Search

Lipid metabolism impacts immune cell differentiation, activation, and functions, modulating inflammatory mediators, energy homeostasis, and cell membrane composition. Despite preclinical evidence, data in humans lack concerning tumors and immunotherapy (IO). We aimed at investigating the correlations between circulating lipids and the outcome of non-small cell lung cancer (NSCLC) patients treated with IO.We identified all patients with advanced NSCLC treated with IO at our Institution with available baseline plasma samples. Fatty acids (FAs) were analyzed through gas chromatography. Survival curves were estimated by the Kaplan-Meier method. Cox multivariate models were constructed through a stepwise procedure, with entry and exit P value set at .2.We identified 112 patients, mostly with performance status 1 (65.2%) and PD-L1≥1% (75.3%). Median progression-free survival (PFS) and overall survival (OS) were 2.8 and 11.0 months, respectively. Multivariable model for survival identified a positive association of circulating free (FFA) C16:0 (P .005) and esterified (EFA) C16:1 (P .030) with PFS, and a positive association of EFA C16:1 (P .001) and EFA C18:0 (P .020) with OS. EFA C16:0 was negatively associated with PFS (P .008).FFA C16:0 and FAs derived from its unsaturation (EFA C16:1) and elongation (EFA C18:0) are associated with a better outcome in NSCLC patients treated with IO. It is conceivable that the ratio among those FAs may modify membrane fluidity and receptor activity, influencing IO efficacy. These data pave the way for the investigation of lipid-modulating strategies in association with IO in NSCLC.

International guidelines recommend severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) vaccine for patients with cancer. A substantial risk of developing vaccine-related autoimmune toxicities could be hypothesised for patients with thymic epithelial tumours (TETs) due to their high risk of autoimmune disorders (ADs). Moreover, a cross-reaction between SARS-CoV-2 spike protein antibodies and various tissue proteins has been shown, and antibodies against nucleoproteins showed overlaps in the autoimmune cross-reaction with antibodies to spike protein. Due to the rarity of TETs, no data addressing this hypothesis are available.Patients with TETs who received SARS-CoV-2 vaccine, treated in 4 referral centres of the Italian Collaborative Group for ThYmic MalignanciEs (TYME) network between February 2021 and September 2021, were interviewed through a standardised 15-items questionnaire in order to describe the safety of SARS-CoV-2 vaccine in patients affected by TETs.Data from 245 doses of vaccine administered to 126 patients (41 = thymic carcinoma, 85 = thymoma; 38 with AD, of which 26 with active AD) were collected. Nine patients had a previous COVID-19-positive swab. No cases of AD reactivation or worsening of a pre-existing AD were seen in the study population. A new diagnosis of myasthenia gravis likely unrelated to the vaccine was made in two patients after the vaccination. Sixty-four patients (51%) experienced a total of 103 adverse events, all G1/G2, most commonly fatigue, new or worsening muscle pain and chills. None AE required patients' hospitalisation.SARS-CoV-2 mRNA vaccines appear to be safe in patients with TET, even in case of active or pre-existing AD.

Background: Hyperprogressive disease (HPD) has been described in ≃14-25% of pretreated non-small cell lung cancer (NSCLC) patients upon single-agent (SA) PD-1/PD-L1 inhibitors (ICI) and has not been reported upon platinum-based chemotherapy (PCT) and ICI combinations. So far, no predictive biomarkers are available for HPD early detection. Methods: NSCLC patients treated with 1st line SA-ICI or PCT-ICI were assessed for HPD and circulating neutrophils. HPD was defined as delta tumor growth rate (TGR) >50% and/or TGR ratio ≥2. Circulating low density neutrophils (LDNs) were assessed by flow cytometry on peripheral blood mononuclear cells (PMBCs). LDNs were defined as CD66b+CD15+ cells among CD11b+ PBMCs and immature subtypes as CD10− LDNs. The LDNs predictive role was assessed by penalized model-based tests. Results: 144 NSCLC patients were included: 75 treated with SA-ICI, 69 with PCT-ICI. In the SA-ICI cohort, HPD occurred in 8 (11%) patients, while progressive disease (PD) and response or stable disease (PR/SD) occurred in 33 (44%) and 34 (45%) of patients respectively. Immature circulating CD10− LDNs were significantly higher in baseline blood samples of HPD patients [median (ME): 39.3, interquartile range (IQR): 28.7] compared to PD [ME: 7.4, IQR: 14.9, p Conclusions: Baseline circulating immature neutrophils characterize HPD upon 1st line SA-ICI and a 30.5% cut-off of immature neutrophils could select NSCLC patients to be addressed to PCT-ICI combinations. Citation Format: Roberto Ferrara, Giuseppe Lo Russo, Chiara Maura Ciniselli, Annamaria Piva, Barbara Bassani, Elena Jachetti, Giuseppina Calareso, Valeria Duroni, Settimio Di Gregorio, Claudia Proto, Arsela Prelaj, Alessandro De Toma, Mario Occhipinti, Marta Brambilla, Sara Manglaviti, Laura Mazzeo, Arturo Rinaldi, Teresa Beninato, Monica Ganzinelli, Filippo De Braud, Marina Chiara Garassino, Paolo Verderio, Mario Paolo Colombo, Sabina Sangaletti. Circulating immature neutrophils early detect hyperprogressive disease upon first-line PD-1/PD-L1 inhibitors in non-small cell lung cancer patients selecting best candidates for platinum-based chemotherapy and PD-1/PD-L1 inhibitors combinations. [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 5506.

Patients affected by inoperable lung cancer, due to great extension or to the presence of metastases, are currently treated with intravenous drugs that act on immune system activation alone or in combination with chemotherapy as first-line treatment. The characteristics of these patients (both their medical history and their blood exams) need to be studied to find out if some of them can help clinicians to predict if they will benefit from the combination of immunotherapy with chemotherapy. The authors collected the data of patients with advanced lung cancer treated in their hospital and found out that a value calculated from their blood exams, collected before the start of treatment and a combination of values named EPSILoN score (which considers patients' clinical condition, their history of tobacco smoking, the presence of metastases in the liver and two blood exam parameters, namely the neutrophil-to-lymphocyte ratio and LDH level) can predict how their disease will evolve during first-line treatment with chemotherapy in combination with immunotherapy.

9057 Background: Immune checkpoint inhibitors (IO) single agent or in combination with platinum chemotherapy (CT-IO) are standard of care for Stage IV non-small cell lung cancer (NSCLC) according to PD-L1 expression. While the efficacy of IO among patients (pts) with common EGFR and ALK alterations appears to be limited, its activity in pts with novel oncogenic drivers alterations is not well characterized. Compared to non-oncogene-addicted NSCLC, the overall response rate (ORR) seems to be similar in BRAF and c-MET altered NSCLC, lower in RET altered NSCLC, while data are less consistent in HER2 and EGFR exon 20 (EGFRex20) altered NSCLCs. Methods: From January 2016 to January 2022, we retrospectively enrolled pts with Stage IV NSCLC that received IO or combination CT– IO in any line, ECOG PS 0 - 2 and detection of MET exon 14 skipping mutations (METex14), BRAF mutations (V600E or non-V600E), RET rearrangement, HER2 point mutations (HER2mut)/exon 20 insertions (HER2ex20) or uncommon EGFR mutations (uEGFRmut)/EGFRex20. A review of clinicopathologic and molecular features and an analysis of response to combination or single-agent IO were conducted. Results: Among sixty-four pts enrolled, 20 (31%) had METex14, 19 (30%) had EGFR alterations [12 (19%) EGFRex20, 7 (11%) uEGFRmut], 8 (12%) had BRAF mutation (3 V600E and 5 non-V600E), 13(20%) had HER2 alterations [7 (11%) HER2ex20, 6 (10%) HER2mut] and 4 (6%) were RET rearranged. 43 received IO single agent and 21 received CT-IO. With a median follow up of 22 months (m), median progression free survival (mPFS) was 5.40 m (0.95 CI 4.73-6.9) overall, 6.77m in CT-IO arm (0.95 CI 5.37-NA) and 5.10m in IO arm (0.95 CI 2.60-6.7), with a trend to better mPFS for CT-IO (p 0.054). Regarding specific mutations irrespectively from treatment arm, NSCLC harboring METex14 showed a mPFS of 5.33 m (0.95 CI, 2.30-13.9), BRAF 9.9 m (0.95 CI, 6.70-NA), EGFR 4.93 m (0.95 CI, 1.80-6.9), HER2 11.4 (0.95 CI, 4.2-NA), RET 5.28 (0.95 CI, 1.42-NA). Disease control rate (DCR) was better in the CT-IO arm vs IO one in the overall population (84.2% vs 50%, p 0.013). Conclusions: Novel driver alterations seem to show a benefit from IO treatments. CT-IO seems to have a better outcome in terms of DCR. Therefore, IO-based treatment should be evaluated also in tumors harboring novel driver alterations. [Table: see text]

