The relationship between surgical margin status and outcomes in sarcoma has been an area of controversy for years. Some question whether a positive margin represents inadequate surgery or perhaps is a marker of aggressive cancer biology. This article reviews the literature regarding the natural history of positive margins and its possible influence on sarcoma recurrence and survival.
There is an inherent complexity to managing patients with soft tissue sarcomas (STS). They are comparatively rare tumors with multiple histologies and anatomic variability. STS may also have unpredictable biologic behavior that can lead to late recurrence or distant metastases. Important variables that influence STS outcomes are divided into 3 main categories. First are patient factors that include age, sex, the anatomic location of the primary tumor, and depth. Second, and probably the most important, are tumor factors such as size, grade, and the specific histology. Third are treatment factors that include the adequacy of the surgical resection and the effect of adjuvant therapies. There is an interaction between all of these prognostic factors. Studies have shown that large, deep, high-grade STS at certain locations (such as head and neck or retroperitoneum) have a significantly increased risk for a positive-margin surgical resection. But the critical question is whether this is caused by bad surgery, a controllable variable, or bad biology, which cannot be controlled. Currently, the complex interaction between STS resection margin status, local recurrence (LR), and survival is not clear.
In 1982, Dr Rosenberg and the National Cancer Institute published a seminal paper that changed the surgical treatment of extremity STS. This was a prospective study in which 43 patients with high-grade tumors were randomized to limb-sparing surgery plus radiation therapy (RT) versus amputation (both groups received systemic chemotherapy). The LR rate for the limb salvage group was 15% (1 successfully salvaged with amputation) compared with no LR in the amputation group ( P = .06). However, there were no differences in 5-year disease-free survival (DFS) (71% vs 78%, P = .75) or overall survival (OS) (83% vs 88%; P = .99). A multivariate analysis of patients with STS from this study and another National Cancer Institute protocol showed that a positive resection margin was significantly associated with an increased risk for LR ( P <.0001). The investigators had several provocative conclusions. First, an increased rate of LR did not seem to affect survival. Perhaps patients with an LR were simply more likely to develop distant metastases even if local control had been obtained with an amputation. In addition, if a negative-margin limb-sparing surgery could be combined with adjuvant radiation and chemotherapy, then long-term survival seemed to be comparable with amputation. Shortly thereafter, the National Institutes of Health formulated a consensus statement recommending limb-sparing surgery for most patients with high-grade extremity STS.
What is the appropriate STS resection margin?
Once limb-sparing surgery became the standard of care, the appropriate STS resection margin needed to be defined. Enneking and others previously classified margins based on the type of surgical resection performed. With intralesional excision, the plane of excision has passed through the tumor, leaving microscopic or macroscopic tumor behind. With marginal excision, the gross tumor is removed within or close to the pseudocapsule with no attempt to remove a cuff of normal tissue. With wide excision, the gross tumor is removed with a rim of normal surrounding tissue. With compartmental (radical) excision, the entire compartment in which a tumor is located is removed. Retrospective studies of STS prognostic factors have correlated this classification with outcome. LR after intralesional and marginal excisions has been unacceptably high. In contrast, compartmental/radical excision, similar to amputation, has lower LR rates but at the expense of poorer limb function and decreased quality of life.
In the modern limb salvage era, the optimal STS resection margin has generically been defined as the amount of tissue necessary to obtain a negative-margin resection. In practice, this is typically at least 1 to 2 cm of grossly normal tissue around the tumor and the pseudocapsule (or less if there is an intact biologic tissue barrier such as fascia). However, especially when combined with RT and a surgical intent to preserve a critical structure (such as vessels, major nerves, or bone), margins of less than 1 cm may still be adequate. As covered in the article by Dr Delaney elsewhere in this issue, wide excision and RT result in acceptable LR rates and good limb functionality in most patients.
