Cancer prevention: Building on successes and failures

By learning from the past, experts hope to improve clinical trials and increase the number of effective agents available for the prevention of cancer.

Cancer prevention is designed to prevent, reverse or slow the development of cancer among healthy people at average or high risk for the disease. Over the past two decades, this growing field has been marked by triumphant clinical trials that led to FDA approvals and, perhaps surprisingly, unsuccessful trials that provided valuable information.

Although some may view these unsuccessful trials as failures, many experts agree that the data drawn from such trials have proven invaluable, shedding light on potentially harmful agents once thought to be useful in cancer prevention and elucidating ways in which clinical trials can be improved upon in the future.

“Our ultimate goal is to prevent or delay a cancer diagnosis long enough so that an individual never gets the disease,” Howard Parnes, MD, chief of the prostate and urologic cancer research group, division of cancer prevention at the National Cancer Institute, told HemOnc Today.

HemOnc Today spoke with several experts to learn more about successful trials, lessons learned from unsuccessful trials, possible improvements to trial design and the anticipated future of cancer prevention.

In 1998, the FDA approved the selective estrogen receptor modulator tamoxifen for the prevention of breast cancer among women at high risk for the disease. At that time, tamoxifen had already been FDA approved for the treatment of breast cancer for almost 20 years. Approval for breast cancer risk reduction was based on results of the Breast Cancer Prevention Trial (BCPT), which began enrolling in 1992.

Howard Parnes, MD, of the NCI, was an investigator with the PCPT and SELECT trials, both cancer chemoprevention trials. Photo by Phillip Rosenburg

The trial was initiated by the National Surgical Adjuvant Breast and Bowel Project and included 13,388 women aged 60 or older, women aged between 35 and 59 with a five-year predicted risk for breast cancer of at least 1.66%, or those with a history of lobular carcinoma in situ. Patients were randomly assigned placebo or tamoxifen 20 mg per day for five years.

Tamoxifen reduced the risk for invasive breast cancer by 49% and noninvasive disease by 50%. The selective estrogen receptor modulator also reduced the occurrence of ER–positive tumors by 69%. However, tamoxifen was also associated with an increased incidence of endometrial cancer among women aged 50 or older.

“We do use tamoxifen in high-risk women who are willing to accept the minor or moderate risk of side effects with the benefit of a 50% reduction in their risk for breast cancer,” said Powell Brown, MD, PhD, professor of medicine at Baylor College of Medicine, associate director of cancer prevention and director of the cancer prevention and population sciences program at the Dan L. Duncan Cancer Center.

A second success in breast cancer prevention came in 2007 when the FDA approved another selective estrogen receptor modulator, raloxifene, for the prevention of invasive breast cancer in postmenopausal women with osteoporosis or those at high risk for the disease.

Results of the Study of Tamoxifen and Raloxifene (STAR) trial demonstrated that raloxifene was as effective as tamoxifen in reducing the risk for breast cancer; both drugs reduced the risk by about 50%. However, compared with tamoxifen, raloxifene was associated with a 36% reduction in uterine cancers — although this was not statistically significant — and a 29% reduction in blood clots.

Success in breast cancer chemoprevention trials can be attributed to the beneficial properties of both tamoxifen and raloxifene; however, some experts acknowledge the pretrial processes that played a role in those victories.

“In the case of tamoxifen, we had trials in the adjuvant setting that demonstrated a decrease in the rate of breast cancer recurrence in the same breast and also the occurrence of a new primary breast cancer in the opposite breast. Those were randomized controlled trials, not observational studies,” Barnett Kramer, MD, MPH, associate director for disease prevention at the National Institutes of Health, told HemOnc Today.

Barnett Kramer

According to Kramer, conducting large-scale studies based on the results of randomized human studies compared with observational data or bench research provided a stronger background of evidence to justify both tamoxifen and raloxifene in the large-scale setting.

Prostate cancer prevention

Prostate cancer prevention has also seen its share of success with studies like the Prostate Cancer Prevention Trial (PCPT) and the Reduction by Dutasteride of Prostate Cancer Events (REDUCE) study, both of which showed a reduction in the risk for disease with the use of a 5-alpha reductase inhibitor.

The PCPT, which was stopped early after reaching the primary endpoint, included 18,882 men who were randomly assigned daily finasteride 5 mg or placebo for seven years. Finasteride was associated with a 24.8% reduction in prostate cancer; however, the drug was also associated with a small increase in high-grade prostate cancer.

