Background
Soft tissue sarcomas are a rare, heterogeneous group of malignancies, accounting for only 1% of all cancer types. Reference Clark, Fisher, Judson and Thomas1 Up to 15–20% of sarcomas arise from the retroperitoneum. Reference Clark, Fisher, Judson and Thomas1,Reference Brennan, Antonescu, Moraco and Singer2 The most common recurrence pattern is locoregional failure and occurs in 25–50% of patients, despite complete surgical resection. Reference Chouliaras, Senehi and Ethun3–Reference Singer, Antonescu, Riedel and Brennan5 Although a previous prospective trial showed no survival benefit with preoperative radiation therapy, Reference Bonvalot, Gronchi and Le Pechoux6 prior retrospective studies showed that those receiving preoperative radiation therapy were more likely to have negative surgical margin and improved survival. Reference Ma, Oladeru, Farrugia, Shekher, Iovoli and Singh7,Reference Nussbaum, Speicher, Gulack, Ganapathi, Englum and Kirsch8 The National Comprehensive Cancer Network (NCCN) guidelines currently recommend the consideration of neoadjuvant therapies, such as chemotherapy (doxorubicin and ifosfamide among others) and radiation (external beam radiation therapy with 50 Gy, with consideration of intraoperative radiation therapy [IORT] or simultaneous integrated boost in select cases), at the discretion of clinicians. 9 However, given the rare incidence of retroperitoneal sarcoma, the comparison of various neoadjuvant therapies has not been evaluated prospectively. To address this knowledge gap, we performed a retrospective, observational cohort study to compare preoperative chemotherapy, radiation and chemoradiation using a nationwide clinical oncology database.
Methods
Our study was approved by institutional review board at the Roswell Park Comprehensive Cancer Center (BDR-131220). It also follows the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline. Reference von Elm, Altman and Egger10
The National Cancer Database (NCDB) Reference Bilimoria, Stewart, Winchester and Ko11 is a nationwide clinical oncology database that captures more than 70% of newly diagnosed cancer cases in the United States and is jointly sponsored by the American College of Surgeons and the American Cancer Society. The NCDB is accessible to investigators from the Commission of Cancer-accredited institutions after obtaining an approval from the American College of Surgeons and the American Cancer Society. The database was queried for patients diagnosed between 2006 and 2015 with non-metastatic, resectable retroperitoneal sarcoma (RPS) treated with neoadjuvant therapies followed by surgery. Follow-up was until the end of 2017. Variables of interest included facility type (academic versus non-academic), facility volume (low, intermediate or high volume), age, gender, Charlson-Deyo comorbidity score (CDS), income, insurance, histology, tumour grade, year of diagnosis, T and N staging, surgery types, surgical margin, postoperative readmission, and postoperative inpatient duration. Income levels and insurance status were included for analysis since they were previously shown to be associated with survival outcomes. Reference Penumarthy, Goldsby, Shiboski, Wustrack, Murphy and Winestone12 Those with metastatic disease, surgery alone and unknown receipt of neoadjuvant therapy were excluded.
The primary endpoint was overall survival (OS), defined as the time interval between diagnosis and the last follow-up or death. OS was evaluated using Kaplan–Meier method, log-rank test and Cox regression multivariable analysis. Interaction term analysis was performed to evaluate differences in treatment effects with respect to facility type, facility volume and histology. Reference Barraclough and Govindan13 Logistic regression multivariable analysis was also performed to evaluate the association of variables with surgical margin status.