Pathologic assessment of the STS resection margin
The pathologic assessment of the resected STS specimen is not only important for the diagnosis (histologic type and grade), which affects prognosis but also the margin status, which determines the need for additional therapy (eg, reresection, RT). Therefore, it is recommended that all STS specimens be evaluated by a pathology team with expertise in this disease. There are also some inherent difficulties and limitations to margin assessment and pathologic evaluation of STS specimens. Ex vivo, the relationship of the tumor to surrounding structures can change greatly. Surrounding muscle fibers do not remain isometric after being transected. This muscular retraction and opening of intermuscular planes can lead to margins being inaccurately defined as close or positive. Therefore, communication between the surgeon and the pathologist who will be grossing the tumor is imperative. Ideally, the surgeon should take the resection specimen to pathology and anatomically orient it for the pathologist. The specimen should be fresh, not formalin fixed, to allow the pathologist to examine/measure the specimen undistorted, take photographs, and perform specialized cytogenetic and molecular diagnostic studies, if indicated. The resection specimen should be measured in 3 dimensions and the closest distance for each margin from the edge of the tumor should be noted. The surgeon should also identify any margins of concern, the anticipated closest margin, and other structures critical to the pathologic assessment, such as margins that contain intact fascia. Margin inking should be performed only after photographs of the specimen are obtained and a thorough gross examination has been performed. Permanent blocks should be taken to definitely include the closest gross resection margin and the deep margin, when appropriate. It is recommended that even wide gross margins (>3 cm) from the tumor be assessed for some histologic subtypes known to infiltrate significant distances (myxofibrosarcoma and epithelioid sarcoma).
In certain circumstances, intraoperative frozen section assessment is informative and helps guide therapy. It may help define clinically suspicious tissues when resecting recurrent or previously irradiated tumors. If there is an intraoperative positive margin, consideration could be given to resection of additional tissue or perhaps intraoperative radiation. However, frozen section analysis can often be time consuming, technically challenging, and inaccurate (eg, atypical spindle cells in an irradiated prior surgical site). Therefore, unless the results of the frozen section will change the intraoperative management, it is not routinely recommended for STS resections.
Pathologic assessment of the STS resection margin
The pathologic assessment of the resected STS specimen is not only important for the diagnosis (histologic type and grade), which affects prognosis but also the margin status, which determines the need for additional therapy (eg, reresection, RT). Therefore, it is recommended that all STS specimens be evaluated by a pathology team with expertise in this disease. There are also some inherent difficulties and limitations to margin assessment and pathologic evaluation of STS specimens. Ex vivo, the relationship of the tumor to surrounding structures can change greatly. Surrounding muscle fibers do not remain isometric after being transected. This muscular retraction and opening of intermuscular planes can lead to margins being inaccurately defined as close or positive. Therefore, communication between the surgeon and the pathologist who will be grossing the tumor is imperative. Ideally, the surgeon should take the resection specimen to pathology and anatomically orient it for the pathologist. The specimen should be fresh, not formalin fixed, to allow the pathologist to examine/measure the specimen undistorted, take photographs, and perform specialized cytogenetic and molecular diagnostic studies, if indicated. The resection specimen should be measured in 3 dimensions and the closest distance for each margin from the edge of the tumor should be noted. The surgeon should also identify any margins of concern, the anticipated closest margin, and other structures critical to the pathologic assessment, such as margins that contain intact fascia. Margin inking should be performed only after photographs of the specimen are obtained and a thorough gross examination has been performed. Permanent blocks should be taken to definitely include the closest gross resection margin and the deep margin, when appropriate. It is recommended that even wide gross margins (>3 cm) from the tumor be assessed for some histologic subtypes known to infiltrate significant distances (myxofibrosarcoma and epithelioid sarcoma).
In certain circumstances, intraoperative frozen section assessment is informative and helps guide therapy. It may help define clinically suspicious tissues when resecting recurrent or previously irradiated tumors. If there is an intraoperative positive margin, consideration could be given to resection of additional tissue or perhaps intraoperative radiation. However, frozen section analysis can often be time consuming, technically challenging, and inaccurate (eg, atypical spindle cells in an irradiated prior surgical site). Therefore, unless the results of the frozen section will change the intraoperative management, it is not routinely recommended for STS resections.
Margin status and LR
Given the rarity and heterogeneity of STS combined with the inconsistent use of adjuvant therapies, accurate data on LR rates are difficult to obtain and interpret. Table 1 summarizes the results of several large STS series. LR rates vary greatly from 7% to 42%. Currently, the expected risk for LR in the modern multimodality STS treatment era should be less than or equal to 20%.
