Evaluating and Balancing the Risk of Breast Cancer-Specific Death and Other Cause-Specific Death in Elderly Breast Cancer Patients

Abstract

Purpose: The dilemma of undertreatment and overtreatment of elderly breast cancer patients is common. This study aimed to investigate clinicopathological features, treatment modalities, and survival in women diagnosed with breast cancer at age 70 years or over, and to assist clinicians in developing individualized treatment plans by balancing the risks of breast cancer-specific death (BCSD) and other cause-specific death (OCSD). Methods: This retrospective study included 420 women who were diagnosed with pathologically confirmed invasive breast cancer at age 70 years or older from January 2008 to December 2015 at Peking University People's Hospital (PKUPH). We collected baseline health status, tumor characteristics, treatment choices, and outcomes and created nomograms for clinicians to estimate individualized BCSD and OCSD risk directly. Results: During a median follow-up of 71.5 months (range 2 to 144 months) in patients with stage I-III tumors, breast cancer specific survival (BCSS) was 92.4% (376/407) and overall survival (OS) was 78.1% (318/407). There were 89 deaths, and 65.2% (58/89) were non-breast cancer related. Upon multivariate analysis by Cox regression model, tumor size, positive lymph nodes, Ki-67, and surgery were independent predictors of BCSS, and comorbidities, positive lymph nodes, Ki-67, surgery, and endocrine therapy were independent predictors of OS. Propensity score weighted (PSW) was applied to analyze therapeutic efficacy, and there was BCSS and OS benefit with surgery (both p < 0.001), BCSS benefit with chemotherapy (p = 0.029), BCSS and OS benefit with endocrine therapy (p = 0.006 and 0.004), and neither BCSS nor OS benefit with radiotherapy (RT) (p = 0.348 and 0.289). Competing-risk nomograms were developed to estimate cumulative mortality probabilities for BCSD and OCSD for individual patients according to clinicopathologic characteristics and treatments. The calibration curves displayed exceptionally, with C-indexes 0.714 for BCSD and 0.717 for OCSD. Conclusions: Older patients had greater risk of dying from non-breast cancer causes. Surgery, chemotherapy, and endocrine therapy were associated with improved survival. Competing risk nomograms allowed individual assessment of BCSD and OCSD, based on clinicopathological characteristics and treatment options, and can be used as a tool to help in choosing appropriate treatment strategies. This study was approved by the Peking University People's Hospital Research Ethics Board on September 4, 2018.

Keywords: balancing; breast cancer; breast cancer-specific death; elderly; other causes-specific death.

Figure 1

(A) Breast surgery with or without local radiotherapy by age. BCS, breast conserving surgery; RT, radiotherapy. (B) Axillary surgery (axillary lymph node dissection or sentinel lymph node biopsy) by age. (C) Regional lymph node radiotherapy by age. (D) Administration of chemotherapy according to age among patients with positive lymph nodes or triple-negative or HER-2 positive tumors. (E) Administration of endocrine therapy by age.

Figure 2

(A) Kaplan-Meier analyses of the effect of chemotherapy on breast cancer specific survival (BCSS) in weighted samples of all patients. (B) BCSS in patients with age from 70 to 74 years. (C) BCSS in patients with age from 75 to 79 years. (D) BCSS in patients age 80 years or older. (E) Kaplan-Meier analyses of the effect chemotherapy on overall survival (OS) in weighted samples. (F) OS in patients with age from 70 to 74 years. (G) OS in patients with age from 75 to 79 years. (H) OS in patients age 80 years or older.

Figure 3

(A) Kaplan-Meier analyses of the effect of endocrine therapy on breast cancer specific survival (BCSS) in weighted samples. (B) Kaplan-Meier analyses of the effect of endocrine therapy on overall survival (OS) in weighted samples.

Figure 4

(A) Kaplan-Meier analyses of different local treatment on breast cancer specific survival (BCSS) in weighted samples. (B) Kaplan-Meier analyses of different local treatment on overall survival (OS) in weighted samples.

Figure 5

Competing risk nomograms predicting 1-, 3-, and 5-year cumulative probabilities for breast cancer-specific death (BCSD) and other cause-specific death (OCSD) in elderly women with breast cancer. (A) Breast cancer-specific death. (B) Other cause-specific death.

Figure 6

Calibration curves for 1-, 3-, and 5-year predictions.