E � cacy and Safety of Trabectedin in Patients with Soft Tissue Sarcoma : A Bicentric Retrospective Analysis

Chaigneau Loïc (  lchaigneau@chu-besancon.fr ) University Hospital of Besançon Jary Marine University Hospital of Besançon Nerich Virginie University Hospital of Besançon Hervieu Alice Centre Georges-François Leclerc Aubry Sébastien University Hospital of Besançon Charon Barra Céline Centre Georges-François Leclerc Meynard Guillaume University Hospital of Besançon Neumann Florent University Hospital of Dijon Kalbacher Elsa University Hospital of Besançon Isambert Nicolas University Hospital of Poitiers


Introduction
Soft Tissue Sarcomas (STS) are a heterogenous group of tumors including more than 80 histological subtypes, the most frequent being liposarcomas and leiomyosarcomas (1,2). However, they represent less than 1% of all malignancies, and the only curative treatment is surgery, with or without radiation. Despite optimal local therapy, the 5-years rate of relapse is 40% (3). At locally advanced or metastatic stage, patients have less than 10% 5-years overall survival (OS). Systemic chemotherapy remains the standard treatment in a palliative intend: doxorubicin and ifosfamide are the most active drugs in STS, providing as a single agent an overall response rate of 15-35% and 18-20%, respectively (4-7). As the disease progresses, several chemotherapies and targeted therapies are available with debatable e cacy. Strategy decision of treatment depends on subtype of sarcoma.
Among the available drugs, trabectedin is an antitumor agent derived from a marine source, the Caribbean tunicate Ecteinascidia turbinata, now synthetically produced (8). The way trabectedin works is complex, occurring during gene transcription. Unlike what is expected for a DNA-damaging agent, sensitivity to trabectedin is correlated to a functional nucleotide excision repair system (9).
In France, trabectedin was available -initially through a compassionate use program (ATU) -since 2003, only in referent centers. After September 2007, trabectedin was approved in the European Union for patients with advanced STS after failure of anthracyclins and ifosfamide or for those who are unsuited to receive such agents. This approval was based on a phase II study demonstrating a e cacy in terms of response and progression free survival (10) .
A routine assessment is often mandatory to con rm the observed e ciency of new drugs in real world settings. Although randomized clinical trials provide essential responses to assess the clinical bene t of a molecule or treatment regimen, the results of these randomized trials may not be applicable in common practice. Indeed, the results are not easily generalizable to a larger and more heterogeneous population.
One must distinguish e cacy, clinical assessment under optimal conditions (based on the inclusion and exclusion criteria speci ed in randomized trials) and effectiveness, clinical assessment under "real-life" conditions of use. Real-world evidence (RWE) and randomized control trial (RCT) are therefore complementary. The latter mainly provides 2 types of information: rst, to what extent the conditions of clinical trials are veri ed in real life: population actually treated, dosage, adherence... Secondarily, they may highlight side effects -particularly delayed -not observed in clinical trials, due to the limited number of patients included and insu cient follow-up.
The purpose of this paper is then to analyse the use of trabectedin at University Hospital Center of Besançon and Georges François Leclerc Cancer Center of Dijon, respectively in Franche-Comté and Burgundy regions (France). This bicentric retrospective analysis describes the bene t of trabectedin, in terms of response, survival and adverse events, in the different types of STS.

Data extraction
All consecutive patients treated with trabectedin at University Hospital Center of Besançon and Georges François Leclerc Cancer Center of Dijon, between January 2002 and August 2018, were retrospectively studied.
All patients with histologically proven STS were eligible for this retrospective analysis. Sarcomas other than STS like GIST, Ewing sarcoma or paediatric sarcoma were excluded. Treatment characteristics were captured from the BPC® software or the pharmacy database (chemotherapy lines, type of chemotherapy, trabectedin dosage, cycle delay, treatment interruption, adverse or severe adverse event…). The patient demographic, the disease characteristics, the treatment e cacy and the patient outcome were collected from the patient's medical record.
For each patient, the age at diagnosis, the sex, the performance status (World Health Organisation criteria), the initial disease characteristics (histological subtype and grading, staging of the disease as per "Fédération Nationale des Centres de Lutte Contre le Cancer" criterion, site of the primary disease and sites of metastases) were recorded. Before beginning the trabectedin regimen, the number of previous lines was recorded, as the date of the last chemotherapy and its type.

Clinical assessment
Only patients who have been treated by at least 2 cycles of trabectedin were included for analysis e cacy. Patients were considered evaluable for e cacy if they had at least one assessment of tumor response.
The primary end-point was progression-free survival (PFS), de ned as the time from the start of chemotherapy (trabectedin) to disease progression or death by any cause. PFS was analysed using the Kaplan-Meier method and was described using the median and 95% con dence interval (CI). The difference between two treatment arms was compared with the use of the log-rank test, with the hazard ratio and its 95% con dence interval calculated from a Cox regression model with a single covariate. Continuous variables were described by the mean standard deviation (SD) and median with range, and qualitative variables by the size and percentage rate. Between two cohorts, qualitative and quantitative variables were respectively compared by the Fisher exact test or the chi-square test and nonparametric Mann-Whitney test. All tests were two-tailed and signi cant at an alpha threshold of 5% (p-value).
Statistical analysis was performed with SAS® software version 9.4 (Cary, NC, USA) Secondary end points were PFS at 3 months, PFS at 6 months, OS, response to the treatment and disease control. OS was de ned as either the time from initiation of chemotherapy (trabectedin) to death from any cause or to last follow-up for survivors. Patients alive on April 01, 2020, the last date of analysis, were censored. Survival curves were estimated by the Kaplan-Meier method and were described using the median and 95% CI. The response rate was primarily assessed by clinical examination or imaging, according to the Response Evaluation Criteria in Solid Tumors (RECIST 1.1) (11). Patients were classi ed as having a disease control if they had Complete Response (CR), Partial Response (PR), or Stable Disease (SD).

Adverse events
All patients who have been treated by at least 1 cycle of trabectedin were included for toxicities analysis. All toxicities were recorded, with their grading, according to the National Cancer Institute -Common Terminology Criteria for Adverse Events v4.0 (CTAE).

Discussion
With the exception of gastrointestinal stromal tumors (GIST), the prognostic of patients treated for locally advanced and metastatic STS remains poor. In all clinical guidelines, management of sarcoma patients should be discussed and carried out within a dedicated multidisciplinary team. Despite a management in expert centres, the median of overall survival is 18 months (12). The various palliative treatments may include surgery (for example, pulmonary metastasis), radiotherapy and especially chemotherapy (13). Nowadays, second lines after failure of doxorubicin and/or ifosfamide include others regimens like dacarbazine alone or in combination, eribulin for liposarcomas, gemcitabine alone or in combination with docetaxel for leiomyosarcomas and undifferentiated pleomorphic sarcomas, paclitaxel for angiosarcomas or pazopanib (14)(15)(16)(17)(18) (19,20). A drug is considered active if it allows 3-and 6month PFS rates > 39% and 14%, respectively, as second-line regimens. In our retrospective study, the 3months PFS rate and the 6-months PFS rate were respectively 48.5% and 30.0%, comparable to the other publications.
To date, one major concern in STS treatment is that no predictive nor prognostic biomarker can be routinely used. In this retrospective analyse, no clinical or tumoral parameter was signi cantly associated with PFS and OS. Trabectedin administration resulted in a median PFS of 3. In terms of toxicities, most of them were expected and manageable, non -cumulative, mainly summarized by haematological and hepatological failures, and consistent with the previous T -SARC randomized trial (23). Thirty four percent of patients used granulocyte colony-stimulating factors which were 28% in the Demetri et al. study (10). Other toxicities were quite similarly to the results of the current publications (35)(36)(37).
This study had some limitations. The retrospective design made comparisons with previous randomized control trial di cult because the study population and follow-up may differ. For example, the enrolment and treatment criteria of our study were less restrictive with more pretreated patients and other histological subtypes than liposarcomas and leiomyosarcomas. In addition, dose reductions, treatment delays and tumor evaluations depended on the physician's choice. The small number of patients and missing data limited the analysis, in particular the association between different histological subtypes and clinical outcomes. Finally, there was no provision for a quality-of-life assessment Despite these limitations, this study is representative of the general population because it included patients treated in 2 centers with different medical situations, and these results could be considered to represent assessment of trabectedin effectiveness in the real-world clinical conditions.

Conclusion
In conclusion, strategies of treatment after conventional drugs remain subject to debate for advanced or metastatic STS. Independently from the necessary need to take the histological and molecular subtypes into account, the choice of the best regimen has to integrate the tumor molecular characterisation, and probably cost effectiveness and quality of life. Nevertheless, trabectedin seems to remain interesting in terms of clinical bene t and low toxicity. It has to be con rmed in larger randomised trials containing recent strategies.

Declarations
Ethics approval and consent to participate This retrospective study was approved by Institutional Review Board of the Regional Cancer Institute (University Hospital of Besancon, France). Written informed consent was obtained from all alive study subjects. The study was performed in accordance with the Declaration of Helsinki.

Consent for publication
Not applicable Availability of data and materials The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.