Relevance of Reference Centers in Sarcoma Care and Quality Item Evaluation: Results from the Prospective Registry of the Spanish Group for Research in Sarcoma (GEIS)

Management of soft‐tissue sarcoma is challenging. This article reports on sarcoma clinical management in cancer centers in Spain, based on information from a prospective registry launched by the Spanish Group for Research in Sarcoma.

Abstract

Background.

Reference centers (RCs) are a key point for improving the survival of patients with soft‐tissue sarcomas (STS). The aim of this study was to evaluate selected items in the management of patients with STS, comparing results between RC and local hospitals (LHs).

Materials and Methods.

Diagnostic and therapeutic data from patients diagnosed between January 2004 and December 2011 were collected. Correlation with outcome was performed.

Results.

A total of 622 sarcomas were analyzed, with a median follow‐up of 40 months. Imaging of primary tumor preoperatively (yes vs. no) correlated with a higher probability of free surgical margins (77.4% versus 53.7%; p = .006). The provenance of the biopsy (RC vs. LH) significantly affected relapse‐free survival (RFS; 3‐year RFS 66% vs. 46%, respectively; p = .019). Likewise, 3‐year RFS was significantly worse in cases with infiltrated (55.6%) or unknown (43.4%) microscopic surgical margins compared with free margins (63.6%; p < .001). Patients managed by RCs had a better 3‐year overall survival compared with those managed by LHs (82% vs. 70.4%, respectively; p = .003). Perioperative chemotherapy in high‐risk STS, more frequently administered in RCs than in LHs, resulted in significantly better 3‐year RFS (66% vs. 44%; p = .011). In addition, patients with stage IV disease treated in RCs survived significantly longer compared with those in LHs (30.4 months vs. 18.5 months; p = .036).

Conclusion.

Our series indicate that selected quality‐of‐care items were accomplished better by RCs over LHs, all with significant prognostic value in patients with STS. Early referral to an RC should be mandatory if the aim is to improve the survival of patients with STS.

Implications for Practice.

This prospective study in patients diagnosed with soft‐tissue sarcoma shows the prognostic impact of reference centers in the management of these patients. The magnitude of this impact encompasses all steps of the process, from the initial management (performing diagnostic biopsy) to the advanced disease setting. This is the first prospective evidence showing improvement in outcomes of patients with metastatic disease when they are managed in centers with expertise. This study provides extra data supporting referral of patients with sarcoma to reference centers.

Introduction

The term “soft‐tissue sarcoma” (STS) encompasses a heterogeneous group of more than 40 different tumor subtypes, accounting for less than 2% of all solid tumors in adults, with an estimated incidence of five new cases per 100,000 per year in Europe [1]. These malignancies can arise in any part of the body, the most frequent sites being limbs (50%–60%) and retroperitoneum (about 15%) [2], [3]. Leiomyosarcomas, undifferentiated pleomorphic sarcomas, and liposarcomas are the most frequent sarcoma subtypes [4]. About 70%–80% of sarcomas are diagnosed in a localized stage, in which the curative intent is the main aim of therapy. Given the heterogeneity and low incidence of this malignancy, its management must be carried out in multidisciplinary teams within reference centers (RCs) [5]. The initial management in a center of expertise has consistently been shown to have an impact on the outcome of these patients in terms both of local control and of survival [6], [7]. As a tool to guide specialists treating patients with sarcoma, several clinical practice guidelines (CPGs) for STS have been developed, their paramount aim being to improve the quality in diagnostic and therapeutic management of patients with STS. In Spain, CPGs for STS were agreed on by consensus and approved for the first time by the Spanish Group for Research in Sarcoma (GEIS) in 2004, and since that date these have been updated periodically [8]. Assessing quality in the therapeutic process of patients with sarcoma is challenging. Regional or national registries could be useful in the quest to identify weaknesses and areas for improvement. Colleagues from the Scandinavian Sarcoma Group are pioneers in this kind of quality assessments, and the Scandinavian register has collected clinical data from patients with sarcoma treated in cancer centers from Norway, Sweden, and Finland from 1986 onwards, assessing the surgical quality, the referral patterns, and the outcome of these patients [9]. The registry has also been useful in assessing the impact of educative programs (regarding early or delayed referral to RCs) and the changes in surgical approach (number of amputations) and outcome over time with the introduction of adjuvant therapies. A population‐based observational study, which ran prospectively in two southwestern French regions in 2006 and 2007, explored compliance with CPGs in the management of patients with sarcoma. Although the preoperative imaging tests for staging were in general adequate, this study identified up to 50% of patients without proper histopathological preoperative diagnosis, especially in those patients managed outside RCs [10]. Likewise, the assessment of compliance with CPGs through medical audit in one particular French region showed a significantly higher conformity to CPGs, as well as better clinical results, if the patients were treated in a center with a multidisciplinary sarcoma committee [11]. Conformity to CPGs was especially low in STS (36%) compared with other types of sarcomas such as gastrointestinal stromal tumors (54.3%), a STS subtype with different biology and management, or bone sarcomas (42.4%). Analyses from the Netherlands Cancer Registry found that high‐volume hospitals had macroscopic residual disease (R2) after surgery in somatic STS less often than low‐volume hospitals [12]. Moreover, a different analysis of conformity to sarcoma CPGs in The Netherlands revealed significantly better adherence in specialized centers than in district hospitals [13].

