Few malignancies have frustrated the persistent efforts of the oncologist as has pancreatic cancer. In a review of 61 published studies that included over 37,000 patients with pancreatic cancer, Gudjonsson observed an overall survival rate of 3.5% for the 4,100 patients who had undergone resection . More recent data indicate continued poor results, with survival rates ranging from 2% to 5% . The annual incidence and overall number of deaths from pancreatic cancer in developing countries are virtually identical; in 1994, an estimated 81,000 new cases were diagnosed, resulting in 78,000 deaths . The incidence of pancreatic cancer rose 1% per year from 1937 until 1973, and has remained unchanged since then .
The dismal reports in the literature result, in part, from the stage of disease at diagnosis . Computed tomography (CT) and magnetic resonance imaging (MRI) have made it easier to define disease stage and determine the diagnosis. Needle biopsies and laparoscopy guided by CT have resulted in fewer unnecessary laparotomies, while biliary stenting, either endoscopically or, when necessary, transhepatically, can lessen the need for a surgical procedure in patients with advanced tumors. Unfortunately, these imaging and diagnostic advances have not yielded detection of earlier-stage disease.
Approximately half of all patients with pancreatic cancer have metastatic disease at the time of diagnosis [4,5], while most of the rest have locally advanced, unresectable disease [6,7]. In general, the overall curative resection rate is 10% to 15%, with most authors reporting an overall 5-year survival rate of 10% [4,8]. The national cancer data base reported a 10% overall survival rate for patients with pancreatic cancer diagnosed since 1990 .
It is clear from more recent surgical data that the survival rate following a Whipple procedure has improved. Several major centers are reporting perioperative mortality of less than 1% for a procedure that some authors previously suggested abandoning due its high complication and death rates [10-12]. This improvement is due, in part, to increased surgical experience, advanced postoperative intensive care techniques, and better preoperative selection with the preopera tive use of abdominal CT, MRI, and laparoscopy to assess local disease extent and resectability.
At Johns Hopkins, perioperative survival for resected patients has shown a steady improvement. In-hospital mortality for patients treated with resection in the 1970s was 30%, vs 0.9% for patients similarly treated in the 1990s. As reported by Yeo et al, this decrease in operative mortality has led, in part, to increased median survival: 7.5 months for patients treated during the 1970s vs 17.5 months for patient treated since 1990 .
This improvement in median survival also results from the increased use of adjuvant postoperative radiation and chemotherapy. Radiation therapy alone, chemotherapy alone, or the two in combination provides palliative therapy for patients with locally advanced, unresectable disease. Radiation alone has been shown to increase median survival from 3 to 6 months , while concurrent radiation and 5-fluorouracil (5-FU) may further increase median survival to 10 months (Table 1) . Yet, reports of 5-year survival among patients managed with nonsurgical therapies remain anecdotal.
At present, the most active single-agent chemotherapy agents are 5-FU and mitomycin(Drug information on mitomycin) (Mutamycin), with each demonstrating response rates in the range of 10% to 30% [15,16]. Data from a number of groups have shown little survival benefit with various single-agent and combination chemotherapies for locally advanced or metastatic pancreatic cancer [17,18]. Palmer et al, however, using 5-FU, Adriamycin, and mitomycin (FAM), reported a median survival of 33 weeks, vs 15 weeks for no treatment, and they suggested that previous studies were negative because they lacked a no-chemotherapy treatment arm .
Our current arsenal of chemotherapeutic agents modestly prolongs life while providing little possibility for cure. Of the reported 150 ten-year survivors of pancreatic cancer in the world literature, only 12 have been cured with nonsurgical therapies . Despite these poor results, many experts consider radiation delivered concurrently with 5-FU to be standard therapy for patients with locally advanced pancreatic cancer.
