(P 168) SYSTEMIC INFLAMMATION INDEX (SII) AS INDEPENDENT PREDICTIVE FACTOR IN PATIENTS WITH L-SARCOMA TREATED WITH TRABECTEDIN - ANALYSIS OF REAL-WORLD EVIDENCE FROM A REFERENCE CENTER

Objective: Treatment of advanced soft tissue sarcoma (STS) remains a great challenge, especially in second and further lines. Trabectedin is a chemotherapeutic agent that showed activity in STS after failure of anthracycline-based treatment, with best outcomes reported in L-sarcomas- liposarcoma (LPS) and leiomyosarcoma (LMS). The aim of the study was to analyze outcomes, prognostic and predictive factors, including biomarkers of systemic inflammation, in patients with L-sarcomas treated in the routine practice.

Methods: We have retrospectively analyzed medical records of all consecutive patients who inititated the treatment with trabectedin between April 2008 and September 2021. Demographic and clinical data have been extracted, and for patients with available full laboratory results following markers of systemic inflammation have been calculated: neutrophil-to-lymphocytes ratio (NLR), platelet-to-lymphocyte ratio (PLR), lymphocyte-to-monocyte ratio (LMR) and systemic inflammation index (SII, platelet x neutrophil/lymphocyte count). Cut off values were selected based on ROC curves if the area under curve (AUC) was above 0.5. For the survival analyses Kaplan-Meier estimator, log-rank test and multivariable Cox regression were performed. A Growth Modulation Index-Based GEISTRA Score was calculated as previously described, assigning 1 point for each variable: metastasis-free interval (MFI) 0–5 months, Karnofsky performance status < 80%, non-L-sarcoma histology, and clinical benefit (CR/PR/SD) of previous line.

Results: We have identified 243 patients – 55.6% female, 53.9% with LMS and 46.1%with LPS. Most patients, 80.7%, received treatment for metastatic disease. and metastasis-free interval (MFI)≤10 months was reported in 32.5%. 174 patients had available laboratory results – median NLR at the start of trabectedin was 2.87, PLR – 174.04, LMR 2.13 and SII 1047.26. Based or ROC curves following cut-off were established: NLR – 2.7, PLR - 180, SII - 765, while LMR was not included due to AUC< 0.5.
Trabectedin was mostly used in 2nd or 3rd line in 35% and 49.8%, respectively. Median number of administered cycles was 6 (range 1-77). As of 31.12.2021, a median follow-up was 13.4 months. Objective responses according to RECIST 1.1 were reported in 10.3% and stable disease in 49.4% of patients. The median PFS was 5.3 (95%CI 3.90-6.62) months, with 6- and 12-month PFS rates of 46.6% and 28.3%. In a multivariate model in all patients LPS (HR 0.71, 95%CI 0.52-0.95) and MFI>10 months (HR 0.71, 95%CI 0.51-0.98) were significantly associated with longer PFS. When analysis was performed in patients with laboratory data - MFI>10 months (HR 0.62, 95%CI 0.42-0.93) and SII≤765 (HR 0.57, 95%CI 0.34-0.94) correlated with PFS. There was no difference in PFS between GEISTRA score 0-1 and 2-4 groups (p=0.198). When SII>765 was added as additional factor to GEISTRA, patients with score 0-1 had significantly longer PFS than those with 2-5 score (7.8 vs 2.3 months, p< 0.001).
The median OS was 19.02 (95%CI 16.36-21.69) months, with 12- and 24-month rates of 64.5% and 37.7%. Considering baseline factors, only female gender (HR 0.63, 95%CI 0.44-0.88) and MFI>10 (HR 0.58, 95%CI 0.41-0.84) were prognostic for OS, while after including on-treatment factors also complete/partial response (HR 0.41, 95%CI 0.23-0.73) and stable disease (HR 0.48, 95%CI 0.34-0.68) on trabectedin were significantly correlated with OS. No correlation of NLR, PLR and SII with OS was found.

Conclusion: Long term-outcomes of treatment with trabectedin in patients with L-STS are comparable to other real-world evidence studies. Metastasis-free interval is significant predictive and prognostic factor in patients treated with trabectedin, while patients with LPS tend to achieve longer PFS, but not OS, comparing to patients with LMS. Moreover, we found that systemic inflammation index SII to significant correlate with PFS and improve prognostication of previously described GEISTRA score.