Abstract
Background:
Given the propensity for lung metastases (LM), NCCN guidelines recommend lung surveillance (LS) with either CXR or CT in high-grade soft tissue sarcoma (HG-STS). Considering survival, diagnostic sensitivity and cost, the optimal modality is unknown.
Methods:
The US Sarcoma Collaborative database (2000-2016) was reviewed for patients who underwent resection of a primary HG-STS. Primary outcome was overall survival (OS). Cost analysis was performed.
Results:
Among 909pts, 83% had truncal/extremity, 17% had retroperitoneal (RP) tumors. Recurrence occurred in 48% of which 54% were LM. LS was performed with CT in 80% and CXR in 20%. Both groups were clinically similar although CT patients had more RP tumors and recurrences. Regardless of modality, 85-90% of LM were detected within the first 2yrs with a similar re-intervention rate. When considering age, tumor size, location, margin status, and receipt of radiation, LM was independently associated with worse OS(HR:4.26; p<0.01) while imaging modality was not(HR:1.01; p=0.97). CXR patients did not have an inferior 5-year OS compared to CT(71vs60%, p<0.01). When analyzing patients in whom no LM was detected, both cohorts had a similar 5-year OS(73vs74%, p=0.42), suggesting CXR was not missing clinically relevant lung nodules. When adhering to a guideline-specified protocol for 2018 projected 4,406 cases, surveillance with CXR for 5yrs results in savings of $5-8M/year to the US healthcare system.
Conclusion:
In this large multicenter study, LS with CXR did not result in worse overall survival when compared to CT. With considerable savings, a CXR-based protocol may optimize resource utilization for LS in HG-STS; prospective trials are needed.
Keywords: Soft-tissue sarcoma, surveillance, lung metastases
Precis
In this study, surveillance with chest x-ray (CXR) for high-grade soft tissue sarcoma was not associated with a worse overall survival when compared to CT. A CXR surveillance protocol would result in savings between $5 million and $8 million for an annual cohort of patients over a 5-year surveillance period.
Introduction
Soft tissue sarcomas are rare tumors which account for 1% of adult malignancies. In 2018, approximately 13,000 people were diagnosed with soft tissue sarcomas in the United States1. During the past three decades, a multimodality approach has been used in the treatment of primary, high-grade soft tissue sarcomas leading to improvements in survival. Despite this, distant recurrences are common, with up to 60% of high-grade soft tissue sarcomas recurring in the lungs2,3. The rate of metastases depends predominantly on tumor grade, and 70% of high-grade soft tissue sarcoma lung metastases will occur within the first two-years after resection4-7.
Considering this rapid progression of high-grade soft tissue sarcomas, prompt detection of lung metastases may improve prognosis given therapeutic interventions currently available. Surgical metastasectomy remains the primary treatment modality for isolated lung metastases, and although no randomized control trial has evaluated its benefit over medical therapy, several retrospective series have demonstrated 3-year survival rates of 40-50% after complete metastasectomy2,3,7-18. Even when resection is not feasible, other lung-directed strategies, such as radiofrequency ablation or stereotactic body radiotherapy have demonstrated acceptable local control rates19-21.
Due to the availability of salvage therapy and its association with improved survival, post-operative lung surveillance is crucial. However, consensus is lacking regarding the optimal imaging modality. Current National Comprehensive Cancer Network (NCCN) guidelines for high-grade soft tissue sarcomas recommend imaging with either chest radiography (CXR) or chest computed tomography (CT)22. Although CXR is easily-accessible and minimizes radiation, the enhanced resolution of CT may improve the sensitivity of detection for lung nodules as small as 3-4 mm23,24. However, patients with <5 mm nodules have been shown to have equivalent survival to those with normal CT scans25. Additionally, the higher false-positive rate for CT may result in costly, unnecessary assessments / procedures with potential increased morbidity and patient anxiety26. Furthermore, the cost between CXR and CT differs by an order of magnitude. Intuitively, elimination of unnecessary CT scans for lung surveillance of high-grade soft tissue sarcoma would result in significant savings to the US-healthcare system.
Considering the potential variability in diagnostic sensitivity, and known differential cost of each modality, the preferred modality of lung surveillance in high-grade soft tissue sarcoma remains unknown. Therefore, the primary objective of this study was to evaluate the difference between CXR and CT lung surveillance after curative resection of high-grade soft tissue sarcoma in regards to overall survival and cost to the US-healthcare system.
