The surgical approach to adult soft tissue tumors has undergone significant changes over the past 10 years. A better understanding of the natural history of the different histologic subtypes,[1, 2] the importance of site, and the different sensitivity to available drugs has opened the way to advances in the individualized treatment of most of these tumors. Given the increasingly personalized nature of therapy, patients with soft tissue sarcoma (STS) should be referred to high-volume sarcoma centers with multispecialty expertise and a dedicated, multidisciplinary sarcoma tumor board.3-5 Ideally, the treatment strategy should be planned at diagnosis, and a decision for surgery should be placed in the context of all other available therapies. The extent of surgery may vary broadly between the different histologic entities; and, at times, surgery may be delayed or even omitted altogether. Although surgery remains the standard and only potentially curative therapy in the management of localized STSs and gastrointestinal stromal tumors (GISTs), it is now postponed to a later line in desmoid-type fibromatosis (DF) if or until other available options fail. Although they belong to the same overarching family of tumors, the natural histories, surgical principles, and sensitivity to locoregional and systemic therapies differ completely among these 3 entities; therefore, they are discussed separately in this review. The treatment of pediatric sarcomas (including primitive peripheral neuroectodermal tumors, alveolar/embryonal rhabdomyosarcomas, and desmoplastic small round cell tumors) is very different, and systemic chemotherapy (CT) is the primary therapy. The discussion of pediatric sarcomas is beyond the scope of the current review.
Sporadic Desmoid-Type Fibromatosis
DF is a rare mesenchymal tumor that accounts for 0.03% of all neoplasms and 3% of all soft tissue tumors. The expected incidence of DF is between 0.2 and 0.4 per 100,000 inhabitants per year. DF never metastasizes but recurs frequently after surgery, generally in a locoregional manner, and can be multifocal. Although most DF arise sporadically, approximately 5% to 10% of these tumors arise in the context of familial adenomatous polyposis (FAP). Sporadic DF predominantly affects young adults, especially females; and, although it may be observed in nearly every part of the body, it often involves the extremities (including pelvic and shoulder girdles), the trunk, and the abdominal cavity (mostly within the mesentery or the pelvis). FAP-associated DF is located predominantly within the abdomen, has no sex predilection, and its natural history is complicated by previous surgery for FAP.
Historically, surgery (either alone or in association with radiation therapy [RT]) was considered the mainstay of treatment, but it was associated with considerable morbidity and high recurrence rates even after apparently adequate local treatment. The observation in more contemporary series that microscopic surgical margins were not associated significantly with a higher local recurrence rate led to a reassessment of overall management, and preservation of function became a priority.10-12 Furthermore, some patients experienced spontaneous stabilization or occasional regression of their tumors, even after multiple recurrences. Therefore, investigators proposed to further limit morbidity by considering an initial observation period in all patients, including those with primary tumors, especially when surgery would have resulted in some function loss.[14, 15]
In parallel with these observations, it became evident that DF is not a single disease but, rather, comprises at least 2 entities with quite different behavior. Some DFs are marked by an indolent, and others can be locally aggressive, but predicting tumor is challenging. The more indolent tumors typically do not recur regardless of margin status because of their natural tendency to regress or remain stable; whereas aggressive lesions do recur despite the adequacy of surgical resection and, of course, they recur more commonly after margin-positive resection.
Several recent retrospective series have consistently reported progression-free survival rates of 50% at 5 years for patients who were managed using a front-line, conservative, “watchful waiting” approach.[14, 15] Although those initial observations demonstrated the potential safety of such an approach, it is important to emphasize that the patients remained under close observation, such that no patient was lost to follow-up, and treatment plans could be altered if tumors progressed. Increase in tumor size was usually mild and occurred in the first 2 years. Spontaneous regressions were also observed in as many as 5% to 10% of patients. There may be sites in which regression is more common and observation may be safer (ie, the abdominal wall). Nevertheless, regression has been observed at all sites. Therefore, it is reasonable to consider watchful waiting as an initial step even when asymptomatic tumors are located at critical sites (ie, mesentery) before undertaking subsequent treatments (Fig. 1).
When the observational approach fails, surgery is still a valid option. When performed, the objective of surgical resection should be to obtain microscopic negative margins, although function preservation—especially for tumors located in the extremities and girdles—should always be taken into consideration; and other alternatives, including RT, should be considered when appropriate. Furthermore, patients with large, sporadic, mesenteric/retroperitoneal DFs may undergo surgical resection anyway just because of the tumor size and possible related symptoms.
Therefore, in the absence of biologic prognostic factors capable of predicting the natural history of the disease, watchful waiting is a reasonable approach to minimize overtreatment and unnecessary morbidity in a subset of patients with desmoid tumors while secondarily also potentially limiting the costs of treatment. Prospective observational studies are currently underway to validate these results and possibly shed more light on the biologic background of this intriguing disease (ClinicalTrials.gov Identifier NCT01801176).
Soft Tissue Sarcomas
STSs have an expected incidence of 5 per 100,000 inhabitants per year. They comprise a family of >50 different histologic subtypes that can arise anywhere in the body.[1, 2] The most common site of origin is the extremities, followed by the trunk, and the retroperitoneum. STSs tend to occur in the sixth decade of life, but there are variations in age predilection according to the different histologies (ie, synovial sarcoma and myxoid liposarcoma in younger patients, leiomyosarcoma and myxofibrosarcoma in older patients).
The basic principle of surgery for STS is that the tumor must be resected en bloc with a cuff of healthy tissue to avoid contamination of the residual tissue from the tumor surface and to remove tumor microsatellites, which may be present in the pseudocapsule. Indeed, most STSs are pushing tumors rather than invading tumors and have a pseudocapsule at their periphery. This pseudocapsule is a reactive zone that separates the tumor tissue from surrounding healthy, noncancerous tissue. Surgical resection along the pseudocapsule is often the most amenable plane of resection. However, resecting the tumor along this plane is associated with an unacceptable rate of local recurrence. Therefore, widening the surgical margins to include en bloc resection of noncancerous, normal tissue beyond the pseudocapsule is recommended and is associated with a lower risk of local recurrence. In limb and trunk wall sarcomas, this basically implies the resection of surrounding soft tissues, mainly muscles, subcutaneous fat, and skin. Tumors abutting bone may include the periosteum as a margin if the bone is not directly invaded. If necessary, vascular resection and reconstruction should be considered for involved vessels. If a critical nerve is encased, then reconstruction with an interposition nerve graft should be considered. In the retroperitoneum, widening surgical margins imply the resection of adherent viscera even when not overtly involved, although a truly negative resection margin is hard to obtain at this site. Then, microscopic evaluation of resection margins is essential to assess the quality of surgical resection and the adequacy of local treatment in extremities and trunk wall tumors, while it's of questionable use in retroperitoneal ones. In essence, the surgeon should try to avoid positive microscopic surgical margins (ie, the presence of tumor cells at the cut surface), because—at least in the extremities and trunk wall—these are clearly associated with a greater risk of local failure and death, especially on the long run, when the quality of surgery becomes the strongest prognostic factor (Table 1).20-23
|Series||Study Period||Median FU, mo||No. of Patients||5-Year Rate, %||10-Year Rate, %|
Controversy over the adequacy of margins has continued for years.,20-23 On one hand, ideal margins may result in unacceptable morbidity (the extreme example being an amputation). Conversely, limiting the width of margins to preserve function may result in unacceptable rates of local recurrence and an increased risk of death, especially for tumors located at critical sites.
With a refined understanding of the natural history of the different sarcomas, their sensitivity to available treatments, and the broader use of reconstructive techniques, surgical resection of STS has been tailored to histologic subtypes and sites of origin. Better functional outcomes in patients with extremity STS and some improvement in local control in those with retroperitoneal STS have been the results.
STS of the Extremities and Trunk Wall
The rates of local recurrence at these sites range from 5% to 10%. Local recurrence is not the leading cause of death, because patients often can be salvaged by a wider reoperation and, in the extremities, with an amputation. The goal of surgery is limb sparing and function sparing while achieving appropriate biologic margins.
Personalizing surgery to histology
For tumors with a very indolent course, the resection margins at initial surgery may be intentionally limited, and more radical resections may be reserved for patients who develop local recurrences. An example is well differentiated liposarcoma (often called atypical lipomatous tumor) in the extremities and trunk wall, for which a marginal excision is the preferred primary approach. The risk of recurrence is in the 30% range, and a wider resection can be planned reasonably at the time of recurrence. Furthermore, because the risks of dedifferentiation and systemic spread are negligible in these tumors, function preservation is a principal goal. Therefore, the prognostic advantage of an initial wide resection is minimal and should be counterbalanced by the functional impairments related to a wider resection, especially in elderly patients who are often affected by this histotype. Radiation therapy is rarely used in this specific subtype, independent of tumor size, because the resulting fibrosis can make any subsequent surgical resection in case of recurrence more challenging.
Another example of a tumor in which limited surgery should be the standard of care is dermatofibrosarcoma protuberans (DFSP). DFSP is a superficial tumor that infiltrates soft tissue for centimeters beyond the obvious margins of the lesion and can recur locally after an inadequate resection. The more common variety of DFSP does not display metastatic behavior. The goal of surgery should still be to achieve negative margins, which often necessitates reconstruction by plastic surgery. Once negative margins are obtained, the risk of failure is virtually none. When cosmesis is an issue, limited positive margins may be acceptable, and a wider resection can be postponed until DFSP locally recurs. Moreover, a limited positive margin does not automatically translate into local recurrence and does not increase the risk of metastatic spread. Radiation therapy is not routinely considered, even in the presence of positive margins. Approximately 5% to 10% of patients with DFSP have a more aggressive fibrosarcomatous variant that may recur locally and metastasize. They should receive the same treatment as patients affected by a “conventional” sarcoma using more aggressive local therapy (including radiation) and should be followed with systematic imaging.
At the other end of the spectrum, myxofibrosarcoma has a high local recurrence rate, especially when negative margins cannot be achieved.[27, 28] These tumors commonly arise in the extremities of elderly individuals. Unlike other sarcomas, when located superficially, the tumor infiltrates through soft tissue (subcutaneous fat and investing fascia) centimeters beyond the ostensible margins of the visible or palpable mass, even at initial presentation. When located intramuscularly, the extension of the infiltration is usually more limited by anatomic barriers, although it has a higher propensity to invade into those anatomic boundaries compared with other histologic subtypes. Myxofibrosarcomas demonstrate a 30% rate of local recurrence and a 16% rate of distant recurrence.[27, 28] Multiple local recurrences have been associated with eventual amputation. Therefore, it is critical to pursue aggressive local therapy. Wide surgical margins (2-4 cm radial margins beyond the clinical boundaries of the palpable mass, especially in more superficial tumors) should be the goal of surgery, which often requires complex wound closure or flap reconstruction by a plastic and reconstructive surgeon as well as potential resection and reconstruction of vessels and/or nerves. Radiation therapy, either preoperatively or postoperatively, may be considered, although its direct impact on this specific histology remains unknown.
Personalizing surgery to response to neoadjuvant therapy
There are tumors for which additional treatments can be delivered in the preoperative setting to facilitate preservation of a critical structure and, possibly, function. This was initially demonstrated in a retrospective series in which surgical resection with “planned” positive margins on critical structures after RT was not associated with a worse outcome compared with wide resection. It is noteworthy that this was not observed when patients underwent simple marginal resection with multifocal residual disease. This initial observation has been confirmed in recent reports: in both retrospective and prospective series, the effect of positive surgical margins on local outcome was lost in patients who received preoperative RT in combination with neoadjuvant CT. The only caution is that preoperative RT is associated with a higher risk of wound complications.[31, 32] Therefore, generous resection of at-risk, superficial soft tissues and proactive wound-care measures may be needed to reduce complication rates.
Although those studies were designed to address the role of various neoadjuvant therapies for STS in general, the specific sensitivity of the different histologic subtypes to different treatments may help in tailoring strategy and optimizing outcomes. An example of this principle is in patients with myxoid liposarcoma. This histology arises predominantly in the lower extremities and is 1 of the most chemosensitive and radiation-sensitive STSs. The low-grade variant can be treated by preoperative RT. The high-grade variant, which is marked by the presence of spindle cell or round cell morphology, may be treated by both anthracycline-based CT regimens and/or RT. If an anthracycline regimen or RT cannot be used, then trabectedin (ecteinascidin 743 or ET-743; Yondelis; developed by Johnson & Johnson [New Brunswick, NJ] and PharmaMar SA [Madrid, Spain]) is an effective option for either form of myxoid liposarcoma, although its availability in the United States is limited.[35, 36] Response to these treatments will allow potentially more conservative operations with very limited local risk.
Another example of customizing surgery to histology and neoadjuvant therapy is angiosarcoma. Angiosarcoma is sensitive to both RT and CT. Different compounds have produced activity in angiosarcoma, including anthracycline, taxanes, and gemcitabine.[37, 38] Because surgery is rarely curative, it should not be considered routinely as first-line therapy for all presentations. The superficial variant is commonly located on the scalp and affects elderly individuals. It usually presents as multifocal disease and, thus, preoperative therapy should be considered. Because of multifocality, curative resection is rarely possible on the scalp without significant morbidity, and even negative margins are no guarantee against local recurrence. When angiosarcoma arises in deep tissues, multifocality is rare and the outcome is usually dominated by distant spread. The treatment approach may include CT, RT, and surgery on a case-by-case basis. Patients with vascular angiosarcoma have a particularly dismal prognosis and also usually are treated using a multimodality approach. For tumors that are not suitable to macroscopically complete surgical resection, definitive RT, possibly with heavy particles, can be considered.
Synovial sarcoma is chemosensitive, particularly to ifosfamide.[39, 40] Because ifosfamide-based CT may be administered in the preoperative setting either concurrent with or sequential to RT, this approach is often considered for patients with large and deep tumors to optimize function preservation and local control. However, CT does not consistently improve overall survival when administered in the (neo)adjuvant setting, so its routine use in the absence of metastatic disease remains controversial for this histology. A similar approach can be also offered to patients with undifferentiated pleomorphic sarcoma (UPS), because responses to anthracycline-based CT and RT have been observed (Fig. 2).
On a general note, whenever a plastic and/or vascular/nerve reconstruction is planned, CT and/or RT preferably should be delivered in the preoperative setting rather than in the postoperative setting, because these treatments may delay or compromise the functional results. Nevertheless, as mentioned above, a greater risk of wound complications should be factored,[31, 32] and proactive wound-care measures may be needed to reduce complication rates.
Finally, for tumors that are resistant to conventional treatments, hyperthermic isolated limb perfusion (ILP) can be an option. This procedure involves placing vascular access catheters into the main artery and vein of the affected extremity and perfusing with high-dose CT (usually melphalan) and tumor necrosis factor alpha (if available) under hyperthermic conditions. Although this treatment is nowadays safe in experienced hands, early and late complication rates of >20% have been reported. High rates of complete responses (range, 15%-30%) and further limb-sparing procedures (80%) are achieved by ILP, but no randomized trials have ever compared it with aggressive limb-sparing resection plus RT for STS. ILP should be considered as potential therapy when other options are not available in appropriately selected patients, and eligible patients should be referred to centers where this therapy is available.
STS of the Retroperitoneum
Local recurrence rates after resection of primary retroperitoneal sarcoma (RPS) range from 20% to 50%, and local rather than distant recurrence is the leading cause of death in half of patients with RPS (Table 2).42-48 Surgery should be aimed at achieving macroscopically complete resection and minimizing microscopically positive margins. This is best done by resecting the tumor en bloc with adherent structures, even if they are not overtly infiltrated (Fig. 3). In essence, this strategy often may include ipsilateral nephrectomy and colectomy; locoregional peritonectomy and myomectomy (partial/total) of the muscle of the lateral/posterior abdominal wall (usually psoas); splenectomy and left pancreatectomy for tumors located on the left upper side; occasionally pancreaticoduodenectomy or hepatectomy for tumors located on the right side; and vascular and bone resection only if vessels/bone are overtly infiltrated. Recent retrospective comparisons of series collected through prospectively maintained databases have suggested that this approach in primary RPS may translate into better local control (roughly 80% vs 50% at 5 years) (Table 2) and potentially better overall survival (roughly 70% vs 50% at 5 years). However, this remains a very controversial issue, and opinion on the extent of uninvolved, contiguous organ resection is divided among sarcoma surgeons at high-volume centers. Preservation of specific organs (ie, kidney, head of the pancreas, and/or liver) should be considered individually based on factors that include histologic subtype, organ invasion, encasement of organ vascular supply, and comorbidities; and specific expertise in the disease is required to make the right decisions. Judgment must also be used in deciding which neurovascular structures to sacrifice, weighing the potential for local control against the potential for long-term dysfunction.
