Managing the risk of toxicity in the treatment of elderly patients with soft tissue sarcomas

ABSTRACT

Introduction

Nearly half of soft tissue sarcomas (STS) occur after the age of 65 years. Treating these patients is a complex issue in the absence of specific guidelines.

Areas covered

This is a narrative review that summarizes current data on the efficacy and the safety of different treatment strategies in this subpopulation.

Expert opinion

Age per se should not be a limiting factor to treatment. Surgery remains the treatment of choice offering the only chance of cure. The potential for benefit from adjuvant therapies must be discussed in the context of expected treatment-related toxicities and impairment of quality of life. Efficacy of systemic treatment in advanced disease did not differ from that in younger patients. However, safety must be considered when selecting treatments. Managing the risk of toxicity requires an assessment of vulnerabilities with validated tools. The Comprehensive geriatric assessment has become increasingly accepted but need to be validated in STS patients. Frailty should not exclude patients from potentially life-saving therapy. The correction of reversible conditions and active supportive care may make the treatment safer. Future studies are warranted to define better the patterns, benefits, risks of existing treatments. New options remain to be identified to reduce toxicity.

KEYWORDS:

• soft tissue sarcomas
• elderly
• toxicity
• efficacy

1. Introduction

Soft tissue sarcomas (STS) are rare malignancies that account for less than 1% of all adult cancers. It is a heterogeneous group of neoplasms gathering over 150 histological subtypes. They can arise at any anatomic site [1,2] making up a large variety of combinations of histology and primary site that vary tremendously in terms of survival and management.

Nearly half of STS occur after the age of 65 years [3]. The incidence strongly increases with age with the highest rates documented in people over 75 years [4,5]. This proportion is expected to rise significantly in the future as the general population ages.

The treatment of older sarcoma patients often represents a significant challenge in daily clinical practice. The physiological age-related changes, comorbidities, compliance to treatment, and the presence of polypharmacy make both treatment activity and toxicity less predictable. A careful assessment of the benefit/risk ratio is necessary and should rely on specific evidence to guide treatment. However, elderly patients are usually excluded or poorly represented in clinical trials, therefore evidence-based decision making, in this population, is usually difficult. Guidelines in this setting of patients exist and are relatively helpful, but they usually do not cover cancer-type specific issues [6,7].

This manuscript aims to capture and summarize current data on the efficacy and the safety of different treatment strategies including surgery, radiotherapy, chemotherapy, and targeted therapy in older patients with STS, and address how to approach the risk of treatment related toxicities in this subpopulation given the availability of limited data.

There is no universally accepted age to define the elderly. In this review, we agreed to consider all studies focusing on or including patients ≥ 65 years old using the classic Medicare definition of the elderly.

2. Epidemiology, clinical features

STS in the elderly exhibit unique demographic and clinical characteristics as compared to those of younger age. Interestingly, differences are also observed across elder subgroups; young-old, old-old, and oldest-old [5]. These variations reflect the diversity of the elderly population with STS and underline the need of an age-stratified approach in the management of older individuals with STS.

Patients over 65 years account for approximately 40% to 50% of all STS cases [3,8]. The incidence increases with age, regardless of sex and race. In the US, according to the Surveillance Epidemiology, and End results (SEER) database, the incidence of STS in 2017 was about 9.6 per 100,000 between the ages of 65 and 74 years, and 13.6 per 100,000 after the age of 75 years. Under the age of 65 years the incidence, is lower accounting for less than 4 per 100,000 person per year. The occurrence of STS steadily increases until the age of 50 years. After the age of 50 the incidence increases much more dramatically. The incidence is higher in older men compared to older women; 11.7 per 100,000 and 7.8 per 100,000, respectively [4] . There is a slight difference across racial groups, with lower incidence in Asians and African Americans [4]. Overall, incidence rates are comparable across the world [9–11]. European registries show similar incidence rates with the highest incidence after 65 years old, 13.1 per 100,000 compared with 2.3 per 100,000 in those aged <65 years [9]. In Asia, data from the Japanese nationwide registry for bone and soft tissue tumors for the period of 2006–2012 show a dramatic increase in the proportion of elderly patients with STS exceeding 50% with a peak around the seventh decade [10]. In China, a population cancer registry shows similar trend, with 42% of STS in individuals aged ≥65 years old and an age-specific incidence at 75–79 years old in women and 80–84 years in men [11].

Mortality from STS in those aged 65 years and over accounts for 53% of all deaths from STS in the US (SEER, 2020), with a median age at death being 66 years old [4]. The mortality rate is slightly higher among older men. It dramatically rises for both sexes after the age of 50 and nearly triples between 65 and after 85 years from 3.38 to 8.81 per 100,000. The causes of this increased mortality in advanced ages remains unknown. It is not clear whether the patients die due to comorbidities or due to more aggressive disease and/or suboptimal treatment.

Although the etiology of STS in higher age is still unclear, it is likely that different and specific risk factors are involved [5]. For the most part, STS in the elderly are sporadic and are attributed to environmental factors [5,12]. Inherited syndromes extending into advanced age, including Von Recklinghausen disease and Gardner’s syndrome are extremely rare [12,13], with almost no genetic predisposition condition over 90 years old [5].

