Abstract
Background/Aim: Almost half of all patients with soft-tissue sarcoma are over 65 years of age, and the proportion of older patients is increasing. Despite this, they have been underrepresented in clinical trials and only limited data are available to guide treatment decisions. The aim of this study was to investigate treatment patterns and outcomes in older patients with soft-tissue sarcoma. Patients and Methods: Patients over 50 years old treated for advanced soft-tissue sarcoma at the Helsinki University Hospital between January 2000 and July 2020 were included. Data on patient and tumor characteristics, treatment, and survival were retrospectively collected. A total of 152 patients were included: 14.5% (n=22) were over 75 years old, 34.2% (n=52) were 65-74 and 51.3% (n=78) were 50-64 years old. Results: The outcomes of the oldest group differed from those of younger patients; they were more likely to receive single-agent treatment as first-line therapy (90.9% vs. 28.8% and 24.4%, p<0.001) and had the lowest relative dose–intensity (70% vs. 88% and 95%, p<0.05). They experienced grade three to four hematological adverse events less frequently (38.1%, 56.9% and 72.7%, respectively, p=0.031), and received fewer lines of treatment (median of 1, 2 and 2, respectively, p=0.01). In patients aged ≥75 years, there was no association between further lines of therapy and improved survival. Compared to the youngest group, the oldest patients had a greater risk of dying (hazard ratio=1.7, 95% confidence interval=1.0-2.8, p=0.041) and their median overall survival was only 7.4 months, compared to 14.3 and 12.9 months in the two younger groups. Conclusion: These findings suggest that older patients tolerate chemotherapy when treatment is tailored to their needs but may not benefit as much as younger patients.
Cancer is said not to discriminate, but it does, at least in part; more than half of all malignant tumors are diagnosed in patients aged over 65 years, and as populations in several countries are aging, the incidence of many cancer types is predicted to increase (1, 2).
Soft-tissue sarcomas (STS) are a heterogenous group of rare tumors that consist of more than 50 different subtypes (3). Like many cancer types, STS disproportionately affect older people, with incidence rising from 1.7 per 100,000 in the population under 50 years old, to 11.3 per 100,000 in those aged 65 years or more (1, 4).
Despite the distribution of STS incidence in the general population, older patients have been severely underrepresented in clinical trials. A meta-analysis of 12 STS trials by European Organization for Research and Treatment of Cancer found that only 12.4% of the trial patients were ≥65 years old (5). This underrepresentation of older patients has left clinicians with limited information to guide treatment decisions. Importantly, even in the highly selected populations of randomized controlled trials, older patients have been reported to have inferior overall survival (OS) and to be more likely to stop treatment due to adverse events (AEs) (5, 6).
Few retrospective studies have described real-world treatment patterns in older patients with STS (7-11). The studies available have reported a wide range of survival, from 6.5 to 13.6 months, among older patients who have received antitumor treatment. They have also confirmed that AEs are prominent in the real-world setting. In a study by Yousaf et al. 38% of patients were hospitalized due to treatment (7).
This study was conducted with the aim of further investigating treatment patterns and outcomes of older patients with STS. Little is known about relative dose intensity (RDI) of treatment in this group, even though a dose–response relationship has been demonstrated in STS (12-14). Therefore, this study also aimed to elucidate patterns and effects of RDI in older patients with STS.
Patients and Methods
Patients treated for STS at the Helsinki University Hospital between January 2000 and July 2020 were identified. Patients with a diagnosis of angiosarcoma, rhabdomyosarcoma, osteosarcoma, chondrosarcoma, Ewing sarcoma or gastrointestinal stromal tumor were excluded, leaving 1,546 patients. Patients were also excluded if they had never received chemotherapy for advanced disease (n=1,325) or if they were <50 years old at the time of advanced disease diagnosis (n=60). During data collection, nine additional patients had to be excluded because their data could not be retrieved. This left 152 patients that were included in this study. Three of these patients lived near a secondary center from which full patient records were not retrievable. These patients were excluded from the toxicity analysis because there was uncertainty whether they had been hospitalized during chemotherapy treatment. A flow chart of patient selection is shown in Figure 1.
Patient selection. STS: Soft-tissue sarcoma.
Age-adjusted Charlson’s comorbidity index (ACCI) was calculated for each patient based on their status at the time of diagnosis of advanced STS. As STS was the focus of this study, it was excluded from the ACCI score. AEs and response to palliative chemotherapy were collected. Common Terminology Criteria for Adverse Events Version 5.0 was used to grade AEs (15). To avoid under-reporting, hospitalizations and deaths within 30 days of last treatment were recorded as treatment-related unless they were reliably contributable to another cause. Response Evaluation Criteria for Solid Tumors (RECIST) Version 1.1 was used to determine response to treatment (16). A radiologist reviewed cases for which the response could not be evaluated from the original reports. The RDI was calculated from the treatments the patients had received during the course, as opposed to intended treatments.
