Adjuvant Radiation Use in Older Women With Early Stage Breast Cancer at Johns Hopkins

Abstract

Purpose

In 2004, The National Comprehensive Cancer Network (NCCN) Guidelines incorporated omission of radiation therapy after breast-conservation surgery in woman ≥70 years old with stage I, estrogen receptor-positive breast cancer who plan to receive endocrine therapy. One study demonstrated wide variation in implementing this change across 13 NCCN institutions. We evaluated the practice pattern at our institution.

Methods

We identified women ≥70 years old treated at our institution from 2009–2014. We calculated radiation therapy omission rate in those meeting the guidelines. We explored associations between radiation therapy omission, year of diagnosis, and patient characteristics with Wilcoxon rank sum tests and Fisher’s exact tests.

Results

A total of 667 women met the inclusion criteria, and 117 (18%) were candidates for radiation therapy omission. Mean age among the 117 was 76.3 years (Range: 70–95). Overall radiation therapy omission rate was 36.8%, but varied greatly by year of diagnosis (Range: 7.7–54.5%). This variation persisted after excluding women who did not receive endocrine therapy (Mean: 39.0%, Range: 0.0–75.0%). Factors associated with higher radiation therapy omission rates included older age and not having pathological nodal evaluation. The radiation therapy omission rate did not vary by race, tumor type, grade, or size.

Conclusions

The implementation of the NCCN guideline has not been consistent at our institution. Our data suggest that other tools should be considered to apply the guidelines more consistently. We have implemented a quality improvement protocol that incorporates life expectancy estimate and geriatric assessment in women meeting the NCCN guideline at our institution.

Background

The United States Census Bureau predicts that the number of individuals over age 65 years will double to nearly 83.7 million by the year 2050, and 17 million people will be over the age of 85 [1]. The majority of cancer patients are over the age of 65, so as the population ages, an even greater proportion of cancer care will be delivered to older adults [2].

In recent years, there have been several randomized control trials focusing on cancer care of older adults, one of which was the CALGB 9343, which demonstrated that women older than age 70 with hormone receptor-positive early stage breast cancer who underwent lumpectomy may not experience a survival benefit from adjuvant radiation therapy in addition to endocrine therapy compared to endocrine therapy alone [3]. At 5-year follow-up, the investigators reported a 4% local recurrence rate in women who received endocrine therapy alone, compared to a 1% rate in those receiving both endocrine therapy and radiation therapy. Importantly, there was no difference in overall survival, time to distant metastasis, or time to mastectomy in either group. Morbidity, however, was higher in the endocrine therapy and radiation therapy group, including breast pain, breast edema, skin changes, and breast retraction. A subsequent analysis with 10 year follow up again demonstrated no difference in overall survival, time to distant metastasis or time to mastectomy in either group. Local recurrence in the endocrine therapy only group was 9% vs. 2% in the endocrine therapy and radiation therapy group, showing that while the recurrence rate did rise over time, at 10 years the maximum recurrence rate remained under 10%. Furthermore, radiation therapy did not impact breast cancer–specific survival [4]. It is also important to note that surgical axillary assessment was not encouraged in this study, and was omitted in 63–64% of patients, establishing a standard that axillary surgery may also be omitted in this subset of women. This landmark study demonstrated that the risk of recurrence in older adults with early stage cancer is low, and the morbidity that is associated with radiation therapy may not outweigh the benefit that radiation provides in reducing local recurrence.

The results from the CALGB 9343 study led to a change in the National Comprehensive Cancer Network (NCCN) Guidelines in 2004 to consider omission of radiation therapy after breast-conservation surgery in woman age 70 years or older with stage I, estrogen receptor-positive breast cancer who plan to receive endocrine therapy [5]. However, a more recent study has demonstrated that the guideline recommendation has not been fully reflected in practice, and that there was great heterogeneity in implementing this change across 13 different NCCN institutions [6]. We evaluated the practice pattern at Johns Hopkins and sought to construct an internal guideline to improve our care for older adults with breast cancer.

