Abstract
Key statements
(i) What is already known about the topic? Soft tissue and bone sarcomas (STS and BS) affect the adolescent and young adult (AYA) population) and often require intensive treatments despite overall poor prognoses. Palliative care teams are multidisciplinary in nature, and these teams address physical, social, spiritual, and psychological suffering while also assisting in complex decision-making and end-of-life discussions. The AYA cancer population has well-defined challenges, including access to care and emotional and mental health needs, but palliative care interventions are not consistently integrated in AYA cancer care.
(ii) What this paper adds Most AYAs with an advanced/metastatic STS or BS diagnosis receive palliative care, but less than half of these patients establish care with a formal palliative care team within a year of their diagnosis. Median overall survival from the time of an advanced/metastatic diagnosis was 700 days for those with palliative care and 500 days for those without, while overall survival at 24 months from the time of advanced or metastatic diagnosis was 43.6% for those with palliative care and 29.2% for those without, although these differences were not statistically significant in this relatively small sample. Palliative care and medical oncology visits each addressed physical symptoms, while end-of-life discussions were more common in palliative care visits but under-documented overall.
(iii) Implications for practice Algorithmic-based models for palliative care referrals or default “opt-out’ palliative care referrals for all AYAs with STS or BS may help normalize early palliative care involvement and remove the negative connotations as a hospice-only specialty. The early presence of the palliative care team for AYAs with STS or BS could allow for the introduction and documentation of end-of-life topics, especially as the palliative care team includes experts in psychology and spiritual care who can help clarify motivations for location of death preferences and other concerns. The complementary nature of medical oncology and palliative care can serve as the introduction point for AYA patients with STS and BS and their families who may have reservations about palliative care involvement.
Introduction
The overarching goal of palliative care is to alleviate the suffering associated with serious illnesses for both patients and families, from the time of diagnosis through the bereavement period. 1 Palliative care teams are multidisciplinary in nature, and these teams address physical, social, spiritual, and psychological suffering while also assisting in complex decision-making and end-of-life discussions. 2 Given these stated aims, palliative care services are often recommended for patients with cancer, with prior work showing that early palliative care team involvement improves quality of life and even survival in patients with lung cancer. 3 Unfortunately, it is not uncommon for palliative care services to be offered too late in the disease course to truly be effective. 4
Sarcomas are a heterogeneous group of neoplasms arising from mesenchymal cells and can affect almost any anatomical site. Broad classifications of sarcomas are soft tissue sarcomas (STS) and bone sarcomas (BS). Despite this heterogeneity, both STS and BS often require intensive treatments, and many patients will either present with or develop metastatic disease with a low likelihood of cure.5,6 A combination of symptoms related to disease, distress and toxicities related to therapies, and poor prognosis seems to provide a clear indication for palliative care involvement. 7 Previous studies have characterized the use of palliative care in patients with STS and BS in both the inpatient and outpatient settings, with improved symptom burden and pain control and decreased stress level. 8 In addition, these studies have shown the differences between oncologic and palliative care encounters, including an emphasis on psychosocial care and coping with palliative care. 9
One notable demographic characteristic of both STS and BS is the burden of disease in the adolescent and young adult (AYA) population, with these diagnoses comprising 8% of all cancers in this group compared to ~1% of all cancers in adults. 10 The AYA cancer population (ages 15–39 at the time of diagnosis) has well-defined challenges, including access to care, fertility concerns, emotional and mental health needs, and emerging independence, all of which suggest a role for palliative care. 11 However, palliative care interventions are not consistently integrated into AYA cancer care. 12 AYAs with cancer are often at the transition of pediatric and adult care, and this transition is evident in palliative care service utilization, as less than one in five pediatric palliative care consultations involve a patient less than 18 years of age, and the average age of the patient seen by adult palliative care specialists is 60–65 years.13,14
The AYA patient with a diagnosis of STS or BS may be in an exceptional position to benefit from palliative care services, given age-related and disease-related concerns. In a prior institutional retrospective review of all patients with deaths related to STS or BS, most were referred to palliative care, but the outcome of this referral was often limited to hospice care. 15 This review also revealed that while medial overall survival (OS) from the time of last therapy was lower for AYAs than in older age groups, AYAs referred to palliative care had longer median survival compared to the older population receiving similar services.
Given this information, we sought to further explore palliative care intervention specifically in the AYA STS and BS population, both through additional granular retrospective review and a qualitative analysis approach to palliative care visits and medical oncology visits at various timepoints in the disease course with a goal to build a framework for collaboration between these specialties in the care of a unique population.
Methods
We first reviewed the electronic health record (EHR) of all patients ages 18–39 with a histological diagnosis of BS or STS receiving care at our sarcoma oncology clinic who died during the time period of April 1, 2019–March 31, 2024. Patients were excluded from the analysis if they only received a one-time treatment opinion or if death was unrelated to BS or STS diagnosis. A 5-year time period was chosen to allow for sufficient data collection.
Records were analyzed from the time of initial referral to the sarcoma oncology clinic until the time of death. Data collected included demographic, tumor-specific, and treatment-related information as well as information related to palliative care involvement (i.e., timing of palliative care involvement, location of contact, number of points of contact). Group comparisons were performed using a two-sample
First and third authors (female medical oncologist and male undergraduate) performed a qualitative analysis of clinical documentation from a randomly selected cohort of 10 patients who had more than one point of contact with palliative care (20 total). Clinical documentation was extracted from the EHR of each patient for each of the following timepoints: initial oncology and palliative care visits following an advanced/metastatic diagnosis, visits surrounding the time of progression of disease or treatment changes, and final oncology and palliative care visits. This study was approved by Michigan Medicine Institutional Review Board or IRB (approval #HUM00068553 on 11 May 2021).
Directed content analysis code book.
Next, investigators reviewed the extracted passages individually for identification of the predetermined coding categories and following individual review, met to achieve consensus through detailed conversation and discussion of discrepancies across coding. Extracted passages that could not be coded into one of the six predetermined categories were coded with another label that captured the essence of the passage. Finally, the investigators compared the extent to which the extracted passages of documentation were supportive of the predetermined categories (i.e., the six components of palliative care and medical oncology visits) versus how much the documentation represented different elements and, in these meetings, the authors achieved data saturation when continued review of coding revealed no new codes or variations on the predetermined codes. Reporting of findings included both the presentation of descriptive evidence in support of and against these predetermined categories, as well as the quantitative comparisons of frequency of codes between types of visits.24,25 Validity measures included the use of triangulation, member-checking with medical oncologists and PC clinicians, and peer debriefing with those external to the study. 26
Results
Quantitative
Hundred patients ages 18–39 with STS or BS known to the sarcoma oncology clinic died during the review period April 1, 2019–March 31, 2024. A total of 37 patients were excluded due to second opinion only, a diagnosis of a low-risk tumor or non-STS or BS-related deaths. As such, a total of 63 patient records were analyzed (Figure 1) with demographic information in Table 2.

Patient selection.
Demographics and tumor-specific details.
MPNST: malignant peripheral nerve sheath tumor.
Males were more represented than females in the AYA population (62% vs 38%), and Ewing sarcoma was the most common diagnosis. Of the 39 palliative care referrals placed, 3 were for hospice care and 8 patients received palliative care through the pediatric palliative care team. Regarding end-of-life care, 22 patients (35%) had an ICU admission in the last 30 days of life with 18 patients passing away in the ICU.
Regarding timing of palliative care referrals for AYAs, 44% (17 patients) received this referral within the first year of their advanced STS or BS diagnosis, while 49% (19 patients) received this referral in the final 60 days of life. The median time from diagnosis to palliative care referral was 395 days (interquartile range (IQR): 647 days). Younger patients in this age group (ages 15–29) were more likely to be referred to palliative care than those aged older than 29 years (OR: 1.6; 95% CI: 0.57–4.47;
Median OS from the time of an advanced or metastatic diagnosis was 700 days for those with palliative care (IQR: 567 days) and 500 days (IQR: 765 days) for those without (ratio of 1.40 with a 95% CI of the ratio 0.84–2.33, Figure 2(a)). Median OS from the time of last treatment was 43 days for those with palliative care (IQR: 78 days) and 44.5 days (IQR: 68.5 days) for those without (ration of 0.97 with a 95% CI of the ratio 0.58–1.61, Figure 2(b)). OS at 24 months from the time of advanced or metastatic diagnosis was 43.6% for those with palliative care (29.2% for those without,

Overall survival. (a) Overall survival from time of advanced/metastatic diagnosis and (b) Overall survival from time of last treatment.
Qualitative (adolescents and young adults)
Directed content analysis was applied to clinical documentation extracted from 10 patients who interacted with palliative care and medical oncology at the following timepoints: initial oncology and palliative care visits following an advanced/metastatic diagnosis, visits surrounding the time of progression of disease or treatment changes, and final oncology and palliative care visits. The six predetermined coding categories were represented in both visit types (Figure 3(a)). For palliative care visits, code frequency rankings (most mentioned to least) were physical symptoms, qualitative of life evaluation, addressing coping mechanisms, support network involvement, end-of-life discussions and prognostic awareness. For medical oncology visits, rankings were physical symptoms, prognostic awareness, quality of life evaluation, support network involvement, assessing coping mechanisms, and end-of-life discussions.

Code frequencies. (A) Coding category frequencies by type of visit and (B) Coding categories addressed per patient record.
In terms of individual patient records, all 10 records included coding for physical symptoms for both palliative care and medical oncology visits (Figure 3(b)). End-of-life discussions were noted in 8 of 10 palliative care encounters while these were coded in 2 of 10 medical oncology encounters. Emerging coding categories present in both types of encounters included communication with other medical teams/specialties and deference to patient wishes while discussions surrounding how to receive information or bad news were noted in only palliative care encounters. Representative extracted passages are shown in Table 3 and were chosen based on early identification by both authors upon record review suggesting consensus and adherence to the noted coding category.
Representative quotes.
Discussion
Main findings of the study
Adolescents and young adults (AYAs) with STS and BS represent a unique population in both the overall rarity of their disease and often poor outcomes as well as age-related psychosocial challenges such as emerging independence and developmental transitions. Palliative care, with its goals of addressing physical, social, spiritual, and psychologic suffering while also assisting in complex decision-making and end-of-life discussions, has the potential to benefit the AYA patient with STS or BS but may not be implemented in a timely manner.
What this study adds
Median OS from the time of an advanced STS or BS diagnosis was ~200 days longer for those who received palliative care. In addition, OS at 24 months also suggested improvement for patients with palliative care at any time. While these differences were not statistically significant in a small sample size, they do suggest a trend warranting additional investigation. Specifically, it may be that palliative care involvement, while not directly influencing sarcoma-directed therapy, provides an additional level of support that helps mitigate symptoms of treatment (pain, nausea, anorexia) or the psychosocial distress that may accompany the disease (anxiety, fear of death). Of note, less than half of the palliative care referrals took place within the first year following an advanced or metastatic diagnosis and close to half of the referrals were made within the final 60 days of life. Taken together, this suggests that palliative care may be viewed primarily as a hospice service, a common misperception among patients.27–29 Algorithmic-based models for palliative care referrals or default “opt-out’ palliative care referrals for all AYAs with STS or BS may help normalize palliative care involvement and remove the negative connotations as a hospice-only specialty.30–32
In our study, the presence of palliative care did not correlate with receiving sarcoma-directed treatment, such as chemotherapy, oral-targeted therapy, immunotherapy, radiation, or surgical intervention in the 30 days preceding death, and OS from the time of last treatment was not appreciably different between the two cohorts. This is in contrast to existing literature demonstrating that palliative care involvement may prolong time off of cancer-directed therapy at the end of life and therefore improve the quality of life during that time without the inherent adverse events and toxicities related to such treatment. 33 These results may be due to the AYA-specific nature of this study and the tendency to continue to offer additional lines of therapy to younger patients with fewer co-morbidities or a strong patient or family desire to pursue more experimental therapies and exhaust all treatment options.34–36
This study also highlighted the frequency of contact with the medical system in the final days of life for AYAs with STS or BS. Most notably, over half of these patients died in the hospital, with about 1/3 dying in the critical care setting. Location of death is an important factor in assessing the quality of end-of-life care patients with cancer.37–39 While most AYAs with cancer do prefer death at home, the more important metric is assessing this preferred location and striving to allow AYA patients the experience of goal-concordant care with respect to the location of death.40,41 The early presence of the palliative care team for AYAs with STS or BS would allow for the introduction of this assessment, as well as fluidity in choices as the disease course changes. The palliative care team includes experts in psychology and spiritual care who are well-suited to help AYAs with STS or BS clarify motivations for location of death preferences well as other end-of-life concerns.
