CHICAGO—A search for the optimal first-line antibody combination to treat untreated metastatic colorectal cancer has ended in a tie, showing that patients had an almost equal overall survival of approximately 29 months for treatment with bevacizumab plus chemotherapy or cetuximab plus chemotherapy.
As reported in the plenary session at the American Society of Clinical Oncology Annual Meeting here, the CALGB/SWOG 80405 study (Abstract LBA3) was technically a negative study in that it did not show a meaningful superiority of one regimen over the other. But, by showing equivalence, it reassured oncologists and their patients that either regimen is effective for untreated metastatic KRAS wild-type adenocarcinoma of the colon or rectum, so that the selection can be based on safety profile, cost, and drug availability.
In the trial, each antibody was paired with FOLFOX and with FOLFIRI chemotherapy.
Experts here hailed the overall survival for each regimen of 29 months as a new benchmark for this patient population, noting that it was accomplished across a broad clinical trials network and that the results would apply in community as well as academic settings.
“Our findings clearly show that the two antibodies—with either FOLFOX or FOLFIRI—are both acceptable, and similarly effective,” said Alan P. Venook, MD, Professor of Medical Oncology and Translational Research at the University of California, San Francisco. “This should reassure doctors and patients facing decisions about treatment selection.”
He said about 75 percent of patients with metastatic colorectal cancer in the United States initially receive bevacizumab-based therapy, while the cetuximab-based therapy is more commonly used in Europe.
In the study, 1,137 patients with untreated metastatic colorectal cancer were randomly assigned to receive bevacizumab plus chemotherapy or cetuximab plus chemotherapy. Selection of chemotherapy was based on physician preference (26.6 percent received FOLFIRI, 73.4 percent had FOLFOX). Median follow-up was 24 months.
There were no significant differences in overall or progression-free survival between the treatment groups. In the bevacizumab-plus-chemotherapy group, overall and progression-free survival times were 29.0 and 10.8 months, respectively, versus 29.9 and 10.4 months, respectively, in the cetuximab-plus-chemotherapy group.
Toxicity was as expected, with grade 3 toxicities in about half of the patients. But there were 10 toxic deaths (2.4%) on therapy.
Predictably rash was associated with cetuximab, occurring in seven percent of patients, but none in the bevacizumab arm. Diarrhea occurred in about 10 percent in each arm of the study.
‘Don't Top Line the Negative’
Speaking at a news conference about the four plenary studies, ASCO 2013-2014 President Clifford A. Hudis, MD, said that “on the one hand there may be some temptation to “top line” this study and call it a negative trial. But more important, he said, is to consider that the outcomes “set an entirely new standard, a high bar, for clinical trials in colorectal cancer—An overall survival of 29 months has never been reported before in this patient population.”
The findings provide reassurance that two widely used regimens “offer good and equivalent survival,” he said. “If you are going to use antibody therapy, you have a choice.” However, he added, for patients who have mutant-KRAS disease there is no choice, that bevacizumab is the drug to use and cetuximab is not an option.
Quality of Life Counts
“When two arms of a study show equal survival, quality of life takes on a new meaning,” Venook said, suggesting that patients might prefer one regimen or the other based on the expected toxicities.
Interestingly, a quality-of-life analysis embedded in the trial, using EORTC Global Quality of Life scores, showed low satisfaction scores from patients in the cetuximab arm, but high quality-of-life scores from the same patients.
Venook said this probably occurred because patients were aware that “the worse the skin rash, the more they are likely to respond” to the drug.
He also noted that fewer than one-third of patients came off study due to progressive disease. “This tells us that patients and physicians were not waiting for disease progression to move on to another therapy,” This may reflect current trends in practice, he said.
Data on response rate, dose intensity, second and later treatments “are still percolating,” Venook said, but he estimated that 88 percent of patients did receive subsequent therapy.
The research was supported in part by the National Cancer Institute, Imclone, Roche, Genentech, Bristol Meyer Squibb, and Eli Lilly.
Statistically Significant Patience
The study took 10 years to complete and Venook said he is often asked why. He explained that only after the study began did researchers learn that KRAS mutations predict for resistance to EGFR antibodies, and also that combination biologics may be harmful as seen in the PACCE and CAIRO-2 studies.
Venook made special mention of the efforts of the trial's chief statistician, Donna Niedzwiecki, PhD, Associate Professor of Biostatistics and Bioinformatics at Duke University School of Medicine: “These results were from the 11th interim analysis, after 14 amendments, and 10 years of e-mails from me to Donna,” he said.
