Best, Worst, and Typical Cancer Survival: New Estimates Tool

Katherine O'Brien

November 15, 2019

LISBON, Portugal — "How much time have I got, Doc?" That's a question never far from the mind of most metastatic breast cancer patients. Unfortunately, accurate estimates are elusive. A new approach mitigates the difficulty of that tough task by offering not one but three estimates.

Belinda Kiely, MD, a medical oncologist and senior research fellow at the University of Sydney, Australia, proposed offering patients best, worst, and typical scenarios derived from clinical trial data. Her system is not formalized as an online tool but is a simple math exercise.

She spoke here at the Advanced Breast Cancer Fifth International Consensus Conference (ABC5).

Kiely said that physicians who give advanced breast cancer patients one estimate, such as 12 months, are wrong 70% to 80% of the time. "Providing patients with a single number estimate of the average survival time is rarely accurate and conveys no hope of a possible longer survival time," she commented.

Kiely's method involves using multiples based in part on overall survival curves from clinical trials (which mirror a patients' circumstances as much as possible) to help convert a physician's estimates into the three scenarios.

In brief, doctors estimate the expected survival time for a patient, divide it by four to get the worst-case scenario, and then multiply by three to get the best-case scenario. The typical scenario is between a half and two times the doctors' estimated survival.

The approach takes into consideration the trickiness of survival statistics.

"Most data on survival times for advanced breast cancer come from clinical trials, where patients must meet specific eligibility criteria," said Kiely. "These patients tend to be younger, fitter, and have fewer other health problems and therefore often live longer than the patients we see in everyday practice."

Although trial data are a starting point in her method, Kiely's proposal offers patients a broader perspective.

Asked about potential barriers to oncologists implementing the "three-scenario" approach, Kiely said fear plays a role: "Nobody wants to be wrong." She added that some doctors don't want to risk upsetting patients and that the typical 15-minute patient encounter leaves little time for detailed prognostic discussions. In general, she's found that younger physicians have been the most receptive to the approach in her native Australia.

Take Your Cue From Patients

Patient response has been positive, as evidenced by a trial conducted Kiely and her team with 33 oncologists who spoke to 146 patients with advanced cancer about their expected survival times.

Each patient was provided with a printed one-page summary of their individual best-case, typical, and worst-case scenarios. Afterward, 91% of the patients said they found the printed information helpful, 88% said it helped them to make plans, and 88% said it improved their understanding. Also, 77% of patients said the scenarios were the same or better than they had expected.

As a practical matter, few cancer patients grasp what "median" means. Kiely said: "If we tell a patient that her estimated median survival time is 6 months, that conveys no hope of a possible longer survival, even though she has a 50% chance of living longer. On the other hand, providing three scenarios helps patients prepare for the possible worst case and, at the same time, hope for the possible best case. This is more helpful for patients making plans and decisions for the future."

She stressed that oncologists should take their cue from individual patients. Some patients aren't looking for detailed survival estimates, and those wishes should be respected. A proactive approach is best. "You should check in with patients to initiate a conversation and [gauge the level of information they want]," Kiely said. "It might be too late when the patient asks."

For patients wanting quantitative prognostic information, Kiely said using simple multiples to arrive at best, worst, and typical survival estimates provides a welcome alternative to consulting Dr Google. Searching online, patients will learn that the median survival for metastatic breast cancer is 2.5 to 3 years. They may erroneously assume that this is their specific timeline, not realizing that subtype, extent of disease, comorbidities, and other factors determine outcomes. By using multiples derived from studies that closely reflect a given patient's situation, oncologists can offer a more nuanced assessment.

Kiely acknowledged that the system isn't necessarily appropriate for every patient, but it can help some deal with practical concerns, such as planning a holiday or determining if the time has come to stop working. "In some cases, these conversations are going to be upsetting, but they are helpful for patients [and should happen sooner rather than later]."

Fatima Cardoso, MD, director, Breast Unit of the Champalimaud Clinical Center in Lisbon, Portugal, who was not involved with the research, said: "Research shows that patients who discuss these issues with their doctor have better quality of life, are less likely to undergo aggressive end-of-life resuscitation, and are less likely to die in the hospital. But at the moment, we also know that many patients are not having these conversations."

Cardoso, who is also chair of the ABC5 conference, added: "Most patients with advanced cancer want some information about how long they are likely to live, although many say they find it difficult to ask this question. The onus is on us as oncologists to start such conversations with our patients. This tool for calculating and sharing the three scenarios gives doctors the help they need to communicate with patients in a realistic and helpful way."

Advanced Breast Cancer Fifth International Consensus Conference (ABC5): Presented November 13, 2019.

Follow Medscape on Facebook, Twitter, Instagram, and YouTube


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: