Adjuvant chemotherapy in elderly patients with early breast cancer. Impact of age and comprehensive geriatric assessment on tumor board proposals

1. Introduction 

Breast cancer is the most common malignancy diagnosed among women in industrialized Countries. Nearly 40% of breast cancers occur in patients aged over 65 years and 25% in patients aged over 70 years [1]. The incidence of breast cancers in this elderly population is expected to increase in the future as a result of the ageing population. In the USA, recent demographic trends suggest that, by 2025, 72% of breast cancers will be diagnosed in elderly women [2]. Therefore, breast cancer in the elderly will represent a major public health issue during the next decades.

However, only a few data are available on the care of elderly breast cancer patients in the adjuvant setting [3], [4]. The lack of evidence-based guidelines in this subset of patients results from an underrepresentation of senior adults in previous large randomized trials over the last years.

Currently, the value of adjuvant chemotherapy in old patients with early breast cancer remains controversial. The EBCTCG meta-analysis suggests that combination chemotherapy improves relapse-free and overall survival in breast cancer patients up to the age of 70, even if this benefit seems to decrease with increasing age. However, only 609 patients ≥70 years old were included in this meta-analysis and significant conclusions on adjuvant chemotherapy benefit become arduous [5]. Similarly, recent data from the CALGB clearly suggested that selected elderly breast cancer women could benefit from adjuvant chemotherapy [6].

In routine practice, several studies revealed a strong association between increasing age and a decreased use of chemotherapy [7], [8]. However, old population is heterogeneous and it is nowadays well accepted that age alone should not be considered as a barrier to the recommendation of adjuvant chemotherapy in elderly patients. Conversely, physicians should take into account a number of other factors including tumor biology, as well as geriatric parameters such as functional and nutritional status, comorbidities and estimation of patients’ life expectancy, provided by comprehensive geriatric assessment [9]. Therefore, an oncogeriatric assessment may be useful in selecting patients who will more likely tolerate chemotherapy especially in adjuvant condition, in order to reduce elderly patients’ undertreatment [10], [11]. Tumor boards arose from the 2005 French “Plan Cancer”, gathering oncologists, radiotherapists, and specialists to establish a treatment plan recommendation for the patients. Within these tumor boards are discussed complex patients files that do not fit into standard guidelines, or which requires additional expertise from a multidisciplinary point of view, including geriatric issues.

In our institution, oncogeriatric assessment has been available since 2006. We undertook this study in order to assess the impact of a geriatric assessment versus age and other prognostic factors, on the tumor board proposal with regard to adjuvant chemotherapy for elderly breast cancer patients. Furthermore, we attempted to identify the factors that possibly influence the choice between adjuvant chemotherapy administration or follow-up.

2. Patients and methods 

We performed, in a two steps procedure, a retrospective review of all consecutive early breast cancer patients’ files that had been discussed in the breast cancer tumor board of the University Hospital of Strasbourg, a tertiary referral centre, between July 2006 and July 2009.

In step #1, patient's files were extracted from the tumor board database, if the diagnosis was breast cancer and if patient's age at diagnosis was ≥70 years old. We excluded patients with either metastatic or recurrent disease as well as patients with a medical history of any other cancer or chemotherapy administration. The variables that were recorded included medical history with significant comorbidities, age, nodal involvement (N+ or N−) and the number of lymph nodes involved, tumor size (pT, according to the UICC TNM classification, 5th edition, 2007), estrogen (ER) and progesterone receptor (PR) status (positive if >10% of cells), HER2 status, tumor grade (according to Scarff–Bloom–Richardson and SBRM) and presence of angioinvasions. We also recorded tumor board proposal (e.g. adjuvant chemotherapy versus follow-up; endocrine therapy recommendation) as well as the geriatric assessment results (namely Balducci's classification of patients into three subgroups: fit, vulnerable and frail). In our institution, we decided to refer patients between 70 and 79 years old with at least one comorbidity as well as all patients >79 years old to the oncogeriatric unit to perform a comprehensive geriatric assessment (CGA). This CGA focused on several domains including nutritional status, motor function, mood, comorbidities, medication, social support as well as residential status. Patients’ functional status was assessed with Lawton's nine item Instrumental Activities of Daily Living (IADL), Katz's Activities of Daily Living (ADL), and the Eastern Cooperative Oncology Group Performance Status (ECOG PS). Screening for depression was performed using the 15-item Geriatric Depression Scale (GDS). Cognitive function was assessed with the Mini-Mental Status (MMS). Nutritional status was evaluated by the Body Mass Index (BMI) as well as the Mini Nutritional Assessment (MNA). Comorbidity was measured with the Cumulative Index Rating Scale-Geriatric (CIRS-G). To assess motor function, patients were asked whether they experienced at least one fall during the previous year. In addition the following demographic data were recorded: age, gender, and living conditions.