Compared with younger patients, elderly tend to have larger and higher-grade tumors at diagnosis. The limb is the most common primary site followed by trunk and intraabdominal sites [5,14]. Sarcomas from head and neck are significantly overrepresented in patients over 90 years [5]. The tumors depth varies between the elderly subgroups. They are more frequently superficial in the oldest of old group, and deep in the other subgroups [5]. In the elderly population, there are almost exclusively sarcomas with complex genomics. Undifferentiated pleomorphic sarcomas (UPS), leiomyosarcomas (LMS), myxofibrosarcomas (MFS), dedifferentiated liposarcomas (DDLPS), and angiosarcomas are the most represented histologies [5,15]. Sarcomas with translocations are rare accounting for less than 3%[5]. The majority of elderly patients present with localized disease at diagnosis [14]. Elderly patients are less frequently metastatic at diagnosis especially the oldest-old patients [5]. However approximately half of patients with initially localized disease will develop distant metastases [3,16].

Overall, older patients with STS have higher recurrence rates [16] and shorter survival, even when adjusting for disease stage [3–5,9,15]. In the RARECARE database, a large population-based series of data collected from 89 European cancer registries, persons aged 65 and over have the lowest survival rates for most sarcoma categories [9]. A similar pattern has been observed in the Japanese registry [10] with lower survival in patients aged ≥75 years old as compared to younger patients (HR = 1.36, (95% CI [02–1.82]). Part of the reason older patients do worse is that tumors in elderly exhibit a more aggressive feature. Undertreatment may also contribute to the poorer prognosis [15]. However, in various analyses, older age has been reported as an independent prognostic factor for patient with STS [3,15].

3. Treatment outcomes: toxicity and efficacy

There are a lot of uncertainties about the efficacy and safety of treatments in elderly patients with STS, since they are substantially underrepresented in clinical trials, especially those with comorbidity or frailty. For example, only 12% of patients enrolled in European Organization for Research and Treatment of Cancer (EORTC) clinical trials were aged >65 years [3]. Subgroup analyses inherently suffer from selection bias, and even when the results of several trials are pooled, data may be insufficient to provide clear conclusions.

The beneficial results achieved with standard treatment in the general population with STS cannot be automatically extrapolated to the elderly, as important differences exist between older and younger patients.

There are limited prospective data. Most data come from retrospective studies. These studies used different age cutoffs to define elderly (>65 years, >75 years, >85 years, or >90 years) which complicate the interpretation of the results, especially since additional clinical and pathological variables interact with patient outcomes across different ages.

3.1. Localized STS

3.1.1. Surgery

Surgery is the primary treatment of choice for patients with localized STS. Studies assessing the surgical management of STS in elderly patients have demonstrated that surgical resection of STS in the elderly can be safely performed and remain the key factor for a superior outcome [17]. Curative surgery has been significantly associated with improved survival among elderly STS patients as compared with those who did not undergo surgery. This correlation is strengthened in the setting of R0 resection [17–19]. However, it has been reported that among patients undergoing surgery, elderly patients have significantly lower incidence of R0 resection [18].

Published results on morbidity and postoperative mortality are conflicting. While Gingrich et al. reported a 3-fold higher 90-day mortality in elderly patients for a low-risk surgical procedure such as those involving extremity STS [18], Lahat and colleagues at the MD Anderson Cancer Center found no significant difference between age groups for postoperative complications with 30% of patients experiencing at least 1 complication and a postoperative mortality rate < 1%, despite including 55% of patients with higher comorbidities with an American Society of Anesthesiologists (ASA) score of 3 or greater [17].

The analysis of two database studies from the SEER registry and the Netherlands cancer registry found a higher sarcoma-specific mortality following surgical resection in patients aged 65 years and over with extremity STS [15,20].

Surgery for retroperitoneal sarcoma in elderly patients is associated with higher rates of morbidity and mortality compared to younger patients. Sourrouille et al. reported a mortality rate of 8% after retroperitoneal sarcoma surgery [21]. Nevertheless, this risk remains acceptable when weighed against the oncological benefit. The main factor influencing surgical morbidity and mortality remains the presence of comorbidities [21].

Despite the improved survival outcome with surgery, elderly patients with STS are less likely to undergo surgery compared with younger patients. In a large national database analysis of 33,859 patients with non-metastatic extremity STS the percentage of patients who did not undergo surgery was significantly higher in older patients compared with younger patients (11.7% versus 6.2%; p < 0.0001) [18]. Multivariate analysis found significant predictors of the likelihood of receiving surgery including younger age, fewer comorbidities, smaller tumor size, treatment at an academic center. Interestingly, the main reported reason for the decision of no surgery was coded as ‘not part of the planned first-course treatment’ (69.2% of cases) meaning that multiple options are considered in upfront setting (chemotherapy or radiation in 56% cases, or no treatment), other reasons were surgery contraindications (12.4%), or surgery refusal (7.8%) [18].

There are no alternatives with curative intent to surgery. Radiation therapy for patients with localized STS who are unable to undergo or refuse surgery, can be used with lower survival results [18,22].

3.1.2. Adjuvant therapy

The choice of adjuvant treatment is particularly difficult. In this setting, it is important to carefully balance the benefits of treatment on lowering the risk of recurrence with the natural life expectancy and the risk of treatment-related toxicities.

The role of neo or adjuvant chemotherapy for patient with STS remains uncertain and controversial. In keeping with the international guidelines which recommend neo/adjuvant chemotherapy as an option for high-risk patients with extremity and truncal STS [23], the potential for benefit from adjuvant chemotherapy in elderly patients must be discussed in the context of expected treatment-related toxicities and impairment of quality of life. The combination of doxorubicin and ifosfamide typically used in the adjuvant setting may not be feasible in elderly patients post resection, due to the potential toxicities and the lower compliance to treatment. Retrospective series and population-based analyses reported a lower use of perioperative chemotherapy in elderly patients, but no significant difference in survival between groups [18]. This clinical finding remains unclear. In view of the uncertainty of survival benefit to justify the toxicity and the morbidity, extreme caution is indicated in treating older STS patients with neo/adjuvant chemotherapy.

Radiotherapy is the standard of care for large tumors, high grade, and deep lesions. It can be administered preoperatively or postoperatively with similar oncologic outcomes but different toxicity profile [23]. Retrospective data demonstrate that elderly patients were less likely to receive radiotherapy, particularly preoperatively [18,24].

Population based studies have reported a survival benefit with radiotherapy which is amplified in patients over the age of 65 years old [24]. Using the SEER database (n = 15,380), Yuen et al. found that patients treated with radiotherapy derived improvement in OS and disease specific survival (DSS) compared to patients who did not receive radiotherapy with a greater magnitude in patients ≥65 years old (n = 5170) [24]. These benefits were most notable after neo-adjuvant radiotherapy. The safety profile of radiotherapy in the elderly has been well described in other cancers. Overall, a greater frequency or severity is reported in older patients as compared to younger patients [25]. However, the concern for potential toxicity should be weighed against the possible benefit in survival.

3.2. Advanced and metastatic STS

3.2.1. Real-world treatment patterns

According to a population-based cohort (n = 4274), approximately 38% of elderly patients with STS were managed by best supportive care (BSC) only [26]. Patients who received active treatment had significantly prolonged overall survival (OS) compared to those managed with BSC (13.6 vs 2.8 months). This does not necessarily mean that chemotherapy is associated with better outcome as we cannot exclude an imbalance in some prognostic factors in this retrospective series. For instance, the improved outcome in patients who undergo chemotherapy may reflect the fact that those who are selected for systemic therapy are likely to have better performance status (PS) than those chosen for BSC only. In this study, the most common regimen used was the combination of gemcitabine and docetaxel (26.5%) followed by doxorubicin (18.8%). Only 11% received second line therapy, most commonly doxorubicin as single agent. The mean age in this study was 77.8 years (SD:7.3; 65–104) with a mean Charlson comorbidity index of 2.8 (SD:2.33). The majority of patients had significant comorbidities; hypertension (69.2%), congestive heart failure (18.3%) and myocardial infraction (6.5%), which may explain the reluctance to use anthracycline [26]. No data on toxicities have been reported.

Garbay and colleagues analyzed a large cluster of advanced-STS in elderly patients (n = 361) [16]. The median age in this series was 79 years (75–98). Over two third of patients had a Charlson comorbidity index score ≥10, and one third had PS ≥2. Among this study population 34% had not received active treatment. Factors associated with BSC option were age ≥80 years, PS ≥2, Charlson comorbidity score ≥10 and metastatic disease. Anthracycline-based regimens was most commonly use in first line (63%). 47% of patients received two or more lines of chemotherapy. Single agent protocols were the preferred (83%). For patients who received chemotherapy, the overall disease control rate was 58.5%. Median OS of patients managed with chemotherapy was significantly superior (10.9 months vs 5.5 months) for patients managed with BSC. Median progression-free survival (PFS) was similar irrespective of therapy: anthracycline-based regimen, non-anthracycline-based regimen, full conventional dose or reduced dose. However single agent regimen, and PS≥2 were associated with worse PFS (3.9 vs 8 months; p < 0.04, and 2.2 vs 4.5 months; p < 0.001, respectively). In terms of toxicity, 32% of patients experienced grade 3 or 4 hematological toxicities, and 16% of patients stopped the treatment because of toxicity. Patients stopping the treatment because of toxicity were more likely to have experienced grade 3 or 4 side effects. No cardiac toxicity was reported. Importantly, patients with PS≥2 were at risk of early death, suggesting the importance of PS in deciding whether to provide or not an active treatment.

In a single-institution study, Mir and colleagues performed a retrospective analysis of a cohort of elderly patients with advanced STS treated at the Institute Gustave Roussy. More than 100 patients aged 65 years or older with adequate performance status and organ function received standard therapy based on doxorubicin. The remaining 26 patients were offered an oral metronomic chemotherapy schedule combining cyclophosphamide (100 mg twice daily) and prednisolone (20 mg daily) from day 1 to day 7 of a 14 days cycle. These 26 patients aged 66–88 years, were either found unfit for doxorubicin-based chemotherapy because of renal function impairment, severe cardiac or respiratory co-morbidities and/or poor PS (n = 20) or were pretreated by doxorubicin (n = 6) [27]. The toxicity profile of metronomic oral cyclophosphamide was favorable. The main grade 3 or 4 toxicities were anemia (7.7%), neutropenia (0 patient), thrombocytopenia (7.7%), and lymphopenia (80.7%) but no opportunist infections occurred. The response rate was 26.9% and the median PFS was 6.8 months [27]. These promising results need to be confirmed by a larger prospective study.

3.2.2. Data from clinical trials

1. First line therapy:

An elderly cohort of 348 patients aged ≥ 65 years old was extracted from the of EORTC database of patients treated with first-line chemotherapy for advanced STS [3]. The participants were generally of excellent PS (0 or 1) which is not representative of an unselected elderly population. Overall, patients received doxorubicin (n = 126; 36%), doxorubicin + ifosfamide (n = 114; 33%), epirubicin (n = 43; 12%), trabectedin (n = 39; 11%) or ifosfamide (n = 26; 7%). Outcomes for PFS and OS in elderly were similar to those in younger patients [3]. No safety data have been reported.

The outcome and the safety of first-line anthracycline based chemotherapy in elderly patients with STS were investigated in a subgroup analysis of SARC021 trial [28]. This trial randomized patients to receive first line doxorubicin or doxorubicin plus evofosfamide. The primary analysis of this study showed no differences in OS or PFS between the two arms [29]. The older group aged ≥65 years old (n = 209) had a baseline PS of 0 or 1. Compared to the younger patients, there was no significant difference in both OS and PFS. However, adverse events (AEs) were significantly more common in older patients. Hematological toxicities (anemia, neutropenia and thrombocytopenia) occurred in 67% of patients vs 48% in the younger group. Sixty three percent of older patients experienced non-hematological AEs (fatigue, and diarrhea). There was no increase in cardiotoxicity in older patients (20% vs 35.8%, p = 0.06). Of note Dexrazoxane was administered to 22 and 12 patients in the younger and the older group respectively. Grade ≥3 AEs were significantly higher in older patients, especially, anemia and neutropenia. Older patients were more likely to discontinue therapy due to toxicity (30% vs 22%; p = 0.0001). Interestingly, quality of life analysis did not identify significant differences between younger and older group. PS 1 or 2 and age ≥ 65 were independently associated with greater hematological toxicities [28].

These data show that there is no significant difference in survival with anthracycline based regimens according to age. However, the increased rates of toxicity highlight the need for less toxic treatments and more effective supportive care.

The EPAZ study is a Phase II randomized non inferiority trial comparing pazopanib with doxorubicin as first-line treatment in elderly patients with metastatic or advanced STS [30]. Of note, the age threshold for defining elderly patients was 60 years old. 120 patients were enrolled, the median age was 71 (60–88). As in other clinical trials, patients were predominantly ‘fit’, only 6.7% had a PS of 2, 15.8% had a Charlson comorbidity index ≥2 and 28.3% had an IADL dependency, which is a major limitation of this study. Based on the available data, it can be concluded that pazopanib is noninferior to doxorubicin and has a different safety profile that may help guide the choice of treatment. Median PFS was comparable between pazopanib and doxorubicin (4.4 vs 3.3 months) with a non-significant trend to a higher PFS in patients ≥ 71 years old treated with pazopanib (HR = 0.62; 95% CI: 0.34–1.12). Median OS for the overall population was 13.9 months. Pazopanib was associated with a median OS of 12.3 months versus 14.3 months in the doxorubicin arm with an HR of 1.08 (95% CI, 0.68–1.72). The incidence of toxicity was similar for both arms. However, the incidence of grade 4 neutropenia (0% vs 56.4%) and febrile neutropenia (0% vs 10.3%) favored pazopanib. Pazopanib was associated with higher incidence of hypertension, hypothyroidism, diarrhea and general deterioration. On the other hand, higher incidence of alopecia, stomatitis and mucosal inflammation was reported with doxorubicin [30]. Thus, the safety profile of the given therapy and the type of administration are to be considered when choosing a treatment.

Oral trofosfamide, an oxazaphosphorine prodrug, has been tested in a randomized phase II trial in elderly patients with untreated metastatic STS [31]. A total of 120 patients were enrolled. The median age was 70 years (60–89). Eighty five percent had an ECOG PS of 0 or 1. Safety analyses revealed a favorable safety profile for trofosfamide. Grade 3 or 4 side-effects were lower in trofosfamide arm (59% vs 30.3% p = 0.005). Trofosfamide caused more grade 1 or 2 dyspnea and fatigue. No hemorrhagic cystitis, nephrotoxicity, or higher grade of neurotoxicity have been reported. The 60-day mortality rate was 8% in both arms. Of note, approximately 10% of patients received trofosfamide for more than 1 year. In terms of efficacy, the primary end-point was met. The 6-month progression free rate determined to be 27.6% (95% CI: 18–39.1%).

A single arm phase II prospective trial investigating a non-pegylated liposomal doxorubicin in advanced STS included 37 patients [32]. The median age was 74 years old (51–87). This report showed high rates of grade 3 and 4 toxicities: neutropenia (62%), leukopenia (27%), febrile neutropenia (16%), thrombopenia (10%), anemia (5%). Grade 3 non-hematological toxicities occurred in 4 (10%) patients. 2 patients presented with cardiac toxicity with one death due to cardiac toxicity.

The TR1US study is phase II trial investigating trabectedin as first-line treatment in elderly patients with advanced STS who are inoperable and unfit to receive standard anthracycline-based chemotherapy due to medical comorbidities (population usually underrepresented in STS trials) [33]. Twenty-four patients with a median age of 79 years (74–83) were enrolled. The study met its co-primary end points showing that trabectedin at standard doses is active and well tolerated. In fact, the PFS at 3 months was 71% (80% CI: 57–81%). Ten patients (42%) experienced clinically limiting toxicity. The most frequent grade 3 or 4 AEs were neutropenia, and fatigue. The results of this study make trabectedin an interesting treatment of choice for elderly patients unfit to receive standard chemotherapy.