In data analyses, patients were divided into age groups. For this, age was defined as patient’s age either at the time of diagnosis of metastases, or at the start date of first-line chemotherapy if palliative chemotherapy was initiated for locally advanced disease.
In the study population, there were 31 patients with a history of metastectomy with curative intent. Most patients included here had experienced relapse after their operation, and hence started palliative chemotherapy but seven patients were redirected to metastectomy after the initiation of palliative chemotherapy treatment or received high-dose radiotherapy to all disease sites after first-line treatment. These patients were included in the analyses of AEs and treatment response but were excluded from the analysis of survival.
Statistical analysis. Differences in disease and treatment characteristics between age groups were analyzed using chi-square and Fisher’s exact tests for categorical variables and Kruskal–Wallis and Mann–Whitney U-test for continuous variables. Kaplan–Meier survival analysis was performed. OS was measured from the start of chemotherapy until death. Patients were censored at the last follow-up date. Cox regression model was used for univariate and multivariate analyses to identify factors predictive of survival. Variables tested included: sex, age, tumor grade, and histology, tumor stage (metastatic or locally advanced), ACCI, RDI during first-line treatment, treatment regimen, and single vs. multi-agent treatment during first-line therapy. Variables associated with survival at p<0.05 were included in the multivariate analysis.
Binary regression was used to investigate the effect of factors on treatment response and AEs. For response, as defined by the RECIST criteria (16), the tested factors included: age, tumor grade, histology, tumor stage (locally advanced or metastatic), relative dose -intensity, treatment regimen, and type of first-line treatment (single vs. multi-agent). For AEs, the factors were: sex, age, ACCI, RDI, treatment regimen, and single vs. multi-agent treatment. Box–Tidwell procedure was used to test the relationship between the logit of dependent variables and continuous independent variables used in binary logistic regression.
All data were analyzed using IBM SPSS Statistics Version 27 (IBM, Armonk, NY, USA).
Results
Patient population. A total of 152 patients in our database fit the inclusion criteria for this study; 51.3% (n=78) of them were 50-64 years old, 34.2% (n=52) were 65-74 years old and 14.5% (n=22) were aged ≥75 years. The mean ages in the three groups were 58.2±4.4, 69.9±2.9, and 78.7±3.2 years. A significantly larger portion of older patients were male (43.6%, 57.7%, and 72.7% in the groups respectively, p=0.041). Most patients in this study had been diagnosed with deep-seated (94.1%) and grade 3 tumors (89.5%). Synchronous metastases had been diagnosed in 20.4% of patients, while 9.2% (n=14) had only locally advanced disease at the start of palliative chemotherapy. The different age groups had no statistically significant differences in these baseline characteristics.
There were numerical differences in the distribution of histological subtypes between the three age groups. However, age group and histological subtype were not significantly associated (p=0.79). Patient demographics are described in Table I.
Patient demographics.
Treatment characteristics. The median number of different chemotherapy lines was two (range 1-5) for the two younger age groups and one (range 1-3) for the ≥75-year-old patients (p=0.01). Treatment regimens for first-line therapy are described in Table II. Three quarters of patients (75%) received anthracycline-based treatment in first-line therapy.
Treatment characteristics during first-line therapy.
First-line treatment regimen and age group were significantly associated (p<0.001). The oldest patients were most likely to receive single-agent treatment (24.4%, 28.8%, and 90.9% in each group, respectively, p<0.001). On the other hand, 65- to 74-year-old patients were significantly more likely to receive anthracycline-based chemotherapy when compared to the youngest patients [hazard ratio (HR)=2.9, p=0.03]. However, a significantly larger proportion of the youngest patients had a history of adjuvant ifosfamide-doxorubicin treatment (29.5%, 15.4%, and 4.5% respectively, p=0.01).
Information on RDI was retrieved for 136 patients. Among these patients, the median RDI was 0.91 (interquartile range=0.78-0.99). The oldest (≥75 years) patients had a significantly lower RDI than the other two groups (median=0.70, interquartile range=0.67-0.83, p<0.05).
Treatment responses. To investigate which factors influenced response to chemotherapy, binary logistic regression was performed comparing patients attaining complete or partial response to those who did not. In univariate analysis, younger age [odds ratio (OR)=1.1, 95% confidence interval (CI)=1.0-1.2], histology of synovial sarcoma (OR=17.5, 95% CI=2.7-111.1), and multi-agent treatment (OR=4.1, 95% CI=1.2-14.5) were significantly associated with response. Multivariate analysis was not performed due to the low number of observations. Treatment responses are described in Table III for different regimens, histological subtypes, and the three age groups.
Treatment response in first-line therapy.
Adverse events. A total of 149 patients were included in the analysis of AEs; 36.9% (n=55) of these patients were hospitalized during first-line treatment, and toxicity led to early cessation in 18.1% of patients (n=27). There were no statistically significant differences in the frequencies of these events between the different age groups.
Age group was significantly associated with grade three to four hematological AEs but not with other AEs: 72.7%, 56.9%, and 38.1% of the 50- to 64-year-old, 65- to 74-year-old, and ≥75-year-old patients, respectively, experienced grade three to four hematological AEs (p=0.031). Logistic regression was performed to investigate if other factors were associated with this outcome. In addition to age, multi-agent chemotherapy treatment was significantly associated with hematological toxicity (OR=2.5, 95% CI=1.2-5.0). However, in multivariate analysis neither of these factors remained significant. AEs are further described in Table IV.
Grade 3-5 adverse events from first-line treatment.
Survival analysis. Of the 145 patients included in the OS analysis, 91% died during the follow-up. The median OS for all patients was 12.9 months (95% CI=11.2-14.7 months). In univariate analysis, factors significantly associated with favorable survival were female sex, French grade two, multi-agent first-line chemotherapy, and histology of leiomyosarcoma. Increasing age was associated with an inferior prognosis (HR=1.02 per additional year, 95% CI=1.00-1.04, p=0.047). In particular, age ≥75 years was associated with inferior survival in comparison to the youngest group (HR=1.68, 95% CI=1.02-2.76, p=0.041) but there was no statistically significant difference between the two younger age groups (HR=1.18, 95% CI=0.81-1.72, p=0.4). The RDI during first-line chemotherapy was not associated with survival in the whole study population nor within the age groups. In multivariate analysis, female sex, French grade two, and leiomyosarcoma histology remained significantly associated with survival. Detailed results from both univariate and multivariate analyses are reported in Table V.
Results of univariate and multivariate analyses of factors significantly associated with overall survival (OS).
Overall, 60.3% (n=88) of the patients included in the OS analysis received more than one line of antitumor treatment. Patients who were diagnosed with progressive disease after first-line treatment, according to RECIST criteria or physician’s evaluation, were analyzed as a separate group to investigate if further lines were associated with improved survival (n=126). Patients were excluded if they had not been diagnosed with progression or had switched treatment due to an AE or cumulative dose limitation. In univariate analysis, patients who had received two or more antitumor treatment lines had a lower risk of dying (HR=0.6, 95% CI=0.41-0.88, p=0.009). However, when patients were divided by age, the number of treatment lines was significantly associated with improved survival only for the 65- to 74-year-old patients. While the 50- to 64 -year-old patients had a trend favoring further lines of therapy, as can be seen in Figure 2, the oldest patients did not share this trend.
Kaplan–Meier analysis of overall survival probability in the whole cohort by age group (A) and according to the number of different antitumor treatment lines in those aged 50-64 years (B), 65-74 years (C), and ≥75 years (D).
Discussion
This study demonstrated that the treatment and outcomes of patients aged ≥75 years differed from younger patients in clinically meaningful ways: The oldest patients received fewer lines of antitumor treatment and had a significantly lower RDI during first-line treatment, despite receiving single-agent treatment more frequently. Age ≥75 years was associated with a greater risk of dying and this group had a median overall survival of only 7.4 months. Importantly, the oldest patients did not show a trend toward improved survival from receiving more than one type of antitumor treatment.
Previous studies have shown that a significant proportion of older patients (38-71.3%) diagnosed with advanced STS never receive systemic treatment (8, 11). The older patients who are treated, on average, receive lighter treatment than younger patients: upfront dose reductions have been reported to be common, affecting 34-44% of patients (8, 10). In addition, single-agent chemotherapy was chosen for the majority of older patients, 80-83%, in contrast to 54.5% in the METASARC cohort which included all adult patients (≥18 years old) (7, 8, 17). In the present study, patients aged ≥75 years had similar first-line treatment patterns to previous reports: 90.9% of them received single-agent treatment in first line and they had a lower RDI than younger patients, 70%. In contrast, only 28.8% of the 65- to 74-year-old patients in the present study received single-agent treatment, and they had a RDI of 88%. These differences from previous reports most likely reflect differences in patient populations.
In a subgroup analysis of the SARC02 trial, older patients were reported to be more likely to experience AEs than younger patients (6). AEs have been reported to be common in real-world cohorts as well: In the study by Yousaf et al. (7) 38% of patients were hospitalized during chemotherapy, and 16% of patients in the study by Garbay et al. (8) stopped treatment due to AEs. While the frequency of these events was similar in our study – 36.9% of patients were hospitalized and 18.1% stopped treatment due to AEs – there were no significant differences between age groups. Age was significantly associated only with hematological AEs: the oldest patients experienced them least frequently. These findings re-enforce the view that with careful patient selection and treatment planning, older patients can tolerate treatment similarly to younger patients.
Older patients treated for advanced STS have been reported to have shorter survival than younger patients (5). Previous real-world cohorts have reported median OS times ranging between 6.5 and 13.6 months (7-10). In line with the previous results, increasing age was adversely associated with survival in the current study. The impact of age was further investigated by comparing the survival of 50- to 64-year-old patients to that of 65- to 74-year-old and ≥75-year-old patients. While those ≥75 years had a greater risk of dying than the youngest group, there was no statistically significant difference between 65- to 74-year-olds and the youngest group. This finding suggests that while an age of 65 years is a useful cut-off to guide geriatric screening, as suggested by the American Society of Clinical Oncology guidelines (18), it is insufficient to stratify prognosis in the STS population.
In addition to younger age, factors that were favorably associated with survival in our study population included female sex, leiomyosarcoma histology, and lower tumor grade. These findings are in line with previous reports (7, 8, 17). In contrast, higher ACCI was not associated with inferior survival in the present study, unlike in the study by Yousaf et al. (7). This discrepancy may be related to differences in the study population, as exemplified by the lower ACCI scores in our study population (three versus eight).
Older studies have reported a dose–response relationship in STS (12-14). In contrast, there was no significant association between RDI and response or survival in the current study. This finding may be affected by the heterogeneity within the study sample. However, it does suggest that dose intensity may not be as important in palliative chemotherapy. This conclusion is supported by previous trials that have found no improvement in survival despite differences in tumor response (13, 19), as well as a study in gastroesophageal cancer that showed that less intensive therapy did not compromise survival in older patients (20).
The oldest patients in our study received fewer lines of treatment than 50- to 74-year-old patients (median of one versus two lines). Additionally, there was no significant association between survival and receiving more than one type of antitumor treatment for this group. This is a striking finding because this type of analysis is affected by selection bias; patients need to stay alive and in sufficient health after the detection of progression to be able to start second-line treatment. Therefore, patients who can start an additional line of treatment might be expected to do better, even on a placebo. The finding that for the oldest patients, there was no association between further lines of treatment and survival suggests that the treatment of this group was not as efficacious.
Our findings do raise important questions about the treatment of older patients with STS. However, there are some limitations that should be taken into account when interpreting these results: the analyzed patients were treated over a long time period during which two new drugs were approved for STS (trabectedin in 2009 and pazopanib in 2013). This, along with the variety of tumor subtypes and treatment regimens, causes heterogeneity in the sample, although it does mirror real-world settings. Incorporating information from comprehensive geriatric screening might possibly have been able to capture some of the sample’s heterogeneity, but during the accrual period, patients did not routinely undergo such screening at our hospital, and we were unable to include this in the analysis. Lastly, it was not possible to calculate the RDI for further lines of treatment because we were not able to reliably retrieve information of per oral drug dosage (mainly pazopanib), and a large proportion of patients received pazopanib after first-line therapy (19% in second-line and 29% in third).
Conclusion
Despite the limitations of this study, there are two important conclusions that can be drawn from these results. Firstly, the finding that patients aged ≥75 years had a greater risk of dying than younger patients highlights the need for research focused specifically on this population – especially as the number of older people is predicted to increase worldwide. This conclusion is supported by our finding that this population did not experience improved survival by receiving more than one type of antitumor treatment. Secondly, it is encouraging that the oldest patients experienced fewer hematological AEs. This supports the argument that older patients can tolerate chemotherapy when it is tailored for them.
Footnotes
Authors’ Contributions
Study concepts: RN, CB, and SJ. Data acquisition: JLe, JL, MS, and ME. Funding acquisition: RN, CB, and MS. Quality control of data: JLe, JL, and SJ. Data analysis and interpretation: JLe, SJ. Statistical analysis: JLe and CB. Article preparation: JLe and SJ. Article editing: JLe, JL, RN, and SJ. Article review: All Authors except for CB due to his unexpected passing.
Conflicts of Interest
The Authors have no relevant financial or non-financial interests to disclose.
Funding
This work was supported by grants from Helsinki and Uusimaa Hospital District and the Ministry of Social and Health Affairs of Finland (No: Y253S00030 and TYH2020203).
- Received April 14, 2024.
- Revision received May 15, 2024.
- Accepted May 20, 2024.
- Copyright © 2024 International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved.
