Fax: (713) 563-2366
Induction chemotherapy selects patients with locally advanced, unresectable pancreatic cancer for optimal benefit from consolidative chemoradiation therapy†
Article first published online: 30 MAY 2007
Copyright © 2007 American Cancer Society
Volume 110, Issue 1, pages 47–55, 1 July 2007
How to Cite
Krishnan, S., Rana, V., Janjan, N. A., Varadhachary, G. R., Abbruzzese, J. L., Das, P., Delclos, M. E., Gould, M. S., Evans, D. B., Wolff, R. A. and Crane, C. H. (2007), Induction chemotherapy selects patients with locally advanced, unresectable pancreatic cancer for optimal benefit from consolidative chemoradiation therapy. Cancer, 110: 47–55. doi: 10.1002/cncr.22735
Presented as a poster discussion at the Annual Meeting of the American Society of Clinical Oncology, Atlanta, Georgia, June 2–6, 2006.
- Issue published online: 15 JUN 2007
- Article first published online: 30 MAY 2007
- Manuscript Accepted: 30 MAR 2007
- Manuscript Revised: 1 MAR 2007
- Manuscript Received: 20 DEC 2006
- pancreatic cancer;
The current study was conducted to determine whether there were differences in outcome for patients with unresectable locally advanced pancreatic cancer (LAPC) who received treatment with chemoradiation therapy (CR) versus induction chemotherapy followed by CR (CCR).
Between December 1993 and July 2005, 323 consecutive patients with LAPC were treated at the authors' institution with radiotherapy and concurrent gemcitabine or fluoropyrimidine chemotherapy. Two hundred forty-seven patients received CR as initial treatment, and 76 patients received a median of 2.5 months of gemcitabine-based induction chemotherapy prior to CR. Most patients received a radiation dose of 30 grays in 10 fractions (85%) concurrently with infusional 5-fluorouracil (41%), gemcitabine (39%), or capecitabine (20%).
The median follow-up was 5.5 months (range, 1–63 months). For all patients, the median overall survival (OS) and progression-free survival (PFS) were 9 months and 5 months, respectively, and the 2-year estimated OS and PFS rates were 9% and 5%, respectively. The median OS and PFS were 8.5 months and 4.2 months, respectively, in the CR group and 11.9 months and 6.4 months, respectively, in the CCR group (both P < .001). The median times to local and distant progression were 6.0 months and 5.6 months, respectively, in the CR group and 8.9 and 9.5 months, respectively, in the CCR group (P = .003 and P = .007, respectively). There was no significant difference in the patterns of failure with the use of induction chemotherapy.
The results from this analysis indicated that, by excluding patients with rapid distant progression, induction chemotherapy may select patients with LAPC for optimal benefit from consolidative CR. The authors believe that this strategy of enriching the population of patients who receive a locoregional treatment modality merits prospective randomized evaluation. Cancer 2007. © 2007 American Cancer Society.
Pancreatic cancer is the third most common gastrointestinal malignancy and the fourth leading cause of cancer death in the U.S.1 Two-thirds of all patients with pancreatic cancer have radiographically detectable metastatic disease at the time of diagnosis. Among patients without metastatic disease, surgery, the only potentially curative treatment, confers a 15% to 25% 5-year overall survival (OS) rate.2 However, <10% of patients are eligible for surgery. Even patients who have undergone surgical resection have up to an 80% risk of both distant and local failure. This highlights the high rate of occult metastases even in seemingly localized disease.3, 4 Patients without evidence of metastatic disease who, nonetheless, are ineligible for surgery comprise a heterogeneous group with disease that is defined as locally advanced pancreatic cancer (LAPC).5
Patients with LAPC frequently are treated with chemoradiotherapy in the U.S. This approach is based on the results from a study that was published approximately 25 years ago in which it was reported that the addition of chemotherapy to radiotherapy nearly doubled the median survival from 5.5 months to 10 months.6 Radiotherapy techniques and fractionation have evolved since then along with chemotherapy regimens. However, there are no universally accepted, standard guidelines for the treatment of patients with LAPC. During the past 5 years, there has been a general shift in the treatment strategy for LAPC at our institution from the fairly routine use of chemoradiotherapy alone (CR) to initial treatment with induction chemotherapy followed by consolidation chemoradiotherapy (CCR). This was based on the recognized metastatic potential of pancreatic cancer and the greater efficacy of newer chemotherapy regimens for metastatic disease. However, the potential benefit of this strategy has not been evaluated formally. Here, we report the results of a systematic analysis of the impact of induction chemotherapy on overall outcomes for patients with LAPC.
