Thursday, April 22, 2010

Predictive Nomogram for Brain Metastases in Metastatic Breast Cancer

In the JCO:

Investigators from MDACC have developed a nomogram which calculates the risk of brain metastases in patients with metastatic breast cancer, validated in the current manuscript with a cohort from Canada. Risk factors are what one might expect (high grade, low age, multiple sites of metastasis, HER2 positivity). Of course this then begs the question of what to do with the information, and opens the door to discussion of a PCI trial in high risk metastatic breast cancer...

Abstract and Link

Nomogram to Predict Subsequent Brain Metastasis in Patients With Metastatic Breast Cancer [Breast Cancer]: "Purpose

Brain metastasis is usually a fatal event in patients with stage IV breast cancer. We hypothesized that its occurrence can be predicted if a clinical nomogram can be developed, thus allowing for selection of enriched patient populations for prevention trials.

Patients and Methods

Electronic medical records of patients with metastatic breast cancer were retrospectively reviewed for the period between January 2000 and February 2007 under a study approved by the institutional review board. A multivariate logistic regression analysis of selected prognostic features was done. A nomogram to predict brain metastasis was constructed and validated in a cohort of 128 patients with brain metastasis treated at the Cross Cancer Institute (Edmonton, Alberta, Canada).


Of 2,136 patients with breast cancer, 362 developed subsequent brain metastasis. Age, grade, negative status of estrogen receptor and human epidermal growth factor receptor 2, number of metastatic sites (one v > one), and short disease-free survival were significantly and independently associated with subsequent brain metastasis. The nomogram showed an area under the receiver operating characteristic curve (AUC) of 0.68 (95% CI, 0.66 to 0.69) in the training set. The validation set showed a good discrimination with an AUC of 0.74 (95% CI, 0.70 to 0.79). The nomogram was well calibrated, with no significant difference between the predicted and the observed probabilities.


We have developed a robust tool that is able to predict subsequent brain metastasis in patients with breast cancer with nonbrain metastatic disease. Selection of an enriched patient population at high risk for brain metastasis will facilitate the design of trials aiming at its prevention.


No comments: