BY ROBERT H. CARLSON
BARCELONA -- Thirty experts in gastrointestinal cancer pooled their opinions on the best current treatments for metastatic colorectal cancer, the results of which were presented here at the European Society for Medical Oncology’s 16th World Congress on Gastrointestinal Cancer.
“Because medical practice is undergoing very rapid changes in data and environment, and rapidly increasing knowledge, recommendations need frequent updates and refinements,” Congress Co-Chair Eric Van Cutsem, MD, PhD, Professor of Internal Medicine at the University of Leuven and Head of the Digestive Oncology Unit at University Hospital Gasthuisberg in Belgium, said in his introduction of the expert discussion report.
In his recap of the panel’s findings, he said that the discussion was initiated with questions sent in advance to the panel members selected, based on their scientific credentials and recognition as opinion leaders from different disciplines. The answers were analyzed beforehand and discussed again by 20 of those experts during the session at the meeting. The decisions were based on evidence from clinical trials, but in many cases were also influenced by clinical experience, he said.
Still, he cautioned, “while the opinions from this discussion can help guide clinicians for treatment choices, they are not official guidelines or true consensus statements.”
Clinically Relevant Categories in Non-Resectable Patients
For the purposes of their review, the panelists divided metastatic colorectal cancer into four categories:
· potentially resectable (borderline);
· not resectable but with heavy tumor burden and aggressive biology; and
For those needing aggressive treatment, the goal is tumor shrinkage to convert to resectable disease. “There is clear evidence that patients should receive all available agents, but the optimal sequence is not known,” Van Cutsem said. “Strategy and continuum of care are important, since 70 to 80 percent of patients will receive second-line therapy, and 50 to 60 percent will receive third-line.”
Chemotherapy Backbone of First-Line Therapy
The experts agreed that a cytotoxic doublet is the appropriate first-line chemotherapy for most patients with metastatic colorectal cancer, with the three most proposed options FOLFOX (fluorouracil-leucovorin-oxaliplatin); CAPOX (capecitabine-oxaliplatin); and FOLFIRI (fluorouracil-leucovorin-irinotecan). A cytotoxic triplet was considered, specifically FOLFOXIRI, but the experts were uncertain for whom it would be appropriate.
He said that for patients with BRAF-mutant disease, a triplet with bevacizumab may be an option, although trials to date have been small. In some cases of downsizing of borderline resectable metastases, the triplet could be an option--again possibly with bevacizumab for RAS-mutant tumors.
In patients not needing aggressive treatment, such as those who are asymptomatic, fluoropyrimidine monotherapy with capecitabine in combination with bevacizumab can be considered.
Combine Biologicals in First-Line Therapy
Biologicals are indicated in first-line treatment in most patients, unless there are contraindications such as organ function involvement, poor performance status, or cardiovascular contraindications. Bevacizumab can be combined with FOLFOX, CAPOX, or FOLFIRI, Van Cutsem said, and anti-EGFR antibodies can be combined with FOLFOX or FOLFIRI.
But the experts recommended against combining a capecitabine-containing regimen with anti-EGFR antibodies: “Extended RAS analysis should be performed in order to determine resistance to anti-EGFR antibodies,” Van Cutsem said.
The panel said there is no unequivocal evidence of superiority of one biological compared with another in first-line treatment--that is, bevacizumab versus anti-EGFR antibodies. “The choice [of a biological] is determined by continuum of care and management strategy and may be influenced by safety and quality-of-life considerations, as well as by reimbursement criteria after discussion with the patient,” van Cutsem said.
In the future there may be further refinement of biological selection based on tumor localization and molecular criteria, he added.
There was uncertainty among the experts about the best cytotoxic-biological combination for potentially resectable patients if conversion is the goal, Van Cutsem said.