1. Second line therapy and later

The efficacy and safety of trabectedin in patients aged > 60 years with pretreated advanced STS have been assessed by a pooled analysis of 5 phase II trials [34]. Data from 350 patients were divided into two cohorts. The younger cohort included 267 patients with a median age of 48 years (19–59 years), and the older cohort had 83 patients with a median age of 65 years (60–81 years) including 24 patients (7%) aged ≥70 years. Most patients were exposed to one or two lines of prior chemotherapy. No significant differences were observed between the younger and older cohorts for response rate (10.1 vs 9.6%), median PFS (2.5 vs 3.7 months) or median OS (13 vs 14 months). However, grade 3/4 AEs were more common in older patients versus younger patients including anemia (19.3 vs 10.1%), neutropenia (60.2 vs 43.6%), fatigue (14.5 vs 6.4%) and febrile neutropenia (1.2 versus 0.4%). Of note, the use of G-CSF was similar in both cohorts (12.7% vs 13.3%). Importantly, no significant differences in efficacy or safety outcomes were found in patients aged 70 or older. Deaths associated with drug-related AEs were infrequent (1.9% and 2.4% of patients in the younger and older cohort, respectively).

A phase III trial has compared trabectedin and dacarbazine in advanced liposarcoma or leiomyosarcoma and demonstrated the superiority of trabectedin [35]. A subgroup analysis of the subgroup of patients aged >65 years (n = 131) showed consistent results with those in the overall population [36]. Trabectedin significantly improved PFS (4.9 vs 1.5 months) with a reduction in the risk of disease progression or death by 60%. However, there was no significant difference in OS (15.1 vs 8.0 months). The safety profile of trabectedin was comparable to the overall study population. However, treatment-related AEs leading to discontinuation of the treatment were increased in the elderly as compared to the overall population (28% vs 17% for trabectedin and 9% vs 8% for dacarbazine).

The results of these studies (Table 1), despite their limitations, provide some evidence for the efficacy and the safety of systemic therapy in elderly patients. Unfortunately, data on how systemic treatments affect the quality of life in older patients with STS are lacking. The E-TRAB trial will analyze quality of life and patient-reported outcomes data in elderly patients with STS [37].

Table 1. Safety and efficacy of systemic therapy in elderly patients with advanced STS

Study	Type	Setting	No of patients	Median Age (Ranges) Years	PS and or GA	Treatments	Efficacy	Toxicities (Grade 3 or 4)
Grosso F et al[^33].	Single-arm phase II	First line	24	79(74–83)	PS 0–1 (n = 21) PS 2 (n = 3)	Trabectedin	RR: n = 16 (80%) SD PFS3: 71% (80%CI; 57%-81%) mPFS:4 months mOS:12 months	CLTs: n = 10 (42%; 80% CI 28%-57%) neutropenia (n = 9, 38%), fatigue (n = 5, 21%), Aminotransferase elevation (n = 5, 21%)
Grünwald Vet al[^30].	Randomized phase II	First line	120	71(60–88)	PS 0–1 (n = 112, 93.3%) PS 2 (n = 5, 6.66%) ≥2 comorbdities (15.8%) ADL impairment (28.3%)	Pazopanib (N = 81)	RR: 12.3% mPFS: 4.4 months (95%CI:2.7–6.0) mOS: 12.3 months (95%CI:8.7–19.8)	Hypertension (25.9%), ALT elevation (7.4%), fatigue (6.2%) neutropenia (5.0%) deterioration of general physical health (3.7%)
						Doxorubicin (N = 39)	RR: 15.4% mPFS:5.3 months (95%CI:1.7–8.2) mOS: 14.3 months (95%CI:8.3–25.9)	Neutropenia (29.7%), Leukopenia (18.9%), Anemia (10.8%), Nausea (5.4%), Diarrhea (5.4%)
Hartmann JTet al[^31].	Randomized phase II	First line	120	70(60–89)	PS 0–1 85% (n = 102) PS 2 15% (n = 18)	Trofosfamide (N = 80)	RR: 48% (n = 36) mPFS: 2.8 months (95%CI:1.7–3.6) mOS: 12.3 months (95% CI: 9.6–16.2)	Leukopenia 9.2%. Anemia 9.2%. Fatigue 5.3% Neutropenia 1.3%. Mucositis 0%
						Doxorubicin (n = 40)	OR: 61.5% (n = 24) mPFS: 4.3 months (95%CI: 2.2–6.3) mOS: 9.8 months (95% CI: 6.7–11.6)	Leukopenia 35.9% Neutropenia 33.3% Anemia 10.3% Fatigue 5.1% Mucositis 2.6%
A. Lopez-Pousa et al[^32].	Single-arm phase II	First line	37	74(51–85)	PS≤2	Doxorubicin GP-pharm	RR: 30% (n = 8) mPFS: 3 months (95% CI: 1.9–7.4) mOS: 14.4 months (95% CI: 6.3–27.1)	Neutropenia 62% Leukopenia 27% Febrile neutropenia 16% Thrombocytopenia 10% Anemia 5% Asthenia 10% Grade 5 cardiotoxicity in one patient
Eugenie Yet al[^28].	Subgroup analysis (phase III)	First line	209	70(65–89)	PS 0–1 (100%)	Doxorubicin (n = 103)	RR: 18% mPFS: 6.1 months (95% CI: 4.7–7.2) mOS:17.0 months (95% CI:12.9–20.6)	Hematological and nonhematological toxicities 83% (n = 85)
						Doxorubicin + evofosfamide (n = 106)	RR: 25% mPFS:6.5 months (95% CI: 4.9–8.2) mOS: 16.2 months (95% CI: 11.5–23.0)	Hematological and nonhematological toxicities 88% (n = 93)
Eugenie Yet al[^3].	Pooled analysis of 12 trials	First line	348	68(67–70)	PS 0–1 89% (n = 311)	Anthracycline-based regimens (n = 283) Non anthracycline regimens (n = 65)	RR: 14.9% mPFS:3.5 months (95% CI: 2.7–4.3) mOS: 10.8 months (95%CI: 9.43–11-83)	-
Jones RL et al[^36].	Subgroup analysis (phase III)	≥2^nd line	131	(65–81)	PS 0–1 100%	Trabectedin (n = 94)	RR: 72% mPFS: 4.9 months (95%CI:3.55–7.13) mOS: 15.1 months (95%CI: 13.14–19.15)	Neutropenia 39% Leukopenia 28% Anemia 18% Thrombocytopenia 20% Aminotransferase elevation 39% Fatigue 12% Nausea 10%
						Dacarbazine (n = 37)	RR:44% mPFS:1.5 months (95%CI:1.38–2.79) mOS: 8.1 months (95%CI:4.57–14.72)	Neutropenia 23% Leukopenia 23% Anemia 34% Thrombocytopenia 34%
Cesne AL et al[^34].	Pooled analysis of 5 phase II trials	≥2^nd line	83	65(60–81)	PS 0–1 100%	Trabectedin	RR:9.6% mPFS: 3.7 months (95%CI:2.1–5.5) mOS:14 months (9.5–23-9)	Neutropenia 60.2% Thrombocytopenia 20% Anemia 19.3% Febrile neutropenia 1.2% Aminotransferase elevation 81% Fatigue 14.5%
Mir O et al[^27].	Retrospective	≥1 first line	26	72(66–88)	Median PS 2 (1–3)	Cyclophosphamide	RR: 26.9% (95%CI:9.9–44.0) mPFS:6.8 months 80–20.6) mOS: 14.1 months (1.2–30.4)	Lymphopenia 80.7% Anemia 8% Thrombocytopenia 8%

