Chemotherapy for Resectable and Advanced Pancreatic Cancer

Carcinoma of the pancreas re mains an imposing threatto health and life, with the majority of patients presenting with advanceddisease and, in most cases, dying from it. Surgical, radiation, and chemotherapytreatments have demonstrated limited efficacy in this disease. Drs. Berlin andRothenberg concisely yet thoroughly review the data on the utility ofchemotherapy at each stage of disease. We would like to expand and emphasize thediscussion in several areas.

Impact of Trial Design on Assessment of Activity

Low frequency of objective remission and difficulty in assessing response inpancreatic cancer has led to the use of survival and clinical benefit responseas primary measures of outcome. Drs. Berlin and Rothenberg correctly callattention to the impact of stage distribution on the survival of patients with"advanced" pancreatic cancer entered into a clinical trial. Asexemplified by the three trials of the gemcitabine/fluorouracil (5-FU)combination, reported differences in survival (medians of 4.4, 7, and 11 months)might be explained by the percentage of enrolled patients with nonmetastaticdisease (0%, 46%, and 61%, respectively). This example serves to remindinvestigators of the need to pay careful attention to disease stage and suggestscaution when making comparisons between studies.

Apparently conflicting study results also contribute to the confusionregarding the utility of adjuvant chemoradiotherapy. Data from theGastrointestinal Tumor Study Group (GITSG) reported a significantly improvedmedian survival of 21 months with adjuvant chemoradiation vs 11 months with bestsupportive care. More recently, the European Organization for Research andTreatment of Cancer (EORTC) found no statistically significant survival benefitwith chemoradiotherapy in their randomized study.

Why did the trials differ in their conclusions? In the EORTC trial, 20% ofpatients randomized to the treatment arm did not receive treatment, which likelyhad an impact on the non-statistically significant but improved mediansurvival of 24.5 months in the treatment arm vs 19 months in the control arm. Inaddition, patients in that trial received only 4 weeks of chemotherapy, whereaspatients in the GITSG trial received 5-FU for 2 years. Perhaps mostimportantly, both studies were underpowered, leading us in the first instance todoubt a positive study, and in the second, to dismiss a clinically meaningfulbenefit due to lack of statistical significance.

We believe adequately powered, well-designed studies with contemporaryregimens administered for an appropriate length of time, if performed, woulddemonstrate a benefit for adjuvant therapy in pancreatic cancer followingcurative resection.

Markers of Response Need to Be Reevaluated

Difficulties with radiographic tumor measurements prompted the development ofa set of criteria for the systematic evaluation of pain, daily analgesic use,performance status, and weight gain, collectively termed clinical benefitresponse. The approval and acceptance of gemcitabine (Gemzar) as therapy foradvanced pancreatic cancer was based principally on improvements in clinicalbenefit response as compared to 5-FU. Clinical benefit response is useful butlimited by its subjectivity, the need for symptoms, awkwardness in itsapplication, and lack of data regarding its reproducibility. An objective markerof clinical benefit that might identify response earlier or more reliably isneeded in this disease.

Several groups have evaluated CA 19-9 as a predictor of disease responseand survival. Heinemann et al prospectively evaluated CA 19-9 levels in 21patients with advanced disease treated with gemcitabine and cisplatin (Platinol).[1]In this study, changes in CA 19-9 levels correlated with radiographic diseaseresponse and with changes in tumor marker levels occurring earlier in the courseof therapy as compared to radiographic changes. Additional reports describe acorrelation between changes in CA 19-9 and survival followingtreatment.[2,3] Thus, CA 19-9 appears to be a reliable, objective measure,rendering it reasonable to include as an assessment of treatment effect infuture studies.

Fluoropyrimidine and Taxanes

The fluoropyrimidine 5-FU was considered standard treatment for pancreaticcancer until 1997, when Burris et al reported the results of a randomized trialcomparing weekly gemcitabine and 5-FU. Gemcitabine was found to be moreeffective in alleviating symptoms (ie, improved clinical benefit response) andprolonging survival than bolus injections of 5-FU. Based on this study, thepalliative benefit of gemcitabine is superior to bolus injections of 5-FU, butthis trial and the strengths of its conclusion can be questioned. Among patientswho consented to the study, 21% were never treated.

The use of clinical benefit response carries the potential pitfalls notedearlier. The objective response rate with both agents was low. The absolutedifference in median survival with gemcitabine (5.65 months) vs 5-FU (4.41months) was quite small and of questionable clinical relevance. Gemcitabine hasnot been compared directly with infusional 5-FU, which is better tolerated andmay be more effective than bolus injections, or to capecitabine (Xeloda), whichmimics infusional 5-FU without the inconvenience of an indwelling venouscatheter and external pump system. Thus, meaningful differences betweengemcitabine and 5-FU, with its versatile administration route and schedule, maybe relatively small.

Data supporting the use of the taxanes in pancreatic cancer, on the otherhand, are disappointing. Recalling that early reports of response rates up to67% with 5-FU were later refuted; response rates of up to 33% with taxanes insmall phase II studies (reported in abstract form only) must be interpreted withcaution. Among three phase II studies reported in detail, overall response ratesranged from 0% to 8%, with median survival ranging from 118 days to 36weeks.[4-6] The systemic activity of these agents appears to be low.

Gemcitabine Combinations

Investigators continue to identify combinations of agents and modalities thatare effective against pancreatic cancer. Although the effect of combinationchemotherapy or combined-modality treatment on advanced disease may be minimal,increased clinical activity identified in that setting might then be transferredto earlier disease, where greater clinical utility may be expected. A higherresponse rate to combination chemotherapy, for example, may translate intobetter control of micrometastatic disease in postoperative adjuvant treatment.

We have concentrated much of our effort on the combination of gemcitabine andcisplatin. Philip et al reported a phase II study of this combination in 42patients, 38 of whom had metastatic disease.[7] The objective response rate of28% and median survival of 7.1 months (95% confidence interval: 6.3 to 9.1months) compared favorably to that reported by Burris et al with single-agentgemcitabine (5.4% response rate, 5.65 months survival). These differences werenoted despite a higher percentage of patients with metastatic disease in thePhilip study. Additional studies with this combination have suggested similarbenefits. Our current trial has added infusional 5-FU to the gemcitabine/cisplatincombination. The relative value of gemcitabine/cisplatin awaits comparison togemcitabine alone or other combinations in phase III trials.

Similarly, the combination of gemcitabine with radiotherapy may offer benefitover either modality alone in patients with advanced disease. In an effort toeffectively address both local and distant disease, we recently completed aphase I study in which 37 patients received up to 42 Gy of radiotherapy to theprimary tumor with weekly gemcitabine at 1 g/m². Despite a reduced dose ofradiotherapy and the elimination of prophylactic nodal radiation, locoregionalcontrol was acceptable, with seven patients experiencing local progression andthree patients, regional progression. Most patients (25) developed distantprogression, emphasizing the need for better systemic disease control. Themedian survival of 11.6 months for the entire population was not affected by thepresence of metastatic disease at study entry.[8] In an attempt to further buildon this regimen, we are currently studying the combination of full-dosegemcitabine and cisplatin with concurrent radiation.

Conclusions

Currently available therapeutic modalities offer modest improvements insymptoms and survival in patients with pancreatic cancer. While certainapproaches have become the "standard of care," we remain dissatisfiedwith the limited options for this group of patients. The introduction of basicand translational research into therapeutic options must accelerate, via furtherdevelopment and support of scientifically based clinical research trials.

References:

1. Heinemann V, Schermuly MM, Stieber P, et al: CA 19-9: A predictor ofresponse in pancreatic cancer treated with gemcitabine and cisplatin. AnticancerRes 19:2433-2436, 1999.

2. Gogas H, Lofts FJ, Evans TRJ, et al: Are serial measurements of CA 19-9useful in predicting response to chemotherapy in patients with inoperableadenocarcinoma of the pancreas? Br J Cancer 77:325-328, 1998.

3. Katz A, Hanlon A, Lanciano R, et al: Prognostic value of CA 19-9 levels inpatients with carcinoma of the pancreas treated with radiotherapy. Int J RadiatOncol Biol Phys 41(2):393-396, 1998.

4. Androulakis N, Kourousis C, Dimopoulos MA, et al: Treatment of pancreaticcancer with docetaxel and granulocyte colony-stimulating factor: A multicenterphase II study. J Clin Oncol 17:1779-1785, 1999.

5. Okada S, Sakata Y, Matsuno S, et al: Phase II study of docetaxel inpatients with metastatic pancreatic cancer: A Japanese cooperative study. Br JCancer 80:438-443, 1999.

6. Whitehead RP, Jacobson J, Brown TD, et al: Phase II trial of paclitaxeland granulocyte colony-stimulating factor in patients with pancreatic carcinoma:A Southwest Oncology Group Study. J Clin Oncol 15:2414-2419, 1997.

7. Philip PA, Zalupski MM, Vaitkevicius VK, et al: Phase II study ofgemcitabine and cisplatin in the treatment of advanced pancreatic cancer. Cancer92(3):569-577, 2001.

8. McGinn CJ, Zalupski MM, Shureiqi I, et al: A phase I trial of radiationdose escalation with concurrent weekly full-dose gemcitabine in patients withadvanced pancreatic cancer. J Clin Oncol. In press.