A New Way of Estimating Metastatic Breast Cancer Survival

First, no one can ever truly know how long any of us will live, and any prognostic estimates from our doctors are just that: estimates. Secondly, think carefully about whether you want this best guess. Whenever I talk with women who are living with Stage IV breast cancer and this topic arises, we think very carefully together about the impact of an answer. Even remembering that no one has a crystal ball or a completely accurate guess, any number has a powerful impact.

Generally speaking, oncologists prefer to avoid answering the How long do I have? question. They know they may well not be right in their response, and they worry about diminishing hope or increasing anxiety and sadness. Sometimes, however, patients really do need at least a best guess. If you are wrestling with questions about when it is time to stop working or whether you should sell your house and move somewhere easier to maintain or maybe move closer to family, this answer can be important. Everyone needs time to complete their legal and financial planning and generally to attend to business. Parents, especially parents of young children, need time to plan.

When forced into a corner and required to respond, most oncologists say something about statistics not being predictive or helpful for any one person and then give a number that is the median survival for a particular set of circumstances. Not everyone remembers the definition of median: that equal numbers of people live shorter and longer than that estimate. Many of us instead hear the number as the length of time remaining.

At a recent meeting in Portugal, Belinda Kiely, MD, an oncologist from Australia, suggested a different way of arriving at a number. She reported first that up to 80% of estimated prognoses are wrong, and that a better way is giving patients three possible numbers: the best, the worst, and the most typical scenarios. These numbers are derived from a little math. The doctor takes the median prognosis from studies that best mirror the individual patient's situation, and then:

• Divide it by 4 to get the worst case
• Multiply it by 3 to get the best case
• Indicate that the most typical is between 2 and 2 ½ times the median number

Her follow-up of patient responses to this system was positive. More than 90% found the numbers, given to them on a printed piece of paper, were helpful, and more than 70% felt they were about as they expected. It seems obvious that giving three numbers is more helpful than a single figure; it respects the unpredictability of cancer, and gives breast cancer patients realistic hope. Each of us can imagine that we will be in the best group — and maybe even surpass that prediction.