For RAS-mutant tumors, a doublet with bevacizumab or FOLFOXIRI with or without bevacizumab could be appropriate. For RAS wild-type tumors, a doublet with an anti-EGFR antibody, or FOLFOXIRI with/without bevacizumab, or even a doublet with bevacizumab, are all options.
“We suggest these patients be reevaluated regularly in order not to overtreat those who are resectable,” van Cutsem said. “Several of the experts, but not all, suggested a change in the cytotoxic doublet if no response is seen on first evaluation in order to maximize the chance of resection.”
The panel voiced concern about the detrimental effects of a doublet plus anti-EGFR antibody in patients categorized as being in non-tumor related performance status 2.
The final point regarding first-line therapy was that the concept of early tumor shrinkage and depth of response are related to survival, but it is unclear how specific this is for a specific drug combination, Van Cutsem said.
Extended RAS Analysis Mandatory
As has become clear with recent research, “KRAS exon-2 analysis is not enough, and extended RAS analysis is mandatory before anti-EGFR antibodies are prescribed,” he said, also referring to an earlier presentation at the meeting in which he said that extended RAS testing finds another 10 to 15 percent of patients with mutated colorectal tumors above what KRAS testing finds.
Molecular analysis of BRAF, KRAS exon-3 and -4, and NRAS exon-2, -3, and -4 should also be performed, the panel said, and there are even experimental studies on taking a new biopsy and retesting after different lines of treatment.
But full genome screening is not recommended. “Molecular analysis is hampered at the moment by lack of quality control on validated testing and sensitive testing methods, and also the cut-off of sensitivity for these molecular markers is not well defined,” Van Cutsem said.
Patients with metastatic colorectal cancer harboring any activating RAS mutation, not only KRAS, are unlikely to benefit from the addition of cetuximab to FOLFOX or to FOLFIRI. But in cases in which cetuximab is appropriate, panitumumab is equally acceptable, as the experts held that these are equally active as single agents in chemorefractory patients, although there are some differences in toxicity patterns.
The panelist said that the combination of cetuximab with irinotecan is more active than cetuximab alone in irinotecan-refractory patients, but that there are concerns about the activity in BRAF-mutant tumors.
“Many of us recommend not to treat BRAF-mutant tumors with anti-EGFR antibodies,” Van Cutsem said. On the other hand, if a patient is refractory to one of the anti-EGFR antibodies, there is no evidence to administer the alternate anti-EGFR antibody.
Optimal Maintenance Treatment is Combination
Maintenance treatment after induction benefits patients with metastatic colorectal cancer, the experts said, and the optimal maintenance treatment is a combination of a fluoropyrimidine (specifically mentioning capecitabine) plus bevacizumab.
“The challenge is to determine in which patients treatment can be stopped for a 'drug holiday,'” Van Cutsem said.
Bevacizumab monotherapy as maintenance is not recommended, since in two clinical trials non-inferiority was not proven. Also, there is limited data on maintenance with anti-EGFR antibodies.
The duration of induction treatment was recommended as six to eight cycles of FOLFOX, or six cycles of CAPOX. But the data on FOLFIRI are not clear, he explained.
Second-Line Treatment Depends on First-Line
In second-line treatment, 70 to 80 percent of the general population of patients with metastatic colorectal cancer are candidates, he said. The cytotoxic agent used in second-line therapy depends on the first-line treatment, but oxaliplatin can follow irinotecan and vice versa.
The experts listed three options in second-line treatment with biologicals:
· Post-progression continuation of bevacizumab with bevacizumab as second-line for patients not previously treated with bevacizumab;
· The VEGF inhibitor aflibercept in combination with FOLFIRI; and
· For RAS wild-type or BRAF wild-type tumors, anti-EGFR antibodies can be used in combination with FOLFIRI or irinotecan.
Unfortunately, no randomized Phase III studies comparing different biologicals are yet available, he noted.
Aflibercept is the preference for fast-progression disease in first-line therapy, although in RAS wild-type tumors the anti-EGFR antibodies are preferred.
