Quality of Life in Long-Term Elderly Early-Stage Breast Cancer Survivors from Spain

1. Introduction

The The prevalence of elderly breast cancer survivors is increasing: for example, in 2022 in the USA, this prevalence was estimated at over 2,200,000 elderly (>70 years) breast cancer survivors [1]. Many of these elderly breast cancer patients become long-term breast cancer survivors (LTBCS) (≥5 years after diagnosis) thanks to earlier detection through screening programmes, improved treatments, and the quality of survivor care. A growing population within this group are very long-term breast cancer survivors (VLTBCS) (≥10 years’ post-diagnosis) [2,3].

Quality of life (QOL) in LTBCS has become a major area of study in clinical work and research [3,4]. It is helpful to study the QOL of LTBCS in order to identify symptoms, address impairments in functioning, improve survivorship care and determine whether certain symptoms and limitations relate more to age than to cancer [5].

QOL is a main treatment aim for elderly breast cancer patients [6], i.e., those aged over seventy years old [7]. Clinicians must understand the specific values and QOL priorities of these patients [8] in order to deliver personalised cancer care [9].

Several QOL studies of elderly LTBCS have been performed. However, more are needed that evaluate elderly-specific QOL areas and have a high participation rate of VLTBCS (since the risk of low participation among elderly survivors is high and more questionnaires are usually completed by those with a better QOL) [2,5,10,11,12]. These LTBCS studies suggest that overall QOL may be comparable to QOL in a control population, though restrictions persist in functional performance and symptoms such as cognitive and emotional impairments, pain and fatigue.

Although a small percentage of breast cancer patients (5-10%) are diagnosed with metastatic cancer, 30% of those diagnosed at initial stages develop metastatic breast cancer [13]. These patients differ from those at initial disease stages since they may have different disease-related characteristics (e.g., metastases) and treatments (e.g., prolonged treatments that may bring long-term toxicity but do not usually aim to cure the disease) that exert a wider psychosocial impact on them and their families owing to the uncertainty due to living with an incurable disease. These characteristics and treatments may have a great impact on their QOL [14].

The aims of this study were to characterise QOL in a sample of elderly early stage LTBCS (>5 years after diagnoses, including a group >20 years after diagnosis) and assess how key demographic and clinical characteristics, together with some general QOL aspects, may affect elderly-specific QOL areas. We expected relatively high QOL scores, with moderate or light limitations in symptoms and emotional and functioning areas. We also expected few differences between groups based on current age, as well as higher QOL in elderly-specific areas in patients that have no limiting comorbidity, do not live alone, and have lower symptoms and higher functioning.

2. Materials and Methods

A consecutive sample of breast cancer patients treated at the Oncology Departments of Hospital Universitario de Navarra were invited by their treating oncologist to participate in this study between June 2021 and November 2023 during their follow-up consultations at the Oncology Departments.

Inclusion criteria were age > 70 years at study entry, primary breast cancer, and stage I-IIIA at diagnoses. They had undergone breast surgery with axillary evaluation by clinical discretion (SLNB, ALND, no surgery) ≥ 5 years earlier (LTBCS) and may have received adjuvant treatments (radiotherapy, chemotherapy and/or endocrine treatment). Exclusion criteria were a second line of treatment, a second tumour during follow-up, relapse during follow-up, a cognitive state that precluded QOL assessment, and a life expectancy of less than three months.

Measures

All patients completed EORTC (European Organisation for Research and Treatment of Cancer) Quality of Life questionnaires QLQ-C30 (Quality of Life core questionnaire) [15] and QLQ-BR42 (Breast cancer module) [16], the latter of which is an updated version of QLQ-B23, and QLQ-ELD14 (elderly cancer patients’ module) [17]. Our group validated QLQ-C30, QLQ-BR23 and QLQ-ELD14[18,19,20] for use in our country. Questionnaires with under 70% of items answered were excluded. Table S1 in the supplementary material shows the main features of these questionnaires. QLQ-C30 evaluates areas common to various tumour sites and their treatment; QLQ-BR42 evaluates aspects specifically related to breast cancer; and QLQ-ELD14 evaluates important age-specific issues for elderly cancer patients (any tumour site).