Methods
Data Source and Study Variables
The US Sarcoma Collaborative (USSC) is a consortium formed to investigate outcomes in soft tissue sarcoma and constitutes eight academic centers (Emory University, Stanford University, Wake Forest University, Medical College of Wisconsin, University of Wisconsin, University of Chicago, The Ohio State University, Washington University). All patients who underwent resection of a primary sarcoma (2000-2016) were identified. In order to mitigate selection bias, this study was limited to high-grade tumors as pathological grade may affect selection of modality for lung surveillance. High-grade was defined as Grade 2 or 3 using the French Federation of Cancer Centers Sarcoma Group (FNCLCC) or as determined in the pathology report by the National Cancer Institute (NCI) system. The analysis was further limited to patients with R0 or R1 resections, without metastatic disease at the time of resection or 30-day operative mortality, and those with lung surveillance data available. Clinicopathologic variables and post-operative outcomes were collected through chart review. As the study was conducted by eight academic institutions, NCCN guidelines for lung surveillance frequency were followed. However, there were no institutional protocols to guide selection of modality of surveillance and this decision was largely based on physician preferences. Patients were considered to have CXR surveillance if they exclusively underwent imaging with CXR throughout the surveillance period. If a patient was transitioned from CXR surveillance to CT at any point in his / her lung surveillance period and prior to the detection of lung metastasis, they were included in the CT surveillance cohort. Once a suspicious lesion was identified with either modality, further diagnostic work-up (cross sectional imaging, percutaneous biopsy or wedge resection) was pursued at the discretion of the treating physician and there were no institutional protocols or algorithms to guide this decision-making. Institutional Review Board approval was confirmed at each institution prior to data collection.
Statistical Analysis
All statistical analyses were performed using the SPSS 22.0 statistical package (IBM Inc., Armonk, NY). Statistical significance was pre-defined as 2-tailed p<0.05. Nominal variables were analyzed with Chi-square or Fisher’s exact test. Continuous variables were analyzed using t-tests or the Wilcoxon signed-rank test. Survival was estimated using the Kaplan–Meier method, and the log-rank test was used for comparison of survival between CXR and CT cohorts. Cox-regression analysis was used to determine the association of clinicopathologic factors with overall survival. A multivariable model was constructed using sequential regression entry with variables statistically associated (p<0.05) with overall survival on univariate analysis.
Cost Model Comparing CXR and CT
A cost model was developed to estimate total cost to the US-healthcare system over a 5-year period using either a CXR or CT-based surveillance protocol. The 2018 incidence data of non-metastatic, high-grade soft tissue sarcoma was determined based on published estimates. This hypothetical cohort was simulated to enter a CXR or CT-based protocol at low-frequency (every six months for four years, then annually) or high-frequency (every three months for the first two years, then every six months for two years, then annually). At each imaging time point there were three potential probabilities: no lung metastasis, true lung metastasis (true positive) or false lung metastasis (false positive). The probability of a true positive was calculated as the modality sensitivity multiplied by the true recurrence rate among those who did not die prior to the imaging time-point. The probability of a false positive is the modality false positive rate (1-specificity) multiplied by 1 minus true recurrence rate plus the death rate among those who did not die prior to the imaging time-point. The cost of each imaging modality or intervention was derived by using the 2018 Medicare Physician Fee Schedule and each service was identified using the Current Procedural Terminology (CPT) code. If a recurrence was detected, the downstream cost of histologic confirmation via wedge resection was also included. Each model was simulated 1000 times, and the average cost to the US-healthcare system at 5-years is reported.
Results
Demographic and Clinicopathologic Characteristics
Among 4,153 patients, 1,093 patients with high-grade soft tissue sarcoma underwent curative-intent resection and of these, 909 had lung surveillance data available. Tumor location included extremities in 71% (n=645), trunk wall in 12% (n=113), and retroperitoneum in 17% (n=151) (Figure 1). Tumor size was <5 cm in 15% (n=137), 5–10 cm in 40% (n=366), >10 cm in 39% (n=351), with a median of 9 cm (IQR 5.5-14.5). Tumors were classified into six main histologic categories as follows: undifferentiated pleomorphic sarcoma/malignant fibrous histiocytoma 39% (n=355), leiomyosarcoma 12% (n=106), myxofibrosarcoma 8% (n=75), dedifferentiated liposarcoma 4% (n=40), synovial sarcoma 5% (n=48), pleomorphic liposarcoma 5% (n=44), and others 27% (n=241). Median follow-up was 33 months.
Figure 1.
Patient population flow diagram. FNCLCC, French Federation of Cancer Centers Sarcoma Group; STS, soft tissue sarcoma; USSC, US Sarcoma Collaborative.
Among those patients who underwent curative resection (R0/R1), 48% (n=432) recurred and of these 34% were local/locoregional (n=149), 55% distant (n=239), and 10% synchronous locoregional / distant (n=42). Of all recurrences, 54% were in the lungs (n=232). Lung surveillance was performed with CXR in 20% (n=197) and CT in 80% (n=771).
Patients who underwent surveillance with CT had more retroperitoneal tumors, a higher proportion of dedifferentiated liposarcoma, and were more likely to have a lung metastasis (p<0.05). Importantly, both imaging modalities detected the majority of the lung metastases within the first two years (CXR: 91%, CT: 85%, p=0.88). Definitive therapy for these included ablation (CXR: 0%, CT: 0.4%), radiation (CXR: 9%, CT: 5%), surgery (CXR: 18%, CT: 40%) and chemotherapy (CXR: 18%, CT: 40%), and both groups had similar intervention rates to treat lung metastasis (p=0.77, Table 1).
Table 1.
Demographic and Clinicopathologic Factors of Patients with High-Grade Soft Tissue Sarcoma
| Variable | All patients (n=1,093) |
Surveyed patients (n=909) |
CXR (n=192) |
CT (n=717) |
p Value (CXR vs CT) |
|---|---|---|---|---|---|
| Demographic | |||||
| Age, y, median (IQR) | 61 (49-72) | 60 (48-71) | 62 (51-76) | 59 (48-71) | 0.02* |
| Sex, n (%) | 0.34 | ||||
| Male | 582 (53) | 480 (53) | 95 (49) | 385 (54) | |
| Female | 511 (47) | 429 (47) | 97 (51) | 332 (46) | |
| Missing | 0 (0) | 0 (0) | 0 (0) | 0 (0) | |
| Race, n (%) | 0.87 | ||||
| White | 818 (75) | 679 (75) | 144 (75) | 535 (75) | |
| Black | 123 (11) | 102 (11) | 22 (11) | 80 (11) | |
| Other | 112 (10) | 100 (11) | 19 (10) | 81 (11) | |
| Missing | 40 (4) | 28 (3) | 7 (4) | 21 (3) | |
| Primary location, n (%) | <0.01* | ||||
| Truncal | 131 (12) | 113 (12) | 18 (9) | 95 (13) | |
| Extremity | 762 (70) | 645 (71) | 157 (82) | 488 (68) | |
| Retroperitoneal | 200 (18) | 151 (17) | 17 (9) | 134 (19) | |
| Missing | 0 (0) | 0 (0) | 0 (0) | 0 (0) | |
| Clinicopathologic factor | |||||
| Tumor size, n (%) | 0.13 | ||||
| < 5 cm | 174 (16) | 137 (15) | 38 (20) | 99 (14) | |
| 5-10 cm | 422 (39) | 366 (40) | 74 (39) | 292 (41) | |
| > 10 cm | 434 (40) | 351 (39) | 70 (36) | 281 (39) | |
| Missing | 63 (5) | 55 (6) | 10 (5) | 45 (6) | |
| Histopathology, n (%) | |||||
| UPS/malignant fibrous histiocytoma | 424 (39) | 355 (39) | 79 (41) | 276 (39) | 0.56 |
| Leiomyosarcoma | 133 (12) | 106 (12) | 18 (9) | 88 (12) | 0.33 |
| Myxofibrosarcoma | 86 (8) | 75 (8) | 23 (12) | 52 (6) | 0.05 |
| Liposarcoma, dedifferentiated | 56 (5) | 40 (4) | 3 (2) | 37 (5) | 0.03* |
| Synovial | 53 (5) | 48 (5) | 11 (6) | 37 (5) | 0.89 |
| Liposarcoma, pleomorphic | 46 (4) | 44 (5) | 10 (5) | 34 (5) | 0.94 |
| Other | 295 (27) | 241 (27) | 48 (25) | 193 (27) | 0.66 |
| Lymph node metastasis, n (%) | 0.69 | ||||
| Negative | 121 (11) | 95 (10) | 9 (5) | 86 (12) | |
| Positive | 25 (2) | 20 (3) | 3 (2) | 17 (2) | |
| Missing | 947 (87) | 794 (87) | 180 (93) | 614 (86) | |
| Lymphovascular | 0.12 | ||||
| invasion, n (%) | |||||
| Negative | 625 (57) | 551 (61) | 130 (68) | 421 (59) | |
| Positive | 54 (5) | 41 (5) | 5 (3) | 36 (5) | |
| Missing | 414 (38) | 317 (35) | 57 (29) | 260 (36) | |
| Final resection status, n (%) | 0.82 | ||||
| R0 | 888 (81) | 755 (83) | 161 (84) | 594 (65) | |
| R1 | 205 (19) | 154 (17) | 31 (16) | 123 (17) | |
| Missing | 0 (0) | 0 (0) | 0 (0) | 0 (0) | |
| Adjuvant multimodal treatment, n (%) | |||||
| Chemotherapy | 346 (32) | 513 (35) | 65 (34) | 254 (36) | 0.78 |
| Radiation | 582 (53) | 513 (57) | 105 (55) | 410 (57) | 0.55 |
| Mode of lung surveillance, n (%) | |||||
| CXR | 192 (18) | 192 (21) | - | - | |
| CT | 717 (66) | 717 (79) | - | - | |
| First recurrence, n (%) | 455 (43) | 432 (48) | 33 (17) | 399 (56) | <0.01* |
| Local/locoregional | 156 (34) | 149 (34) | 16 (48) | 133 (33) | 0.3 |
| Distant | 249 (55) | 239 (55) | 15 (45) | 224 (56) | 0.3 |
| Both (locoregional + distant) | 48 (11) | 42 (10) | 2 (6) | 40 (10) | 0.3 |
| Lung metastasis† | 234 (79) | 232 (83) | 11 (65) | 221 (83) | <0.01* |
| Median follow-up (months) | 41 | 33 | 48 | 43 | 0.128 |
| Lung metastasis | |||||
| Timing of detection, n (%) | - | - | 0.88 | ||
| < 2 y | - | - | 10 (91) | 188 (85) | |
| 2-5 y | - | - | 1 (9) | 27 (12) | |
| > 5 y | - | - | 0 (0) | 1 (0.4) | |
| Intervention type, n (%) | - | - | 0.77 | ||
| Ablation | - | - | 0 (0) | 1 (0.4) | |
| Radiation | - | - | 1 (9) | 12 (5) | |
| Surgery | - | - | 5 (45) | 94 (43) | |
| Chemotherapy | - | - | 2 (18) | 88 (40) |
Significant.