PS = Performance status, GA = geriatric assessment, RR = Response rate, SD = stable diseases, PFS3 = Progression free survival at 3 months, mPFS = median progression free survival, mOS = median overall survival, CLTs = Clinically limiting toxicities, AEs = Adverse Events. ADL = Activities of daily living, ALT = alanine aminotransferase

4. Assessment of vulnerabilities and tools to predict treatment related toxicities

4.1. Geriatric assessment

Chronological age should not be a barrier to the use of potentially curative or palliative life-prolonging therapy. Indeed, based on the studies mentioned above, the outcome of elderly patients with STS treated with specific therapy was significantly better than patients receiving BSC, and close to that observed in younger patients. However, since the elderly population is heterogenous, it is essential to identify those who are more likely to benefit from specific treatment. Several studies have shown that geriatric assessment has the potential to evaluate the balance of benefits and harms of performing or omitting specific treatments in older patients with cancer [6,7].

The Comprehensive Geriatric Assessment (CGA) represents the standard. It is a multidimensional and interdisciplinary procedure to identify vulnerabilities. A complete assessment must include functional status, comorbidity, cognition, mental health status, nutrition, social status and support, fatigue, and assessment for polypharmacy and presence of geriatric syndromes. Various tools are available to measure these domains [7]. The CGA helps to determine, fit, vulnerable or frail patients. It has the ability to detect impairments not identified in routine oncology assessment, predict survival, treatment-related toxicity and influence the treatment choice. In a systematic review, Hamaker et al. showed that the CGA modified the treatment plan in 8% to 54% of older patients with cancer [38]. However, there are no studies exploring the impact of CGA on treatment decision in patients with STS.

The CGA is a strong predictor of adverse events in geriatric oncology, nevertheless it has not been specifically studied in STS. The only geriatric STS trial that incorporated a prospective geriatric assessment was the EPAZ trial [30]. However, no stratifications on the geriatric assessment has been provided in the publication. An ongoing trial ET-TRAB by the German Interdisciplinary Sarcoma Group (NCT03022448) is analyzing the predictive value of CGA with regards to treatment-related toxicities in elderly patients with advanced STS who are unfit to receive anthracycline-based chemotherapy and treated with trabectedin as first line therapy [37].

Particular domains, such as nutrition, functional status, comorbidities, and cognition are significantly associated with treatment-related toxicity, interruption of chemotherapy and mortality across several cancer types [39,40]. The impact of these domains on STS treatment-related toxicities has not been investigated. Based on the available literature, the only domains assessed in the elderly sarcoma population are comorbidities in addition to the functional status using the classical PS. These have shown that most elderly STS patients have accompanying comorbidities (>60%) [16], and poor general condition with a PS≥2 (>30%) [16]. These parameters significantly affect prognosis, outcome, and treatment decision in elderly patients with STS [16]. However, it is important to note that daily living dependence scales including activities of daily living (ADL), and instrumental activities of daily living (IADLs) are more sensitive than a PS scale to detect functional problems and are more efficient to predict survival and toxicity in elderly patients with cancer. In fact, Jolly TA, D et al., reported that geriatric assessment of functional domain according to ADL and IADL was able to identify deficits that could affect treatment tolerance and outcome in 23% of patients who were assessed with adequate PS [41]. Furthermore, Hurria A et al. have found that Karnofsky PS was not predictive of toxicity to chemotherapy [42].

Malnutrition is a substantial domain to consider. Data on elderly STS patients are insufficient. In their study, Yousaf N et al. included 120 patients aged ≥65 years old receiving first line chemotherapy for advanced STS [43]. The authors demonstrated that a low serum albumin was significantly associated with increased risk of hospitalization due to toxicity probably due to pharmacokinetic changes with an elevation of free-drug level. However, hypoalbuminemia does not necessarily mean a nutritional impairment as it could be due to several causes. Screening for malnutrition is a fundamental part of geriatric assessment and validated tools such as the Mini Nutrition Assessment (MNA) should be used [6].