The treating oncologist assessed toxicity using selected items from version 5.0 of the NCI Common Toxicity Criteria scale [21], limiting comorbidity using the Charlson questionnaire [22], and performance status using the Karnofsky scale [23]. These assessments were conducted at longer sessions on days scheduled for habitual consultations in the oncology departments. QOL assessments were also conducted at these sessions to avoid the burden of extra sessions for patients.

Data collection procedures

Patients were invited to complete the QOL questionnaires twice: once at study entry and again one year later to assess changes in QOL scores after one year. Questionnaires were completed at consultations (rather than by telephone or mail to decrease the risk of missing data). This study was approved by the Regional Ethics Committee for Biomedical Research of Navarra and followed the recommendations of the Declaration of Helsinki (project number 2019/57). All patients provided signed written informed consent.

Statistical analysis

Descriptive statistics were used to summarise the demographic and clinical characteristics and the QOL scores in the three questionnaires at the two assessments. Comparisons between two groups based on age at assessment (<75 years and >75 years) were performed using χ² for categorical variables and Kruskal-Wallis for continuous variables (p<0.05). Comparisons in the three QOL questionnaires (first assessment) were adjusted by time since surgery using linear regression models.

Differences in the QOL scores (first assessment) between groups based on demographic and clinical variables (age at treatment, cohabitation, radiotherapy (yes/no) and limiting comorbidity) were studied. All analyses were conducted in bivariate analysis (Kruskal-Wallis) and complemented by adjusting for time since surgery (linear regression models). Cocks et al.’s criteria [24] were used to interpret differences between groups in the QLQ-C30.

Changes between the two within-patient QOL assessments were estimated (Wilcoxon test). Differences in the QLQ-C30 were interpreted with Cocks et al. criteria [25] (which could be trivial, small, medium, or large in both comparisons).

Multivariate analyses (linear regression models, backward regression method) were then performed at the first assessment with the QLQ-ELD14 scales of ‘mobility limitations’ and ‘maintaining purpose’ (having a positive outlook in life, being motivated to continue with activities) as dependent variables and, as determinants, the demographic (cohabitation) and clinical (breast and axillary surgery, chemotherapy, radiotherapy, endocrine treatment, comorbidity) variables studied in the previous group comparisons, the six QLQ-ELD14 scales (since these showed more limitations than the other two QOL questionnaires), and selected QLQ-C30 scales, i.e., emotional functioning, fatigue and pain (which are key determinants of QOL in breast cancer [26]). Age at evaluation and years since beginning of treatment were included in all three models. The QLQ-ELD14 scales for mobility limitations and maintaining purpose were studied due to their relevance for elderly patients and because they frequently showed differences in the group comparisons in our study.

3. Results

3.1. Characteristics of the sample

A total of 316 women out of 334 candidates invited to participate in this study (participation rate 94.6%) performed the first assessment. Reasons for not completing the questionnaires were administrative failure, i.e., changes to days programmed for follow-up interviews (eleven cases) and participant refusal (seven cases). The second assessment was performed by 305 patients. Reasons for not performing the second assessment were relapse (six cases), death (three cases), and administrative failure (two cases). The mean age of the total sample was 75.8 years. The mean number of years since start of treatment was 12.0 (ranging from 5 to 33). Also, 27.6% of patients had more than 16 years of follow-up and 51.6% were 60–69 years old when treatment began. Most patients (57.3%) lived with another person, 46.2% had limiting comorbidity, and 83.8% had received endocrine treatment (Table 1).

3.2. QOL Scores

QOL scores were high in most QOL areas. Moderate limitations (>30 points) occurred in the QLQ-BR42 in sexual functioning/enjoyment and musculoskeletal symptoms, and in three QLQ-ELD14 areas: worries about others, maintaining purpose, and joint stiffness. Slight limitations (20–29) occurred in the QLQ-C30 in Global QOL, emotional functioning, fatigue, pain and insomnia; in the QLQ-BR42 in future perspective and breast satisfaction; and in the QLQ-ELD14 in future worries and family support (Table 2). None of the QLQ-C30 scores were under the QLQ-C30 threshold for clinical significance [27]. QLQ-C30 scores in the general sample were in line with the reference values for women aged ≥70 in Spain[28]. Small differences[24] in favour of our sample appeared in role, social functioning and constipation (7–8 points) and medium differences appeared in Global QOL (11 points).