Single Agents in Third-Line Treatment
In third-line and later lines, cetuximab or panitumumab are recommended in RAS wild-type and BRAF wild-type tumors. These are equally active as single agents, Van Cutsem said, but there is no unequivocal evidence to administer the alternate anti-EGFR antibody if a patient is refractory to one of the anti-EGFR antibodies.
Regorafenib is recommended for third-line therapy in patients pretreated with fluoropyrimidines, oxaliplatin, irinotecan, bevacizumab, or anti-EGFR antibodies.
The recommended starting dose for regorafenib is 160 mg/day on days 1 to 21 of a four-week cycle, Van Cutsem said, adding, though, that some of the experts have doubts about whether this is the optimal starting dose. “The reality is that in some regions, many physicians may start with a lower dose and then increase if there is no toxicity.” That is a matter for further exploration, he said, but frequent and close monitoring for toxicity are clearly recommended.
The 30 members on the expert panel were: Eric van Cutsem, Mario Dicato, Josep Tabernero, Nadir Arber, Dirk Arnold, Johanna Bendell, Jordan Berlin, Andrés Cervantes, Fortunato Ciardiello, David Cunningham, Stefano Cascinu, Aimery de Gramont, Eduardo Diaz-Rubio, Michel Ducreux, Richard Goldberg, Thomas Gruenberger, Karin Haustermans, Volker Heinemann, Herbert Hurwitz, Roberto Labianca, Heinz-Josef Lenz, Fotios Loupakis, Bernard Nordlinger, Philippe Rougier, Hans-Joachim Schmoll, Daniel Sargent, Alberto Sobrero, Thomas Seufferlein, Julian Taib, and John Zalcberg.
Alternate Opinion from Aimery de Gramont
In an interview at the end of the meeting, panel member Aimery de Gramont, MD, Head of the Internal Medicine-Oncology Department at Saint-Antoine Hospital in Paris, said he agreed with the majority of the report but felt that some of the statements could be improved upon.
Regarding the four patient groups that the expert report focused, de Gramont offered a suggestion about the third group, borderline resectable patients: “If borderline resectable patients could not be resected, then we don't know what happens to them.” Trials and statistics are looking at the unresectable population as well as the resected population in the borderline resectable population, but not at the unresected population in the borderline group.
This leads, he said, to what he called an “inverse Will Rogers effect” – “If you take the best unresectable borderline patients out of the unresectable group, you will decrease the results in the group of the surgical patients because those are not the best surgical patients. And you will also decrease the results in the unresectable group, because they are the best of that group.”
De Gramont said he would prefer a new idea, that resection might be a primary objective in trials in unresectable patients.
He said he also thought that “very low” front-line treatment may not be the best strategy for patients in the fourth group, those who are asymptomatic with non-aggressive unresectable tumors.
“For the patient who cannot receive aggressive therapy, if they have an indolent tumor, the deepness of response would be a way to improve survival. But if you do not choose an active treatment, you will not have a deepness of response, so it might not be the optimal way to treat even the patient with good prognosis.”
On the subject of chemotherapy, de Gramont differed from the expert recommendation on the number of cycles of oxaliplatin-based chemotherapy, saying he would not give more than six cycles of oxaliplatin (in FOLFOX).
“We have demonstrated that six is enough,” he said. “In all cancers outside of GI, I don't believe they give more than six cycles of platinum compounds. The patient cannot be reintroduced to treatment because of the neuropathy if he receives more than six cycles, so we miss the effect of reintroduction.”
On the issue of maintenance issue, de Gramont said he believes bevacizumab can be used as maintenance monotherapy: “Bevacizumab alone has been shown to be non-inferior in survival to bevacizumab-capecitabine. If the patient has any toxicity with capecitabine, I believe they can go on with bevacizumab alone.”
All these issues, he said, will need discussions in the panel's future meetings.