Percentage in parentheses calculated with distant metastases (either distant alone or locoregional + distant) as the denominator.
CXR, chest x-ray; IQR, interquartile range; UPS, undifferentiated pleomorphic sarcoma.
Lung Metastases and Survival
On univariate Cox regression analysis, older age, retroperitoneal tumors, tumor size ≥ 5 cm, positive margin status, presence of lymphovascular invasion, and positive lymph node status were associated with worse overall survival. Lung metastasis was also strongly associated with worse overall survival (HR 3.91; 95%CI 3.11-4.92, p<0.01), while lung surveillance with CXR was not associated with inferior overall survival when compared to CT (HR 0.62; 95%CI 0.45-0.85, p<0.01). On multivariable Cox regression analysis, when controlling for age, tumor location, tumor size, margin status, and receipt of radiation, lung metastasis remained an independent predictor of worse overall survival (HR 4.26; 95%CI 3.28-5.53, p<0.01), while lung surveillance modality had no effect on overall survival (HR 1.01; 95%CI 0.71-1.43, p=0.97, Table 2).
Table 2.
Clinicopathologic Factors Associated with Overall Survival in R0/R1 Resections
| Variable | Univariable Cox regression | Multivariable
Cox regression |
||
|---|---|---|---|---|
| HR (95% CI) | p Value | HR (95% CI) | p Value | |
| Age, y | 1.02 (1.01-1.03) | <0.01* | 1.02 (1.01-1.03) | <0.01* |
| Sex | ||||
| Male | Reference | |||
| Female | 1.06 (0.85-1.34) | 0.60 | ||
| Race | ||||
| White | Reference | |||
| African American | 1.05 (0.73-1.51) | 0.78 | ||
| Other | 0.90 (0.61-1.33) | 0.60 | ||
| Primary location | ||||
| Truncal/extremity | Reference | Reference | ||
| Retroperitoneal | 0.64 (0.48-0.83) | <0.01* | 1.32 (0.97-1.80) | 0.08 |
| Tumor size | ||||
| < 5 cm | Reference | Reference | ||
| 5-10 cm | 1.79 (1.156-2.77) | <0.01* | 1.38 (0.87-2.18) | 0.11 |
| > 10 cm | 3.03 (1.98-4.62) | <0.01* | 2.13 (1.36-3.34) | <0.01* |
| Histopathology | ||||
| UPS/malignant fibrous histiocytoma | 0.86 (0.68-1.10) | 0.23 | ||
| Leiomyosarcoma | 1.06 (0.75-1.49) | 0.74 | ||
| Myxofibrosarcoma | 0.81 (0.51-1.29) | 0.37 | ||
| Liposarcoma, dedifferentiated | 1.29 (0.79-2.11) | 0.31 | ||
| Synovial | 0.79 (0.46-1.35) | 0.39 | ||
| Liposarcoma, pleomorphic | 1.16 (0.73-1.85) | 0.53 | ||
| Other | 1.18 (0.92-1.51) | 0.20 | ||
| Lymph node metastasis | ||||
| Negative | Reference | |||
| Positive | 2.61 (1.37-4.98) | <0.01* | ||
| Lymphovascular invasion | ||||
| Absent | Reference | |||
| Present | 1.94 (1.25-3.03) | <0.01* | ||
| Final margin status | ||||
| Negative (R0) | Reference | Reference | ||
| Positive (R1) | 1.64 (1.25-2.15) | <0.01* | 1.79 (1.34-2.39) | <0.01* |
| Multimodal treatment | ||||
| Radiation | 0.79 (0.63-0.99) | 0.04* | 0.75 (0.58-0.96) | 0.11 |
| Chemotherapy | 0.91 (0.72-1.16) | 0.44 | ||
| Recurrence | ||||
| No recurrence | Reference | Reference | ||
| Recurrence | 7.1 (5.22-9.66) | <0.01* | - | - |
| Lung metastasis | 3.91 (3.11-4.92) | <0.01* | 4.26 (3.28-5.53) | <0.01* |
| Lung surveillance modality | ||||
| CXR | Reference | Reference | ||
| CT | 1.61 (1.17-2.21) | <0.01* | 1.01 (0.71-1.43) | 0.97 |
Significant.