MATERIALS AND METHODS
Between December 1993 and July 2005, 370 consecutive patients received chemoradiotherapy for nonmetastatic LAPC, which initially deemed unresectable for cure, at the University of Texas M. D. Anderson Cancer Center. Among those patients, 47 were excluded from the current analysis, because they received protocol therapy with bevacizumab during radiotherapy and were reported previously.7 The remaining patients received traditional radiosensitizers (fluoropyrimidines or gemcitabine) with their radiotherapy. Seventy-six patients received CCR. The remaining 247 patients received CR. Patient characteristics are summarized in Table 1. The Institutional Review Board approved this retrospective analysis.
|Characteristic||No. of patients (%)||P|
|All patients (N = 323)||CR (N = 247)||CCR (N = 76)|
|Men||183 (57)||144 (58)||39 (51)|
|Women||140 (43)||103 (42)||37 (49)||.28|
|White||250 (77)||193 (78)||57 (75)|
|African American||23 (7)||21 (9)||2 (3)|
|Hispanic||38 (12)||27 (10)||11 (14)|
|Others||12 (4)||6 (2)||6 (8)||.26|
|Jaundice||189 (58)||139 (56)||50 (66)||.14|
|Abdominal pain||210 (65)||166 (67)||44 (57)||.14|
|Back pain||97 (30)||81 (33)||16 (21)||.05|
|Change in bowel pattern||104 (32)||91 (36)||13 (17)||<.001|
|Fatigue||129 (40)||93 (37)||36 (47)||.13|
|Anorexia||141 (44)||86 (34)||55 (72)||<.001|
|Nausea/vomiting||58 (22)||40 (22)||18 (24)||.75|
|≤12||78 (25)||60 (25)||18 (24)|
|>12||235 (75)||177 (75)||58 (76)||.77|
|≤10||212 (86)||165 (85)||47 (92)|
|>10||34 (14)||30 (15)||4 (8)||.20|
|CA 19–9, U/mL|
|<100||40 (26)||19 (22)||21 (31)|
|≥100||116 (74)||69 (78)||47 (69)||.19|
|% Weight loss|
|<5||94 (36)||77 (39)||17 (25)|
|≥5||170 (64)||120 (61)||50 (75)||.04|
|≤80||197 (61)||146 (59)||51 (67)|
|>80||126 (39)||101 (41)||25 (33)||.21|
|Head||250 (77)||192 (78)||58 (76)|
|Body||39 (12)||30 (12)||9 (12)|
|Head/body||19 (6)||14 (5)||5 (7)|
|Body/tail||8 (3)||7 (3)||1 (1)|
|Tail||7 (2)||4 (2)||3 (4)|
|Well differentiated||18 (17)||17 (20)||1 (6)|
|Moderately differentiated||40 (39)||31 (36)||9 (53)|
|Poorly differentiated||45 (44)||38 (44)||7 (41)|
|Adenocarcinoma||309 (96)||238 (97)||71 (93)|
|Mucinous||11 (3)||6 (2)||5 (7)|
|Signet ring||1 (<1)||1 (<1)||—|
|Giant cell||1 (<1)||1 (<1)||—|
|Adenosquamous||1 (<1)||1 (<1)||—|
|5-FU||133 (41)||133 (53)|
|Gemcitabine||127 (39)||81 (34)||46 (60)|
|Capecitabine||63 (20)||33 (13)||30 (40)|
|Radiation dose, Gy*|
|30||284 (88)||220 (89)||64 (84)|
|50.4||39 (12)||27 (11)||12 (16)|
Patients were evaluated by a dedicated multidisciplinary team, which included a medical oncologist, a radiation oncologist, and a surgical oncologist. Tumors that extended to the celiac axis or the superior mesenteric artery or tumors that occluded the superior mesenteric venous (SMV)-portal venous confluence were deemed locally advanced and unresectable based on a review of the computed tomography (CT) images. Tumors that involved only the SMV were deemed potentially resectable at our institution and, thus, were excluded from the study.
