Surgery in soft tissue sarcoma: the thin line between a surgical or more conservative approach

Surgery is the primary treatment for localized, clinically resectable soft tissue sarcoma (STS), and is recommended for advanced or metastatic STS when complete resection is feasible. In some situations, however, there is thin line between selecting a surgical or more conservative approach. Multiple factors need to be considered before deciding to proceed with surgery. This review provides several examples of settings where surgery may not be best patient care and discusses key factors to consider in the decision-making process.

Surgical abstention

Although, for many years, surgery was the standard primary treatment option for desmoid tumors, a more conservative management approach has recently been adopted. The current primary approach is active surveillance, with surgery considered only after failure of medical therapies [1,2]. The benefits of a conservative approach were first identified in two retrospective studies of desmoid tumors. The first study found that 3-year event-free survival was similar in patients who underwent R0 surgery (with or without radiotherapy) and those who were simply observed [3]. Subsequently, a bi-institutional study reported a 5-year progression-free survival (PFS) rate of 50% in patients managed according to a ‘wait and see’ approach, similar to that of patients who had received medical therapy (mainly chemotherapy and hormonal therapy) [4]. Spontaneous regression of desmoid tumors has been demonstrated in a randomized controlled trial (RCT) and reported in several retrospective studies. In a placebo-controlled RCT of sorafenib, the regression rate with placebo was 20% [5]. Two retrospective studies observed a spontaneous regression rate of 27% (29 of 102) in abdominal wall desmoid tumors [6] and of 22% (27 of 122) in nonabdominal wall desmoid tumors [7]. An Italian prospective study of 108 desmoid tumors described an initial spontaneous regression rate of 21%, with another 20% of tumors spontaneously regressing after initial progression [8].

A conservative management approach also applies to other STS types. For example, aggressive angiomyxoma usually occurs in the pelvis and perineum of young women. Surgery in such a delicate area can be avoided, as the tumor can remain stable for several years and may benefit from hormonal therapy [9,10]. Another specific type in which surgical abstention should be considered is epithelioid hemangioendothelioma (EHE). Figure 1 shows tumor evolution in a non-operated patient who continued to maintain disease control 3 years after undergoing isolated limb perfusion. It has been observed that hepatic EHE can remain stable for years or even spontaneously reduce. A computed tomography (CT) evaluation of 15 patients with hepatic EHE found that six patients (40%) had reduction and five patients (33%) had no change in the growth of the largest tumor during a median follow-up of 36 months [11]. For this reason, it is controversial to consider patients with hepatic EHE as candidates for liver transplants.

Inoperability

From a technical viewpoint it can be stated that, nowadays, almost any tumor is resectable. However, there are situations of advanced disease where surgery may not be worthwhile, as often occurs with retroperitoneal sarcomas that are prone to central vascular encasement (celiac trunk, mesenteric axis, suprarenal aorta, retrohepatic vena cava with extension into the right atrium). Other determinants of inoperability include unfavorable histological type and patient-related factors such as poor performance status. A retrospective analysis of patients with primary retroperitoneal sarcomas (n = 322) treated at a single center in Italy reported that 11.5% (n = 37) did not undergo surgical resection with curative intent. The main reasons precluding surgery were technically unresectable tumors (n = 19) mainly due to involvement of celiaco-mesenteric vessels, and comorbidities/poor performance status (n = 18) [12]. A key consideration in situations of uncertainty is that: ‘the more complex the surgery, the less likely it will benefit the patient’. However, since some patients do benefit from complex surgeries, each case must be evaluated with the proper perspective, paying special attention to cases of major vascular encasement, low performance status or unfavorable histological type. Recently, the Trans-Atlantic Australasian RetroPeritoneal Sarcoma Working Group (TARPSWG) updated consensus recommendations regarding the management of primary retroperitoneal sarcoma [13]. This group is working toward defining the term inoperability as no established criteria currently exist for its definition.

Locoregional recurrence

Factors to consider when deciding to perform surgery after disease recurrence are disease-free interval, tumor growth rate, histological type and subtype, likelihood of a macroscopically complete resection and response to chemotherapy.

In a retrospective study of primary retroperitoneal sarcoma, histology (well-differentiated liposarcoma), disease-free interval and surgical resection were found to be predictive of post-relapse overall survival (OS) [14]. A high local recurrence growth rate (>0.9 cm/month) is associated with poor outcomes in recurrent retroperitoneal liposarcoma [15]. The majority of patients with retroperitoneal sarcoma who undergo surgery for a first local recurrence will develop a second local recurrence [16], and those who undergo surgery for a second local recurrence will typically develop a third local recurrence [17]. However, any arguments against performing surgery are complicated by the fact that about 20% of operated patients with recurrent retroperitoneal sarcoma will be disease free at 5 years, underscoring the need for careful patient selection based on disease-free interval, complete resectability and histological type.

