33.4.1 Surgery

The surgical principles for management of skull base sarcomas are not materially different from other tumors that occur at these sites. The metastatic spread of sarcomas seldom involves lymph nodes, and extensive lymph node dissections that may be common for the treatment of carcinomas are not required during sarcoma resections in the absence of clinical lymph node involvement.11 Wide local excision with microscopically negative margins is the standard of care for sarcomas of both soft tissue and bone, though the confines of the skull base can make this difficult to achieve.18 In the International Collaborative Study Group series of anterior skull base sarcomas, involved sites included the orbit in 53% of patients and intracranial extension in 28% of patients.11 Consequently, although adequate surgery alone does remain an acceptable standard of care in those patients in whom it can be accomplished, multimodality therapy has an important role in the management of patients whose tumors are in proximity to or involving vital structures or who are otherwise not candidates for a wide excision.

33.4.2 Radiation

Multiple clinical reports have led to the routine use of adjuvant radiation therapy to aid in the local control of extremity soft tissue sarcomas. Two randomized prospective studies have demonstrated that administration of radiation in the adjuvant setting reduces the rate of local recurrence in extremity and or chest wall sarcomas, though neither demonstrated improvements in overall survival.19 ^,​ 20

There are fewer prospective data regarding the utility of radiation in patients who have head and neck sarcomas, though several retrospective studies support its use. Tran and colleagues at UCLA reported on 164 cases of head and neck sarcomas seen over a 33-year period. Patients receiving adjuvant radiation had a locoregional control rate of 90%, compared with 52% for those receiving surgery without radiation.21 Also in support of adjuvant radiation therapy, a case series from the Princess Margaret Hospital suggested that head and neck sarcoma patients who had clear surgical margins had local relapse rates similar to those of patients with microscopically positive margins who received postoperative radiation therapy (26% versus 30%, respectively, at 5 years).22 Data regarding radiation in skull base sarcomas is rare. In the series published by Prabhu and colleagues at MD Anderson Cancer Center, radiation was delivered postoperatively to 62%. The researchers were unable to identify a progression-free survival (PFS) or overall survival (OS) benefit, though admittedly the numbers involved in this study were small.23

The utility of preoperative rather than postoperative radiation has also been investigated. Possible advantages to preoperative treatment include lower required dose, smaller radiation field, and excision of treated tissues at the time of surgery. These considerations may allow sparing of radiosensitive structures near the tumor and minimization of late toxicities.24 ^,​ 25 A prospective randomized study conducted by the National Cancer Institute of Canada Clinical Trials Group (NCIC CTG) in extremity sarcomas demonstrated a higher risk of wound healing complications in patients undergoing preoperative versus postoperative radiotherapy (35% vs. 17%) with similar disease outcomes.24 Subsequent follow-up demonstrated that patients in the postoperative therapy cohort experienced greater fibrosis and trended to experience more edema and joint stiffness.26

A prospective series of patients receiving preoperative radiation in head and neck sarcoma and using the same criteria for definitions of wound complications from the NCIC CTG study was reported by the Princess Margaret Hospital. There the investigators documented a 20% rate of major wound complications, lower than that seen in extremity sarcoma studies. No patients had loss of vision, laryngeal function, or any neurological sequelae attributable to radiation therapy. The authors postulated that increased vascularity in the head and neck, and increased use of flaps for wound closure owing to cosmetic and functional considerations, might explain the lower-than-expected rate of wound complications.27 Ultimately, the decision to favor preoperative versus postoperative radiation is a complex one that must take into account the expected sensitivity of the tumor to radiotherapy, the toxicity to adjacent tissues, expected margins, and the cosmetic and functional results of surgery. Additionally, many surgeons, including several at our own center, prefer to operate prior to radiation treatment. We suggest that the appropriate sequencing of therapies should be carefully considered for any patient who has base of skull sarcomas, ideally in a multidisciplinary conference setting at a specialty center.

33.4.3 Chemotherapy

The use of chemotherapy to treat unresectable or metastatic sarcoma has a history dating back to the 1960s. At that time, most sarcomas were treated with regimens that were most active against rhabdomyosarcomas and other pediatric sarcomas, with marginal activity in the subtypes that are more common in adults. Indeed, the employment of drugs such as vincristine, actinomycin, and cyclophosphamide in the treatment of pediatric rhabdomyosarcoma is as much a standard of care in 2017 as it was 50 years ago. For adult sarcomas, these drugs have limited activity, and the introduction of drugs such as doxorubicin and dacarbazine in the 1970s was the beginning of the divergence in treatment of adult sarcomas from pediatric tumors.

The choice of regimen is often impacted by goals of treatment. When given in the neoadjuvant setting, response rate (likelihood of a tumor shrinking) is considered of paramount importance in decreasing the likelihood of distant metastatic recurrence and facilitating an adequate resection. Whether this approach routinely succeeds in decreasing the extent of surgery is questionable. In one report of neoadjuvant chemotherapy for head and neck sarcomas, even with a response rate of 34%, the scope of surgery was lessened in only 13% of patients; 26% of patients had progression of disease on chemotherapy, and as a result 9% of patients had an increased scope of surgery.28 This illustrates that the decision to employ neoadjuvant chemotherapy is one that should take into account the patient’s risk of subsequently developing metastatic disease, the feasibility of resection prior to neoadjuvant treatment, and the patient’s ability to tolerate intensive chemotherapy.

33.4.4 Doxorubicin

Doxorubicin has been used in the treatment of sarcoma for nearly half a century and, despite new drug approvals in 2012 and 2015 for the treatment of soft tissue sarcoma, remains the most efficacious agent in terms of response rate and impact on survival.29 It has been noted that, unlike other tumors for which the drug is commonly used (including breast cancer, leukemias, and lymphomas), the administration of higher doses of doxorubicin to patients with sarcomas leads to higher response rates.30 When used as a single agent, responses are low according to response evaluation criteria in solid tumors (RECIST)—in the range of 14% in a large study in 2014.31 Consequently, the search for effective combinations has been of interest since the drug was introduced, starting with the alkylating agent dacarbazine.29 Multiple trials have shown increased response rate to the combination of doxorubicin and dacarbazine over doxorubicin alone, but none of these has definitively shown an improvement in overall survival.32 ^,​ 33 ^,​ 34 A more modern combination, felt to be most active, is the combination of dose-intensive doxorubicin and ifosfamide, which was shown to have a response rate of 66% in the initial description and 26% by RECIST in a recent study, with a trend toward improvement in overall survival over doxorubicin alone.31 ^,​ 35

Whether doxorubicin and ifosfamide in combination improve overall survival over single-agent therapy remains controversial. The previously cited study, intended to settle this question, was powered to find a 10% improvement in overall survival. They found a 9% improvement, which missed the cutoff for statistical significance.36 Additional evidence for combination chemotherapy can be found in the recently reported METASARC retrospective study conducted by the French Sarcoma Group, which looked at 2,165 patients treated in coordination with three sarcoma reference centers, with expert pathology review of all cases. In this study, patients who received combination chemotherapy in the front-line setting had improved overall survival over those who did not, to a statistically significant degree. Some of this effect might result from the selection of more fit patients for combination therapy—patients who might have done better regardless. Even so, this large and well-curated database provides the best available look at the natural history of patients treated for sarcoma and, in our opinion, is compelling evidence of the superiority of combination chemotherapy in appropriate patients. What is generally agreed on is that for patients who require tumor shrinkage to achieve palliation or to proceed to surgery, doxorubicin and ifosfamide in combination is the most active regimen. This has been incorporated into National Comprehensive Care Network (NCCN) guidelines.

The most recently examined doxorubicin-based combination uses a new anti-PDGFRα monoclonal antibody, olaratumab. The drug received accelerated approval from the FDA in 2016 for use in combination with doxorubicin on the basis of a 12-month improvement in overall survival over doxorubicin alone in a randomized phase II study. Interestingly, response rate was only modestly improved (18.2 vs. 11.9%), and the improvement in progression-free survival was also small (6.6 vs. 4.1 months).37 A larger confirmatory phase III study has accrued but has not yet been reported. In view of its impressive overall survival benefits, this combination has gained widespread acceptance as front-line treatment for patients who have metastatic disease. For fit patients, we still favor doxorubicin and ifosfamide as front-line therapy considering the limited data available to support the survival benefit seen in the randomized doxorubicin and olaratumab phase II study.

33.4.5 Gemcitabine and Docetaxel

With the approval of gemcitabine for pancreas cancer and docetaxel for breast cancer in 1996, several phase II studies examined the activity of the drugs in sarcomas.38 ^,​ 39 ^,​ 40 ^,​ 41 ^,​ 42 ^,​ 43 ^,​ 44 Gemcitabine clearly has single-agent activity, with response rates ranging from 3 to 18% in several phase II studies.38 ^,​ 39 ^,​ 40 ^,​ 41 The response rate to single-agent docetaxel was more modest, ranging from 0 to 22%.42 ^,​ 43 ^,​ 45 Considering the nonoverlapping toxicity, the combination of the two drugs was examined. Retrospective and early phase II studies suggested that the combination might result in synergistic activity, with response rates of 14 to 53%.46 ^,​ 47 It was not clear whether this improved efficacy over gemcitabine alone was the result of optimal administration of the gemcitabine with use of a slower fixed-dose-rate infusion or of the addition of the marginally active docetaxel. To address this question, two randomized studies have been conducted. The first, conducted by the Sarcoma Alliance for Research Collaboration (SARC), enrolled patients who had a variety of soft tissue sarcomas. To minimize the number of patients required to get a statistically significant result, the trial employed Bayesian adaptive randomization to assign patients to treatment groups and determine when to stop the study. A total of 122 patients who had a variety of soft tissue sarcomas were assigned to receive either gemcitabine alone or gemcitabine and docetaxel. With the adaptive randomization, 73 patients received combination therapy and 49 received gemcitabine alone. Median progression-free survival was 6.2 months for the combination arm and 3.0 months for the single-agent arm. Overall survival was also improved, at 17.9 versus 11.5 months. Importantly, combination therapy was associated with more significant side effects and many more treatment discontinuations as a result of toxicity.48

In apparent conflict with these results is a second randomized study, conducted by the French Sarcoma Group, that randomized patients who had uterine and nonuterine leiomyosarcoma to gemcitabine and docetaxel versus gemcitabine alone. This study demonstrated no improvement in progression-free or overall survival with the addition of docetaxel to single-agent gemcitabine.49 It should be noted, however, that the study was limited to patients who had leiomyosarcoma, whereas the patients who had the best results on the SARC study had unclassified pleomorphic sarcomas (called malignant fibrous histiocytoma at the time of the study). Furthermore, the French study used more gemcitabine in the control arm than the SARC study did.

Our approach has been to use gemcitabine and docetaxel in combination in fit patients, in most cases as a second-line treatment after doxorubicin-containing combination therapy. However, considering the mixed evidence regarding the importance of docetaxel in the combination, we are quick to dose reduce or eliminate the drug in patients of advanced age or marginal performance status or who have docetaxel-related toxicity.

33.4.6 Trabectedin

Approved in the United States in 2015 and in Europe in 2007, trabectedin has become an important treatment option, particularly for patients who have pretreated liposarcomas and leiomyosarcomas. It also has clear activity in several other sarcoma subtypes. The drug was initially isolated from a marine tunicate and is now produced synthetically. Its main mechanism of action appears to involve binding of the minor groove of DNA, though other mechanisms have also been implicated.50 A number of phase II studies resulted in the approval of trabectedin for the treatment of soft tissue sarcomas in Europe in 2007.51 ^,​ 52 ^,​ 53 ^,​ 54 Based on evidence of increased efficacy in leiomyosarcomas and liposarcomas, a phase III study was conducted that randomized patients who had those tumor types to treatment with trabectedin or dacarbazine. The study did not meet its primary endpoint of improved overall survival, but it did demonstrate significantly improved progression-free survival at 4.5 months for trabectedin and 1.5 months for the dacarbazine control.55 Based on these data, the FDA approved trabectedin in 2015 for patients who have liposarcomas and leiomyosarcomas previously treated with anthracycline based therapy.

Trabectedin has been of particular interest in patients who have chromosomal translocation-associated sarcomas, especially myxoid liposarcoma. A phase III study that randomized patients who had translocation-associated sarcomas to treatment with trabectedin versus doxorubicin has been conducted. Nearly half of the patients on the trial had myxoid liposarcoma. Ultimately, the study was not able to demonstrate a difference in PFS or OS, because too many of the patients received additional therapy prior to progression of disease and were subsequently censored. The study was ultimately underpowered, so definitive conclusions about relative efficacy of the two drugs in the front line cannot be drawn, although it is notable that 23.5% of patients in the trabectedin arm who were previously deemed unresectable proceeded to surgery.56 Consequently, trabectedin is often used earlier than gemcitabine or other salvage therapies when treating myxoid liposarcoma, and to a lesser extent other translocation-associated tumors.