Treatment characteristics are summarized in Table 1. Patients in the CCR group received a median of 2.5 months of induction chemotherapy prior to chemoradiation. Most patients (N = 276; 85%) received the median radiation dose of 30 grays (Gy) in 10 fractions over 2 weeks (3 Gy per fraction) using a 4-field technique. Thirty-four patients (11%) received a more traditional radiation regimen of 45 Gy in 25 fractions with a 5.4-Gy boost to 50.4 Gy over 5 or 6 weeks (1.8 Gy per fraction). In 2001, we eliminated the routine use of elective lymph node irradiation in patients with LAPC to reduce toxicity. Fields targeted the primary tumor and regional lymph nodes in the majority of patients (N = 224; 69%); and, in the remaining patients, fields targeted the primary tumor only (N = 99; 31%).
During induction chemotherapy, most patients (N = 56; 74%) received a combination of gemcitabine and cisplatin, and 12 patients (15%) received gemcitabine alone. The dose of gemcitabine was 750 mg/m2 (range, 450–1000 mg/m2) weekly for patients who received gemcitabine alone and every 2 weeks for those who received gemcitabine and cisplatin. The dose of cisplatin was 25 to 30 mg/m2 every 2 weeks.
Concurrent chemotherapy with radiotherapy for all patients consisted of 5-fluorouracil (5-FU) (41%), gemcitabine (39%), and capecitabine (20%). Gemcitabine was administered at 350 to 400 mg/m2 infused over 30 minutes once a week for 7 weeks starting 24 to 48 hours after the first radiation dose. 5-FU was given as a continuous infusion of 300 mg/m2 daily. Capecitabine was given at 800 to 900 mg/m2 in divided doses twice daily on days when radiotherapy was administered. Chemotherapy dose adjustments were based on previously published criteria.8
Patients were scheduled for follow-up visits with a medical or radiation oncologist at least every 3 to 4 months for the first 2 years. Abdominopelvic CT scans and chest x-rays were obtained at these visits to monitor diseases status. The use of serum markers (CA 19-9) during follow-up was variable and dependent on the individual physician. Therapy was individualized at the time of progression.
The endpoints of this study were OS, progression-free survival (PFS), local progression, distant progression, and toxicity. PFS, local progression, and distant progression were determined retrospectively using formal interpretations of all follow-up radiographic imaging data reviewed by diagnostic radiologists. Wherever lesions were reported as indeterminate, these were not coded as progression. Evidence of progression at the primary site on radiographic imaging was considered local progression. Evidence of progression at distant sites included the development of radiographically visible peritoneal, liver, lung, and other systemic metastases. Severe toxicity was coded using the criteria reported previously.8
Survival was calculated from the first day of treatment (chemotherapy for CCR patients and chemoradiotherapy for CR patients). PFS was censored at the date of the last follow-up on record if no recurrences were observed and the patient remained alive. All statistics are actuarial and were calculated from the date of initial treatment. The significance of differences in proportions was calculated with a chi-square test, and the differences in means were calculated with a Student t test. Survival probabilities were estimated nonparametrically using the Kaplan-Meier product-limit method.9 Comparisons between groups were performed with a log-rank test.10 Each variable that was identified as statistically significant on univariate analysis was used in the multivariate model. Cox proportional-hazards modeling was used to examine the effect of induction chemotherapy and other factors on the time to progression and survival.11 All tests were 2-sided, and P values ≤.05 were considered statistically significant.
