randomized phase II study of cetuximab (Ce) or bevacizumab (B) in combination with gemcitabine (G) and in combination with capecitabine (Ca) and radiation (XRT) as adjuvant therapy (Adj Tx) for patients (pts) with completely- resected pancreatic adenocarcinoma (PC). J. D. Berlin 1 , P Catalano 2 , Y Feng 2 , A. Lowy 3 , A.W. Blackstock 4 , P. A. Philip 5 , R.R. McWilliams 6 , J. Abbruzzese 7 , A.B. Benson, III 8 1) Vanderbilt Ingram Cancer Center, Nashville, TN; 2) Dana Farber Cancer Institute, Boston, MA; 3) University of California San Diego, San Diego, CA; 4) Wake Forest University, Winston-Salem, NC; 5) Wayne State University, Detroit, MI; 6) Mayo Clinic, Rochester, MN 7) MD Anderson Cancer Center,
Transcript
ECOG 2204: An intergroup randomized phase II study of cetuximab (Ce) or bevacizumab
(B) in combination with gemcitabine (G) and in combination with capecitabine (Ca) and
radiation (XRT) as adjuvant therapy (Adj Tx) for patients (pts) with completely-resected
pancreatic adenocarcinoma (PC).
J. D. Berlin1, P Catalano2, Y Feng2, A. Lowy3, A.W. Blackstock4, P. A. Philip5, R.R. McWilliams6, J.
Abbruzzese7, A.B. Benson, III8
1) Vanderbilt Ingram Cancer Center, Nashville, TN; 2) Dana Farber Cancer Institute, Boston, MA; 3) University of California San Diego, San Diego,
CA; 4) Wake Forest University, Winston-Salem, NC; 5) Wayne State University, Detroit, MI; 6) Mayo Clinic, Rochester, MN 7) MD Anderson
Cancer Center, Houston, TX; 8) Northwestern University, Chicago, IL
Background: At the time this study was developed, Ce and B were being investigated as part of tx for advanced PC. Concern existed over safety of adding Ce or B to standard adj tx if either drug proved to be active in advanced PC. Methods: Pts with resected PC, ECOG PS 0-1 and adequate bone marrow and liver function were eligible to be randomized to either Ce (arm 1) or B (arm 2) in combination with a standard regimen of G given before and after capecitabine (625 mg/m2 bid on days of XRT only) + XRT (50.4 Gy). CE (400 mg/m2 day 1, then 250 mg/m2 weekly), or B (5 mg/kg q 2 weeks until end of XRT, then 10 mg/kg q 2 weeks), was given throughout therapy. Primary endpoint was rate of pre-specified toxicities (tox) of concern assessed prior to start of XRT and at end of all therapy. This study had a 2-stage design with interim analysis at 25 pts per arm. A total of 126 pts were required assuming a 5% ineligibility and 85% power to detect a 35% rate of specific toxicites. Disease-free (DFS) and overall survival (OS) were secondary endpoints. Results: 137 pts were enrolled, 129 eligible for analysis, 67 on arm 1 and 62 on arm 2. Median age was 60 and 81% had pancreas head primary on both arms. PS (0/1/2) was 40/60/0% in arm 1 and 48/48/3% for arm 2. 65.7% of pts on arm 1 and 54.8% on arm 2 completed adj tx per protocol. 9% on arm 1 and 12.9% on arm 2 recurred while on therapy and 14.9% on arm 1 and 22.6% on arm 2 stopped treatment due to tox/complications. 1 treatment-related death on arm 2 was due to colon perforation. DFS at 2 years was 16% on arm 1 and 22% on arm 2. OS at two years was 35% on arm 1 and 37% on arm 2. Prior to start of XRT, 9% in arm 1 and 2% in arm 2 experienced prespeciified grade 3/4 tox of concern. At end of all therapy 27% on arm 1 and 23% on arm 2 had pre-specified gr 3/4 tox of concern, mostly hematologic. Conclusions: Both arms were safe and fairly well tolerated. Over 10% of pts recurred during adj tx even without planned disease assessments. There is no evidence to develop either arm further in adj tx of PC. Efforts should focus on developing new agents and defining whether or not XRT has a role in adj tx of PC.
Abstract
Background• At the time this study was designed,
cetuximab and bevacizumab were being tested in metastatic pancreas cancer
• There were no data on safety of either drug in patients with pancreas cancer post-resection for either drug
• Additionally, there was no data for either drug in combination with fluoropyrimidine chemotherapy and radiation to the pancreas/pancreatic bed
Objectives/Endpoints• Primary Endpoint: Toxicity of either of the two
regimens– Specific list of toxicities of concern– Measured at two timepoints
• Secondary Endpoints– Assess Safety Profiles– Collect tissue specimens for further analysis– Evaluate Disease-free and Overall Survival for the
two regimens• Evaluate 2-year Overall Survival
– Correlate changes in serum amphiregulin and/or TGF alpha to survival and DFS
Eligibility• Patients with history of pancreas cancer s/p R0 or R1 resection
– Surgery completed > 4 and ≤ 8 weeks prior– R2 resection ineligible– Prospective collection of margin status including retroperitoneal
resection• Adequate hematologic, hepatic and renal function• Age ≥ 18 years• Ability to Understand and provide Informed Consent• No prior chemotherapy or radiation
– Also no EGFR or VEGF inhibitors in the past• ECOG PS 0-2• No cardiovascular or cerebrovascular disease• No unhealed wounds• No full dose anticoagulation
Study Schema
• R0 or R1 resection allowed; tissue requested• Standardized margin definitions given• Serum for TGF alpha obtained
Gemcitabine 1000 mg/m2
qw x 3 of 4 (1 cycles)
Gemcitabine 1000 mg/m2
qw x 3 of 4 (1 cycles)
Gemcitabine 1000 mg/m2
qw x 3 of 4 (3 cycles)
Gemcitabine 1000 mg/m2
qw x 3 of 4 (3 cycles)
RT x 5040 Gy +Capecitabine
825 mg/m2 bid M→F
RT x 5040 Gy +Capecitabine
825 mg/m2 bid M→F
Cetuximab 400 mg/m2 wk 1, then 250 mg/m2 weekly
Bevacizumab 5 mg/kg q2w Beva 10 mg/kg q2w
RANDOMIZE
Bevacizumab and Cetuximab provided through NCI/CTEP
Statistical Design• Randomized, open-label phase II design
– Note that it was written into the protocol that at least half the accrual had to come from community/affiliate sites to limit bias from high volume institutions
– Stratified only for R0 vs R1 resection– 2 stage design: evaluation after 25 patients per arm
completed first cycle of chemotherapy (timepoint 1)• Primary Endpoint: Toxicity, at two timepoints
– Evaluated after 4 weeks of gemcitabine + either bevacizumab or cetuximab
– Evaluated after all therapy completed– Institutions were required to evaluate and summarize
toxicity at end of each portion of treatment• End of first cycle of chemo prior to chemo/xrt• End of chemo/xrt prior to chemo alone • End of all therapy
Primary Endpoint: ToxicityTimepoint 1
– Grade ¾ toxicity rate based on Table 1 <20% is acceptable
– 120 patients (60 per arm) needed to randomize to have an 85% power to detect an unacceptable toxicity rate of 35% with a one-sided 9% level exact binomial test for time point (1) = end of first cycle of chemotherapy prior to start of chemo/xrt
Primary Endpoint: ToxicityTimepoint 2
– Grade ¾ toxicity rate based on Table 1 <25% is acceptable
– 120 patients (60 per arm) needed to randomize to have an 85% power to detect an unacceptable toxicity rate of 45% with a one-sided 9% level exact binomial test for time point (2) = end of all therapy
Statistics Continued• Interim Analysis:
– With 25 patients per arm, the probabilities of observing at least one toxicity with corresponding true rates of 5% and 1% are, respectively, 72% and 22%
• Survival: 87% power with a maximum one-sided type I error of 8% using an exact binomial test in each arm to reject a 2-year overall survival of .37 in favor of .52 or higher.
• 126 total patients needed assuming 5% ineligibility
Table 1: Toxicities of ConcernAny Grade 5 Toxicity
Grade 4+ toxicities Grade 3+ toxicities Other toxicities