Immune-checkpoint inhibitors (ICIs) have significantly improved outcome of advanced non-small cell lung cancer (aNSCLC) patients. However, their efficacy remains uncertain in uncommon histologies (UH).Data from ICI treated aNSCLC patients (April,2013-January,2021) in one Institution were retrospectively collected. Univariate and multivariate survival analyses were estimated by Kaplan-Meier and Cox proportional hazards regression model, respectively. Objective response rate (ORR) and disease control rate (DCR) were assessed.Of 375 patients, 79 (21.1%) had UH: 19 (24.1%) sarcomatoid carcinoma, 15 (19.0%) mucinous adenocarcinoma, 10 (12.6%) enteric adenocarcinoma, 8 (10.1%) adenocarcinoma not otherwise specified, 7 (8.9%) large-cell neuroendocrine carcinoma, 6 (7.6%) mixed histology non-adenosquamous, 5 (6.3%) adenosquamous carcinoma, 9 (11.4%) other UH. In UH group, programmed death-ligand 1 (PD-L1)1%, 1-49%, ≥50% and unknown expression were reported in 27.8%, 22.8%, 31.7% and 17.7% patients respectively and ICI was the second/further-line in the majority of patients. After a median follow-up of 35.64 months (m), median progression-free survival (mPFS) was 2.5 m in UH [95% CI 2.2-2.9 m] versus (vs.) 2.7 m in CH [95% CI 2.3-3.2 m, P-value = .584]; median overall survival (mOS) was 8.8 m [95% CI 4.9-12.6 m] vs. 9.7 m [95% CI 8.0-11.3 m, P-value = .653]. At multivariate analyses only ECOG PS was a confirmed prognostic factor in UH. ORR and DCR were 25.3% and 40.5% in UH vs. 21.6% and 49.5% in CH [P-value = .493 and .155 respectively].No significant differences were detected between UH and CH groups. Prospective trials are needed to understand ICIs role in UH population.

Background Radically resected early uterine leiomyosarcoma (eULMS) is still marked by a poor prognosis. Adjuvant strategies investigated up to now have not been corroborated by controlled studies. We retrospectively reviewed the clinical outcome of eULMS patients treated with adjuvant anthracycline-based or gemcitabine-based chemotherapy at two Italian reference centers. Methods In this explorative, retrospective, cohort analysis, we included all the consecutive patients with radically resected eULMS treated at two centers between 1997 and 2017. Results A total of 109 consecutive patients were included. Sixty-six (60%) received an anthracycline-based regimen, whereas 43 (40%) received a gemcitabine-based regimen. Median disease-free survival (DFS) was 41.3 months with anthracycline-based regimens compared to 20.9 months with gemcitabine-based regimens (HR: 0.49; 95% CI: 0.30–0.80; P = 0.004). In the multivariable model, anthracycline-based regimens were independently associated with a better DFS. No difference in terms of overall survival was observed. Conclusions DFS was not the same by using an anthracycline-based or a gemcitabine-based adjuvant chemotherapy for patients with radically resected eULMS. The results of our study are in line with recent prospective controlled evidence in limb and superficial trunk soft tissue sarcomas. The role of anthracycline-based adjuvant chemotherapy should still be viewed as a research issue in eULMS.

Exocrine pancreatic neoplasms represent up to 95% of pancreatic cancers (PCs) and are widely recognized among the most lethal solid cancers, with a very poor 5-year survival rate of 5%-10%. The remaining < 5% of PCs are neuroendocrine tumors that are usually characterized by a better prognosis, with a median overall survival of 3.6 years. The most common type of PC is pancreatic ductal adenocarcinoma (PDAC), which accounts for roughly 85% of all exocrine PCs. However up to 10% of exocrine PCs have rare histotypes, which are still poorly understood. These subtypes can be distinguished from PDAC in terms of pathology, imaging, clinical presentation and prognosis. Additionally, due to their rarity, any knowledge regarding these specific histotypes is mostly based on case reports and a small series of retrospective analyses. Therefore, treatment strategies are generally deduced from those used for PDAC, even if these patients are often excluded or not clearly represented in clinical trials for PDAC. For these reasons, it is essential to collect as much information as possible on the management of PC, as assimilating it with PDAC may lead to the potential mistreatment of these patients. Here, we report the most significant literature regarding the epidemiology, typical presentation, possible treatment strategies, and prognosis of the most relevant histotypes among rare PCs.

Introduction: In advanced Non-Small Cell Lung Cancer (NSCLC), Programmed Death Ligand 1 (PD-L1) remains the only used biomarker to candidate patients to immunotherapy (IO) with many limits. Given the complex dynamics of the immune system it is improbable that a single biomarker could be able to profile prediction with high accuracy. A promising solution cope with this complexity is provided by Artificial Intelligence (AI) and Machine Learning (ML), which are techniques able to analyse and interpret big multifactorial data. The present study aims at using AI tools to improve response and efficacy prediction in NSCLC patients treated with IO.MethodsReal world data (clinical data, PD-L1, histology, molecular, lab tests) and the blood microRNA signature classifier (MSC), which include 24 different microRNAs, were used. Patients were divided into responders (R), who obtained a complete or partial response or stable disease as best response, and non-responders (NR), who experienced progressive or hyperprogressive disease and those who died before the first radiologic evaluation. Moreover, we used the same data to determine if the overall survival of the patients was likely to be shorter or longer than 24 months from baseline IO. For A literature review and forward feature selection technique was used to extract a specific subset of the patients’ data. To develop the final predictive model, different ML methods have been tested, i.e., Feedforward Neural Network (FFNN), Logistic Regression (LR), K-nearest neighbours (K-NN), Support Vector Machines (SVM), and Random Forest (RF).Results 200 patients were included. 164 out of 200 (i.e., only those patients with PD-L1 data available) were considered in the model, 73 (44.5%) were R and 91 (55.5%) NR. Overall, the best model was the LR and included 5 features: 2 clinical features including the ECOG performance status and IO-line of therapy; 1 tissue feature such as PD-L1 tumour expression; and 2 blood features including the MSC test and the neutrophil-to-lymphocyte ratio (NLR). The model predicting R/NR of the patient achieves accuracy ACC= 0.756, F1 score F1=0.722, and Area Under the ROC Curve AUC=0.82. The use of the PD-L1 alone has an ACC=0.655. The accuracy of the ML models excluding some of the features from the model were as follow: without PD-L1 value (ACC=0.726), MSC (ACC=0.750), and both PD-L1 and MSC (ACC=0.707), i.e., considering only clinical features. At data cut-off (Nov 2020), median Overall Survival (mOS) for R was 38.5 months (m) (95%IC 23.9 - 53.1) vs 3.8 m (95%IC 2.8 - 4.7) for NR, with pConclusionsThe results suggest that the integration of multifactorial data provided by ML techniques is a useful tool to improve personalized selection of NSCLC patients candidates to IO. In particular, compared to PD-L1 alone the expected improvement was around 10%. In particular, the model shows that the higher the ECOG, NLR value, IO-line, and MSC test level the lower the response, and the higher PD-L1 the higher the response. Considering the difference in survival among R and NR groups, these results suggest that the model can also be used to indirectly predict survival. Moreover, a second model was able to predict long survival patients with good accuracy.

