Soft tissue sarcomas of the popliteal fossa

A single-institution retrospective review




Soft tissue sarcomas (STSs) arising from the popliteal fossa present a challenge with regard to local control of primary tumors. Due to concerns of functional morbidity and neurovascular compromise, there is debate about what represents the best therapy for these patients.


We conducted a retrospective medical record review of patients treated at The University of Texas M. D. Anderson Cancer Center for STS of the popliteal fossa from 1990 to 2008.


There were 47 eligible patients, 28 of whom were male and 19 of whom were female. Synovial sarcoma was the most common diagnosis, with 12 cases. Most patients had T2b tumors (31 patients; 66%). The median duration of follow-up was 3.8 years (range, 0.6-17.9 years). The 5- and 10-year overall survival rates were 63% and 51%, respectively. Metastasis at diagnosis was associated with poorer overall survival (5-year overall survival, 74% versus 13%; P<.001) and poorer recurrence-free survival (5-year recurrence-free survival, 51% versus 0%; P<.001) on univariate analysis. Radiation therapy improved local recurrence-free survival (5-year local recurrence-free survival, 56% versus 17%; P = .004), whereas a trend was observed for surgical margin status (P = .07). Tumor size and neurovascular involvement did not influence outcome. Twenty-two patients had recurrent disease, with 15 patients having local recurrence, and 16 patients died from progressive disease.


Radiation therapy may play an important role in the treatment of popliteal fossa STS, but further study is needed to better define the best clinical application. Additional study is needed to re-evaluate association of surgical margin status and outcome. Cancer 2011; © 2010 American Cancer Society.

Soft tissue sarcomas (STSs) are uncommon tumors of mesenchymal origin and represent approximately 1% of all adult cancers.1 The extremities represent one of the most common primary sites for STS, but the popliteal fossa is a relatively rare anatomic site of primary tumor presentation that presents unique management challenges.2 Treatment typically involves a multidisciplinary approach, but adequate surgical resection remains the mainstay of primary local therapy for STS, and previous studies have reported that tumor size and grade are important factors for survival.3-5

Treatment of STS of the popliteal fossa is a challenge due to the proximity of important neural and vascular structures, which prevents wide margins of resection. Aggressive local control, such as the addition of radiation therapy either before or after resection, may have increased risk of functional morbidity. However, there are limited data regarding the clinical presentation, characteristics, and outcome of patients with this challenging presentation of STS. Most studies report small numbers of patients with popliteal tumors. In addition, the data are unclear regarding neurovascular involvement in relation to clinical outcome. The primary goal of this study was to examine our institution's experience with patients diagnosed and treated for STS of the popliteal fossa, which represents the largest reported series to date.


After obtaining institutional review board approval, medical records were reviewed for patients diagnosed and treated for STS of the popliteal fossa at The University of Texas M. D. Anderson Cancer Center from 1990 to 2008. Patients with metastatic popliteal fossa lesions from another primary site and patients seen only for consultation on treatment options were excluded from analysis. Patient characteristics that were reviewed included age at diagnosis, sex, histology, and disease extent. Tumors were classified according to the American Joint Committee on Cancer's tumor-node-metastasis classification system.6 Tumor size was categorized as T1a (<5 cm and localized to superficial tissue), T1b (<5 cm and involving the deep tissue of the lower extremity), T2a (≥5 cm and localized to superficial tissue), or T2b (≥5 cm and involving the deep tissue). Treatment characteristics included treatment received and duration of follow-up. Clinical endpoints included time to treatment failure from date of diagnosis and site of treatment failure. Treatment failure was defined as being the total duration of time from date of initial diagnosis to date of progression or recurrence of disease. Duration of follow-up was defined as the time from initial diagnosis to the last documented clinical follow-up. Nerve and/or blood vessel involvement was defined as either 1) complete encasement of the structure as an operative finding or 2) invasion of the structure identified at time of pathological evaluation of the tumor. Final surgical margin status was defined as clear if 1) no tumor could be identified grossly or microscopically involving any margin of the pathological specimen, no matter how thin the rim of normal tissue, or 2) no tumor cells could be identified in specimens from a wide local re-excision.

