Chemotherapy use near the end-of-life in patients with metastatic breast cancer

Although various options are available for the treatment of patients with metastatic breast cancer, chemotherapy is commonly used in later stages of the disease. Very few data however inform on the patterns of chemotherapy use during the final month of life, which may negatively impact quality of life and cause unnecessary toxicity. With this retrospective descriptive analysis, we analyze the types of chemotherapy used during the final month of life and the associated patient clinical and demographic characteristics in two European countries, Sweden, and Greece. Being one of the few studies that have surveyed chemotherapy use in breast cancer patients at this stage of the disease, it provides important background for mitigation strategies.

Metastatic breast cancer (MBC) is with few exceptions an incurable disease, where treatment aims at prolonging survival and improving quality of life by relieving symptoms. Several treatment approaches are available for MBC, including endocrine manipulation, radiotherapy, and chemotherapy. Advances in the understanding of BC biology in recent decades and the deciphering of the plasticity of the clonal architecture of BC under the natural selection pressure applied by anticancer therapy [1‐3] have led to the development and approval of several targeted agents, such as anti-HER2 (human epidermal growth factor receptor 2) therapy [4], mTOR- (mechanistic target of rapamycin) and CDK4/6 (cyclin-dependent kinase 4/6) inhibitors [5], as well as immunotherapy through checkpoint inhibition [6].

Despite these improvements, the vast majority of MBC patients will receive non-targeted cytotoxic therapy during the disease trajectory, most often in the form of continuous mono-chemotherapy until disease progression, at which point treatment is changed to a non-cross resistant agent. There is limited randomized evidence on the treatment of this typically heavily pretreated population, with eribulin being the only single agent demonstrating an overall survival (OS) benefit in patients that had received at least two chemotherapy lines for advanced disease [7]. It is also important to note that the heavily pretreated MBC population likely differs from that of patients enrolled in clinical trials, the latter generally being fit, younger and without significant comorbidities. This poses a difficulty in translating results from clinical trials to the general MBC population where the observed median OS is shorter [8], and clearly supports a need to gain knowledge about real-life use of chemotherapy in late-stage MBC. This lack of evidence and consistent clinical guidelines may expose patients to unnecessarily aggressive and ultimately futile treatment, which can negatively impact the patients’ quality of life [9].

To date, the use of chemotherapy in MBC patients near the end-of-life has not been extensively studied, and there are no well-described predictors for the administration of futile chemotherapy in this patient group. Furthermore, the lack of consistent guidelines for this population may result in treatment differences between cancer centers, with variations being possibly susceptible to cultural differences between different countries. In this study, we explore the real-life use of chemotherapy near the end-of-life and the characteristics of MBC patients treated at this stage in areas of Sweden (Stockholm) and Greece (Athens and Heraklion), in order to highlight possible factors that could help in decision-making and identify cultural differences between Northern and Southern Europe that may drive palliative care.

Data on treatment of MBC near the end-of-life are scarce, making patient management more intuition-based rather than data-driven. The use of “futile” chemotherapy may lead to impaired quality of life for patients in the last stage of MBC, underscoring the need to recognize how widespread this phenomenon is in an effort to eventually mitigate it. In this retrospective cohort study, we describe the real-world use of chemotherapy near the end-of-life in areas of Sweden and Greece. The Swedish cohort was identified from a population-based registry with adequate completeness, providing a representative sample of the MBC population. Data on the Greek cohort were extracted from hospital charts, which represents a selected population fit enough to receive active cancer treatment. Consequently, results from the two cohorts are not comparable with each other, especially considering that approximately a fifth of the Swedish patients never received any active oncologic treatment but were nevertheless included in the registry.

The median OS in the Swedish cohort was 16.96 months, which represents a marginal improvement when compared to the results from previous decades in the same geographic area [8]. This slow improvement reflects on the one hand the therapeutic advances, but also the increasing use of adjuvant polychemotherapy and endocrine therapy, which in turn greatly decreases the number of metastatic relapses but enriches them with treatment-resistant patients [12]. However, OS was clearly worse as compared to results reported in modern clinical trials, corresponding to the fact that participants in population-based studies are generally less fit, older and with more comorbidities than that of clinical trials [13], but also that many of the patients included in the survival analysis did not receive any treatment. In contrast, the median OS in the Greek cohort was in line with outcomes reported from contemporary trials, which is likely attributed to the selected population which was considered fit enough to receive anticancer treatment.

