Outcome of surgically treated octogenarians with breast cancerLHM Tan, J Bate, K McNamara, AR Carmichael
Department of Breast Surgery, Russells Hall Hospital, Dudley, West Midlands, United Kingdom
Context: Breast cancer is the commonest cancer in women worldwide and its incidence increases with increasing age. In an era of evidence-based practice, there is a paucity of evidence relating to the management of breast cancer in an oncogeriatric population. The authors define oncogeriatrics as cancer in patients aged 80 years or more. Aim: The study aimed to evaluate the survival benefit of surgically managed octogenarians with breast cancer. Materials and Methods: This was a retrospective study of octogenarians diagnosed with breast cancer over a 5-year period and who were treated surgically. Kaplan-Meier survival analysis was used to determine the overall survival. The differences in survival were tested using the logrank (Mantel-Cox) test. A P-value of <0.05 was considered to be statistically significant. Results: One hundred patients were included (median age- 84 years, median follow up 3.3 years). A validated adult comorbidity evaluation-27 (ACE-27) index score system was used to characterize patient comorbidities. Fourteen percent of patients had severe comorbidities, 55% nonsevere, 11% no comorbidities, and 20% with unknown comorbidities. The estrogen receptor was positive in 67% of tumors. Eighty-four percent had mastectomy and 15% had wide local excision. Sixty-eight percent had axillary lymph node dissection, 10% had sentinel lymph node biopsy, and 5% had no axillary surgery. The majority (72%) of the tumors were pathologically T1 or T2 tumor. The Nottingham Prognostic Index (NPI) mean score was 4.4. The Kaplan-Meier survival analysis showed a median survival of 5 years. Forty-eight percent died during the observation period, with 54.2% of this group dying of causes unrelated to breast cancer. Conclusion: The surgically treated octogenarians in our sample had an acceptable survival outcome
Keywords: Breast cancer surgery in octogenarians, elderly with breast cancer, octogenarians with breast cancer
Breast cancer is the most common cancer in women worldwide.  Breast cancer incidence increases with increasing age,  and together with increasing life expectancy, our societies can expect a significant proportion of healthcare resources to be needed to to manage this problem. Surgical intervention for estrogen receptor positive breast cancers in elderly women especially among those aged 80 years or older remains contentious and inconsistent. ,, There is a significant body of evidence to show that older breast cancer patients are less likely to receive standard treatment than younger patients.  In 2003, a UK based audit showed that 55% of all women aged over 80 years were being treated with primary hormonal treatment for operable breast cancer.  A study from the United States also demonstrated that patients with breast cancer aged 80 years or older were less likely to receive standard of care as recommended in consensus guidelines. 
Even with limited level I evidence that surgery impacts on the ultimate survival of the elderly with hormone-responsive cancers, there is a general consensus that the management of breast cancer in older women should not be age specific. ,,,,,, A review published by the Cochrane Collaboration in 2008, concluded that surgery offered better local control in medically fit older women (70 years plus) with estrogen receptor-unselected operable breast cancer independent of the type of surgery. However, it further concluded that surgery did not result in a significantly better overall survival.  Subsequently In 2009, the UK National Institute for Health and Clinical Excellence (NICE) governance recommended treating patients with breast cancer irrespective of age with surgery and appropriate systemic therapy, rather than endocrine therapy alone unless significant comorbidities precluded surgery. 
The burden of comorbidities in the elderly plays an important role in evaluating eligibility for most beneficial interventions as well as prognosis. However, the tools to record comorbidity in elderly are not universally used, which therefore result in variation of treatment from current best practice guidelines when compared to younger patients.  Nonetheless, surgery under local or regional anesthetic approaches when elderly patients are deemed unfit for general anesthesia may still be appropriate for local control of progressive disease.  More recently in April 2012, the International Society of Geriatric Oncology (SIOG) and the European Society of Breast Cancer Specialist (EUSOMA) recommended that standard breast surgery should be offered to the older patients. . It was also advised that adjuvant systemic treatment like chemotherapy and trastuzumab should not be age specific. 
