Select Excerpts from the Interview

Track 2

DR LOVE: Can you discuss the background to ECOG trial 5202 in patients with Stage II disease?

DR BENSON: A great deal of emphasis has been placed on laboratory correlative work linked to outcome data from randomized clinical trials. Correlative laboratory studies can be designed in several ways, although the most typical evaluation to date is a retrospective analysis of patient tumor specimens obtained from patients in clinical trials for which outcome data are already available.

One area that has been evaluated extensively — mostly in the retrospective arena — is the phenomenon of 18q deletion, which has been associated with a much poorer prognosis. Under the direction of Stan Hamilton’s laboratory, ECOG examined tumor blocks from the primary tumors of individuals who had entered one of two of the early Intergroup adjuvant trials, which were 5-FU based.

Unfortunately, because these are older trials, it was impossible to collect all of the tumor specimens we needed. However, we did have sufficient tumor specimens to conduct a retrospective analysis, which demonstrated that those patients who retain the 18q allele have a far better survivorship than those who do not ( Jen 1994).

We also looked at microsatellite instability (MSI) in this series of patients, and it appeared that patients with MSI and evidence of the TGF-beta mutation had a superior survivorship compared to those who did not (Watanabe 2001).

In the cooperative group setting, randomized trials for patients with Stage III disease who receive a 5-FU-based regimen report five-year survival rates near 65 percent. Among patients with 18q deletion — in other words, a poor-risk group — five-year survival was only 50 percent. This correlates with survival statistics for patients who had surgery alone. The implication is that the 5-FU-based regimen had no impact on outcome.

On the other hand, those so-called good-risk patients who retained 18q or had MSI with the TGF-beta mutation had a survival rate of 75 percent at five years. What we don’t know in this population is whether this group would have done just as well with surgery alone or if the chemotherapy had some impact. We cannot answer those questions.

Track 4-5

DR LOVE: How are the high- and low-risk patients identified in trial 5202?

DR BENSON: The laboratory correlative data were felt to provide enough evidence to conduct a hypothesis-driven trial. The Intergroup elected to evaluate patients’ primary tumor specimens for the 18q allele as well as for MSI. The specimens are collected as close to the time of surgery as possible.

When the result is known for the Stage II patients who are asked to participate in the trial, patients are assigned to either the high- or low-risk group. The low-risk group will include those individuals who retain 18q. Our projected survivorship for this group is nearly 90 percent. Of course, this trial will tell us if we’re right or wrong in that assumption.

In the high-risk group, another suggestion based on retrospective data indicated that a 5-FU/leucovorin regimen would not be likely to have a major impact. Given the MOSAIC trial data implying that a worse prognostic group tends to receive a greater benefit from FOLFOX (André 2004), it was elected to incorporate FOLFOX into the randomization for the high-risk patients.

So the Intergroup strategy for both colon and rectal cancer is to randomly assign these patients between FOLFOX and FOLFOX with bevacizumab. This is what is being done for Stage II patients in the 5202 trial (4.1).

5202 is now active, and we encourage patients and physicians to participate since it is really the first large-scale prospective randomized trial to determine whether a molecular parameter in colon cancer can aid us in making treatment decisions. In addition, we will be storing tissue to evaluate other potential prognostic and predictive markers once we have outcome data available.

Select publications