Despite the relatively poor prognosis of LMS, one of the more common soft-tissue sarcoma subtypes, there is a paucity of data describing the metastatic pattern, prognostic factors, and efficacy of specific staging or surveillance imaging modalities [10–13]. We therefore sought to answer three questions: (1) What is the anatomical distribution and frequency of metastatic disease in truncal/extremity LMS? (2) Are factors such as age at primary diagnosis, tumor grade, tumor size, and primary tumor location associated with metastatic risk or overall survival after primary diagnosis of truncal/extremity LMS? (3) Is imaging modality associated with a greater frequency of metastatic disease detection? Here, we show that truncal/extremity LMS has a high rate of metastasis, with significantly higher rates in older patients and those with higher grade tumors at initial diagnosis. We show that truncal/extremity LMS has a high rate of metastases to extra-pulmonary sites, and the use of PET/CT imaging, both at initial staging and throughout surveillance, was associated with a greater rate of metastatic tumor detection.
This study has several limitations, including its retrospective nature and non-randomized design. As a retrospective study, we were reliant on the accuracy of patient records. Further, as we collected data from our single cancer institution, bias in patient referral should be considered. As a retrospective, non-randomized investigation, this study was not designed to parse out the utility of each imaging modality for surveillance or progression of disease/response to treatment. The use of imaging modalities is often influenced factors such as patient insurance. Also, this study was not designed to evaluate how the type and frequency of imaging studies impact clinical outcomes such as survival and quality of life.
In this retrospective analysis, we identified 73 patients with truncal or extremity LMS. Consistent with the findings of others, the lung was the most common metastatic site [6, 7], with a relatively high rate of metastatic disease arising in the abdomen, thorax, visceral organs, bone, and skin or soft tissues. We show that grade III tumors are significantly more likely to metastasize to the lungs compared to low-grade tumors, with an odds ratio of 9.74 (p = 0.046). The 5-year overall survival rate in this cohort was 59%, while Shoushtari et al. reported a 5-year survival rate of under 25% in an analysis of only metastatic LMS`, and Lamm et al. reported a 5-year survival rate of 44.4% [6, 7]. The lower 5-year survival rates previously reported likely represent the difference in proportions of high vs low-grade LMS. The largest study focusing on LMS to date, conducted by Shoushtari and colleagues in 2016, provides a description of the overall survival and response to systemic therapy in extrauterine metastatic LMS [6] but lacks a precise description of metastatic pattern. Further, in 2014, Lamm et al. compared uterine to non-uterine LMS, showing that lungs are the most common metastatic site in both uterine and non-uterine LMS, with initial metastatic disease serving as a prognostic factor for overall survival [7]. More recently, Lee et al. compared the response to radiation treatment of truncal/extremity LMS versus non-LMS soft-tissue sarcomas. Lee and colleagues describe and similar average age and tumor size of 63 years and 6.0 cm, respectively [11]. Similarly, Gladdy et al. describe a median age of 57 years, and average tumor size of 6.0 cm and identified high grade tumors as predictive of disease-specific survival [10].
In this study, 10.0% of patients were classified as grade I, compared to 3% and 2% in the studies by Shoushtari et al. and Lamm et al [6, 7]. In our analysis of metastatic prognosticators, we identified high-grade tumors (II/III) and patients over the age of 50 showed significantly greater rates of metastasis, consistent with current reports of ULMS and NULMS [4–7]. We also show that tumor size is not associated with metastatic risk (OR = 1.04, p = 0.515), and primary tumor location, either flank/pelvis or chest wall/spine, does not influence overall survival (HR = 1.37, p = 0.412, HR = 1.73, p = 0.174, respectively).
Effective detection, diagnosis, and surveillance are vital in the treatment and management of LMS. Analyses of efficacy of radiological staging and surveillance modalities for LMS are also sorely needed. Here, we found a significantly worse overall survival in patients that underwent more frequent radiological surveillance. We believe this represents positive clinical decision-making, in that patients that were deemed to have more aggressive tumors at diagnosis subsequently underwent more frequent surveillance. While Chest CT with or without contrast remains the benchmark for assessing lung metastases, there is conflicting evidence for obtaining an abdomen/pelvis CT for staging [9]. Reports of incidence rates of metastatic disease in the abdomen or pelvis vary – one report from a single institution suggests a 16.0% incidence rate [14], supporting routine abdomen/pelvis CT, while another retrospective review reported only a 2.9% incidence rate; this would argue against routine use of abdomen/pelvis CT for staging and monitoring in the setting of soft-tissue sarcoma of the extremity [15]. There is also a growing role for PET/CT for staging, surveillance, and gauging treatment response of soft-tissue sarcomas [16–18]. In two studies of LMS, tumor 18F-FDG uptake, as measured by the maximum standardized uptake value (SUVmax) was a powerful prognostic factor for overall survival correlating with tumor grade and size [17, 19]. In this study, we show that the risk of metastatic disease to any extra-pulmonary area after primary LMS diagnosis occurred in 36/73 patients (49.3%). Of those 50 patients that developed metastatic disease, 35 (72.0%) developed metastatic sites outside of the lungs, supporting the use of imaging that extends beyond routine chest CT. Despite the potential strength of PET/CT imaging, there have been relatively few reported series, with low case numbers to justify the routine application of PET/CT imaging of LMS. In this study, we provide the largest comparison known to date of metastatic tumor detection between PET/CT imaging and CT CAP in a cohort of patients with truncal/extremity LMS.
In comparing the rate of tumor detection between the two most common and comparable imaging modalities, CT CAP and PET/CT, we found that PET/CT imaging detected significantly more tumors both at initial staging (Chi-square = 4.7, p = 0.03), and throughout surveillance/treatment response (Chi-square = 11.32, p < 0.001). In this study, we show that metastases to extra-pulmonary sites are relatively frequent, suggesting a particular important utility for PET/CT imaging in detecting metastatic disease at sites that are not readily detected on conventional CT imaging. In particular, PET/CT imaging had a greater rate of detecting tumors in the abdomen/visceral organs, and skin/soft tissue – areas that might be missed using CT Chest/Abdomen/Pelvis imaging. However, caution is warranted as Hensley et al. have suggested that PET has not been shown to be superior for staging of uterine and ovarian LMS, and may miss small volume lung metastatic tumors, often necessitation chest CT imaging in conjunction with whole body PET/CT imaging [20]. Moreover, our study was not designed to parse out the utility of each imaging modality for surveillance or progression of disease/response to treatment. The use of PET/CT versus CT CAP was not always dictated by care algorithms. Further investigations are needed to elucidate the clinical benefit of specific staging and surveillance techniques and timing.