Highlights

• 9% of Receivers and 2% of Decliners experienced a DFS limiting event, while 3% of Receivers and 2% of Decliners died.

• After controlling for stage at diagnosis and type of oncology care received, no difference was found on the Adjusted Hazard ratio of DFS or mortality between Receivers and Decliners.

• Better clinical predictors among Decliners may be related to no rate differences in recurrence and Mortality between Decliners than Receivers.

Introduction

Approximately 3.5 million women have breast cancer in the U.S. and this number goes up each year; in 2019, an estimated 268, 600 women will be newly diagnosed with breast cancer and 41,760 will die from it [1,2]. However, among these women, about 6-13.8% of women with breast cancer willingly decline recommended chemotherapy [3-5], 9% radiotherapy [6] and 14% hormone therapy [7]. Limited evidences suggest that when women did not receive surgery, relative risk of estimated 10-year mortality was 1.58 times higher than among Receivers, while the relative mortality was 1.54 times higher when women did not receive chemotherapy [8]. 

In another study, the risk for having another breast cancer among woman who did not receive recommended systemic treatments (i.e., chemotherapy, hormone therapy) was 1.9 times, and risk for death was 1.7 times during an average 7.3 years of follow-up [5]. Recently, Johnson et al. found that 5-year survival rate of breast cancer patient was 58.1% among women who did not receive all treatments, whereas it was 86.6% for women who received all treatments; Cox hazard ratio was 5.68 [9]. However, lack of information still existed on the recurrence and mortality of women who did receive surgery but choose not to receive at least one adjuvant therapy (i.e., chemotherapy, radiotherapy, hormone therapy) after surgery. Overall objective of this study was to compare differences in recurrence and mortality between Receivers and Decliners. In this paper, Receivers indicates women who received all doctor recommended medical breast cancer treatments including surgery, chemotherapy, radiotherapy, and hormone therapy. Decliners are those who voluntarily declined all or part of the doctor recommended adjuvant therapy (i.e., chemotherapy, radiotherapy, and hormone therapy) after receiving surgery.

Methods

Design, sample, setting, and procedures

This study used longitudinal and correlational study design. This study reports results from the secondary analysis of the baseline and 5-year follow up data from the Breast Cancer Integrative Oncology Study, in which 427 women were recruited through integrative oncology clinics in greater Seattle area and the Cancer Surveillance System (CSS) registry in Western Washington State. Sample criteria included: 1) 18 years of age or older, 2) biopsy-pathology verified diagnosis of breast cancer or ductal carcinoma in situ, 3) surgery was conducted, 4) their doctor recommended at least one adjuvant treatment among chemotherapy, radiotherapy, and hormone therapy, 5) clear indication of receiving or declining recommended treatment, and 6) clear abstracted records of the following dates: diagnosis, last visit with medical oncologists or recurrence or second primary cancer diagnosis (for analysis of DFS), and death as applicable.

Instruments

Data sources

Data were collected from participants self-reports (household income and comorbidity), oncologists medical chart review (diagnosis date, last visit date with oncologists, and recurrence date), and Cancer Surveillance System (CSS) registry (age and stage at diagnosis, ethnicity, marital status, Estrogen and Progesterone Receptors status, side of breast cancer, receiving/declining recommended treatments, and last contact/death date if indicated) in Western Washington State.

Data analysis

All data were analyzed using SPSS 20 software and R Version 3.2.2 [10]. The comparison of survival and DFS between Receivers and Decliners was analyzed with Cox proportional hazards regression, using the coxph function in R. For the analysis of DFS time was computed as going from date of diagnosis to time of DFS limiting event or death for those who did die but did not recur (n = 2). Those who did not die or recur were censored at the time last known to be alive and not recurred, i.e. the last visit to the oncologist. Analysis of survival was similar, except that time of death was used as the outcome. These analyses controlled for stage of cancer at diagnosis and type of oncology care received. Kaplan-Meier survival and DFS curves were generated using the survfit function in R.

Ethical Considerations

The original study obtained the permission from the Institutional Human Subjects Review Committee of the Bastyr University and Fred Hutchinson Cancer Research Center and each subject provided informed written consent before participation. For the current study, the committees from Bastyr University, Fred Hutchinson Cancer Research Center and the University of Washington approved to conduct secondary data analysis using a de-identified data set.

Sample Characteristics

As shown in Table 1, Decliners tended to have less household income, earlier stage of breast cancer at the time of diagnosis, more estrogen positive receptor status, and more frequently received integrative naturopathic oncology care than Receivers. No other differences were found.

Recurrence and Mortality 

As depicted in Table 2, overall 31 women (30 Receivers and 1 Decliner) experienced a DFS limiting event, commonly a recurrence of cancer and 13 women (12 Receivers and 1 Decliners) died during 5 years of follow-up. The recurrence rate was 9% (30/334) among Receivers and 2% (1/41) among Decliners, while mortality was 3% (12/379) among Receivers and 2% (1/48) among Decliners. After controlling for stage at diagnosis and type of oncology care received, no difference was found on the Adjusted Hazard ratio of DFS or mortality between Receivers and Decliners. Adjusted Hazard ratio of Decliners relative to Receivers was 0.29 (95% CI; 0.04 – 2.22, p = 0.22) for DFS and 0.50 (95% CI: 0.04, 6.49, p = 0.59) for Mortality.

Figure 1 shows the Kaplan-Meyer curves for comparing disease free survival time between Receivers and Decliners, which was not statistically significant. The x-axis represents time since diagnosis and y-axis represents the fraction of women who are alive and disease free at each time.

All women (100% or 1.0 fraction) start at the top of the y-axis, indicating all women, at the time of diagnosis, have not experienced DFS limiting event or death. The lines go down at the times at which a woman in that group died or experienced a DFS limiting event. Women who did not die or recur are shown in the plot as a small vertical tick mark which indicates the time at which that woman was censored, i.e. the time at which she was last known to be alive and disease-free. Red thin lines represent Decliners and black thin lines represent Receivers.

Figure 1: Disease free survival between Receivers and Decliners.

Discussion

 Overall, the recurrence rate was 9% among Receivers and 2% among Decliners, while the mortality rate was 3% among Receivers and 2% among Decliners. These rates are much lower than Saquib et al.s finding; they found recurrence rate of 6.63% for Receivers and 19.21% for Decliners and mortality rate of 10.14% for Receivers and 11.30% for Decliners [5]. The Adjusted Hazard Ratio of 0.29 for recurrence and 0.50 for mortality in the current study is much lower than the ratio of 2.35 for recurrence and 3.05 - 5.68 for mortality found in previous studies [5,9]. This difference may be related to data collection time; Saquib et als data were from 1995 - 2000 and Johnson et al.s data were from 2004 - 2013, whereas our data were from 2012 - 2017.

Table 1: Comparison of Descriptive Statistics between Receivers and Decliners (N = 427).

Table 2: Adjusted Hazard Ratios (HR) of time to recurrence and all-cause mortality among Receivers and Decliner.