Systemic treatment of metastatic soft tissue sarcoma

INTRODUCTION — Soft tissue sarcomas (STS) are a heterogeneous group of rare tumors that arise from mesenchymal cells at all body sites. The malignant precursor cell(s) can differentiate along one or several lineages, such as muscle, adipose, fibrous, cartilage, nerve, or vascular tissue. These tumors arise most often in the limbs (particularly the lower extremity), followed in order of frequency by the abdominal cavity/retroperitoneum, the trunk/thoracic region, and the head and neck. (See "Clinical presentation, histopathology, diagnostic evaluation, and staging of soft tissue sarcoma", section on 'Clinical presentation' and "Clinical presentation, histopathology, diagnostic evaluation, and staging of soft tissue sarcoma", section on 'Introduction'.)

There are more than 50 histologic subtypes of STS, many of which are associated with distinctive clinical profiles, responses to individual therapies, and prognoses. Even within a single histologic subtype, a heterogenous array of translocations or other molecular changes may be observed. While in the past these tumors were all "lumped" together and treated similarly, consensus is emerging that selection of treatment should be histology driven, particularly in the setting of advanced disease. (See 'Histology-driven treatment' below.)

Systemic chemotherapy for metastatic non-GIST (gastrointestinal stromal tumor) and nonuterine STS will be reviewed here. Primary management of localized STS, adjuvant and neoadjuvant chemotherapy, surgical management of metastatic STS, and systemic therapy for certain specific types of STS (GIST, rhabdomyosarcoma, Ewing sarcoma, uterine sarcomas, desmoids, dermatofibrosarcoma protuberans [DFSP], Kaposi sarcoma, and select uncommon sarcoma subtypes) are discussed separately:

●(See "Surgical resection of primary soft tissue sarcoma of the extremities".)

●(See "Overview of multimodality treatment for primary soft tissue sarcoma of the extremities and superficial trunk".)

●(See "Breast sarcoma: Treatment".)

●(See "Treatment and prognosis of uterine leiomyosarcoma".)

●(See "Clinical features, evaluation, and treatment of retroperitoneal soft tissue sarcoma".)

●(See "Head and neck sarcomas".)

●(See "Adjuvant and neoadjuvant chemotherapy for soft tissue sarcoma of the extremities".)

●(See "Surgical treatment and other localized therapy for metastatic soft tissue sarcoma".)

●(See "Tyrosine kinase inhibitor therapy for advanced gastrointestinal stromal tumors".)

●(See "Rhabdomyosarcoma in childhood, adolescence, and adulthood: Treatment".)

●(See "Treatment of Ewing sarcoma".)

●(See "Treatment and prognosis of uterine leiomyosarcoma".)

●(See "Desmoid tumors: Systemic therapy".)

●(See "Dermatofibrosarcoma protuberans: Treatment".)

●(See "Classic Kaposi sarcoma: Clinical features, staging, diagnosis, and treatment" and "AIDS-related Kaposi sarcoma: Staging and treatment".)

●(See "Uncommon sarcoma subtypes".)

GENERAL PRINCIPLES

Patterns of spread — While local complications from primary or recurrent sarcomas can cause significant morbidity and occasional mortality, the most life-threatening aspect of sarcomas is their propensity for hematogenous dissemination. The pattern of tumor spread varies according to tumor type and location:

●For most STS of the extremity, chest wall, or head and neck, the primary metastatic site is the lungs [1,2]. However, there are exceptions. Extrapulmonary metastases to the spine, bony pelvis and other bones, retroperitoneum, and paraspinous soft tissues predominate with myxoid/round cell liposarcomas, although lung metastases develop eventually in almost all [3]. (See "Clinical presentation, histopathology, diagnostic evaluation, and staging of soft tissue sarcoma", section on 'Pattern of spread' and "Clinical presentation, histopathology, diagnostic evaluation, and staging of soft tissue sarcoma", section on 'Pattern of recurrence'.)

●With retroperitoneal and visceral sarcomas, the primary site of failure is local. These tumors spread hematogenously to the liver, and less commonly, distant metastases outside of the abdomen/pelvis can also occur [4].

●Spread to locoregional lymph nodes is rare, except with clear cell and epithelioid sarcomas, angiosarcomas, synovial sarcomas, and alveolar rhabdomyosarcoma [5].

Natural history of metastatic disease and implications for treatment — The majority of patients who develop metastatic STS are incurable; however, therapeutic nihilism is unwarranted for the following reasons:

●Potentially curative options should be sought in appropriate patients so that the opportunity for cure is not overlooked. As an example, in selected patients, resection of pulmonary metastases is feasible, with a reported five-year survival of 25 to 40 percent [6,7]. (See "Surgical treatment and other localized therapy for metastatic soft tissue sarcoma".)

●For patients with unresectable disease, judicious use of systemic therapy provides meaningful palliation and may prolong survival. The selection of systemic therapy must be individualized and based on several factors, including the histology and biologic behavior of the disease, as well as the health status and preferences of the patient. (See 'Overview of the therapeutic approach' below.)

●Median survival after development of distant metastases is 12 to 19 months, and 20 to 25 percent of patients with unresectable disease are still alive at two to three years; however, these statistics can vary significantly based on primary histologic subtype and evolving treatment paradigms [1,8,9]. The prognostic factors for more prolonged survival differ from those predicting response to chemotherapy [8], suggesting that survival is more dependent on disease biology than solely on treatment-associated considerations. This variability in the natural history of metastatic STS can be illustrated by the following reports:

•In a report in an era before gastrointestinal stromal tumors (GISTs) were recognized as a distinct clinical entity, outcomes from over 2000 patients treated with anthracycline-based chemotherapy for advanced STS were examined [8]. Median survival was approximately one year for the entire cohort, regardless of the specific regimen used. The likelihood of a chemotherapy response was greater in younger patients with a high-grade non-leiomyosarcoma (LMS) histology and no hepatic metastases, while longer median survival was predicted by a good performance status, a low-grade histology, the absence of liver metastases, and a longer recurrence-free interval.

•In a more modern series, 488 patients were treated with first-line chemotherapy for advanced and/or metastatic STS [10]. In multivariate analysis, age <40 years, liposarcoma or synovial sarcoma histology, lack of bone metastases, and combination rather than single-agent therapy were associated with better outcome.

These principles can be applied to therapeutic decision-making. As an example, for some patients with asymptomatic, low-grade, unresectable disease (eg, intra-abdominal well differentiated liposarcoma), it might be reasonable to follow the patient without active chemotherapy. Conversely, for patients with a high-grade chemotherapy-sensitive tumor, such as synovial sarcoma or myxoid/round cell liposarcoma, early use of combination chemotherapy may be preferable.

Endpoints to define benefit — Unresectable metastatic STS is, with rare exception, a fatal disease eventually. A few patients may enter prolonged remission. However, for the majority of patients with metastatic STS, chemotherapy is administered with palliative intent to decrease tumor bulk, diminish symptoms, improve quality of life, and prolong survival.

The specific endpoints that best reflect benefit from systemic therapy in metastatic STS remain unclear. Objective response rates (as judged by a decrease in the size of the measurable lesions) are increasingly considered to be poor surrogates for benefit in this family of cancers. Even in cases where systemic treatment successfully induces massive tumor cell kill in vivo, hyalinized acellular tissue remains, leading to a falsely negative assessment of the true response to treatment and an underestimation of antitumor efficacy. Therefore, in clinical practice, lack of progression is often used as a measure of clinical benefit. The "disconnect" between objective tumor response and disease stabilization is particularly evident in studies of molecularly targeted therapies and drugs such as trabectedin.

Increasing attention is being paid to other important indicators of clinical outcomes, such as progression-free survival (PFS), the progression-free rate at a specific time point, the percentage survival at a given time point, overall survival, and disease stabilization. A therapeutic agent that is associated with low objective antitumor response may slow tumor progression and prolong survival. Stabilization of disease is increasingly viewed as a realistic endpoint for metastatic STS. However, the generation of clinical and radiologic data to support these alternative endpoints requires rigorous attention to the consistency of follow-up.

Histology-driven treatment — Most studies of chemotherapy for metastatic STS have been hampered by the admixture of a variety of histologic subtypes in the analysis of outcome, making it difficult to assess the clinical activity of any given treatment in a specific histology. The interpretation of clinical trial results will be heavily weighted by the distribution of histologic subtypes. This, in turn, complicates the assessment of chemotherapy efficacy, making it impossible to determine whether high or low response rates are due to the specific treatment or to the specific population under study [11].

These data have led to the emergence of the concept of histology-driven treatment, rather than a "one size fits all" approach to therapy, in patients with metastatic STS. As examples:

●Synovial sarcomas and myxoid/round cell liposarcomas are among the more chemotherapy-sensitive subtypes when specific agents are used [12-15]. In particular, myxoid/round cell liposarcomas tend to be sensitive to doxorubicin-based chemotherapy and to trabectedin, and synovial sarcoma appears sensitive to alkylating agents, such as ifosfamide [15]. In contrast, other subtypes, such as clear cell sarcoma and fibromyxoid sarcoma (Evans tumor), appear to have lower response rates to conventional anthracycline-based and ifosfamide-based chemotherapy [8,16]. GISTs are well recognized for the inactivity of standard cytotoxic agents.

●Trabectedin appears to have activity in LMS and liposarcomas (particularly the myxoid/round cell subtype), and perhaps other sarcomas, while eribulin has the greatest activity in dedifferentiated or pleomorphic liposarcoma. (See 'Trabectedin' below and 'Eribulin' below.)

●In contrast with virtually every other histology, angiosarcomas are sensitive to single-agent taxanes, with relatively high objective response rates and the possibility of prolonged disease control. (See "Head and neck sarcomas", section on 'Treatment (resectable disease)' and 'Taxanes (angiosarcoma)' below.)

●VEGF receptor inhibitors such as pazopanib or sunitinib are active in patients with solitary fibrous tumor (SFT)/hemangiopericytoma, alveolar soft part sarcoma, and extraskeletal myxoid chondrosarcoma (at least those that carry the characteristic EWSR1-NR4A3 fusion gene [17]).

•(See 'Pazopanib' below.)

•(See "Solitary fibrous tumor", section on 'Progressive disease'.)

•(See "Uncommon sarcoma subtypes", section on 'Alveolar soft part sarcoma'.)

•(See "Uncommon sarcoma subtypes", section on 'Extraskeletal myxoid chondrosarcoma'.)

●Benefit has also been shown for the mechanistic (mammalian) target of rapamycin (mTOR) inhibitor nab-sirolimus in patients with neoplasms with perivascular epithelioid cell differentiation (PEComas) that are characterized by dysregulated mTOR signaling. (See 'Nab-sirolimus' below.)

●Pexidartinib, which targets the colony-stimulating factor 1 receptor (CSF1R), is effective in patients with tenosynovial giant cell tumor (TGCT). CSF1R has biologic relevance in TGCT because overexpression results in tumorigenesis. Targeted inhibition of this pathway represents a logical therapeutic approach. (See "Treatment for tenosynovial giant cell tumor and other benign neoplasms affecting soft tissue and bone", section on 'Tenosynovial giant cell tumor'.)

●There are data supporting the activity of imatinib in patients with metastatic dermatofibrosarcoma protuberans (DFSP) and of imatinib, sorafenib, and pazopanib in patients with desmoid tumors. (See "Dermatofibrosarcoma protuberans: Treatment", section on 'Imatinib' and "Desmoid tumors: Systemic therapy", section on 'Noncytotoxic approaches'.)

For most STS tumor types that are anthracycline sensitive, anthracycline with or without ifosfamide is considered an appropriate first-line therapy. However, histology-driven treatment is recommended as initial therapy for some anthracycline-resistant histologies (eg, alveolar soft part sarcoma, SFT, clear cell sarcoma, PEComa, and DFSP). (See 'Initial therapy' below.)

OVERVIEW OF THE THERAPEUTIC APPROACH

Initial therapy — Enrollment in a clinical trial is always preferred if available. If a clinical trial option does not exist, the following represents our general approach to initial therapy for metastatic STS types other than gastrointestinal stromal tumor (GIST) or desmoid tumor, as outlined in the algorithm (algorithm 1).

Patients should initially be assessed for their potential responsiveness to doxorubicin to gauge whether an anthracycline should be used as part of first-line therapy or if other agents should be considered first based on histology.

Anthracycline-sensitive histologies — For patients with a good performance status, minimal comorbidity, and an STS histology that is known to have at least some sensitivity to anthracyclines (ie, liposarcoma, leiomyosarcoma [LMS], epithelioid sarcoma [ES], synovial sarcoma, angiosarcoma, undifferentiated pleomorphic sarcoma, malignant peripheral nerve sheath tumor), doxorubicin with or without ifosfamide represents a first-line standard of care off study. If a patient is symptomatic and the immediate goal is tumor shrinkage, we will typically employ doxorubicin plus ifosfamide rather than doxorubicin alone. (See 'Doxorubicin-based regimens' below and "Uncommon sarcoma subtypes", section on 'Epithelioid sarcoma'.)

●An exception to this general rule may be angiosarcoma. Angiosarcomas are anthracycline sensitive, and at some institutions, patients with angiosarcoma are treated initially with an anthracycline-based regimen. However, angiosarcomas can also be highly sensitive to taxanes, and some clinicians may choose a taxane-based regimen for initial therapy. (See 'Taxanes (angiosarcoma)' below.)

●For patients with ES and indolent disease, tazemetostat is a reasonable alternative to an anthracycline-based regimen as initial therapy, and it has regulatory approval in this setting. For those with more aggressive disease, we prefer initial treatment with chemotherapy. (See "Uncommon sarcoma subtypes", section on 'Tazemetostat'.)

●Whether initial therapy with a gemcitabine-based regimen is preferable to doxorubicin with or without ifosfamide is uncertain. Given the data from the randomized phase III GeDDis trial (which showed similar progression-free survival [PFS] and overall survival, along with less toxicity, for single-agent doxorubicin in the first line [18]), we usually employ doxorubicin with or without ifosfamide in the first line for anthracycline-sensitive histologic subtypes. However, a gemcitabine-based combination may be offered as initial therapy in a patient for whom an anthracycline is relatively contraindicated (eg, clinical heart failure, prior treatment with >400 mg/m² doxorubicin in the adjuvant setting). (See 'Gemcitabine plus docetaxel versus doxorubicin' below.)

●Patients with a poorer performance status or extensive comorbidity who remain eligible for chemotherapy can be considered for treatment with pegylated liposomal doxorubicin (PLD), gemcitabine alone, or a gemcitabine-based combination.

●For older adults (age ≥60 years) with select anthracycline-sensitive histologies (eg, LMS, undifferentiated pleomorphic sarcoma, and angiosarcomas) who cannot tolerate or wish to avoid the potential toxicities of doxorubicin, pazopanib is one appropriate alternative option for initial therapy. However, data are limited for this approach, and other alternative chemotherapy agents with more supportive evidence are also available, such as PLD for all anthracycline-sensitive histologies, and paclitaxel or gemcitabine for those with angiosarcoma. (See 'Pazopanib' below.)

Anthracycline-resistant histologies — For advanced or metastatic STS histologies that are not sensitive to anthracyclines, other options may be offered as initial therapy, even though these may represent off-label use of existing drugs that are approved for other conditions:

●For patients with indolent histologies such as alveolar soft part sarcoma and extraskeletal myxoid chondrosarcoma, we suggest observation rather than initiation of systemic therapy. For those with symptomatic or progressive disease, options include pazopanib or sunitinib if a clinical trial option does not exist.

•(See 'Sunitinib' below and 'Pazopanib' below.)

•(See "Uncommon sarcoma subtypes", section on 'Alveolar soft part sarcoma'.)

•(See "Uncommon sarcoma subtypes", section on 'Extraskeletal myxoid chondrosarcoma'.)

●For SFT, active agents include dacarbazine with or without doxorubicin, temozolomide plus bevacizumab, pazopanib, or sunitinib; the choice between these agents is discussed separately. (See "Solitary fibrous tumor", section on 'Advanced and metastatic disease'.)

●For patients with malignant perivascular epithelioid cell differentiation (PEComas), we suggest initial therapy with nab-sirolimus rather than other systemic agents. For those without access to nab-sirolimus, other mTOR inhibitors (sirolimus, everolimus, temsirolimus) are alternatives. (See 'Nab-sirolimus' below.)

●For patients with tenosynovial giant cell tumor (TGCT) with relapsed, recurrent, or unresectable disease, options include pexidartinib or clinical trials. (See "Treatment for tenosynovial giant cell tumor and other benign neoplasms affecting soft tissue and bone", section on 'Pexidartinib (CSF1R inhibitor)'.)

●For unresectable, recurrent, or metastatic dermatofibrosarcoma protuberans (DFSP), imatinib may be indicated. (See 'Imatinib' below and "Dermatofibrosarcoma protuberans: Treatment", section on 'Treatment of locally advanced, recurrent, and metastatic disease'.)

Treatment at progression — For most patients with progression on the initial regimen, we prefer enrollment in a clinical trial if one is available. If protocol treatment is not available or is declined, our recommendations for therapy at progression for patients who retain a good performance status are histology driven:

●Trabectedin is approved in the United States for use in advanced LMS and liposarcoma after failure of anthracycline-based therapy. Most of the patients with LMS that were enrolled in the pivotal trial leading to trabectedin approval in the United States had received both doxorubicin-based and gemcitabine-based therapy. Myxoid/round cell liposarcoma appears particularly sensitive to trabectedin, and it is preferred for this specific histology in the second line. (See 'Trabectedin' below.)

●Eribulin is approved for use in advanced liposarcoma in the United States and for LMS and liposarcoma in other countries. Dedifferentiated liposarcoma and pleomorphic liposarcoma in particular appear to be more sensitive to eribulin than to trabectedin, although these agents have not been directly compared. As a result, eribulin is preferred for this specific histology in the second line. (See 'Eribulin' below.)

●Pazopanib is approved for advanced STS other than liposarcoma or GIST that has failed anthracycline therapy. Pazopanib has specific activity in solitary fibrous tumor (SFT)/hemangiopericytoma, alveolar soft part sarcoma, LMS, and angiosarcoma. (See 'Pazopanib' below.)

●For patients with advanced progressive angiosarcoma who received initial anthracycline-based therapy, weekly single-agent paclitaxel is a good option. Single-agent gemcitabine or a gemcitabine-based combination is also a reasonable option. (See 'Taxanes (angiosarcoma)' below.)

●Other options for second-line treatment and beyond include PLD, an ifosfamide-containing regimen, gemcitabine, or a gemcitabine-based combination. (See 'Pegylated liposomal doxorubicin' below and 'Ifosfamide' below and 'Gemcitabine and other agents' below and 'Gemcitabine-based combinations' below.)

•In our experience, undifferentiated pleomorphic sarcoma is the one histologic subtype that responds better to gemcitabine plus docetaxel than to other chemotherapy combinations, and we would choose this combination for second-line therapy after failure of initial doxorubicin with or without ifosfamide. In addition, gemcitabine plus docetaxel has activity in LMS, and we would choose it as second-line therapy in this population.

●Next-generation sequencing may identify candidates for molecularly targeted approaches. These include larotrectinib or entrectinib for individuals whose tumors have gene fusions involving one of the neurotrophic tyrosine receptor kinase (NTRK) genes, and pembrolizumab for the rare individual whose STS has high levels of microsatellite instability/deficient mismatch repair. (See 'Next-generation sequencing' below.)

EFFICACY OF CYTOTOXIC CHEMOTHERAPY — Most of the trials of conventional cytotoxic therapy were conducted in a variety of patients with different histologic subtypes of STS, evaluating various doses and schedules; few have analyzed outcomes separately according to histology.

Single-agent therapy — The only single agents that are consistently associated with response rates of more than 20 percent in metastatic STS are doxorubicin, epirubicin, and ifosfamide. An important exception is single-agent paclitaxel in angiosarcoma. (See 'Taxanes (angiosarcoma)' below.)

Even for these agents, the range of objective activity between various small (and even larger) trials is impressive, demonstrating the variability in disease sensitivity (noted above) and the fact that any given STS patient population could serve as an important confounding variable for interpretation of drug efficacy.

Doxorubicin — The sensitivity of STS to systemic chemotherapy was first demonstrated with single-agent doxorubicin in the early 1970s [19], and subsequent studies suggested a dose-response relationship. The threshold dose for optimal activity appears be ≥60 mg/m² per cycle, usually administered once every three weeks, with lower doses associated with inferior antitumor activity [20]. It is difficult to demonstrate a clinically meaningful dose-response relationship with single-agent doxorubicin at doses beyond 75 mg/m² per cycle, which has become the standard dose.

Even in modern multi-institutional series using 70 to 80 mg/m² per cycle, there is significant variability in reported response rates, which range from 10 to 25 percent [20-26]. The vast majority are partial rather than complete responses [27].

Doxorubicin is associated with reversible myelosuppression, mucositis, alopecia, nausea and vomiting, and both acute and chronic cardiotoxicity. This drug has a relatively narrow therapeutic index; even small variations of dose beyond 75 mg/m² can drastically worsen patient tolerance, and even 75 mg/m² is considered by many patients to be significantly more toxic than 60 mg/m². Infusional, rather than bolus, administration reduces the likelihood of cardiotoxicity. Other methods to diminish cardiotoxicity include the concomitant use of the cardioprotectant dexrazoxane or the use of liposome-encapsulated doxorubicin. (See "Clinical manifestations, diagnosis, and treatment of anthracycline-induced cardiotoxicity" and "Risk and prevention of anthracycline cardiotoxicity".)

Pegylated liposomal doxorubicin — Liposomal encapsulation appears to improve the side effect profile of doxorubicin, thereby increasing the therapeutic index. The most widely used preparation (pegylated liposomal doxorubicin [PLD; Doxil in the United States and Caelyx in Europe]) is a large liposome with polyethylene glycol (PEG) anchored within the lipid bilayer, which acts as a hydrophilic coating to prolong the circulating half-life of the liposome by preventing degradation within the reticuloendothelial system. The toxicity profile is somewhat different from nonencapsulated doxorubicin, mainly consisting of an acute hypersensitivity-like reaction (in approximately 8 percent of patients), palmar-plantar erythrodysesthesia (hand-foot syndrome), and esophagitis. (See "Cutaneous adverse effects of conventional chemotherapy agents".)