Overall Treatment Outcomes
The median follow-up was 5.5 months (range, 1–63 months) for all patients and 6.7 months (range, 1–63 months) among the 44 surviving patients. For the whole group, the median OS was 9.1 months (range, 1–78 months), and the median PFS was 5.0 months (range, 1–63 months) (Fig. 1). The estimated OS rates at the end of 1 year and 2 years were 28% and 9%, respectively, and the estimated PFS rates at the end of 1 year and 2 years were 13% and 5%, respectively (Fig. 1). At the time of analysis, 279 patients had died, 275 from primary cancer and 4 from other causes. Nine patients (3%; 7 patients from the CR group and 2 patients from the CCR group) were able to undergo margin-negative resections. These patients had a median OS of 29.4 months (range, 5.6–63 months) and a median PFS of 20 months (range, 5.6–63 months). With a median interval of 0.6 months between diagnosis and treatment, the median OS and PFS from the date of diagnosis were 10 months and 5.7 months, respectively.
Table 2 provides the results from a univariate analysis of all factors that were considered prognostic for survival outcomes. The factors that were analyzed included age, sex, weight loss, Karnofsky performance status (KPS), hemoglobin (Hgb), radiation fractionation, concurrent chemotherapy regimen, and use of induction chemotherapy. On univariate analysis, the prognostic factors for improved OS were KPS >80 (P < .001), weight loss <5% in the preceding 3 months (P = .024), Hgb ≥12 g/dL (P = .004), and the use of induction chemotherapy (P < .001). The significant prognostic factors for PFS were the use of induction chemotherapy (P < .001), KPS >80 (P = .004), and Hgb ≥12 g/dL (P = .001). On multivariate analysis (Table 3), the use of induction chemotherapy (P < .001), KPS (P < .001), and Hgb (P = .006) remained the only independent prognostic factors for OS. Similarly, the use of induction chemotherapy (P < .001), KPS (P = .006), and Hgb (P = .003) were independent prognostic factors for PFS on multivariate analysis.
|Outcome||HR (95% CI)||P|
|Induction chemotherapy||0.58 (0.42–0.77)||<.001|
|KPS ≥80||0.62 (0.49–0.80)||<.001|
|<5% weight loss||0.73 (0.55–0.96)||.024|
|Hgb ≥12 g/dL||0.89 (0.82–0.96)||.004|
|Induction chemotherapy||0.62 (0.47–0.82)||<.001|
|KPS >80||0.71 (0.56–0.89)||.004|
|Hgb ≥12 g/dL||0.88 (0.82–0.95)||.001|
|Induction chemotherapy||0.53 (0.38–0.71)||<.001|
|KPS >80||0.61 (0.47–0.78)||<.001|
|Hgb ≥12 g/dL||0.90 (0.83–0.97)||.006|
|Induction chemotherapy||0.60 (0.45–0.78)||<.001|
|KPS >80||0.71 (0.55–0.90)||.006|
|Hgb ≥12 g/dL||0.89 (0.83–0.96)||.003|
|Variable||Treatment group: No. of patients (%)|
|Sites of initial failure|
|CR (N = 247)||52 (21)||72 (29)||52 (21)||71 (29)|
|CCR (N = 76)||19 (25)||27 (36)||7 (9)||23 (30)|
|Sites of any failure|
|CR (N = 247)||117 (47)||132 (53)|
|CCR (N = 76)||31 (41)||41 (54)|
Outcomes Analysis by Treatment Group
The use of induction chemotherapy was an independent prognostic factor for OS and PFS. The estimated OS was 8.5 months in the CR group and 11.9 months in the CCR group (P < .001) (Fig. 2). The median PFS was 4.2 months in the CR group and 6.4 months in the CCR group (P < .001) (Fig. 3). The median time to local progression was 6 months in the CR group and 8.9 in the CCR group (P = .003). The median time to distant progression was 5.6 months in the CR group and 9.5 months in the CCR group (P = .007).