There are no studies exploring the use of CGA for pre-surgery assessment in STS surgery field. For radiotherapy, a recent study in 50 elderly patients with lung or head and neck cancers receiving radiation found that individuals with IADL problems were more likely to develop severe patient-reported toxicities during radiation treatment with decline in health-related quality of life [44].

Geriatric assessment is a fundamental part of the management of elderly patients with cancer. However, the CGA is very difficult to implement in daily practice for all patients. It is time consuming and requires interdisciplinary collaboration with geriatricians. Several screening tools to select patients who could benefit from a CGA have been developed such as the G8, the Flemish version of the Triage Risk Screening Tool (fTRST) and the Vulnerable Elders Survey-13 (VES-13) [6,45]. These are rapid triage tools that can be used in a busy setting. They are highly sensitive, but none of them are specific and need to be supplemented by a CGA if positive. The G8 is one the most studied and practical tools, and it could be filled out by patients or included in the initial nursing assessment. The ET-TRAB trial is investigating the role of G8 screening tool as a predictive factor for unplanned hospitalizations, grade 4 toxicities and early death within the first 6 months in elderly patients with advanced STS [37]. The results of this trial are awaited.

4.2. Tools to predict treatment related toxicity

On the basis of CGA, comprehensive risk tools to predict chemotherapy toxicity in older patients with cancer have been developed.

The Chemotherapy Risk Assessment Scale for High-Age Patients (CRASH) score consists of two components, a hematologic toxicity score and non-hematologic toxicity score [46]. The need for assistance by the IADL scale, diastolic blood pressure, lactate dehydrogenase, and toxicity of the chemotherapy regimen are predictive of hematologic toxicity. PS, cognition by Mini-Mental Status Score, malnutrition by MNA score, and toxicity of chemotherapy regimen are predictive of non-hematologic toxicity.

The CARG prediction tool is another chemotherapy toxicity score proposed by Cancer and Aging Research Group (CARG) [42,47]. This score combines different variables; age, cancer type, chemotherapy dosing, number of drugs, hemoglobin, creatinine clearance, as well as geriatric assessment parameters including hearing, number of falls in previous 6 months, the need for assistance with taking medications, the ability to walk 1 block, and social activity. The predictive value of CARD score will be investigated in ET-TRAB trial in elderly patients with STS [37].

Few data have been reported with regards to surgery and radiotherapy. For surgery, besides the standardized pre-surgery assessment, elderly patients require a geriatric assessment. The pre-operative Assessment of Cancer in the Elderly (PACE) is a multiparametric score that includes a CGA, Brief Fatigue Inventory (BFI), PS, and ASA score, has been demonstrated to be predictive of 30-day postoperative morbidity [48]. For instance, an international Society of Geriatric Oncology (SIOG) prospective study conducted in 460 patients ≥70 years with various cancers who received PACE prior to surgery found that dependent ADL and IADL, fatigue and PS were associated with 50% increase in the relative risk of post-operative complications [48]. There is no specific score for sarcoma surgery.

5. Active supportive care and treatment adjustment

The fear of toxicity and unexpected side effects are the main factors of the undertreatment of older patients. However, it has been demonstrated that with appropriate supportive care and a close monitoring, treatments can be made safer.

5.1. Active supportive care

The most frequent toxic effects of chemotherapy in elderly patients with STS are neutropenia, anemia, mucositis and cardiotoxicity. This greater susceptibility may be explained by physiological changes and the higher prevalence of comorbidities, nutritional impairments, and polypharmacy in older patients.

Correction of reversible conditions is a fundamental part before stating active treatment for cancer patients. All patients should be screened for nutritional risk. Supportive measures to prevent chemotherapy side effects should be applied in accordance with general established guidelines [49]. Prophylactic therapy with granulocyte colony-stimulating factor is recommended for patients ≥65 years old [50]. Anemia has been reported as a significant prognostic factor in STS [43]. Any cause of anemia should be treated accordingly. The management of chemotherapy-induced anemia and the use of erythropoietic agents is the same as for the general population [51]. Preventive measures are important in reducing the severity of mucositis. Maintenance of optimal nutritional support during treatment, and daily oral hygiene routine are the key strategies to reduce mucosal injury [52].

Supportive measures to prevent hematological and non-hematological toxicities for patients on targeted therapy are similar to that for chemotherapy.

5.2. Treatment adjustment

Dose adjustment is frequently considered in elderly patients in order to prevent treatment-related toxicities and hospitalization. In their study, Garbay et al. found that 37% of elderly patients with advanced STS received a reduced dosage of chemotherapy with no negative impact on PFS [16]. Prospective trials are needed in order to evaluate whether reduced dose regimens provide acceptable results for the elderly especially the frail ones.

