“So how long do I have to live?”

A reader with cancer recently sent me a letter with this question, which on its face seems simple enough: “My oncologist told me that the five-year survival rate for my cancer is 45 percent. What does this actually mean for me?”

I am familiar with this question from my own cancer diagnosis three decades ago, when a doctor gave me 10-year survival odds that were dismal. The first two questions racing through someone’s mind after getting a cancer diagnosis are “Is this terminal?” and “How long do I have?”

There’s much to be unpacked when confronted with survival odds — including statistics, the medical treatments available, disease progression and existential queries about life. Tackling them requires a combination of math and science, and the answers are not completely satisfying, in part because statistics won’t tell an individual patient what really lies ahead.

Researcher Melissa Troester, co-leader of the cancer epidemiology program at the University of North Carolina Lineberger Comprehensive Cancer Center in Chapel Hill, addressed the reader’s question this way: “Either they will be surviving at five years or they won’t be. It’s yes or no, not 45 percent.”

But she said the math can provide some starting guidelines that can be helpful. For instance, a 45 percent five-year survival rate, said Donald Berry, a biostatistician at MD Anderson Cancer Center in Houston, means that out of 100 patients historically diagnosed with a cancer “apparently the same” as the letter writer’s, 45 would be expected to be alive after five years. It’s a generalized prognosis, though, which Berry says may not take into account tumor type, personal genomics, health history, diet, level of physical activity and much more. In other words, it’s a big bucket that an individual may or may not fit into. As Berry put it, “Time of survival cannot be predicted perfectly.”

Which leads us to the science. Doctors can be “very bad at prognosis, at guessing how long people will live,” explained Andrew Vickers, a biostatistician at Memorial Sloan Kettering Cancer Center in New York, whose work centers on assessing the clinical value of predictive tools. “That’s because they rely on their clinical experience instead of the data.”

Berry said that “a disease may have a subgroup [of patients] that survives much longer” than statistics would suggest, usually for reasons not well understood. Pancreatic cancer, for example, he said, has an overall five-year survival rate of less than 10 percent. But Apple founder Steve Jobs learned he had a rare neuroendocrine type of the disease, which has a five-year survival rate of about 60 percent. Jobs lived eight years after his diagnosis.

Historical survival rates are also group trends that don’t account for patient individuality. Many other variables — such as age, overall health and how well an individual cancer responds to treatment — will affect a person’s chances of survival. They also don’t take into account how a recent scientific breakthrough can dramatically improve odds in a way that a 10-year average cannot show. That was definitely the case for me. The discovery of a highly effective chemotherapy cocktail not long before I began my treatment for testicular cancer dramatically shifted prospects in a way that the 10-year survival odds hadn’t yet accounted for.

When I asked Vickers the reader’s question, he quickly replied: “How individualized is that? How much does it take into account tumor characteristics and general health? How far has it spread?” Vickers has built a set of prediction tools called “nomograms” that he says “can be used to predict cancer outcomes or assess risk based on specific characteristics of a patient and of his or her disease.”

The online tools cover more than a dozen cancers. Take colorectal cancer, for which seven data points predict survival following surgery. Among them: gender, age, staging and the number of positive (cancerous) lymph nodes. The result is pure math: Given your input, there is an “X” percent chance you’ll be alive in five years.

But here’s the crapshoot: A patient’s survival odds may be better or worse when individual criteria and characteristics are considered. As much as we say we want to know more, are we ready to live with what we learn? That’s a highly personal decision.

With that I went back to Troester, who is a critic of the Memorial Sloan Kettering model’s reliance on clinical data. What about someone’s body mass index, diet, overall health and physical activity? These are “all things that interact together to determine someone’s prognosis,” she said. (“I’m sure every single model we have in medicine could be improved because there are other things that it could take into account. However, when making a prediction for an individual patient, you don’t have the luxury of saying, ‘Well, let’s do some more research and come up with something better in the future,’ ” Vickers replied. “You have to make a prediction now, and the best way to do that is to use whatever happens to be the most accurate tool around.”)

Troester also pointed out that “at the five-year mark, that whole prediction process starts anew.” By that she means that the longer you live, the longer you are likely to live.

Troester also raised one last important point: Patients are likely to get the highest-quality treatment at a designated comprehensive cancer center, and higher quality translates into longer survival. A 2015 study by Memorial Sloan Kettering researchers found “large survival differences” between patients treated for cancer at specialized cancer centers vs. those treated at community hospitals. To boost your odds with a cancer diagnosis (especially a rare or complicated cancer), it’s worth trying to get to a comprehensive cancer center — or at least to a hospital that does a high volume of cancer cases, especially of the sort of cancer that you have.

In the end, here’s what I know: Some of us want to predict the future. Others prefer to stay in the here-and-now. Some of us will be statistical outliers. Others will die of another cause. I’m reminded of Stephen Jay Gould, the noted evolutionary biologist, who was diagnosed with abdominal mesothelioma in 1982. As he famously wrote: The disease was “incurable, with a median mortality of only eight months after discovery.” Gould lived another two decades, dying of an unrelated cancer.

In other words: Plan for the worst. Hope for the best. Don’t expect stats to predict your future. And most important, live your life to the max.

This is part of an occasional series that answers patient questions. Send queries to stevenpetrow@gmail.com.