In Spain, GEIS designed a registry program as a tool to evaluate the compliance of some quality items in the diagnostic and therapeutic processes of somatic STS. In addition, plans were set in place to assess the impact of hospital type (reference center vs. local hospital) in clinical outcomes for the period 2004–2011.

Materials and Methods

A web‐based registry was designed to prospectively collect diagnostic and therapeutic processes in patients diagnosed with STS in Spain between January 2004 and December 2011. Thirty‐one hospitals (RCs and local hospitals [LHs]) were involved in this program nationwide. Physicians or specifically trained staff filled in the registry, providing data related to demographics, tumor characteristics, diagnosis, treatment, and follow‐up from patients with localized STS. In detail, data fields included information regarding demographics (gender, Karnofsky score at diagnosis), specific dates (appearance of first symptom or sign, performance of biopsy, local pathologic report, imaging test, different treatments, recurrence, death, and last follow‐up), diagnosis (imaging tests for primary tumor location and metastasis workup, sarcoma subtype, grade, size), type of biopsy (excisional, incisional, core biopsy, or fine needle aspiration [FNA]) and therapeutic approach (type of surgical resection, status of surgical margins, perioperative treatments). We defined RCs as those hospitals that had a multidisciplinary team experienced in sarcoma care, along with a weekly operative sarcoma board committee, a minimum number of 70 patients with STS managed per year, and at the very least a defined regional referral policy in relation to nearby local hospitals (those not fulfilling these criteria) and primary health care. For the comparative analysis of relapse‐free survival (RFS) in relation to biopsy provenance, only data originating from the Hospital Son Espases (HSE) were taken into account because, as a pilot experience, this hospital had collected data concerning the particular medical team (LH or RC) in charge of biopsy procedure and performance of the first surgery. This made it possible to focus on the differences in clinical outcome between patients who had been referred after excisional biopsy and those who had been referred to the RC before biopsy.

A query‐based task was used for remote data cleaning, and on‐site monitoring was carried out for leading recruiting centers. A minimum data set for selecting patients was established; it included the following data: diagnosis (biopsy type, biopsy date, sarcoma subtype, grade), staging procedures (date and type of different imaging tests), treatment (date and type of different treatments), treatment outcome (margins status, response), and follow‐up (date of relapse and death if applicable, date of last follow‐up, status of last follow‐up). For assessment of diagnostic and therapeutic adequacy, only data coming from patients with localized STS arising in extremities (EE) or trunk wall (TW) were eligible. Patients had to be diagnosed within the 2004–2011 period. The items selected to assess compliance with CPG were type of biopsy, percentage of biopsies carried out in local hospitals (in the HSE reference center region), imaging tests for primary tumor before surgery, tumors larger than 5 cm initially resected elsewhere than a reference center, surgical margins, percentage of re‐resections, and the use of adjuvant treatments. The quality items selected to assess variations between the first period (1994–2000) and the current second period (2004–2011) were percentage of adequate biopsies (defined as core biopsy or incisional in tumors >5 cm) and positive margins in the first surgical resection. Statistical analysis was performed using SPSS (IBM Corporation, Armonk, NY). Comparison between categorical variables was made with the chi‐square or Fisher's exact test as appropriate. For time‐to‐event variables (i.e., RFS, overall survival [OS]), the Kaplan‐Meier method was used, and the log‐rank test was employed to compare groups. Multivariate analysis with the variables that proved to be significant in univariate analysis was performed according to the Cox proportional hazard regression model. All reported p values were two sided, and statistical significance was defined at p < .05. Approval from the ethics committee was obtained for this study.