Descriptive statistics were used to summarize the patient characteristics. The Fisher exact test was used to compare patients with respect to sex, race, treatment type, primary disease site, and recurrence status. Cox proportional hazards regression models were used to evaluate overall survival (OS), recurrence-free survival (RFS), and local recurrence-free survival (LRFS) in patients while accounting for various potential prognostic factors.7 OS, RFS, and LRFS were measured from diagnosis date to date of defined event or date of last follow-up. The product-limit method of Kaplan and Meier was used to estimate survival based on presence of neurovascular involvement, metastases, radiation therapy history, tumor stage, and surgical margin status.8



A total of 60 patients were identified as having STS in the popliteal fossa, 13 of whom were excluded from analysis because their primary tumors were not originating from the popliteal fossa. Therefore, 47 patients were eligible for analysis (Table 1), 28 of whom were male and 19 of whom were female. The mean age at time of diagnosis was 44 years (range, 12-78 years). Thirty-four patients were Caucasian; 5 patients were Hispanic/Latino; 3 patients were black; and 5 patients were of other race/ethnicity.

Table 1. Clinical and Demographic Characteristics of Patients
CharacteristicsNo. (%)
  • a

    Seven patients required resection.

  • b

    Four patients required resection.

 Female19 (40)
 Male28 (60)
 Synovial sarcoma12 (26)
 Malignant fibrous histiocytoma9 (19)
 UndifferentiatedSarcoma8 (17)
 Liposarcoma8 (17)
 Chondrosarcoma2 (4)
 Desmoid fibromatosis2 (4)
 Other6 (13)
Tumor stage
 T1a or T1 (unknown a/b)5 (11)
 T1b5 (11)
 T2a or T2 (unknown a/b)5 (11)
 T2b31 (66)
 Unknown1 (2)
Nodal involvement
 No46 (98)
 Yes1 (2)
Metastatic disease
 No39 (83)
 Yes8 (17)
Neural involvement
 No24 (51)
 Yes11a (23)
 Unknown12 (26)
Vessel involvement
 No27 (58)
 Yes8b (17)
 Unknown12 (26)

Synovial sarcoma was the most common histological diagnosis and was identified in 12 (26%) patients. The next most common diagnoses were malignant fibrous histiocytoma in 9 (19%) patients, liposarcoma in 8 (17%) patients, and undifferentiated sarcoma in 8 (17%) patients. There were 36 (77%) patients with T2 tumors, and 31 (66%) patients had large, invasive tumors (T2b). Only 1 (2%) patient had nodal involvement, whereas 8 (17%) patients had metastases at presentation, with the lungs being involved in all 8 patients.

Involvement of neural structures could be evaluated in 35 patients, 11 (31%) of whom had involvement of local neural structures based on operative findings at the time of definitive surgery. Six patients required neurolysis of the involved neural structure. Vascular involvement could be assessed in 35 patients, 8 (23%) of whom had blood vessel involvement based on encasement of a vessel at the time of definitive surgery; 3 of these patients required ligation of the vessel, and 1 patient had pathological confirmation of vascular invasion of tumor.


A total of 44 patients underwent definitive surgical treatment for their tumors, including 4 patients who underwent above-the-knee amputative surgery. Three patients did not undergo surgical treatment due to metastatic disease. Therefore, the rate of limb salvage was 91% (40 of 44 patients) (Table 2). Fourteen patients underwent a second surgical procedure in an attempt to completely resect the primary tumor, 1 of whom underwent above-the-knee amputative surgery. After excluding the 4 patients who had amputative surgery, final surgical margin status could be assessed in 40 patients. Negative surgical margins were achieved in 24 (60%) patients, whereas 11 (28%) patients had positive surgical margins. Five patients had surgical margin status that was indeterminate based on available records. When reviewing the 13 patients who underwent a second surgical procedure that was nonamputative, 10 (77%) patients were able to achieve negative surgical margin status, whereas 2 (15%) patients still had involved margins.