33.4.7 Eribulin

Approved for metastatic breast cancer in 2010, eribulin was approved for treatment of liposarcomas in 2015. The drug is an inhibitor of microtubule polymerization that is a derivative of a compound isolated from a marine sponge. An additional postulated mechanism of action is modulation of the tumor microenvironment and the metastatic potential of tumor cells.57 Treatment using the agent is associated with increased number of microvessels and possible improved delivery of subsequent lines of therapy.58

A phase II study of eribulin conducted by the European Organization for Research and Treatment of Cancer (EORTC) demonstrated activity of the drug in patients who had synovial sarcoma, leiomyosarcoma, or adipocytic sarcoma. Based on a customary cutoff of 12 week PFS > 30% as the threshold for advancing a drug in development for sarcoma, patients who had liposarcoma or leiomyosarcoma were enrolled in a randomized phase III study of eribulin versus dacarbazine. The study demonstrated a 2-month improvement in overall survival in those treated with eribulin without a significantly improved progression-free survival or response rate. Interestingly, preplanned subgroup analysis suggested that liposarcomas had improved overall survival and that leiomyosarcomas had not.59 Accordingly, the FDA approved eribulin in 2016 for the treatment of liposarcomas only rather than the entire studied population. Whether the lack of improvement in outcome for leiomyosarcomas was due to better-than-expected efficacy of dacarbazine in that group or poor efficacy of eribulin is unclear.

Subsequently presented data from the same phase III study showed that the survival benefit was significantly more pronounced (15.6 vs. 8.4 months) in liposarcomas alone than when leiomyosarcomas were included in the analysis. In contrast to the original study, this report demonstrated that in liposarcomas alone, PFS was modestly, though statistically significantly, improved after treatment with eribulin at 2.9 versus 1.7 months.60

33.4.8 Pazopanib

Pazopanib is a multitargeted tyrosine kinase inhibitor approved for the treatment of soft tissue sarcomas. The phase III PALETTE study, which led to the approval of pazopanib, excluded patients who had liposarcomas owing to lack of activity in an earlier study. The phase III trial demonstrated a 3-month improvement in progression-free survival with pazopanib over placebo (4.6 vs. 1.6 months), the primary endpoint of the study, but failed to demonstrate improved median overall survival. Six percent of patients in the pazopanib arm experienced a partial response, though a more substantial 67% of patients achieved stable disease (in contrast to 38% with placebo). Despite the modest response rate, the disease stabilization is of clinical interest. With limited options for patients who have metastatic or unresectable disease, pazopanib remains an important and commonly used palliative treatment option for patients who have progressed through other therapies.

33.5.1 Osteosarcoma

Although many varieties of soft tissue sarcoma can extend to involve the skull base, osteosarcomas are an exception in that they can arise from the skull base itself. This is a rare presentation, with head and neck osteosarcomas accounting for 10% of osteosarcoma cases. Of these, approximately 80 to 85% arise from the maxilla and mandible (jaw). The behavior of these tumors is distinct from the more common osteosarcoma of the long bones. The rate of distant metastatic disease is low, approximately 7 to 12% in two large single-institution reports. Local recurrence is more problematic, with rates of 18 to 34% in the same two case series.61 ^,​ 62 Predictors of improved local control include tumors < 4 cm in size, fibrous histology, low histologic grade, and favorable response to chemotherapy. Positive margins are also intuitively associated with high risk for local relapse and are correlated with worsened overall survival. Retrospective data support the use of postoperative radiation therapy for patients who have positive margins at the time of surgery, demonstrating improved local control rates.61

Response to chemotherapy in head and neck sarcomas is also less robust than in more common limb osteosarcomas. In the Memorial Sloan Kettering series of patients receiving neoadjuvant chemotherapy, a relatively high 73% of patients had histologically poor responses to neoadjuvant treatment. Combined with lower metastatic potential, this apparently decreased chemosensitivity makes the routine use of neoadjuvant chemotherapy, which is standard for patients with osteosarcomas at other sites, less appealing for those who have jaw primaries.