CXR, chest x-ray; HR, hazard ratio; UPS, undifferentiated pleomorphic sarcoma.
Survival Analysis by Lung Surveillance Modality
On log-rank analysis, patients in the CXR cohort did not have an inferior 5-year lung-specific recurrence free survival (CXR: 93% vs CT: 62%, p<0.01; Figure 2A) and 5-year overall survival (CXR: 71% vs CT: 60%, p<0.01; Figure 2B). However, when analyzing patients in whom no lung metastasis was detected, both imaging cohorts had identical 5-year overall survival (CXR: 74% vs CT: 73%, p=0.42; Figure 2C), suggesting that patients undergoing surveillance with CXR were not subjected to false negative imaging for clinically relevant lesions which otherwise would have resulted in decreased overall survival.
Figure 2.
Kaplan-Meier curves of patients with pulmonary metastasis compared according to chest imaging modality. (A) Lung-specific recurrence-free survival, (B) overall survival, and (C) overall survival excluding patients with documented lung metastases. CXR, chest x-ray.
Cost Analysis of Lung Surveillance Modalities
Estimated 2018 incidence data for non-metastatic, high-grade soft tissue sarcoma are shown in Table 3 and cost data are shown in Table 4. Over a 5-year surveillance period, a CXR-based protocol compared to CT results in a savings of $5,525,413-$7,853,732 to the US-healthcare system based on the 2018 Medicare Physician Fee Schedule, depending on whether a low or high-frequency strategy is used, respectively (Table 5).
Table 3.
Cost Model Assumptions
| STS incidence | n | % |
|---|---|---|
| Incidence of STS (2018) | 13,0401 | - |
| Non-metastatic STS | 10,43237 | 80 |
| High-grade STS | 6,67633 | 64 |
| Retroperitoneal STS | 2,67133 | 40 |
| Trunk STS | 66833 | 10 |
| Extremity STS | 1,06833 | 16 |
| Final cohort | 4,406 | - |
STS, soft tissue sarcoma.
Table 4.
Cost Data
| Modality | CPT code | Cost, $ |
|---|---|---|
| Chest x-ray | 71046 | 30.96 |
| CT | 72178 | 183.96 |
| Video-assisted thoracoscopic wedge resection | 32666 | 904.31 |
Cost of each imaging modality or intervention was derived by using the 2018 Medicare Physician Fee Schedule.
Table 5.
Cost Model Results for a Five-Year Surveillance Period for High-Grade Soft Tissue Sarcoma
| Surveillance protocol | Low frequency surveillance | High frequency surveillance |
|---|---|---|
| Chest x-ray, $, mean ± SD | 2,333,224 ± 32,057.71 | 2,985,268 ± 33,626.64 |
| CT, $, mean ± SD | 7,858,637 ± 62,783.94 | 10,839,000 ± 77,141.49 |
| Savings, $ | 5,525,413 | 7,853,732 |
Discussion
Nearly 60% of patients with high-grade soft tissue sarcoma will develop lung metastases after curative-intent resection with the risk of recurrence being greatest within two years of surgery27,28. This study’s findings are concordant with those in the literature with a lung metastasis rate of 52% (Table 1). Additionally, as previously known, our results demonstrate that lung metastasis is associated with a worse prognosis. Given this high rate of recurrence, the associated impact on survival, and the availability of salvage therapy, NCCN guidelines provide clear recommendations for lung surveillance. However, the optimal modality is unknown and either CXR or CT are accepted. Our results demonstrate that lung surveillance with CXR is not associated with an inferior overall survival compared to CT. Furthermore, depending on the frequency of imaging, a CXR-based protocol affords a potential cost savings of $5-8 million over a 5-year period to the US-healthcare system.
Several small studies and a randomized controlled trial have evaluated the optimal modality of lung surveillance in STS28-31. In a prospective, single-institution study, Puri et al. demonstrated that at a median follow-up of 42 months, surveillance with CXR after resection of extremity soft tissue sarcoma did not lead to worse survival when compared to CT30. Additionally, a retrospective study by Whooley et al., evaluated the effectiveness of follow-up testing for detecting distant recurrences of extremity soft tissue sarcoma and showed that 83% of asymptomatic lung metastasis were detected by CXR32.