Patterns of Failure
Of the 323 patients in our series, 231 patients developed recurrent disease. The anatomic sites of recurrence (occurring at any time after initial treatment) and the sites of first failure are shown in Table 3. Local progression, with or without distant progression, was observed in 117 patients (47%) in the CR group and in 31 patients (41%) in the CCR group (P = .31). Isolated local progression occurred in 44 patients (18%) in the CR group and in 12 patients (16%) in the CCR group (P = .68). Distant disease as the only component of recurrence was not significantly different between the CR group and the CCR group (24% vs 29%, respectively; P = .38). The most frequent sites of distant failure in both groups were liver, peritoneum, and lung. There was no significant difference in the patterns of failure with the use of induction chemotherapy.
Toxicity During Chemotherapy and/or Chemoradiotherapy
In the CCR group, 15 of 76 patients (20%) developed severe acute toxicity. One patient developed gemcitabine-induced pneumonitis, and the remaining patients were admitted for supportive care for intractable nausea/vomiting and failure to thrive. In the CR group, 22 of 81 patients (27%) who received gemcitabine and 10 of 133 patients (8%) who received 5-FU developed severe acute toxicity (P < .001). Only 1 patient who received capecitabine developed severe toxicity and required total parenteral nutrition for intractable nausea and vomiting. Among the 10 patients who received 5-FU, 8 patients had to be admitted for supportive care, and 2 patients had gastrointestinal bleeding. The patients who had severe toxicity from gemcitabine included 4 patients with gastrointestinal bleeding and 14 patients who were admitted for the management of intractable nausea/vomiting.
To our knowledge, the current study is the first to analyze systematically the association between induction chemotherapy, treatment outcomes, and patterns of failure after chemoradiotherapy for patients with LAPC. We report that induction gemcitabine-based chemotherapy followed by chemoradiotherapy appears to provide more promising clinical outcomes than chemoradiotherapy alone for LAPC. There is no difference in local or distant metastasis rates between the 2 groups. We believe the improvement in survival is achieved by eliminating patients who have micrometastases that progress during the induction phase, thereby enriching the population of patients for treatment with a localized therapy, such as CCR.
Chemoradiotherapy with 5-FU has been used as the standard treatment regimen since early randomized trials by the Gastrointestinal Trials Study Group demonstrated nearly 2 decades ago that the combination of chemotherapy and radiotherapy was more effective than chemotherapy or radiotherapy alone.6, 12 Multiple attempts have been made to improve outcomes further in this patient population.
One strategy has been the use of newer chemotherapeutic regimens with greater systemic activity as radiosensitizers. Although few initial trials had higher toxicities with gemcitabine at doses from 400 to 1000 mg/m2 per week,13–15 the toxicity of gemcitabine can be reduced by using a lower dose of gemcitabine or by reducing the radiation dose and volume.8, 16, 17 Other cytotoxic agents, either alone or in combination, have failed to demonstrate convincing and consistent improvements in clinical outcomes.18–25 Newer targeted agents that target various stages of tumor growth and spread may prove useful in combination with standard chemotherapy and/or radiotherapy.7, 26–28
An alternative strategy that has been employed is radiation dose-intensification.29, 30 However, the proximity of pancreatic cancers to duodenal and other bowel loops limits the ability to safely escalate the dose of radiotherapy. An offshoot of this approach is one in which the escalated dose of radiation is confined to the tumor surrounding major vessels or the retroperitoneal surface in an attempt to render the tumor surgically resectable with negative margins. Early reports suggest that this is technically feasible, but its efficacy remains to be established.31
Another strategy that has been employed is sequential therapy. A French Phase III trial that compared initial CR (intermittent cisplatin and infusional 5-FU) followed by gemcitabine versus gemcitabine alone in patients with LAPC recently was reported.32 Patients who were treated with this strategy of chemoradiotherapy followed by chemotherapy reportedly had inferior survival compared with patients who received chemotherapy alone (1-year overall survival, 24% vs 51%; median survival, 8.4 months vs 14.3 months; P = .014).