Liposomal anthracyclines are active in STS, but it is unclear if they are as efficacious as unencapsulated doxorubicin [21,28-30]. In a randomized phase II trial, PLD (50 mg/m² every four weeks) was compared with unencapsulated doxorubicin (75 mg/m² every three weeks) [21]. Although PLD was well tolerated and appeared to have similar efficacy compared with doxorubicin, the response rates in both arms were low (10 and 9 percent, respectively). Of note, there was a high proportion of gastrointestinal stromal tumor (GIST) in this study population, emphasizing the importance of disease subtype as a confounding factor in studies of chemotherapy efficacy. (See 'Histology-driven treatment' above.)

On the basis of this trial, rigorous clinical trial rules would have discarded both unencapsulated doxorubicin and PLD as putatively "inactive" agents in STS, with response rates <20 percent. This would be a major error since doxorubicin has important clinical activity, proven over the past 30 years, in patients with advanced STS. These data underscore the poor correlation between objective response rate and clinical benefit, and the need for caution in interpreting any clinical trial of a new drug as "negative" based on objective tumor shrinkage. (See 'Endpoints to define benefit' above.)

Others report response rates of 50 percent or higher with PLD, which are durable in many patients [31]. PLD is a widely used agent for metastatic STS, particularly outside of the United States and in patients aged 65 or older.

Similar to paclitaxel, PLD also has activity in angiosarcoma [30,32,33]. (See 'Taxanes (angiosarcoma)' below.)

Epirubicin — Compared with doxorubicin, epirubicin is less cardiotoxic on an mg-to-mg basis. At least two trials have directly compared epirubicin with single-agent doxorubicin for STS [25,34]:

●A European Organisation for Research and Treatment of Cancer (EORTC) trial randomly assigned 334 patients with untreated metastatic STS to doxorubicin (75 mg/m² on day 1) or to epirubicin on one of two different schedules (50 mg/m² daily for three days or 150 mg/m² on day 1) [34]. Neither epirubicin schedule was associated with a better response rate or survival, and cardiovascular and hematologic toxicity were worse than with doxorubicin.

●In the second trial, 210 patients received either epirubicin or doxorubicin (both dosed at 75 mg/m²) once every three weeks [25]. There was a slight trend toward a lower response rate with epirubicin (18 versus 25 percent), but response duration and median survival were similar.

Ifosfamide — Single-agent ifosfamide has similar antitumor activity to doxorubicin. Response rates range from 7 to 41 percent (average 25 percent) among patients who previously failed a doxorubicin-based regimen [11,35-42].

Ifosfamide is challenging to administer, especially considering the variety of published doses and schedules [11,42]. Ifosfamide can be delivered over one or several consecutive days; infusional schedules may be less toxic than bolus administration [37-40,42,43]. Ifosfamide-related toxicities differ from those associated with doxorubicin; they include hemorrhagic cystitis, renal tubular acidosis, salt-wasting nephropathy, and central nervous system toxicity.

Concurrent administration of the uroprotectant mesna decreases hemorrhagic cystitis. Mesna binds to the active metabolite acrolein, decreasing the incidence of hemorrhagic cystitis (a different metabolite, chloroacetaldehyde, is thought to be responsible for the nephrotoxicity). Mesna is usually administered immediately following, and four and eight hours after each ifosfamide dose. (See "Chemotherapy and radiation-related hemorrhagic cystitis in cancer patients" and "Ifosfamide nephrotoxicity".)

A dose-response relationship has been shown for ifosfamide in metastatic STS; the threshold is approximately 6 g/m² per cycle [11,39]. Additional responses can be demonstrated with ≥10 g/m² per cycle [44,45].

At least one trial directly compared single-agent doxorubicin (75 mg/m² every three weeks) with two different doses of ifosfamide (3 g/m² over four hours daily for three days or 9 g/m² over 72 hours by continuous infusion) in patients with advanced STS [46]. Antitumor efficacy was similar, and toxicity was worse in both ifosfamide groups.

Taxanes (angiosarcoma) — Taxanes as single agents are relatively inactive, except in angiosarcoma. Paclitaxel is particularly useful for advanced angiosarcoma [30,47-52]. Weekly therapy was more active than every-three-week therapy in one report [51].

Doxorubicin (including PLD) is also an active agent for angiosarcoma [53], and whether the antitumor efficacy of taxanes surpasses that of doxorubicin for this histology is unclear [54]. (See 'Pegylated liposomal doxorubicin' above.)

Gemcitabine and other agents — Other conventional cytotoxic drugs with modest antitumor activity include vinorelbine [55,56], dacarbazine, and temozolomide (particularly for leiomyosarcoma [LMS]) [57-60]. All are associated with response rates <20 percent, but additional patients have stable disease, some for prolonged periods.

There are conflicting data as to the extent of efficacy of single-agent gemcitabine [61-68]:

●In one report, gemcitabine administered by fixed-dose-rate infusion (mg/m²/min) was associated with a partial response in 4 of 10 patients with non-gastrointestinal LMS [62]. Gemcitabine is also an active agent in patients with angiosarcoma [68].

●Other studies suggest little activity for gemcitabine monotherapy, both in previously treated and untreated patients [64-66]. In most of these studies, the drug was administered over 30 minutes, but a preliminary report of a French trial noted an objective response rate of only 5 percent in the 20 patients treated with fixed-dose-rate gemcitabine; two-thirds had prolonged periods of stable disease [67].

●In contrast, combinations of gemcitabine (administered by dose rate infusion) with docetaxel, vinorelbine, or dacarbazine appear to be active in LMS of uterine and gastrointestinal origin, as well as in other histologies. (See 'Gemcitabine-based combinations' below.)

Combination chemotherapy — Many different combination regimens have been studied in patients with metastatic disease; most include doxorubicin and an alkylating agent. Several of these regimens are outlined in the table (table 1) (see "Treatment protocols for soft tissue and bone sarcoma"):

●Doxorubicin plus ifosfamide with mesna (AIM) [22,23,69,70], or AIM with dacarbazine (ie, mesna, doxorubicin, ifosfamide, and dacarbazine [MAID]) [71-73]

●Gemcitabine plus either docetaxel [74,75], vinorelbine [76], or dacarbazine [77]

●Doxorubicin plus dacarbazine [20,71,78,79]

These moderately intensive combination chemotherapy regimens are associated with overall response rates in the range of 16 to 46 percent, with complete responses in 5 to 10 percent. Approximately one-third of the complete responders (ie, 1 to 3 percent of patients with advanced STS overall) are long-term disease-free survivors [72,80-83].

Doxorubicin-based regimens — Several randomized studies and a pooled analysis have compared various doxorubicin-based combinations with each other and with single-agent doxorubicin [9,20,22,23,26,71,78-80,84-87].

In most randomized prospective trials, combination regimens delivered at conventional doses, such as MAID, AIM (table 2 and table 3 and table 4), and doxorubicin plus dacarbazine, are associated with higher response rates (range 18 to 46 percent) than are seen with single-agent doxorubicin (typically 12 to 18 percent). However, fewer than 10 percent are complete, and responses are generally short lived (median approximately eight months). Combination therapy also does not improve overall survival and is associated with greater toxicity than single-agent doxorubicin. For patients receiving combination therapy with AIM, the choice of specific doses and schedule is a function of patient age, comorbidities, and institution-specific standards. Similarly, the addition of olaratumab, a platelet-derived growth factor receptor alpha (PDGFRA) inhibitor, to single-agent doxorubicin did not improve survival and this agent was withdrawn from the market [84,88,89].

Gemcitabine-based combinations

Gemcitabine plus docetaxel — Activity for gemcitabine plus docetaxel (table 5) was initially reported in a study of 34 patients with LMS of uterine (n = 29) or gastrointestinal (n = 5) origin [74]; 53 percent had an objective response, including two of the five with gastrointestinal LMS. Contemporary data suggest that this regimen may have a broader range of activity, although response rates appear to be lower in histologies other than LMS [75,88,89]. As an example, a retrospective review of 133 patients treated with gemcitabine plus docetaxel included 76 LMS and 57 other histologies [88]. Only 17 percent were chemotherapy naive. The overall response rate was 18 percent (24 and 10 percent for LMS and other histologies, respectively). At 12 and 24 months, 51 and 15 percent of patients were still alive, respectively, although this may represent, at least in part, differences in tumor biology between cohorts rather than differences in treatment efficacy.

Lower doses (particularly of docetaxel) may be needed for patients with prior radiation to fields encompassing large amounts of marrow.

Gemcitabine plus docetaxel versus doxorubicin — Whether initial therapy with a gemcitabine-based regimen is preferable to anthracycline-based therapy for patients with a good performance status and any histologic type of STS (including LMS) is uncertain. However, given the report of data from the randomized GeDDis trial (which showed no significant difference in PFS or overall survival, but less toxicity, with single-agent doxorubicin [18]), a gemcitabine-based combination is typically not employed in the first-line setting for anthracycline-sensitive histologies. However, some clinicians still prefer gemcitabine plus docetaxel over doxorubicin-based therapy, particularly for metastatic uterine LMS. (See "Treatment and prognosis of uterine leiomyosarcoma", section on 'Docetaxel and gemcitabine'.)

The GeDDis trial compared combined gemcitabine plus docetaxel with doxorubicin alone in 257 previously untreated patients with a variety of histologies of advanced/metastatic STS [18]. Doxorubicin was less toxic and easier to administer than gemcitabine plus docetaxel, and the proportion of patients who were progression free at 24 weeks (the primary endpoint) was identical (46 percent in both groups); median overall survival slightly favored doxorubicin (76 versus 67 weeks, HR for death 1.14, 95% CI 0.83-1.57).

Results were not stratified according to histologic subtype. It should be noted that undifferentiated pleomorphic sarcoma/malignant fibrous histiocytoma was underrepresented in this trial. In our experience, undifferentiated pleomorphic sarcoma/malignant fibrous histiocytoma is the one histologic subtype that responds better to gemcitabine plus docetaxel, and we would choose this combination for second-line therapy after failure of an initial doxorubicin-based regimen in this group. (See 'Treatment at progression' above.)

It is also important to note that the doses of gemcitabine plus docetaxel used in the GeDDis trial were lower than the standard doses used in the United States. This has been a notable criticism of the results of this trial, particularly as applied to patients with LMS.

Gemcitabine plus docetaxel versus gemcitabine — As noted above, gemcitabine alone is an active single agent in metastatic STS when administered as a fixed-rate infusion. The superiority of gemcitabine plus docetaxel over gemcitabine alone was addressed in an open-label multicenter phase II trial comparing gemcitabine with and without docetaxel that used a novel Bayesian adaptive randomization strategy [90]. This allowed for continuous outcomes monitoring, incrementally assigning more patients to the superior treatment arm while accounting for possible treatment-subgroup interactions (eg, histology, prior RT versus none, performance status [91]).

The treatment arms were gemcitabine alone (1200 mg/m² on days 1 and 8; 900 mg/m² in patients with prior pelvic RT) or gemcitabine (900 mg/m² on days 1 and 8; 675 mg/m² if prior pelvic RT) plus docetaxel (100 mg/m² on day 8; 75 mg/m² if prior pelvic RT), with courses repeated every 21 days. In both arms, gemcitabine was given via fixed-dose-rate infusion (10 mg/m²/min). All patients received growth factor support.

Forty-nine patients were adaptively randomized to gemcitabine alone, and 73 were adaptively randomized to gemcitabine plus docetaxel. The primary endpoint (a composite endpoint of tumor response [complete plus objective responses plus stable disease] for at least 24 weeks) was met by 27 and 32 percent of patients receiving gemcitabine alone and combined therapy, respectively. The objective response rate with combined therapy was higher (16 versus 8 percent), as was PFS (6.2 versus 3 months) and overall survival (17.9 versus 11.5 months). Patients with LMS and undifferentiated/unclassified STS (previously included in the broad category of "malignant fibrous histiocytoma," a subset of which is the undifferentiated pleomorphic sarcoma variant [92]) appeared to derive the most benefit. However, these benefits came at the cost of greater toxicity (particularly edema and constitutional symptoms). Significantly more patients receiving combined therapy discontinued treatment due to toxicity. (See "Clinical presentation, histopathology, diagnostic evaluation, and staging of soft tissue sarcoma", section on 'Histopathology'.)

On the other hand, the benefit of combined therapy over gemcitabine alone, particularly for nonuterine LMS, was called into question by the results of a randomized phase II trial in which 90 patients with metastatic LMS of uterine (n = 46) or nonuterine (n = 44) origin were randomly assigned to gemcitabine alone (1000 mg/m² on days 1, 8, and 15 of each 28-day cycle) or gemcitabine plus docetaxel (100 mg/m² on day 8 of a 21-day cycle), with primary prophylaxis with a granulocyte colony-stimulating factor (G-CSF) [93]. Among patients with uterine LMS, the response rate was only modestly higher with combined therapy (24 versus 19 percent), and median PFS was not better (4.7 versus 5.5 months). Among those with nonuterine LMS, the response rate to combined therapy was low (5 versus 14 percent with gemcitabine alone), and median PFS was only 3.8 months (versus 6.3 months for gemcitabine monotherapy). Patients receiving single-agent gemcitabine experienced less toxicity.

Although the authors concluded that both treatments were similarly efficacious, at least for uterine LMS, the trial was conducted as a parallel set of two phase II studies, and the study lacked power to determine which treatment was better [94]. An important finding was that a high proportion of patients in both arms (>40 percent, highest in the nonuterine LMS group) had durable stable disease, and the six-month PFS rate in all groups was ≥47 percent.

Dose intensification — In view of the demonstrated dose-response relationship with doxorubicin and ifosfamide, higher-than-standard-dose chemotherapy has been studied in STS. Higher doses of doxorubicin and ifosfamide can be administered in combination regimens with the support of hematopoietic growth factors [69,85,95-100] or autologous hematopoietic stem cells [101-105]; however, there is no clear overall survival benefit for higher as compared with standard dosing regimens [106].

These approaches must be considered investigational for metastatic STS and performed only in the context of innovative exploratory new trials or appropriately controlled phase III clinical trials.

NEWER TREATMENT STRATEGIES — Given the limited efficacy of conventional cytotoxic chemotherapy, STS remains fertile ground for the field of drug development. Clinical trials in a number of areas have shown promise in metastatic STS.

Performing valid clinical trials in STS is challenging because the different histologic types can behave differently; as a result, clinical trials should be stratified by histologic subtype for adequate interpretation of results. In order to generate studies of sufficient size and power, large-scale collaborations are required on a national and international level. Such collaborations are already in place in the United States and among the nations of Europe (in Scandinavia, Italy, and Canada). With these collaborations, it is hoped that further research will rapidly translate research findings into the novel therapeutics that are so desperately required by patients with sarcoma.

Treatments approved in the United States, Europe, or Canada

Imatinib — The most dramatic example of translation of molecular understanding of STS into a novel therapy is with the use of the selective tyrosine kinase inhibitor (TKI) imatinib to treat gastrointestinal stromal tumors (GISTs). Unfortunately, this agent is of limited utility for treatment of non-GIST STS, with the exception of dermatofibrosarcoma protuberans (DFSP) and tenosynovial giant cell tumor (TGCT). The use of imatinib for these specific histologic soft tissue tumor types is discussed in detail separately. (See "Tyrosine kinase inhibitor therapy for advanced gastrointestinal stromal tumors" and "Treatment for tenosynovial giant cell tumor and other benign neoplasms affecting soft tissue and bone", section on 'Tenosynovial giant cell tumor' and "Dermatofibrosarcoma protuberans: Epidemiology, pathogenesis, clinical presentation, diagnosis, and staging".)

Trabectedin — Trabectedin (ecteinascidin 743) is a US Food and Drug Administration (FDA)-approved option for patients with advanced leiomyosarcoma (LMS) and liposarcoma previously treated with an anthracycline-containing chemotherapy regimen [107]. It is particularly effective in those with myxoid/round cell liposarcoma. Trabectedin is also approved by the European Medicines Agency (EMA) for patients with advanced STS experiencing disease progression on doxorubicin and ifosfamide [108,109]. Dosing and administration considerations are discussed below. (See 'Dosing and administration' below.)

In a phase III trial (ET743-SAR-3007), 577 patients with advanced LMS or liposarcoma previously treated with anthracycline-based chemotherapy were randomly assigned to trabectedin versus dacarbazine [110,111]. Approximately three-fourths of those enrolled had LMS (423 patients), and the remaining one-third had liposarcomas (154 patients). After a median follow-up of approximately 21 months, relative to dacarbazine, trabectedin demonstrated improved progression-free survival (PFS) but similar overall survival (median PFS 4.2 versus 1.5 months, hazard ratio [HR] 0.55, 95% CI 0.44-0.70; median overall survival 13.7 versus 13.1 months, HR 0.93, 95% CI 0.75-1.15). The similar overall survival results may have been confounded by the use of post-study antineoplastic therapy in both treatment arms. Grade ≥3 toxicities for trabectedin included neutropenia, thrombocytopenia, anemia, and transient elevations in aminotransferases. Nine treatment-related deaths were reported due to infection, rhabdomyolysis, and renal failure.

The efficacy of trabectedin in LMS, liposarcoma, and translocation-related sarcomas is discussed elsewhere. (See 'Leiomyosarcoma, liposarcoma, and translocation-related sarcomas' below.)

●Mechanism of action – Trabectedin kills cells by poisoning the DNA nucleotide excision repair machinery [112]. Trabectedin was originally isolated from the Caribbean sea sponge Ecteinascidia turbinata, but it is now synthetically produced. Trabectedin represents a unique form of molecularly targeted therapy, particularly in myxoid liposarcomas, due to the presence of a unique fusion oncoprotein that results from chromosomal translocation in this histologic subtype. Trabectedin interferes with the ability of the fusion protein to bind to DNA promoter regions, resulting in antineoplastic effects [113]. (See "Pathogenetic factors in soft tissue and bone sarcomas", section on 'Myxoid liposarcomas'.)

Trabectedin is an active agent for advanced STS, although the objective response rate by conventional criteria is fairly low [114-118]. Experience with trabectedin has underscored the relevance of stable disease as a beneficial endpoint for metastatic STS in general [108,114-117,119].

Dosing and administration — Trabectedin is administered at 1.5 mg/m² intravenously over 24 hours on day 1 of a 21-day cycle. Dexamethasone pretreatment (20 mg intravenously 30 minutes prior to each dose) is recommended to reduce hepatotoxicity risk [120,121]. Indications for therapy are discussed above. (See 'Trabectedin' above.)

Trabectedin carries a warning for risk of severe and fatal neutropenic sepsis; rhabdomyolysis and hepatotoxicity; skin and soft tissue necrosis following extravasation [122]; and heart failure. (See "Cardiotoxicity of cancer chemotherapy agents other than anthracyclines, HER2-targeted agents, and fluoropyrimidines", section on 'Trabectedin' and "Chemotherapy hepatotoxicity and dose modification in patients with liver disease: Conventional cytotoxic agents", section on 'Trabectedin'.)

The superiority of every-three-week dosing over weekly dosing of trabectedin was established in a randomized phase II trial of patients with previously treated advanced liposarcoma or LMS. In this study, median time to tumor progression was improved with administration of trabectedin at 1.5 mg/m² intravenously over 24 hours on day 1 of a 21-day cycle versus 0.58 mg/m² intravenously over three hours on days 1, 8, and 15 of a 28-day cycle (3.7 versus 2.3 months, HR 0.73, 95% CI 0.55-0.97) [108].

Leiomyosarcoma, liposarcoma, and translocation-related sarcomas — The specific efficacies of trabectedin for LMS, liposarcoma, and translocation-related sarcomas come from a randomized phase III trial (ET743-SAR-3007) as well as phase II studies in both treatment-naïve and chemotherapy-refractory patients [115-117,123-125]. (See 'Trabectedin' above and 'Dosing and administration' above.)

●Leiomyosarcoma – In the phase III trial (ET743-SAR-3007), among those with pretreated LMS, trabectedin resulted in the following, relative to dacarbazine [110,111]:

•Similar overall survival but improved PFS (median overall survival 14.1 versus 13.6 months, HR 0.89, 95% CI 0.69-1.15; median PFS 4.3 versus 1.6 months, HR 0.56, 95% CI 0.42-0.73).

•Improved PFS in both nonuterine and uterine disease (nonuterine disease, median 4.9 versus 1.6 months, HR 0.58, 95% CI 0.37-0.92; uterine disease, 4.0 versus 1.5 months, HR 0.58, 95% CI 0.41-0.81).

•Improved clinical benefit rates (ie, objective disease response plus stable disease rate; 37 versus 20 percent).

For patients with LMS, we do not routinely offer trabectedin in combination with doxorubicin. Although an initial observational study in patients with treatment-naïve LMS receiving trabectedin plus doxorubicin suggested high rates of objective response and disease control with a tolerable toxicity profile, a subsequent randomized phase II trial failed to confirm a survival benefit or tolerable toxicity profile for the addition of trabectedin to doxorubicin in 115 patients with advanced STS of various histologic subtypes [126,127]. In a subgroup analysis of the latter study, relative to other histologic subtypes, patients with LMS (approximately one-third of the study population) receiving trabectedin had higher median PFS and overall survival (7 versus 3.9 months and 24.2 versus 10.3 months, respectively) [127]. Larger randomized trials in patients with LMS are needed before the combination of trabectedin plus doxorubicin can be considered a standard chemotherapy regimen.

●Liposarcoma – Trabectedin has particular clinical efficacy in liposarcoma, particularly those of myxoid/round cell histology [110,111,128,129].

In the phase III trial (ET743-SAR-3007), among those with pretreated liposarcoma, trabectedin resulted in the following, relative to dacarbazine [110,111]:

•Similar overall survival but improved PFS (median overall survival 13.1 versus 12.6 months, HR 1.05, 95% CI 0.69-1.60; median PFS 3.0 versus 1.5 months, HR 0.55, 95% CI 0.34-0.87).