The current study differs from the existing literature in that it further establishes the utility of CXR for lung surveillance after resection of high-grade soft tissue sarcoma, a subset of sarcoma that has been deemed high-risk for lung metastases. On Kaplan-Meier analysis, patients in the CXR cohort had a superior 5-year lung-specific recurrence-free survival (Figure 2A). It naturally follows to question whether this observation in recurrence-free survival is related to a decreased diagnostic sensitivity of the CXR modality, and hence a higher false-negative rate and inability to detect a metastasis. However, if patients surveyed with CXR had lung metastases that were not detected and therefore not treated, this cohort would likely have had a decreased overall survival when compared to the CT cohort. In contrast, the CXR cohort had an improved 5-year overall survival (Figure 2B). In order to further investigate this observation, survival analysis was repeated after excluding patients in whom no lung metastasis was detected which demonstrated near identical 5-year overall survival between both imaging cohorts (CXR: 73 vs CT: 74%, p=0.42; Figure 2C). Given the known poor prognosis of untreated lung metastases, this finding suggests that CXR is not associated with a high false-negative rate of clinically significant nodules which would otherwise have led to a worse overall survival when compared to CT. These results were further supported with multivariable Cox regression which demonstrated that surveillance modality was not associated with decreased overall survival (HR: 1.01; 95%CI 0.71-1.4; p=0.97), when considering age, tumor size, tumor location, margin status, and receipt of adjuvant radiation. Thus, it appears that CXR provides an adequate detection threshold for clinically significant lung nodules.
The decreased survival observed in the CT cohort is a result of selection bias, namely unidentified factors that influenced the decision to survey with CT versus CXR. Given this study’s retrospective design, these factors cannot be accurately identified. One potential explanation is that patients in the CT cohort were more likely to have primary retroperitoneal soft tissue sarcomas. It is well established that retroperitoneal sarcomas have a high propensity for early local recurrence, and that this local progression can be the main driver of disease-specific death33. Given that these patients generally undergo local abdominopelvic surveillance with CT, it is likely that lung surveillance would have been pursued with the same modality.
Notably, for both imaging cohorts, the majority of lung metastases were detected within the first 2-years (Table 1), a finding that is in accord with other series34. Furthermore, there was no difference in the intervention frequency and type pursued, including ablation, radiation, surgery, or chemotherapy, for these lung metastases (p>0.05). Therefore, it does not seem that CT surveillance results in an earlier diagnosis of lung metastases and more prompt treatment, particularly since the overall survival was not superior to CXR surveillance.
Although cost should not be the primary driver of our decision-making, the importance of being thoughtful about the cost of each intervention is ever-increasing. Cost-effectiveness is commonly cited in studies on surveillance strategies for soft tissue sarcoma. However, few have examined the actual costs to the US-healthcare system of follow-up surveillance according to NCCN guidelines which recommend chest imaging with either CXR or CT every 3-6 months for 2-3 years, then every 6 months for the next two years, and then annually22. A review by Goel et al. in 2004 summarized literature on the topic from 1982 to 2003 and found wide disparity in costs of 54 methods of following STS patients35. The financial analysis in our study, which is based on current NCCN guidelines and takes into account sensitivity and specificity of each modality, demonstrates that a CXR-based protocol could lead up to $5-8 million in savings to the US-healthcare system per 5-year surveillance period, depending on whether a low or high-frequency surveillance strategy is employed. Notably, when compared to CT, CXRs provide a decreased radiation exposure and are more readily available, particularly at smaller community-based practices.
The limitations of this work stem from its retrospective design and lack of granular data regarding frequency of surveillance. It should also be noted that this study includes only patients who were selected for surveillance with either CXR or CT. This decision is inherently subject to bias as patients chosen to undergo surveillance with CT may have been deemed to be higher risk for distant recurrence. Indeed, in 2003, Sakata et al. examined whether tumor grade and size accounted for variation in follow-up of STS. The authors found that office visits, labs and imaging were ordered more frequently with increasing tumor size and grade36. In an effort to reduce this selection bias, this study was limited to high-grade soft tissue sarcoma. However, it is difficult to account for all clinicopathologic differences between each group that could introduce bias. Furthermore, there are several limitations to our cost model resulting from our assumptions. The incidence data is derived from published estimates. Additionally, cost data was estimated using Medicare payments as a proxy, given the interest in estimating cost to the US-healthcare system. Although these costs may change over time and certainly vary per institution, the economic impact of a surveillance protocol using CXR is nonetheless compelling. The results of this study are not attempting to propose a protocol for generalized acceptance, but simply suggesting that the modality of lung surveillance in high-grade soft tissue sarcoma may include CXR with no associated decrease in survival. A prospective clinical trial is being developed to identify whether CXR is the optimal lung surveillance modality after resection of high-grade soft tissue sarcoma by providing a non-inferior survival as compared to CT at a reduced financial burden to the US-healthcare system.
Conclusion
In this large multicenter study, lung surveillance with CXR did not result in worse overall survival when compared to CT. Considering a potential cost savings of $5-8 million per 5-year surveillance period, a CXR-based protocol may optimize resource utilization for lung surveillance in patients after resection of high-grade soft tissue sarcoma.
