Dr. Asad Umar, acting chief of the Gastrointestinal and Other Cancers Research Group in NCI's Division of Cancer Prevention, offers his insights into the implications of two major colorectal cancer chemoprevention studies presented at the 2008 AACR annual meeting.

What is your reaction to the APC and DFMO/sulindac trial results?

NCI's Division of Cancer Prevention funded the DFMO/sulindac study, and we were confident that this was a well-designed study. We now see that this is an excellent combination for preventing the recurrence of colorectal adenomas, particularly advanced adenomas, which have the highest risk of becoming cancer.

The big question is always: How will the animal model results translate to human trials? In this case, the animal models predicted accurately. Nonsteroidal anti-inflammatory drugs like celecoxib and sulindac, and the ornithine decarboxylase inhibitor DFMO, have all shown efficacy as single agents, and there was a synergistic effect when they were given in combinations in animal chemoprevention studies.

In the APC trial, the efficacy did drop after participants stopped taking the drug, but there was still a statistically significant reduction in adenomas after being off the drug for two years. And, importantly, the strongest effect was seen in advanced adenomas, the polyps most likely to progress into cancer.

What are your thoughts about the cardiac affects related to celecoxib in the APC trial?

The NCI-funded meta-safety analysis by Dr. [Scott D.] Solomon and colleagues, which includes the APC trial, provides a better understanding of celecoxib's safety profile. By pooling the safety data from six different trials, we can better assess the cardiovascular risk of celecoxib.

We saw that this risk was greatest in those with pre-existing cardiac risk factors. In those patients without pre-existing risk factors at baseline, there was no significant risk of a cardiovascular adverse event, even with a high celecoxib dose. This result should be interpreted with some caution because the number of individuals in the low cardiac-risk category was small. However, this is an important finding because we can stratify patients according to their underlying cardiac risk factors in future studies.

We now have two approaches that may effectively prevent colorectal cancer precursors. What are the next steps?

There are a couple of things that we need to think about. How do we best design the next celecoxib trials to stratify populations into appropriate risk categories and minimize the risk of cardiovascular toxicity? We need to identify the population in which the benefits of polyp prevention will be greater than any potential risk of taking celecoxib.

We also need to better define the lowest effective dose and dosing regimen. Participants in the PreSAP trial, who were given 400 mg of celecoxib once a day, experienced fewer instances of cardiovascular toxicity compared to patients who took 200 mg twice a day or 400 mg twice a day. We can design the next trial to evaluate a once-a-day regimen with a lower dose or an every-other-day regimen. We also might want to conduct additional animal studies to see if we can decrease toxicity and maintain efficacy for adenoma prevention with a "drug holiday" approach.

As for DFMO/sulindac, Dr. [Frank] Meyskens and his colleagues are planning a larger trial of this combination. To get regulatory approval, we'll need to show that these drugs, as single agents, are effective. Future trials will need to include several treatment arms: sulindac alone, DFMO alone, and DFMO plus sulindac.