For the current study, we analyzed a new strategy of sequentially administering chemotherapy followed by chemoradiotherapy for LAPC. This approach of introducing the systemic treatment (chemotherapy) prior to locoregional treatment (surgery or radiotherapy) has been investigated in other gastrointestinal cancers, such as esophageal, gastric, rectal, and anal cancers, with varying degrees of success.33–37 The high incidence of occult micrometastatic disease at the time of diagnosis and the more gratifying responses observed with newer chemotherapeutic agents compared with older, single-agent fluoropyrimidines makes this approach particularly beneficial for the treatment of LAPC. The efficacy of this strategy does not necessarily rely on the increased systemic potency of the induction chemotherapy regimen, but the induction regimen merely may serve as a therapeutic screening test for the inherent biology of LAPC that initially is deemed to have no radiographically identifiable metastatic disease. Patients who fail this screening test because of systemic progression are most likely to be patients who already had micrometastatic disease and were never optimal candidates for a locoregional treatment such as chemoradiotherapy. This strategy serves to select preferentially for CCR the patients who are most likely to benefit from a locoregional therapy. In this series, the incorporation of induction chemotherapy prior to chemoradiotherapy did not seem to increase toxicity or the ability to administer the planned chemoradiation regimen.
There are other theoretical advantages to the induction approach other than the eradication of micrometastatic disease and the prediction of biologic behavior of tumors. It may be postulated that chemotherapy causes shrinkage of tumors, rendering them more likely to respond to subsequent chemoradiotherapy; that drug delivery is better in untreated, well-vascularized tumors; and that the response to induction chemotherapy may guide the choice of chemotherapy regimen used either with radiotherapy and/or subsequently.
Limitations of our study include those common to retrospective studies that use data abstracted from patient records. These data sources tend to capture incomplete information regarding toxicity, although the incompleteness most likely is distributed equally across both the CR group and the CCR group. Because they were retrospective in nature, the data on tumor markers, tumor grade, and reasons for unresectability were not available for all patients. We acknowledge that 1 source of bias is the inability to identify or include patients who progressed while they were undergoing induction chemotherapy. Our experience with resectable pancreatic cancers has demonstrated that approximately 25% of patients who receive neoadjuvant chemoradiotherapy are found to have disease progression (predominantly liver metastases) at the time of restaging prior to pancreaticoduodenectomy.38, 39 We anticipate that a similar (or lesser) number of patients may have been excluded from analysis in the CCR group because of disease progression. The ability to identify and exclude patients who have rapidly progressive disease prior to chemoradiotherapy improves the median survival duration of the group of patients who undergo chemoradiation. Therefore, this series represents the treatment outcomes for patients who received chemoradiation following this selection strategy rather than that for all patients who received induction chemotherapy. An additional source of bias is the possibility that patients who had more symptomatic (and, thus, possibly more advanced) disease were chosen for CR, whereas patients who had more favorable constitutional and/or tumor-related characteristics were chosen for CCR. We have attempted to account for all known confounders, and there did not seem to be a noticeable difference in their distribution across these 2 groups. However, this does not discount the possibility that unknown confounders could have served as a source of bias. The changes in imaging quality (improvement in the ability to detect small metastases with newer, cross-sectional CT imaging and contrast-enhancement protocols) and radiotherapy techniques and the use of additional supportive care measures in later years (such as the more prevalent use of newer generation antiemetics and colony-stimulating factors) during this period may have introduced unknown confounding factors. Finally, the fact that the majority of induction chemotherapy patients received gemcitabine as the radiosensitizing chemotherapy regimen should not influence the outcomes, because previous studies from our institution have demonstrated that the type of concurrent chemotherapy regimen does not independently change survival.8
The major strength of this study is the large sample size of patients who were treated by a dedicated multidisciplinary team at a single institution. The multidisciplinary approach engenders considerable uniformity in 1) accurate radiographic staging, 2) safe and efficient means of obtaining a tissue diagnosis and relieving biliary obstruction, 3) adherence to a strict definition of resectability, 4) consistent treatment approaches (chemotherapy doses and techniques), and 5) standardized posttreatment follow-up. In addition, the systematic analysis of patterns of failure aids our comprehension of the mechanism of action of induction chemotherapy. To our knowledge, the current report is the only comparison of chemoradiation alone versus induction chemotherapy in patients with LAPC, and it may serve as a means to generate a testable hypothesis that will need to be evaluated prospectively.