|5-Year Rate, %|
|Series||Study Period||Median FU, mo||No. of Patients||No. Who Underwent Complete Resection||LRFS||OS|
|Ferrario & Karakousis 2003||1977-2001||41||79||99||43||65|
The majority of STSs in the retroperitoneum are made up of 4 well characterized histologic subtypes: well differentiated liposarcoma, dedifferentiated liposarcoma, leiomyosarcoma, and solitary fibrous tumor. They have different natural histories and patterns of failures, which should be factored into the decision-making process when planning the treatment strategy. When the local approach is optimized, well differentiated liposarcoma only recurs locally; leiomyosarcoma predominantly metastasizes hematogenously; dedifferentiated liposarcoma can either recur locally or metastasize; and solitary fibrous tumor seldom recurs locally or metastasizes, unless a sarcomatous component comes into play. It is important to consider extended surgery for all RPS subtypes, but other adjuvant or neoadjuvant strategies are sorely needed to improve long-term results (especially for patients with dedifferentiated liposarcoma and leiomyosarcoma).
Preoperative RT in resectable tumors is being investigated in a currently accruing, prospective, randomized clinical trial (European Organization for Research and Treatment of Cancer study 62092; STRASS). The objective of preoperative treatments is not to change the extent of surgery but, rather, to reduce locoregional failure, which is the predominant cause of death.
Postoperative/adjuvant external-beam radiation after complete gross resection has limited value and is associated with significant short-term and long-term toxicities. Therefore, it has fallen out of favor.
Surgery for locoregional recurrence may be offered on an individual basis, especially to patients who are affected by low-grade liposarcoma, those who have a long disease-free interval between initial resection and subsequent recurrence, or those who experience a response to therapy. There is little benefit to incomplete resections for recurrent disease with the possible exception of patients who have well differentiated liposarcoma. Palliative surgery may have short-term symptomatic benefit, but this rarely lasts longer than 90 days.
Gastrointestinal Stromal Tumors
GISTs have an expected incidence of 1.5 per 100,000 inhabitants per year, representing from 0.1% to 3% of all gastrointestinal malignancies, although they are the commonest mesenchymal malignancy of the gastrointestinal tract. They commonly arise in the stomach (50%-70% of tumors), small intestine (25%-35% of tumors), colon and rectum (5%-10% of tumors), mesentery or omentum (7% of tumors), and esophagus (<5% of tumors).
In the last 12 years, the understanding and treatment of GISTs have witnessed remarkable advances because of 2 key developments: 1) the identification of constitutively active signals (oncogenic mutation of the c-KIT and platelet-derived growth factor alpha [PDGFRA] genes encoding receptor tyrosine kinases),[51, 52] and 2) the development of therapeutic agents that suppress tumor growth by specifically targeting and inhibiting this signal, resulting in improved outcomes (imatinib mesylate, sunitinib malate, and regorafenib).53-55 The advent of effective therapy has dramatically improved outcomes for patients with GIST. Nevertheless, surgery remains the only potentially curative therapy for patients who have localized GISTs.
In general, surgery for GIST is a wedge or segmental resection of the involved gastric or intestinal tract, with margins that can be less wide than those required for an adenocarcinoma. Lymphadenectomy is not routinely required, because lymph nodes are rarely involved (in adult patients) and, thus, are resected only when they are clinically detectable. Sometimes, a more extensive resection is needed (total gastrectomy for a large, proximal, gastric GIST; pancreaticoduodenectomy for a periampullary GIST; or abdominoperineal resection for a low rectal GIST).
