Sarcoma Surveillance: A Review of Current Evidence and Guidelines

Cara A. Cipriano, MD, MS; Eugene Jang, MD, MS; Wakenda Tyler, MD, MPH


J Am Acad Orthop Surg. 2020;28(4):145-156. 

In This Article

Review of Evidence

Local Recurrence Surveillance: Considerations

LR after treatment of sarcomas in the extremities is associated with increased morbidity and mortality. Re-excision of LR can be a relatively minor procedure if the recurrence is detected early, but larger recurrences can require a major operation involving complex reconstruction or amputation. The STS literature also suggests that the occurrence of LR independently predicts higher rates of subsequent metastasis and reduced overall survival.[2] However, if LR can be detected and treated before metastasis, patients' overall survival is markedly better than if metastasis has already occurred.[3] Thus, surveillance for LR can theoretically improve survival and reduce morbidity of treatment.

A number of risk factors are associated with the incidence of LR after sarcoma and, therefore, justify more intense local surveillance. Time since initial treatment is a strong predictive factor for both bone sarcoma and STS; most LRs take place within the first 5 years, and frequency continues to decline with time.[4,5] Histologic grade of the tumor also affects the LR rate in extremity soft-tissue and bone sarcomas (Figures 1 and 2). A retrospective study of 105 patients with STS by Sugiura et al[3] identified tumors deep to the fascia or located in the upper extremity/trunk as higher risk of LR. In addition, several studies have found that close or positive margins are strongly associated with the risk of LR as well as higher mortality in both bone sarcoma and STS, although the precise definition of an appropriate margin remains debated.[2,3,8,9] When positive margins result from unplanned resections, timely and appropriate treatment with an adequate wide resection results in a return to the baseline rate of LR; however, the associated surgical morbidity is often higher.[10,11] Local contamination as a result of tumor seeding during biopsy can also potentially contribute to LR. Barrientos-Ruiz et al[12] histologically examined biopsy tract resections and found a higher incidence of contamination (cell seeding) in cases of open compared with percutaneous biopsies; moreover, they observed a higher rate of LR in the cases with evidence of contamination. Conversely, Binitie et al[13] found no difference in disease-free survival when the tract of a percutaneous biopsy was not excised, as long as patients subsequently received radiation. Taken together, these findings suggest that percutaneous biopsy combined with radiation does not produce clinically significant local contamination, but open biopsy can potentially increase the risk of LR.

Figure 1.

Graphs showing the rates of local recurrence (LR) for high- and low-grade soft-tissue sarcoma (STS). A, For high-grade STS, the incidence of LR is highest in the first 2 years after treatment. B, For low-grade soft-tissue sarcoma, the rate is more constant, with late recurrences occurring more frequently than in high-grade tumors.5 Although higher grades of STS have shown a clear association with higher rates of LR, the evidence to date for primary bone sarcomas has not revealed a similar link between grade and the incidence of LR.6

Figure 2.

Kaplan-Meier plot demonstrating the rates of local recurrence (LR) for high-, intermediate-, and low-grade bone sarcomas. LR appeared to occur more frequently in higher grade sarcomas during the first 3 years after treatment, but with the number available for study no notable differences were found between rates of LR among tumor grades. LR after 5 years was rare, and no instances of LR of low-grade sarcomas beyond 4 years were found.7

Local Recurrence Surveillance: Modalities

Several approaches to monitoring for local recurrence have been proposed. Physical examination remains an essential component of surveillance, and patient self-examination has been shown to be highly effective.[14] Radiographs are a relatively low-cost and low-risk imaging modality that represents an important means of identifying LR of bone sarcomas. However, they are less effective in detecting soft-tissue LR.[4] MRI is therefore also heavily used, especially for STS, although its benefit over physical examination has not been established. Several studies have suggested that the majority of LR are identified by patients, and that MRI should be reserved for tumors not easily evaluated by physical examination.[14–16] Bone scan is not applicable for STS and nonspecific for bone sarcomas, especially after skeletal reconstruction. Finally, positron-emission tomography (PET) in conjunction with CT is often used for staging purposes and has also been assessed as a surveillance tool. The evidence to date does not demonstrate a notable difference between receiver operating characteristic curves of PET/CT versus MRI for detecting LR of STS.[17] These data, combined with its increased cost, do not support the use of PET imaging for routine surveillance at this time.

