Fax: (416) 946-4442
High-risk extracranial chondrosarcoma†
Long-term results of surgery and radiation therapy
Article first published online: 18 JAN 2011
Copyright © 2011 American Cancer Society
Volume 117, Issue 11, pages 2513–2519, 1 June 2011
How to Cite
Goda, J. S., Ferguson, P. C., O'Sullivan, B., Catton, C. N., Griffin, A. M., Wunder, J. S., Bell, R. S., Kandel, R. A. and Chung, P. W. (2011), High-risk extracranial chondrosarcoma. Cancer, 117: 2513–2519. doi: 10.1002/cncr.25806
Presented at the 51st Annual Meeting of the American Society for Therapeutic Radiology and Oncology, Chicago, Illinois, November 1-5, 2009.
- Issue published online: 19 MAY 2011
- Article first published online: 18 JAN 2011
- Manuscript Accepted: 12 OCT 2010
- Manuscript Revised: 25 SEP 2010
- Manuscript Received: 21 JUL 2010
- radiation therapy;
A study was undertaken to evaluate results of surgery and radiotherapy (RT) for high-risk extracranial chondrosarcomas.
Between 1986 and 2006, 60 patients underwent surgery and RT for extracranial high-risk chondrosarcoma. Preoperative RT (median, 50 gray [Gy]) and postoperative RT (median, 60 Gy) were used in 40% and 60% patients, respectively. Sites included pelvis/lower extremity (48%), chest wall (22%), spine/paraspinal (17%), and head and neck (13%). Overall, median tumor size was 7 cm (range, 1-22 cm), and tumor grade was I, II, and III in 22%, 64%, and 14% of cases, respectively.
Pathologically clear surgical margins (R0) were present in 50%, microscopic positive margins (R1) in 28%, and gross positive margins (R2) in 13%, half of whom had clinically detectable residual disease; surgical margin was unknown in 8%. Median follow-up was 75 months (range, 5-230 months). The crude local control rate was 90%. Patients with R0, R1, and R2 resections had local control of 100%, 94%, and 42%, respectively. Of the 8 cases that had R2 resection, 3 experienced uncontrolled progression, but 5 patients had stable disease with long-term follow-up. The 10-year overall survival, progression-free survival, and cause-specific survival were 86%, 80.5%, and 89.4%, respectively. Younger age and grade III tumors were associated with worse progression-free survival (P = .03 and .0003, respectively).
Although surgery with complete resection is paramount in management of chondrosarcoma, RT is a useful adjuvant treatment and appears to offer excellent and durable local control where wide surgical resection is difficult to accomplish. Cancer 2011. © 2011 American Cancer Society.
Low- and intermediate-grade disease in patients with chondrosarcoma may have a relatively indolent natural history, with a low risk of metastasis. In such patients, optimal local control assumes much significance, as it is local disease that is the cause of morbidity. In those with high-grade conventional chondrosarcoma, and relatively uncommon mesenchymal chondrosarcoma, a rapid growth rate is often seen, and metastatic disease also is more likely to occur.1-6 The prognosis of chondrosarcoma patients depends not only on the tumor site, type of surgical resection, and the resulting resection margins, but also on the grade of the tumor.7-11 Management of chondrosarcoma is challenging, considering that they have been reported to generally respond poorly to other cancer treatments such as chemotherapy and radiotherapy (RT).8, 12, 13 Local disease control of chondrosarcoma is obtained by surgical excision with wide margins, which is considered to be the most effective treatment for these tumors.7 However, the location or the size of the primary tumor may cause surgical resection with such margins to be complicated or possible only with excessive morbidity. Moreover, certain adverse features such as grade or positive resection margins14, 15 may lead to a reduction in local control, and thus further therapy may be required in such cases. In this regard, radiation treatment can be considered as an attractive adjunctive therapy in these so-called high-risk situations. There is no standard definition for what would be considered high-risk chondrosarcoma, but lesions located at complex sites where complete resection is an anticipated problem and/or lesions with close or involved surgical margins and/or high-grade lesions could be considered as such. In this regard, cranial or skull-base chondrosarcoma can be considered a high-risk lesion based on tumor location and difficulty with radical resection. The use of RT, either in the adjuvant or primary setting, has been established as standard care.16-18 In addition, the combination of surgery followed by adjuvant RT in extremity soft tissue sarcomas has yielded dividends in optimizing cure rates and quality of life.19, 20 Studies have shown this strategy to have improved local control with limb preservation compared with surgery alone.20-24 The same strategy in extracranial chondrosarcoma using both surgery and RT in those at high risk of local recurrence when treated with surgery alone may yield similar dividends.
The objective of this retrospective report was to assess the outcome of patients with high-risk extracranial chondrosarcoma treated with combined surgery and RT.
MATERIALS AND METHODS
After research ethics board approval, we retrospectively analyzed the charts of 310 consecutive patients in our database with biopsy-proven chondrosarcoma between 1986 and 2006. Patients who had intracranial or skull base disease were excluded from the analysis, except for 1 patient who had primary disease in the maxillary sinus and nasopharynx with extension to the skull-base who was categorized under primary head and neck chondrosarcoma. Patients presenting with recurrent disease were also excluded. Staging studies included routine radiography, computed tomography, magnetic resonance imaging, and bone scan as appropriate for the primary site and to exclude metastatic disease. High-risk chondrosarcoma was defined as follows: lesions at complex locations where complete resection was an anticipated problem and/or resection with close/involved margins and/or high-grade lesions. Combination of 1 or more of the above factors was considered for treating patients with combined modality therapy. The timing of RT depended on the clinical assessment of the treating surgical oncologist in conjunction with the radiation oncologist. Patients underwent preoperative RT in situations where it was anticipated that there was a high probability of a positive resection margin, particularly where more extensive surgery would result in excess morbidity. Otherwise postoperative RT was delivered. Although grading is important for prognostication, the system of grading chondrosarcomas is intrinsically subjective.14 For the purpose of analyzing outcome, we condensed the grading system into a broader 2-tier system, defining low grade as grade I and II, and high grade to be grade III.
Patients were followed regularly at 3- to 6-month intervals with clinical examination, chest radiographs, and appropriate imaging of the primary anatomic site, as needed.
The endpoints considered were progression-free survival (PFS), local control, overall survival (OS), and cause-specific survival (CSS). Time was calculated from the date of completing combined modality therapy to the event of interest. For OS, death from any cause was considered. For CSS, death as a result of disease or treatment complication was considered. For PFS, time to first failure or death from disease was taken into consideration. The survival plots for PFS, CSS, and OS were estimated using the Kaplan-Meier method. The differences in survival estimates were tested using log-rank test. Statistical significance was considered at P < .05.
Of the 310 patients, 60 with extracranial chondrosarcomas received combined modality therapy between 1986 and 2006; their characteristics are shown in Table 1. The median age at diagnosis was 42.5 years (range, 18-84 years). Male preponderance was evident, with a male:female ratio of 2:1. Of the 24 patients who had lower extremity tumors, 11 had true lower extremity lesions, whereas 13 had pelvic lesions. Of the 5 patients with upper extremity lesions, 3 had lesions in the scapula, whereas 1 each had the primary tumor located in the clavicle and radius, respectively. Pure extraskeletal chondrosarcoma occurred in 11 patients. Patients with primary head and neck chondrosarcoma had diverse subsite involvement (nasal septum 2, larynx with hyoid 2, maxilla 2, and mandible 2).
|Characteristics||Patients, N = 60|
|Age at diagnosis, y|
|Tumor size (maximum dimension), cm|
|Head and neck sites||8||13%|
|Grade of tumor|
|Margin status after surgery|
|Resection with clear margins (R0)||30||50%|
|Resection with microscopic margins (R1)||17||28.3%|
|Resection with gross residual disease (R2)||8||13.3%|
|Resection status not known||5||8.4%|
|Mean dose, 49.9 Gy|
|Median dose, 50 Gy|
|Mean dose, 58.8 Gy|
|Median dose, 60 Gy|
Of 60 patients, 36 (60%) had postoperative RT with a median dose of 60 gray (Gy) (range, 39.6-70.6 Gy). The other 24 (40%) patients received preoperative RT; median dose was 50 Gy (range, 46-50 Gy). Clear (R0), microscopic positive (R1), and gross positive (R2) resection margins were obtained in 30, 17, and 8 patients, respectively. There were 5 patients for whom resection margin status was not available.
The median follow-up was 75 months (range, 5-230 months). At last follow-up, 47 (78%) of 60 patients were alive without disease, 5 (8%) of 60 were alive with disease, 4 (6.7%) of 60 had died of disease, 2 (3%) patients had died of unrelated causes, and 2 (3%) patients were lost to follow-up with their final disease status unknown. The 5- and 10-year PFS, CSS, and OS were 85% and 81%, 95% and 89%, and 91% and 86%, respectively. On univariate analysis, patients with low-grade (grade I and II) tumors and older patients (age ≥45 years) fared better in terms of PFS (P = .005 and P = .03, respectively). Tumor size (≤8 cm vs >8 cm), primary tumor site, soft tissue involvement, and timing of RT were not statistically significant factors for PFS. Tumor grade was a statistically significant prognostic factor for OS (P = .009) (Fig. 1).
Local control and patterns of relapse
The crude local control rate was 90% at 10 years (Fig. 2). As expected, local control was suboptimal (42%) in the presence of R2 resection (Fig. 3). Of 8 patients who had R2 resection on pathological examination, 3 experienced uncontrolled progression (2 pelvis and 1 thoracic spine), whereas 5 patients had stable local disease (3 cervical spine, 1 lumbar spine, and 1 upper extremity) with long-term follow-up (68-140 months).
The pattern of relapse and outcome of relapsed patients is summarized in Table 2. A total of 12 (20%) patients had either relapse or progression of disease. Median time to relapse was 29.5 months (range, 8-126 months). Five of 8 patients who had high-grade tumors had distant metastasis, whereas this occurred in only 3 of 52 patients with low-grade tumors. All 7 distant relapses (with 2 of these patients relapsing beyond 5 years) were pulmonary metastases, of which 1 patient subsequently relapsed in spine, soft tissue, and brain.
|Patient Number||Age/ Sex||Primary Site||Grade/ Histology||Surgical Margins||Treatment||RT Dose, cGy||Relapse Site(s)||Time to Relapse, mo||No. of Relapses||Treatment||Status at Last Follow-up||OS, mo|
|1||59/M||Pelvis||Grade 1/conventional||R1||S+ART||6000||Pelvis||12||3||S+RT||Alive (Ds)||227|
|2||38/M||Larynx+hyoid||Grade 3/conventional||R1||S+ART||7000||Lung||12||3||S||Alive (Ds)||119|
|3||34/F||Tibia||Grade 3/mesenchymal||R1||S+ART||6600||Lung||28||1||S||Alive (NED)||70|
|4||37/M||Paraspinal||Grade 2/conventional||R0||S+ART||3960||Lung||31||1||S||Alive (NED)||44|
|5||39/M||Scapula||Grade 1/conventional||R1||S+ART||6600||Lung, brain, spine, soft tissue||72||3||S||Alive (Ds)||103|
|6||40/F||Scapula||Grade 3/myxoid||R0||S +ART||6000||Lung||8||1||S+RT+CT||Alive (Ds)||15|
|7||44/F||Pelvis||Grade 2/conventional||R0||NRT+S||5600||Pelvis||PD (local)||—||S||Alive (NED)||193|
|8||32/M||Pelvis||Grade 3/myxoid||R2||S+ART||—||Pelvis, lung||PD (local), relapse (lung)||1||NA||Dead (Ds)||9|
|9||40/M||Chest wall||Grade 2/myxoid||R0||NRT+S||5000||Lung||126||1||S||Alive (NED)||144|
|10||27/M||Thoracic spine||Grade 3/conventional||NA||S+ART||4800||Spine||48||1||S+RT||Alive (NED)||182|
|11||39/M||Thoracic spine||Grade 2/conventional||R2||NRT+ S||4600||Spine||PD (local)||—||None||Alive (Ds)||2|
|12||18/M||Pelvis||Grade 2/myxoid||R2||S +ART||6600||Pelvis||PD (local)||—||NA||Alive (Ds)||10|
Salvage treatment after relapse
Of the 9 patients who received salvage surgery, 3 also received further RT. In total, 5 patients who had salvage treatment for their relapse had their disease controlled at last follow-up. The median OS in the subset of patients who received salvage treatment after local/ systemic relapse was 86.5 months (range, 2-227 months).