To reduce selection bias, propensity score matching was performed based on the nearest neighbour method in a 1:1 ratio without a replacement and with a caliper distance of 0·1 of the standard deviation of the logit of the propensity score. Reference Austin14 The standardised differences of variables were less than 0·1, indicating adequate match. Reference Haukoos and Lewis15 To address immortal time bias, those who survived less than 6 months after diagnosis were excluded as a conditional landmark. Analyses were performed with R-software version 3.6.1 (R Project for Statistical Computing). Drs Ma and Singh had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Results
A total of 1,253 patients with a median age of 60 years (interquartile range [IQR] 51–69) met our inclusion criteria. Of those, 210 patients (17%) received chemoradiation, 850 patients (68%) received radiation and 193 patients (15%) received chemotherapy. The median follow-up was 36·1 months (IQR 19·9–63·9). For those who underwent radiation therapy, its median total dose was 50·0 Gy (IQR 48·8–50·4). When compared between those treated with chemoradiation or radiation therapy, they both received comparable radiation doses (p = 0·40). On multivariable analysis, when compared to preoperative chemoradiation, preoperative radiation was not associated with improved OS (hazards ratio [HR] 0·98, 95% CI 0·76–1·25, p = 0·84), while preoperative chemotherapy was associated with worse OS (HR 1·64, 95% CI 1·24–2·18, p < 0·001).
On Cox multivariable analysis, other factors associated with worse OS were recent years of diagnosis (2011–15 versus 2006–10; HR 1·30, 95% CI 1·08-1·58, p = 0·007), elderly age (≥ 65 versus <65; HR 1·32, 95% CI 1·10-1·59, p = 0·003), male gender (HR 1·20, 95% CI 1·00-1·43, p = 0·047), lower income (HR 1·19, 95% CI 1·00-1·43, p = 0·049), poorly differentiated (HR 2·53, 95% CI 1·84-3·48, p < 0·001) or other tumour grade (HR 2·44, 95% CI 1·75-3·40, p < 0·001), and positive surgical margin (HR 1·37, 95% CI 1·12-1·68, p = 0·002). No interaction of preoperative therapy was observed with facility type (interaction p = 0·08), facility volume (interaction p = 0·89) or histology (interaction p = 0·49). On logistic multivariable analysis, histology was the only factor associated with surgical margin status (low grade liposarcoma, not otherwise specified: odds ratio [OR] 2·93, 95% CI 1·93-4·46, p < 0·001; intermediate grade liposarcoma: OR 0·37, 95% CI 0·15-0·78, p = 0·02; high-grade liposarcoma: OR 2·62, 95% CI 1·82-3·78, p < 0·001) when compared to leiomyosarcoma. After propensity score matching, all baseline characteristics were well balanced (Table 1). Similar findings were observed in 199 matched pairs for preoperative radiation (HR 1·00, 95% CI 0·74-1·36, p = 0·98; Figure 1) and 128 matched pairs for preoperative chemotherapy (HR 1·51, 95% CI 1·04-2·19, p = 0·03; Figure 2) when compared to preoperative chemoradiation.
Baseline characteristics for cohorts after matching
Abbreviations: NOS, Not otherwise specified.
Kaplan–Meier survival curves for preoperative chemoradiation versus radiation.
Preop: preoperative; CRT: chemoradiation; RT: radiation.
Kaplan–Meier survival curves for preoperative chemoradiation versus chemotherapy
Preop: preoperative; CRT: chemoradiation; CT: chemotherapy.
Discussion
This is the first report to compare preoperative chemotherapy, radiation and chemoradiation using a national oncology database. While preoperative radiation and chemoradiation had comparable survival outcomes, preoperative chemotherapy was associated with worse survival outcomes compared to chemoradiation.
Several single-arm prospective trials involving approximately 30–80 patients have investigated chemotherapy as a radiosensitiser and showed preoperative chemoradiation (up to 50·4 Gy with either ifosfamide or doxorubicin) to be feasible in select patients with the majority of patients receiving grossly complete surgical resection. Reference Gronchi, De Paoli and Dani16,Reference Pisters, Ballo and Fenstermacher17 However, although survival outcomes from the prospective trial were consistent with our study, up to one-third of patients from the trial could not complete preoperative chemoradiation due to toxicity and local failure still occurred in over 40% of cases. Reference De Sanctis, Giordano and Colombo18 Significant locoregional failure despite the use of chemoradiation may explain a lack of OS benefits seen with preoperative chemoradiation in our study.