Acknowledgments
Support: Supported in part by the National Center for Advancing Translational Sciences of the NIH under Award Number TL1TR002382 and the Katz Foundation. Research reported in this publication was supported in part by the Biostatistics and Bioinformatics Shared Resource of Winship Cancer Institute of Emory University and NIH/NCI under award number P30CA138292.
Footnotes
Disclosure Information: Nothing to disclose.
Disclaimer: The content is solely the responsibility of the authors and does not necessarily represent the official views of NIH.
Presented at the Society of Surgical Oncology Annual Cancer Symposium, San Diego, CA, March 2019.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- 1.Siegel RL, Miller KD, Jemal A. Cancer statistics, 2018. CA Cancer J Clin. 2018;68(1):7–30. [DOI] [PubMed] [Google Scholar]
- 2.Feig BW, Ching CD. The M.D. Anderson Surgical Oncology Handbook. Philadelphia: Wolters Kluwer/Lippincott Williams & Wilkins Health; 2012. [Google Scholar]
- 3.Cariboni U, De Sanctis R, Giaretta M, et al. Survival Outcome and Prognostic Factors After Pulmonary Metastasectomy in Sarcoma Patients: A 18-Year Experience at a Single High-volume Referral Center. Am J Clin Oncol. 2019;42(1):6–11. [DOI] [PubMed] [Google Scholar]
- 4.Markhede G, Angervall L, Stener B. A multivariate analysis of the prognosis after surgical treatment of malignant soft-tissue tumors. Cancer. 1982;49(8):1721–1733. [DOI] [PubMed] [Google Scholar]
- 5.Mandard AM, Petiot JF, Marnay J, et al. Prognostic factors in soft tissue sarcomas. A multivariate analysis of 109 cases. Cancer. 1989;63(7):1437–1451. [DOI] [PubMed] [Google Scholar]
- 6.Coindre JM, Terrier P, Guillou L, et al. Predictive value of grade for metastasis development in the main histologic types of adult soft tissue sarcomas: a study of 1240 patients from the French Federation of Cancer Centers Sarcoma Group. Cancer. 2001;91(10):1914–1926. [DOI] [PubMed] [Google Scholar]
- 7.Rehders A, Hosch SB, Scheunemann P, et al. Benefit of surgical treatment of lung metastasis in soft tissue sarcoma. Arch Surg. 2007;142(1):70–75; discission 76. [DOI] [PubMed] [Google Scholar]
- 8.Casson AG, Putnam JB, Natarajan G, et al. Five-year survival after pulmonary metastasectomy for adult soft tissue sarcoma. Cancer. 1992;69(3):662–668. [DOI] [PubMed] [Google Scholar]
- 9.Billingsley KG, Burt ME, Jara E, et al. Pulmonary metastases from soft tissue sarcoma: analysis of patterns of diseases and postmetastasis survival. Ann Surg. 1999;229(5):602–610; discussion 610–602. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Garcia Franco CE, Algarra SM, Ezcurra AT, et al. Long-term results after resection for soft tissue sarcoma pulmonary metastases. Interact Cardiovasc Thorac Surg. 2009;9(2):223–226. [DOI] [PubMed] [Google Scholar]
- 11.Pastorino U, Buyse M, Friedel G, et al. Long-term results of lung metastasectomy: prognostic analyses based on 5206 cases. J Thorac Cardiovasc Surg. 1997;113(1):37–49. [DOI] [PubMed] [Google Scholar]
- 12.Jablons D, Steinberg SM, Roth J, et al. Metastasectomy for soft tissue sarcoma. Further evidence for efficacy and prognostic indicators. J Thorac Cardiovasc Surg. 1989;97(5):695–705. [PubMed] [Google Scholar]
- 13.Gadd MA, Casper ES, Woodruff JM, et al. Development and treatment of pulmonary metastases in adult patients with extremity soft tissue sarcoma. Ann Surg. 1993;218(6):705–712. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Verazin GT, Warneke JA, Driscoll DL, et al. Resection of lung metastases from soft-tissue sarcomas. A multivariate analysis. Arch Surg. 1992;127(12):1407–1411. [DOI] [PubMed] [Google Scholar]
- 15.Roth JA, Putnam JB Jr., Wesley MN, Rosenberg SA. Differing determinants of prognosis following resection of pulmonary metastases from osteogenic and soft tissue sarcoma patients. Cancer. 1985;55(6): 1361–1366. [DOI] [PubMed] [Google Scholar]
- 16.Mentzer SJ, Antman KH, Attinger C, et al. Selected benefits of thoracotomy and chemotherapy for sarcoma metastatic to the lung. J Surg Oncol. 1993;53(1):54–59. [DOI] [PubMed] [Google Scholar]
- 17.Saltzman DA, Snyder CL, Ferrell KL, et al. Aggressive metastasectomy for pulmonic sarcomatous metastases: a follow-up study. Am J Surg. 1993;166(5):543–547. [DOI] [PubMed] [Google Scholar]