Patients were characterized by the geriatric assessment as fit, vulnerable and frail. Patients with at least one major comorbidity, or geriatric syndrome or dependence were considered as frail. Patients without any comorbidity, geriatric syndrome and dependence were categorized into the fit patients subgroup. The remaining patients were considered as vulnerable.

The tumor board proposal was confronted to the current breast cancer treatment guidelines. Because of the lack of specific breast cancer treatment guidelines for elderly patients we applied our inter-regional guidelines for the appropriate use of adjuvant chemotherapy in patients aged <70 years. These guidelines are based on national and international recommendations and were currently used in North-east regions of France during the study. To account for adjuvant chemotherapy, we considered (as in younger patients) the following risk factors for disease recurrence: tumor size ≥20mm, high tumor grade (SBR III or SBRM 3–5/5), nodal involvement and ER/PR negative status. All cases that did not match the guidelines (irrelevant follow-up or chemotherapy proposal) were individually reviewed and reasons advocated for these deviations were recorded.

In step #2 we went back to medical files to investigate whether physicians who took care of patients actually followed the tumor board proposal. Reasons for treatment deviations or protocol modifications were recorded appropriately.

3. Results 

3.1. Patients’ characteristics 

Of all early breast cancer patients referred to our institution between July 2006 and July 2009 we identified a total of 192 patients aged over 70 years, which files had been discussed in our breast cancer tumor board and which characteristics appear in Table 1. The median age was 75.1 ranging from 70 to 98 years old. Advanced age was significantly associated with larger tumors. Other prognostic factors such as ER/PR and HER2 status, pathological subtype, nodal involvement, angioinvasion as well as tumor grade were similarly distributed among the whole study population.

Table 1. Patient characteristics.

	Variable	Age 70–74 (n=87)		Age 75–79 (n=54)		Age 80–84 (n=35)		Age ≥85 (n=16)
		No. of patients	%	No. of patients	%	No. of patients	%	No. of patients	%	P
	Diagnostic circumstances
	Clinical examination	37	42.5	34	63.0	20	57.1	12	75.0	0.025
	Abnormal mammogram	50	57.5	20	37.0	15	42.9	4	25.0
	Histological subtype
	Invasive ductal carcinoma	56	64.4	38	70.4	21	60.0	10	62.5	0.076
	Invasive lobular carcinoma	13	14.9	6	11.1	7	20.0	5	31.3
	Other subtypes	18	20.7	10	18.5	7	20.0	1	6.2
	Pathological tumor size
	pT1	58	66.7	25	46.3	14	40.0	5	31.3	<0.001
	pT2	28	32.2	27	50.0	19	54.2	6	37.5
	pT3	1	1.2	2	3.7	1	2.9	3	18.7
	pT4	0	0.0	0	0.0	1	2.9	2	12.5
	Pathological lymphatic invasion
	pN0	62	71.3	34	63.0	21	60.0	9	56.3	0.416
	pN1	19	21.8	16	29.6	9	25.7	2	12.5
	pN2	3	3.4	3	5.6	1	2.9	3	18.8
	pN3	3	3.4	1	1.8	2	5.7	0	0.0
	Unknown	0	0.0	0	0.0	2	5.7	2	12.5
	Angioinvasion
	Yes	20	23.0	15	27.8	7	20.0	7	43.8	0.284
	No	67	77.0	39	72.7	28	80.0	9	56.2
	Estrogen receptor
	ER+	61	70.1	44	81.5	20	57.1	10	62.5	0.388
	ER−	8	9.2	6	11.1	6	17.1	3	18.8
	Unknown	18	20.7	4	4.6	9	25.7	3	18.8
	Progesteron receptor									0.159
	PR+	36	41.4	36	66.7	15	42.9	8	50.0
	PR−	33	37.9	14	25.9	11	31.4	5	31.2
	Unknown	18	20.7	4	7.4	9	25.7	3	18.8
	HER2 status (IHC)									0.864
	HER2+	6	6.8	3	5.5	1	2.9	1	6.3
	HER2−	61	70.2	44	81.5	24	68.6	12	75.0
	Unknown	20	23.0	7	13.0	10	28.5	3	18.7
	Tumor grade (SBRM)
	SBRM1	14	16.1	2	3.7	3	8.6	2	12.5	0.365
	SBRM2	33	37.9	20	37.0	16	45.7	6	37.5
	SBRM3	18	20.7	12	22.2	4	11.4	1	6.2
	SBRM4	14	16.1	11	20.4	7	20.0	3	18.8
	SBRM5	4	4.6	7	13.0	4	11.4	2	12.5
	Unknown	4	4.6	2	3.7	1	2.9	2	12.5