BackgroundHyperprogression (HPD) has been described in ≃14–26% of NSCLC patients upon single-agent ICI1 and has not been reported upon ICI and platinum-based chemotherapy (PCT) combinations. Both high circulating neutrophils2 and senescent T-cells3 correlated with HPD, however the exact neutrophils-T-cells interplay and the role of specific neutrophils subsets in driving HPD is unknown.MethodsNSCLC patients treated with 1st line ICI as single-agent or in combination with PCT were assessed for HPD and circulating neutrophils’ phenotype. HPD required 3 assessment (2 before ICI, 1 upon ICI) and was defined as delta tumor growth rate (TGR) (TGR upon ICI – TGR before ICI) >50% and/or TGR ratio (TGR upon ICI/TGR before ICI) ≥2. Circulating low density neutrophils (LDNs) subtypes were assessed by flow cytometry on peripheral blood mononuclear cells (PMBCs). LDNs were defined as CD66b+CD15+ cells among CD11b+ PBMCs. Immature subtypes were defined as CD10- LDNs. T-cells were isolated from healthy donors and cocultured with patients‘ LDNs to characterize the neutrophils-T-cells interplay. LDNs subtypes were isolated from patients and treated in-vitro with cisplatin to assess cell death.Results46 NSCLC patients were treated with single-agent ICI and 17 with PCT+ICI (table 1). In the ICI single-agent cohort, PD and HPD occurred in 21 (41%) and 4 (9%) patients. Before ICI start, HPD patients had significantly higher median% of circulating immature CD10- LDNs neutrophils [43.5 (min 29.5; max 82.6) vs 10.3 (min 0.1; max 79.4), p=0.01] compared to PD patients (figure 1). In the ICI-PCT cohort no HPD was reported. 5 patients had baseline CD10- LDNs ≥ 43.5% (median% of CD10- LDNs in HPD patients upon single-agent ICI), 2 of them had stable disease and 3 PD upon ICI-PCT. In these 5 patients, CD10- LDNs significantly decreased during ICI-PCT compared to what observed in HPD patients upon single-agent ICI [median variation -43.4 (min -67.6, max -31.6) vs +6.9 (min -33, max +44), p= 0.03] (figure 2). After 7 days of coculture with T-cells, immature CD10- LDNs significantly reduced T-cells survival and promoted a T-cell senescent phenotype (CD28 loss, CD57 gain) impairing T-cells proliferation and increasing IFN-gamma production (figure 3). Cisplatin treatment significantly increased necrotic cell death among CD10- LDNs compared to CD10+ LDNs (figure 4).Abstract 15 Table 1Patients characteristics in the single-agent ICI and PCT+ICI cohortsAbstract 15 Figure 1Patterns of response, progression and hyperprogression to single-agent ICI and correlation with mature (CD10+) or immature (CD10-) LDNs’ subtypesAbstract 15 Figure 2Neutrophils dynamic variation upon treatment (5 patients with high baseline CD10- LDNs and PD/SD to chemo-ICI vs 4 patients with HPD upon single-agent ICI)Abstract 15 Figure 3Mature (CD10+) and immature (CD10-) LDNs in coculture with T-cells from an healthy donor differently influence CD8 and CD4 T-cells survival, proliferation, IFN-gamma production and expression of a senescent (CD28- CD57+) phenotype.Abstract 15 Figure 4Cisplatin in-vitro treatment at different concentration (1 uM and 5 uM) increases necrotic (7AAD expression) cell death preferentially of immature (CD10-) rather than mature (CD10+) LDNsConclusionsHigher baseline immature CD10- LDNs impair T-cell survival and promote T-cell senescence being a circulating biomarker of HPD upon single-agent ICI. The addition of PCT prevents HPD by inducing immature neutrophils cell death.ReferencesFerrara R, Mezquita L, Texier M, Lahmar J, Audigier-Valette C, Tessonnier L, et al. Hyperprogressive disease in patients with advanced non-small cell lung cancer treated with pd-1/pd-l1 inhibitors or with single-agent chemotherapy. JAMA Oncol 2018;4:1543–52. https://doi.org/10.1001/jamaoncol.2018.3676Kim Y, Kim CH, Lee HY, Lee S-H, Kim HS, Lee S, et al. Comprehensive clinical and genetic characterization of hyperprogression based on volumetry in advanced non-small cell lung cancer treated with immune checkpoint inhibitor. J Thorac Oncol 2019;14:1608–18. https://doi.org/10.1016/j.jtho.2019.05.033Ferrara R, Naigeon M, Auclin E, Duchemann B, Cassard L, Jouniaux JM, et al. Circulating T-cell immunosenescence in advanced non-small cell lung cancer patients treated with single agent PD-1/PD-L1 inhibitors or platinum-based chemotherapy. Clin Cancer Res 2020. https://doi.org/10.1158/1078-0432.CCR-20-1420.Ethics ApprovalPatients blood was obtained after signature of informed consent and within an observational prospective study (INT 22_15) approved by local Institutional Ethical Committee.

