Metastatic appendicular soft tissue sarcoma: treatment and survival outcomes of 2,553 patients from the SEER database

Statistical analysis

Demographic, clinical, and treatment characteristics were analyzed using descriptive statistics. Logistic regression with the Cox proportional hazards model was performed to determine the role of QT and RT on disease-specific survival. One-, two-, and five-year DSS were calculated using the Kaplan-Meier method. Differences in survival were compared using Log-rank analysis. A P value ≤ 0.05 was considered statistically significant. All statistical analyses were performed with Stata (StataCorp LLC, Texas, USA).

Patient demographic and clinical characteristics

We initially classified the tumors that most commonly presented with metastasis at diagnosis by histology [Figure 1A].

Figure 1. (A) Total number of patients with metastatic disease at presentation according to tumor histology. (B) Metastatic rate of each tumor histology. Dx: Diagnosis; UPS: undifferentiated pleomorphic sarcoma.

The six most common STS presenting with metastasis at diagnosis were leiomyosarcoma, undifferentiated pleomorphic sarcoma (UPS), liposarcoma, synovial sarcoma, spindle cell sarcoma, and alveolar rhabdomyosarcoma (A-RMS). We additionally analyzed the metastatic rates of these tumors by dividing the number of cases presenting with metastases by the total number of patients diagnosed with that STS histology. Alveolar soft part sarcoma and alveolar rhabdomyosarcoma had the highest metastatic rates, presenting with metastasis in 43.9% and 34.6% of cases, respectively [Figure 1B]. Leiomyosarcoma, which showed the highest number of metastases, had a metastatic rate of 10.3%.

For our demographic analysis, we compared patients with and without metastasis at presentation in the general STS population and for each of the six most common tumor subtypes. A total of 2,553 and 20,130 patients with and without metastasis at diagnosis were included. Sex distribution was similar in both groups, slightly favoring the male population (P = 0.89) [Table 1].

Sub-population patterns differed between the metastatic and non-metastatic groups. Afro-American patients accounted for only 11.1% of the non-metastatic group but 15.6% of the metastatic group (P < 0.001). The lower limb was the most common location in both groups and tumors located in the pelvis were more commonly in the metastatic (33.2%) than in the non-metastatic (16%) group (P < 0.001). Tumors in the metastatic group were significantly larger (median size: 111 mm) than those in the non-metastatic group (73 mm). Likewise, LN compromise was more common in patients with metastasis at diagnosis than in those without (21.9% and 2.3%, respectively). Median survival was 10 months for patients with metastatic STS and 48 months for those without metastasis (P < 0.001).

We conducted a separate analysis of demographic and clinical characteristics by histology to see if general trends remained stable across the most common tumor subtypes [Tables 2 and 3]

Metastatic patterns of the most common STS

In the entire cohort, lung was the most common site of metastases (77%), followed by bone (29.7%), distant LNs (17.8%), and liver (14.5%) [Table 4]. Metastases to other sites, which included any location other than the lung, bone, liver, brain, and distant LNs, occurred in 26.2% of cases.

Different metastatic patterns were seen when patients were stratified by histology. In alveolar rhabdomyosarcoma, only 45.9% of patients with metastasis at diagnosis had lung compromise (P < 0.001). Instead, patients with this tumor presented with bone metastases in 85.2% of cases; in all other histologic subtypes, this incidence ranged between 14.4% and 30.2% [Table 4]. Liver metastases ranged from 3.3% in alveolar rhabdomyosarcoma to 30.2% in leiomyosarcoma. Brain metastases were rare, with a frequency below 5 percent across all histologies.

Treatment outcomes in patients with and without metastatic STS

Surgery at the primary site was performed in 42% and 88.9% of patients with and without metastatic disease, respectively [Table 5]. Amputation was performed in 14% of patients with metastasis at diagnosis and 5.9% of those without (P < 0.001). Chemotherapy was administered in 60.4% and 19.5% of patients with and without metastasis at diagnosis (P < 0.001). Radiotherapy was given in 42.4% of patients with metastasis and 47% of those without (P < 0.001).

A regression analysis was performed to assess whether radiotherapy and chemotherapy were associated with improved disease-specific survival [Table 6]. On multivariate analysis, radiotherapy (HR = 0.788) and chemotherapy (HR = 0.755) were both associated with increased survival.

We compared DSS in patients with metastasis at diagnosis treated with radiotherapy or chemotherapy, in addition to surgery. Patients treated with surgery, RT and chemotherapy had a higher DSS than those treated with only surgery and chemotherapy [Figure 2A, P < 0.001]. The addition of chemotherapy to RT and surgery only showed a survival benefit at the 12- and 24-month marks [Figure 2B, P < 0.001]. No difference in survival between groups was seen at the 5- and 10-year follow-up marks.

Figure 2. Kaplan-Meier disease-specific survival for patients with metastatic soft tissue sarcoma according to (A) radiotherapy and (B) chemotherapy treatment. Differences were significant for both curves (Log-rank test: P < 0.001). QT: Chemotherapy, RT: radiotherapy.

Disease-specific survival according to metastatic patterns and histologic subtypes

Disease-specific survival by STS histology and metastatic pattern were additionally analyzed [Table 7].

Patients with alveolar rhabdomyosarcoma had a higher DSS than other STS subtypes [Figure 3] (P < 0.05).

Figure 3. Kaplan-Meier disease-specific survival for patients with metastatic soft tissue sarcoma according to tumor histology. UPS: Undifferentiated pleomorphic sarcoma.

Patients with metastatic UPS showed the worst survival pattern of all histologic subtypes. Five-year DSS for all histologic subtypes was less than 20%.

An additional analysis of the most common metastatic patterns was conducted. Five-year DSS was considerably lower in patients with metastasis to ≥ 2 organs. Five-year DSS was 5.6% for patients with metastases to the bone and lung and 7.2% for those with metastases to the liver and lung [Table 7]. A trend towards better DSS in patients with only liver metastasis was detected [Figure 4]; this was not significant in log-rank analysis.

Figure 4. Kaplan-Meier disease-specific survival for patients with metastatic soft tissue sarcoma according to metastatic pattern.

Patient demographic characteristics

Our study found that 8.9% of patients diagnosed with STS of the extremities and pelvis presented with metastasis at diagnosis. This is consistent with the 7% to 12% rate reported in the literature^[15,16]. Alveolar soft part sarcoma (ASPS) showed the highest rate of metastatic disease at presentation among all tumors. Although extremely rare, accounting for less than 1% of all STS, this tumor is extremely aggressive and 15% of patients with disseminated disease develop brain metastases^[17]. This was in contrast with liposarcoma, the second most common STS in our cohort, which displayed the lowest rate of metastasis at diagnosis (2.9%). However, due to the high frequency of this histology, metastases caused by liposarcoma ranked third in absolute incidence. The metastatic rate of liposarcoma found by our study was lower than in previous studies^[18]. In a long-term cohort of 133 patients with liposarcoma, Knebel et al. found that 22.6% developed metastases in the course of the disease^[18]. We consider that the lower rate found in our study is due to the SEER database only capturing metastatic events at diagnosis.

Age distribution favored patients in their late fifth decade of life for most STS subtypes. The exception was rhabdomyosarcoma, which is the most common STS in children and accounts for half of all STS and 3.5% of all tumors diagnosed in children^[19]. Interestingly, we found that patients with alveolar rhabdomyosarcoma presenting with metastases were significantly older than those without. Literature on alveolar rhabdomyosarcoma has reported that adults generally fare worse than pediatric patients^[20,21]. Within the pediatric population, a cohort study of 2,343 patients with A-RMS by Joshi et al. found that adolescents (≥ 10 years) had lower failure-free survival than children aged 1-9 years^[21]. In light of our findings, we consider this survival difference might be partially explained by a higher rate of metastasis in teenagers and adolescents.

Soft tissue sarcomas most commonly affect patients of European descent. However, we found that the frequency of Afro-American patients increased in the metastatic cohort. On histologic sub-analysis, differences in sub-population distribution patterns were seen in patients with leiomyosarcoma and UPS. Similar racial disparities in metastatic status have been reported in primary breast, colorectal, and prostate cancer^[22]. Although the underlying causes of racial disparities remain unclear, differences in socioeconomic background and access to healthcare have been reported to be important factors^[23].

Across all STS histologies, the lower limb was the most common location for patients without metastatic disease. Moreover, STS located in the pelvis increased in frequency in patients with metastasis at diagnosis. Pelvic sarcomas have already been recognized for their larger size at presentation, early involvement of adjacent structures and difficulty obtaining wide margins due to anatomic location^[24]. We consider that these factors might explain the higher rates of metastatic disease seen in pelvic tumors at presentation^[25].

Metastatic patterns of the most common STS

Metastatic soft tissue sarcomas (STS) typically affect the lungs as the first site of metastasis, accounting for about 70% of cases^[26]. However, some STS subtypes exhibit different metastatic patterns. For example, myxoid liposarcoma is known to present with extrapulmonary metastases in 85% of cases, while alveolar soft part sarcoma (ASPS) often metastasizes to the brain^[3,27]. In our study, only 45.9% of patients with metastatic STS showed lung involvement, while involvement of the bones and regional LNs was seen in 85.2% and 62.1% of cases, respectively. The extensive variability in metastatic patterns between STS subtypes reflects the group heterogeneity, especially with regards to molecular profile^[28]. Indeed, unlike other neoplasms such as carcinomas, soft tissue sarcomas encompass a varied array of tumors derived from a mesenchymal progenitor. Although current STS staging guidelines recommend chest imaging at diagnosis to rule out pulmonary metastases^[5], these guidelines are based on the most common STS types and may not capture the unique metastatic patterns of rarer subtypes. Thus, additional imaging studies tailored to each tumor's specific metastatic patterns should be considered after a histopathological diagnosis is made.

While much of the literature on metastatic STS focuses on lung metastasis, liver metastases also represent a significant concern. In our study, 14.5% of patients presenting with metastatic STS will present with liver metastases. This figure was even higher in leiomyosarcoma, where 30% of metastatic patients showed liver involvement. Furthermore, a study by Jaques et al. found that in 65 patients with STS with metastatic disease, 85% of those with liver metastases had leiomyosarcoma as primary tumor^[29]. While retroperitoneal and/or visceral leiomyosarcomas are commonly studied in metastatic STS, it is worth noting that appendicular leiomyosarcomas also frequently metastasize to the liver^[30]. In the setting of STS that tend to metastasize to the liver such as leiomyosarcoma, early detection of liver metastases is paramount due to the reported survival benefit of hepatic metastasectomy in soft tissue sarcomas^[31]. Although retrospective in nature, this nationwide study by Grimme et al. reported a 53.9% 3-year overall survival after hepatic metastasectomy^[31]. However, further prospective studies are required to compare surgical resection of hepatic metastases with available chemotherapy regimens.

Patients with alveolar rhabdomyosarcoma differed from all other tumor subtypes as they were more likely to present with bone metastases (85.2%) rather than lung metastases (45.9%). A study from the Intergroup Rhabdomyosarcoma Study group reported lung (47%), bone marrow (38%), bone (34%), and distant LN (26%) as the most common metastatic sites in this tumor^[32]. Furthermore, they found that patients with bone metastasis had a lower 3-year event-free survival (14%) than those with lung metastases (24% for combined lung and extrapulmonary metastasis). Although our study did not identify a survival difference between patients with isolated bone or lung metastases, our analysis was restricted to metastasis at the time of diagnosis and may not represent the course of all patients who develop bone metastasis.

Treatment outcomes in patients with and without metastatic STS

Limb-sparing surgery has been widely adopted as the standard treatment for localized, appendicular STS^[33]. In metastatic STS, however, surgery to the primary site is not often performed and treatment usually takes a palliative approach. This is reflected in our findings, with only 42% of patients with metastatic disease at presentation having surgery at the primary site. Likewise, surgery to the distant site is rare in this population as a clear survival benefit has only been demonstrated in patients with metachronous lung metastasis^[34]. In our study, we found that only 11.7% of patients with metastatic disease at diagnosis had surgery at the distant site. Of 1,503 patients with metastatic disease, 77% presented with lung metastasis and only 11.7% underwent surgery at this distant site.

Radiotherapy was used in 42.4% of patients with metastatic disease and was associated with longer survival (HR = 0.755). Although the use of radiation was almost as common in patients with metastases as in those without, we consider that the scope of treatment was different. Neo-adjuvant radiotherapy in localized appendicular STS is an effective technique to optimize tumor control in the primary site^[9]. In metastatic disease, however, radiation therapy is often used for palliative reasons or local control of metastasis in patients not suitable for surgery^[35]. Our analysis was restricted to the assessment of treatment patterns and due to the retrospective nature of the database, no cause-effect conclusions can be inferred.

Our study also reported that long-term DSS was low for patients with metastatic STS regardless of the use of chemotherapy [Figure 2B]. Although higher 1-year and 2-year survival rates were seen in patients receiving chemotherapy, 10-year DSS remained below 25% in both groups. Our findings emphasize how, unlike in other malignancies such as breast carcinoma, systemic treatment falls short of achieving more reliable disease remission or cure^[36]. This can be partially explained by the lack of major breakthroughs in systemic treatment, which has relied on the combination of doxorubicin and ifosfamide for the last 30 years^[8].

Limitations

Our study presented several limitations, primarily due to the nature of the SEER database. First, analysis of metastatic disease was restricted to metastases detected at diagnosis. As the majority of metastases present in the course of the disease, our findings cannot be generalized to all patients with metastatic STS. Second, analysis of metastases by location was limited to the sites included in the database: lung, liver, brain, bone, and distant LN. Third, there was no available information on the chemotherapy and/or radiotherapy treatment plans and the intent (curative vs. palliative) of therapy. Fourth, as the variable “surgery at the distant site” did not specify the exact surgical location in patients with multiorgan metastatic disease, the site of surgery could not be confirmed.

Conclusions

Extensive heterogeneity in rates of metastatic disease at presentation and organ involvement patterns exists between different STS histologies. Although most STSs present with metastasis involving the lung, alveolar rhabdomyosarcoma most commonly affects the bones and lymph nodes. Disease-specific survival remained poor for patients with metastatic disease at presentation regardless of (neo)-adjuvant radiotherapy or chemotherapy.