•Improved PFS in those with myxoid/round cell histology (median 5.6 versus 1.5 months, HR 0.41, 95% CI 0.17-0.98) but not in those with dedifferentiated or pleomorphic histology (median 2.2 versus 1.9 months, HR 0.68, 95% CI 0.37-1.25, and 1.5 versus 1.4 months, HR 0.33, 95% CI 0.07-1.64, respectively).

•Clinical benefit rates were numerically higher but not statistically significant (28 versus 15 percent).

●Translocation-related sarcomas – Trabectedin has demonstrated activity in translocation-related sarcomas, including myxoid/round cell liposarcomas that have translocations [124,130]. (See "Pathogenetic factors in soft tissue and bone sarcomas", section on 'Chromosomal translocations'.)

In an open-label phase II trial, 76 patients with advanced, translocation-related, predominantly pretreated sarcoma (including myxoid/round cell liposarcoma and synovial sarcoma) were randomized to trabectedin (1.2 mg/m² over 24 hours on day 1 every 21 days) versus best supportive care [124]. After a median follow-up of approximately nine months, trabectedin improved PFS (5.6 versus 0.9 months, HR 0.07, 95% CI 0.03-0.16). Common grade ≥3 toxicities for trabectedin included nausea (9 percent), decreased appetite (8 percent), febrile neutropenia (11 percent), and increased alanine aminotransferase (47 percent). [115-117,123,124].

Other histologies — We do not use trabectedin for histotypes other than liposarcoma, LMS, or translocation-related sarcomas. Efficacy in histotypes other than LMS, liposarcoma, and translocation-associated sarcomas was addressed in the phase III T-SAR trial [131]. In this study, 103 patients with advanced STS unresponsive to or intolerant of previous chemotherapy were randomly assigned to trabectedin or best supportive care. Sixty percent of tumors were liposarcoma or LMS, and the remainder were other histotypes. In those with non-liposarcoma/LMS histotypes, trabectedin had no objective tumor responses relative to the liposarcoma/LMS group (0 versus 19 percent) and had similar PFS to those receiving best supportive care (median 1.8 versus 1.5 months). In contrast, for those with liposarcoma/LMS, trabectedin demonstrated improved PFS relative to best supportive care (median 5.1 versus 1.4 months).

Pazopanib — Pazopanib is a multitargeted, orally active small molecule inhibitor of several tyrosine kinases, including vascular endothelial growth factor receptor (VEGFR) and platelet-derived growth factor receptor alpha (PDGFRA) and beta (PDGFRB). We typically offer pazopanib as subsequent therapy to patients with advanced STS (other than adipocytic tumors such as liposarcoma, or GIST) who progress on prior chemotherapy or other approved agents, as it has regulatory approval in this setting. (See 'Subsequent therapy' below.)

Pazopanib is one available initial treatment option in older adults with select anthracycline-sensitive histologies who cannot tolerate or wish to avoid the potential toxicities of doxorubicin, although data are limited to only one randomized phase II trial [132] and other alternative treatment options are also available. (See 'Initial therapy (anthracycline-sensitive histologies)' below.)

Initial therapy (anthracycline-sensitive histologies) — In most older adults who are willing and able to tolerate anthracycline-based chemotherapy, doxorubicin remains our preferred initial therapy. However, pazopanib is one available initial treatment option for older adults (age ≥60 years) with select treatment-naïve, anthracycline-sensitive histologies (eg, LMS, undifferentiated pleomorphic sarcoma, and angiosarcoma) who cannot tolerate or wish to avoid the potential toxicities of doxorubicin [132,133]. (See 'Doxorubicin' above.)

Based on data from a single randomized phase II trial of older adults with advanced STS, initial therapy with pazopanib demonstrated noninferior PFS and less myelotoxicity compared with doxorubicin [132]. However, this is the only randomized trial to evaluate pazopanib in this setting [132], and other well-tolerated agents with more supporting evidence are also available in this patient population, such as pegylated liposomal doxorubicin (PLD) for all anthracycline-sensitive histologies and paclitaxel or gemcitabine for those with angiosarcomas. (See 'Doxorubicin' above and 'Taxanes (angiosarcoma)' above and 'Gemcitabine and other agents' above.)

We do not offer pazopanib to patients with liposarcoma as initial therapy, as several phase II trials suggest no benefit for pazopanib in these patients [132,134].

Given the limited available data, clinicians who offer pazopanib as initial therapy in this setting should provide a risk-benefit discussion about the unique toxicity profiles of each agent. For example, compared with doxorubicin, pazopanib has lower rates of myelosuppression, alopecia, and mucosal inflammation, but higher rates of common toxicities seen with antiangiogenic agents (eg, hypertension, hypothyroidism, and diarrhea). Pazopanib may be offered initially at a lower dose (400 to 600 mg daily) and uptitrated as tolerated in order to reduce such toxicities in older patients. (See "Toxicity of molecularly targeted antiangiogenic agents: Cardiovascular effects" and "Toxicity of molecularly targeted antiangiogenic agents: Non-cardiovascular effects".)

In an open-label phase II trial (EPAZ), 120 older adults (age ≥60 years) with advanced, unresectable STS (including those with LMS, undifferentiated pleomorphic sarcoma, liposarcoma, and angiosarcomas) without previous exposure to systemic therapy were randomly assigned to either pazopanib or doxorubicin for up to six cycles [132]. A majority of patients were fit without significant comorbidities; approximately one-third had impairments in activities of daily living, and importantly only 16 percent had two or more comorbidities. Compared with doxorubicin, pazopanib demonstrated noninferior PFS (median 4.4 versus 5.3 months, HR 1.00, 95% CI 0.65-1.53) and similar overall survival (OS; median 12 versus 14 months, HR 1.08, 95% CI 0.68-1.72). Objective response rates were similar between the two agents (12 versus 15 percent). In subgroup analyses, although data are limited by small numbers of patients, pazopanib was more effective than doxorubicin among those younger than age 71, but less effective among those with liposarcoma and those with Eastern Cooperative Oncology Group (ECOG) performance status (table 6) of 2.

Overall toxicity rates were similar between the two groups (91 versus 94 percent) and were consistent with the unique toxicity profiles of each agent. For example, compared with doxorubicin, rates of myelosuppression were lower with pazopanib, including grade 4 neutropenia (0 versus 56 percent) and febrile neutropenia (0 versus 10 percent). Pazopanib also demonstrated lower rates of alopecia (3 versus 54 percent), stomatitis (4 versus 19 percent), and mucosal inflammation (12 versus 24 percent), but higher rates of hypertension (37 versus 8 percent), hypothyroidism (14 versus 0 percent), diarrhea (43 versus 14 percent), and general deterioration (14 versus 3 percent).

Subsequent therapy

●Pazopanib – We typically offer single-agent pazopanib as subsequent therapy to patients who progress on other approved lines of therapy, as it has regulatory approval in this setting.

Pazopanib has activity for some histologic subtypes that are not anthracycline sensitive (eg, solitary fibrous tumor [SFT]/hemangiopericytoma, alveolar soft part sarcoma [135], and extraskeletal myxoid chondrosarcoma [136]). (See "Solitary fibrous tumor", section on 'Pazopanib' and "Uncommon sarcoma subtypes", section on 'Alveolar soft part sarcoma' and "Uncommon sarcoma subtypes", section on 'Extraskeletal myxoid chondrosarcoma'.)

Single-agent pazopanib showed activity in phase II trials that included various STS subtypes [134,137]. Pazopanib met the primary endpoint for activity in three of the four histology-specific cohorts (LMS, synovial sarcoma, and other eligible STS types). However, pazopanib demonstrated low response rates in liposarcoma, which was also confirmed in a separate phase II trial [138].

Based on those results, a worldwide, randomized, double-blinded, phase III study (the PALETTE trial) was designed by the European Organisation for Research and Treatment of Cancer (EORTC) and other investigators to compare pazopanib (800 mg daily) with placebo in 369 patients with a variety of histologic subtypes (LMS, fibrosarcoma, synovial sarcoma, malignant peripheral nerve sheath tumor, vascular STS, and sarcoma not otherwise specified, but not adipocytic sarcomas or GIST) whose disease had progressed during or after first-line chemotherapy (including an anthracycline) [139]. Median PFS was significantly higher in the pazopanib group (4.6 versus 1.6 months), and benefit was consistent across all histologic subtypes. There was no significant difference in overall survival (12.5 versus 10.7 months, HR 0.86, 95% CI 0.67-1.1). The best overall response was partial response in 6 versus 0 percent of the pazopanib and placebo groups, respectively, and stable disease in 67 versus 38 percent.

The most common grade 3 or 4 treatment-related toxicities were fatigue, hypertension, diarrhea, anorexia, and transient elevations in liver function tests. A drop in left ventricular ejection fraction occurred in 16 patients treated with pazopanib, compared with three cases in the placebo group (6.5 versus 2.4 percent); only three cases were symptomatic. Venous thromboembolism was more common in the pazopanib group (5 versus 2 percent, all grades). In addition, pneumothorax occurred in eight patients in the pazopanib group (3 percent), possibly due to necrosis of pleural lesions. This toxicity profile did not translate into a significantly worse global health status during treatment [140]. Treatment-related toxicity in patients treated with TKIs that target vascular endothelial growth factor (VEGF) is discussed in more detail separately. (See "Toxicity of molecularly targeted antiangiogenic agents: Cardiovascular effects" and "Toxicity of molecularly targeted antiangiogenic agents: Non-cardiovascular effects".)

Based on these data, pazopanib was approved in the United States for treatment of patients with advanced STS (but not adipocytic tumors or GIST) who had received prior chemotherapy. It was also approved by the EMA.

The efficacy of pazopanib in advanced GIST is addressed separately. (See "Tyrosine kinase inhibitor therapy for advanced gastrointestinal stromal tumors", section on 'Alternative targeted agents'.)

●Pazopanib plus gemcitabine – The addition of gemcitabine to pazopanib improved progression-free survival rates in a randomized phase II trial of patients with soft tissue sarcoma (predominantly leiomyosarcoma and liposarcoma) who had progressed on prior anthracycline and/or ifosfamide-based therapy [141]. However, further data are necessary before incorporating this combination into routine clinical practice.

Toxicity and administration — Given the risk for potentially fatal hepatotoxicity, close monitoring of liver function tests is recommended, particularly in the first nine weeks of therapy with pazopanib. (See "Chemotherapy hepatotoxicity and dose modification in patients with liver disease: Molecularly targeted agents", section on 'Pazopanib'.)

The solubility of pazopanib is pH dependent, and an elevated gastric pH (such as might occur with concomitant use of gastric acid suppressive agents) may decrease bioavailability [142,143]. The FDA-approved prescribing information recommends against concomitant use of proton pump inhibitors or histamine H2 receptor antagonists in patients receiving pazopanib [144]. Short-acting antacids could be offered in place of these agents, but dosing of antacids and pazopanib should be separated by several hours.

Eribulin — Eribulin is an approved option for treatment of unresectable or metastatic liposarcoma in patients who received prior anthracycline-containing chemotherapy in the United States, Canada, and Europe and for LMS and liposarcoma in Japan and the Philippines.

