Trabectedin, a semisynthetic DNA-binding protein that was initially isolated from the marine ascidian Ecteinascidia turbinata, exerts an antineoplastic effect by binding to the minor groove of DNA and bending the DNA towards the main groove (1).

Trabectedin was first approved for treatment of soft-tissue sarcoma (STS) in Europe (2), based on successful single-arm phase II trials (3-5). Data from subsequent clinical use in Europe suggested that two subsets of STS patients, with leiomyosarcoma/liposarcoma (hereafter referred to as L-sarcoma) or chromosome translocation-related sarcoma, would benefit most from trabectedin treatment (6,7). A randomised phase III trial in patients with advanced L-sarcoma who had received prior therapy with an anthracycline and at least one additional systemic regimen demonstrated that trabectedin gave longer progression-free survival (PFS) of 4.2 months compared to dacarbazine (1.5 months) (8). These results supported the approval of trabectedin for clinical use in the U.S. (9). In a Japanese randomised phase II study in STS patients that had chromosomal translocations, trabectedin reduced the risk of disease progression compared to best supportive care (BSC) after chemotherapy, achieving a median PFS of 5.6 months for patients treated with trabectedin compared to 0.9 months for BSC (10). Based on these results, trabectedin was approved for the treatment of all types of STS in Japan.

Despite these advances, prognostic predictors for trabectedin in patients with STS have not been identified. Like trabectedin, eribulin mesylate, a microtubule-targeting agent, is also used as a second- or later-treatment option for STS. We recently reported that a low baseline neutrophil-to-lymphocyte ratio (NLR) predicts the efficacy of eribulin treatment in patients with STS (11). Several previous reports have suggested that the baseline NLR may reflect the antitumour immunity status and associated prognosis of cancer patients (12-14). Changes in NLR were also found to be prognostic in several retrospective studies in patients with non-small cell lung cancer, renal cell carcinoma, and gastric cancer treated by immune checkpoint inhibitors (15-17). Herein, we explored factors that predict the efficacy of trabectedin, including NLR and changes in NLR, for patients with STS, particularly non-round cell sarcoma.

Patients. We retrospectively analysed prospectively collected data from 39 consecutive patients with recurrent or metastatic non-round cell sarcoma who began treatment with trabectedin at the Cancer Institute Hospital of Japanese Foundation for Cancer Research between October 2012 and December 2019. The database comprised the following patient characteristics: age, sex, histological diagnosis, location of the primary tumour, gene translocation, Eastern Cooperative Oncology Group performance status (ECOG PS), number of previous systemic chemotherapies, and the absolute neutrophil count (ANC) and absolute lymphocyte count (ALC) of blood samples collected within a week before (pre-) and after (post-) the first infusion. These factors were categorised as follows. Age: <40 years and ≥40 years, or age <65 years and ≥65 years; histology: L-sarcoma (leiomyosarcoma and liposarcoma) and non-L-sarcoma; location of the primary tumour: extremities and non-extremities; chromosomal translocation: positive and negative; ECOG PS: 0 and ≥1; number of previous systemic chemotherapies: 0-1 and ≥2; ALC: <1,500 cells/μl and ≥1,500 cells/μl; NLR (calculated as ANC divided by ALC): <3.0 and ≥3.0; and changes in NLR (ΔNLR; calculated as post-NLR divided by pre-NLR): <0.5 and ≥0.5.

Trabectedin was administered at a dose of 1.2 mg/m² as the standard approved dose in Japan, and subsequent administration was continued at the same dose every 3 weeks. Dose reductions to 1.0 and 0.8 mg/m² or prolonged administration intervals were permitted until improvement of adverse events at the physician’s discretion. Dosing was adjusted or discontinued depending on the condition of each individual patient. All treatment was continued until the occurrence of unacceptable adverse effects or disease progression.

The requirement for informed consent was waived because the data were reported anonymously. This study was approved by the Institutional Review Board of the Cancer Institute Hospital of Japanese Foundation for Cancer Research.

