Risk Prediction Model Could Guide Metastatic Breast Cancer Decisions

A novel prediction model can help predict long-term overall survival among breast cancer patients with distant metastases, according to a new study. The model could be used as a decision-making tool for physicians and patients.

“A reliable prediction of the expected survival in breast cancer patients with distant metastasis is a critical basis for appropriate treatment selection,” wrote study authors led by Hyeong-Gon Moon, MD, PhD, of Seoul National University College of Medicine in South Korea. “However, accurate prediction of survival in a newly diagnosed metastatic breast cancer patient is one of the most difficult challenges that physicians face.”

To develop a method to address that challenge, the researchers analyzed data from 547 patients at a single institution; a validation cohort of 254 additional patients from four Korean teaching hospitals was also included. The results were published online ahead of print in Annals of Oncology.

Based on the initial cohort, a post-metastasis overall survival (PMOS) score was developed that ranged from 0 to 8. It included six factors with varying weight: primary tumor stage, hormone receptor (HR) status, and Ki-67 expression level; and metastasis-specific factors including the duration of disease-free interval, the site of the metastasis, and the presence of symptoms related to the metastatic site.

The median duration of the PMOS in the full cohort was 31 months, and each of those factors was an independent prognostic factor for PMOS. (Tumor stage and PMOS were only significantly associated in those with early relapses.)

The patients were divided into four PMOS score groups; some of the higher risk characteristics that raised a patient’s score included brain or multiple metastatic sites, high Ki-67 expression, and HR positivity. “The PMOS score in patients with distant metastasis clearly separated groups of different survival outcomes (P < .0001),” the authors wrote. The patients in the group with the lowest PMOS score (0 or 1) had a median survival of 71 months, compared with only 12 months for those in the group with the highest scores (6 to 8).

The score “successfully predicted the post-metastasis survival” in the validation cohort as well, with the two highest groups showing significantly poorer survival than the lowest (P < .0001).

Those patients with higher scores, the authors wrote, “are unlikely to benefit from intensive multidisciplinary therapy aimed at clinical remission since their disease may progress rapidly.”