Table 2. Treatment and Outcome Characteristics of Patients
 No. (%)
  • a

    Four patients required amputative surgery.

  • b

    Median radiation dose: 50 Gy (range, 20-66 Gy).

  • c

    Median follow-up: 3.8 years (range, 0.6-17.9).

Treatment received
 Surgery44a (94)
 Chemotherapy23 (49)
 Radiation therapyb37 (79)
Final surgical margin
 Negative24 (51)
 Positive11 (23)
 Indeterminate5 (11)
 Unknown3 (6)
 Excluded due to amputation history4 (8)
Wound complication
 None31 (66)
 Delayed healing5 (11)
 Wound infection2 (4)
 Graft failure1 (2)
 Chronic pain1 (2)
 Seroma1 (2)
 Unknown6 (13)
 None23 (49)
 Locoregional14 (30)
 Synovial sarcoma4
 Unclassified sarcoma4
 Malignant fibrous histiocytoma2
 Distant7 (15)
 Malignant fibrous histiocytoma3
 Synovial sarcoma2
 Site unknown1 (2)
 Unknown status2 (4)
 Alive20 (43)
 Dead19 (40)
 Synovial sarcoma6
 Unclassified sarcoma5
 Other histology5
 Other causes3
 Lost to follow-upc8 (17)

Chemotherapy was administered for 23 (49%) patients. The combination of doxorubicin with ifosfamide was the most common initial chemotherapy regimen used (13 patients [57%]). Nine other patients received different initial chemotherapy regimens, but in general these regimens also contained doxorubicin with an alkylating agent, such as doxorubicin with cyclophosphamide and dacarbazine (4 patients). One patient received doxorubicin with ifosfamide and dacarbazine, and another patient received doxorubicin with dacarbazine. One patient received high-dose ifosfamide. Another patient received vincristine, doxorubicin, cyclophosphamide, etoposide, and ifosfamide. Chemotherapy regimen could not be determined for 1 patient.

Thirty-seven (79%) patients received radiation therapy, and the median dose delivered was 50 Gy (range, 20-66 Gy). Sixteen (34%) patients received preoperative radiation therapy with a median dose of 50 Gy (range, 20-60 Gy), and 21 patients received postoperative radiation therapy with a median dose of 60 Gy (range, 50-66 Gy). Of the 16 patients that received preoperative radiation therapy, 11 (69%) achieved negative surgical margins after definitive surgery, whereas 4 (25%) patients had positive surgical margins and 1 (6%) patient had progressive disease that required amputative surgery. For patients who received postoperative radiation therapy, brachytherapy was used in 2 patients, and 1 patient initially started with intraoperative radiation therapy and then completed radiation treatment postoperatively. Twelve patients who received radiation therapy had negative surgical margin status, 9 (75%) of whom had T2 tumors.


The median duration of clinical follow-up was 3.8 years (range, 0.6-17.9 years). Sixteen patients (34%) died of their disease. Two patients died from non-disease-related causes, and 1 patient died from complications of the surgical procedure (chronic delayed surgical wound healing complicated by poorly controlled diabetes mellitus, which eventually resulted in wound infection and fatal septic shock). The 5- and 10-year OS rates for the entire cohort were 63% (95% confidence interval [CI], 46%-75%) and 51% (95% CI, 33%-66%), respectively. Univariate analysis determined that only metastatic disease at time of diagnosis was significantly associated with OS (P<.001), whereas tumor stage, neurovascular involvement, and final surgical margin status were not significant (Table 3). Patients with nonmetastatic disease at time of diagnosis had a 5-year OS of 74% (95% CI, 55%-86%) versus 13% for patients with metastatic disease (95% CI, 1%-42%) (Figure 1). Metastases at diagnosis was also the only significant factor associated with RFS (P<.001) (Figure 2), with patients having nonmetastatic disease demonstrating a 5-year RFS of 51% (95% CI, 33%-67%) versus 0% for patients with metastatic disease.