Eribulin inhibits microtubules via a mechanism that is distinct from other microtubule-targeting agents, such as taxanes. An initial phase II trial suggested that eribulin was active in a variety of sarcoma histotypes, including LMS, adipocytic sarcoma, synovial sarcoma, and others [145].

The specific efficacy of eribulin over dacarbazine in advanced liposarcoma and LMS was subsequently confirmed in a randomized phase III trial, with greatest benefits for those with liposarcoma [146]. In this trial, 452 patients with advanced LMS or liposarcoma previously treated with an anthracycline were randomly assigned to eribulin (1.4 mg/m² intravenously on days 1 and 8) or dacarbazine (between 850 and 1200 mg/m² intravenously on day 1) every 21 days [146-148]. Approximately two-thirds of those enrolled had LMS (309 patients), and the remaining one-third had liposarcomas (143 patients). Over one-half of those with LMS had nonuterine disease (57 percent).

At a median follow-up of 31 months, relative to dacarbazine, eribulin demonstrated the following:

●Improved overall survival in the entire study population (median 13.5 versus 11.5 months, HR 0.77, 95% CI 0.62-0.95), particularly in those with liposarcoma (15.6 versus 8.4 months, HR 0.51, 95% CI 0.35-0.75) [146,147].

●Similar overall survival in those with LMS (median 12.7 versus 13.0 months, HR 0.93, 95% CI 0.71-1.2) [148].

●Similar PFS in the entire study population (median 2.6 months in both arms) and in those with LMS (2.2 versus 2.6 months, HR 1.07, 95% CI 0.84-1.38).

●Modestly improved PFS in those with liposarcoma (median 2.9 versus 1.7 months, HR 0.52, 95% CI 0.35-0.78).

However, eribulin increased treatment-related toxicity relative to chemotherapy in all patients (67 versus 56 percent), including higher rates of neutropenia (43 versus 24 percent), pyrexia (28 versus 14 percent), peripheral sensory neuropathy (21 versus 4 percent), and alopecia (35 versus 3 percent) [146,148].

The use of eribulin in advanced or metastatic uterine LMS is discussed separately. (See "Treatment and prognosis of uterine leiomyosarcoma", section on 'Eribulin'.)

Nab-sirolimus — For patients with advanced unresectable or metastatic malignant neoplasm with perivascular epithelioid cell differentiation (PEComa), we suggest initial therapy with nab-sirolimus over other systemic agents. For patients living in areas where nab-sirolimus is not approved or unavailable, other mTOR inhibitors (sirolimus, everolimus, and temsirolimus) are reasonable alternatives.

The management (including sirolimus and supportive care measures) of other diseases in the PEComa family of tumors, such as lymphangioleiomyomatosis (LAM) and renal angiomyolipoma, are discussed separately. (See "Sporadic lymphangioleiomyomatosis: Treatment and prognosis" and "Renal angiomyolipomas (AMLs): Management".)

Malignant PEComa is a rare epithelioid malignancy that typically arises in the gastrointestinal tract, retroperitoneum, uterus, or somatic soft tissues and intimately related to blood vessel walls. These tumors may exhibit an aggressive clinical course with local recurrences and distant metastases, most commonly in the lung, and respond poorly to systemic chemotherapy [149]. Malignant PEComas belong to the PEComa family of tumors which also includes more benign tumors such as LAM and renal angiomyolipoma. This family of mesenchymal neoplasms shares certain pathologic features including myomelanocytic differentiation; involvement of the perivascular epithelioid cell [150-152]; and in many cases, inherited (in those with tuberous sclerosis complex) or sporadic mutations in the TSC1 or TSC2 genes which dysregulate the mechanistic (mammalian) target of rapamycin (mTOR) signaling pathway [153-158]. (See "Sporadic lymphangioleiomyomatosis: Epidemiology and pathogenesis" and "Renal angiomyolipomas (AMLs): Epidemiology, pathogenesis, clinical manifestations, and diagnosis" and "Tuberous sclerosis complex: Genetics, clinical features, and diagnosis".)

In an open-label phase II trial (AMPECT), nab-sirolimus, an mTOR inhibitor, was evaluated in 31 patients with locally advanced unresectable or metastatic malignant PEComa [159]. Nab-sirolimus was administered intravenously at 100 mg/m² on days 1 and 8 of a 21-day cycle until disease progression or unacceptable toxicity. In preliminary results, overall responses were seen in 12 patients (39 percent), including two complete responders [159]. Among the 12 patients with treatment response, a majority (58 percent) had durable responses lasting two years or more. Patients with a TSC2 mutation were more likely to respond to therapy, with objective responses seen in eight of nine patients with TSC2-positive tumors (89 percent).

Treatment was well tolerated overall. Grade 1 or 2 interstitial lung disease/noninfectious pneumonitis occurred in 18 percent of patients. Grade ≥3 toxicities included stomatitis (18 percent), infections (12 percent), dehydration (6 percent), diarrhea, vomiting, constipation, fatigue, edema, musculoskeletal pain, hypertension, hemorrhage, and insomnia (3 percent each).  

Based on these data, the US Food and Drug Administration (FDA) approved nab-sirolimus for adult patients with locally advanced unresectable or metastatic malignant PEComa [160]. Although nab-sirolimus is the first agent to be specifically approved for this disease, it may not be approved or available in all countries. In this situation, other mTOR inhibitors (sirolimus, everolimus, temsirolimus) are reasonable alternatives to nab-sirolimus, since they also have shown activity in malignant PEComa [153,161-163].

Tazemetostat — Tazemetostat, a small molecule inhibitor of enhancer of zeste homolog 2 (EZH2), has clinical activity in patients with epithelioid sarcomas. Further details on the use of tazemetostat in these patients are discussed separately. (See "Uncommon sarcoma subtypes", section on 'Tazemetostat'.)

Off-label treatments approved in other settings

Sunitinib — Sunitinib is an approved drug for GIST. However, sunitinib does not have regulatory approval for other non-GIST STS histologic subtypes, due to limited activity in these tumors [164,165]. (See "Tyrosine kinase inhibitor therapy for advanced gastrointestinal stromal tumors", section on 'Sunitinib'.)

Sunitinib has clinical activity in the following STS subtypes, which are discussed separately.

●Alveolar soft part sarcoma (see "Uncommon sarcoma subtypes", section on 'Alveolar soft part sarcoma')

●Clear cell sarcoma (see "Uncommon sarcoma subtypes", section on 'Clear cell sarcoma')

●Solitary fibrous tumor/hemangiopericytoma (see "Solitary fibrous tumor", section on 'Alternative options')

Sorafenib — Sorafenib is another multitargeted TKI that has limited activity in metastatic non-GIST STS, as evidenced by the following reports:

●In a phase II trial of 120 patients with six different histologic types of STS who received sorafenib 400 mg twice daily, there was one objective partial response among 37 LMS, one complete and four partial responses among 37 angiosarcomas (14 percent), and no objective responses in the malignant peripheral nerve sheath tumors, malignant fibrous histiocytomas (most of which are now reclassified as undifferentiated/unclassified STS [92]), synovial sarcomas, and other histotypes [166]. Treatment-related toxicity was not trivial, despite use of the FDA-approved dose and schedule. Over 60 percent of patients required dose reduction, the majority due to dermatologic toxicity. There were three grade 5 toxicities (one gastrointestinal hemorrhage, one intestinal perforation, and one fatal tension pneumothorax in a patient with pulmonary metastases).

●More modest activity against angiosarcoma was seen in another phase II trial of 51 patients with advanced vascular sarcomas, high-grade liposarcoma, and LMS [167]. Six of the eight patients with some form of "vascular sarcoma" had prolonged periods of stable disease in this nonrandomized trial, although no objective antitumor responses were observed. Median PFS in this group of vascular sarcomas was five months, compared with two to three months for the other sarcoma histologies. Despite the same dose, the toxicity profile was more favorable than that seen in the prior study; 50 percent required at least one dose reduction, but there were no grade 5 toxicities.

Regorafenib — Regorafenib is an orally active inhibitor of angiogenic (including VEGFR 1, 2, and 3), stromal, and oncogenic receptor tyrosine kinases. It is structurally similar to sorafenib and targets a variety of kinases implicated in angiogenic and tumor growth-promoting pathways.

Regorafenib activity varies by histologic subtype of STS. In our view, regorafenib and similar broad-spectrum VEGFR TKIs show promising activity in subsets of refractory nonadipocytic STS. As examples:

●Gastrointestinal stromal tumors – Regorafenib is approved for refractory GIST but not for non-GIST STS subtypes. (See "Tyrosine kinase inhibitor therapy for advanced gastrointestinal stromal tumors", section on 'Regorafenib'.)

●Extraskeletal osteosarcoma – Several early phase clinical trials have suggested some activity for regorafenib over placebo in refractory osteosarcoma, including extraskeletal cases. (See "Chemotherapy and radiation therapy in the management of osteosarcoma", section on 'Multitargeted kinase inhibitors'.)

●Liposarcoma – Regorafenib has limited activity in treatment-refractory advanced or metastatic liposarcoma, based on data from a phase II, randomized, placebo-controlled trial (SARC024) [168].

●Other histologies – Regorafenib is active in nonadipocytic STS. This was illustrated in a double-blind, randomized, placebo-controlled trial of 182 patients with previously treated liposarcoma, LMS, synovial sarcoma, and other types of STS [169,170]. Regorafenib improved PFS in all subgroups but liposarcoma. The most common adverse events, seen in more than one-third of the regorafenib-treated patients, were asthenia, diarrhea, mucositis, acral erythema, anorexia, and arterial hypertension. Importantly, only six patients had received prior pazopanib. Additional experience with regorafenib for STS subtypes (other than liposarcoma and GIST) after failure of pazopanib is needed.

Bevacizumab — Bevacizumab is a monoclonal antibody (MoAb) targeting VEGF. The combination of bevacizumab plus doxorubicin was evaluated in 17 anthracycline-naive patients with metastatic STS [171]. Although there were only two partial responses (lasting 21 to 36 weeks, both in patients with uterine LMS), 11 patients had stable disease for 12 weeks or more. Of concern, six patients developed grade 2 cardiac toxicity at cumulative doxorubicin doses of 75 to 300 mg/m².

Given that efficacy was not substantially different from that expected with doxorubicin alone and the worrisome cardiotoxicity, bevacizumab has been largely abandoned, except in highly vascular tumors, such as angiosarcoma, or in combination with temozolomide in malignant SFT. However, the available data are quite limited:

●In a phase II trial of bevacizumab in 32 patients with metastatic or locally advanced angiosarcoma or epithelioid hemangioendothelioma, four patients (two with angiosarcoma and two with epithelioid hemangioendothelioma) had a partial response (17 percent), while 50 percent had tumor stabilization, with a mean time to tumor progression of 26 weeks [172].

●The multicenter randomized phase II ANGIOTAX-PLUS trial evaluated weekly paclitaxel with or without bevacizumab in 50 patients with advanced primary or radiation-induced angiosarcoma, 16 of whom had received prior anthracycline chemotherapy [173]. The addition of bevacizumab did not lead to higher rates of being progression free at six months (57 versus 54 percent), longer median PFS, or longer overall survival (15.9 versus 19.5 months).