Worse mortality with preoperative chemotherapy in our study is consistent with a prior NCDB report that included 163 patients treated with chemotherapy. Reference Miura, Charlson and Gamblin19 This observation may be likely due to mortality secondary to locoregional, rather than distant recurrences. Reference Singer, Antonescu, Riedel and Brennan5,Reference Linehan, Lewis, Leung and Brennan20–Reference Stojadinovic, Yeh and Brennan22 Several multimodality approaches involving chemotherapy (ifosfamide, doxorubicin and etoposide) were investigated, such as regional hyperthermia with favourable local control and survival outcomes, in prospective trials, one of which was a phase III trial involving 341 patients. Reference Issels, Lindner and Verweij23,Reference Wendtner, Abdel-Rahman and Krych24 However, the use of hyperthermia was not available in the NCDB for comparison. In our study, intermediate grade liposarcoma was associated with negative margin, which was consistent with literature suggesting that myxoid and round cell liposarcoma are sensitive to radiation and chemotherapy. Reference Chowdhry, Goldberg and DeLaney25–Reference Pitson, Robinson and Wilke28 Ongoing trials, such as Surgery With Or Without Neoadjuvant Chemotherapy in High-Risk Retroperitoneal Sarcoma II trial (STRASS II; NCT04031677), are further investigating the role of preoperative chemotherapy.
Limitations of our study are related to the retrospective nature of the NCDB. In addition, clinically relevant factors, such as performance status, chemotherapy agents and toxicity profiles, were not captured in the NCDB. Unmeasured confounding and selection bias may be present despite propensity score matching. In our analysis, postoperative readmissions and duration of postoperative inpatient admission were also matched as proxy measures for postoperative performance status. Reference Stitzenberg, Chang, Smith and Nielsen29 Our findings also may not be generalisable to other populations that were not included in the NCDB.
While our report suggests radiation may be an integral part of preoperative modalities, further studies are warranted to optimise preoperative therapies to improve local control and survival outcomes among patients with retroperitoneal sarcoma. For example, the role of intensity-modulated proton therapy and simultaneous integrated boost to high-risk tumour regions is being currently investigated in an ongoing trial. Reference DeLaney, Chen and Baldini30 In addition, given favourable outcomes from incorporating IORT, Reference Pisters, Ballo and Fenstermacher17,Reference Alektiar, Hu, Anderson, Brennan and Harrison31,Reference Yoon, Chen and Kirsch32 the European Society of Radiotherapy and Oncology (ESTRO) Reference Roeder, Morillo, Saleh-Ebrahimi, Calvo, Poortmans and Ferrer Albiach33 and the NCCN 9 guidelines both recommend the consideration of IORT in select patients. Ultimately, systematic multidisciplinary discussions are important to tailor treatment options based on individual patient and tumour characteristics, institutional areas of expertise, and shared decision-making with patients in the context of their treatment goals. Reference Blay, Soibinet and Penel34
Conclusion
Our study suggested an OS benefit in using preoperative chemoradiation compared to chemotherapy alone, but OS outcomes were comparable between preoperative chemoradiation and radiation alone.
Acknowledgements
The National Cancer Database is a joint project of the Commission on Cancer of the American College of Surgeons and the American Cancer Society. The data used in the study are derived from a de-identified NCDB file. The American College of Surgeons and the Commission on Cancer have not verified and are not responsible for the analytic or statistical methodology employed, or the conclusions drawn from these data by the investigator.
Financial Support
Dr Singh is currently supported by the National Cancer Institute (grant number 5P30CA016056–42) Cancer Center Support Grant.
Conflicts of Interest
The authors declare none.
Ethical Standards
The authors assert that all procedures contributing to this work comply with the ethical standards of the relevant national guidelines on human experimentation and with the Helsinki Declaration of 1975, as revised in 2008, and have been approved by Clinical Research Services and Office and Office of Research Subject Protection at Roswell Park Cancer Institute. The authors assert that all procedures contributing to this work comply with the ethical standards of the relevant national guides on the care and have been approved by Clinical Research Services and Office and Office of Research Subject Protection at Roswell Park Cancer Institute under the study number STUDY00000621/BDR 099918.
Availability of Data and Materials
The data that support the findings of this study are available from the Commission on Cancer of the American College of Surgeons and the American Cancer Society. The data are publicly accessible by any investigator affiliated with the Fellow of the American College of Surgeons applying to gain access. More information can be found at http://ncdbpuf.facs.org/.