“Over the last five years or so, subsequent analyses have shown that this increase was due, at least in part, to better detection of high-grade disease,” said Parnes, a researcher on the PCPT. “We now know that finasteride increases both overall detection of prostate cancer and the detection of high-grade prostate cancer. Therefore, the initial study results, reported in The New England Journal of Medicine in 2003, underestimated the benefits and overestimated the harms of finasteride.”

The findings of the PCPT were supported and strengthened by the recent data from the REDUCE trial, which demonstrated a similar reduction in prostate cancer prevalence with no significant incidence in high-grade disease.

REDUCE included men with PSA between 2.5 ng/mL and 10 ng/mL (aged 50 to 59) or between 3.0 ng/mL and 10 ng/mL (aged 60 to 75) at initial screening. All men had one negative prostate biopsy within six months prior to study entry. Participants were randomly assigned dutasteride or placebo; 10 core biopsies were performed at two and four years.

Dutasteride was associated with a 23% reduction in the risk for prostate cancer, and there was no overall increase in the prevalence of high-grade disease.

“For now, our successes have been with targeting pathways: the hormonal pathway in the case of breast cancer and the hormonal pathway in the case of prostate cancer,” Kramer said. “The more general gunshot approach with vitamins and minerals to date has not proven successful, at least not in a relatively well-nourished population such as the United States.”

Negative trials, valuable lessons

To date, cancer prevention studies of vitamins and micronutrients have not only proven unsuccessful in the United States but also have shed light on previously unknown toxicities associated with supplements and the people potentially at highest risk. Although studies like CARET, SELECT, ATBC and the Physicians’ Health Study II did not meet their primary endpoints, many experts agree that the universal messages they conveyed are important: Vitamins and minerals may not be as innocuous as once believed and prevention trials must be based on the strongest evidence.

“The bar should be high, which means that when you’re dealing with healthy people you should be very meticulous, not only looking for evidence of benefit but also looking for evidence of harm. The most efficient way to do that is through randomized controlled trials,” Kramer said.

Randomized, controlled trials of vitamins and micronutrients were largely based on observational and epidemiologic data, however. According to William N. William, MD, assistant professor in the department of thoracic/head and neck medical oncology at The University of Texas M.D. Anderson Cancer Center, such observational and epidemiologic studies are not adequately designed to look at whether or not certain vitamins should be given for prevention. When planning large, prospective, definitive studies, it is always best to have good epidemiologic data supported by strong biological rationale, which is obtained through extensive preclinical testing. William also acknowledges, however, that when some of the studies were designed years ago, there was a lack of good preclinical models to test prevention drugs.

William N. William

The SELECT trial, which examined the use of 400 IU vitamin E and 200 mcg selenium for the prevention of prostate cancer, demonstrated that these agents given either alone or in combination did not prevent prostate cancer. Instead, vitamin E was associated with a small increase in prostate cancer and selenium with a small increase in diabetes. “These were most likely chance associations rather than a true cause and effect relationship given that these trends were not statistically significant and, even more importantly, were not observed in men who received both selenium and vitamin E,” Parnes said.

“The rationale for SELECT was based not only on epidemiological observations and laboratory studies, but on secondary analyses of randomized, clinical trials such as ATBC,” said Parnes, also a researcher on SELECT. “It is important to remember that we did the trial precisely because we did not know if it would work; that is the reason for doing clinical trials.”

Brown agreed. He noted the importance of recognizing that there is no area of medicine or drug development that has 100% success, which results in a lower number of successful phase-1 trials of cancer drugs. Experts should, however, recognize that there is not enough knowledge to select ‘homeruns’ every time, he said.

Although SELECT was not a “positive” trial, according to Parnes, “putting nutritional supplements to the test in randomized clinical trials provides an important public health service, considering the huge number of people who take these agents every day, despite a paucity of evidence regarding both efficacy and toxicity.”

Like SELECT, the Alpha-Tocopherol, Beta-Carotene (ATBC) Cancer Prevention Trial tested vitamin E and beta-carotene in a population of 29,133 male smokers in Finland to determine whether either supplement had anticancer effects.

Vitamin E had no effect on lung cancer incidence or overall mortality, but beta-carotene was associated with an 18% increase in lung cancer incidence and an 8% increase in overall mortality. The effects of beta-carotene were stronger among men with a modest alcohol intake and those who smoked at least 20 cigarettes per day. However, vitamin E was associated with 32% fewer prostate cancer cases and 41% fewer deaths from prostate cancer. This secondary finding led directly to the inclusion of vitamin E in SELECT.

Results from the CARET study confirmed the findings from the ATBC, demonstrating a 12% increase in the risk for lung cancer and an 8% increase in death from any cause among participants assigned to beta-carotene.

“I consider CARET, in one sense, a triumph of evidence-based medicine because we got a surprise: We found that smokers may actually be at increased risk for lung cancer with what was thought previously to be an innocuous intervention,” Kramer said. “It turns out beta carotene probably does convey harm.”

Vitamins E and C were studied in the Physician’s Health Study II, a randomized clinical trial designed to assess the effect of these agents on prostate and total cancer incidence. The recently published results revealed no evidence of benefit for these vitamins in the study population of middle-aged or older men.

“The mistake and haste we’ve seen with some of these studies is taking drugs that were not tested very well in smaller settings and jumping into phase-3 studies. That’s where we need to use a little pause,” said Edward Kim, MD, assistant professor of thoracic/head & neck medical oncology at The University of Texas M.D. Anderson Cancer Center.

Improving cancer prevention trials

According to experts like Kim, there are several ways in which cancer prevention trials can improve. The first step is to examine prevention strategies in an optimized patient population.

“In many of the studies we conduct, whether in prevention or therapeutics, we don’t see a big difference or meaningful improvement with the newer drugs because there are a good number of patients in the study that don’t derive as good of a benefit, but there is a certain population that does,” he said.

Other experts like William, who co-authored a paper with Kim that was published in Nature this year, agreed that patients at very high risk for cancer should be targeted for clinical trials. Doing so would maximize trial design by lowering the number of patients necessary for enrollment, which in turn would shorten trial time. In addition, including patients who will respond the best to any given treatment or intervention and excluding those who are at highest risk for adverse events may improve trial outcomes.

“In essence, what we’re trying to do is personalize cancer chemoprevention in the same way cancer chemotherapy is being personalized today,” William said.

In addition to personalized prevention, Kim said that it is crucial to identify a biomarker as an intermediate endpoint. Because invasive cancer takes time to develop and become clinically evident, identifying a sensitive marker to serve as an intermediate endpoint would result in a reduction of required follow-up time.

However, using lessons of the past and learning more about the nature of different cancers may be one of the most invaluable strategies. According to Kramer, learning more about molecular pathways will make experts smarter, though the time when those insights could lead to successes are hard to predict.

The future of cancer prevention

Although the future of cancer prevention is uncertain, researchers continue to examine a multitude of interventions for the prevention of cancer. Despite the lack of luck with vitamins and micronutrients, most agree that they may have a role in the future of cancer prevention. However, molecularly targeted agents and the use of combined agents may be the most promising.

“Cancer has a very complicated biology and using narrowly targeted single agents is probably not going to be the major path to success,” Parnes said. “If you have a target that is both necessary and sufficient, such as BCR-ABL translocation in CML, it’s a different story. But solid tumors are genetically more complicated than that, so it seems likely that combinations of agents will need to be developed to maximize the potential of cancer prevention.”

According to Kim, treatment and prevention have lined up side by side. “We’re finding more drugs for treatment that are safer and easier to take that have lower side effect profiles and demonstrated activity, which is helping with prevention.”

And though the road to prevention has been marked with more failures than successes, many experts agree that the future of cancer prevention is bright.

“Cancer prevention has been effective. We’re going to need to educate the population that in the future, cancer prevention interventions may not be applied broadly to the population but may be best applied to the highest risk group,” Brown said.

For more information:

• Alpha-tocopherol, beta-carotene cancer prevention trial. National Cancer Institute website. www.cancer.gov
• Andriole. #LBA1. Presented at: American Urologic Association Annual Meeting; April 25-30, 2009; Chicago.
• Beta-carotene supplements confirmed as harmful to those at risk for lung cancer. National Cancer Institute website. www.cancer.gov
• Cancer prevention. National Cancer Institute website. www.cancer.gov
• Fisher. J Natl Cancer Inst. 1998;90:1371-1388.
• Gaziano. JAMA. 2009;301:52-62.
• Hampton. JAMA. 2005;294:29-31.
• Lippman. Cancer Res. 2002;62:5119-5125.
• Lippman. JAMA. 2009;301:39-51.
• Study of tamoxifen and raloxifene (STAR) trial. National Cancer Institute website. www.cancer.gov/STAR
• Tamoxifen: Questions and answers. National Cancer Institute website. www.cancer.gov/cancertopics/factsheet/therapy/tamoxifen
• William. Nature. 2009;doi:10.1038/nrd2663
• William. Cancer prevention in the 21st century. Presented at: 9th Annual Oncology Update: Advances and Controversies; January 17-21, 2009; Park City, Utah.