3.3. Group Differences and Changes Within a Year

Differences between the two age groups (<75, >75 years) appeared in demographic and clinical variables: the >75 years group showed a higher frequency of widows, people who lived alone, axillary node dissection and comorbidity, as well as a lower mean Karnofsky. With regard to the QOL questionnaires, the >75 years group showed lower QOL than the <75 years group (after adjusting for years since beginning of treatment) in the QLQ-C30 in summary score, physical functioning, cognitive functioning, global QOL, fatigue, insomnia, appetite loss and constipation (differences of 5–8 points, i.e., small differences in all cases) as well as in the QLQ-ELD14 in mobility, maintaining purpose, joint stiffness and family support (differences of 8–14 points) (Table 1 and Table 2).

Differences between groups based on age at treatment appeared in physical and cognitive functioning (QLQ-C30; small differences), sexual functioning and breast satisfaction (QLQ-BR42), and mobility, maintaining purpose and family support (QLQ-ELD14), with better QOL in patients who were younger at time of treatment (Table S2 to S4 in supplementary material). Differences between groups based on cohabitation appeared in dyspnoea (QLQ-C30; small difference), sexual functioning (QLQ-BR42), and maintaining purpose (QLQ-ELD14), with better QOL in those living with a partner. Differences between groups based on radiotherapy (yes/no) and limiting comorbidity appeared in dyspnoea (QLQ-C30; small difference), sexual functioning (QLQ-BR42), and maintaining purpose and family support (QLQ-ELD14), with better QOL in those who had not received radiotherapy and had no limiting comorbidity (Tables S2–S4).

Regarding the assessment of changes within a year of life, significant differences were observed in six QLQ-C30 areas (0.7–3.8 points, i.e., trivial differences)[25], in two QLQ-BR42 areas (0.7–2.0 points), and in one QLQ-ELD14 area (0.9 points). Scores were higher in the first assessment in all areas except for sexual functioning (Supplementary Table 5).

3.4. Multivairiate Analyses

The multivariate model fitted to assess factors associated with mobility limitations showed that the presence of limiting comorbidity, fatigue and joint stiffness are risk factors of reporting higher scores in mobility limitations, whereas maintaining purpose had a protective effect. Similarly, the model assessing factors associated with maintaining purpose showed that living with a partner, having high scores in emotional functioning and having pain and family support are related to higher scores, whereas not having axillary surgery and having mobility limitations and future worriers are related to lower scores (Table 3).

4. Discussion

QOL scores in a sample of Spanish elderly early stage LTBCS were high in most QOL areas. The main limitations appeared in sexuality areas, skeletal symptoms, and three elderly-specific scales. QLQ-30 scores were in line with Spanish general population reference values [28]. Differences between age groups (<75, >75 years) showed few QOL differences. These mainly involved lower QOL in the >75 years group. Combinations of lower symptoms and higher functioning with cohabitation and clinical variables were related to higher mobility limitations and higher maintaining purpose.

Levels of compliance were high since 94.6% of the candidates agreed to participate, 91.4% filled in the second assessment, and all QOL questionnaires had over 70% of their items answered. This indicates that the study was well accepted, especially considering that lower levels of compliance were expected in elderly patients [29]. The large number of VLTBCS patients should be highlighted (48.1%, including 27.6% of patients with >16 years’ follow up). Levels of comorbidity were in line with expectations for the age of this population [ 11].

QOL scores in the global sample were high. Patients with relapse in the follow-up period and those with a new tumour were excluded. This may have influenced these high QOL scores. Low scores in sexual functioning and enjoyment were also observed in elderly LTBCS from our institution [20]. Limitations in sexual functioning and enjoyment were higher (20 points higher in both sexuality scales) than those found in LTBCS (5–20 years after diagnosis) who had been diagnosed in adolescence or young adulthood [30]. Differences may be related to the patients’ ages. The fact that limitations in key areas such as fatigue, pain, emotional functioning, family support and insomnia are slight (20-29 points) is positive since high long-lasting limitations in these areas could heavily affect the QOL of elderly LTBCS [31]. Also encouraging is the fact that future perspective and future worries showed only slight limitations since worry of recurrence is considered to play a key role in the QOL of breast cancer survivors [32]. It is important that cognitive functioning showed high scores. This is because the patients’ high age might have favoured lower functioning and because any cancer-related cognitive impairment (CRCI) is reported to potentially have a significant impact on the QOL of survivors [33].