Results

A total of 1,470 patients with sarcoma were registered in the online registry after approval by ethics committees and institution review boards. A subset of 848 cases was excluded in the current analysis, for several reasons: 358 cases lacked essential data, 286 cases were diagnosed as visceral sarcomas, and 204 cases were diagnosed outside the time period of the study. A final number of 622 cases were thus considered for the current analyses.

With a median follow‐up of 40 months (range, 1–97), 95 patients (28.6%) with completely resected localized EE or TW sarcomas had relapsed, the actuarial 3‐ and 5‐year RFS rates being 75.5% and 72.1%, respectively. One hundred thirty‐five patients of the whole series had died, with actuarial 3‐ and 5‐year OS rates of 80.8% and 77%, respectively. The median of RFS or OS had not yet been reached.

Thirty‐one hospitals were involved in this registry program, and only two of them (HSE and Hospital Sant Pau [HSP]) met the predefined criteria for an RC. A total of 285 patients (46%) were initially or eventually managed by a RC, whereas 337 patients (54%) were managed elsewhere (local hospital). Patients with STS located in limbs or trunk wall accounted for 70%, and high‐grade (grade 2 and 3) STS made up 59% of the series (Table 1). The main characteristics of the analyzed patients were comparable to the whole series (patients were excluded for the above reasons; supplemental online Table 1). The most frequent sarcoma subtypes were liposarcomas (21.6%), undifferentiated pleomorphic sarcoma (17.4%), leiomyosarcoma (11.7%), and synovial sarcoma (7.8%). With regard to diagnostic procedures (Table 2), excisional biopsy was the most frequent diagnostic method for EE and TW sarcomas (46% of procedures). Cases with incisional and core biopsies, considered appropriate methods for sarcoma diagnosis, made up 44% of the series. Excisional biopsy was the most frequent diagnostic method, even in tumors larger than 5 cm (excisional, 37%, vs. core biopsy, 35%) or in deep tumors (excisional, 39%, vs. core biopsy, 33%).

Table 1. Clinical and tumor characteristics (somatic sarcoma, n = 622).

Table 2. Diagnostic and therapeutic procedures.

Abbreviations: FNA, fine needle aspiration; HSE, Hospital Son Espases; MRI, magnetic resonance imaging.

With respect to those patients registered in HSE, 59 (33%) underwent biopsy in a local hospital, and 120 (67%) were referred before any biopsy procedure to an RC. Similarly, focusing on the site of first surgery for HSE, 54 (30%) of cases were resected in local hospitals and 125 (70%) in the RC. With regard to microscopic surgical margins of the series, the percentage distributions for negative, positive, and unknown were 38%, 24%, and 38%, respectively.

Correlation Between Adequacy of Biopsy and Staging with Surgical Margins

The probability of microscopic positive surgical margins was significantly associated with biopsy type; these were higher (36%) in excisional procedures than in core biopsies (11%; p < .001). FNA is not recommended as a diagnostic procedure in the GEIS CPG for soft tissue sarcoma; however, it represents 8% of procedures, and it is related to twice the number of cases of positive margins (22%) than core biopsies. The percentage of unknown surgical microscopic margins, according to local pathologic reports, is as high as 32% (Table 3).

Table 3. Correlation between type of biopsy with surgical margin status.

p ≤ .001.

^aThirteen cases lacked information about type of biopsy.

Abbreviation: FNA, fine needle aspiration.

Variations over time of biopsy types, comparing the first (1994–2000) and second (2004–2011) registry programs in the diagnostic method for cases larger than 5 cm and located in limbs or trunk wall, were as follows: excisional biopsies changed from 50% to 32%; incisional from 28% to 21%; core biopsy from 11% to 39%, and FNA from 11% to 9%. However, this reduction of excisional biopsies had hardly any influence in obtaining a lower percentage of positive surgical margins. For the whole somatic series, the variation over time of excisional biopsies was less marked; this is probably related to the fact that a high percentage of excisional biopsies is in retroperitoneal sarcomas (higher than 65%; Table 4).

Table 4. Surgical margin and excisional biopsy variations across first and second registry programs.

It is noteworthy that cases with magnetic resonance imaging (MRI) or computed tomography (CT) scans of primary tumor, carried out before the first surgery, were also significantly related to a higher probability of free microscopic surgical margins: 77.4%, compared with 53.7% in patients without presurgical CT or MRI of primary tumors (p = .006). Focusing on limb sarcomas, 75% of cases managed by RCs had documented MRI of primary tumors, compared with only 19% for cases managed in local hospitals.

Clinical Impact of Biopsy Provenance and Surgical Margins

We found a significant correlation between the type of hospital in charge of the biopsy (RC or LH) and the status of the surgical margin. Focusing on the subset of 173 patients managed in HSE, when the biopsy was performed in a local hospital, there were 42.4% of patients with positive surgical margins in the first surgery, whereas this figure decreased to 21.3% when the biopsy was carried out in the RC, HSE (p = .002). Importantly, the actuarial 3‐year RFS was significantly better in patients biopsied in the RC, 66% (95% confidence interval [CI], 56.1%–75.9%), compared with patients biopsied in a local center, 46.4% (95% CI, 31.9%–60.9%; p = .019; Fig. 1A).

Figure 1.

Survival curves in accordance to center in charge of biopsy and margin status. (A): Kaplan‐Meier curves for relapse‐free survival (RFS) that compare biopsy provenance (reference centers vs. local hospitals) in cases registered by Hospital Son Espases. (B): Curves for RFS in the whole series according to surgical margins. (C) and (D): Curves for RFS (C) and overall survival (D) in the cases registered in reference centers according to surgical margins.

The actuarial 3‐year RFS in regard to surgical microscopic margins at first surgery was as follows: 63.6% (95% CI, 56.8%–70.4%) for free margins, 55.6% (95% CI, 46.2%–65.0%) for infiltrated, and 43.4% (95% CI, 35.2%–51.6%) for unknown (p < .001; Fig. 1B). In this latter subset, the microscopic margin status was omitted in the surgical pathologic report.

Clinical Outcome According to Center (Reference or Local Center)

Cases treated with definitive surgery in local hospitals had similar features to those treated in RCs with respect to size, gender, and type of presentation (symptomatic or asymptomatic). However, in the two RCs, there were more grade 3 cases than in local hospitals (43% vs. 32%; p = .002).

As a general rule, both RCs protocoled re‐resection for cases with infiltrated surgical margins at first surgery. We therefore deemed it relevant to analyze the survival impact, if any, of margin status at the first surgery in the subset of 285 cases registered in RCs (HSE and HSP). With regard to RFS, infiltrated margins or unknown margins status was associated with a significantly worse 3‐year RFS, 52% (95% CI, 38%–66%) or 50% (95% CI, 38%–62%) respectively, compared with cases with free margins, 73% (95% CI, 64%–82%; p = .002; Fig. 1C). Furthermore, there were significant differences in overall survival as well; the actuarial 3‐year OS rates were 82% (95% CI, 71%–93%), 73% (95% CI, 63%–83%) and 90% (95% CI, 83%–97%) for infiltrated, unknown, and free surgical margins respectively (p = .003; Fig. 1D).

There was a clear trend towards a better median of RFS if the patient were referred to RC at any time: 63.3 months (95% CI, 44.9–81.7) compared with 39.6 months (95% CI, 19.8–59.4) for local center, p = .10. Moreover, the 3‐year actuarial overall survival was significantly better if the patient were managed in an RC rather than in a local center: 82% (95% CI, 74–90) versus 70.4% (95% CI, 64.7–76.1; p = .003; Fig. 2). Recurrences were seen in 44% and 56% of patients in RC and LH, respectively. The profile of such recurrences was clearly different between center types; there were relatively more local relapses in patients registered by RCs (60% of 44%) than in those registered by local hospitals (41.7% of 56%). The number of distant metastases was higher in local hospitals (40% of 56%) than in RC (24.7% of 44%; p = .058). The amount of coincident local and distant recurrences was similar in both subsets.

Figure 2.

Kaplan‐Meier curves for overall survival in relation to center type (reference center vs. local hospital).