- 18.Choong PF, Pritchard DJ, Rock MG, et al. Survival after pulmonary metastasectomy in soft tissue sarcoma. Prognostic factors in 214 patients. Acta Orthop Scand. 1995;66(6): 561–568. [DOI] [PubMed] [Google Scholar]
- 19.Nakamura T, Matsumine A, Yamakado K, et al. Lung radiofrequency ablation in patients with pulmonary metastases from musculoskeletal sarcomas [corrected]. Cancer. 2009;115(16):3774–3781. [DOI] [PubMed] [Google Scholar]
- 20.Navarria P, Ascolese AM, Cozzi L, et al. Stereotactic body radiation therapy for lung metastases from soft tissue sarcoma. Eur J Cancer. 2015;51(5):668–674. [DOI] [PubMed] [Google Scholar]
- 21.Dhakal S, Corbin KS, Milano MT, et al. Stereotactic body radiotherapy for pulmonary metastases from soft-tissue sarcomas: excellent local lesion control and improved patient survival. Int J Radiat Oncol Biol Phys. 2012;82(2):940–945. [DOI] [PubMed] [Google Scholar]
- 22.National Comprehensive Cancer Network (U.S.). The complete library of NCCN oncology practice guidelines. In: 2000 ed. Rockledge, PA: NCCN,; 2000. [Google Scholar]
- 23.Patel DB, Matcuk GR Jr. Imaging of soft tissue sarcomas. Chinese clinical oncology. 2018;7(4):35. [DOI] [PubMed] [Google Scholar]
- 24.Peloschek P, Sailer J, Weber M, et al. Pulmonary nodules: sensitivity of maximum intensity projection versus that of volume rendering of 3D multidetector CT data. Radiology. 2007;243(2):561–569. [DOI] [PubMed] [Google Scholar]
- 25.Rissing S, Rougraff BT, Davis K. Indeterminate pulmonary nodules in patients with sarcoma affect survival. Clinical orthopaedics and related research. 2007;459:118–121. [DOI] [PubMed] [Google Scholar]
- 26.Marten K, Engelke C, Seyfarth T, et al. Computer-aided detection of pulmonary nodules: influence of nodule characteristics on detection performance. Clinical radiology. 2005;60(2): 196–206. [DOI] [PubMed] [Google Scholar]
- 27.Kane JM 3rd. Surveillance strategies for patients following surgical resection of soft tissue sarcomas. Curr Opin Oncol. 2004;16(4):328–332. [DOI] [PubMed] [Google Scholar]
- 28.Miller BJ, Carmody Soni EE, Reith JD, et al. CT scans for pulmonary surveillance may be overused in lower-grade sarcoma. The Iowa orthopaedic journal. 2012;32:28–34. [PMC free article] [PubMed] [Google Scholar]
- 29.Chou YS, Liu CY, Chen WM, et al. Follow-up after primary treatment of soft tissue sarcoma of extremities: impact of frequency of follow-up imaging on disease-specific survival. J Surg Oncol. 2012;106(2):155–161. [DOI] [PubMed] [Google Scholar]
- 30.Puri A, Gulia A, Hawaldar R, et al. Does intensity of surveillance affect survival after surgery for sarcomas? Results of a randomized noninferiority trial. Clinical orthopaedics and related research. 2014;472(5):1568–1575. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Johnson FE, Sakata K, Sarkar S, et al. Patient surveillance after treatment for soft-tissue sarcoma. Int J Oncol. 2011;38(1):233–239. [PubMed] [Google Scholar]
- 32.Whooley BP, Gibbs JF, Mooney MM, et al. Primary extremity sarcoma: what is the appropriate follow-up? Ann Surg Oncol. 2000;7(1):9–14. [DOI] [PubMed] [Google Scholar]
- 33.Brennan MF, Antonescu CR, Moraco N, Singer S. Lessons learned from the study of 10,000 patients with soft tissue sarcoma. Ann Surg. 2014;260(3):416–421; discussion 421–412. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Tseng WW, Amini B, Madewell JE. Follow-up of the soft tissue sarcoma patient. J Surg Oncol. 2015;111(5):641–645. [DOI] [PubMed] [Google Scholar]
- 35.Goel A, Christy ME, Virgo KS, et al. Costs of follow-up after potentially curative treatment for extremity soft-tissue sarcoma. Int J Oncol. 2004;25(2):429–435. [PubMed] [Google Scholar]
- 36.Sakata K, Johnson FE, Beitler AL, et al. Extremity soft tissue sarcoma patient follow-up: tumor grade and size affect surveillance strategies after potentially curative surgery. Int J Oncol. 2003;22(6):1335–1343. [PubMed] [Google Scholar]
- 37.Italiano A, Mathoulin-Pelissier S, Cesne AL, et al. Trends in survival for patients with metastatic soft-tissue sarcoma. Cancer. 2011;117(5):1049–1054. [DOI] [PubMed] [Google Scholar]