The use of chemotherapy without bevacizumab represents a standard option for angiosarcoma. Based on the ANGIOTAX-PLUS study, there is no role for combining paclitaxel with bevacizumab.

CDK4/6 inhibitors — Although a majority (>90 percent) of well-differentiated or dedifferentiated liposarcomas have amplification of cyclin-dependent kinase (CDK) 4, various early phase trials have suggested only a modest degree of benefit from selective CDK4/CDK6 inhibitors such as palbociclib [174], ribociclib [175,176], and abemaciclib [177]. The use of these agents remains investigational in those with liposarcoma.

Checkpoint inhibitor immunotherapy — For patients with advanced STS of any histologic type with high levels of microsatellite instability (MSI-H), we suggest the use of checkpoint inhibitor immunotherapy. (See 'Microsatellite instability and pembrolizumab' below.)

For other patients with STS, we do not consider immunotherapy to be a standard clinical therapy, and its use remains investigational. Several observational and phase II trials have suggested a potential benefit for these agents, such as pembrolizumab plus axitinib in patients with alveolar soft part sarcomas [178]; nivolumab with or without ipilimumab in patients with undifferentiated pleomorphic sarcoma [179,180], and pembrolizumab in those with myxofibrosarcoma, undifferentiated pleomorphic sarcoma, poorly differentiated or dedifferentiated liposarcoma, and angiosarcoma [181-183].

●Immunotherapy with a VEGF inhibitor – Pembrolizumab and the VEGF inhibitor axitinib have demonstrated increased activity in STS, with the most promising efficacy noted in alveolar soft part sarcomas. (See "Uncommon sarcoma subtypes", section on 'Alveolar soft part sarcoma'.)

●Nivolumab with or without ipilimumab – Nivolumab, a fully human anti-programmed cell death 1 (PD-1) MoAb, was administered with or without ipilimumab, a humanized anti-cytotoxic T lymphocyte-associated antigen 4 (CTLA-4) MoAb, in an open-label, randomized, phase II trial of 85 patients with advanced STS or bone sarcoma after failing prior regimens [180]. There was minimal activity with nivolumab alone (response rate 5.3 percent, median PFS 2.6 months), but there was encouraging antitumor activity with combined therapy (response rate 16 percent, median PFS 4.5 months). Treatment was relatively well tolerated, with 14 percent of patients receiving combined therapy experiencing grade 3 or 4 treatment-related adverse effects. In particular, a signal seen in undifferentiated pleomorphic sarcoma indicates that at least some forms of sarcoma are worth investigating further with combination PD-1/CTLA-4 inhibition.

●Pembrolizumab – Pembrolizumab, another anti-PD-1 MoAb, was administered in the SARC028 trial, an open-label phase II trial conducted with 10 patients in each of four cohorts: undifferentiated pleomorphic sarcoma, poorly differentiated/dedifferentiated liposarcoma, synovial sarcoma, and LMS [181]. The objective response rate in the entire cohort was 18 percent, and 12-week PFS was 55 percent. Clinical activity varied by histologic type, with a 40 percent (4 of 10) objective response rate in undifferentiated pleomorphic sarcoma, 20 percent in poorly differentiated/dedifferentiated liposarcoma, 10 percent in synovial sarcoma, and no responses in LMS. Expansion cohorts in undifferentiated pleomorphic sarcoma and liposarcoma show evidence of activity consistent with the original study [182].

In patients with angiosarcoma, observational data also suggest a potential unique sensitivity to pembrolizumab [183], although this approach remains investigational. (See "Head and neck sarcomas", section on 'Angiosarcoma'.)

Further information about the molecular mechanisms of action of immunotherapy is discussed separately. (See "Principles of cancer immunotherapy".)

Next-generation sequencing — In a small subset of patients with STS, next generation sequencing (NGS) may identify potentially actionable targets for treatment and, in some cases, confirm the diagnosis of specific sarcoma subtypes that express characteristic translocations. However, we do not routinely offer NGS for sarcoma subtypes that consistently express specific molecular alterations (eg, well-differentiated or differentiated liposarcomas).  

Several ongoing trials (eg, the National Cancer Institute [NCI] Molecular Analysis for Therapy Choice [MATCH] and the American Society of Clinical Oncology [ASCO] Targeted Agent and Profiling Utilization Registry [TAPUR] trials) are using next-generation sequencing of multiple genes (gene panel tests) to identify molecular abnormalities in the tumors of patients with refractory cancers that may potentially match molecularly targeted therapies that are either in clinical trials or approved for treatment of other cancer types. Two such gene panel tests (the Memorial Sloan-Kettering Cancer Center Integrated Mutation Profiling of Actionable Cancer Targets [MSK-IMPACT] and the FoundationOne CDx [F1CDx]) are FDA approved in the United States. When NGS is used, we prefer sequencing panels that capture both mutations and translocations, as the latter can provide evidence of a specific sarcoma subtype when the diagnosis is in question. (See "Next-generation DNA sequencing (NGS): Principles and clinical applications", section on 'Cancer screening and management'.)

The potential impact of next-generation sequencing in advanced STS and bone sarcoma was addressed in an analysis of 5635 patients with 56 different histologies [184]. In a preliminary report, treatment-linked alterations known to respond to FDA-approved or study drugs were identified in 16 and 7 percent of patients, respectively, and 42 percent had a treatment-linked alteration that was eligible for the MATCH trial, the TAPUR trial, or another trial. Of the 107 patients from Memorial Sloan-Kettering Cancer Center with clinical data available, 60 had at least one "treatment-linked" alteration, and 31 were enrolled in a matched trial; 26 were ineligible or lacked access to a trial.

These and other data on the frequency of potentially targetable alterations in adult patients with advanced STS need to be interpreted cautiously [185] as the precise frequency of actionable mutations leading to treatment changes that actually impact outcomes remains unknown but is very low in our experience.

This likelihood of a therapeutically useful outcome from genomic sequencing is under active investigation, but available data suggest that it varies widely by sarcoma histology. For example, the vast majority of GISTs are defined by alterations in KIT or PDGFRA, and in this case, mutation analysis is helpful to define the patients who will most likely benefit from imatinib (either in the adjuvant or metastatic setting). (See "Tyrosine kinase inhibitor therapy for advanced gastrointestinal stromal tumors".)

In contrast, well-differentiated/dedifferentiated liposarcomas nearly universally harbor mouse double minute 2 homolog (MDM2) and CDK4 amplification/overexpression, with few other molecular alterations observed, but this does not impact the choice of therapy. Next-generation sequencing is not generally indicated to prove MDM2 or CDK4 amplification, and any patient with well-differentiated/dedifferentiated liposarcomas should be encouraged to explore clinical trials of drugs targeting these pathways, irrespective of whether his or her tumor has been sequenced. Similarly, sporadic desmoid tumors are usually defined by a point mutation in the beta-catenin gene (CTNNB1), but treatment does not depend on having this sequencing result available. (See "Pathogenetic factors in soft tissue and bone sarcomas", section on 'MDM2 gene' and "Pathogenetic factors in soft tissue and bone sarcomas", section on 'CDK4 gene' and "Desmoid tumors: Systemic therapy".)

NTRK fusion genes — Fewer than 1 percent of unselected STS have gene fusions involving one of the neurotrophic tyrosine receptor kinase (NTRK) genes [186-189]. Infantile fibrosarcoma, a very rare NTRK-translocation sarcoma nearly exclusively occurring in children under age 2, is one of the index tumors evaluated using these agents. Inflammatory myofibroblastic tumors also express NTRK gene fusions at lower frequencies [190]. NTRK gene fusions are otherwise very rare in other sarcoma subtypes. (See "TRK fusion-positive cancers and TRK inhibitor therapy" and "Uncommon sarcoma subtypes", section on 'Inflammatory myofibroblastic tumor'.)

Two highly selective NTRK inhibitors, larotrectinib and entrectinib, are approved by the FDA for use in adults and children with solid tumors with an NTRK gene fusion without a known acquired resistance mutation [191,192]. Patients who are candidates for NTRK inhibitors have metastatic disease, have disease that is unresectable or likely to result in severe postoperative morbidity with no satisfactory alternative treatments or have progressive disease following initial treatment. Confirmation of an NTRK molecular alteration is critical prior to the use of larotrectinib or entrectinib.

Between these two agents, we typically prefer larotrectinib because it has more durable responses and a better toxicity profile than entrectinib (which is associated with cardiac toxicity and skeletal fractures). These issues are discussed in detail elsewhere. (See "TRK fusion-positive cancers and TRK inhibitor therapy", section on 'Side effects'.)

Dosing recommendations and toxicity for larotrectinib and entrectinib are discussed separately. (See "TRK fusion-positive cancers and TRK inhibitor therapy", section on 'Dosing' and "TRK fusion-positive cancers and TRK inhibitor therapy", section on 'Side effects'.)

●Larotrectinib – The efficacy of larotrectinib was shown in a combined analysis of three phase I and II trials of 159 adult and pediatric patients with various NTRK fusion-positive malignancies; among these patients, 69 had STS [189,193]. Objective responses were seen in 27 of 28 patients with infantile fibrosarcoma (96 percent), including three patients with complete responses; all four patients with gastrointestinal stromal tumors (100 percent); and 29 of 31 patients with other STS histologies (81 percent) [189]. The longest duration of response was seen in one patient with sarcoma (44 months). Although patients were eligible for treatment based on the presence of an NTRK gene fusion, a limitation of this study was that there was no central histology review of this patient population.

●Entrectinib – The efficacy of entrectinib was demonstrated in a pooled analysis of three multicenter, single-arm, open-label trials (ALKA, STARTRK-1, and STARTRK-2) conducted in 54 patients with relapsed advanced or metastatic solid tumors harboring an NTRK gene fusion [192,194]. Among the 13 patients with sarcoma, the overall response rate was 46 percent, and the duration of response ranged between 3 and 15 months.

Further data on the efficacy of larotrectinib and entrectinib in tumors other than STS and their associated toxicities are discussed separately. (See "TRK fusion-positive cancers and TRK inhibitor therapy", section on 'Treatment with TRK inhibitors'.)

Microsatellite instability and pembrolizumab — Between 1 and 5 percent of STS have MSI-H, which is the biologic footprint of deficient mismatch repair (dMMR) [195-198] and may predict benefit from immune checkpoint inhibitors that target PD-1 or programmed cell death ligand 1 (PD-L1). (See "Principles of cancer immunotherapy", section on 'PD-1 and PD ligand 1/2' and "Tissue-agnostic cancer therapy: DNA mismatch repair deficiency, tumor mutational burden, and response to immune checkpoint blockade in solid tumors".)

The US Food and Drug Administration (FDA) approved pembrolizumab, a PD-1 inhibitor, for treatment of a variety of advanced solid tumors (including sarcomas) that are MSI-H or dMMR, that progressed following prior treatment, and for which there are no satisfactory alternative treatment options.

An important point is that MSI-H may indicate the presence of Lynch syndrome, an inherited condition that predisposes to several cancers, including possibly STS [199]. Given that Lynch syndrome is more prevalent than previously thought, all patients with an MSI-H/dMMR STS should be referred for germline genetic assessment for Lynch syndrome, regardless of family history [196]. (See "Lynch syndrome (hereditary nonpolyposis colorectal cancer): Clinical manifestations and diagnosis", section on 'Microsatellite instability testing'.)