Anthracyclines are the cornerstone of advanced STS treatment. The higher risk of anthracycline-induced cardiotoxicity and myelosuppression in the elderly patients ultimately limit the potential of the drug in this group [53]. According to the previous studies, standard doses of anthracycline can be used efficiently in selected elderly patients with advanced STS. However, considering the high risk of cardiomyopathy and congestive heart failure, it is important to closely monitor cardiac function and cumulative dose limits. Early retrospective studies showed that smaller doses of doxorubicin caused less cardiotoxicity [53,54]. However, empirical dose reduction of anthracycline may impact its activity. A dose-response correlation in patients with advanced STS has been reported with a higher response rates at dose >60 mg/m2 every 3 weeks [55]. Dexrazoxane, an iron chelator, reduces the incidence of anthracycline-induced cardiotoxicity with no adverse effect on survival, as demonstrated in numerous clinical trials [56]. However, this must be taken with caution as it was not studied primarily in elderly patients with STS. In addition, a retrospective study in STS patients reported an apparent increase in myelosuppression with higher rates of grade 3 and 4 toxicities; leukopenia (56.5% vs 28.7%; p = 0.0014), neutropenia (69.6% vs 24.1%; p = 0.0001), as well as more frequent febrile neutropenia (52.2% vs 20.7%, p = 0.0004) and dose reduction (39.1% vs 19.5%; p = 0.0221) when dexrazoxane was used [57]. Continuous infusion of doxorubicin over 48 hours or more have been shown to increase the cumulative doxorubicin dose that can be safely given to patients. However, a recent study conducted in sarcoma patients aged 23–71 years old has shown that both continuous schedule or dexrazoxane do not completely improve subclinical doxorubicin-induced cardiotoxicity as detected by sensitive biomarkers such as high-sensitivity troponin [58].

Liposomal doxorubicin has better tolerability profile in young adult with advanced STS as compared to free doxorubicin, but has not been tested in elderly patients [58]. Trabectedin remains the only approved alternative to doxorubicin for first line therapy in patients unsuited to receive anthracycline. Given at doses 1.3–1.5 mg/m2, maximum of 2.6 mg over 24 hours every 3 weeks with adequate dexamethasone premedication, trabectedin showed an acceptable toxicity profile in vulnerable and frail elderly patients [33].

The use of oral agents (pazopanib [30] or metronomic cyclophosphamide [27]) is appealing in older patients. They offer convenience and may be preferred by patients [59]. However, adherence to oral treatments may be particularly challenging especially in the presence of cognitive impairment. It is also particularly important to consider concomitant medications due to the risk of interactions. Finally, comorbidities and patient’s preferences should align with characteristics and expected side effects.

6. Conclusion

The proportion of elderly patients with STS is increasing continuously. The management of these patients requires careful considerations. All decisions should consider physiological age, estimated life expectancy, risks, benefits, and the patient preferences. The role of CGA in treatment assignment and risk stratification of elderly patients with STS is still unclear. Future studies must be representative of the population of interest and should include CGA.

7. Expert opinion

The key to balance efficacy with toxicity in the elderly is to risk stratify older patients with regard to potential toxicity: Low risk patients may receive standard therapy. However, vulnerable patients, who could suffer excessive toxicities, may be proposed alternative options.

To provide such recommendations, only a randomized controlled trial which stratify patients by specific geriatric characteristics has the potential of answering this question. The main issue with this trial design in older individuals is to achieve the target accrual, and this is especially true for a relatively rare disease such as STS. Within this context, expanding the eligibility criteria with prospective subgroup analyses, single arm phase II trials, and prospective observational studies could be practical alternatives to guide management in this population. It is also important to consider including geriatric assessment tools in these studies and further collaboration between geriatric and oncology researchers. In fact, the feasibility of CGA in a multicenter trial setting has been well established by geriatric trials outside STS (lung cancer). This principle is very important for STS trials that are usually of a collaborative nature. It is also important to choose appropriate endpoints which are more relevant for this population of patients such as the quality of life.

There are two notable examples that geriatric trials in STS are feasible; the phase II TR1US trial that generated the evidence for treating vulnerable and frail patients with trabectedin in first line setting for advanced STS. Another example is the ongoing non-interventional E-TRAB (GISG-13) study including older patients with advanced STS treated with trabectedin and in which a CGA is performed in order to predict for safety. These studies, despite their limitations, represent a major step in the evolution of geriatric research in the field of STS and mark the beginning of an era of inclusion of older patients in STS trials.

Article highlights

• Almost half of patients diagnosed with STS are older than 65 years; however, this group is largely excluded from or underrepresented in clinical trials.

• Compared with their younger counterpart, older patients with STS receive less intensive treatment. Indeed, older age is associated with lower use of radiotherapy, chemotherapy and less extensive surgery.

• Treatment efficacy in older patients with STS did not differ from that in younger patients.

• Available evidence supports treating elderly patients with STS with similar treatment as that for younger patients after a careful geriatric evaluation.

• As some elderly patients may not be suitable for standard therapies, alternative options with better tolerability profile are clearly needed. The majority of evidence has been generated with trabectedin, the only approved agent for first line use in elderly patients with advanced STS unsuited to receive doxorubicin.

• Managing the risk of toxicity requires an assessment of vulnerabilities with appropriate tools. All decisions should consider physiological age, estimated life expectancy, risks, benefits, and the patient preferences.

This box summarizes key points contained in the article.

Declaration of interest

R Jones has received funding from/served as a consultant for Adaptimmune, Athenex, Bayer, Boehringer Ingelheim, Blueprint, Clinigen, Eisai, Epizyme, Daichii, Deciphera, Immunedesign, Lilly, Merck, Pharmamar, Springworks, Tracon and Upto Date. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

Reviewer disclosures

Peer reviewers on this manuscript have no relevant financial or other relationships to disclose.

Additional information

Funding

This paper was supported by the Royal Marsden/Institute of Cancer Research National Institute for Health Research Biomedical Research Centre.