The management of localized STS underwent a paradigm shift in the 1980s when it was shown that limb salvage surgery plus radiation was equivalent to amputation in both disease free survival (DFS) and OS [1]. Subsequently, Yang et al. randomized patients to limb salvage surgery alone, or to surgery with the addition of 4,500 cGy of radiation with a 1,800 cGy boost to the tumor bed. After median follow-up of over 9 years, local recurrence (LR) rates in patients with low and high grade STS undergoing limb salvage surgery alone were 33 % and 19 % respectively [2]. With the addition of adjuvant radiation there was a statistically significant decrease in LR rates to 3.8 % and 0 % respectively. Patients who experienced a LR were subsequently treated with either amputation or wide local re-excision followed by radiation. The OS in both groups was not statistically different. External beam radiation can be administered neoadjuvantly or adjuvantly with similar disease control rates [3]. In the only randomized control trial of either approach, patients were either given a preoperative dose of 50 Gy or a postoperative dose of 66 Gy. Patients who received pre-operative radiation were more likely to have wound healing complications (35 % versus 17 %, P = 0.01). The local recurrence, regional and distant failure rates, and progression free survival (PFS) were not different between the two groups. In a follow-up to this study higher rates of long term morbidity such as subcutaneous fibrosis (48.2 % versus 31.5 %, P = 0.07), edema (23.3 % versus 15.1 %, P = 0.51) and joint stiffness (23.2 % versus 17.8 %, P = 0.26) were noted in the postoperative arm at 2 years following treatment [4]. A recent report on preoperative image-guided intensity-modulated radiation therapy (IMRT), allowing for precise targeting of tumor and less radiation exposure to normal tissue demonstrated a 30.5 % rate of wound complications similar to the historical study [5]. However, it appeared that primary wound closure was more attainable with IMRT (55 of 59 patients [93.2 %] versus 50 of 70 patients [71.4 %]; P = 0.002).

There may be some patients who do not require adjunctive radiation therapy. In a retrospective analysis of 74 patients who underwent limb salvage surgery without radiation, the 10 year local control rate was 93 %. Patients with a close resection margin (less than 1 cm) had a local control rate of 87 % versus 100 % in those with margins 1 cm or greater. There was no association between local control rate and grade, size, site or depth of tumor [6]. A separate Surveillance, Epidemiology, and End Results (SEER) analysis of patients with small (less than 5 cm) STS showed no benefit for radiation in sarcoma specific survival or OS regardless of the grade of tumor [7]. Other studies have shown that older age, recurrent disease at presentation, positive margins and histologic subtype of STS impact LR risk [8]. Therefore, it is often difficult to predict an individual’s risk of LR clinically. For this reason, a nomogram attempting to quantify the risk of LR after limb sparing surgery has been developed [9]. This nomogram has not been validated however.

Unlike patients with extremity sarcomas, patients with primary retroperitoneal tumors are more likely to experience LR [10]. In a series by Jaques and colleagues, of 114 patients with resected retroperitoneal sarcoma, LR after complete resection occurred in 44 % of patients [11]. There are no randomized controlled trials of radiation therapy in retroperitoneal sarcoma. Similar to the extremity sarcoma data, in retrospective series, radiation appears to decrease the risk of LR but has no impact on OS [12–15]. Postoperative radiation therapy is often difficult to administer in this location as bowel can fall back into the radiation field once the mass has been resected. Long term results are available from two studies of high risk patients (n = 72) who received preoperative radiation (median dose, 45 Gy; range, 18.0–50.4 Gy) followed by surgical resection 4–8 weeks later. Eighty-nine percent of patients received the entire planned radiation dose, with discontinuation in the others due to reasons such as progression of disease and bowel ischemia. Seventy-nine percent of patients were able to undergo laparotomy, 95 % of which were able to undergo an R0 or R1 resection. Of the patients who were able to complete radiation and undergo gross total resection (n = 54), 28 patients developed recurrent disease, with local failure in 20 of these. The 5-year LR-free survival rate in this study was 60 % with a median OS that exceeded 60 months. The authors therefore suggested that further study of this approach was warranted given that survival appeared to exceed historical controls [16]. An ongoing phase III study by the EORTC is assessing the role of radiation therapy in decreasing the risk of local recurrence in this group of patients (NCT01344018). Patients are being randomized to receive surgery alone or preoperative radiation followed by surgical resection.

The use of adjuvant chemotherapy in STS has not been uniformly accepted, owing to the heterogeneity of the disease, and lack of uniformity in study design. In 1997 the Sarcoma Meta-Analysis Collaboration (SMAC) performed a meta-analysis of 14 randomized trials (n = 1568). The chemotherapy regimen in all of the included studies was doxorubicin based, either as a single agent or in a combination, administered after surgical local control. The hazard ratios (HR) for LR-free interval, distant recurrence-free interval and overall recurrence-free survival were 0.73 [95 % CI, 0.56–0.94; P = 0.016], 0.70 [95 % CI, 0.57–0.85; P = 0.0003] and 0.75 [95 % CI, 0.64–0.87; P = 0.001] respectively favoring the chemotherapy arm. However, there was no statistically significant benefit in OS for chemotherapy in the analysis. In a pre-planned analysis of the extremity only group (n = 886), the HR for OS was 0.80 (P = 0.029), with an absolute benefit of 7 % favoring the chemotherapy group [17]. In 2007, Pervaiz and colleagues published an updated meta-analysis, which included the original 14 trials, plus an additional 4 trials that included ifosfamide, a very effective drug in the metastatic setting. In this study the benefit of chemotherapy remained for local and distant recurrence with an absolute risk reduction of 10 % [95 % CI, 5–15 %; P ≤ 0.01] in overall recurrence favoring the chemotherapy group. In addition, a benefit for OS was seen. The hazard ratio for risk of death was 0.77 [95 % CI, 0.64–0.93; P = 0.01] (absolute risk reduction 6 % [95 % CI, 2–11 % P = 0.003]). In the group of patients who received doxorubicin and ifosfamide (n = 414), the absolute risk reduction in death was 11 % [95 % CI, 3–19 %; P = 0.01] [18]. The OS improvement with doxorubicin and ifosfamide in the updated meta-analysis led investigators to formally study this combination (EORTC 62931). Patients were randomized after surgery to no further therapy or five cycles of doxorubicin and ifosfamide. There was no benefit for chemotherapy in relapse free survival or OS [19]. The lower dose of ifosfamide used (5 g/m²/cycle) and the inclusion of low grade sarcomas as well as tumors less than 10 cm could be reasons why the trial was negative. This study was included in a separate meta-analysis published in 2008 which still showed a statistically significant improvement in DFS and OS at 5 years (OR, 0.71; 95 % CI, 0.54–0.85; P = 0.000) and (OR, 0.79; 95 % CI, 0.66–0.94; P = 0.005) respectively for adjuvant chemotherapy. The OS benefit was not maintained at 10 years however (OR, 0.87; 95 % CI, 0.72–1.04; P = 0.12) [20]. If one chooses to administer adjuvant chemotherapy, the appropriate number of cycles to use remains unclear. A study of three pre-operative cycles of epirubicin and ifosfamide versus the same number of pre-operative cycles with an additional two cycles post-operatively showed no difference between the arms for the primary objective of 5-year OS [21].

The only randomized control trial of neoadjuvant chemotherapy was EORTC 62874 in which 134 high risk patients were randomized to three cycles of pre-operative chemotherapy with doxorubicin 50 mg/m² and ifosfamide 5 g/m² or surgery alone. Patients were defined as high risk if their tumors were either 8 cm or larger, or grade II or III. Patients with grade II or III tumors that were locally recurrent or had inadequate initial surgery were also eligible. Accrual was slow, and the study was terminated early. Although the study was not adequately powered, there was no difference in 5-year DFS or OS between the two groups. The lower doses of doxorubicin (cumulative dose of 150 mg/m²) and ifosfamide (cumulative dose 15 g/m²) may have once again contributed to the negative results [22].

Most studies of neoadjuvant and adjuvant chemotherapy focus on patients with extremity tumors. In fact, in the original SMAC meta-analysis, it was this group of patients who derived the most benefit from adjuvant chemotherapy [17]. To understand the benefit of neoadjuvant chemotherapy in retroperitoneal sarcomas, Donahue and colleagues collected data on 55 patients with high grade, primary retroperitoneal tumors who had received neoadjuvant therapy. Chemotherapy agents included doxorubicin, ifosfamide or gemcitabine and docetaxel. All patients had surgical resection, and may have also received radiation. The 5-year disease specific survival for this cohort was 47 %. This was compared to the expected survival with surgery alone as predicted by an internally validated STS nomogram [23] developed by Memorial Sloan Kettering Cancer Center; there was no statistical difference. Interestingly, in the 25 % of patients who had necrosis greater than or equal to 95 % pathologically at time of resection, the 5-year DSS was 83 %, significantly improved compared to expected per the nomogram (P = 0.018) [24].