Pulmonary Metastasis Surveillance: Considerations

The lungs are the most common site of sarcoma metastases, and as such, PMs have notable implications on mortality.[18] Studies have found that in STS, the presence of >5 PMs at time of diagnosis markedly reduces median survival (22 versus 55 months),[19] and the number and distribution of subcentimeter lung nodules predicts survival in young sarcoma patients.[20] A growing body of evidence supports the role of interventions for limited PMs as a means of improving survival in both bone sarcoma and STS.[21,22] The potential to improve survival through early detection of PM may justify the practice of intense chest surveillance for patients who are likely to benefit from treatment. However, the prognosis of metastatic sarcoma remains poor, and many patients are not eligible for curative treatments, and so, in these situations, the benefits of surveillance are less clear.

Generally, established risk factors for sarcoma PM include high histologic grade, location deep to the fascia, and larger tumor size. Additional risk factors have been described for patients with STS (close/positive margins, LR after appropriate resection, increased patient age, male sex, and <90% necrosis after preoperative treatment)[2,3] as well as osteosarcoma (tumor location, histologic response to chemotherapy, patient age, and laboratory values such as alkaline phosphatase).[23,24] As with LR, the time from surgery influences the risk of PM. For both soft-tissue and bone sarcomas, high-grade tumors are most likely to metastasize within 2 years of follow-up, whereas low-grade tumors metastasize at a lower and more constant rate (Figures 3 and 4). Thus, to maximize the yield of PM detection per examination, the frequency of pulmonary surveillance should be dictated by the tumor grade and time since treatment.[7]

Figure 3.

Graphs showing the rates of pulmonary metastasis (PM) for high- and low-grade soft-tissue sarcoma (STS). A, For high-grade STS, the incidence of pulmonary metastasis was markedly higher in the initial 2 years after treatment. B, For low-grade STS, the incidence was more constant over time, with late PM occurring at higher rates than low-grade tumors.5

Figure 4.

Kaplan-Meier plot demonstrating the rates of pulmonary metastasis (PM) for high-, intermediate-, and low-grade bone sarcomas. Higher metastatic rates were observed in higher grade sarcomas. Although the frequency of events generally decreased over time for all groups, most PMs for high grade occurred within the first 2 years, whereas the rate was more stable for low- and intermediate-grade sarcomas. New metastases were rare after 4 years for low-grade tumors and after 10 years for intermediate- or high-grade tumors. Of note, osteosarcomas (OSA) that were histologically classified as intermediate grade were noted to follow the pattern of high-grade sarcomas and were therefore considered grade 3 in this analysis.7

Pulmonary Metastasis Surveillance: Modalities

The use of radiograph versus CT for pulmonary surveillance is a highly debated topic in orthopaedic oncology. Chest CTs, when compared with chest radiographs, offer the ability to detect smaller lung metastases; however, the clinical relevance of this sensitivity has been questioned. Some studies have shown that the increased sensitivity of chest CT translates into a statistically significant increase in overall survival for osteosarcoma PM.[25] By contrast, level I evidence published by Puri et al demonstrated no disease-free or overall survival benefit from CT compared with radiograph; in addition, in a follow-up study, they found no advantage to imaging every 6 versus 3 months.[14,26]