Given the paucity of evidence to support surgical management, versus endocrine therapy in elderly patients especially in hormone receptors positive cancers, the present study was carried out .
A retrospective study of patients aged 80 years or more treated with surgery for breast cancer in a single large teaching district general hospital diagnosed over a 5-year period (2004-2008) were included. Each breast cancer diagnosed was considered as a case for our cohort as some patients had bilateral cancer. Our population cohort was identified from data collected by the West Midlands Cancer Intelligence Unit (WMCIU), who provided verified data on dates of deaths and causes of deaths related or unrelated to breast cancer. The data on comorbidity was collected from hospital case notes.The study was approved by the Institutional Review Board.
The comorbidities for this cohort of population were identified and graded using a validated adult comorbidity evaluation-27 (ACE-27) index score, which is a 27-item comorbidity for patients with any cancers. Overall comorbidities' scores were classified as: 3-severe; 2-moderate; 1-mild; 0-none, and 9-unknown.  Comorbidity measured with the ACE-27 index score and the Nottingham prognostic index (NPI) score used for characterizing the pathology of disease. 
The primary endpoints were overall survival and the cause of death related or unrelated to breast cancer. Secondary endpoints included mean survival difference in the different comorbidity groups and the survival difference between the surgically treated (ST) and the sole primary endocrine treated (PET) group.
Descriptive analyses for the non-normally distributed continuous data were presented in proportion, median, and its corresponding range. Kaplan-Meier survival curves were constructed for survival analysis. Kaplan-Meier survival curves for different comorbidities' groups were compared with a logrank (Mantel-Cox) test. P < 0.05 was considered to be statistically significant. All analyses were performed using GraphPad PRISM Version 5.0d biostatistics software.
There were 192 cases who satisfied the inclusion criteria. Ninety-two cases were treated with sole primary endocrine treatment and 100 cases were treated surgically. The surgical cohort was treated between February 2004 and February 2011. The median number of days from the date of diagnosis to the date of surgery was 30.0 days (range 0-1,615 days). The median follow-up was 1,201 days or 3.3 years (range 14-2,627 in days or 0.0-7.2 in years) from initial surgery, with the end of follow-up date taken as date of death or was censored as the end of data collection on 29 April 2011.
The characteristics of this cohort are shown in [Table 1] classified according to patient's preoperative, intraoperative, and postoperative characteristics.
Patient's preoperative characteristics
Patient's preoperative characteristics were classified according to age and comorbidities as shown on [Table 1]. The median age of patients at operation was 84.0 years (range 80-95 years). The ACE-27 index score showed that the majority of the cohort had nonsevere comorbidities. ACE-27 index graded 3 (severe comorbidity) in 14%; graded 1 or 2 (nonsevere comorbidity) in 55%; graded 0 (no comorbidities) in 11%, and graded 9 (unknown comorbidity) in 20%. Patient's preoperative tumor characteristics were classified according to estrogen receptor status, of which 67% were estrogen receptor positive.
Intraoperative and postoperative tumor characteristics
The type of surgery received by the cohort is shown in [Table 1]. Ninety-seven percent of our cohort received a single operation and 3% had two operations, where one patient needed completion mastectomy and two patients needed further excision of margins. The majority (84%) had breast conservation surgery and axillary dissection (68%). Amongst the axillary dissection group, nearly half (48.5%) had positive lymph node(s) involvement. Out of 33, 14 (42.4%) had one micrometastatic or macrometastatic lymph node; whereas, 19/33 (57.6%) had more than one lymph node positive.
