Due to new developments regarding systemic anticancer treatment (SACT), patients with metastasized and inoperable or locally advanced and inoperable solid tumors may have a survival time of several years [1]. However, the course of the disease is hard to predict and can sometimes expand to several decades. The goal of antitumor therapy in advanced, metastatic cancer is usually not cure, but symptom control and prolongation of survival. A considerable percentage (20–50%) of patients diagnosed with advanced cancer undergo antitumor therapy within 30 days prior to death [1]. However, increasing evidence suggests that this treatment approach is neither effective nor useful in most cases [2, 3].

The American Society of Clinical Oncology (ASCO) recommends that end-of-life (EOL) anti-cancer therapy should only be given according to recommended guidelines [4••]. These guidelines recommend stopping anticancer therapy in patients with advanced solid tumors if the performance status (PS) is ECOG (Eastern Cooperative Oncology Group) 3–4, no benefit from prior evidence-based therapies has been observed, the patients are ineligible for clinical trials, or the chance of success is expected to be very low [4••]. Many studies have shown that patients who meet the above criteria do not benefit from antitumor therapy [4••]. Exceptions are patients who, for example, have special disease characteristics (e.g., certain mutations) that predict response to specific therapeutic measures. In any case, supportive and palliative measures should be implemented in patients with advanced metastatic cancer [5]. From the early 1980s, there have been recommendations not to give chemotherapy to patients with a poor ECOG PS (from ECOG 3). In most cases, a poor PS indicates poorer response, shorter survival, and increased toxicity due to chemotherapy [6, 7]. Moreover, there is little evidence of treatment success in patients with poor PS, since many studies reviewing new drugs usually only test patients with a good PS [4••].

In most patients with solid metastatic tumors, the chance of treatment success after unsuccessful third-line therapy is very low, but the probability of toxicity is high. Therefore, ASCO guidelines do not recommend further antitumor therapy after an unsuccessful third line of treatment [4••].

The first major report on mortality within 30 days of chemotherapy was conducted by the National Confidential Enquiry into Patient Outcome and Death (NCEPOD) in the UK [8••]. The reported mortality rate within 30 days of SACT was 2% and has been used as the historical benchmark [8••]. Following this report, numerous centers have published data on SACT given within 30 days of death [9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27]. Comparisons between studies are difficult due to differences in the types of tumors included/excluded, treatment with curative/palliative intent, and treatment modality. Furthermore, the studies reported different outcome measures, with most reporting the number of deaths within 30 days of treatment as a proportion of all patients who received treatment and, less frequently, as a proportion of all deaths. After the NCEPOD report, the Christie Cancer Centre in the UK implemented the recommendation to review all deaths within 30 days of SACT at morbidity and mortality meetings and reassess progress through an audit process [8••]. Over a 4-year period, this practice did not reduce the rate of deaths within 30 days of SACT and had a minor, statistically insignificant reduction in the rate of treatment-related deaths [21]. In contrast, Wilson et al. reported two audits conducted at Auckland Hospital 6 years apart [18]. Mortality within 30 days of chemotherapy treatment decreased slightly, with rates of 2.8% in 2009 and 2.2% in 2015. They proposed a series of clinical interventions to inform this improvement implementation plan [18]. An Australian study examined the use of aggressive treatments, including SACT, in the last 30 days of life while considering new treatments like immune checkpoint inhibitors. The authors proposed a quality improvement plan that has been shown to increase the use of palliative care referrals and reduce the use of SACT [28].

Despite the limited evidence for the efficacy of therapy beyond third line, this is not infrequently administered. At a large university facility in Michigan, as many as 50% of all patients with solid tumors received chemotherapy in the last 2 weeks of life [29].

A study published in 2014 was able to show that patients with metastatic cancer and poor general condition who were still receiving chemotherapy in the last months of life were much more likely to undergo intensive care measures (ventilation, resuscitation, or both) than EOL patients no longer receiving chemotherapy [30]. In addition, chemotherapy patients were transferred to a hospice very late or died less frequently at home [30]. Both the intensive medical measures and late transfer to hospice are associated with a poorer QOL at EOL [31]. It is not appropriate to draw a general conclusion that chemotherapy is futile for seriously ill patients in advanced stages. Nevertheless, it is crucial to consider that purely palliative symptom control in the sense of “best supportive care” can be a favorable alternative for numerous patients.

There have been significant recent developments in the treatment of cancer with many new therapy options demonstrating clinical evidence for improved survival and QOL. SACT includes cytotoxic chemotherapy, endocrine or hormonal agents, targeted or biologic agents, and immune checkpoint inhibitors. Non-chemotherapy treatments are often associated with simpler routes of administration, less but not negligible adverse effect profiles, and the potential of profound and durable clinical responses. This has made the decision-making process for commencing, continuing, and ceasing SACT more complex and requires a careful consideration of key factors, specifically disease biology, patient and family expectations, and clinician bias.