During the last month of life, 23.2% of patients in the Swedish cohort received at least one chemotherapy treatment. Expectedly, younger patients and those in better condition—using albumin as a surrogate [11]—were more often treated during this late phase of the disease. This is consistent with observations regarding tolerance of chemotherapy among elderly patients due to age-related developments in human physiology including decline in liver and kidney function and marrow reserves, which affects clinical decision-making [14‐16]. Other factors which may have influenced this under-representation of elderly patients, such as refusal of anticancer treatment—which is far more common among older and more ill patients [17] and delays in seeking medical assistance, in referral to specialized oncology care or in treatment initiation cannot of course be excluded. An even higher usage of chemotherapy in general and introduction of a new regimen in particular was noted in the Greek cohort. These rates are considerably higher compared to a previous study by Wallington et al. where 30-day mortality following the last administration of chemotherapy in patients with MBC was 7% [18]. Although cross-trial comparisons can be misleading since differences in methodologies, data sources and study endpoints may account for the observed variations in chemotherapy use, cross-cultural differences in both physician and patient perceptions and expectations concerning the disease, treatment and prognosis possibly affect patient management. In addition, the choice of chemotherapeutics used during the last of month of life exhibited significant variation, a fact that is consistent with, and caused by, the lack of evidence-based recommendations regarding heavily pretreated patients. In the Swedish cohort, chemotherapy was more commonly given as monotherapy administered orally. This may be the preferable choice for late-stage patients wanting to avoid regular contacts with the hospital for intravenous therapy. Such a preference was not noted in the Greek cohort though, further highlighting the different approaches employed.

Previous studies have reported on the frequency of aggressive interventions, including chemotherapy administration, in patients with advanced malignancies with varying results. Although two studies from Canada showed that only 6.3% and 22.4% of patients received chemotherapy or experienced aggressive medical interventions (such as emergency department visits or intensive care admissions) respectively, both endpoints were observed more frequently among patients with MBC [19, 20]. The latter was confirmed by another study reporting high rates of aggressive end-of-life care among patients with MBC [21]. Interestingly, chemotherapy use at the last month of life in patients with multiple tumor types was reported in a fourth of all patients in two studies from Sweden and Italy, which is similar to the Swedish cohort in the present study [22, 23], although the number of patients with MBC was too low to allow for meaningful comparisons.

Following the surveying of chemotherapy use at the end-of-life, the next step is efforts aiming to limit it, considering the clear potential benefit: overtreating patients with palliative chemotherapy does not improve survival [24], exposes patients to unnecessary toxicity, negatively impacts their quality of life and has been shown to be associated with aggressive interventions near the end-of-life [25]. Importantly, mutually reinforcing attitudes by both patients and physicians may further aggravate this phenomenon, indicating that physician-centered interventions might not adequately remedy this phenomenon [26]. As a result, mitigating efforts should also focus on patient education, frank communication and frequent discussions on treatment goals, the disease course and eventual prognosis, although the latter is notoriously hard to predict [27].

Besides selection bias, inherent in retrospective studies, several weaknesses of this study should be acknowledged. Considerable data missingness and differences in collected variables between the two cohorts do not permit any meaningful comparisons of possible factors that affect the decision to continue with chemotherapy at this stage of the disease. Furthermore, data regarding the cause of death were not collected. Both toxic death due to palliative chemotherapy for MBC, although a relatively rare event [18], and competing non-cancer related causes of death may have affected the results so that the true rates of “futile” chemotherapy administration may be lower than those presented. Finally, factors such as receptor status at the metastatic setting which has been shown to differ compared to the primary tumor [28], socioeconomic status, healthcare setting, treating physician’s experience and others that could affect chemotherapy use were not evaluated.

In conclusion, use of chemotherapy near the end-of-life was fairly common, especially among younger patients, while a wide variety of agents was shown to be used in this setting. Prospective assessment of mitigation strategies is warranted.