Search

Angiosarcoma (AS) is a rare disease with a dismal prognosis. The treatment landscape and prognostic factors for advanced AS, including locally advanced, unresectable, and metastatic disease remain elusive. The Asian Sarcoma Consortium is an international collaborative effort to understand the sarcoma treatment landscape in Asia. We undertook a retrospective chart review of AS patients seen in 8 sarcoma academic centers across Asia. Patients with complete clinical, treatment, and follow‐up data were enrolled. Overall, 276 advanced AS patients were included into this study; 84 (30%) of the patients had metachronous metastatic AS. The median age was 67 y; primary sites of AS was cutaneous in 55% and visceral in 45% of patients. In total, 143 (52%) patients received at least 1 line of systemic chemotherapy. The most common first‐line chemotherapy regimen used was paclitaxel (47.6%) followed by liposomal doxorubicin (19.6%). The median overall survival (OS) was 7.8 mo. Significant prognostic factors for OS included age > 65 (hazard ratio (HR) 1.54, P = .006), male gender (HR 1.39, P = .02), and a cutaneous primary AS site (HR 0.63, P = .004). The median progression‐free survival (PFS) for first‐line chemotherapy was 3.4 mo. PFS for single vs combination or paclitaxel vs liposomal doxorubicin chemotherapy regimens were comparable. This study provides an insight into the treatment patterns and prognostic factors of advanced AS patients in Asia. Prognosis of advanced AS remains poor. Data from this study serve as a benchmark for future clinical study design., With a total of 276 advanced angiosarcoma patients, we provided the complete clinicopathological demographics, treatment patterns, and prognostic outcomes of advanced angiosarcoma in Asia.

Contemporary data of peripheral T-cell lymphoma (PTCL) and natural-killer/T-cell lymphoma (NKTL) patients treated with ifosfamide, carboplatin and etoposide (ICE) are limited.We performed a retrospective analysis to estimate outcomes of ICE-treated PTCL and NKTL patients at three tertiary cancer centres in Singapore.Patients were identified through lymphoma databases from National Cancer Centre Singapore (NCCS), National University Hospital, Singapore (NUHS), and Singapore General Hospital (SGH). Responses and survival outcomes were determined from electronic medical records. A total of 75 patients with a median age of 50 were included. ICE was used as first-line treatment in 14 patients (19%) and as subsequent lines of treatment in 61 patients (81%). The overall response rates (ORR) for all patients was 63% (40% complete response [CR]). The ORR and CR in the first line were 86% and 64% respectively. At a median follow-up duration of 71.0 months, the median progression-free (PFS) and overall survival (OS) for all patients were 4.4 months (95%CI, 2.7-6.0) and 16 months (95%CI, 8.3-45.4) respectively.In summary, ICE showed high ORR but poor PFS in relapsed/refractory PTCL and NKTL. ORR of ICE in the first line setting appears better than real-world CHOP data and warrants further study.

Angiosarcomas are rare, clinically aggressive tumors with limited treatment options and a dismal prognosis. We analyzed angiosarcomas from 68 patients, integrating information from multiomic sequencing, NanoString immuno-oncology profiling, and multiplex immunohistochemistry and immunofluorescence for tumor-infiltrating immune cells. Through whole-genome sequencing (n = 18), 50% of the cutaneous head and neck angiosarcomas exhibited higher tumor mutation burden (TMB) and UV mutational signatures; others were mutationally quiet and non–UV driven. NanoString profiling revealed 3 distinct patient clusters represented by lack (clusters 1 and 2) or enrichment (cluster 3) of immune-related signaling and immune cells. Neutrophils (CD15(+)), macrophages (CD68(+)), cytotoxic T cells (CD8(+)), Tregs (FOXP3(+)), and PD-L1(+) cells were enriched in cluster 3 relative to clusters 2 and 1. Likewise, tumor inflammation signature (TIS) scores were highest in cluster 3 (7.54 vs. 6.71 vs. 5.75, respectively; P < 0.0001). Head and neck angiosarcomas were predominant in clusters 1 and 3, providing the rationale for checkpoint immunotherapy, especially in the latter subgroup with both high TMB and TIS scores. Cluster 2 was enriched for secondary angiosarcomas and exhibited higher expression of DNMT1, BRD3/4, MYC, HRAS, and PDGFRB, in keeping with the upregulation of epigenetic and oncogenic signaling pathways amenable to targeted therapies. Molecular and immunological dissection of angiosarcomas may provide insights into opportunities for precision medicine.

Introduction The role of surveillance imaging (SI) in patients with peripheral T-cell lymphoma (PTCL) in first complete remission (CR1) is unclear. Materials and Methods Patients with PTCL were identified through prospectively maintained T-cell lymphoma databases from the National Cancer Centre Singapore/Singapore General Hospital and Weill-Cornell Medical College after institutional review board approval. Patients with leukemia or indolent, composite, and cutaneous lymphomas were excluded. The patients' medical records were retrospectively reviewed to determine the frequency and type of SI used. Of those with relapse, the method of relapse detection and data on symptoms, signs, and elevated lactate dehydrogenase LDH were extracted. Results A total of 338 patients were included in the present study. In the first year after achieving CR1, patients had an average of 1.2 and a median of 1 SI performed (range, 0-4). In the second year after achieving CR1, they had an average of 0.78 and a median of 1 SI performed (range, 0-4). Of the 135 patients who achieved CR1, 61 (45%) developed a relapse. Relapses were detected before SI in 48 (84%), and 9 patients had relapses detected during routine SI. Of the 9 patients whose relapses were detected during planned SI, only 3 did not have any symptoms or signs suggestive of relapsed disease. Of these 3 patients, 2 had angioimmunoblastic T-cell lymphoma and 1 had natural killer/T-cell lymphoma. Conclusion Most PTCL relapses were detected before planned SI, and most patients had symptoms with relapse. Only 3 patients (5.2%) were completely asymptomatic at relapse, suggesting a limited utility of routine imaging for detecting PTCL relapses.

Primary mediastinal large B‐cell lymphoma (PMBCL) is a distinct clinico‐pathological subtype of diffuse large B‐cell lymphoma with unclear prognostic factors and limited clinical data. Optimal treatment and role for radiotherapy is not fully defined. We performed a multicenter retrospective review of 124 patients with newly diagnosed PMBCL between 2001 and 2016. Treatment regimens were R‐CHOP (n = 41), R‐CHOP + RT (n = 37), and DA‐EPOCH‐R (n = 46). 6% (n = 3) in the DA‐EPOCH‐R group received RT. With a median follow up of 45 months, the overall 5‐year OS and PFS was 89.4% and 82.4%, respectively. The type of chemo‐radiotherapy regimen, B symptoms and Ann‐Arbor staging showed a significant association with OS on univariate analysis but only B symptoms remained prognostic (P = 0.012) after multivariate analysis. The chemo‐radiotherapy regimen, Japanese IPI and Ann‐Arbor stage was significantly associated with PFS in univariate analysis, but only chemo‐radiotherapy regimen remained significant (P = 0.02) after multivariate analysis. Patients who received R‐CHOP + RT or DA‐EPOCH‐R had better PFS than those receiving R‐CHOP alone, with 5‐year PFS of 90% vs 88.5% vs 56%, respectively (P = 0.02). In the subgroup analysis of patients with bulk (n = 71), R‐CHOP alone (n = 21) had inferior 5‐year PFS 56.6% compared to those who received R‐CHOP + RT (n = 23) 91.3% or DA‐EPOCH‐R (n = 27) 92.6% (P = 0.007). In contrast, in patients without bulk (n = 42), there was no impact of treatment regimen on PFS (P = 0.25). In conclusion, R‐CHOP + RT and DA‐EPOCH‐R provide excellent outcomes in patients with PMBCL. In patients with bulky disease, the use of DA‐EPOCH‐R may be preferable as it allows omission of RT without reduction in efficacy.

Purpose The value of screening for hepatitis B virus (HBV) infection before chemotherapy for nonhematopoietic solid tumors remains unsettled. We evaluated the cost effectiveness of universal screening before systemic therapy for sarcomas, including GI stromal tumors (GISTs). Patients and Methods Drawing from the National Cancer Centre Singapore database of 1,039 patients with sarcomas, we analyzed the clinical records of 485 patients who received systemic therapy. Using a Markov model, we compared the cost effectiveness of a screen-all versus screen-none strategy in this population. Results A total of 237 patients were screened for HBV infection. No patients developed HBV reactivation during chemotherapy. The incremental cost-effectiveness ratio per quality-adjusted life-year (QALY) of offering HBV screening to all patients with sarcomas and patients with GISTs exceeded the cost-effectiveness threshold of SG$100,000 per QALY. This result was robust in one-way sensitivity analysis. Our results show that only changes in mortality rate secondary to HBV reactivation could make the incremental cost-effectiveness ratio cross the cost-effectiveness threshold. Conclusion Universal HBV screening in patients with sarcomas or GISTs undergoing chemotherapy is not cost effective at a willingness to pay of SG$100,000 per QALY and may not be required.

11047 Background: Angiosarcomas are rare, clinically-aggressive endothelial tumors with remarkable heterogeneity in terms of anatomical and etiological origins, amongst which UV-radiation and oncogenic viruses have been implicated. Methods: The genomic and transcriptomic landscapes of human angiosarcomas were examined using whole genome sequencing (n = 18) and RNAseq (n = 14). Results were validated in an extended cohort of FFPE tissue (n = 57) on various platforms. Results: We observed recurrent mutations in known angiosarcoma-associated genes including TP53, KDR, POT1 and PTPRB. The median tumor mutation burden (TMB) was 1.95 mutations per megabase and was highest in a subset of head and neck angiosarcomas (HN-AS, n = 6) harboring UV signatures. The remaining cases, including seven HN-AS, contained a significantly lower TMB (5.04 vs 0.91, p= 0.0007). In corroboration, UV exposures were validated by cyclobutane pyrimidine dimer immunohistochemistry in 15 of 35 (42.9%) HN-AS and cutaneous angiosarcomas, but not in visceral or breast angiosarcomas. Gene set enrichment analysis (GSEA) revealed upregulation of “cell cycle” and “inflammation” pathways in angiosarcomas. Notably, drug inhibition of overexpressed kinases including aurora kinase, polo-like kinase and checkpoint kinase-1 potently reduced viability in two angiosarcoma cell lines, achieving IC50s of < 1µM. Metagenomic analysis detected viral reads mapping to the human herpesvirus-7 (HHV-7) genome, which were verified by PCR and immunohistochemistry. Genomic integration of HHV-7 was demonstrated at its flanking DRR telomeric repeat sequence adjoining the subtelomeric region of chromosome 17p. Overall, HHV-7 was detected in 38 of 57 (66.7%) angiosarcomas and inversely correlated with UV exposure ( p= 0.01). GSEA and Nanostring IO360 profiling showed that “inflammation” pathways were enriched in viral-positive angiosarcomas, while “cell cycle” pathways were enriched in viral-negative cases. Compared with HHV-7 negative cases, HHV-7 positive angiosarcomas were associated with high tumor inflammation scores (TIS ≥ median 6.74) (73.3% vs 25.0%, p = 0.02). Conclusions: We identified distinct subtypes of angiosarcoma characterized by HHV7-positive “inflammation” and UV-associated “cell cycle” phenotypes.

Soft-tissue sarcomas require tailored and multidisciplinary treatment and management. However, little is known about how sarcomas are treated and managed throughout the Asia-Pacific region.MEDLINE was systematically searched using prespecified criteria. Publications (previous 10 years) that reported tumour characteristics, treatment patterns, survival outcomes, and/or safety outcomes of patients with soft-tissue sarcoma were selected. Exclusion criteria were studies of patients18 years of age; ≤ 10 patients; countries other than Australia, Hong Kong, Indonesia, Korea, Malaysia, New Zealand, Philippines, Singapore, Taiwan, or Thailand;20% benign tumours; sarcomas located in bones or joints; gastrointestinal stromal tumour; Kaposi's sarcoma; or not reporting relevant outcomes.Of the 1,822 publications retrieved, 35 (32 studies) were included. Nearly all patients (98%, 1,992/2,024; 31 studies) were treated with surgery, and more studies used adjuvant radiotherapy than chemotherapy (24 vs 17 studies). Survival outcomes and recurrence rates varied among the studies because of the different histotypes, sites, and disease stages assessed. Only 5 studies reported safety findings.These findings highlight the lack of specific data available about soft-tissue sarcomas in the Asia-Pacific region. Better efforts to understand how the sarcoma is managed and treated will help improve patient outcomes in the region.

BACKGROUND Advanced uterine leiomyosarcoma (ULMS) is an incurable disease. A significant percentage of cases of ULMS express estrogen and/or progesterone receptors (ER and/or PR). To the authors' knowledge, the role of estrogen suppression in disease management is not known. METHODS The authors performed a single-arm phase 2 study of the aromatase inhibitor letrozole at a dose of 2.5 mg daily in patients with unresectable ULMS with ER and/or PR expression confirmed by immunohistochemistry. Tumor assessments were performed at baseline, 6 weeks, 12 weeks, and every 8 weeks thereafter. Toxicity was monitored throughout treatment. The primary endpoint was the progression-free survival at 12 weeks. RESULTS A total of 27 patients was accrued, with a median of 2 prior treatment regimens (range, 0-9 treatment regimens). The median duration of protocol treatment was 2.2 months (range, 0.4 months-9.9 months). The 12-week progression-free survival rate was 50% (90% confidence interval, 30%-67%). The best response was stable disease in 14 patients (54%; 90% CI, 36%-71%). Three patients, all of whom had tumors expressing ER and PR in > 90% of tumor cells, continued to receive letrozole for > 24 weeks. The most common reason for treatment discontinuation was disease progression (85%). Letrozole was found to be well tolerated. CONCLUSIONS Letrozole met protocol-defined criteria as an agent with activity in patients with advanced ULMS. Patients with the longest progression-free survival rate were those whose tumors strongly and diffusely expressed ER and PR. Cancer 2014;120:738–743. © 2013 American Cancer Society.

Leiomyosarcomas (LMS) comprise 25% of soft tissue sarcomas. Recent reports suggest differences in treatment outcomes between uterine (uLMS) and extrauterine (eLMS) disease that may reflect distinct disease biologies. We sought to identify prognostic factors in LMS and clinicopathologic differences between uLMS and eLMS.This is a single-center retrospective study evaluating 97 eligible patients treated for LMS between 2002 and 2010.Median follow-up was 21.2 months. uLMS affected 53% of patients, and was less common beyond age 60 years compared with eLMS (10% vs. 37%, P = 0.002). Seventy-two percent of patients presented with nonmetastatic disease. Of these, 94% underwent curative surgery, among whom more uLMS patients achieved negative surgical margins (90% vs. 45%, P = 0.003). There were no significant differences in adjuvant therapy use and relapse patterns between uLMS and eLMS. Half of metastatic patients received palliative chemotherapy, among whom 76% received anthracycline-based chemotherapy in first line to which response rate was 31%. Median overall survival was 45.2 months, 49.8 months in uLMS, and 40.5 months in eLMS (P = 0.294). Among patients without metastases, median survival was 60.8 months (77.3 vs. 48.1 mo in uLMS and eLMS, respectively, P = 0.194). In metastatic disease, median survival was 20.7 months (22.0 vs. 17.5 mo in uLMS and eLMS, respectively, P = 0.936). Advanced disease stage, bone metastases and lack of metastasectomy prognosticated for inferior survival.While demonstrating interesting clinicopathologic differences, the evidence for uLMS and eLMS being biologically distinct remains inconclusive. Disease stage is prognostically most important in LMS.

This study aimed to describe the epidemiology and risk factors for cytomegalovirus (CMV) infection and end-organ disease in patients with lymphoma undergoing potentially curative or salvage therapy. We retrospectively reviewed 534 patients with lymphoma treated at an Asian tertiary cancer center between January 2007 and December 2010. Overall, 48 patients (9.0%) experienced CMV infection, with 12 patients (25.0%) being further diagnosed with CMV end-organ disease. Many patients with CMV infection were male, with poor performance status, non-Hodgkin lymphoma and advanced disease, and received rituximab use. Moreover, patients receiving rituximab and HyperCVAD (hyperfractionated cyclophosphamide, vincristine, doxorubicin and dexamethasone) regimens had a high rate of CMV end-organ disease. In Asian patients with lymphoma receiving curative or salvage therapy, CMV infection was relatively common (9.0%). Most of these were likely to be reactivation in nature. A small group, especially those on rituximab or HyperCVAD, developed CMV end-organ disease (12/534). Such patients should be monitored closely for CMV end-organ disease. Alternatively, prophylaxis should be studied.

Background: Well-differentiated/dedifferentiated liposarcoma (WDLPS/DDLPS) is characterized by a consistent coamplification of human double minute 2 homolog (HDM2) and cyclin-dependent kinase 4 (CDK4) which inactivates the tumor suppressor pathways p53 and Rb, respectively. HDM201 is a selective inhibitor of the p53-HDM2 interaction and ribociclib is a CDK4/6 inhibitor. Preclinical studies suggested a synergy in in vitro and in vivo models of WDLPS/DDLPS and both agents have demonstrated single-agent clinical activity in solid tumors. We aim to determine the optimal dose and regimen of HDM201 + ribociclib and to assess the preliminary antitumor activity of this combination in patients (pts) with liposarcoma. Methods: In this multicenter, open-label, Phase Ib ongoing study, pts with locally advanced or metastatic liposarcoma that had unequivocally progressed on, or despite prior systemic therapy were treated orally with HDM201 + ribociclib. Three treatment regimens were explored: Regimen (Reg) 1 (HDM201 + ribociclib daily for the first 2 wks [QD 1st 2 wks] in a 4-wk cycle), Reg 4 (HDM201 every 3 wks + ribociclib QD 1st 2 wks in a 3-wk cycle), and Reg 5 (HDM201 every 4 wks + ribociclib QD 1st 2 wks in a 4-wk cycle). Results: As of Nov 21, 2017, 74 pts received HDM201 + ribociclib (Reg 1 n=26; Reg 4 n=29; Reg 5 n=19); 12 pts (6 each in Reg 4 and 5) were still receiving treatment. Ten pts (Reg 1 n=3; Reg 4 n=6; Reg 5 n=1) discontinued treatment due to adverse events (AEs) and 2 pts (1 each in Reg 4 and 5) died. The most common AE of any grade, regardless of cause, reported across regimens (Reg 1; Reg 4; Reg 5) was nausea (81%; 76%; 63%) which was mainly Grade 1/2 and not dose limiting. Common Grade 3/4 AEs regardless of cause included neutropenia (39%; 52%; 42%), thrombocytopenia (35%; 45%; 42%), anemia (27%; 17%; 21%), leukopenia (27%; 28%; 37%), and lymphopenia (15%; 14%; 21%). Dose-limiting toxicities were reported in 16 pts (2; 9; 5) and all except 1 (prolonged QT) were hematologic (including neutropenia [n=5], thrombocytopenia [n=4], febrile neutropenia, and anemia [n=2 each]). Partial responses were observed in 3 (4%) pts (2 in Reg 4; 1 in Reg 5). Stable disease was achieved by 36 (49%) pts (11 in Reg 1; 16 in Reg 4; 9 in Reg 5). The median progression-free survival (PFS) was 2.7 mo (95% confidence interval [CI]: 1.9-8.2 mo) in Reg 1, 4.8 mo (95% CI: 3.9 mo-not reached) in Reg 4, and 2.1 mo (95% CI: 1.4 mo-not reached) in Reg 5. Conclusions: HDM201 + ribociclib demonstrated a manageable safety profile and preliminary efficacy in pts with locally advanced or metastatic WDLPS/DDLPS, with hematologic toxicities being dose limiting. The median PFS in Reg 4 compares favorably with single-agent CDK4 inhibitors and, alongside a tolerable safety profile, suggests further exploration of this regimen may be warranted in Phase II studies of HDM201 + ribociclib in this patient population. Citation Format: Albiruni Abdul Razak, Sebastian Bauer, Jean-Yves Blay, Richard Quek, Cristina Suárez, Chia-Chi Lin, Marie L. Hütter-Krönke, Ricardo Cubedo, Stephane Ferretti, Christophe Meille, Ensar Halilovic, Giorgia Clementi, Maria Santos-Rosa, Nelson Guerreiro, Astrid Jullion, Claire Fabre, Antoine Italiano. Results of a dose- and regimen-finding Phase Ib study of HDM201 in combination with ribociclib in patients with locally advanced or metastatic liposarcoma [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr CT009.

This study attempted to evaluate the usefulness of the International Prognostic Index (IPI) as a prognostic model in patients treated with R-CHOP (rituximab, cyclophosphamide, vincristine, adriamycin and prednisolone) chemotherapy. We compared 279 patients with DLBCL. Among them, 183 received CHOP while 96 received R-CHOP. Results showed that there were no statistically significant differences between the two groups of patients in terms of both the patient and the lymphoma characteristics. The estimated 2-year survival was significantly higher among patients treated with R-CHOP compared to CHOP alone (85.6% vs. 64.7%, P = 0.004). Both the IPI and age-adjusted IPI were less useful as prognostic models in patients receiving R-CHOP compared to CHOP. In the multivariate analysis, ageor= 60, elevated serum LDH, low serum albumin and advanced stages of disease were each independently associated with decreased survival in patients treated with CHOP. In contrast, among those treated with R-CHOP, only male sex and advanced stage of disease were each independently associated with decreased survival. Using these two factors, patients treated with R-CHOP could be separated into three prognostic groups with 5-year estimated survival ranging from 47% to 100% (P0.0001). In summary, we can conclude that with the significant improvement in survival following the use of rituximab, the relevance of previously recognized prognostic factors has to be reassessed and re-evaluated.

Background GI stromal tumours (GISTs) are clinically heterogenous exhibiting varying degrees of disease aggressiveness in individual patients. Objectives We sought to identify genetic alterations associated with high-risk GIST, explore their molecular consequences, and test their utility as prognostic markers. Designs Exome sequencing of 18 GISTs was performed (9 patients with high-risk/metastatic and 5 patients with low/intermediate-risk), corresponding to 11 primary and 7 metastatic tumours. Candidate alterations were validated by prevalence screening in an independent patient cohort (n=120). Functional consequences of SETD2 mutations were investigated in primary tissues and cell lines. Transcriptomic profiles for 8 GISTs (4 SETD2 mutated, 4 SETD2 wild type) and DNA methylation profiles for 22 GISTs (10 SETD2 mutated, 12 SETD2 wild type) were analysed. Statistical associations between molecular, clinicopathological factors, and relapse-free survival were determined. Results High-risk GISTs harboured increased numbers of somatic mutations compared with low-risk GISTs (25.2 mutations/high-risk cases vs 6.8 mutations/low-risk cases; two sample t test p=3.1×10 −5 ). Somatic alterations in the SETD2 histone modifier gene occurred in 3 out of 9 high-risk/metastatic cases but no low/intermediate-risk cases. Prevalence screening identified additional SETD2 mutations in 7 out of 80 high-risk/metastatic cases but no low/intermediate-risk cases (n=29). Combined, the frequency of SETD2 mutations was 11.2% (10/89) and 0% (0/34) in high-risk and low-risk GISTs respectively. SETD2 mutant GISTs exhibited decreased H3K36me3 expression while SETD2 silencing promoted DNA damage in GIST-T1 cells. In gastric GISTs, SETD2 mutations were associated with overexpression of HOXC cluster genes and a DNA methylation signature of hypomethylated heterochromatin. Gastric GISTs with SETD2 mutations, or GISTs with hypomethylated heterochromatin, showed significantly shorter relapse-free survival on univariate analysis (log rank p=4.1×10 −5 ). Conclusions Our data suggest that SETD2 is a novel GIST tumour suppressor gene associated with disease progression. Assessing SETD2 genetic status and SETD2 -associated epigenomic phenotypes may guide risk stratification and provide insights into mechanisms of GIST clinical aggressiveness.

The aim of this study was to characterize primary central nervous system lymphoma (PCNSL) among Asian patients and to determine their outcomes with different therapeutic modalities.Between 1990 and 2005, 37 patients with PCNSL were analyzed within 5 different treatment groups: radiotherapy alone (arm A), combined chemoradiation (arm B), chemotherapy alone with methotrexate (MTX)or = 1 g/m2 (arm C), miscellaneous therapy (arm D), and best supportive care (arm E).The median age at presentation was 59 years, and the majority of patients were male (68%). All patients had aggressive diffuse large B-cell lymphoma. The number of patients in arms A to E were 6, 16, 3, 8 and 4, respectively. The overall median survival was 7.4 months, 54.1 months, not reached, 8.9 months and 0.9 months, respectively. Use of MTX 1-2.5 g/m2 per cycle and an ECOG performance status of 0-2 were each associated with improved survival on univariate analysis (p = 0.022 and p = 0.049, respectively). Compared to radiotherapy alone, use of combined chemo-radiation was associated with a trend towards improved overall median survival (7.4 vs. 54.1 months, p = 0.058).The clinical characteristics and treatment outcomes in our Asian patients were comparable to those reported in Western series. Use of MTX 1-2.5 g/m2 per cycle and an ECOG performance status of 0-2 were associated with improved survival.

11049 Background: Recent studies suggest that markers of systemic inflammation such as blood neutrophil-to-lymphocyte ratio (NLR) may be prognostic for various cancers, though its clinical utility has not been widely accepted. This study aims to investigate its clinical relevance in patients (pts) with soft tissue sarcoma (STS). Methods: Five hundred and twenty-nine pts with localized STS who had available pre-operative blood counts at the time of diagnosis were retrospectively examined. An optimal cutoff for high NLR ( > 2.5) in predicting overall survival (OS) and relapse-free survival (RFS) in pts who underwent curative surgery (n = 473) was determined using receiver operating curve analyses. Cutoffs for platelet-lymphocyte ratios (PLR, > 180) and lymphocyte-monocyte ratios (LMR, < 3.6) were similarly obtained. Survival analysis was performed using the Kaplan-Meier method and multivariate Cox proportional models. Median follow-up was 40 months. Results: A high NLR was present in 311 (58.8%) pts, which was significantly associated with tumor grade ( p< 0.0001), depth ( p= 0.003) and size > 5 cm ( p= 0.0242), but not with age at diagnosis, sex or ethnicity. High NLR was associated with both worse OS (HR 1.78; 95%CI 1.28-2.47; p= 0.0005) and RFS (HR 1.54; 95%CI 1.17-2.03; p= 0.0019), as were age at diagnosis, tumor grade, size, PLR and LMR. In multivariate models adjusted for clinicopathological predictors of survival, only NLR, in addition to tumor grade and size, were independently associated with worse OS (HR 1.52; 95%CI 1.09-2.11; p= 0.0131) and RFS (HR 1.42; 95%CI 1.08-1.85; p= 0.0114). Analysis of survival according to American Joint Committee on Cancer (AJCC) stages subdivided as NLR-high and NLR-low revealed a significant worse prognosis for NLR-high subgroups ( p< 0.0001), with a 2.2-fold and 1.5-fold higher risk of death within stages II (HR 2.20; 95%CI 1.20-4.01; p= 0.0103) and III (HR 1.55; 95%CI 1.01-2.37; p= 0.0459), respectively. Conclusions: High NLR is an independent marker of poor prognosis among pts with localized STS. Inclusion of NLR as a classifier into the AJCC staging of STS may improve estimation of survival.

11046 Background: An increasing number and proportion of cancer patients with apparently localised disease are treated with chemotherapy and radiation therapy in contemporary oncology practice. In a pilot study of radiation-induced sarcoma (RIS) patients, we demonstrated that chemotherapy was associated with a reduced time to development of RIS. We now present an international multi-centre collaborative study to validate this association. Methods: This was a retrospective cohort study of RIS cases across five large international sarcoma centres between the 1st January, 2000 to 31st December, 2014. The primary endpoint was time to development of RIS, defined as the date of diagnosis of the first malignancy to date of the RIS diagnosis. We also assessed the relationship between chemotherapy, patient and cancer characteristics, and time to RIS. Results: We identified 419 patients with RIS, who were predominantly diagnosed with their first malignancy at adulthood. The median interval from the index cancer to development of RIS for the entire cohort was 11 years (range 1-64). Chemotherapy for the first malignancy was associated with a shorter time to RIS development (HR 1·37; 95% CI: 1·08-1·72; P = 0·009). In the multi-variable model, older age (HR 2·11; CI 1·83-2·43; P < 0·001) and chemotherapy for the first malignancy (HR 1·61; CI 1·26-2·05; P < 0·001) were independently associated with a shorter time to RIS. Anthracyclines and alkylating agents significantly contribute to the effect. Conclusions: This study confirms an association between chemotherapy given for the first malignancy and a shorter time to development of a RIS. Our data highlights the importance of vigilance in surveillance for RIS after chemotherapy and radiation therapy, particularly in younger patients who also have a longer potential time to develop a second malignancy.

11069 Background: Guidelines recommend routine surveillance imaging in patients (pts) following curative resection of STS. However the benefit of such an approach is unclear. We sought to evaluate the utility of a surveillance imaging strategy in pts with localized STS treated with curative intent. Methods: Pts with localized non-indolent STS, seen between 2010 – 2016, who had undergone surgery with R0/R1 surgical margins were included. Epidemiology, treatment and relapse data were collected as was the mode of detection. We defined optimal surveillance as CT/ MRI performed at least 6-mthly following surgery; suboptimal surveillance was defined as CT/ MRI imaging performed less frequently than 6mthly. Results: Of 294 pts included, 31% (n = 92) vs 34% (n = 100) vs 35% (n = 102) had optimal, suboptimal and no routine CT/MRI surveillance imaging respectively. At a median follow-up of 27mths (range 0-79), 36% (n = 105) experienced a relapse; 43% (n = 45) local and 57% (n = 60) had metastatic relapse. More relapses were noted in the optimal surveillance group, 57% (n = 52) vs 28% (n = 28) and 25% (n = 25) in the suboptimal and no surveillance groups respectively (p < 0.001). Within each cohort, relapses detected directly by routine surveillance imaging vs outside of surveillance imaging were as follows: 35% (n = 32) / 22% (n = 20) in the optimal, 17% (n = 17) / 11% (n = 11) in the suboptimal and 0 / 25% (n = 25) in the no surveillance arms respectively. Comparing the 3 strategies, the proportion of pts who then went on to receive curative resection/ metastacectomy was not significantly different, 38% (n = 20), 57% (n = 16) and 32% (n = 8) of relapses, in the optimal vs suboptimal vs no surveillance cohorts respectively (p = 0.1). Notably, routine surveillance imaging directly leading to curative resection occurred only in 15% (n = 14) of pts in the optimal and 9% (n = 9) in the suboptimal surveillance groups. Conclusions: While an intensive routine CT/MRI surveillance imaging strategy detected more recurrences, the impact it has on subsequent resection is less certain. Optimal frequency of surveillance imaging remains unclear.

Background: Angiosarcoma (AS) is an uncommon soft tissue sarcoma with dismal prognosis that presents either cutaneously (C-AS) or non-cutaneously (NC-AS). We compared the clinical features and treatment outcomes between these 2 groups. Methods: A single-centre study evaluating 60 AS patients between 2002 and 2012 was performed. Results: The median age was 70 years. C-AS of the scalp or face comprised 66% of patients. C-AS patients were older than NC-AS (median age 74 vs. 56 years; p < 0.001). Proportionately more C-AS patients presented with non-metastatic disease (86 vs. 50%; p = 0.007). Amongst resected C-AS and NC-AS patients, rates of positive surgical margins (53 vs. 50%; p = 1.00) and adjuvant therapy (25 vs. 43%; p = 0.626) were not significantly different, though proportionately fewer C-AS patients relapsed (36 vs. 78%; p = 0.038). Paclitaxel was the most common agent in first line palliative systemic therapy, achieving an objective response rate of 56%. Median overall survival was 11.2 months, with no significant difference between C-AS and NC-AS (11.3 vs. 9.8 months; p = 0.895). Conclusion: Distinct from AS in the West, our series demonstrates a clear preponderance of scalp AS. Disparities in clinical characteristics between C-AS and NC-AS did not translate into survival differences.

e19076Background: Patients with lymphoma experience fevers due to infections (with or without bacteremia), malignant fever or as a complication of cancer treatment. This study was designed to evalu...

Learning Objectives After completing this course, the reader will be able to: Compare concordance indices of the older and the revised FIGO staging systems.Identify gaps in the current FIGO staging system. CME This article is available for continuing medical education credit at CME.TheOncologist.com Background. In 2008, the Federation of Gynecology and Obstetrics (FIGO) revised their 1988 staging system for uterine leiomyosarcomas. In this article, we compare performance of the 2008 and 1988 FIGO systems. Methods. Individual case data were manually culled. Staging was retrospectively assessed according to revised and 1998 FIGO criteria. Overall survival distribution was assessed by the Kaplan-Meier method. Harrell's concordance index was used to assess the discriminative ability of a fitted Cox model to predict overall survival. Results. A total of 110 cases of uterine leiomyosarcomas were reviewed and data from 88 patients were analyzed. In all, 71% of cases were classified as stage I, 7% as stage II, 3% as stage III, and 19% as stage IV under the revised FIGO staging system. Nine patients (10.2%) were downstaged and none were upstaged. The revised FIGO system did not show a significant improvement over the 1988 FIGO system in the ability to discriminate the risk of death of patients between stages, with concordance indexes of 0.70 and 0.71, respectively. Most patients were classified as stage I with age, tumor grade, tumor size, and lymphovascular invasion as prognostic factors. Conclusion. The 2008 revised FIGO staging system for uterine leiomyosarcomas does not perform better than the 1988 system for uterine endometrial carcinomas. A better staging system is needed for these cases.

Well-established clinicopathological variables used in the risk stratification of gastrointestinal stromal tumor (GIST) may not completely predict rectal GIST, an uncommon and poorly studied GIST subset. The aim of the present study was to determine the patterns of relapse and morbidities associated with recurrence in rectal GIST. A single-institution retrospective study between 2002 and 2011 was conducted, identifying 9 patients (8%) with localized rectal GIST, while comparing small intestinal (n=37) and gastric (n=63) GIST (median age, 60 years). Rectal GIST tumors were smaller compared to small intestinal/gastric GIST (P=0.044). The number of mitoses per 50 high-power field (HPF) did not differ by primary site. In general, 73% of patients were high-risk, as defined by the National Institutes of Health (NIH) consensus criteria, however, only 25% received adjuvant imatinib. Fewer rectal GIST patients achieved negative surgical margins compared to small intestinal/gastric GIST (67 vs. 92%; P=0.054). Of the 9 patients with localized rectal GIST 6 had peri-operative tumor rupture, anastomotic breakdown or required anal sphincter-compromising surgery. At the time of the first relapse, 83% of the recurrences were local failures for rectal GIST, compared to 21% for small intestinal/gastric GIST (P=0.005). The median relapse-free survival was 51 months for the entire cohort, and 54, 36 and 56 months for rectal, small intestinal and gastric GIST, respectively (P=0.468). Rectal GIST was found to be associated with high rates of local relapse and significant morbidity, despite being significantly smaller compared to GIST of other sites. A multimodality peri-operative therapeutic approach may be required to improve outcomes.

Several pre-clinical studies report that statins interfere with the surface binding of rituximab to CD20. This study investigated the effects of statins in patients with diffuse large B-cell lymphoma (DLBCL) receiving rituximab-based chemoimmunotherapy, and the impact of commonly used drugs, metformin and aspirin, on the clinical outcomes of patients receiving chemoimmunotherapy. We included 213 patients with DLBCL who received rituximab-based chemoimmunotherapy. Details of statin, metformin, and aspirin use and initiation of chemoimmunotherapy were abstracted from medical records. All patients received rituximab, and 47 (22.1%) were taking statins. The median age of patients receiving statins was significantly higher compared to those who did not (p 0.001). Response rates between patients receiving and not receiving statins were not significantly different (85.1% vs. 87.3%; p = 0.688). Event-free survival (EFS) was not significantly different (p = 0.352). Overall survival was lower in patients receiving statins compared to those who did not (p = 0.036). However, it was no longer significant after adjustment for age (p = 0.140). Metformin had no impact on the response rate (p = 0.268), EFS (p = 0.574), and overall survival (p = 0.141). Aspirin had no impact on the response rate (p = 0.784), EFS (p = 0.836), and overall survival (p = 0.779). Statins do not interfere with rituximab, and need not be withheld during rituximab administration. Larger studies are needed to confirm the impact of metformin and aspirin on patients with DLBCL receiving chemoimmunotherapy.

This retrospective study aimed to evaluate the clinical characteristics and prognostic factors of Asian patients with primary mediastinal large B-cell lymphoma (PMBCL) and to determine the role of rituximab in this entity. Forty-one consecutive patients from 1997 to 2009 were included: 14 received CHOP chemotherapy (cyclophosphamide, doxorubicin, vincristine, prednisolone), while 27 more recently treated patients received CHOP with rituximab (R-CHOP). All patients with a complete or partial response received consolidation involved field radiotherapy (RT). After a median follow-up of 31.2 months (104.4 months for CHOP and 28.8 months for R-CHOP), the overall survival (OS) and progression-free survival (PFS) for R-CHOP- and CHOP-treated patients were 87% vs. 57% and 88% vs. 36%, respectively. R-CHOP resulted in an improvement of PFS (hazard ratio [HR] 8.27, 95% confidence interval [CI] 2.23-30.74, p = 0.002) and OS (HR 4.20, 95% CI 1.05-16.8, p = 0.04). Nineteen patients had positron emission tomography/computed tomography (PET/CT) evaluation after six cycles of R-CHOP (metabolic complete response 13, partial metabolic response five, and metabolic progression one). All five patients with a metabolic partial response received RT instead of intensive salvage chemotherapy; four remained progression-free. In patients with PMBCL, R-CHOP in combination with involved field radiotherapy portended a 3-year OS rate of 87%, which is comparable to historical survival rates with more intensive chemotherapy regimens.

Purpose: Preclinical models demonstrate synergistic antitumor activity with combination blockade of mTOR and IGF-1R signaling. We aimed to determine the safety, tolerability, and recommended phase II dose (RP2D) of the combination of figitumumab, a fully human IgG2 anti-insulin-like growth factor-1 receptor (IGF-1R) monoclonal antibody (Pfizer) and the mTOR inhibitor, everolimus (Novartis). Pharmacokinetics and preliminary antitumor effects of the combination were evaluated. Experimental Design: Phase I trial in patients with advanced sarcomas and other solid tumors. Initial cohort combined full phase 2 dose figitumumab (20 mg/kg IV every 21 days) with full dose everolimus (10 mg orally once daily). Intercohort dose de-escalation was planned for unacceptable toxicities. Dose modifications were allowed beyond cycle 1. Results: No DLTs were observed in the initial cohort during cycle one, therefore full dose figitumumab and everolimus was declared the RP2D. In total, 21 patients were enrolled on study. Most toxicities were grade 1 or 2, and were similar to reported toxicities of the single agents. Mucositis was the most frequently observed grade 3 toxicity. Median time on study was 104 days (range 17–300). Of 18 patients evaluable for response, best response was partial response in 1 patient with malignant solitary fibrous tumor and, stable disease in 15 patients. There were no apparent pharmacokinetic interactions between everolimus and figitumumab. Conclusions: Combination figitumumab plus everolimus at full doses appears safe and well tolerated with no unexpected toxicities. Dose reductions in everolimus may be required after prolonged drug administration. This regimen exhibits interesting antitumor activity warranting further investigation. Clin Cancer Res; 17(4); 1–9. ©2010 AACR.

According to several prognostic indices and studies, age ≥ 60 has consistently been described as an independent poor prognostic factor for survival for patients with lymphoma [1-3]. With improvemen...

10519 Background: Primary and secondary mutations in GIST affect prognosis and treatment response and impacts treatment decision making. A single driver mutation (primary) is found amongst exons of...

Background Trabectedin and gemcitabine/docetaxel are routinely used in the treatment of advanced sarcoma after failure of anthracycline and/or ifosfamide. The optimum regimen remains unclear. We aimed to assess the efficacy and safety of trabectedin versus gemcitabine/docetaxel in patients with sarcoma after failure of standard therapy. Methods Twenty-nine patients with advanced sarcoma who had received trabectedin or gemcitabine/docetaxel after failure of anthracycline and/or ifosfamide were included in this single-centre study. Three patients received both regimens and were included in the initial treatment group for the purpose of this analysis. Per institutional guidelines, imaging studies were repeated after 2–3 cycles of treatment; response rates and drug-emergent toxicity were evaluated using Response Evaluation Criteria in Solid Tumors (RECIST) (version 1.1) and CTACE (version 4), respectively. Findings Sixteen and 13 patients received trabectedin and gemcitabine/docetaxel, respectively. Median age (53 versus 56 years) and sex (44% versus 46% males) were not significantly different between the two groups. Patients in both cohorts had previously received a median of 1 line of chemotherapy (range 1–3). A median of two cycles were administered in both cohorts (range 1–10 and 1–6, respectively). Median administered dose of trabectedin was 1.2 mg/ m 2 (range 0.9–1.5 mg/ m 2 ) and, per institutional practice, gemcitabine was administered routinely at 900 mg/ m 2 on days 1 and 8 and docetaxel at 75 mg/ m 2 day 8, supported by growth factors. With the exception of self-limiting grade 3/4 ALT elevations in 50% of patients in the trabectedin group versus none in the gemcitabine/docetaxel ( p = 0.003), both treatments were generally well tolerated with no significant differences in toxicity profiles. Objective response rates were 6% (trabectedin) versus none (gemcitabine/docetaxel). Overall median PFS was 1.5 months with no significant difference in the two groups (1.4 versus 1.9 months, respectively); 31% versus 15% of patients in the trabectedin and gemcitabine/docetaxel groups, respectively, did not have disease progression at 4 months. Interpretation Encouraging anti-tumour activity was noted in both treatment groups with subsets of patients achieving prolonged clinical benefit. Prospective studies are urgently needed to clarify the optimum strategy in this clinical setting. Funding National Cancer Centre Research Fund and Janssen Singapore. The authors declared no conflicts of interest.

8543 Background: The role of routine surveillance imaging (RSI) in patients with peripheral T-cell lymphoma (PTCL) in first complete remission (CR1) is unclear. RSI offers the theoretical advantage...

10051 Background: STS in geriatric pts is not well studied. We evaluated prognostic factors and clinical outcomes in elderly pts with HG STS. Methods: Single centre retrospective study. G-HG STS pts defined as age > 65 yrs seen in our centre from 2002 - 2011 with complete medical records were identified. Charlson age-comorbidity score was assessed for each pt. Results: 116 pts from 4 most common HG STS histo-types representing 69% of pts in the geriatric STS cohort were analysed; leiomyosarcoma (LMS, 14%), non well-differentiated liposarcoma (nWD-LPS, 9%), angiosarcoma (AS, 30%), and undifferentiated pleomorphic sarcoma (UPS, 47%). Median age was 72 yrs, 81% presented with localised disease; of 78% of these localised pts who had curative surgery, 49% received adjuvant therapy, of whom 92% had radiotherapy (RT) only. AS arises more commonly from the head/neck region (p

8078 Background: Current methods used to prognosticate patients with follicular lymphoma (FL) include the Follicular Lymphoma International Prognostic Index (FLIPI) and tumor grade. However, they do not provide information on the biological and molecular features of FL. Recent data suggest that high tumor-associated macrophages (TAM) content may be an adverse prognostic factor in FL. However, TAM consists of two main subtypes, M1 and M2, with the former related to pro-inflammatory properties while the latter, anti-inflammatory functions. Currently, the prognostic impact of each individual subtype has not been elucidated and this may be more important than looking at TAM alone. Methods: Tumor specimens of 98 patients with FL diagnosed between 2005 and 2009 were investigated using immunohistochemistry and fluorescence in-situ hybridization (FISH) using break-apart FISH probes targeting BCL2, BCL6, MYC and IgH genes. Tumor specimens expressing a higher proportion of CD68+/CD163- in one high power field were defined as M1 subtype and those that expressed a greater proportion of CD68+/CD163+ were denoted as M2 subtype. Results: Amongst the 98 patients, 60 (61%) were of the M2 subtype, 59% presented with advanced disease, 47% were grade 3 and the median age was 59 years (21 – 88 years). Baseline characteristics including grade, stage and recurrent translocations did not significantly differ between the M1 and M2 groups. Similarly, there was also no difference in terms of treatment received (both Rituximab and chemotherapy) between the two groups. Despite these similarities, with a median follow-up of 3.2 years, the 3-year overall survival (OS) was significantly different (M1, 100% vs M2, 85%; p < 0.01). All 11 patients who died were of the M2 subtype (p=0.009). Conclusions: Our findings show that despite similar patient characteristics and treatment, there was, however, a statistically significant difference in OS, with the M2 subtype demonstrating an inferior outcome. The M2 TAM subtype rather than TAM alone may be a more important prognostic marker in FL, which could possibly be explained by its known biological anti-inflammatory function.

1593 Background: Recently, several studies evaluating the effect of ultraviolet light exposure on the risk of developing lymphoma have yielded conflicting results, in part due to seasonal and latitude variation. Further, most of these studies also lack long term follow up and could not systemically address the influence of sunlight exposure on survival. In this prospective, epidemiological study, we explored the relationship between past sunlight exposure and survival in lymphoma patients in tropical Singapore, where sunlight exposure is constant throughout the year. Methods: Singapore Lymphoma Study was established in 2005. A total of 384 patients with histologically confirmed lymphoma were followed from date of diagnosis through November 2011. Interviewers were trained to conduct in-person interviews to collect detailed information, including every occupational period that lasted at least 1 year and number of hours worked outdoor from 9am to 5pm using a structured questionnaire. Descriptive, multivariate and survival analysis (Kaplan Meier; age, gender, race and education level adjusted) were performed. Results: Among all lymphoma subtypes, the most prevalent was diffuse large B cell lymphoma, 47.7%, followed by Hodgkin lymphoma, 14.6%, follicular lymphoma, 12.2% and T cell lymphoma, 6.5%. 97.1% had black or brown eye color and 49.6% had fair skin. Overall survival of lymphoma and NHL subtype decreased significantly with increasing lifetime occupational hours (ptrend=0.01 and 0.03). Chinese was associated with longer survival with increasing daily sunlight exposure during childhood and adolescent years, whether at school or outdoor leisure (ptrend=0.04) and there were no evidence for non-Chinese (Malays or Indians). Patients with lighter eye color results in shorter survival (HR=2.91, 95% CI: 1.15-7.34, p=0.02). Conclusions: We found leisure sunlight exposure during childhood and adolescent years to be beneficial to the survival of Chinese lymphoma patients and increasing lifetime cumulative occupational exposure is associated with increasing risk of death. This data contributes to the growing research on the effects of sunlight exposure on lymphoma patients.