The current report describes the efficacy of combined modality therapy with surgery and RT in a cohort of patients with localized extracranial chondrosarcoma treated over a period of 20 years. Patients had excellent local control rates that compare well with surgical series reported in the literature.14, 15 Although the role of RT in chondrosarcoma is controversial, given the risk of local recurrence if surgery was used as the sole method of management in this patient population, the addition of RT might improve local control rates. As might be expected, in our cohort of patients, local control was a direct function of resection margins and OS a function of the grade of the disease.
Traditionally RT has had a limited role in the radical management of chondrosarcomas, as they are considered to be radioresistant tumors.8, 12, 13 Thus, most series have reported on control rates with surgery alone, with few reporting the role of RT as a part of multimodality management.15, 25 The more recent use of proton therapy or mixed proton/photon therapy for skull-base and spinal sarcomas, in both radical and adjuvant settings, has allowed higher doses to be delivered, thereby overcoming resistance and improving local control rates.16, 26 The emerging evidence of efficacy of these radiation techniques in skull-base chondrosarcoma coupled with success achieved using combined modality approach (RT and surgery) for soft tissue sarcomas, which have also been considered relatively radio-resistant tumors,20, 23, 27 makes a case for the use of RT as an adjuvant modality, for selected cases, in chondrosarcoma.
Reported series of chondrosarcoma patients from the past 3 decades have focused on outcome in patients treated with surgery alone and are presented in Table 3. OS at 10 years was reported to be in the range of 50% to 80% and disease-free survival at 10 years between 28% and 88% with surgery alone for various primary sites. Of note is that in these studies, the majority of the patients had pelvic chondrosarcomas, which portend a poorer prognosis. This has been attributed to the anatomic complexity of the pelvis and often close proximity of the tumor to the surrounding neurovascular and visceral structures, resulting in suboptimal surgical resection margins.
|Study||Study Period||Median/ Mean FU, mo||Site(s) of Involvement||No. of Patients||Treatment||5-Year/10-Year DFS/DSS/PFS/MFS||5-Year/ 10-Year OS|
|Evans 19779||1948-1974||NR||Entire skeleton except skull||71||Surgery alone||NR||77%/67%|
|Pritchard 1980||1909-1975||Mean 138||Limb, limb girdle, and mobile spine||280||Surgery alone||10-year DFS <1950 28%, 10-year DFS >1950 70%||59%/46%|
|Sheth 199636||1970-1992||Median 115||Pelvis||67||Surgery alone||NR||—/52%|
|Bjornsson 199814||1911-1990||Median 150||Long bones and limb girdle||344||Surgery alone||5-year DSS 84%, 10-year DSS 80%||77%/66%|
|Koch 2000 (NCDB, USA)||1985-1995||NR||Head and neck||400||Surgery (59.5%), surgery + RT (21%)||5-year DSS 87.2%, 10-year DSS 70.6%||77.6%/64.3%|
|Bergh 200134||1967-1999||Mean 156||Pelvis, sacrum, and mobile spine||69||Surgery alonea||NR||72%/67%|
|Pring 200115||1975-1996||Median 140||Pelvis||64||Surgery alone||5-year DFS 69%, 10-year DFS 67%||82%/80%|
|Griffin 200837||1989-2004||NR||Scapula||24||Surgery aloneb||10-year MFS 88%||NR|
|Present study||1986-2006||Median 75||Extracranial||60||Surgery + RT||10-year PFS 80.5%||—/86%|
The majority of our patient cohort had primary skeletal tumors with or without soft tissue involvement, and pure extraskeletal chondrosarcomas were observed in 11 of 60 patients. Of these 11 patients who had pure extraskeletal soft tissue disease, 7 (64%) patients had myxoid histology, whereas the rest had conventional histology. Although they have been included in the analysis along with skeletal chondrosarcomas, it is not clear if extraskeletal myxoid chondrosarcomas (EMCs) behave in a similar manner to true skeletal myxoid chondrosarcomas (SMCs) or to soft tissue sarcomas. It is generally debated whether true myxoid chondrosarcoma of bone exists. The 2002 World Health Organization classification classified this entity in the “tumors of uncertain differentiation” category, although molecular studies suggest that EMC is a distinct and separate entity from SMC.29 These tumors may have a distinct natural history as compared with skeletal conventional and skeletal myxoid chondrosarcomas. Data on EMC have been limited to case reports and small series. However, a retrospective study from 2 centers studying the role of surgery and chemotherapy in 87 patients with EMC suggests that this histology is associated with inferior outcomes (local recurrence rate of 37%, distant metastasis rate of 26%, and 10-year OS of 65%);28 another small comparative study of SMC and EMC has also shown poorer prognosis in patients with EMC, with a higher rate of distant metastasis.29
A report on head and neck chondrosarcoma from the United States derived data from the National Cancer Database, where the majority of patients had received surgery alone (59.5%), whereas a subset of patients (21%) had received combined surgery and adjuvant RT, and reported a 10-year PFS of 70.6% and OS of 64.3%.30 Although this study did not specifically comment on the efficacy of adjuvant RT in its cohort of patients, the majority of patients who had RT were those with positive margins after surgery and those with myxoid or mesenchymal subtypes.