Methods

We identified women age 70 years or older at time of diagnosis who received some or all of their care for breast cancer at Johns Hopkins from 2009–2014. Only women who met the inclusion criteria of the CALGB 9343 trial were included in our study. These criteria include tumor size of ≤ 2 cm, negative axillary lymph nodes (pathologically negative or clinically negative if lymph nodes were not sampled), positive estrogen or progesterone receptor status, age ≥70 years at time of diagnosis, lumpectomy with negative resection margins, and no other history of cancer other than in situ cervical or non-melanoma skin cancer within 5 years. Women were excluded from the study if they received neoadjuvant therapy, did not undergo surgery, or had tumor that was human epidermal growth factor receptor 2 (HER2) positive. Women who were seen as a consult or a second opinion without receiving their care at Johns Hopkins were excluded from the study.

All information was obtained retrospectively through chart review through an IRB-approved database. We evaluated factors that could affect the omission of radiation therapy, including year of diagnosis, age, tumor size, race, tumor histology, tumor grade, and nodal status (pathologically negative vs. clinically negative but unsampled).

Patient characteristics were compared between patients who did and did not receive radiation therapy using Fisher’s exact testing for categorical measures and Wilcoxon rank sum tests for continuous measures. The omission rate was estimated over the entire study period of 2009–2014 and also separately by both year and age (dichotomized at 75) for all patients and also for the subset of women who were able to tolerate endocrine therapy. Rates are reported as proportion with exact 95% binomial confidence intervals. The log odds of omission for each additional year was estimated using logistic regression with year of screening as a continuous independent variable. Analyses were completed with R version 3.2.2 (reference: R Core Team (2015). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. URL. https://www.R-project.org/.)

Results

A total of 667 women aged 70 years or older sought treatment for their breast cancer at our institution from 2009–2014. Of these, 117 (18%) were candidates for radiation therapy omission based on the NCCN guidelines and were included in the study cohort. The reasons for exclusion from the study are listed in the Table 1. The majority of patients were excluded because of positive lymph nodes, large tumor size, presence of metastatic disease, hormone receptor-negative disease, or if they underwent mastectomy.

Table 1.

Women aged 70 years or older treated at Johns Hopkins from 2009–2014 who were excluded from the analysis

	N
Total number of women ≥70 years old	667
Positive lymph node	172
Tumor >2 cm	113
Metastatic disease	82
Hormone receptor negative	64
Mastectomy	38
DCIS only	30
Did not receive care at Johns Hopkins	17
HER2 positive	14
Lost to follow up	9
Did not undergo surgery	5
Neoadjuvant therapy	6
Number of women who were included in the study	117

cm: centimeters

DCIS: ductal carcinoma in situ

HER2: human epidermal growth factor 2

N: number.

The patient characteristics of the study cohort are listed in Table 2. Mean age was 76.3 years (Range 70–95), the majority were Caucasian (83/117, 72.2%), the median tumor size was 1 cm (Range 0.2–2), the majority of tumor histology were ductal (87/117, 74.4%) and grade II (77/117, 66.4%), and 21 (17.9%) of the women had clinically negative lymph nodes and did not undergo surgical lymph node evaluation.

Table 2.

Characteristics of patients, overall and by receipt of radiation therapy. P values for differences between patients who did and did not receive radiation therapy.

	All Patients	Received RT	Omission of RT	P
	N = 117	N = 74	N = 43
Age – mean (SD)	76.3 (5.68)	74.7 (4.61)	78.9 (6.41)	< 0.001
Age -median (range)	75 (70, 95)	73 (70, 87)	78 (70, 95)	< 0.001
Age -no. (%)
70–74	58 (49.6)	44 (59.5)	14 (32.6)	0.007
75 or older	59 (50.4)	30 (40.5)	29 (67.4)
Race -no. (%)
Caucasian	83 (72.2)	54 (74)	29 (69)	0.408
African American Asian	25 (21.7)	13 (17.8)	12 (28.6)
Other	1 (0.9)	1 (1.4)	0 (0)
Unknown	6 (5.2)	5 (6.8)	1 (2.4)
	2	1	1
Tumor size (cm) -median (range)	1 (0.2, 2)	1.1 (0.2, 2)	1 (0.2, 1.8)	0.362
Tumor grade -no. (%)
I	30 (25.9)	18 (24.7)	12 (27.9)	0.675
II	77 (66.4)	48 (65.8)	29 (67.4)
III	9 (7.8)	7 (9.6)	2 (4.7)
Unknown	1	1	0
Tumor histology -no. (%)
Ductal	87 (74.4)	55 (74.3)	32 (74.4)	0.353
Lobular	17 (14.5)	13 (17.6)	4 (9.3)
Mixed	11 (9.4)	5 (6.8)	6 (14)
Other	2 (1.7)	1 (1.4)	1 (2.3)
LN Biopsy -no. (%)
Not done, clinically negative	21 (17.9)	6 (8.1)	15 (34.9)	< 0.001
Done, negative	96 (82.1)	68 (91.9)	28 (65.1)