From Three Study Arms to Two
Between November 2005 and March 2012, a total of 2,334 KRAS wild-type patients received FOLFIRI or a modified version of FOLFOX6 and were then randomly assigned to either one dose of cetuximab at 400 mg/m2 followed by 250 mg/m2 weekly, or to receive bevacizumab at 5 mg/kg twice weekly.
The original study protocol included unselected patients receiving FOLFIRI or modified FOLFOX6 and randomized to receive cetuximab, bevacizumab, or both. But after 1,420 patients were accrued, the study was amended, and only patients with KRAS wild type tumors (codon 12 and 13) were included, and the combination cetuximab-bevacizumab arm was deleted.
Treatment continued until disease progression, death, unacceptable toxicity, or curative surgery. Treatment holidays of four weeks were permitted, and subsequent treatment was not mandated.
For the 1,137 patients (333 pre-amended eligible retrospective KRAS test, and 804 post-amended), the median follow-up was 24 months. The patients' median age was a relatively young 59, and 61 percent were male.
The chemotherapy-bevacizumab arm included 559 patients, and the chemotherapy-cetuximab arm 578.
Discussant: Follow-up May Show Survival Differences
The Discussant for the study, Josep Tabernero, MD, PhD, Head of Medical Oncology and Director of Clinical Research at Vall d'Hebron University Hospital and Institute of Oncology in Barcelona, said the increase in overall survival for patients with advanced colorectal cancer has come slowly—from 13 months in the era of chemotherapy alone, to 23 months in the best trials of chemotherapy with biological agents, and now almost 30 months in CALGB 80405.
Tabernero suggested that an expanded analysis of CALGB 80405 may show more efficacy for one regimen over the other, citing the German FIRE-3 trial of FOLFIRI combined with cetuximab or with bevacizumab, reported at last year's ASCO Annual Meeting (Abstract LBA3506).
FIRE-3 had almost identical overall response rates for FOLFIRI/cetuximab and FOLFIRI/bevacizumab (62% vs. 58%, respectively) and progression-free survival times (10.0 vs. 10.3 months, respectively). But contrary to the CALGB 80405 results, overall survival was significantly longer in patients treated with FOLFIRI/cetuximab (28.7 vs. 25 months).
“It is difficult to understand the difference in overall survival, considering the almost equal response rates and progression-free survival,” Tabernero said. He urged the CALGB 80405 researchers to conduct independent reviews of overall response rates and depth of response, to help understand why similar clinical trials had different survival results.
‘Not Practice Changing Yet’
Asked for her opinion for this article, Smitha S. Krishnamurthi, MD, Associate Professor of Medicine in the Division of Hematology and Oncology at University Hospitals Seidman Cancer Center and Case Western Reserve University, noted that oncologists have been waiting a long time for the results. “The results are very clear cut, with a practical study design, so physicians can discuss this with their patients,” she said.
But the findings will not change her practice much for the moment, she said, because she is waiting for results of extended RAS testing from CALGB 80405, which may weigh more heavily in favor of one regimen over another.
Her current approach is to offer FOLFIRI-cetuximab to patients who are RAS wild type, based on the survival difference seen in the FIRE trial. But for patients who prefer FOLFOX, she said she would feel comfortable offering either FOLFOX-cetuximab, or FOLFOX-bevacizumab, based on the results of CALGB 80405 or FOLFOX-panitumumab based on the PRIME study, while for RAS-mutant patients she would recommend either FOLFOX-bevacizumab or FOLFIRI-bevacizumab.
“FOLFOX has been the most popular chemotherapy backbone in the U.S. and Canada because of the side effects—patients tend to not lose their hair, to have less nausea and less diarrhea,” she said. “They will have more cold sensitivity and risk of peripheral neuropathy [with FOLFOX], but most patients seem to prefer that over the nausea and hair loss seen with FOLFIRI.”
Does Choice of Chemo Regimen Matter?
In his plenary presentation, Venook said that another question to be considered from CALGB 80405 is whether the choice of chemotherapy mattered, or do these biological therapies act the same with different chemotherapies?
He said among the 835 FOLFOX patients, median overall survival was 30.1 months when combined with cetuximab, versus 26.9 months with bevacizumab.
“It is interesting to note that these curves diverged at two years,” he said. “The difference was not quite statistically significantly different, but I would argue that FOLFOX-cetuximab is a reasonable option for patients in first-line advanced colorectal cancer.”
With fewer patients in the FOLFIRI arm—302—the difference between cetuximab at 28.9 months and bevacizumab at 33.4 months was also not statistically different, but it was noteworthy that the divergence was again at two years, he said.
Venook made a final point, that 124 study patients (11%) were actually rendered disease free, with median overall survival exceeding five and a half years: “This is obviously a selected group, but it means there is a subset of patients with metastatic colorectal cancer who will do exceedingly well, and we need to take that message home with us.”