The guidelines of ASCO explicitly mention “cancer-directed therapies,” which include anti-tumor treatments, and not exclusively chemotherapy [4••]. However, the recommendations are derived almost exclusively from chemotherapy studies. Compared to conventional chemotherapy, fewer studies have investigated the appropriate time to discontinue treatment with new, so-called targeted drugs, such as tyrosine kinase inhibitors or antibodies, based on the above criteria. Furthermore, new oral therapies are initiated more quickly and terminated later.

A systematic review of the literature evaluated a total of 49 studies relating to anticancer therapy at the EOL [32•]. None of the studies provided criteria for stopping oral anticancer therapy. In most cases, therapy was discontinued when intolerable side effects occurred. However, the personal experience of the treating physicians with the respective treatments also influenced the continuation of the therapy. In some studies, the decision to terminate the therapy was made by the patient. In other studies, the decision to continue the therapy was made by the physicians based on the success of the therapy. The literature review demonstrated an imperative necessity for precise guidelines regarding the termination of anti-cancer therapy, and current research is inadequate regarding the new generation of drugs such as tyrosine kinase inhibitors (TKIs) [32•]. Studies to determine criteria for the duration of therapy are necessary for these new therapeutic measures, especially oral antitumor therapies [33].

A recent comprehensive literature search was conducted to find relevant publications on chemotherapy cessation [34•]. The authors retrieved a total of 2700 records, and 141 were identified as eligible for inclusion in their review. Staples et al. found that palliative chemotherapy did not achieve the goal of tumor-related symptom reduction for patients who had experienced progressive disease with more than two prior lines of chemotherapy [34•].

In our own review, we identified 37 studies as relevant for a narrative review (Fig. 1; see supplement). Table 1 includes studies which evaluate the impact of SACT on cancer patients receiving therapy in the last weeks of life (Table 1). Table 2 includes studies which report on predictive factors that are associated with the receipt of SACT at the EOL of cancer patients (Table 2).

Flow chart of literature research: The data that support the findings of this review article are available upon request. See supplement

Bao et al. investigated the association between chemotherapy use in patients with stage IV pancreatic cancer and healthcare use, as well as Medicare and out-of-pocket costs in the last 30 days of life [35]. The study concluded that chemotherapy use among older patients diagnosed with metastatic pancreatic cancer leads to increased healthcare utilization and higher patient out-of-pocket costs near death [35]. According to a retrospective study that analyzed the use of palliative SACT for patients with advanced non-small cell lung cancer, 22% of patients received this therapy within 30 days of death. This was associated with reduced access to palliative care, higher rates of in-hospital death, a decreased use of voluntary assisted dying and palliative sedation, and a shorter median overall survival of 4.0 months compared to 9.0 months for patients who did not receive the therapy close to death [36].

Eastern Cooperative Oncology Group Performance Status (ECOG-PS) and Glasgow Prognostic Score were significant prognostic factors regarding the effectiveness of aggressive EOL chemotherapy in a study performed by Hiramoto et al. [13]. Furthermore, patients who died within 30 days after chemotherapy had a higher rate of nausea, vomiting, and hydration [13].

An Irish study found that a significant percentage of patients received treatment in the last 12, 4, and 2 weeks of life, which was associated with more hospital admissions, procedures, and in-hospital deaths [37]. Patients who received chemotherapy were also referred to specialist palliative care services at a later stage than those who did not [37].

A study by Mathew et al. reviewed medical charts of metastatic breast cancer patients who received immune checkpoint inhibitor (ICI) treatment within 30 days of death [38]. Out of 97 patients evaluated over a 4-year period, 40% received only one dose of ICI during that period. Over 50% had a poor ECOG-PS, and more than 60% were hospitalized. Additionally, 65% visited the emergency department, 20% required intensive care unit admission, and 25% died in the hospital [38]. The study’s findings add to the current knowledge of using ICI treatment for advanced cancer patients in their EOL stage [38].

A study analyzing the use of palliative chemotherapy in advanced non-hematological cancer patients revealed that 29.5% of all patients received chemotherapy in the last month before death. There was a strong association between last month before death palliative chemotherapy and place of death (critical care vs. regular ward) as well as mode of admission (emergency room vs. outpatient department vs. direct transferred from other health care facilities) [39].

In a Swedish study, 23% of patients received chemotherapy during their last month of life, leading to shorter survival time, frequent hospital admissions, less documented decisions to cease treatment, and less deaths at home [29].

