Phase III trials comparing capecitabine
to infusional 5-FU

Our editorial in the Journal of Clinical Oncology questioned whether we need more data on substituting oral fluoropyrimidines for infusional fluorouracil with radiotherapy for rectal cancer (Crane 2004).

The X-ACT study, an adjuvant trial in patients with node-positive colon cancer, compared 5-FU/leucovorin versus capecitabine. Trial data that were initially presented at ASCO in 2004 demonstrated that capecitabine is an acceptable alternative and, in fact, may be better than IV 5-FU/leucovorin (Cassidy 2004).

I believe capecitabine is basically interchangeable with intravenous 5-FU and that the argument not to use it because of a lack of Phase III data is an excuse, not a reason. I also think that reimbursement issues are a factor. The “5-FU versus 5-FU question” has been investigated over the last 15 years in randomized cooperative group studies involving 4,000 patients, costing $10 million and without any benefit to any patients currently. While I believe we always need Phase III data, we do have Phase III data answering this question from the X-ACT trial.

The NSABP designed R-04 to compare capecitabine with venous infusional fluorouracil in patients receiving preoperative radiotherapy for locally advanced rectal cancer, but I don’t believe such a trial is necessary. There was a meeting about this, and the study design was changed to incorporate oxaliplatin, which I believe is our only opportunity to understand whether that drug will benefit such patients. The final design is a two-by-two randomization of infusional 5-FU versus capecitabine with a second randomization to oxaliplatin or not.

I believe everyone will agree that the amended design is better. If I had to guess what this trial would show, my guess is that capecitabine will be equally effective but less toxic than infusional 5-FU and that oxaliplatin will improve response but not long-term outcome.

Selection of agents for preoperative chemoradiation therapy

In most cases, single-agent capecitabine is adequate for preoperative chemoradiation. When another agent is added, such as irinotecan or oxaliplatin, we introduce levels of toxicity that, in my view, are not acceptable. The exception may be the worst cases — such as patients with large T3 tumors and positive nodes or obstructing T4 lesions where resection would be difficult. In such cases, I believe oxaliplatin improves the response rate, but I’m not sure that will translate to a meaningful clinical benefit for the whole population. It’s clear that oxaliplatin and irinotecan add gastrointestinal toxicity on the order of 25 to 30 percent and some hematologic toxicity.

At MD Anderson, we use capecitabine rather than infusional 5-FU in this setting. Capecitabine is more convenient, and while the toxicity profiles of the agents are similar, I believe the quality of life for the patients and the doctors and nurses is improved with capecitabine. Patients can remain active during treatment with capecitabine, and they don’t have a risk of infection or a five to 10 percent risk of thrombosis from a peripheral line.

At ASCO 2005, European data were presented from a randomized trial comparing preoperative radiotherapy with versus without 5-FU for the treatment of T3-4 rectal cancer (Gerard 2005). While in the United States, the standard is to give some form of 5-FU. This study failed to demonstrate a huge impact with 5-FU — with only single-digit improvements — so 5-FU may not even be necessary.

Neoadjuvant bevacizumab in the treatment of rectal cancer

Bevacizumab has been proven in many disease sites to improve the effects of chemotherapy. Approximately three years ago, before it was approved with radiation therapy, we had the opportunity to investigate this agent. We conducted a Phase I trial of 50 patients with pancreatic cancer (ID02-146) who received capecitabine, radiation therapy and bevacizumab, and the results were very exciting. In the patients who received five mg/kg of bevacizumab every two weeks, which was the final recommended dose, we saw a 50 percent partial response rate. Six of the 12 patients had their tumors shrink by 50 percent, which is a “high bar” endpoint for pancreatic cancer.

The regimen was well tolerated, and the RTOG is now conducting a Phase II study with bevacizumab, capecitabine and radiation therapy in patients with locally advanced pancreatic cancer that cannot be surgically excised (RTOG- 0411). At MD Anderson, we currently have a neoadjuvant Phase II study with the same regimen in patients presenting with locally advanced rectal cancer (2003-0832; [3.1]).

Investigators at Mass General published a Phase I trial in Nature Medicine and presented it at ASCO in 2004. In this trial, patients with primary rectal cancer received neoadjuvant bevacizumab, 5-FU and radiotherapy (Willett 2004a, 2004b). It was initially reported that five out of six patients had either microscopic residual or complete pathologic responses to the preoperative regimen, but I know from personal communication that these results are holding up, and now 11 out of 12 patients have had this response. In addition, no surgical catastrophes have been encountered following this regimen as long as six weeks elapse before the patient undergoes surgery.

These data open a lot of doors for the future of these patients and chemoradiation in general. In clinical trials, we will be evaluating bevacizumab’s ability to enhance the effect of radiation therapy. One of our focuses at MD Anderson is organ preservation, and with bevacizumab, instead of removing radiation therapy, from the neoadjuvant treatment equation, this agent, when used with radiation therapy, may lessen how radical a surgery needs to be. I want to stress that this is investigational, but the responses are better, and I believe they will also translate into better local control.

Select publications

Dr Crane is an Associate Professor, Program Director and Section Chief of the Gastrointestinal Section in the Department of Radiation Oncology at The University of Texas MD Anderson Cancer Center in Houston, Texas.