No differences were detected in the distribution of Eastern Cooperative Oncology Group (ECOG) performance status between RCs and LHs. For RCs there were 83.5% and 16.5% patients with ECOG scores 0–1 and >1 respectively, and for LHs there were 86.6% and 13.4% patients with ECOG scores 0–1 and >1 respectively (p = .308).

The Impact of Radiation Therapy and Chemotherapy

We did not observe any difference in radiotherapy administration between RC and LH; in general, the compliance was higher than 80% in stage III STS. However, there was greater variability for chemotherapy. With respect to high‐risk localized STS in the series (limb sarcomas, tumors larger than 5 cm, high grade, and deep tumors), 76 patients received perioperative chemotherapy and 62 did not. There were significantly more patients receiving perioperative chemotherapy (67%) in RCs than in local hospitals (45%) in these high‐risk localized tumors (p = .01). The 3‐year RFS was significantly better for patients treated with perioperative chemotherapy: 66% (95% CI, 60%–72%) versus 44% (95% CI, 36%–52%; p = .011; Fig. 3A). In addition, a multivariate analysis was performed by including the tumor size (≤10 or >10 cm) in the model, along with the grade (2 vs. 3) and perioperative chemotherapy (yes vs. no). The only independent prognostic factor was perioperative chemotherapy, with a hazard ratio of 2.057 (95% CI, 1.164–3.636; p = .013). Chemotherapy consisted in a combination of anthracyclines (usually epirubicin 120 mg/m² per cycle) plus ifosfamide (usually 9 g/m² per cycle). Furthermore, in additional data from this registry program we analyzed whether the outcome of patients with metastases at diagnosis was different depending on the center type. The most noteworthy observation was that patients treated in RCs had a significantly better median of OS compared with those in LHs: 30.4 months (95% CI, 22.5–38.3) versus 18.5 months (95% CI 13.3–23.6; p = .036; Fig. 3B).

Figure 3.

Survival curves in accordance to perioperative chemotherapy administration (in localized disease) and center type (in advance disease). (A): Kaplan‐Meier curves for relapse‐free survival in patients with high‐risk localized soft‐tissue sarcomas of extremities according to perioperative chemotherapy administration. (B): Curves for overall survival in metastatic cases in relation to center type.

Discussion

We have presented the results of a prospective multicenter national registry program launched by GEIS, used as a tool to assess several quality criteria related to sarcoma care and its evolution over time in different hospitals of Spain.

Given that LHs and RCs were involved in this project, we were able to make comparisons between these two center categories. Importantly, as RCs frequently receive patients who have already been submitted to unplanned surgeries elsewhere, and as these patients are eventually registered by RCs, it is obvious that this fact would lead to the paradox that RCs will have a higher number of excisional biopsies and positive surgical margins. To circumvent this misleading statistic, in one RC, data concerning the center that performed the diagnostic biopsy were collected and monitored.

It is a concerning fact that excisional biopsy is still the most frequent biopsy type, even in large and deep tumors. This indicates that there were still a substantial number of unplanned surgeries that led to a reversal of the rational sequence of obtaining diagnosis first of all and subsequently designing the therapeutic plan, the latter being an issue to be addressed by a multidisciplinary team. In our series, 46% of somatic sarcomas were diagnosed through excisional biopsy; this figure compares unfavorably with other recent series that exhibit a range of 26%–45% of excisional biopsies [14], [15], [16]. Nevertheless, there was a trend towards improvement in the number of excisional biopsies in large tumors (>5 cm) of limbs and trunk wall; there were fewer of these over time, decreasing from 50% to 31% in our series.

The significant correlation between biopsy type and positivity of microscopic surgical margins is well known; the higher the percentage of excisional biopsy, the greater the probability of positive margins. In this sense, our study showed significant correlation; however, the rate of positive margins after excisional biopsy was extremely low in comparison with other series, which range from 85% to 100% [15], [17], [18]. This discordance could be explained, at least partially, by the prominent percentage of unknown microscopic surgical margins in our series. In fact, cases with unknown surgical margins had the worst prognosis, in terms of OS or RFS, even compared with cases of positive margins in our analyses. Data on margins were collected from surgical pathologic reports and were not really improvable by query tasks, because they were generated at the time of surgical specimen management by pathologists. In any case, in our country there is clearly room for improvement in the handling of surgical specimens as far as surgical margin status is concerned.