| Series | Time Frame | Patients (n) | Follow-Up (y) | High Grade (%) | Deep (%) | Radiation (%) | Amputation (%) | LR (%) |
|---|---|---|---|---|---|---|---|---|
| Stotter et al, 1990 | 1982–1987 | 175 | 3 | 60 | 77 | 55 | 0 | 42 |
| Pisters et al, 1996 | 1982–1994 | 1041 | 4 | 65 | 76 | 40 | 10 | 17 |
| Coindre et al, 1996 | 1980–1989 | 546 | 5 | 84 | 79 | 56 | 4 | 29 |
| Yang et al, 1998 | 1983–1991 | 132 | 10 | 70 | NR | 50 | 0 | 11 |
| Baldini et al, 1999 | 1970–1994 | 74 | 10.5 | 49 | 66 | 0 | 0 | 7 |
| Karakousis and Driscoll, 1999 | 1977–1994 | 194 | 3 | 86 | 88 | 42 | 7 | 15 |
| Trovik et al, 2001 | 1986–1995 | 1331 | 6 | 78 | 66 | 24 | 10 | 17 |
| Zagars et al, 2003 | 1960–1999 | 1225 | 9.5 | 71 | N/A | >95 | 0 | 41 |
| Lahat et al, 2008 | 1997–2007 | 1091 | 4.4 | 67 | 64 | 44 | N/A | 16 |
Most of the STS literature shows that a positive resection margin is associated with an increased risk for LR ( Table 2 ). Biologically, this makes sense, but what is more intriguing is the natural history of a positive-margin resection. Most positive margins do not lead to an LR. In a Memorial Sloan Kettering Cancer Center (MSKCC) analysis of 2084 patients with STS (all anatomic sites), 22% (n = 460) had a positive resection margin. The LR rate for a positive margin was 28%. Conversely, that means that 72% of positive-margin resections did not result in an LR. Although the LR rate was significantly higher for positive versus negative-margin resections (risk ratio [RR] 2.4), 15% of negative margins still had an LR. In a similar large series from MD Anderson Cancer Center (MDACC) of 1091 patients with STS, 22.6% of patients (n = 247) had a positive resection margin despite all gross tumor being clinically resected. However, only 15.9% of all 1091 patients (n = 173) developed an LR. Consequently, although a positive resection margin places a patient with STS at a higher risk for LR, it is still more likely that an LR will not occur.
| Series | Time Frame | Patients (n) | Margin Classification | LR | Survival |
|---|---|---|---|---|---|
| Tanabe et al | 1970–1987 | 95 | Positive vs negative | Microscopic positive surgical margin or intraoperative tumor violation had increased LR | Neither positive margin nor LR adversely affected survival |
| Lewis et al | 1982–1994 | 495 | Positive vs negative | On univariate and multivariate analysis, LR was associated with positive microscopic margins (RR 2.1) | On multivariate analysis, a positive microscopic margin had a worse metastasis-free survival |
| Pisters et al | 1982–1994 | 1041 | Positive (within 1 mm) vs negative | Positive margin was an independent prognostic factor for LR (RR 1.8) | On multivariate analysis, positive margin was an adverse factor for DSS |
| Lewis et al | 1982–1995 | 911 | Positive vs negative | On univariate and multivariate, a positive microscopic margin was associated with increased LR | Positive microscopic margin was associated with distant recurrence and decreased DSS |
| Zagars et al | 1960–1999 | 1225 | Positive, negative, or uncertain | On univariate and multivariate analysis, margin status was a major factor contributing to local control (RR 2.5) | Positive margin was associated with decreased disease-free survival |
| Stojadinovic et al | 1982–2000 | 2084 | Positive vs negative | Positive margin nearly doubled the risk of LR | Positive margin increased the risk of disease-related death |
| Dickinson et al | 1987–2002 | 279 | Contaminated, >20 mm, 10–19 mm, 5–9 mm, 1–4 mm, <1 mm | Contaminated margins had higher rates of LR | Failure to obtain an uncontaminated margin was associated with decreased OS |
| Liu et al | 1997–2007 | 181 | 0–1 mm, 1–4 mm, 5–9 mm, 10–19 mm, 20–29 mm, ≥30 mm | Margin<10 mm was an independent risk factor for LR | Margin<10 mm was associated with decreased metastasis-free and DSS |
| Novais et al | 1995–2008 | 248 | Positive at ink, ≤2 mm, >2 mm but ≤2 cm, >2 cm | Margin≤2 mm was associated with increased LR | Inadequate surgical margin (≤2 mm) was associated with decreased OS |
Related posts:
Prognostic Factors and Staging for Soft Tissue Sarcomas: An Update
Adjuvant Chemotherapy for Soft Tissue Sarcomas
Pulmonary Metastasectomy for Soft Tissue Sarcoma
Isolated Regional Therapy for Advanced Extremity Soft Tissue Sarcomas
Updates on the Management of Gastrointestinal Stromal Tumors
Sarcomas and the Immune System: Implications for Therapeutic Strategies
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