Our qualitative analysis utilized a directed content analysis approach based on palliative care and medical oncology literature and revealed that while there is a degree of overlap in the topics addressed by medical oncology and palliative care for AYAs with STS and BS, there are noted differences that suggest a potentially synergistic role for each specialty. For example, medical oncology providers more frequently address prognostic awareness in documenting this routinely may allow the palliative care providers to better frame their discussions of coping mechanisms or end-of-life topics. The complementary nature of medical oncology and palliative care can serve as the introduction point for AYA patients with STS and BS and their families who may have reservations about palliative care involvement.42,43 The directed content analytic methodology allows for frequency determination, and future studies may incorporate other qualitative approaches to explore the coding categories and themes in more depth.
Our analysis also suggested an overall paucity of end-of-life discussions by both palliative care and medical oncology teams. This does not inherently mean that these discussions are not taking place outside of clinical documentation. However, lack of documentation can present communication challenges with medical teams outside the medical oncology/palliative care dyad, including emergency and critical care providers who are often involved in care when AYA patients may not be able to verbalize their end-of-life wishes.44,45 Additional qualitative analytic approaches outside of directed content analysis may be applied to the documentation passages in this study to better understand the nature of the conversations that focus on end-of-life care.
Strengths and limitations of the study
AYAs are an under-represented group in medical literature in general and when AYA patients with cancer are discussed in research, it is often as a homogenous group. A strength of this study is that it focuses on one group of AYA patients with cancer, specifically those with STS and BS. The AYA cancer population is heterogeneous, with needs differing based on disease state. This study gives a granular description of palliative care for AYAs with STS and BS, specifically, recognizing the nuances and unique needs of AYA cancer subsets.
Limitations of this study include the age spectrum represented. While the median age of death for patients in this study was 28 years old, only five patients (8%) were 18 years of age or younger. The AYA population includes patients aged 15–39, and this study may not fully reflect the palliative care experiences of pediatric patients. We also note that a single-center study presents limitations as both medical oncology and palliative care processes may vary across institutions. We acknowledge that this single center is one that operates within the context of the current United States healthcare structure and as such, the definitions of critical care, end-of-life care and hospice may not be generalizable to all healthcare systems. Finally, our qualitative analysis is limited by the nature of medical documentation, noting that not all aspects of a clinical encounter are captured in written documentation, including sensitive topics such as end-of-life planning. Future research would be strengthened by presenting the emergent themes of this study to patients living with STS and BS to assess their resonance and incorporating in-depth interviews with patients and families to further explore and confirm validity. This future research will also allow for the incorporation of additional qualitative methodologies to explore emergent themes outside of the predetermined coding categories presented herein.
Conclusion
This study explored palliative care intervention specifically in the AYA STS and BS population through both a detailed retrospective review and a directed qualitative analysis of palliative care visits and medical oncology visits. Integration of palliative care suggested potential improved survival from the time of an advanced or metastatic STS or BS diagnosis, despite initiation for many patients in the final 60 days of life, which for some included death in the hospital setting, including critical care units. Qualitative analysis revealed a lack of discussions surrounding end-of-life goals. Overall, this study suggests that AYAs with STS and BS may especially benefit from early engagement of palliative care services both in terms of OS and elucidation and respect for end-of-life wishes. Future directions will be aimed at promoting this early engagement through patient and provider education and optimization of referral patterns to identify vulnerable patients.
