Hyman Muss, MD
The first thing that Hyman B. Muss, MD, considers before he designs an adjuvant treatment strategy for an older patient isn’t tumor biology or metastases or anything else about her breast cancer.
Instead, he starts his deliberations by considering how long she’d be likely to live if her cancer magically disappeared.
He then considers whether his patient’s top priority is maximizing her own longevity or maximizing something else.
It is only after he answers these questions that Muss turns his attention to the specifics of cancer biology and the likely costs and benefits of various treatment regimens.
“In many cases, the final recommendation is to stick with standard of care,” said Muss, who runs the geriatric oncology program at UNC-Chapel Hill’s Lineberger Cancer Center. “In many other cases, however, the standard of care for a particular cancer doesn’t make sense for a particular patient.”
The standard-of-care model, which assumes that all patients with comparable cancers should get the same “best” treatment, works for younger cancer patients because oncologists can safely assume that most patients will live long enough to enjoy the full benefit of any cancer treatment and that most patients will endure any reasonable amount of suffering to live as long as possible.
These assumptions often prove to be false for older breast cancer patients, particularly those with other health issues. “Many older patients care less about minimizing the risk of cancer death than they care about minimizing the risk that they’ll become a burden to their families or that they’ll lose their cognitive function,” said Muss, who has authored dozens of papers on geriatric oncology. “Many other older patients are very likely to die from other causes far too quickly to justify the use toxic treatments that only outperform milder alternatives in the very long run.”Using Validated Tools
Muss believes that the single most important factor in selecting treatment regimens for geriatric patients is a good estimate of each patient’s cancer-free life expectancy. It is a frame through which all other considerations must be viewed, a frame that allows oncologists to quickly determine which treatment strategies do and don’t pay off quickly enough to warrant serious consideration.
Age alone is not enough to estimate patient longevity, Muss warns, not even if oncologists informally subtract a bit to account for each of a patient’s health problems.
“An unusually healthy 80-year-old’s life expectancy might exceed 10 years, while a diabetic 80-year-old with heart problems might have a 50% chance of dying within the year,” Muss said. “Many health-related factors must be considered, along with the relationships among those factors, and it would be a very challenging calculation for oncologists to perform—if there weren’t websites that do the calculations for them.”
Muss particularly recommends the calculator at eprognosis.ucsf.edu, which uses several validated tools to transform individual patient data into predictions about that patient’s chance of dying within set periods of time.
Mere probabilities do not, of course, provide a crisp delineation between justifiable and absurd treatment strategies. Any given patient might be the miracle, the one who starts with a 95% chance of dying almost immediately from heart failure yet somehow gets 20 good years from aggressive cancer treatment.
Nevertheless, Muss urges oncologists with geriatric patients to secure longevity estimates before they weigh treatment options, and he urges them to make such estimates a primary consideration, particularly when deciding to use chemotherapy instead of or in addition to milder treatments.
That’s not to say to oncologists should never “give patients a chance” by prescribing a standard of care that’s unlikely to outperform alternatives before patients die of something else. If the treatment in question is relatively mild (and none of the alternatives work better in the short term), it may well be the proper choice.
If, on the other hand, the treatment in question is chemotherapy and the alternatives are not, long shots are far harder to justify.
“Probabilities multiply, so the chance that two reasonably improbable things will both happen is very low—far lower than people intuitively feel it to be,” Muss said.
Consider, for example, a chemotherapy that provides a 5% survival benefit over endocrine therapy at the 10-year mark, but no benefit at the 5-year mark, and a patient with a 90% chance of dying from other causes within the decade. The chemotherapy would have a 0.5% chance of benefitting the patient and a 99.5% chance of producing needless suffering.
The accuracy of such calculations obviously hinges upon the accuracy of the underlying estimates.