3.3. Factors influencing breast cancer tumor board proposal 

The tumor board proposal with regard to either adjuvant chemotherapy or simple follow-up was compared to the current breast cancer treatment guidelines in order to identify deviations from optimal therapy, regardless of age. Among the 192 patients, 118 had at least one or more risk factors, hence justifying the discussion adjuvant chemotherapy. Out of these 118 patients, 70 (59%) were actually recommended with adjuvant chemotherapy after tumor board discussion. 136 patients presented an hormone responsive tumor. An endocrine therapy based on aromatase inhibitors was recommended to all of these patients by the tumor board.

In the univariate analysis (Table 2), factors associated with adjuvant chemotherapy recommendation included age alongside the usual aforementioned factors. However geriatric assessment results, namely the classification of patients into three subgroups: fit, vulnerable and frail patients were not significantly associated with adjuvant chemotherapy recommendation. In the multivariate logistic regression model, including age, histological subtype, tumor grade, tumor size, nodal status, ER status, HER2 status as well as geriatric assessment results, age remained an independent factor associated with a lower rate of adjuvant chemotherapy proposal, especially after 80 years old (Table 3). The most important factors governing the decision of administering adjuvant chemotherapy were large tumor size, lymph node involvement and presence of angioinvasions. Furthermore all patients with a positive HER2 status received a proposal for adjuvant chemotherapy (including trastuzumab).

Table 2. Factors associated with adjuvant chemotherapy recommendations in univariate analysis.

	Variable	Adjuvant chemotherapy recommendation (n=74)		Follow-up recommendation (n=118)
		No. of patients	%	No. of patients	%	P
	Histological subtype
	Invasive ductal carcinoma	55	74.3	70	59.3	0.014
	Invasive lobular carcinoma	12	16.2	19	16.1
	Other subtypes	7	9.5	29	24.6
	Pathological tumor size					<0.0001
	pT1	22	29.7	80	67.8
	pT2	48	64.9	32	27.1
	pT3	3	4.1	4	3.4
	pT4	1	1.4	2	1.7
	Pathological lymphatic invasion
	pN0	24	32.4	102	86.4	<0.0001
	pN1	35	47.3	11	9.3
	pN2	9	12.2	1	0.8
	pN3	6	8.1	0	0.0
	Unknown	0	0.0	4	3.4
	Angioinvasion
	Yes	38	51.4	11	9.3	<0.0001
	No	36	48.6	107	90.7
	Estrogen receptor
	ER+	46	62.2	89	75.4	0.014
	ER−	14	18.9	9	7.6
	Unknown	14	18.9	20	17.0
	Progesteron Receptor					0.346
	PR+	33	44.6	62	52.5
	PR−	27	36.5	36	30.5
	Unknown	14	18.9	20	16.9
	HER2 status (IHC)
	HER2+	11	14.9	0	0.0	<0.0001
	HER2−	47	63.5	94	79.7
	Unknown	16	21.6	24	20.3
	Tumor grade (SBRM)					<0.0001
	SBRM1	1	1.4	19	16.1
	SBRM2	15	20.3	60	50.8
	SBRM3	20	27.0	15	12.7
	SBRM4	22	29.7	13	11.0
	SBRM5	11	14.9	6	5.1
	Unknown	5	6.8	5	4.2
	Age					0.009
	70–74	39	52.7	48	40.7
	75–79	25	33.8	29	24.6
	80–84	8	10.8	27	22.9
	≥85	2	2.7	14	11.9
	Geriatric assessment results					0.052
	Fit	20	27.0	16	13.6
	Vulnerable	16	21.6	31	26.3
	Frail	2	2.7	8	6.8
	Not applicable	40	51.3	63	53.4