Considering the lack of chemosensitivity and lower rates of distant metastatic disease, surgical resection without radiation or chemotherapy as up-front management of amenable tumors is an acceptable standard of care for jaw osteosarcomas. For tumors located in the mandible, where anatomy allows for relatively straightforward margin-negative resection, this is the most commonly applied treatment strategy. Tumors originating in the maxilla and other nonmandible skull sites, including the skull base, are often selected for neoadjuvant treatment on the basis of anticipated difficulty in obtaining negative margins. The data demonstrating improved outcomes treated in this fashion are limited, though Memorial Sloan Kettering demonstrated that patients who received neoadjuvant treatment had decreased rates of distant metastatic disease compared with historical controls.62 Prognosis varies based on the case series, with 5-year overall survival rates ranging from 55 to 63% in several larger case series.61 ^,​ 62 ^,​ 63

The precise natural history of osteosarcomas arising in bones of the skull other than the mandible and maxilla is poorly characterized. We have seen metastatic disease from such tumors and thus treat osteosarcoma of the skull outside the jaw as if it were similar to that of other skeletal sites.

There is no universal agreement on optimal chemotherapy regimen for nongnathic osteosarcoma. Most regimens, including those used at our center, incorporate doxorubicin and cisplatin (AP). Many others also incorporate methotrexate into the treatment, resulting in the MAP regimen. Whether methotrexate improves outcomes when added to doxorubicin and cisplatin in adult patients is not clear, with two randomized studies having demonstrated no improvement in survival.64 ^,​ 65 Nevertheless, based on the approach in the pediatric population, it is often employed in adult osteosarcoma patients. Typically, after a course of neoadjuvant therapy, the tumor is resected and additional adjuvant chemotherapy is administered. Here also, there is no clear standard of care, and many centers continue with MAP irrespective of the pathologic response seen at the time of surgery. They cite the EUROAMOS-1 trial, which demonstrated that intensification of adjuvant chemotherapy with ifosfamide and etoposide in addition to MAP in patients who had suboptimal (< 90%) necrosis did not improve outcomes over those who simply received additional MAP.66

The approach at our center has been to use doxorubicin and cisplatin in the neoadjuvant setting. Based on the level of tumor necrosis, we treat either with additional anthracycline-based therapy or, in poor responders, with high-dose ifosfamide followed by high-dose methotrexate. In our hands, this intensified approach in patients who have otherwise poorer prognoses results in similar outcomes irrespective of the level of necrosis.67 ^,​ 68

33.5.2 Rhabdomyosarcoma

Rhabdomyosarcoma is a rare tumor most commonly seen in children, in whom it constitutes more than 50% of soft tissue sarcomas—in contrast to adults, in whom the figure is closer to 3%. The childhood form of this disease carries a better prognosis, with cure rates as high as 70% when using multimodality treatment. In adults, the prognosis is poorer, with a 5-year OS of 45.7% in a large cohort of patients treated at the Istituto Nazionale Tumori in Milan, Italy. Of the 171 adult patients in this series, 53% had primary tumors in the head and neck. In this report, patients who received optimal chemotherapy (defined as a regimen containing an anthracycline along with cyclophosphamide or ifosfamide, given for at least eight cycles) and complete surgical resection, with radiation in the event that the tumor was incompletely resected, had a better outcome, with a 5-year OS of 61.5%. Illustrating the importance of appropriate chemotherapy in this disease, patients who did not receive chemotherapy had a 5-year relapse- and event-free survival of 39.4%, compared with 64.1% in the group that received optimal chemotherapy.69 Thus, unlike other sarcomas of the head and neck, for which local therapy generally takes precedence, chemotherapy is a critical component of the treatment of this tumor type. Considering the apparent chemosensitivity, neoadjuvant treatment’s likelihood of making surgery less extensive is likely to be higher than for other sarcomas.

Our approach to patients who have rhabdomyosarcoma is to use doxorubicin and ifosfamide at doses similar to other soft tissue sarcomas, with the addition of vincristine. Many other centers use vincristine, doxorubicin, and cyclophosphamide, often alternating with ifosfamide and etoposide. Actinomycin-D is more effective in these tumors than most adult sarcomas and can be used in place of doxorubicin, particularly when patients have received doxorubicin in the past. In the second line, topoisomerase inhibitors paired with alkylating agents are frequently used.