Statistical analysis. PFS and overall survival (OS) were estimated using the Kaplan–Meier method and the log-rank test. Data were censored on January 31, 2021. Patients who were lost to follow-up were censored at the date of last contact or follow-up. PFS was calculated from the date of trabectedin initiation to the date of disease progression or death from any cause. OS was calculated from the date of trabectedin initiation to the date of death from any cause. Patients who were alive on January 31, 2021 were censored for OS analysis. Tumour response was evaluated according to the Response Evaluation Criteria in Solid Tumours, version 1.1 (8), based on computed tomographic findings. The best overall response was assessed as complete response (CR), partial response (PR), stable disease (SD), or progressive disease (PD). The overall response corresponded to the sum of the CR and PR, and disease control corresponded to the sum of the CR, PR, and SD. We performed univariate and multivariate analyses to estimate factors potentially prognostic of PFS and OS; we calculated hazard ratios (HRs) using a Cox proportional hazards model for PFS and OS. The level of significance was set to p<0.1 for the univariate analysis and p<0.05 for the multivariate analysis and was two-sided. All statistical analyses were performed with EZR (Saitama Medical Center, Jichi Medical University, Saitama, Japan), which is a graphical user interface for R (www.r-project.org). More precisely, it is a modified version of R Commander designed to add statistical functions that are frequently used in biostatistics (18).

Patient characteristics. A total of 39 patients with non-round cell sarcoma were treated with trabectedin between October 2012 and December 2019. The 39-patient cohort included 19 men, and the median age was 49 years (range=21-74 years). The median duration of observation was 17.2 months (range=0.5-64.0 months). The patient characteristics are shown in Table I. The median pre-treatment and post-treatment ANCs were 3,070 cells/μl (range=1,930-28,830 cells/μl) and 2,010 cells/μl (range=570-34,770 cells/μl), respectively, and the median pre-treatment and post-treatment ALCs were 1,230 cells/μl (range=410-3,590 cells/μl) and 950 cells/μl (range=400-3,030 cells/μl), respectively. The median pre-treatment and post-treatment NLRs were 2.71 (range=0.89-20.59) and 2.04 (range=0.66-31.63), respectively.

Clinical efficacy of trabectedin. The objective response rate was 12% (n=5) and the disease control rate was 59% (n=23) (Table II). Trabectedin was withdrawn in four patients without evaluation of best objective response because of deterioration of their general conditions. The median PFS and OS were 6.5 months [95% confidence interval (CI)=2.6-5.2] and 18.0 months (95% CI=12.5-27.2), respectively (Figure 1).

Predictive factors for OS and PFS. As shown in Table III, multivariate analysis indicated that age ≥40 years (HR=0.33, 95% CI=0.15-0.71; p=0.0050) and ΔNLR <0.5 (HR=2.40, 95% CI=1.01-5.72; p=0.048) were independent predictors of OS. Gender, histology, location of primary lesion, gene translocation, ECOG PS, number of previous chemotherapies, pre-ALC and pre-NLR, and post-ALC and post-NLR were not associated with OS. Moreover, as shown in Table IV, multivariate analysis indicated that age ≥40 (HR=0.23, 95% CI=0.10-0.52; p<0.001) was an independent predictor of longer PFS. Gender, histology, location of primary lesion, gene translocation, ECOG PS, number of previous chemotherapies, pre-ALC and pre-NLR, post-ALC and post-NLR, and ΔNLR were not associated with PFS.

In this study, we investigated predictive factors of trabectedin monotherapy in patients with non-round cell sarcoma, including NLR and changes in NLR. Notably, we identified ΔNLR <0.5 and age ≥40 as independent predictors of OS. Age ≥40 was also identified as an independent predictor of prolonged PFS.

To the best of our knowledge, no other study has evaluated NLR in trabectedin-treated patients with sarcoma. In the present study, change in NLR was identified as an independent predictive marker of OS. Pre-treatment NLR was not a predictive factor of trabectedin efficacy in patients with non-round cell sarcoma, unlike eribulin for STS patients. This disparity between the two drugs may be due to differences in their mechanisms of cytotoxicity and in their influences on antitumour immunity status. The inhibition of trans-activated transcription and DNA repair proteins are hallmarks of trabectedin cytotoxicity. Inhibition of active transcription is achieved by interaction of trabectedin with RNA polymerase II, which subsequently alters the production of cytokines and chemokines by tumour-associated macrophages (1). Eribulin, a synthetic analogue of halichondrin B that was originally isolated from the marine sponge Halichondria okadai, selectively binds to the polymerisation site of microtubules with high affinity and inhibits microtubule growth. Its potent microtubule-depolymerising properties and its profound effects on the tumour microenvironment differ from those of other microtubule-targeting agents such as paclitaxel (19). Further research into the cytotoxic mechanisms of eribulin and trabectedin is warranted in order to understand their disparate effects in sarcoma patients.

In the present study, age ≥40 years was also an independent predictor of longer OS and PFS. As shown in Table V, histological features between the two groups were not significantly different (p=0.99). A previous study of trabectedin in children with STS, including 22% with non-round cell sarcoma, did not demonstrate sufficient antitumour activity (20). Our data are consistent with this previous result. Patients aged <40 years are defined as adolescent and young adult (AYA) patients, and there are several differences between AYA and older patients with STS in regard to presentation, treatment, and survival (21). Moreover, the same pathologically diagnosed sarcoma has a different profile of genetic mutations in young patients compared to adults (22). Further investigation of the potential biological mechanisms underlying STS in these two age groups is warranted.

Our study shows that histology features and chromosomal translocation were not associated with efficacy of trabectedin. These data were consistent with our previous retrospective report that evaluated differences in the efficacy and safety of trabectedin for patients with STS separated by histological subtype or chromosomal translocation (23).

Several limitations of this study should be acknowledged. First, this was a retrospective study with a small number of patients, and a selection bias may have resulted from physician subjectivity when determining which patients should receive trabectedin at which time. Second, the NLR value is variable, not only because of tumour factors but also owing to infection, corticosteroids, radiotherapy, or other physiological stressors. Although we used a cut-off value of 3.0 for NLR and 0.5 for ΔNLR according to the findings of previous studies (11,17), the appropriate cut-off value is still under debate. We plan to continue accumulating data from a greater number of patients to inform future studies.

In conclusion, this retrospective study evaluated predictive factors of trabectedin monotherapy in non-round cell sarcoma patients. Notably, we found that changes in NLR (ΔNLR ≥0.5) and age ≥40 years were independent predictors of prolonged OS.

YS reports personal fees from ONO Pharmaceutical Co., Ltd., Bristol Myers Squibb Company, MSD KK, and Taiho Pharmaceutical Co., Ltd., outside the submitted work. NF and JT report personal fees from Eisai Co., Ltd. ST reports grants and personal fees from Bristol Myers Squibb Company, grants and personal fees from ONO Pharmaceutical Co., Ltd., grants and personal fees from MSD KK, grants and personal fees from AstraZeneca plc, grants and personal fees from Chugai Pharmaceutical Co., Ltd., and grants and personal fees from Bayer AG, outside the submitted work. The other Authors report no competing interests to disclose.

YS designed the study and wrote the manuscript. KN designed the study and revised the manuscript. KK, NF, XW, TU, AO, NH, MY, MO, TJ, KH, YF, TT, KA, SM, and TT contributed critical revisions of the manuscript. All Authors read and approved the final version of the manuscript.

We thank the medical staff of the Department of Medical Oncology and Orthopedic Oncology at the Cancer Institute Hospital of the Japanese Foundation for Cancer Research for their support during this study. We thank Michelle McRae, PhD, from Edanz Group (https://jp.edanz.com/ac) for editing a draft of this manuscript.