Table 3. Logistic regression data.

	Variable	Multivariate odds ratio	95% CI	P value
	Age category, years
	70–74	1	Ref	Ref
	75–79	0.341	0.0932–1.242	0.103
	80–84	0.009	0.0006–0.154	0.001
	≥85	0.003	0.0001–0.092	<0.001
	Tumor size
	pT1	1	Ref	Ref
	pT2–pT4	20.78	4.99–86.43	<0.001
	Nodal involvement
	pN0	1	Ref	Ref
	pN1–3	32.71	7.49–142.79	<0.001
	Angioinvasions
	No	1	Ref	Ref
	Yes	19.51	4.22–90.1	<0.001

3.4. Patients’ geriatric assessment 

In order to investigate why geriatric assessment was insufficiently taken into account for adjuvant chemotherapy recommendation, we took a closer look onto this subset of patients. Among the 192 elderly EBC patients, 93 patients (48%) underwent a geriatric assessment. The median age of these patients was 79.1 (range 71–98). Our results suggest an association between increasing age and geriatric assessment score. According to geriatric assessment, patients fell into three subgroups: 47 patients were considered fit, 36 patients were categorized in the subgroup of vulnerable patients, and 10 patients were considered as frail (Table 4). However, as only 10 patients fell into the latter group, we cannot exclude a bias in either patients’ recruitment or possibly presentation in tumor board.

Table 4. Geriatric assessment results.

	Variable	Fit patients (Balducci 1)	Vulnerable patients (Balducci 2)	Frail patients (Balducci 3)
	No. of patients	36	47	10
	Median age	75.4	80.3	87.4
	Appropriate tumor board recommendation	86% (n=31)	55% (n=26)	80% (n=8)
	Guideline concordance for adjuvant chemotherapy administration	69% (n=26)	42% (n=20)	100% (n=10)

3.5. “Appropriate” treatment recommendation and geriatric assessments results (Table 5) 

We investigated the accordance between tumor board proposals and current breast cancer treatment guidelines, taking into account geriatric assessment conclusions. We defined “appropriate” the recommendation of chemotherapy in fit patients presenting one or more risk factors and follow-up for frail patients or those who presented no risk factors for breast cancer relapse. Currently, there are no data supporting the benefit of adjuvant chemotherapy in vulnerable patients. Nevertheless, we considered chemotherapy as an appropriate treatment for vulnerable patients with at least one risk factor in order to identify the reasons advocated by the tumor board to explain the deviation from guidelines.

Table 5. Adjuvant chemotherapy recommendations and comprehensive geriatric assessment.

