Fax: (313) 576-8699
Patterns of care in a population-based sample of soft tissue sarcoma patients in the United States
Article first published online: 24 APR 2009
Copyright © 2009 American Cancer Society
Volume 115, Issue 12, pages 2744–2754, 15 June 2009
How to Cite
Gadgeel, S. M., Harlan, L. C., Zeruto, C. A., Osswald, M. and Schwartz, A. G. (2009), Patterns of care in a population-based sample of soft tissue sarcoma patients in the United States. Cancer, 115: 2744–2754. doi: 10.1002/cncr.24307
- Issue published online: 4 JUN 2009
- Article first published online: 24 APR 2009
- Manuscript Accepted: 3 DEC 2008
- Manuscript Revised: 19 NOV 2008
- Manuscript Received: 22 SEP 2008
- soft tissue sarcomas;
- patterns of care
Soft tissue sarcomas (STS) are relatively uncommon tumors. Data regarding the patterns of care of patients with STS and its consistency with available guidelines are relatively sparse. The authors conducted a detailed analysis of STS patients diagnosed in 2002 and sampled from the Surveillance, Epidemiology, and End Results registries.
The authors sampled 1369 patients with invasive sarcomas. Hospital records were reabstracted, and treating physicians were contacted to verify the therapy provided to each patient.
The median age of patients was 60 years. There was a slight male predominance among the patients with nongynecologic sarcomas. Fifty percent of the patients had localized stage sarcoma. Most patients received surgery, but negative margins were obtained in only 50% of patients. Complete resection was less frequent in patients ≥50 years old. Radiation therapy was used in 53% of patients with extremity sarcomas but in only 20% to 30% of the patients with sarcomas at other sites. About 27% of all patients received chemotherapy. Tumor grade was significantly associated with the use of radiation and chemotherapy. Surgical resection, tumor grade, tumor size, use of radiation therapy, and age significantly influenced survival.
Patterns of care of STS differ based on the site of the tumor. The patterns of care for extremity sarcomas are fairly consistent with the available recommendations, but the patterns of care for other sites are less consistent. In addition to certain tumor characteristics, age of the patient was significantly associated with therapy and patient outcome. Cancer 2009. © 2009 American Cancer Society.
Soft tissue sarcomas (STS) are uncommon cancers, representing <1% of all malignant neoplasms, with an annual incidence of approximately 10,390 cases and about 3680 deaths per year.1 Management of these tumors requires a multidisciplinary approach.
Guidelines for the management of STS have been published by different institutions, including the National Comprehensive Cancer Network.2 Separate management guidelines have been provided for different anatomic sites.3 For extremity sarcomas, the guidelines recommend surgery for low-grade tumors, with postoperative radiation recommended for tumors >5 cm. In high-grade tumors, radiation with or without chemotherapy is recommended in addition to optimal surgical resection. Limb salvage is considered an important objective and generally guides treatment decisions.
For nonextremity sarcomas, surgical resection is considered the primary therapy. Postoperative radiation is recommended in patients with high-grade sarcomas. In unresectable tumors, multiple different approaches are suggested, including observation, palliative surgery, radiation, and chemotherapy.
Management of STS can be challenging because of differences in anatomic sites and the rarity of the tumors. Treatment is provided by physicians from different disciplines, increasing the variability in therapy. Data regarding patterns of care of STS from the United States are relatively sparse,4-7 and most data are from case series from major cancer centers with expertise in treating sarcomas.8-11 These data sets may not reflect the general patterns of care. In addition, many of the reports have presented analysis of all STS together or focused on extremity sarcomas.
We therefore conducted this analysis of STS patients using data from a population-based sample of the patterns of care study by the National Cancer Institute (NCI) to determine the patterns of care and assess the outcomes of these patients by anatomic site.
MATERIALS AND METHODS
The NCI's Surveillance, Epidemiology, and End Results (SEER) is a population-based cancer registry system that collects comprehensive information on demographics, tumor characteristics, and therapy and maintains follow-up of patients for vital status on all cancers occurring in defined geographic region. Because data collection is primarily conducted in hospitals, surgical centers, and radiation facilities, the systemic therapy information is incomplete. The NCI annually conducts the patterns of care studies to obtain detailed information on therapy for selected cancers. In 2002 patients with invasive sarcoma, histology codes 8800-8991, 9040-9044, were identified by participating registries (the states of Connecticut, Iowa, New Mexico, New Jersey, Utah, and Louisiana and the metropolitan areas of San Francisco/Oakland, San Jose/Monterey, Seattle/Puget Sound, Atlanta, Detroit, Los Angeles County, and the remaining area of the State of California). Patients aged <20 years, diagnosed at autopsy or on the death certificate, or who had been previously diagnosed with cancer, other than nonmelanoma skin cancer, were ineligible. The remaining patients were stratified by registry, race/ethnicity, and sex, and a random sample of cases was selected. We oversampled non-Hispanic black and Hispanic patients and selected all cases of Asians, Pacific Islanders, American Indians, and Alaskan Natives to obtain more stable estimates.
Medical records were reabstracted, and each patient's treating physician was asked to verify the therapy given. The physicians were also asked to identify any other provider who might have treated the patient. These providers were then contacted. A centralized training was provided for the individual at each registry who was primarily responsible for conducting the study to ensure consistency of abstracting and coding of data. A 5% sample of case records were reabstracted for quality control. Comorbid conditions documented in the hospital medical record at the time of the most definitive therapy were recorded and coded centrally by a single registered health information technologist. The data were analyzed separately based on tumor location, because this is known to influence therapeutic recommendations.
Data analyses were performed using SAS and SUDAAN (Research Triangle Institute, Research Triangle Park, NC) and were weighted to reflect the SEER population from which the data were obtained. We present adjusted percentages of patients receiving radiation or chemotherapy. Adjusted percentages were obtained from the estimates of the probability for each individual receiving treatment in the logistic regression models. These percentages were directly standardized to the distribution of the covariates among the entire weighted sample.12 Cancer-caused mortality was analyzed using Cox regression models based on a follow-up through December 31, 2005. The statistical significance tests were assessed using the Wald-type F statistics. All P values were 2-sided.
A total of 1369 patients were selected: 345 extremity sarcomas, 304 gynecologic sarcomas, and 720 sarcomas of other sites (Table 1). The median age at diagnosis was 60 years. The tumor size was smaller in extremity sarcoma patients. Patients with gynecologic sarcomas more often had advanced stage and patients with extremity sarcomas more often had localized stage at diagnosis. Among patients who underwent surgery, lymph node dissections were performed in 4%, 24%, and 45% of patients with extremity, all other sites, and gynecologic sarcomas, respectively.
|Extremity||GYN||All Other Sites||Total|
|No. of Cases||345||304||720||1369|
|Other histology code||8.5||45.5||24.6||25.5|
|SEER historic stage|
|Well or moderately differentiated||27.9||18.3||24.5||23.9|
|Poorly or not differentiated||43.1||49.3||27.9||36.8|
|Not stated or N/A||29.0||32.5||47.5||39.3|
|Lymph nodes status|
|No node dissection done||94.4||59.3||78.9||78.1|
|All nodes examined negative||3.7||31.0||15.4||16.2|
|Size of primary tumor|
|Anatomic depth of tumor|
Most patients had surgery (Table 2). Radiation therapy was used more commonly in patients with sarcoma of the extremities. Doxorubicin was the most commonly administered chemotherapeutic agent for patients with sarcoma of the extremity and ifosfamide for gynecologic sarcoma, whereas imatinib was given to slightly >10% of patients with sarcomas of other sites.
|Extremity||GYN||All Other Sites||All|
|Margins free of tumor||61.9*||53.2||46.4||51.9|
|Tumor at margins||23.3||9.0||18.7||17.6|
Localized and Regional Stage Extremity Sarcoma
Treatments received varied by tumor characteristics (Table 3). Overall, >42% of the extremity sarcoma patients had surgery only. Amputation was performed in 5% of the patients. Patients with clear margins were more often treated with adjuvant radiation than patients with other margin status. Patients with tumors at the margins were more often treated with adjuvant radiation plus chemotherapy. More than half of the patients with tumors <5 cm received surgery alone, whereas adjuvant radiation therapy and chemotherapy were used more frequently in patients with larger tumors. Patients with well or moderately well differentiated tumors were most often treated with surgery alone (57%), although 39% had radiation in addition. Patients with poorly or undifferentiated tumors had more intensive therapy, with surgery and radiation (42%) or surgery, radiation, and chemotherapy (31%).
|Region||Therapy||Pathologic Margins||Tumor Size||Grade/Differentiation|
|No Resection||Clear||Tumor Margin||Not Stated||≤5 cm||5.1-10||>10 cm||Not Stated||Well or Moderate||Poor or Undifferentiated||Unknown||All|
|All other sites||None||41.0||0.0||0.0||0.0||0.9||3.2||4.7||3.8||1.6||4.1||3.1||2.9|
|Unknown if surgery||0.0||0.0||0.0||1.0||0.0||0.3||0.0||0.7||0.0||0.0||0.4||0.2|
Localized and Regional Stage Gynecologic Sarcoma
Of patients with localized or regional gynecologic sarcomas, 56% with clear margins had no treatment beyond surgery. Nearly 29% of patients with tumor at the margins received adjuvant radiation and chemotherapy. Women with tumors >10 cm were more often given adjuvant chemotherapy or radiation therapy, but only 1% of patients with large tumors received all 3 modalities. Striking differences were seen by tumor grade. More than 88% of women with well or moderately well differentiated tumors had surgery alone, compared with only 38% patients with poorly or undifferentiated tumors.
Localized and Regional Stage Sarcoma of All Other Sites (Not Extremity or Gynecologic)
Nearly 70% of patients with tumors of other sites who had clear margins received only surgery. Even in patients with tumor at the margins, 56% received only surgery. As the size or grade of the tumor increased, the percentage of patients receiving surgery plus chemotherapy increased.
Radiation and Chemotherapy
The use of radiation varied by site and stage of disease (Table 4). Patients with sarcoma of an extremity were more likely to receive radiation than were the other 2 sites. The use of chemotherapy increased as stage increased. In each of the sarcoma sites, patients with poorly or undifferentiated tumors were more likely to receive chemotherapy.
We used multiple logistic regression analysis to determine factors associated with radiation and chemotherapy by sarcoma site (Table 5). The grade of the tumor was associated with the use of radiation therapy for all 3 sarcomas sites, whereas pathologic margin status and depth of the tumor were not. Larger tumors were associated with less radiation use in patients with sarcomas of other sites. Age of younger than 50 years was associated with significantly less radiation use for patients with extremity sarcomas; however, the opposite was true for sarcoma of other sites.
|Pred Prob||P||Pred Prob||P||Pred Prob||P||Pred Prob||P||Pred Prob||P||Pred Prob||P|
|Age at diagnosis, y||.01*||.76||.006*||.04*||.25||.43|
|Chemotherapy or radiation||.002*||.21||.001*||.0001*||.23||.001*|
Chemotherapy use in extremity sarcomas was associated with tumor depth, grade, age, and radiation. Women with gynecologic sarcomas were given chemotherapy more often if they had any residual tumor at the margins, or large, poorly differentiated, or undifferentiated tumors. Patients with sarcomas of other sites were treated with chemotherapy more frequently if they did not have a surgical resection, or had large or poorly differentiated tumors. A Charlson comorbidity score of 1 or more was associated with less chemotherapy in patients with gynecologic sarcomas and sarcomas of other sites. Patients with extremity sarcomas and sarcomas of other sites who had radiation were more likely to receive chemotherapy.
The maximum follow-up on these patients was 47 months. Survival was associated with tumor size, grade, pathologic margins, and age in all 3 sites of sarcomas (Table 6). For all sites, not being able to undergo surgical resection was associated with worse survival. Patients with tumors that were poorly differentiated or undifferentiated had poor survival, as did patients aged 50 years or older. Radiation therapy was associated with improved survival in extremity and gynecologic sarcomas. However, chemotherapy use was not associated with decreased risk of cancer death.
|P||Hazards||95% CI||P||Hazards||95% CI||P||Hazards||95% CI|
This is the first study to examine treatment of sarcoma in a large population-based data set. Grade and size of the tumor are recognized prognostic factors and are important variables in determining the stage of STS. However, in this dataset a high proportion of patients did not have the grade or size of the tumor reported. This was particularly true of gynecological sarcomas and sarcomas of other sites. These data are similar to a report from the National Cancer Database on STS, which reported unknown grade in >50% of the patients and unknown stage in >30% of the patients.7 The same data set also reported a decline in the percentage of patients with unknown grade from 1988 to 1993. It is possible that underreporting of grade and size of tumor in STS in sites other than extremities has less impact on therapeutic decisions than in extremity sarcoma. Lymph node metastases in STS are known to be uncommon except for certain histologic subtypes such as rhabdomyosarcomas, and lymph node sampling or dissection is not recommended. Lymph node assessment occurred more frequently in gynecological sarcomas than sarcomas of other sites.
In extremity sarcomas, 62% of patients underwent surgery with complete resection. This is lower than the complete resection rate of 80% reported in a series from a single institution by Weitz et al.8 Limb salvage is an important objective of the management of extremity sarcomas, and in this series only 5% underwent an amputation. However, it is unclear if complete resection of the tumor was not pursued in some patients to achieve the objective of limb salvage.
On the basis of data from randomized trials, adjuvant radiotherapy is recommended in most patients with extremity STS.13, 14 The rate of radiation therapy in our data is similar to that reported in other series.15, 16 Radiation was used more frequently for higher grade tumors and trended toward higher use in larger tumors, but was not associated with margin status.
Only 24% of extremity sarcoma patients received chemotherapy, and even among patients with distant stage sarcomas, chemotherapy was used in <50% of the patients. In the series by Weitz et al, which focused on extremity sarcomas, chemotherapy use varied at different time periods from 9% to 32%.8 Chemotherapy use in extremity sarcomas was significantly influenced by tumor depth and tumor grade, as suggested by the guidelines. Chemotherapy was also associated with higher use of radiation, probably reflective of the use of chemotherapy in patients with adverse prognostic features. Data regarding benefits of adjuvant chemotherapy in STS patients are disparate. Results of 2 recent trials evaluating the combination of an anthracycline and ifosfamide differ regarding the benefits of adjuvant therapy.17, 18 Previous meta-analysis has suggested disease-free survival advantage from adjuvant chemotherapy, and the most recent meta-analysis that included trials with ifosfamide-based regimens showed both relapse-free and overall survival.19, 20 The guidelines do recommend consideration of adjuvant chemotherapy, particularly in high-grade tumors. In advanced disease, data demonstrating the benefits of chemotherapy in STS are limited.21 The limited benefits of adjuvant chemotherapy and the paucity of data regarding the benefits of chemotherapy in advanced stage STS may have contributed to the limited use of chemotherapy in these patients.
Age of the patient was significantly associated with radiation and chemotherapy in extremity sarcomas, with more radiation and less chemotherapy in older patients. The lower use of chemotherapy in elderly patients may be explained by concerns about chemotherapy, particularly agents such as doxorubicin. However, the reasons for age-based disparity in the use of radiation are unclear. The current guidelines do not distinguish therapeutic recommendations based on age of the patient. The number of patients with STS who are older is expected to rise in the coming years, with the aging of the population. It is therefore imperative that available options for systemic therapy are applicable in the older patient population with STS.
Among patients with sarcomas of other sites, only 46% underwent complete resection, and 18% had no surgery. Case series reports of retroperitoneal and deep trunk STS from different institutions report a negative margin resection rate of about 45%.11, 22 Surgical resection at sites such as the retroperitoneum can be challenging because of the proximity of the tumors to surrounding vital organs and the large size of the tumors at diagnosis. Despite a higher proportion of patients with sarcomas of other sites not undergoing surgery, radiation was used only in 23% of the patients. This lower use of radiation may also be related to difficulty in delivering radiation at these sites compared with the extremities, as suggested by the significantly lower use of radiation in patients with larger tumors. Local relapse rates are known to be higher in patients with sarcomas at other sites,22 probably because of limitations in delivering local therapeutic modalities. Some reports have suggested that preoperative radiation, at sites such as the retroperitoneum, may be more feasible because of less toxicity.23, 24 Intraoperative radiation has also been used in patients with retroperitoneal sarcomas to improve efficacy and reduce toxicity. However, the availability of such techniques is limited to specialized centers.
The percentage of patients with sarcomas at other sites receiving chemotherapy is lower than that reported in single institution case series.10, 22 Chemotherapy use in these patients was associated with poor prognostic features, such as lack of resection and larger tumor size. In many patients, site and tumor size preclude surgery and radiation, making chemotherapy the only therapeutic option. In recent years, the introduction of imatinib mesylate has improved the outcomes of patients with gastrointestinal stromal tumors. However, the response to chemotherapy in these sites is generally limited. Thus, investigation of novel drugs for the management of these tumors is necessary, because drug therapy may be the only option in these patient.
Gynecological sarcomas were analyzed as a separate category, because the location of the tumors and the physicians involved in the care of these patients differ from STS at other sites. Complete resection was achieved in 53% of the patients, whereas 11% did not undergo resection. The available literature regarding complete resection of gynecological sarcomas is extremely limited; therefore, we cannot determine whether this complete resection rate is comparable to others. Thirty-one percent of gynecological sarcomas received radiation, a rate similar to reports on uterine sarcomas from single institutions.25, 26 The only factor that predicted use of radiation was tumor grade. Tumor margin status was not significantly associated with radiation use. Adjuvant radiation has been shown to reduce local recurrence in retrospective reviews of uterine sarcomas and is recommended in resected patients. Data regarding use of chemotherapy is also limited, although the guidelines do recommend its use in patients with high-grade tumors and metastatic disease. In this analysis, chemotherapy use was associated with adverse prognostic features such as residual tumor after surgery, larger tumors, and high-grade tumors.
Interpretation of the survival analysis is limited to the available follow-up time for these patients. Surgical resection, tumor size, and tumor grade were associated with cancer-specific survival in each of the sites. These are recognized prognostic factors for STS, although tumor grade has not been consistently shown to be associated with survival in sarcomas at sites other than the extremities.11, 27 Age >50 years was also a significant factor in cancer-related death for sarcomas at all sites. Patient age has been reported as a factor for local control and disease-specific survival in other series.9, 16 Specific reasons for this age-related difference in outcomes of patients with STS are unclear. In this analysis, age did influence use of different treatment modalities and therefore may have resulted in inferior survival in these patients.
Radiation was associated with improved survival in patients with extremity and gynecological sarcomas. Previous results have shown that radiation does decrease local recurrence but does not increase survival in extremity sarcomas.13 However, patients with local recurrence are known to be at a higher risk of developing distant metastases, an important cause of death. Therefore, improved local control observed with radiation may influence long-term survival. The rate of complete resection in clinical trials and in retrospective reviews from centers with expertise in treating sarcomas may be higher than the complete resection rate in this analysis. It is possible that use of radiation may influence cancer-related survival in patients who have less than adequate surgery, and therefore use of radiation was associated with improved survival in this analysis. Use of radiation did not influence survival in patients with sarcomas at other sites.
There are certain limitations of these data. Although SEER is an integration of population-based, quality-controlled registries that cover 26% of the US population, the data from SEER are still only a sample of adult sarcoma cases occurring in the United States. The SEER program is comparable on measures of poverty and education, but is slightly more urban and has more foreign-born individuals than the general US population. The registry does contain all sarcomas occurring in a defined geographic area, and therefore may better reflect the sarcoma patient population than hospital-based registries, which are susceptible to referral patterns. We therefore believe that this data set is a good reflection of the actual patterns of care of sarcoma patients in the United States.
In summary, this analysis reveals that patterns of care for STS differ significantly based on the site of the tumor. For extremity sarcomas, the patterns of care are fairly consistent with the recommendations, but the patterns of care in other sites are less consistent. In addition, age of the patient was a significant factor in determining therapy and outcome. Educational efforts directed toward adherence to management guidelines for physicians treating sarcoma might improve outcomes. Referral of patients with sarcoma to centers of excellence with a multidisciplinary team might further improve outcomes of these patients.
Conflict of Interest Disclosures
Supported in part by grants N01-PC-35,133, N01-PC-35,135, N01-PC-35,141, N01-PC-35,136, N01-PC-35,137, N01-PC-35,138, N01-PC-35,139, N01-PC-35,142, N01-PC-35,143, N01-PC-35,145, N01-PC-54,402, N01-PC-54,404, and N01-PC-54,405.
- 2National Comprehensive Cancer Network. NCCN Clinical Practice Guidelines in Oncology. Available at: http://www.nccn.org/professionals/physician_gls/PDF/sarcoma.pdf Accessed on April 13, 2009.
- 3National Comprehensive Cancer Network. Uterine neoplasms. Available at: http://www.nccn.org/professionals/physician_gls/PDF/uterine.pdf Accessed on April 13, 2009.
- 18Adjuvant chemotherapy (CT) with doxorubicin and ifosfamide in resected soft tissue sarcoma (STS): interim analysis of a randomized phase III trial [abstract]. J Clin Oncol. 2007; 25( 18 suppl). Abstract 10008., , , et al.
- 19Adjuvant chemotherapy for localised resectable soft tissue sarcoma in adults. Sarcoma Meta-analysis Collaboration (SMAC). Cochrane Database Syst Rev. 2000; 2: CD001419.
- 21Doxorubicin-based chemotherapy for the palliative treatment of adult patients with locally advanced or metastatic soft tissue sarcoma. Cochrane Database Syst Rev. 2003; 3: CD003293., , , et al.