CT scans are also not without controversy and risk. The increased sensitivity of chest CT is also accompanied by a higher false-positive rate, with a 33% rate of finding false-positive nodules after two annual chest CTs, versus just 15% when using chest radiograph over the same period.[27] Furthermore, radiation exposure from a CT scan is two orders of magnitude higher than radiograph and may pose a notable cumulative risk of malignancy.[28] In a 2016 Musculoskeletal Tumor Society (MSTS) survey, 62.9% of respondents indicated that patients have expressed concerns regarding radiation exposure from surveillance within the past year.[29] PET imaging has been suggested as a means of avoiding the radiation exposure of CT for pulmonary surveillance, and 18-FDG-PET has been shown to have high sensitivity and intermediate specificity for lung nodules >1 cm. However, PET is expensive and has relatively low sensitivity for subcentimeter lung nodules,[30] which limits its utility for pulmonary surveillance in its current technological state.

Currently, the use of radiograph versus CT varies according to training and the geographic region. For example, European surgeons are more likely to use chest radiographs for surveillance, whereas surgeons in the United States more frequently use chest CT.[29,31,32] Despite this practice pattern of increased chest CT usage in the United States, the results from an MSTS survey indicated that one-third of respondents felt that CT is overutilized for high-grade tumors and two-third felt CT is overutilized for low-grade tumors.[33] The use of separate protocols for high- and low-risk surveillance, with fewer routine CTs for the latter, would potentially be more cost effective, radiation sparing, and noninferior in terms of accuracy.[32,34] The specific diagnosis and patient characteristics should also be taken into consideration, as these factors impact whether early detection of PM can improve outcomes. For example, if effective treatments are not available, or if the patient is not eligible for existing treatments due to age or comorbidities, early detection of PM becomes less clinically relevant, and a less intense surveillance plan may be justified.

Extrapulmonary Metastasis Surveillance: Considerations

EMs from sarcoma are rare and are associated with a worse prognosis than PM alone (15 versus 38 months).[19] Historically, the presence of EM has been considered a contraindication for resection of PM, but recent research suggests that resection of both PM and EM results in survival (37.8 months) that is similar to those with only PM treated with lung metastasectomy (35.5 months). Furthermore, both of the aforementioned groups had markedly improved survival when compared with those with both PM and EM that only had lung metastasectomy and left the EMs untreated (13.5 months).[33] Therefore, timely diagnosis of EMs after sarcoma treatment may have a tangible effect on treatment choices and patient survival.

Although the incidence and risk factors for EM are not entirely understood, higher grade STS is associated with a higher incidence of distant EM.[35] Certain sarcoma histologies also exhibit a greater predilection for EM; in particular, myxoid-round cell liposarcoma is known to metastasize to the retroperitoneum, abdominal wall, and abdominal cavity.[36,37] Solitary fibrous tumors can also require long-term surveillance because of benign-acting masses reportedly transforming and metastasizing decades after initial presentation.[38] Although hematologic spread predominates in most sarcomas, regional lymphatic spread is known to occur in synovial sarcoma, angiosarcoma, rhabdomyosarcoma, clear-cell chondrosarcoma, and epithelioid sarcoma; therefore, extrapulmonary monitoring may be justified in these diagnoses. Finally, distant EMs of osteosarcoma are rare but have been described in the bones, pleura, and heart. It has been hypothesized that newer medical treatments may be responsible for the apparent increase in incidence of EM of osteosarcoma in recent history.[39] If this proves to be the case, more routine extrapulmonary surveillance may justified in certain situations.

Pulmonary Metastasis Surveillance: Modalities

Because relatively few indications are noted for surveillance for EM aside from the histologies listed previously, no strong evidence exists to support the use of routine CT of the chest, abdomen, and pelvis in STS surveillance.[40] Similarly, surveillance of extrapulmonary sites is not generally recommended for osteosarcoma, although bone scans represent an accurate methodology for detecting the subsets of EM from osteosarcoma that are calcified.[41] There may be a role for whole-body PET scan for surveillance for EM in the future, but as of yet limited evidence exists to support this practice.