Compliance with CPGs for imaging tests, particularly MRI of primary tumor, was notably higher in RCs than in LHs (75% vs. 19%) in our series, something that has already been reported [9]. More importantly, performing imaging tests before surgery resulted in a significantly higher probability of free margins. These differences are clear even if many of cases registered by RCs were initially staged and resected in LHs and then referred, a circumstance that clearly penalizes the RC (our registry was not designed to distinguish the center in charge of performing the staging studies).

An outstanding piece of data coming from our analyses is that biopsy provenance has a significant impact on RFS. Biopsy execution prior to referral entailed a significantly worse actuarial 3‐year RFS compared with cases biopsied in RC. This is so despite that for those cases with positive surgical margins referred from LH, re‐surgery is a standard procedure in the RC involved in this analysis (HSE). Other authors too have found survival impact for cases referred after unplanned surgeries, with the prognostic impact being higher in cases with residual disease at the time of re‐excision [19], [20], [21], [22]. Data of microscopical residual disease at re‐excision were not analyzed in our series, so we cannot confirm this finding as critical for worse RFS. In series with shorter follow‐up, such as ours, the impact on OS could be unperceived. It is well recognized that microscopic surgical margin is a time‐dependent prognostic variable in localized STS. However, its prognostic value on survival has a later influence than other variables such as grade or size; that time‐dependent influence is detected in studies with a larger number of patients and longer follow‐up [23], [24]. This result reinforces the need to refer patients to the RC before any biopsy procedure takes place. The message that early referral is the key to success in sarcoma survival outcome is included in the CPG of GEIS, as well as in all European CPGs on sarcomas [5], [8], [25], [26], [27]. The first surgery is carried out with significantly better quality by an RC, where a multidisciplinary team definitively designs the best therapeutic strategy [7], [28], [29], [30].

The distribution of different relative recurrences between center types is critical to understanding, on one hand, that a substantial number of patients had been referred to RCs after a local recurrence or after an excisional biopsy. That understanding helps us to see that these events might skew the results for RC, because these kinds of patients were registered by an RC in the end, rather than by the original LH. In fact, worse RFS had previously been demonstrated for patients referred after biopsy performance. In addition, this phenomenon may explain why RCs have not reached a favorable statistically significant difference in RFS.

The role of perioperative chemotherapy has been a subject of debate in STS for decades. Our CPG recommends the use of chemotherapy (adjuvant or more importantly neoadjuvant) in high‐risk localized STS of limbs or trunk wall. The concept of high‐risk is important (high grade, tumor larger than 5 cm, and deep tumors), as well as the selection of highest dose intensity for the two most active drugs: doxorubicin/epirubicin and ifosfamide. Taking into account these two crucial concepts, two randomized trials have obtained similar survival curves above 65% [31], [32], and this outcome was maintained in the long term [33]. More recently, the standard three courses of full‐dose epirubicin and ifosfamide were demonstrated to be significantly better than the experimental arm, in terms of disease‐free survival and overall survival, when compared with histotype‐tailored schemes [34]. Our results, in the same population and with the identical scheme, reproduce the outcome of prospective randomized trials, at least in RFS. Even with a short follow‐up, the overall result observed in our series emphasizes the value of perioperative chemotherapy (mostly neoadjuvant) within the terms mentioned above. Moreover, the multivariate analysis reinforces the relevance of full‐dose perioperative chemotherapy in the high‐risk subset, because only this variable resulted in an independent prognostic value. Other recent unsuccessful trials neither selected homogeneous high‐risk STS nor selected a full‐dose scheme of the two most active drugs [35].

The demonstration of significant survival benefit for those stage IV cases treated in RCs compared with those treated in LHs is meaningful. In CPGs the referral of patients, even in advanced disease, is recommended [25], because the multidisciplinary team strategy for designing the most appropriate therapeutic plan is still crucial. However, to the best of our knowledge, this is the first time that the hospital type (RC vs. LH) has been evaluated as a prognostic variable in advanced STS. The significant difference favoring RCs resulted in a survival benefit of one year as a median. The magnitude of this benefit is similar to that achieved by olaratumab, a finding that led to accelerated approval by the U.S. Food and Drug Administration and the European Medicines Agency [36]. In recent times, tertiary hospitals have been rolling out new operative sarcoma boards, which might improve the quality of care results in sarcoma care in our country.

Limitations derived from relatively low compliance with data entry are acknowledged by the authors of the study. The main reason for this is related to the fact that, this is, in the end, an academic and discretionary study. In addition, the database was highly demanding, and only a few study coordinators were trained well enough to record data. All of the above factors led to there being a more limited number of valid files.

Conclusion

The results of this study, including patients from different types of hospitals, give support to the assertion that an early referral to an RC is critical for improvement in the survival of patients with STS. Higher compliance with the most appropriate biopsy type, workup of imaging tests, and chemotherapy use are some examples of factors that may explain the significantly favorable survival outcome in RC, and it is clear that multidisciplinary teams are crucial in designing the most suitable therapeutic plan. Hopefully, the Ministry of Health's designation in July 2017 of RCs for sarcoma care will lead to clear progress in survival results in patients with sarcoma in Spain.

See http://www.TheOncologist.com for supplemental material available online.

Author Contributions

Conception/design: Javier Martin‐Broto, Josefina Cruz, Claudia Valverde, Angeles Sala, Antonio Lopez‐Pousa

Provision of study material or patients: Javier Martin‐Broto, Nadia Hindi, Josefina Cruz, Javier Martinez‐Trufero, Claudia Valverde, Luis M. De Sande, Angeles Sala, Lorena Bellido, Ana De Juan, Jordi Rubió‐Casadevall, Roberto Diaz‐Beveridge, Ricardo Cubedo, Oscar Tendero, Diego Salinas, Isidro Gracia, Rafael Ramos, Silvia Baguè, Antonio Gutierrez, José Duran‐Moreno, Antonio Lopez‐Pousa

Collection and/or assembly of data: Javier Martin‐Broto, Nadia Hindi, Josefina Cruz, Javier Martinez‐Trufero, Claudia Valverde, Luis M. De Sande, Angeles Sala, Lorena Bellido, Ana De Juan, Jordi Rubió‐Casadevall, Roberto Diaz‐Beveridge, Ricardo Cubedo, Oscar Tendero, Diego Salinas, Isidro Gracia, Rafael Ramos, Silvia Baguè, Antonio Gutierrez, José Duran‐Moreno, Antonio Lopez‐Pousa

Data analysis and interpretation: Javier Martin‐Broto, Nadia Hindi, Josefina Cruz, Javier Martinez‐Trufero, Claudia Valverde, Luis M. De Sande, Angeles Sala, Lorena Bellido, Ana De Juan, Jordi Rubió‐Casadevall, Roberto Diaz‐Beveridge, Ricardo Cubedo, Oscar Tendero, Diego Salinas, Isidro Gracia, Rafael Ramos, Silvia Baguè, Antonio Gutierrez, José Duran‐Moreno, Antonio Lopez‐Pousa

Manuscript writing: Javier Martin‐Broto, Nadia Hindi, Josefina Cruz, Javier Martinez‐Trufero, Claudia Valverde, Luis M. De Sande, Angeles Sala, Lorena Bellido, Ana De Juan, Jordi Rubió‐Casadevall, Roberto Diaz‐Beveridge, Ricardo Cubedo, Oscar Tendero, Diego Salinas, Isidro Gracia, Rafael Ramos, Silvia Baguè, Antonio Gutierrez, José Duran‐Moreno, Antonio Lopez‐Pousa

Final approval of manuscript: Javier Martin‐Broto, Nadia Hindi, Josefina Cruz, Javier Martinez‐Trufero, Claudia Valverde, Luis M. De Sande, Angeles Sala, Lorena Bellido, Ana De Juan, Jordi Rubió‐Casadevall, Roberto Diaz‐Beveridge, Ricardo Cubedo, Oscar Tendero, Diego Salinas, Isidro Gracia, Rafael Ramos, Silvia Baguè, Antonio Gutierrez, José Duran‐Moreno, Antonio Lopez‐Pousa

Disclosures

Javier Martin‐Broto: Pharmamar, Eli Lilly & Co., Novartis, Bayer (C/A), Pharmamar, Eisai, Novartis (RF); Nadia Hindi: Pharmamar, Novartis, Eisai (RF). The other authors indicated no financial relationships.

(C/A) Consulting/advisory relationship; (RF) Research funding; (E) Employment; (ET) Expert testimony; (H) Honoraria received; (OI) Ownership interests; (IP) Intellectual property rights/inventor/patent holder; (SAB) Scientific advisory board