Postoperative tumor biological features as shown in [Table 1], showed that the majority (72%) of the tumors were T1 and T2 tumors. In 81 cases, the size of tumors was known, with a median size of 23.5 mm (range 10-120 mm). Forty-seven percent had intermediate grade tumors and 27% had high grade tumors. The NPI mean score was 4.4 (standard deviation 1.263). Patients with positive estrogen receptor received adjuvant hormonal therapy and 31% had adjuvant radiotherapy. None of our cohort of patients had adjuvant chemotherapy.
Primary endpoints-outcome data
Primary outcome data analysis using the Kaplan-Meier survival curve showed our cohort had an overall median survival of 5 years [Figure 1], with a survival probability estimate of 73.8% and 52.6% at year 2 and 5 post surgery, respectively. Secondary outcomes data analyses for cause of death of our cohort are shown in [Table 2] and [Table 3] according to age group and ACE-27 index score group respectively. It showed that, 48 of our cohort died during the observational period of which 45.8% (22/48) died from causes related to breast cancer and 54.2% (26/48) died from causes unrelated to breast cancer.
Secondary endpoints - other analysis data-difference in ACE-27 comorbidities
In secondary data analysis, the Kaplan-Meier curves for survival for the different ACE-27 groups were compared. The median survival for patients with ACE-27 index of no. comorbidities (n = 11) was undefined (as more than 50% of the subjects were alive at the end of the study). The median survival for nonsevere (that is mild and moderate) comorbidities (n = 55) was 4.5 years. The median survival for the group of patients with ACE-27 index of severe comorbidities (n = 14) was 5.0 years. As for the unknown comorbidity group, the median survival (n = 20) was 3.3 years. It is evident that the numbers in the three groups of comorbidities; that is no comorbidities, nonsevere comorbidities, and severe comorbidities groups are not comparable and vastly different. The median survival for these three groups was not statistically significant with a P-value of 0.05 using the logrank (Mantel-Cox) test.
Other analysis data-difference in surgical treated and sole primary endocrine treated group
The baseline characteristic of the ST octogenarians in terms of age and comorbidities was statistically significantly different from the PET octogenarians, as shown in [Table 4]. Amongst the ST group, 67% had estrogen receptor positive compared to 100% of the sole PET group. When comparing the survival of ST and sole PET groups, there was no statistically significant difference in overall survival with a P-value of 0.12 using the logrank (Mantel-Cox) test. During the observational period, 37.0% (34/92) died in the sole PET group. Amongst the 34 patients who died, 20.6% died related to breast cancer, 64.7% died unrelated to breast cancer, and 14.7% died from an unknown cause of death [Table 5].
Oncogeriatric surgical practice remains a significant challenge. Age-related physiological changes can influence tolerance to standard cancer therapy and shift the overall risk-benefit ratio of oncogeriatric surgical management. Our study of octogenarians with breast cancers treated with surgery showed an overall median survival benefit of 5 years, with Kaplan-Meier survival probability estimates of 73.8% and 52.6% at year 2 and 5 post surgery, respectively. This has been shown to be similar to a retrospective observational cohort study done in Netherlands involving 346 elderly breast cancer patients aged 75 years or older, treated with either primary endocrine treatment (113) or primary surgical treatment (233). Amongst the primary surgical treatment group, the numbers of survivors for primary surgical therapy breast cancer patients was 84.1% at 2 years and 61.8% at 5 years follow-up.  In our cohort, 84.0% had breast conservation surgery and 68.0% had axillary lymph node dissection surgery. The types of breast and axillary operations performed were also similar to the experience of the Institute of Curie in Paris with 538 patients aged 70 years or older. Their patients were grouped into three age groups: 70-75 years; 75-80 years; and 80 years or older. There were 94 patients in the 80 years or older group, whereby 62.2% of these 94 patients had breast conservation operation and 77% had axillary dissection surgery. 
The limitations of this study are that it was retrospective with some resultant data being incomplete. There is also a confounding bias, as data for neoadjuvant therapy was not included in the study. There may also be limitations to the general validity of the study's conclusions due to selection bias, as the surgically treated octogenarians with breast cancer in our cohort were already selected for surgical treatment and considered fit for surgery.