Phase II studies of single agent doxorubicin or ifosfamide in combination with radiation have demonstrated feasibility and relatively high response rates. Owing to the increased radiation sensitivity of STS when chemotherapy is given concurrently, full doses of either agent are not generally used. Doxoroubicin has been administered at 12 mg/m²/day over 5 days every 2 weeks [25] or ifosfamide 12 g/m² by continuous infusion over 5 days every 3 weeks for three cycles combined with external beam radiation at doses of 50 Gy [25, 26]. When radiation was combined with ifosfamide, the pathologic response rates were greater or equal to 95 % in 28 % of patients [26].

The higher response rates achieved with multi-agent chemotherapy in the metastatic setting led investigators to study this same approach in patients with localized disease. A single institution study of preoperative doxorubicin, ifosfamide, dacarbazine (MAID) for three cycles interdigitated with radiation 44 Gy (two sets of 22 Gy) followed by surgery suggested improved outcomes. To be eligible patients were required to have a high grade extremity STS larger or equal to 8 cm. A total of 83.3 % of patients received all six cycles of chemotherapy. Outcomes were compared to a cohort of patients from an existing database with similar tumor size, grade, age and era of treatment who had received adjunctive radiation only. The 5-year local control, freedom from distant metastases, DFS and OS in the chemotherapy group were 92 %, 75 %, 70 %, 87 %, statistically improved over the historical controls. However, the rate of febrile neutropenia requiring hospitalization was 25 %, and 29 % of patients had wound healing complications. One patient developed a myelodysplastic syndrome [27]. The RTOG then studied a very similar regimen in 66 patients with high grade sarcoma. Only 59 % of patients were able to complete all planned chemotherapy and there were three chemotherapy related deaths [28]. Long-term follow-up of this group of patients showed that the 5-year DFS and OS were 56.1 % [95 % CI, 43.9–68.3 %] and 71.2 % [95 % CI, 60.0–82.5 %], lower than what was achieved in the single institution study [29]. In retroperitoneal sarcomas, two small studies have shown safety and feasibility of combined modality therapy. Most grade 3 or higher toxicity was gastrointestinal in nature (nausea, vomiting, diarrhea, anorexia), but hematologic toxicity, dermatologic toxicity and stomatitis was also noted in a few patients [30, 31]. One study did demonstrate an R0/R1 resection rate of 90 %, however, 17 % of patients progressed on chemoradiation rendering them unresectable [30]. All patients had been initially considered resectable.

Patients may present initially with tumors that are considered unresectable. In this scenario, chemotherapy or radiation may be used to try and downstage the tumor prior to an attempt at surgery. If the disease is still not resectable, other techniques such as limb infusion or definitive radiation may be used. Finally, amputation still has a role in the management of patients with STS.

Definitive radiation can also be offered to patients who have unresectable tumors or who are not medically fit for surgery. In general, a higher dose of radiation must be used in STS than for epithelial tumors. Series have shown 5-year local control rates and survival of 43.5 % and 28.4 % respectively when doses of 64 Gy or higher are used. In patients with tumors less than 5 cm local control rates approached 90 %, whereas in patients with tumors greater than 10 cm local control rates were only 30 % [42]. Brachytherapy is a form of radiation which involves the placement of catheters in the operative bed during a surgical procedure. It allows for higher doses of radiation to be directed to the tumor, and sparing of normal adjacent tissue. There have been no randomized control trials comparing external beam radiation therapy with brachytherapy in the management of patients with early stage sarcoma. For patients with tumors that have been previously irradiated, it is often difficult to administer further radiation therapy to the area and brachytherapy can be of use. A retrospective review of 26 patients with recurrent STS all of whom had previously received external beam radiation therapy showed that 100 % of patients were able to attain a margin negative resection after undergoing brachytherapy. Five patients had a wound complication and nine patients did ultimately develop a local recurrence despite this approach [43].

Neoadjvuant therapy can be used to assist inconverting unresectable tumors and/or limb salvage. In the previous EORTC 62874 study limb salvage was achieved in 88 % of patients [22]. Unfortunately, most data collected on this approach is done so retrospectively and therefore subject to selection bias, as patients with the most aggressive tumors are more likely to receive chemotherapy. At least two smaller studies have demonstrated that neoadjuvant chemotherapy was useful in downstaging tumors and allowing for limb salvage. Azzarelli et al. showed that neoadjuvant chemotherapy was useful in avoiding amputation in four patients with large high grade STS [44]. Similarly, Meric and colleagues were able to show that with neoadjuvant chemotherapy 13 % of patients were downstaged sufficiently to change the scope of the operation performed. Unfortunately, another 9 % of patients had tumor progression requiring a more aggressive operation than was initially planned [45].

The selection of chemotherapy for a patient with STS must depend on the particular sarcoma histology as certain subtypes of STS such as gastrointestinal stromal tumor (GIST) or dermatofibrosarcoma protuberans (DFSP) are considered relatively resistant to traditional cyotostatic agents.

The sensitivity of sarcomas to doxorubicin has been known for decades [46]. Response rates for single agent therapy range from 9 % to 27 % [47, 48]. There is a strong dose-response curve, with higher response rates in patients who receive greater than or equal to 60 mg/m² per dose [49, 50]. Pegylated liposomal doxorubicin (Doxil or Caelyx) is a formulation of doxorubicin in which a polyethylene glycol layer surrounds doxorubicin containing liposomes. Phase II studies of this agent show similar response rates to that of doxorubicin [48]. Clinically there is particular interest in Doxil for patients with angiosarcoma based on partial and complete responses seen in case reports and retrospective series [51, 52].

The single agent activity of dacarbazine in unselected STS groups is 18 % [53]. Temozolomide is an oral agent and pro-drug of the active metabolite of dacarbazine, but does not require hepatic activation. When given at a dose of 75 mg/m² or 100 mg/m² continuously for 6 weeks, temozolomide is an active agent, with a response rate of 15.5 % by WHO criteria in STS [54]. This activity is not seen with different dosing schedules [55–57]. Patients with leiomyosarcoma (LMS) tend to be particularly sensitive to these agents. In a study of dacarbazine at various doses in second or third line STS, of the three partial responses, two were in patients with LMS [58]. ORR of 45.5 % have also been noted with temozolomide in patients with gynecological LMS.

Ifosfamide is an alkylating agent with similar single agent activity as doxorubicin [59, 60]. A dose-response curve also exists for this agent, as patients who progress on ifosfamide at doses less than or equal to 10 g/m² show responses when exposed to high-dose ifosfamide (doses >10 g/m²) [61–64]. Ifosfamide appears to be particularly active in synovial sarcoma, based on data from retrospective and small patient series. Rosen and colleagues treated 13 patients with pulmonary metastases from synovial sarcoma with high dose ifosfamide (14–18 g/m² per cycle). All patients had an objective response to therapy and four patients had a CR [63]. Median OS was 20 months (range 2–43 months). Three of the patients were able to undergo metastasectomy, rendering them disease free.

Both ifosfamide and dacarbazine have been added to doxorubicin with an increase in response rate in some studies, but without improvement in OS [50, 65–67]. In a multicenter randomized trial of doxorubicin and dacarbazine versus the same combination plus ifosfamide and mesna, the addition of ifosfamide improved ORR from 17 % to 32 % (P < 0.002), but median survival in both groups was similar (12 versus 13 months) and there was worsening hematologic toxicity [66]. In this study partial response was defined as a reduction of the product of perpendicular diameters of all measurable lesions for at least 4 weeks by at least 50 %. Subsequent use of hematopoietic growth factors have allowed for dose escalation and shortened duration of neutropenia [68, 69]. However, in at least one study this dose escalation did not improve outcomes [70]. Currently, most clinicians would therefore reserve combination chemotherapy for patients with good performance status who are either symptomatic from their disease or in whom a complete response could be anticipated. Single agent chemotherapy for palliation and potential prolongation of life would therefore be recommended in patients with widespread disease. Recent data are available from the prospective randomized European Organization for Research and Treatment of Cancer (EORTC 62012) trial comparing single-agent doxorubicin to the combination of doxorubicin and ifosfamide in patients with unresectable or metastatic sarcoma in the first-line setting. There was no significant difference in OS between groups (12.8 vs 14.3 months, HR 0.83 [95.5% CI 0.67-1.03]) for doxorubicin and the combination respectively. Median PFS was higher for the combination arm (4.6 vs 7.4 months, HR 0·74 [95% CI 0.60-0.90]) as was overall response rate (26% vs 14%, P<0.0006) for doxorubicin and the combination respectively.[71].

