By 2030, the U.S. population over the age of 65 will reach 70 million, and older patients face a significantly increased risk of developing cancer and/or dying from cancer. Two-thirds of all cancer patients are over 65, and aging cancer patients are the largest recipients of chemotherapy.
Given comorbidities, mental status, and other issues, older cancer patients are not always administered the most aggressive treatment, particularly if the regimen includes adjuvant chemotherapy. The efficacy and toxicity of adjuvant therapy in patients more than 70 years of age has been the source of controversy, but recent research has indicated that adjuvant chemotherapy can be quite successful when used selectively (see "Doublet chemo proves superior to single-agent therapy in older patients with advanced non-small-cell lung cancer"). The question remains: What is the best way to decide if an older patient is a candidate for adjuvant chemotherapy?
At an ASCO 2010 education session, geriatric medicine specialists Arash Naeim, MD, and Supriya Gupta Mohile, MD, MS, highlighted strategies that oncologists can use to determine if an older patient is fit for adjuvant therapy.
Dr. Mohile, an assistant professor of medicine in the hematology/oncology unit at the University of Rochester in N.Y. outlined how to estimate life expectancy and the benefit of adjuvant therapy based on comorbidities. Dr. Naeim, who is the director of geriatric oncology at the Jonsson Comprehensive Cancer Center, University of California, Los Angeles, discussed the importance of shared decision-making for adjuvant therapy.
Estimating life expectancy
Dr. Mohile began by posing and answering the question: Why estimate life expectancy? "Life expectancy of elders at any specific chronologic age is heterogeneous," she said. "Although some fit elders benefit from therapy, not all elders—especially in the age group [75 and older] that we're discussing—will live long enough to reap benefits from treatment. So we need to focus our improvement efforts on providing evidence-based treatment to elders who are most likely to benefit from adjuvant therapy."
But that kind of evidence-based treatment knowledge is lacking, especially because clinical trials generally don't include older patients. Dr. Mohile cited data from a study that looked at the underrepresentation of older patients in multiple Southwest Oncology Group trials and found a significant discrepancy between the percentage of elderly entered into trials (25%) vs the percentage of elderly seen in the cancer community (63%). This was particularly true in breast cancer trials where only 9% of patients in clinical trials were over the age of 65, compared with 50% of the patients seen in the cancer population, Dr. Mohile pointed out (N Engl J Med 341:2061-2067, 1999).
Variable underlying health problems paired with an uncertainty about life expectancy may create a bias against the older patient, she said. The best way to get a handle on these issues is to estimate life expectancy. While there are many life expectancy tables available, Dr. Mohile expressed a preference for one developed by Louise C. Walter, MD, and Kenneth E. Covinsky, MD, MPH (JAMA 285:2750-2756, 2001).
Dr. Walter and Dr. Covinsky looked at different chronologic ages in five-year increments and then calculated the average life expectancy as a function of underlying health status. "So a 75-year-old at very poor health only has an average life expectancy of five years vs a 75-year-old who's very fit [and] has an average life expectancy of 14 years. And this variability continues as one ages," Dr. Mohile explained. "If we think about five years being an endpoint to which one needs to live in order to benefit from adjuvant therapy, you can see that the 75-year-old at the lowest health has about a five-year life expectancy. An 85-year-old in average health also has about a five-year life expectancy, and a 95-year-old who is very fit has a five-year life expectancy." The take-home message here is that it's not just age that can determine life expectancy; health status also needs to be considered.