QLQ-C30 scores were in line with >73 years early-stage breast cancer survivors (who had received treatment three years earlier). These patients showed higher emotional functioning (14 points, i.e., medium difference), an area in which our patients also showed rather high QOL [34]. QLQ-C30 scores were in line with VLTBCS (10 years after surgery) who had received radiotherapy (no CT) [3]. Medium differences appeared in favour of our study in dyspnoea (13 points) and insomnia (21 points) [24]. Patients had a lower mean age, and younger survivors tend to consider cancer as a bigger threat [3]. QLQ-C30 and QLQ-BR42 scores were in line with >68 years early-stage breast cancer survivors from our institution (who had finished treatment three years earlier) [35]. Scores in the breast module were higher (10-20 points) than those in a study of 70–99 year-old LTBCS (5-15 years) whose disease may have been active [11]. Scores in the QLQ-30 were in line with Spanish general population reference values, which indicates that the patients adapted well to their situation [28]. Similar comparisons with the general population were found in other studies: e.g., in >65 years LTBCS (10 years after diagnoses) [4] and in disease-free 70-99 years LTBCS (5–16 years after diagnosis) [9].

Comparisons between age-based groups showed differences in few QOL areas. These mainly involved lower QOL in the >75 years group. Like the differences in demographic and clinical variables (cohabitation and performance status), these may be related to the ageing process. Interestingly, the size of these QOL differences is not high (and are small in the QLQ-C30 areas), which indicates that the patients in the eldest group in our sample adapted well to their situation. In line with our study, Gao et al. [3] found lower physical functioning, cognitive functioning and global QOL in the >76 years group of LTBCS (10 years after treatment) and higher fatigue, insomnia and constipation than in the 65–75 years group. The better QOL in those living with a partner highlights the role of social support in elderly patients [36]. The few differences between patients who had and had not received radiotherapy were mainly found in areas not related to the treatment. These results are in line with studies that found that RT does not seem to compromise the overall QOL of elderly breast cancer patients [37]. The few differences between the two assessments confirm the high QOL scores and suggest the stability of the patients’ clinical status.

Related to determinants of the two QLQ-ELD14 areas, comorbidity has also been found to be a QOL determinant in elderly breast cancer survivors [39]. Comorbidity is more prevalent and is considered to have a more important role with ageing [2,40]. Elderly-specific QOL areas have been found to be frequent determinants in our study and to show QOL limitations, which suggests that specific supportive interventions are needed for these areas in this age group [41]. Other areas, such as cohabitation, support, emotional functioning, fatigue and pain, have also been found to be QOL determinants in LTBCS (5 years [40,42] and 10 years [3] after diagnosis).

Our study has several strengths that may add relevant information to previous studies. One of these is the patients’ age (mean 75.8 years) as more information in this age group could be useful. Another strength is the large number of QOL areas we assessed. In this study we administered a general cancer questionnaire and a breast cancer questionnaire as well as an elderly-specific questionnaire whereas other studies have assessed just general QOL areas. Note also that the patients came from a “real-world” setting: unlike clinical trials, studies in the real world tend to provide more QOL information because the proportion of patients who assess QOL is higher and more QOL areas are evaluated [43].^.Finally, the participation rate was very high, which limits the possibility of selection bias. However, this study also has limitations. For example, we did not have a baseline (pre-treatment) assessment with which to compare our results. Also, patients with relapse during follow-up were excluded, which may have favoured higher QOL scores (health survivors bias) [11]. Finally, the sample came from just one cultural area.

5. Conclusions

QOL in these elderly LTBCS was high in both the whole sample and the two age groups after controlling for years since beginning of treatment, which indicates that the patients offered a positive balance of their situation and expectations. These results are highly relevant owing to the sample’s high mean age and the fact that intense, long-lasting limitations that could permanently inhibit the patients’ ability to return to a pre-cancer lifestyle [44] were not found.

Our results support previous reports that, for most elderly early-stage breast cancer survivors, cancer-specific well-being and general emotional health do not change substantially after breast cancer diagnosis [29]. Our results can provide professionals and patients with more information about potential general and elderly-specific late effects and chronic long-term sequalae that could be treated in a comprehensive survivorship care setting [9,41,45].