In patients with local recurrence, a response to chemotherapy is also a good indicator for recommending further resection and may ‘upgrade’ negative prognostic factors such as multifocal recurrence soon after initial surgery. Conversely, disease progression after chemotherapy is a predictor of poor outcomes as illustrated in a retrospective review of 81 sarcoma patients who received chemotherapy for lung metastases (first site of recurrence), then underwent pulmonary metastasectomy. Five-year survival rates after metastasectomy in patients whose disease progressed after neoadjuvant chemotherapy were noticeably worse compared with patients without measurable disease progression (0 vs 32%) [18].

An ongoing Italian multicenter Phase II study is aiming to confirm the activity of trabectedin as second- or further-line treatment in terms of slowing tumor growth in patients with retroperitoneal leiomyosarcoma and well differentiated/dedifferentiated liposarcoma. Among patients considered amenable to surgery, tumor tissue specimens collected before and after treatment are being compared to characterize the biological features of the tumor associated with different response patterns to trabectedin (ClinicalTrials.gov: NCT02247544) [19].

Metastatic STS

In metastatic STS, the decision to perform surgery is multifactorial. The timing of surgical resection is made on a case-by-case basis, as indicated by the following case studies.

Case 1: retroperitoneal leiomyosarcoma with lung metastases

A 66-year-old male patient presented with retroperitoneal leiomyosarcoma and synchronous lung metastases. Doxorubicin plus dacarbazine (×2 cycles) produced stable disease, which was followed by a minor response to gemcitabine plus dacarbazine (×5 cycles). The primary tumor was resected; however, the patient experienced hepatic and lung progression after 2 months. Rechallenge with gemcitabine plus dacarbazine (×4 cycles) elicited a partial response. The patient was switched to trabectedin monotherapy. He also received liver transarterial chemoembolization and radiotherapy on a single progressing lesion in the liver. 6 years after diagnosis and 5 years post-surgery, he continues to receive trabectedin monotherapy. In this case, the decision to resect the primary tumor, but not metastatic disease, was correct.

Case 2: retroperitoneal dedifferentiated liposarcoma

Case 2 involves a 54-year-old male patient with left retroperitoneal dedifferentiated liposarcoma. He underwent en bloc partial resection of the tumor and left kidney in a community hospital, followed by complete resection of residual disease at our center in Milan. Six months later, lung metastases appeared which were treated with long infusional ifosfamide (×5 cycles). The lung metastases were resected; however, 5 months later, two diaphragmatic metastases were detected which were treated with definitive radiotherapy. 22 months later, the patient developed a new lung metastasis, an endocaval metastasis and a locoregional recurrence invading the body of the L2 lumbar vertebra. Doxorubicin plus olaratumab (×6 cycles) produced stable disease. The switch to trabectedin monotherapy (×31 cycles) led to a progressive partial response. Due to histology and the technical difficulty of surgery required to resect recurrent disease, the decision was taken to continue trabectedin until progression. At the time of writing, the patient was alive with disease 7 years after diagnosis and 3 years after beginning trabectedin.

Case 3: uterine leiomyosarcoma

Case 3 describes a 44-year-old female with uterine leiomyosarcoma. Primary disease was treated with surgery, radiotherapy and adjuvant chemotherapy (gemcitabine plus docetaxel ×4 cycles). 4 years later she developed peritoneal metastases which were treated with doxorubicin plus dacarbazine for a minor response and subsequent surgical removal of visible disease. A single new peritoneal nodule appeared 1 year later which was also resected. 2 years later, three new peritoneal deposits occurred and were surgically removed. 2 years later, a single lung metastasis was detected which was surgically removed. At 14 years after diagnosis and 3 years after her last resection, there is no evidence of disease.

Conclusion

Although surgery is the primary treatment modality for localized resectable STS, certain cases demand a more conservative approach. As has been demonstrated with desmoid tumors and a few other unusual histological types, surgical abstention can sometimes translate to better patient care. Surgery of some tumors is not recommended for technical or patient-related reasons or because the tumors are prone to multiple recurrences despite surgery. A decision to proceed or not proceed with surgery should always be based on multidisciplinary discussion. Factors in favor of surgery in advanced disease are isolated oligometastatic disease, long disease-free interval, suitable histology, response to chemotherapy and high probability of a complete resection.