy cannot be made easily.
The evaluation of response by standard end points (ie, tumor shrinkage) poses some difficulties in patients with pancreatic cancer.[2,18] The primary tumor is not easily distinguished from the normal pancreas and adjacent organs because of its retroperitoneal position, infiltration of adjacent tissues, and local inflammatory reaction/fibrosis. The standard of imaging available in the 1960s and 70s would have made it difficult to assess primary tumors; this may account for the higher response rates reported in older vs more recent studies. These factors also may explain, in part, the wide variation in response rates documented even with the same regimens.[2,18,19]
In diseases such as pancreatic cancer, in which chemotherapeutic agents have modest activity and do not appear to improve survival, the major goal of treatment is palliation. We therefore need to assess disease-related symptoms, as well as the trade-off between the improvement produced by treatment and the toxicity of that therapy. In the past, however, few trials have attempted to formally assess the palliative effects of treatment.[7,20] End points to assess treatment palliation should include changes in quality of life, performance status,[3,11] and disease-related symptoms.
A recent roundtable discussion focused on the utility of validated tools to describe disease-related symptoms and quality of life and the application of these tools in the evaluation of new therapies in pancreatic cancer patients. The principles derived from this discussion have been incorporated into two recent trials of the new agent gemcitabine (Gemzar). These studies have shown gemcitabine to have modest antineoplastic activity in this disease, but to produce more meaningful palliative benefit in treated patients.[21,22]
Gemcitabine (2´,2´-difluorodeoxycytidine, dFdC [Gemzar]) is a novel nucleoside deoxycytidine (dC) analog with structural and metabolic similarities to cytarabine. Both gemcitabine and cytarabine have substitutions at the 2´-position of the sugar moiety (Figure 1) and require intracellular phosphorylation to become active.
Cellular Metabolism and Mechanism of Action
On entry into the cell, gemcitabine is phosphorylated in a stepwise fashion by the enzyme dC kinase, first to gemcitabine monophosphate (dFdCMP), and then to the diphosphate (dFdCDP) and triphosphate forms (dFdCTP; see Figure 2). This enzyme is the rate-limiting step in the activation of gemcitabine,[24,25] with the triphosphate being the major product.[26,27] Kinetic studies in human cancer cells have demonstrated an enhanced accumulation and a prolonged intracellular elimination half-life of gemcitabine triphosphate relative to cytarabine triphosphate.[28,29]
Deamination is the major form of elimination of gemcitabine, where-by the enzyme cytidine deaminase converts gemcitabine to the biologically inert 2´,2´-difluorodeoxyuridine (dFdU) (Figure 2).[30,31] Deamination of gemcitabine occurs at one-tenth of the rate of the normal substrate dCMP. Deoxycytidine monophosphate deaminase acts on gemcitabine to produce 2´,2´-difluorodeoxyuridine monophosphate (dFdUMP), which can be phosphorylated (Figure 2).
The phosphorylated metabolites of gemcitabine inhibit DNA synthesis by two mechanisms: masked chain termination and substrate depletion. The first process depends on incorporation of gemcitabine into DNA. In vitro studies have confirmed that gemcitabine triphosphate serves as a substrate for DNA polymerase-alpha and -epsilon in a strictly competitive fashion with deoxycytidine triphosphate. Once gemcitabine triphosphate has been incorporated into the phosphodiester backbone, only one additional deoxynucleotide can be inserted before the DNA polymerases cease further elongation. This process is referred to as masked chain termination. Once gemcitabine is in such a position atthe 3´-terminus, its excision by proofreading exonucleases is extremely difficult. Chain termination is followed by DNA fragmentation and subsequent apoptosis.
Depletion of DNA substrate is achieved by inhibition of the enzyme ribonucleotide reductase by gemcitabine diphosphate. This enzyme is a major source of deoxynucleotides for DNA synthesis. In vitro studies in both human and nonhuman cell lines exposed to gemcitabine have demonstrated a decrease in the concentration of deoxynucleotides, which is dependent on dose and length of exposure.[28,30,35]