An ideal induction chemotherapy regimen should have high efficacy in controlling occult metastatic disease, the ability to shrink tumors before consolidative locoregional therapy, convenient dosing schedule, minimal toxicity, and optimal duration to avoid undue delay of locoregional therapy. One area of concern is that a prolonged course of systemic chemotherapy, especially when gratifying responses in tumor markers are noted, eventually may cause local progression or systemic failure, rendering the locoregional therapy (chemoradiotherapy) less meaningful. To ensure that a possible window of opportunity to integrate chemoradiotherapy into this treatment paradigm is not lost, it may be worth stipulating that CCR should be administered after approximately 2 to 3 months of induction chemotherapy, with the optimal duration guided by assessment of CA 19-9 and radiographic images. Such a strategy of chemoradiation after chemotherapy has been adopted in the postoperative setting in a recently reported Phase III trial (Radiation Therapy Oncology Group 9704) in which patients received adjuvant prechemoradation and postchemoradiation chemotherapy for resected pancreatic adenocarcinoma.40 Integrating chemotherapy and chemoradiotherapy in this manner for patients with LAPC is worthy of prospective evaluation.
In conclusion, the use of induction chemotherapy prior to consolidative chemoradiation in patients with LAPC may select patients who are most likely to benefit from a local treatment modality. This strategy of enriching the population of patients who receive a locoregional treatment modality deserves prospective randomized evaluation.
- 6Therapy of locally unresectable pancreatic carcinoma: a randomized comparison of high dose (6000 rads) radiation alone, moderate dose radiation (4000 rads + 5-fluorouracil), and high dose radiation + 5-fluorouracil: the Gastrointestinal Tumor Study Group. Cancer. 1981; 48: 1705–1710., , , et al.
- 10Regression models and life tables [with discussion]. J R Stat Soc B. 1972; 34: 187–220..
- 11Asymptotically efficient rank invariant procedures (with discussion). J R Stat Soc. 1972; Series A 135: 185., .
- 23A randomized phase III study of radiotherapy alone or with 5-fluorouracil and mitomycin-C in patients with locally advanced adenocarcinoma of the pancreas: Eastern Cooperative Oncology Group study E8282. Int J Radiat Oncol Biol Phys. 2005; 62: 1345–1350., , , et al.
- 32Phase III trial comparing initial chemoradiotherapy (intermittent cisplatin and infusional 5-FU) followed by gemcitabine vs gemcitabine alone in patients with locally advanced non metastatic pancreatic cancer: an FFCD-SFRO study [abstract]. J Clin Oncol 2006. 2006; 24: 4008., , , et al.
- 37Intergroup RTOG 98–11: a phase III randomized study of 5-fluorouracil (5-FU), mitomycin, and radiotherapy versus 5-fluorouracil, cisplatin and radiotherapy in carcinoma of the anal canal [abstract]. J Clin Oncol. 2006; 24: 4009., , , et al.
- 40RTOG 9704 a phase III study of adjuvant pre and post chemoradiation (CRT) 5-FU vs gemcitabine (G) for resected pancreatic adenocarcinoma [abstract]. J Clin Oncol. 2006; 24: 4007., , , et al.