	Tumor board recommendation	Fit patients (n=36)		Vulnerable patients (n=47)		Frail patients (n=10)
		Chemo-therapy (n=20)	Follow-up (n=16)	Chemo-therapy (n=16)	Follow-up (n= 31)	Chemo-therapy (n=2)	Follow-up (n=8)
	Histological subtype
	Invasive ductal carcinoma	14	9	12	18	2	7
	Invasive lobular carcinoma	2	0	3	7	0	1
	Other subtypes	4	7	1	6	0	0
	Pathological tumor size
	pT1	4	11	2	15	1	1
	pT2	14	5	13	12	0	6
	pT3	1	0	1	3	1	0
	pT4	1	0	0	1	0	1
	Pathological lymphatic invasion
	pN0	4	14	4	24	0	5
	pN1	13	1	8	6	0	1
	pN2	2	0	2	1	2	0
	pN3	1	0	2	0	0	0
	Unknown	0	1	0	0	0	2
	Angioinvasion
	Yes	12	2	7	2	2	3
	No	8	14	9	29	0	5
	Estrogen receptor
	ER+	14	11	9	21	1	5
	ER−	4	1	2	6	1	2
	Unknown	2	4	5	4	0	1
	Progesteron receptor
	PR+	10	8	6	8	1	5
	PR−	8	4	5	4	1	2
	Unknown	2	4	5	4	0	1
	HER2 status (IHC)
	HER2+	6	0	0	0	1	0
	HER2−	12	11	10	26	1	7
	Unknown	2	4	6	5	0	1
	Tumor grade (SBRM)
	SBRM1	0	4	1	3	1	0
	SBRM2	4	8	2	15	0	1
	SBRM3	6	3	3	1	0	2
	SBRM4	4	1	6	6	1	3
	SBRM5	4	0	3	4	0	1
	Unknown	2	0	1	0	0	0
	Age
	70–74	8	4	5	2	0	0
	75–79	9	5	7	8	0	2
	80–84	3	7	4	14	0	0
	≥85	0	0	0	7	2	6

Five fit patients were inadequately advocated to follow-up despite one or more risk factors, therefore accounting for an 86% rate of guidelines concordance. Moreover, only 55% of vulnerable patients received an appropriate treatment recommendation. Finally follow-up was appropriately proposed to 80% of frail patients. Two frail patients were inadequately proposed with adjuvant chemotherapy.

3.6. Physicians insufficiently apply tumor board proposals in the subset of fit patients 

In order to assess to what extent tumor board proposals actually translated into the clinic, we investigated the rate of appropriate treatment implementation in fit and vulnerable patients and the reasons for not administering adjuvant chemotherapy in patients with one or more risk factors. Geriatric assessment conclusions, as well as tumor board proposals were available when oncologists had to take the final decision with regard to adjuvant therapy. Nevertheless, only 69% of all fit patients received finally an appropriate treatment. Hence, 10 patients did not receive adjuvant chemotherapy even when deemed as appropriate by current breast cancer treatment guidelines (Table 6). Among the various reasons advocated by physicians to explain the deviation from guidelines, the most frequent were age and comorbidities despite the fact that patients were considered fit patients after geriatric assessment. Only 20 patients from the 47 vulnerable patients received finally an appropriate treatment whereas 27 patients did not receive appropriate adjuvant chemotherapy. When it came to frail patients, none received adjuvant chemotherapy, in accordance with Balducci's algorithm for the management of elderly cancer patients.

Table 6. Reasons advocated to not recommend or not start the recommended adjuvant chemotherapy in fit and vulnerable patients.

	Variable	Fit patients (n=10)	Vulnerable patients (n=27)
	Old age	2	5
	Comorbidities	3	9
	Patients’ refusal	1	1
	Tumor biology	2	3
	Unknown	2	9

4. Discussion 

Breast cancer is mostly a condition that arises in elderly patients. Accordingly, nearly 25% of breast cancers occur in women aged >70 years old [1]. Numerous data including the EBCTCG meta-analysis suggest that the benefit of adjuvant chemotherapy decreases with advancing age, nonetheless, recent studies suggest a possible benefit of adjuvant chemotherapy in senior adults [5], [6]. Several retrospective studies also suggest that in routine practice, patients >65 years old are less likely to receive adjuvant chemotherapy than women <50 years [3], [12]. In all these series, age itself and physicians’ attitude did take over objective prognostic factors and geriatric parameters when it came to recommend adjuvant chemotherapy in elderly women. Moreover the recent CALGB data suggest that age alone should not be considered as a barrier to the administration of chemotherapy and that various factors including tumor biology as well as patients health status should be taken into account for selecting women who will more likely benefit from adjuvant chemotherapy. Therefore, there is some uncertainty to whether adjuvant chemotherapy translates into an overall survival benefit after 70 years old. In the setting of the increasing life expectancy, especially in women, the challenge is to balance between pros and contras of chemotherapy, keeping in mind that a fit 70 years old patient might live some additional 15 years. In this perspective one can favourably consider the benefits of adjuvant chemotherapy, at least with regard to prolonged disease-free, if not overall survival. Our results, in accordance with other published data, suggest a strong association between increasing age and the diagnosis of breast cancer at more advanced stages (e.g. clinically apparent) [13], [14]. Indeed, the breast cancer screening program in France focuses on women between 50 and 75 years old, who undergo mammography every two years. There is no recommendation for systematic mammography after 75 years old, although patients may ask for additional screening. Hence, elderly patients are more prone to present, even in the setting of early disease, with larger tumors, an important issue to consider for recommending adjuvant chemotherapy. Conversely, although geriatric assessment certainly reduces elderly breast cancer undertreatment in selecting good candidates for adjuvant therapy, there might be an age limit beyond which adjuvant therapy becomes irrelevant. However, we lack long-term follow-up with regard to adjuvant chemotherapy in elderly breast cancer patients. Such studies are definitely warranted. Tools such as Adjuvant!Online are currently used by medical oncologists to estimate the risk of relapse or cancer-related death in breast cancer patients.