SPECIAL CONSIDERATIONS DURING THE COVID-19 PANDEMIC — The COVID-19 pandemic has increased the complexity of cancer care. Important issues include balancing the risk from treatment delay versus harm from COVID-19, minimizing the use of immunosuppressive cancer treatments whenever possible, mitigating the negative impacts of social distancing during care delivery, and appropriately and fairly allocating limited health care resources. These and other recommendations for cancer care during active phases of the COVID-19 pandemic are discussed separately. (See "COVID-19: Considerations in patients with cancer".)

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Soft tissue sarcoma".)

INFORMATION FOR PATIENTS — UpToDate offers two types of patient education materials, "The Basics" and "Beyond the Basics." The Basics patient education pieces are written in plain language, at the 5^th to 6^th grade reading level, and they answer the four or five key questions a patient might have about a given condition. These articles are best for patients who want a general overview and who prefer short, easy-to-read materials. Beyond the Basics patient education pieces are longer, more sophisticated, and more detailed. These articles are written at the 10^th to 12^th grade reading level and are best for patients who want in-depth information and are comfortable with some medical jargon.

Here are the patient education articles that are relevant to this topic. We encourage you to print or e-mail these topics to your patients. (You can also locate patient education articles on a variety of subjects by searching on "patient info" and the keyword(s) of interest.)

●Basics topics (see "Patient education: Soft tissue sarcoma (The Basics)")

SUMMARY AND RECOMMENDATIONS

●Histology-driven treatment – Soft tissue sarcomas (STS) are a heterogeneous group of rare tumors arising from mesenchymal cells at all body sites. Consensus is emerging that selection of treatment should be histology-driven, particularly in the setting of advanced disease. (See 'Histology-driven treatment' above.)

●Goals of therapy – For a majority of patients with metastatic STS, chemotherapy is administered with palliative intent, with the goals of decreasing tumor bulk, diminishing symptoms, improving quality of life, and prolonging survival. Objective response rate (as judged by a decrease in the size of the measurable lesions) is a poor surrogate for benefit in metastatic STS, particularly for molecularly targeted therapies and drugs such as trabectedin. (See 'Endpoints to define benefit' above.)

●Impact of natural history of STS on therapy – The natural history of unresectable metastatic disease is variable and is influenced by disease biology as well as treatment. For some patients with asymptomatic, low-grade, unresectable disease (eg, low-grade intraabdominal leiomyosarcoma [LMS]), it might be reasonable to follow the patient without active chemotherapy. Conversely, for patients with a high-grade chemotherapy-sensitive tumor, such as synovial sarcoma or liposarcoma, early use of combination chemotherapy may be preferable. (See 'Natural history of metastatic disease and implications for treatment' above.)

●Clinical trials versus conventional systemic therapy – Patients with advanced unresectable nonuterine STS and non-GIST (gastrointestinal stromal tumors) are appropriate candidates for clinical trials to identify more active treatment approaches. If participation in clinical trials is not feasible, conventional systemic therapy is an appropriate option for selected patients.

•The treatment of patients with advanced uterine LMS and GIST is discussed separately. (See "Treatment and prognosis of uterine leiomyosarcoma" and "Tyrosine kinase inhibitor therapy for advanced gastrointestinal stromal tumors".)

•The treatment of advanced dermatofibrosarcoma protuberans (DFSP), desmoid tumors, rhabdomyosarcoma, solitary fibrous tumor (SFT), Ewing sarcoma, tenosynovial giant cell tumor (TGCT), and various uncommon sarcoma subtypes is discussed separately. (See "Dermatofibrosarcoma protuberans: Treatment" and "Desmoid tumors: Systemic therapy" and "Rhabdomyosarcoma in childhood, adolescence, and adulthood: Treatment" and "Solitary fibrous tumor" and "Treatment of Ewing sarcoma" and "Treatment for tenosynovial giant cell tumor and other benign neoplasms affecting soft tissue and bone", section on 'Tenosynovial giant cell tumor' and "Uncommon sarcoma subtypes".)

●Choice of initial therapy – Patients should be assessed for potential tumor sensitivity to anthracyclines to determine whether they should be used as part of initial therapy or whether alternative agents should be offered. (See 'Initial therapy' above.)

●Anthracycline-sensitive histologies – For most patients with a good performance status, minimal comorbidity, and an STS histology that is known to have at least some sensitivity to anthracyclines (eg, LMS, myxoid/round cell liposarcoma, epithelioid sarcoma [ES], dedifferentiated and pleomorphic liposarcoma, synovial sarcoma, angiosarcoma, undifferentiated pleomorphic sarcoma, and malignant peripheral nerve sheath tumor), we suggest doxorubicin, with or without ifosfamide, rather than other therapies (Grade 2C). Combination therapy with doxorubicin plus ifosfamide may be preferred over doxorubicin alone for symptomatic patients who are in need of the most rapid possible tumor response. (See 'Doxorubicin-based regimens' above and 'Gemcitabine plus docetaxel versus doxorubicin' above.)

•A gemcitabine-based combination could be offered as an alternative in patients with a relative contraindication to anthracyclines (eg, clinical heart failure, prior treatment with >450 mg/m² doxorubicin in the adjuvant setting).

•Paclitaxel is a reasonable alternative to anthracyclines for initial treatment of patients with angiosarcoma. (See 'Gemcitabine-based combinations' above and 'Taxanes (angiosarcoma)' above.)

•Tazemetostat is an alternative to anthracyclines for initial treatment of patients with ES and indolent disease, and it has regulatory approval in this setting. (See "Uncommon sarcoma subtypes", section on 'Tazemetostat'.)

●Older adults, declined, performance status and comorbidities – For patients with a poorer performance status or extensive comorbidity who are eligible for chemotherapy, alternative chemotherapy options include pegylated liposomal doxorubicin (PLD), gemcitabine alone, or a gemcitabine-based combination. (See 'Pegylated liposomal doxorubicin' above and 'Gemcitabine and other agents' above and 'Gemcitabine-based combinations' above.)

•For older adults (age ≥60 years) with select anthracycline-sensitive histologies (eg, LMS, undifferentiated pleomorphic sarcoma, and angiosarcoma) who cannot tolerate or wish to avoid the potential toxicities of doxorubicin, the antiangiogenic agent pazopanib is an appropriate alternative option for initial therapy. However, other agents are also available in this patient population such as PLD for all anthracycline-sensitive histologies, and paclitaxel or gemcitabine for those with angiosarcoma. (See 'Pazopanib' above.)

●Anthracycline-resistant histologies – For patients with advanced or metastatic STS histologies that are not sensitive to anthracyclines, other options may be offered as initial therapy (see 'Initial therapy' above):

•For patients with alveolar soft part sarcoma or extraskeletal myxoid chondrosarcoma with symptomatic or progressive disease, options include pazopanib or sunitinib if a clinical trial option does not exist. (See 'Pazopanib' above and 'Sunitinib' above and "Uncommon sarcoma subtypes", section on 'Alveolar soft part sarcoma' and "Uncommon sarcoma subtypes", section on 'Extraskeletal myxoid chondrosarcoma'.)

•For patients with SFT, active agents include dacarbazine with or without doxorubicin, temozolomide plus bevacizumab, pazopanib, or sunitinib; the choice between these agents is discussed separately. (See "Solitary fibrous tumor", section on 'Advanced and metastatic disease'.)

•For patients with malignant neoplasm with perivascular epithelioid cell differentiation (PEComa), we suggest initial therapy with nab-sirolimus rather than other systemic agents (Grade 2C). For those without access to nab-sirolimus, other mTOR inhibitors (sirolimus, everolimus, temsirolimus) are reasonable alternatives. (See 'Nab-sirolimus' above and 'Histology-driven treatment' above.)

•For patients with TGCT with relapsed, recurrent, or unresectable disease, options include pexidartinib or clinical trials. (See "Treatment for tenosynovial giant cell tumor and other benign neoplasms affecting soft tissue and bone", section on 'Pexidartinib (CSF1R inhibitor)'.)

•For patients with unresectable, recurrent, or metastatic DFSP, imatinib may be indicated. (See 'Imatinib' above and "Dermatofibrosarcoma protuberans: Treatment", section on 'Treatment of locally advanced, recurrent, and metastatic disease'.)

●No role for high-dose chemotherapy with stem cell rescue – High-dose chemotherapy with stem cell rescue is not a standard approach to treatment of advanced STS of the usual adult types, and we recommend that this approach not be pursued outside of the context of a clinical trial (Grade 1A). (See 'Dose intensification' above.)

●Treatment at disease progression – For patients with progression on an initial doxorubicin-based regimen who retain a good performance status, we prefer enrollment in a clinical trial if one is available. If protocol treatment is not available or is declined, our approach to subsequent therapy in patients who retain a good performance status is histology-driven:

•Gemcitabine-based therapy, typically in combination with docetaxel, is a reasonable option, particularly for undifferentiated pleomorphic sarcoma/malignant fibrous histiocytoma and LMS. (See 'Gemcitabine plus docetaxel' above.)

•Trabectedin is approved for advanced LMS and liposarcoma in many countries, including the United States. It is particularly active in myxoid/round cell liposarcoma and is the preferred second-line agent for this histology. (See 'Trabectedin' above.)

•Eribulin is approved for use in liposarcoma in the United States and for liposarcoma and LMS in other countries. It may have more activity in pleomorphic and dedifferentiated liposarcoma than trabectedin, and it is preferred over trabectedin for second-line therapy. (See 'Eribulin' above.)

•Pazopanib is approved in many countries for advanced sarcomas other than liposarcoma or GIST. It may have more activity in LMS, synovial sarcoma, angiosarcoma, and SFT than in other histologic types. (See 'Pazopanib' above.)

•For patients with angiosarcoma, a weekly taxane is a good option for subsequent therapy if an anthracycline was used as initial therapy. (See 'Taxanes (angiosarcoma)' above.)

•For patients with ES who progress on initial anthracycline-based regimens, tazemetostat is an available treatment option. (See "Uncommon sarcoma subtypes", section on 'Tazemetostat'.)

•Other options for second-line treatment and beyond in patients who retain a good performance status include PLD, ifosfamide alone or with doxorubicin (if combination therapy was not given first line), gemcitabine alone, or a gemcitabine-based combination. (See 'Pegylated liposomal doxorubicin' above and 'Ifosfamide' above and 'Gemcitabine and other agents' above and 'Gemcitabine-based combinations' above and 'Doxorubicin-based regimens' above.)

•Next-generation sequencing infrequently identifies candidates for molecularly targeted agents approved by the US Food and Drug Administration (FDA). These agents include larotrectinib or entrectinib for individuals whose tumors have gene fusions without known resistance mutations involving one of the neurotrophic tyrosine receptor kinase (NTRK) genes, and pembrolizumab for the rare individual whose STS has high levels of microsatellite instability/deficient mismatch repair. (See 'Next-generation sequencing' above.)

ACKNOWLEDGMENT — The UpToDate editorial staff acknowledges George Demetri, MD, who contributed to an earlier version of this topic review.