Indeed, all GISTs that measure >2 cm in greatest dimension should be resected when possible, because none of these can be considered “benign.” The management of GISTs that measure <2 cm in greatest dimension is more questionable.57-59 Although the low risk of progression of GISTs that measure <2 cm offers the possibility of a more conservative approach, a reliable mitotic index cannot be determined by biopsy or fine-needle aspiration, thus preventing the identification of those at higher risk. Therefore, both observation and resection can well be considered for GISTs that measure from 1 cm to 2 cm in greatest dimension, and the risks and benefits of one versus the other should be discussed with the patient. Endoscopic resection of small gastric GISTs could be an option in these presentations. The risk of perforation may be low in highly experienced hands, although the decision is made on a case-by-case basis. Regardless of their size, any small GISTs that are symptomatic (eg bleeding from erosions through the mucosa) or that increase in size on serial follow-up should be resected.
Open or laparoscopic/laparoscopy-assisted resection of a primary GIST should be performed according to standard oncologic principles. Although primary GISTs may demonstrate inflammatory adhesions to surrounding organs, true invasion is not frequent. The goal of surgery is a macroscopically complete, en bloc excision. A macroscopically complete resection with negative or microscopically positive margins (R0 or R1 resection, respectively) is associated with a better prognosis than a macroscopically incomplete resection (R2 resection). Available series have demonstrated that patients who undergo R1 resection have no greater risk of local failure than those who undergo R0 resection, with or without adjuvant imatinib; therefore, re-excision for microscopically positive margins is rarely if ever indicated.[60, 61] One exception is rectal GISTs, in which microscopic surgical margins appear to be associated with a greater risk of local failure and death. In general, however, local relapse after R0 surgery is very unlikely in patients with GISTs. In discussing margins with the patient and the pathologist, it is important to understand that most exophytic tumors are simply covered by a thin layer of peritoneum over a portion of the tumor; thus, it is not surprising that there is a high rate of peritoneal relapse even after R0/R1 surgery. Tumor rupture or violation of the tumor capsule during surgery is associated with a very high risk of recurrence, essentially equivalent to that of metastatic disease, and thus should be avoided.
Given the extensive use of adjuvant therapy with imatinib in high-risk populations and the activity of the drug, the use of preoperative imatinib should be considered before undertaking any major surgical procedure. This is particularly true for tumors located at critical sites, such as the esophagogastric junction, the duodenum, and the rectum (Fig. 4), reducing the need for a complex operation. Imatinib may also be considered for large tumors at risk of rupture or for large tumors in which tumor shrinkage may enable a minimally invasive approach. Retrospective single-institution and multiple-institution series as well as a few prospective studies have demonstrated that this approach is safe and can minimize surgical morbidity.64-66 Currently, neoadjuvant imatinib is recommended by all available guidelines whenever surgical morbidity is not considered to be minimal.[4, 67] The only drawback is that it is impossible to calculate the risk of recurrence and, thus, to adapt to the need for further adjuvant therapy, because the mitotic rate in the surgical specimen is inevitably influenced by preoperative therapy. Therefore, all patients who receive neoadjuvant imatinib are considered to be at high risk and receive treatment for a total of at least 3 years.
In the rare syndromic GIST (either sarcosine dehydrogenase [SDH]-deficient or neurofibromin 1 [NF1]-related), tumors are often multifocal and are confined either to the stomach (SDH-deficient GIST) or the small bowel (NF1-related GIST).68-70 The extent of surgery should be determined on a case-by-case basis, taking into account the risk of recurrence, the lack of benefit from available tyrosine kinase inhibitors, and the actual behavior of the underlying disease.
Soft tissue tumors may arise in a variety of body sites and within a variety of tissues. The management of DF, STS, and GIST requires a thorough understanding of the biology of the different diseases and molecular subtypes as well as the constraints of specific anatomic site. Treatment planning should include multidisciplinary consultation to determine optimal therapy, taking into consideration tumor histology, site and extent of the disease, its natural history and sensitivity to available treatments, surgical challenges, and of course, the wishes of the patient.