e18546 Background: We described EBV & HHV8 co-infection rates, histology subtypes (particularly T-cell), prognosticators and outcomes of Asian ARL and compared them to Western series. Methods: We studied clinicopathologic features of 46 HAART-era ARL diagnosed from 1998-2011 in an Asian Tertiary Cancer Centre. To study HIV effect on survival, we applied multivariate analysis to HIV-DLBCL matched against 453 de novo DLBCL from 2000-2008. Results: The 46 patients’ characteristics are presented in the table below. Of note, germinal centre B-cell (GCB) subtype predominated in HIV-DLBCL with evaluable tissue. In multivariate analysis of HIV-DLBCL and matched de novo DLBCL, HIV status did not predict survival (HR1.33, p=0.63); CR was the only significant prognosticator (HR0.30, p=0.048). Conclusions: 1. EBV+ rates appeared to surpass prior Western ARL series. Although concomitant EBV infection appeared to correlate with inferior survival (HR34.5), it was not statistically significant. 2. Alb

Background Trabectedin has recently been approved for treatment of soft tissue sarcoma (STS) and has shown interesting anti-tumour activity in a wide spectrum of bony sarcomas. Published literature about its use in Asian patients is limited. Methods Twenty consecutive patients with histology-confirmed sarcoma treated with trabectedin at our institution were included in this single-centre retrospective study. Per institutional guidelines, imaging studies were done after 2–3 cycles of treatment; response rates and drug-emergent toxicities were evaluated using Response Evaluation Criteria in Solid Tumors (RECIST) (version 1.1) and CTACE (version 4), respectively. Findings Median age was 57 years (range, 22–72), 60% of patients were female and 95% had STS. Common histotypes included undifferentiated pleomorphic sarcoma (UPS, 30%), leiomyosarcoma (LMS, 20%), and liposarcoma (LPS, 15%). Eighty-five per cent of patients had bulky primary tumours (>5 cm); the lung was the most common metastatic site (44%). Patients had previously been given a median of 1 line of chemotherapy (range 0–4). A total of 59 cycles of trabectedin was administered with a median dose of 1.2 mg/ m 2 (range 0.6–1.5 mg/ m 2 ) and a median of two cycles (range 1–10). At a median follow-up of 4 months, 2 patients were still on drug treatment; the most common reason for discontinuing trabectedin was disease progression. Objective response rate was 5% (patient with LMS) and 25% of patients achieved stable disease as the best response. Three patients (15%) received more than six cycles of trabectedin; 20% of patients did not have evidence of disease progression at 6 months (2 patients with LPS, 1 each with LMS and UPS). Median PFS was 1.5 months. Trabectedin was generally well tolerated; the most common grade 3/4 toxicity was self-limiting ALT/AST elevations. Interpretation Trabectedin is safe and well tolerated in Asian patients with sarcoma. A significant subset of patients achieved prolonged clinical benefit. Biomarker development is urgently needed to better identify this group of patients. Funding National Cancer Centre Research Fund and Janssen Singapore. The authors declared no conflicts of interest.

10095 Background: LMS is a narrowly defined subset of soft tissue sarcoma with heterogeneous clinical presentation. Recent reports suggest differences in treatment responses between uterine (uLMS) and extrauterine (euLMS) disease. We sought to identify prognostic factors in LMS patients (pts) and study the impact of primary site on clinical outcomes. Methods: Single centre retrospective study. 97 eligible LMS pts seen in our centre from 2002 - 2010 with complete medical records were identified from our sarcoma database of 790 pts. Results: Median follow-up was 20.7 months (mths), 51% of pts had uLMS. Although median age in uLMS and euLMS was similar (49 vs 54 yrs), uLMS was less common beyond 60 yrs (10% vs 38%, p=0.002). Surgical resection rates (92% vs 79%), adjuvant therapy use (56% vs 42%) and metastatic disease at diagnosis (25% vs 27%) were not significantly different between uLMS and euLMS respectively. Amongst pts who underwent resection of disease, more uLMS pts had negative surgical margins (90%...

10091 Background: Although the entity of MC-AS is well recognized, non MC-AS is less well reported and differences between the two are not well studied. We aimed to compare the clinical presentatio...

TO THEEDITOR:Wereadwithinterestthismeta-anaylsisbyZiepert etal 1 involving1,062patientswithdiffuselargeB-celllymphomaaccrued fromthreeprospectivephaseII/IIItrials:MinT(Mab-TheraInternational Trial), RICOVER-60 (cyclophosphamide, doxorubicin, vincristine, and prednisone plus rituximab [R-CHOP] for patients older than age 60 years),andMegaCHOEP(dose-escalatedregimenofcyclophosphamide, doxorubicin,vincristine,etoposide,andprednisone)trials. 2-4 Theauthors affirms the prognostic relevance of the International Prognostic Index (IPI)scoreforallthreeendpointsofprogression-freesurvival,event-free survival,andoverallsurvival.Thus,theyconcludedthatIPIshouldremain the major tool for risk stratification for patients with diffuse large B-cell lymphoma in the era of rituximab. WhileweagreewithZiepertetal 1 thattheIPIwillremainanimpor

10002 Background: Preclinical models demonstrate synergistic antitumor activity with combination blockade of mTOR and IGF-IR signaling. We aimed to determine the tolerability and recommended phase ...

10091 Background: Interstitial hypertension may contribute to chemo-resistance in STS. Sorafenib is a small molecule tyrosine kinase inhibitor with potent activity against c-RAF/B-RAF, VEGFR-2, and...

14089 Background: Effective 1st line chemotherapy that is easily delivered and tolerable is important in patients with advanced gastric cancer (AGC). This study aimed to evaluate the toxicity and efficacy of combination capecitabine and oxaliplatin (XELOX), given 3 weekly, in patients with AGC. Methods: Twenty-seven patients with AGC, were treated with XELOX at the National Cancer Centre Singapore between June 2003 and December 2005. Capecitabine was given 2000 mg/m2/day, days 1–14, and oxaliplatin was given 130 mg/m2 on day 1, every 21 days. Results: The median age was 61 years with 22% of patients ≥ 70 years. One-third of patients had ECOG 2 - 4, and 33% of patients had signet ring cell histology. Median doses of capecitabine and oxaliplatin achieved were 1500 mg/m2/day (1200 - 2400) and 130mg/m2 (80 - 130) respectively. A median of 5 cycles (range 1 - 8) of XELOX was administered. All 27 patients had toxicity assessment. Toxicities were infrequent and mild. There were 3 cases of grade 3/4 toxicities (thrombocytopenia, neutropenia and fatigue). Twenty-two patients had measurable disease for response evaluation. Complete (CR) and partial (PR) response rates were 9 and 50% respectively. Overall response rate was 59%. Chemo-naïve patients had better responses (80% vs 25%). Four of the remaining 5 patients had reduction in non-measurable disease or in tumor marker levels. Of the 9 patients with ECOG > 1, there was radiological CR in 1 patient and 5 PR; 1 had normalization of CEA, another reversal of disseminated intravascular coagulation and the third decreased ascites. The median time to progression and overall survival were 5.9 months and 7.6 months respectively. Conclusions: XELOX regimen is safe, active and well tolerated in AGC even in patients with poor ECOG status. It is easily administered without requiring a central venous catheter. This regimen should be investigated further in a prospective manner. No significant financial relationships to disclose.