Figure 1.

Kaplan-Meier curves comparing overall survival between patients with nonmetastatic disease (M0) and patients with metastatic disease (M1) are shown (P<.001).

Figure 2.

Kaplan-Meier curves comparing recurrence-free survival between patients with nonmetastatic disease (M0) and patients with metastatic disease (M1) are shown (P<.001).

Table 3. Factors Associated With Survival
5-Year OS95% CIP5-Year RFS95% CIP5-Year LRFS95% CIP
  1. OS indicates overall survival; RFS, recurrence-free survival; LRFS, local recurrence-free survival; CI, confidence interval.

 T2b66%45%-80% 41%23%-59% 55%34%-72% 
Nerve involvement
 Yes81%42%-95% 41%13%-68% 61%27%-84% 
Vessel involvement
 Yes75%32%-93% 25%4%-56% 50%15%-78% 
Final margin
 Positive73%53%-90% 33%8%-61% 33%8%-61% 
 Yes13%1%-42% 0% 
Radiation therapy

Forty-five patients could be evaluated for initial treatment failure, and recurrence or metastasis of disease was observed in 22 (49%) patients at a median duration of 14.5 months (range, 2-77 months). Locoregional treatment failure was observed in 14 (31%) patients, including 3 patients who experienced local progression of disease during initial treatment. Metastatic recurrence was noted in 7 (16%) patients, and the lungs were the site of recurrence in each case. One patient had documented recurrence of disease, but there were insufficient records to determine the site of recurrence. The characteristics of recurrence and outcome based on tumor histology are outlined in Table 2.

For LRFS, the patient with unknown site of recurrence and patients who received amputative surgery were excluded from analysis. Patients who received radiation therapy had significant improvement in outcome (Table 3 and Figure 3), with a 5-year LRFS of 56% (95% CI, 36%-71%) compared with 17% for patients who did not receive radiation therapy (95% CI, 1%-52%) (P = .004). A trend was also noted with final surgical margin status (P = .07) (Table 3), but tumor stage and neurovascular involvement did not influence LRFS.

Figure 3.

Kaplan-Meier curves comparing local recurrence-free survival between patients who received radiation therapy (XRT) and patients who did not receive radiation therapy (No XRT) are shown (P = .004). One patient excluded from XRT group due to unknown site of recurrence. Three amputees and 1 patient with unknown recurrence history excluded from No XRT group.

Ten (21%) patients developed surgical complications. The most common complication experienced was delayed wound healing, which was noted in 5 patients. Two patients had wound infection, and there was 1 case each of graft failure, persistent postoperative pain, and postoperative seroma. As mentioned previously, 1 patient had a comorbidity with poorly controlled diabetes mellitus, which likely contributed to wound complications and culminated in a wound infection with fatal septic shock. The patient died 3.5 months after the definitive surgical procedure.

Long-term orthopedic complications could be determined in 17 patients who were still receiving active clinical follow-up. The median duration of follow-up in this group was 64 months (range, 13-215 months). One patient underwent above-the-knee amputative surgery and did not report any problems. One patient had persistent ankle weakness. This patient had a synovial sarcoma that was found to be encasing the popliteal nerve and vessels at the time of definitive surgery. The neurovascular structure was dissected away from the tumor, and the patient also received 54 Gy of adjuvant radiation therapy. Another patient had persistent lymphedema of the lower extremity after 60 Gy of adjuvant radiation therapy. Fourteen patients did not report any significant long-term limitations in range of motion of the knee, and 12 of these patients had received radiation therapy.