The definition of surgical adequacy varies in different series, although most are guided by the criteria described by Enneking, which considers adequate margins as “wide excision” (removal of tumor together with margin of normal tissue completely encircling the tumor) or “radical excision” (complete removal of tumor, including all the involved muscles from their origin to insertion and all the involved bone from joint to joint).7, 31, 32 This is considered to be the most important prognostic factor associated with local control of disease.7, 11, 15, 33 The issue of need for RT in patients with R0 resections in our series could be debated; however, many patients were treated preoperatively because of concerns with regard to the ability to obtain a wide R0 resection akin to a policy of preoperative RT in soft tissue sarcoma.14, 15, 34
Overall, patients who had gross residual disease (R2) fared poorly, supporting the need for adequate gross surgical resection in this disease; however, the local control rate for such patients without RT might have been anticipated to be worse than that achieved in our study. Thus, RT as an adjunct to surgery might be most beneficial to that group of patients where there is microscopic contamination, that is, in R1 resections or possibly in R0 resection with very close resection margins. The burden of disease may, in these situations, be low enough to allow successful residual tumor cell kill by moderately high doses of RT.
Grade of tumor is considered to be an important determinant of distant failure and ultimately survival.14, 15, 30 The influence of grade on probability of local control is not clear, and this is confounded by many studies documenting only PFS without distinguishing between local and distant relapse, which is true for our study as well. As expected, PFS and OS for patients with high-grade tumors were worse and consistent with other studies.9, 15, 34
In a study by Pritchard et al, larger primary tumor size was shown to be associated with a statistically significant difference in the 10-year OS, with a cutoff at 10 cm.33 In our current series, we were not able to demonstrate a difference based on tumor size with 8 cm used as a cutoff (10-year PFS of 87% for tumors ≤8 cm vs 78% for tumors >8 cm, P = .77). Somewhat similar results were observed by Wirbel et al in his series of pelvic chondrosarcomas, where tumor volume was not a significant factor affecting the rate of local recurrences. In this study, the local recurrence rates were 22% for large-volume tumors (>500 mL) and 20% for small-volume tumors (<500 mL).35
As with all retrospective series, patient selection significantly influenced outcome in our patient cohort. The small number of patients may have influenced the finding that tumor size was not a predictor for outcome. Considering that all these tumors were felt to have a high risk of local recurrence if treated with surgery alone, combined surgery and RT seems appropriate; this is reflected in the local control rates, which compare favorably to series where surgery was the sole modality of treatment. We recognize that although these data are compelling, only a randomized trial will truly define the role of (neo)adjuvant RT in this setting.
Currently, we offer RT preoperatively (50 Gy) to patients with tumors where positive margins are anticipated or in the postoperative setting (66-70 Gy) to those with unexpected positive margins and where further surgery would result in significant morbidity. There are few series evaluating the role of RT in chondrosarcoma in the extracranial setting. Our data suggest that although complete resection remains the primary treatment in the management of chondrosarcoma, the addition of RT in this disease appears to provide excellent and durable local control in those at high risk of local failure when treated with surgery alone.
CONFLICT OF INTEREST DISCLOSURES
The authors made no disclosures.
- 6Dahlin's Bone Tumours. General Aspects and Data on 11087 cases. 5th ed. Philadelphia, PA: Lippincott-Raven; 1996: 109-115.. . In: