HOLLYWOOD, Florida—Preliminary data from the TREE-2 randomized study of first-line regimens for metastatic colorectal cancer show that bevacizumab (Avastin) can be safely added to oxaliplatin (Eloxatin)/fluoro-pyrimidine regimens and that the addition of bevacizumab significantly improves response rates.

The data also show that bolus fluorouracil (5-FU) is less active than infusional 5-FU in this setting. Howard S. Hochster, MD, of New York University School of Medicine, reported the early results at the 2005 Gastrointestinal Cancers Symposium (abstract 241).

The TREE-2 study is a continuation of TREE-1, which is a randomized phase II study of three oxaliplatin/fluoropyrimidine regimens in metastatic colorectal cancer (see Table 1):

■ modified FOLFOX6: bolus plus infusional 5-FU/leucovorin (LV)

■ bFOL: bolus 5-FU/low-dose LV

■CapeOx: oral capecitabine (Xeloda)/oxaliplatin

After publication of studies showing that adding bevacizumab improves survival of patients treated with irinotecan (Camptosar)/5-FU/LV (IFL) by about 35% (N Engl J Med 350:2335-2342, 2004), the TREE-1 protocol was amended in November 2003 to add bevacizumab to all three study regimens.

"This randomized, multicenter study is the first to determine the safety and tolerability of oxaliplatin/fluoropyrimidine regimens (bolus, infusional, and oral) in combination with bevacizumab for first-line treatment of metastatic colorectal cancer," Dr. Hochster said. "The hypothesis was that bevacizumab plus oxaliplatin/fluoropyrimidine regimens will provide similar incremental efficacy and comparable safety as with the IFL regimen."

The study enrolled patients with measurable, untreated metastatic colorectal cancer and ECOG performance status of 1 or lower. The primary endpoint was incidence of grade 3-4 toxicities on each arm during the first 12 weeks of therapy. Secondary endpoints included standard efficacy measures.

Toxicity

Dr. Hochster reported that 223 patients were accrued between November 2003 and March 2004 in TREE-2, of whom 213 were treated. He reported toxicity data for 71 patients treated with FOLFOX-bevacizumab (FOLFOX-B), 70 treated with bFOL-B, and 72 treated with CapeOx-B. "Addition of bevacizumab to all regimens was well tolerated, and no significant additive toxicities were noted," he said.

The absolute number of patients with specific bevacizumab-related toxicities included two (2.8%) on each arm with gastrointestinal perforations, and impaired would healing in five (7%) FOLFOX-B patients, two (2.5%) bFOL-B patients, and five (4.9%) CapeOx-B patients.

There were no treatment-related deaths with FOLFOX-B, three (4.5%) with bFOL-B, and two (2.8%) with CapeOx-B.

Dr. Hochster noted that compared with TREE-1, CapeOx tolerability improved on TREE-2 when a lower cape-citabine dose of 850 mg/m² twice daily was used (compared with 1,000 mg/m² twice daily on TREE-1), with no decrease in efficacy.

Based on preliminary analysis of response data in the TREE-2 cohort, Dr. Hochster and his colleagues found that FOLFOX-B appears to be more active than bFOL-B, and that FOLFOX-B and CapeOx-B are approximately equivalent in efficacy (see Table 2).