Gemcitabine is a nucleoside analogue with activity in STS that is dependent on the method of administration due to the formation of the metabolite gemcitabine triphosphate. Activity of the combination of gemcitabine and docetaxel was first reported in patients with advanced LMS [72]. Docetaxel is a microtubule inhibitor in the taxane family. The activity of single agent docetaxel in STS is poor, with one study showing a 0 % response rate [73, 74]. In patients with angiosarcoma or Kaposi’s sarcoma, however, another microtubule inhibitor, paclitaxel, has shown clinical benefit [75, 76]. Preclinical data have demonstrated synergistic activity of gemcitabine followed by docetaxel [77]. This combination was subsequently tested in 34 patients with unresectable LMS after failure of 0–2 prior chemotherapy regimens [72]. Gemcitabine was given at 900 mg/m² over 90 min on days 1 and 8 of a 21 day cycle. Docetaxel was given on day 8 only at a dose of 100 mg/m². The ORR by RECIST was 53 % with a PFS of 5.6 months. Although the majority of these patients had a uterine sarcoma, there were five patients with a non-uterine LMS, two of whom had an objective response. In a follow-up study by the Gynecology Oncology Group, the same combination was tested in patients with advanced uterine LMS in the first line setting [78]. The ORR was 35.8 % (RECIST), with a PFS of 4.4 months and OS of more than 16 months. The high response rates seen in LMS prompted investigators to study this combination in a broad range of STS histologies. A phase II randomized trial of fixed dose rate gemcitabine versus fixed dose rate gemcitabine in combination with docetaxel enrolled 122 previously treated patients [79]. Median PFS and OS were 6.2 and 17.9 months for the gemcitabine and docetaxel group and 3 and 11.5 months for the gemcitabine alone group, supporting the synergistic activity of these two drugs. Additional responses were seen in high-grade undifferentiated pleomorphic sarcomas, pleomorphic liposarcoma and rhabdomyosarcoma. In other retrospective data additional responses have also been seen in angiosarcomas, osteosarcomas, malignant peripheral nerve sheath tumors and Ewing’s sarcoma [77].

The activity of single agent paclitaxel in unselected sarcoma subtypes is poor. Anecdotal data suggesting activity of paclitaxel in patients with scalp angiosarcoma [80] led to a prospective phase II study by the French Sarcoma Group. Patients with metastatic or advanced angiosarcoma including non-cutaneous/visceral disease were given paclitaxel 80 mg/m² weekly for 3 weeks out of 4. An ORR of 19 % by RECIST after six cycles was observed [75]. Median time to progression was 4 months with OS of 8 months. The drug was well tolerated with grade 3 and 4 toxicities related to cytopenias, nausea and vomiting, fatigue, CNS toxicity, and mucositis. There was one death due to thrombocytopenia. The authors concluded that weekly paclitaxel was well tolerated and showed clinical benefit in patients with angiosarcoma.

Trabectedin (ET-743; Johnson and Johnson) is a marine derived alkaloid that uniquely binds DNA through the minor groove. It is approved in Europe for STS patients who have failed prior anthracycline therapy. Two phase II trials from the US and Europe investigated the 1.5 mg/m² dose as a 24 h continuous infusion in patients with previously treated metastatic STS. The ORR ranged from 4 % to 8 % by WHO criteria with PFS of less than 2 months [81, 82]. Data from phase II and compassionate use trials, have confirmed these findings with response rates of 4–14 % and clinical benefit rates of 14–52 % in pretreated patients [83–85]. The response to single agent trabectedin in the first line setting parallels that of the combination of doxorubicin and ifosfamide [86]. In 36 patients with metastatic STS, trabectedin given at the previous dose and schedule demonstrated an ORR of 17.1 % by WHO criteria, with a PFS of 1.6 months and OS of 15.8 months [86]. Common toxicities of trabectedin include cytopenias and a reversible transaminitis which can be attenuated with the use of prophylactic dexamethasone [87]. Patients with liposarcomas and LMS appear to be particularly sensitive to this agent [88–90], possibly due to deficient homologous recombination repair pathways in these subtypes [91, 92

Stay updated, free articles. Join our Telegram channel

Join