Among 437 patients with gastrointestinal cancer, 67% received CT within 3 months of death and 28% within 1 month. Patients receiving CT within 1 month had poorer OS and received less palliative care [40].

A study by Saito et al. examined the impact of continuing chemotherapy on survival and EOL care for 7879 Medicare-enrolled patients aged 65 or older who survived at least 3 months after diagnosis of advanced non-small cell lung cancer [41]. Three different statistical approaches showed no additional survival benefit from continuing chemotherapy within 14 days of death but was associated with a decreased likelihood of receiving hospice care [41].

An investigation of 604 patients of whom 50% received chemotherapy in the last 3 months revealed that these patients had a higher rate of unplanned hospital admissions and emergency room visits [42].

A retrospective study investigating the prevalence and outcomes of EOL chemotherapy in advanced solid cancer patients revealed that of the 3350 decedents included, 5.3%, 11.6%, and 25.0% received EOL chemotherapy within the last 2 weeks, 1 month, and 2 months of life respectively [43]. Receiving EOL chemotherapy was associated with inferior OS, more intensive treatments, and hospital death. However, subgroup analysis suggested that receiving oral agents was associated with better outcomes [43].

A study analyzing the use of targeted therapies in the last month of life for non-small-cell lung cancer patients found that 21.3% of patients received targeted therapy within 30 days of death [44]. Younger patients, patients with adenocarcinoma, longer survival post-diagnosis, and patients treated by respiratory physicians or oncologists, were more likely to receive targeted therapy [44]. According to the results of this study, targeted therapy at EOL should be a quality-of-care indicator [44].

Data of a large Taiwanese trial including 49,920 patients who underwent palliative chemotherapy were analyzed. Palliative chemotherapy was shown to be associated with more aggressive EOL care, including emergency room visits, intensive care unit admissions, endotracheal intubation, and less hospice service at the EOL [45].

Another study analyzed EOL chemotherapy in small cell lung cancer patients and documented an association with shorter survival and more aggressive care [46]. Younger patients were more likely to receive EOL chemotherapy [46].

Altogether, these studies identified several different factors that were related to the impact of SACT on health and healthcare-related outcomes, including an increase in hospital admissions, emergency department visits, in-hospital deaths, and fewer days in hospice care. Some studies demonstrated a decreased use of medical aid in dying and palliative sedation, and a shorter median OS. Patients were referred later to specialist palliative care services. Overall, palliative chemotherapy was associated with more aggressive EOL care, including emergency room visits, intensive care unit admissions, and endotracheal intubation. Some of the studies found that younger patients were more likely to receive EOL chemotherapy.

According to Allen et al., 9–12% of lung and pancreatic cancer patients received EOL chemotherapy, with a higher prevalence in males with advanced metastatic cancer [47]. This was associated with higher rates of deaths within an acute care facility [47].

In analyzing how often patients with metastatic colorectal cancer started a new chemotherapy regimen within 90 days of death, a Canadian study showed that of the 511 patients who received chemotherapy, 25.8% started chemotherapy near EOL [48]. Factors such as comorbidity index score and oncology PS predicted for initiation of chemotherapy near EOL [48].

A recent study showed that 97 patients had ICI treatment in their EOL stage over 4 years. Forty percent received a single dose within the last 30 days before death [49]. Over 50% had an ECOG-PS of 2 or higher, and 17% had a PS of 3. Hospitalization rate was over 60%, and 25% died in hospital [49].

A large trial analyzed the administration of EOL chemotherapy and targeted therapy in patients with breast, lung, colorectal, or prostate cancer and time of death between 2007 and 2013 [50]. The administration of chemotherapy within 14 days of EOL declined, while it increased within 4 to 6 months of EOL. Targeted therapy use remained stable. Physician-level variation accounted for 5.19% of the variation in 14-day EOL chemotherapy [50]. This study suggests that national benchmarking can effectively reduce EOL chemotherapy use [50].

An Italian study investigated the use of anticancer drugs and palliative care services in EOL care for 55,625 cancer patients who died between 2017 and 2020 [51•]. The study found that the use of anticancer drugs was inversely associated with receiving palliative care services in the last month of life [51•]. The authors suggest that the implementation of adequate prognostic tools is key to improve the appropriateness of EOL care [51•].

A Turkish study examined the use of palliative chemotherapy near the EOL [52]. Of the 373 patients who died from stage IV solid tumors between 2010 and 2011, 23.9% received chemotherapy in the last month of life, and 10.5% received chemotherapy in the last 14 days [52]. Age, recent diagnosis, and PS influenced the likelihood of receiving chemotherapy in the last month of life [52].

An earlier study analyzed 255 patients who received palliative chemotherapy and died between 2002 and 2006 [53]. Lack of information about palliative care units, presence of symptoms, and age ≤ 45 years were predictive factors for receiving palliative chemotherapy in this study [53].