The sub-analyses of this study failed to show a statistically significant difference in the median survival between the severe, nonsevere, and no comorbidity groups. Therefore, the outcome of survival in our study may not be truly representative of the heterogeneous group of elderly women with breast cancers aged 80 years or older. The loss of follow-up may have also affected the validity of the study; however, we found significant variation in the offering of surgical treatment in this cohort of octogenarian population, that is, 66.0% (66/100) of patients in ST group and 25% (23/92) in PET group had no or nonsevere comorbidity.
In concordance with the Cochrane Collaboration Review, our data did not demonstrate a statistically significant survival advantage to the ST group as compared to PET group. These patients groups were not randomized and the baseline characteristics of the patients in the two groups were statistically significantly different. The PET group was older and had worse ACE-27 comorbidities. This is consistent with the Cochrane Collaboration review published in 2008, which concluded that surgery did not result in a significantly better overall survival. 
Forty-eight percent of our surgically treated cohort died during the observational period, of which 45.8% of these patients died related to breast cancer and 54.2% died unrelated to breast cancer. When considering predictors of competing mortality in early breast cancer, it was suggested that when an individual died from causes unrelated to breast cancer it is no longer at risk for progression of cancer or death from breast cancer. The risks of death unrelated to breast cancer or its therapy is considered a competing risk of death in older population.  A study analysis of a prospectively registered data of two cohorts of primary and surgically treated elderly women with breast cancer have a higher chance of death due to other causes unrelated to breast cancer. This was shown to be similar for the primary endocrine and primary surgically treated patients in terms of tumor control and deaths related to breast cancers.  This is in contrast to a randomized controlled study that showed increasing age was associated with a higher mortality related to breast cancer among postmenopausal women with hormone-receptor positive breast cancer. 
In our cohort, none of the patients underwent adjuvant chemotherapy despite 57.6% having more than one involved node on axillary clearance and 18.0% having estrogen receptor negative tumors. The International Society in Geriatric Oncology suggested considering anthracycline-based adjuvant chemotherapy in high risk older women with node-positive or hormone-receptor negative breast cancers as it has been shown to improve survival from breast cancer in fit elderly patients. 
Another confounder is that HER-2 status was not included. Prevalence of HER-2 positive tumors in over 70-year-olds varies according to different case series.  It was suggested that adjuvant trastuzumab with or following adjuvant chemotherapy should not be withheld in older fit women provided that major cardiac dysfunctions and other relevant cardiac comorbidities are excluded.  The risks of adverse events from trastuzumab in association with chemotherapy for HER-2 expressing tumor in unfit patients is however still largely unknown. 
The reasons for the high percentage of oncogeriatric patients not receiving standard treatment are multifactorial: Age, life expectancy, competing cause of death, comorbidities, functional autonomy, and personal choice. It is recognized that elderly patients are a heterogeneous group. Therefore, better understanding in research and more research in evaluating the frailty of older patients with cancer and a comprehensive multidimensional geriatric assessment is required. , These principles can be applied to aiding the design of therapeutic decision making in oncology care in the enlarging older population.  At present there is no comprehensive assessment tool in place in routine clinical practice on which best treatment decisions can be based to ensure elderly patients receive the best treatment option; be it primary endocrine treatment or primary surgical treatment.
However, overall survival benefit should not be the only primary goal in managing elderly patients with breast cancer. Further research including randomized controlled trials in understanding this heterogeneous group of breast oncogeriatric patients is required in order to develop standardized tools in oncogeriatric preoperative assessment, taking age, life expectancy, disease-free survival, competing cause of mortality, comorbidities, tumor biology, functional autonomy, and patient's choice into consideration to achieve the best overall individualized risk-benefit of surgery delivered multidisciplinary collaboration.
West Midlands Cancer Intelligence Unit.
[Table 1], [Table 2], [Table 3], [Table 4], [Table 5]