There have been limited advances toward improving the overall survival rate after resection with adjuvant therapies. Tepper et al observed a local tumor recurrence rate of 50% in curatively resected patients, suggesting a possible survival advantage to improving local control with radiation therapy . In randomized data from the Gastrointestinal Tumor Study Group (GITSG), postoperative radiation given concurrently with 5-FU not only resulted in a local control advantage over surgery alone but also produced a median survival advantage (Table 2) . Data from this trial at 5 and 10 years continue to show a survival advantage for patients receiving adjuvant therapy; 5- and 10-year survival rates for the surgery-alone arm were 5% and 0%, respectively, vs 19% at both 5 and 10 years for patients receiving adjuvant chemoradiation . These data showing benefit with adjuvant therapy are supported by nonrandomized studies from the University of Pennsylvania and Johns Hopkins [10,23].
Numerous clinical trials have been performed with other innovative radiation treatment techniques aimed at improving local control, including neutron therapy , iodine-125 implantation , and intraoperative radiation therapy [26,27]. Often, an improvement in local control over conventional external-beam therapy has been observed, but this has not translated into an overall survival advantage.
Although the gains with surgery plus adjuvant therapy have been modest, most consider standard therapy for patients with operable pancreatic cancer to be resection followed by postoperative radiation therapy with concurrent 5-FU. Because of the marginal advances made thus far with standard oncologic therapies, a number of researchers have focused on developing other strategies for the treatment of pancreatic cancer. In this article, we will discuss some of the more innovative of these approaches and their potential limitations in the management of pancreatic cancer.
Somatostatin and its Analogs
Somatostatin(Drug information on somatostatin) is a cyclic peptide hormone widely distributed throughout the gastrointestinal system , and can act as a potent inhibitor of tumor cell growth . It has been postulated that somatostatin and its analogs inhibit cell growth by triggering signal transduction pathways that negatively control cell growth or by downregulating the stimuli responsible for tumor growth . Szende et al recently demonstrated that somatostatin may induce tumor regression through a mechanism associated with programmed cell death . Unfortunately, clinical experience has demonstrated no significant benefit of somatostatin monotherapy in the treatment of pancreatic cancer [31,32].
More compelling clinical results were reported by Ebert et al with high-dose octreotide(Drug information on octreotide) (Sandostatin), an analog of somatostatin with an increased duration of action. In a cohort of patients with advanced pancreatic cancer, a low-dose octreotide regimen resulted in a median survival of only 3 months, whereas the high-dose produced a median survival of 6 months .
The presence of estrogen receptors in neoplastic mammary tissue and the subsequent tumor responses observed following antiestrogen treatment have been well documented in the literature . Pancreatic carcinomas also possess estrogen receptors; therefore, it is plausible that a similar approach would result in an effective therapy for pancreatic cancer . Unfortunately, most of the clinical data, including a randomized trial reported by Taylor et al, have not demonstrated an improved median survival with tamoxifen(Drug information on tamoxifen) alone over placebo .
One of the few positive results comes from a case-control study of 80 patients with metastatic pancreatic cancer reported by Wong et al, which demonstrated a modest improvement in median survival for patients receiving tamoxifen therapy (7 vs 3 months) . Rosenberg et al reported a median survival advantage for patients receiving a combined octreotide-tamoxifen regimen (12 months, vs 3 months) for an untreated matched cohort) . Although these data are somewhat encouraging, tamoxifen and octreotide do not appear to have great overall effectiveness in pancreatic cancer.
Luteinizing Hormone-Releasing Hormone Agonist
The use of leuprolide (Lupron), a luteinizing hormone-releasing hormone agonist, alone or in combination with somatostatin, has demonstrated in vitro and in vivo activity in pancreatic cancers in hamsters . In light of these and other supporting data , Zaniboni et al conducted a phase II trial to test the combination of leuprolide and tamoxifen in 15 patients with pancreatic cancer. No objective responses were observed, and median survival was a disappointing 5 months .
In conclusion, the clinical impact of hormonal therapy in pancreatic cancer, either alone or in combination with other agents, appears limited.