A retrospective analysis of 300 patients from 2011 to 2016 identified ECOG-PS and GPS as significant prognostic factors at the last administration of chemotherapy [13]. Within 14 and 30 days, 5.3% and 16.7%, respectively, died. Those with PS 2–4 and GPS 2 had median survival of 38 days. A higher prevalence of nausea, vomiting, and hydration was found in those who died within 30 days [13].

In a study by Hui et al., predictive factors for targeted agents were younger age and hematologic cancers, whereas predictive factors for chemotherapy were hematologic cancers and survival < 6 months between diagnosis and death [54].

Out of 5759 patients who died of cancer between 2013 and 2020 in Japan, 4.8% received anticancer therapy within 14 days of death. Patients aged 60 and above had a higher probability of receiving anticancer therapy near EOL [55].

In a retrospective study of adult cancer patients who died in a hospital, the authors investigated factors associated with a shorter time from last chemotherapy to death (TLCD) [56]. Patients with better PS and under a medical oncology service had a shorter TLCD, while those with no palliative care involvement had a higher risk of dying from treatment-related complications and receiving more aggressive EOL care [56]. Palliative care involvement was associated with longer TLCD [56].

In a Swedish study, age and albumin levels were associated with chemotherapy use. This study concluded that EOL chemotherapy may harm patient QOL [57].

A Taiwanese study revealed that factors such as male gender, younger age, lower comorbidity levels, certain cancer types, and being cared for by a medical oncologist or in a teaching hospital were associated with higher rates of continued chemotherapy. Regional healthcare resources did not have an impact on chemotherapy continuation [58].

A study assessing care quality for metastatic colorectal cancer patients revealed that 9.8% started new treatment in the last 30 days of life. Age and no advanced treatment were linked to overuse of SACT [14]. Mathew et al. found that younger age at diagnosis of metastatic breast cancer and the number of metastatic organ systems involved were related to application of SACT at the EOL [38].

A retrospective cohort study found that male gender, younger age, breast cancer, melanoma, prostate cancer, and cancer diagnosis within the last 6 months of life were associated with SACT in the last 30 days of life [59].

An interesting finding was described by Mieras et al. who found that a predictive factor for tyrosine kinase inhibitors (TKIs) for NSCLC was the age of the prescribing oncologist (< 51 years) [60].

Younger age and high level of education was found to be a predictive factor for the prescription of SACT in the last 30 days of life [61].

In Austria, there were slight variations in cancer treatments during hospital stays within 30 days before death, with the highest proportion of patients receiving chemotherapy in the state of Salzburg (10.4%) and the lowest in Vorarlberg (5.3%) and Burgenland (5.4%) [62••].

A very large retrospective register study analyzed data from 279,846 patients with solid metastatic cancers and found that 19.5% received chemotherapy in the last month of life [63]. Factors associated with this included male gender, younger age, lower comorbidities, and having chemosensitive tumors or cancers of the skin, head and neck, female genital organs, or respiratory system. Patients treated at comprehensive cancer centers, private hospitals, or rehabilitation care facilities, and those receiving a higher annual volume of chemotherapy, were also more likely to receive chemotherapy [63]. The lack of a palliative care unit was also associated with receiving chemotherapy in the last 30 days of life [63].

Sheng et al. found that ≥ 56 years of age, cancer type, metastatic disease, high education level, two or more lines of treatment, and patients of general hospitals were associated with SACT at the EOL [43].

Several factors were found to be predictive of receiving chemotherapy in the last 30 days of life in a large retrospective analysis, including younger age, late disease stage, adenocarcinomas, longer survival after diagnosis, later year of death, having a treating pulmonologist or oncologist, younger treating physician, higher lung cancer case volumes of treating physicians, and being treated in district or non-public hospitals, teaching hospitals, or hospitals with higher lung cancer case volumes [44].

A retrospective analysis identified younger age, female gender, marriage, high income level, small hospital size, lack of hospice facility, and responsive cancer types as predictive factors to receive SACT in the last month of life [64]. Male gender was a predictive factor for receiving SACT in a study by Zdenkowski et al. [24].

A retrospective analysis of medical records of 605 metastatic cancer decedents of which 98 patients were treated with palliative chemotherapy in the last month of life found an association of age > 50 years with application of SACT in the last 30 days of life [65].

Altogether, the selected studies found various predictive factors that were linked to enhanced use of SACT at the EOL including sex, younger age, lower comorbidity levels, higher level of education of patients, and even age of the prescribing oncologist (< 51 years). Other factors such as poor PS, high ECOG, or claim of palliative care services were correlated to reduced administration of SACT at the EOL.