Sarcomas of the popliteal fossa represent a therapeutic challenge. Due to the potential for involvement of adjacent neurovascular structures, there is concern whether radical surgical approaches, such as amputative surgery, are required for patients with STS of the popliteal fossa or whether a combination of surgery and radiation therapy can be used effectively instead.9 Several studies have demonstrated that the use of aggressive multimodal therapy can produce limb-sparing rates of 65%-96%, and we observed a limb-sparing rate of 91% in our cohort.10-14 We found that the addition of radiation therapy to nonamputative surgical resection was associated with better local control than surgery alone in this cohort. However, because there can be selection bias associated with deciding which patients receive radiation therapy, additional data are needed to clarify the indications for radiation therapy. In addition, there are limited data regarding other clinical factors that may be prognostic of outcome, because most studies have small cohorts of popliteal sarcoma patients that are grouped with patients having tumors in other flexor fossa, such as the antecubital space.

Due to the close proximity of neurovascular structures, achieving wide surgical margins of resection of >1 cm can be difficult. Analysis of our cohort demonstrated that final surgical margin status did not reach statistical significance as a risk factor for OS, but a trend was noted with regard to LRFS. However, given the small number of patients analyzed, it would be of interest to re-evaluate the potential association of final surgical margin status in a larger study cohort.

There are limited data regarding detailed analysis of neurovascular involvement. Hohenberger et al15 described their experience treating 20 patients with STS invading neurovascular structures as determined via angiography or operative finding. Unfortunately, only 4 patients had popliteal fossa tumors, and 2 of those patients were being treated for recurrent disease. Other studies have not evaluated neurovascular involvement as an independent risk factor for analysis.

The evaluation of our cohort did not reveal any significant association between neurovascular involvement and clinical outcome, and only 4 patients underwent amputative surgery. However, our analysis may have been influenced by 2 factors. First, a significant number of patients had insufficient records to determine the status of neurovascular involvement, which decreased our total number of patients evaluable for these factors. Second, we defined neurovascular involvement based on operative findings at the time of definitive surgery or pathological evaluation after definitive surgery, which was not uniform for all patients because some patients had upfront surgery whereas other patients received neoadjuvant therapy. Thus, future studies may require additional clarification to define level of neurovascular involvement.

Radiation therapy is an essential component of limb-sparing local therapy for many presentations of STS. Randomized controlled trials have consistently demonstrated that radiation therapy cannot be omitted from the local therapy management strategy when a less than radical resection is planned.16, 17 There are select primary tumor presentations for which radiation therapy may be omitted in the absence of radical resection, such as low-grade disease and small, superficial high-grade tumors.18 However, omission of radiation therapy usually rests on the principle that wide-margin surgical extirpation of the tumor obviates the need for treatment intensification with radiation therapy for a relatively low-risk tumor. In the popliteal fossa, wide margins are difficult to obtain, and in the absence of amputation, radiation therapy warrants consideration. Our findings and those of others are consistent with the conclusion that radiation therapy is beneficial in the treatment of STS of the popliteal fossa.17 The timing of radiation therapy, whether it be preoperative or postoperative, is a matter of debate and depends on the specifics of individual case presentation. Investigators, including those reporting a randomized controlled trial to study the issue, have found that local control and survival outcomes are the same for both preoperative and postoperative radiation therapy.19, 20 However, many studies have associated postoperative radiation therapy with an increased risk of long-term complications (eg, fibrosis or impaired function), likely due to the increased dose and radiation field size.21-23 This finding is of particular concern in the popliteal region,24 and when appropriate, treatment planning should include consideration that a lower preoperative dose of 50 Gy is less likely to result in joint fibrosis than the required 60 Gy or more in the postoperative setting. Issues such as decreased mobility and bone fractures have also been observed.25, 26 Of note, omission of radiation therapy should be seriously considered in lieu of more radical surgery in pediatric patients, who have an increased risk of leg-length discrepancy when radiation therapy is administered.27

In conclusion, the use of aggressive multimodal therapy is beneficial for the majority of patients with STS of the popliteal fossa in allowing for limb-sparing local management. Radiation therapy appears to play an important role in disease control, but further study is necessary to better define the best clinical practice of radiation therapy for certain subgroups (eg, pediatric patients). Further study is also needed to provide a more uniform definition of neurovascular involvement that may provide prognostic information on clinical outcome.


This work was supported in part by a Cancer Center Support Grant (National Cancer Institute Grant P30 CA016672).