Indeed, this model is very sensitive to parameters such as clinical evaluation, and integrates comorbidity data. However, the mathematical model was built on a young (<70 years old) breast cancer patients subset, and further validated on a larger population including 26.2% of patients between 66 and 75 years old, whereas only 9.2% of patients were older than 75. Hence, this tool might be more relevant to women of less than 70 years. Futures directions in this field should focus on incorporating geriatric assessment results to estimate more accurately the benefits of adjuvant chemotherapy in elderly breast cancer patients. Indeed, elderly patients should be – as are the young patients – informed about pros and contras of chemotherapy, including survival estimates.

Quality of life is another important issue to consider in the setting of adjuvant therapy, especially when there is uncertainty to whether the treatment is beneficial or not for an individual. Chemotherapy-induced toxicity such as neuropathy, weight loss or fatigue are of importance and may deteriorate the patients’ quality of life before any benefit is seen in terms of disease relapse. Although geriatric assessment is helpful in precociously detecting patients at risk, future work on adjuvant chemotherapy for the elderly should necessarily incorporate quality of life assessments.

In the present study, we assessed the impact of geriatric assessment among other prognostic factors usually considered in tumor boards to recommend or not adjuvant chemotherapy in breast cancer patients. In routine practice, the main factors that influence adjuvant chemotherapy recommendation are prognostic factors such a tumor size and nodal involvement. Surprisingly, ER/PR status seemed to be insufficiently considered by the physicians in tumor boards. Indeed, recent data suggest that the major benefit of standard adjuvant chemotherapy occurred in patients with hormone-receptor negative tumors [6]. This finding is in accordance with the previous published Oxford Overview. ER/PR status seems to be one of the most important marker which should be taken into account for recommending adjuvant chemotherapy. Moreover, in the next decades, prognosis prediction by different tools such as MammaPrint, Oncotype DX may be helpful in identifying patients who might benefit from adjuvant chemotherapy in addition to endocrine therapy.

Our results suggest that CGA is insufficiently taken into account in routine practice when it comes to recommend but also to administer adjuvant chemotherapy. Conversely, very few frail patients were given chemotherapy in our study. This probably reflects a pragmatic attitude, where physicians (with or without geriatric assessment) deemed their patients not suitable for chemotherapy anyway. Therefore, and as previously mentioned, our results might be biased on this point if these patients’ files were not discussed in tumor boards. Indeed, a subset of fit patients (14%) were not recommended chemotherapy. Furthermore, only 69% of fit patients finally received an appropriate chemotherapy. These results suggest that despite a favourable CGA, the physicians remain reluctant to administer adjuvant chemotherapy. We have previously investigated the current opinions and needs toward geriatric oncology among primary care physicians and specialists, and shown that the latter category expressed a strong need for continuous medical education on this topic [15]. Indeed, specialists having cancer patients under their care stated that CME was of outstanding (37%) and average (49%) for their practice. Our results are strongly consistent with these prior findings. To this end, we suggest that an educational effort should be conducted toward oncologists with regard to the geriatric assessment results significance. Hence, elderly “fit” patients should be less undertreated and receive adjuvant chemotherapy as much as young patients do.

Conflict of interest statement 

The authors have no conflict of interest.