LN: lymph node

RT: radiation therapy

cm: centimeter

Overall, radiation therapy was omitted in 36.8% of women (43/117). Thus, the majority of women (63.2%) underwent radiation therapy after the guideline change. The percentage of women for whom radiation therapy was omitted varied greatly from year to year (Range 7.7–54.5%, Table 3, Figure 1). Even after excluding women who did not receive aromatase inhibitor or tamoxifen, this large variation was still present (Range 9.1–66.7%, Table 3). There is a general reduction in the omission rate from 2009–2014 [odds ratio = 0.77, 95% CI 0.61 to 0.97, p = 0.02], though the relationship is not strictly linear, with an increase in 2014 (Figure 1).

Table 3.

Proportion of radiation omission with exact 95% confidence intervals, overall and by year of screening and age.

	All patients	Patients who received ET	Patients aged 70–74	Patients aged 75 or older
All years	36.8 [28, 46.2]	35.5 [24.9, 47.3]	24.1 [13.9, 37.2]	49.2 [35.9, 62.5]
2009	54.5 [32.2, 75.6]	55.6 [21.2, 86.3]	54.5 [23.4, 83.3]	54.5 [23.4, 83.3]
2010	47.4 [24.4, 71.1]	40 [16.3, 67.7]	16.7 [0.4, 64.1]	61.5 [31.6, 86.1]
2011	41.2 [18.4, 67.1]	66.7 [29.9, 92.5]	30 [6.7, 65.2]	57.1 [18.4, 90.1]
2012	25.9 [11.1, 46.3]	25 [8.7, 49.1]	11.8 [1.5, 36.4]	50 [18.7, 81.3]
2013	7.7 [0.2, 36]	9.1 [0.2, 41.3]	0 [0, 52.2]	12.5 [0.3, 52.7]
2014	36.8 [16.3, 61.6]	33.3 [9.9, 65.1]	22.2 [2.8, 60]	50 [18.7, 81.3]

ET: endocrine therapy

Figure 1.

Radiation therapy omission rates by year and categories of age

Older age was associated with higher radiation therapy omission rate (Table 2, Figure 1). When stratified by age, radiation therapy was omitted in 24.1% of women who are aged 70–74 years compared to 49.2% among women who are 75 years or older (p= 0.007, Table 2). Radiation therapy was more likely to be omitted in women who did not undergo pathological nodal evaluation compared to women who did (71.4% vs. 29.2%, p < 0.001, Table 2). The radiation therapy omission rate did not vary by race, tumor type, tumor grade, or tumor size (Table 2).

Among the women who received radiation therapy, the majority of patients received hypofractionated whole breast irradiation (63/74; 85%). Eleven percent (8/74) of the women received conventionally fractionated whole breast irradiation, and very few received partial breast irradiation (3/74; 4%).

Discussion

The implementation of the new NCCN guideline based on the CALGB 9343 trial was not consistent from year to year at our institution. This inconsistency cannot be completely explained by patients’ preference to not receive endocrine therapy or their inability to tolerate endocrine therapy. Despite the fact that the change in guidelines to allow omission of radiation therapy was based on category I evidence, the majority of our patients over age 70 continue to receive radiation therapy. Even though women over age 75 were less likely to receive radiation therapy at our institution, 25% of women who were 80 years or older still received radiation therapy.

We also found that the women who did not undergo a lymph node biopsy were more likely to not receive radiation therapy. We can infer from this result that perhaps women who were perceived to be frailer or to have a shorter life expectancy might have received less aggressive care from both a surgical and radiation oncology perspective. It is unclear, however, how the frailty or life expectancy assessments were made. We presume that these assessments were most likely subjective in nature and not based on objective measurement criteria. Our data suggest that despite the publication of the NCCN guideline, there remains uncertainty as to how to apply the guideline to the appropriate cohort of patients. Since age is not an accurate proxy for life expectancy, clinician assessment of overall health and life expectancy could explain the inconsistencies in our practice pattern and why the radiation therapy omission rate remains low.

Another factor that could explain our practice pattern is the concern for local recurrence. In CALGB 9343 there was a small increase in local recurrence in the group who did not receive radiation therapy. If a patient develops a local recurrence at a more advanced age, she may not be able to tolerate the needed therapy at that time if her functional reserve has declined over time. Also, the CALGB trial includes follow-up data up to 10 years. If a provider believes that a woman may have a life expectancy greater than 10 years, one may argue that radiation therapy would be appropriate for that patient since we do not have data beyond 10 years, and the overall breast cancer literature suggests that recurrence risks continue to increase slowly beyond 10 years [7]. This suggests that we need tools to help determine both life expectancy as well as risk of functional decline among older adults in applying the NCCN guideline.

In order to accomplish a more consistent approach to therapy in this group of patients, we have constructed a quality improvement protocol at our institution that incorporates tools to assess life expectancy and functional status in woman age 70 years or older. E-prognosis and the Vulnerable Elders-13 Survey (VES13) are the two tools we have selected for the protocol. Eprognosis is a group of calculators based on validated geriatric prognostic indices that predicts 5 and 10- year mortality [8–11]. The VES13 is an abbreviated geriatric assessment that has been validated to predict 5- year functional decline among community dwelling older adults [12–13]. We hope that these tools will provide additional objective assessment of an older patient in an efficient and easy-to-use manner that can be realistically applied in the current workflow of an oncology clinic. The tools will be used for patient selection both for surgical assessment of the axilla, as well as the use of radiation.

Several other studies have also shown similar findings and support a need for a consistent approach. McCormick and colleagues found a significant percentage of patients who were still receiving radiation therapy after the guideline change and only a small decrease in radiation therapy use (9%) after the guideline change [6]. Analogous to our study, they found that radiation therapy omission was associated with omission of axillary surgery and older age.

A study of Medicare patients by Soulos et al. has similarly demonstrated that radiation therapy omission rate only decreased by 4% after 2004, when the guideline was changed [14]. Even among the patients whose life expectancy was <5 years, 40.7% of the patients still received radiation therapy after the guideline change (compared to 44.4% prior to the guideline change) [14].

Another recent study, by Rutter et al. was more hopeful, demonstrating a trend towards decreased use of radiation therapy use as well as a decrease in the intensity of radiation therapy. However the decrease is small (6.6%) and a significant percentage of patients (64.7%) continue to receive radiation therapy after the 10-year follow-up study was published. Rutter and colleagues also suggested that further efforts are needed to develop tools to aid decision-making [15].

Finally, Chu and colleagues published one of the largest retrospective studies, reviewing a total of 205,860 patients who met the CALGB 9343 trial criteria using the National Cancer Database from 1998–2012. They reported only a 2.95% decrease in radiation use after the CALGB 9343 trial was published, and that almost one third of the patients over age 85 still received radiation therapy [16].

From the collective data of our study and others, it is clear that in order to apply the evidence from the CALGB 9343 trial to individual patients appropriately, providers need validated tools to assess an older patient beyond age and comorbidities. Our study has several limitations. First, this was a retrospective study collecting data from chart review. Therefore the findings described in the study are dependent on the completeness of the data in the electric medical record system at our institution. Since the detail of documentation varied amongst providers, we are limited in the number of factors involved in decision-making that we were able to assess. Second, the sample size was limited by the fact that this was a single center study since we were interested in the practice pattern at our institution compared to the national trend. Finally, we assessed the practice pattern over a 6-year period (from 2009–2014). It is possible that other external influences that were not accounted for could have impacted the practice pattern.

In conclusion, the implementation of the NCCN guideline to consider omission of radiation therapy after breast-conservation surgery in woman age70 years or older with stage I, estrogen receptor-positive breast cancer who plan to receive endocrine therapy is not consistent at our institution. Our results suggest that other tools should be employed to apply the guidelines more consistently. To achieve this, we have developed a Quality Improvement Protocol that incorporates life expectancy estimate and a brief geriatric assessment to the treatment of all woman age 70 years or older treated in our breast cancer clinics. We will be evaluating the impact of our quality improvement protocol on radiation omission rate and patient outcomes. We hope that our protocol will serve as a step towards advancing our ability to accurately risk stratify older patients with cancer. Further study is required to evaluate the impact of this new protocol on our practice pattern and patient outcomes.