9033 Background: In advanced NSCLC (aNSCLC) with PD-L1 < 50% chemo-immunotherapy is the standard of care. Although the activity of single agent pembrolizumab was reported, no biomarkers have been identified able to select patients who mostly benefit. The trial aimed to identify immune biomarkers associated with PFS in aNSCLC patients with PD-L1 levels < 50% treated with first-line pembrolizumab. Here we reported for the first time the clinical outcomes of the trial and circulating immune profiling (CIP) biomarkers correlated to PFS. Methods: This phase II trial was conducted at Fondazione IRCCS Istituto Nazionale dei Tumori of Milan. Eligible patients were previously untreated stage IIIB-IV NSCLC, EGFR and ALK wild type with PD-L1 < 50%, PS 0-2. PD-L1 was assessed with 22-C3 antibody (Dako). Patients received pembrolizumab 200 mg every 3 weeks until 35 cycles, disease progression or unacceptable toxicity. The primary endpoint was the association of immune biomarkers with PFS. OS, ORR, DCR, DoR and safety were secondary endpoints. CIP was performed by determination of absolute cell counts for 36 innate and adaptive immunity subsets through multiparametric flow cytometry on freshly isolated whole blood samples at baseline. An orthoblique principal components-based clustering approach and multivariable Cox regression model adjusted for clinical variables were used to analyse CIP variables and their correlation with clinical endpoints. Results: From May 2018 to October 2020, of 87 screened, 65 patients were enrolled. The median age was 70.9 years, most patients were male (67.7%), smoker (87.7%), non-squamous (83.1%), PDL1 + (70.8%). 12 patients (18.5%) had PS 2. During a median follow-up of 26.4 months (mo), 51 radiological progressions, 46 deaths and 60 PFS events were observed. The median PFS was 2.9 mo (95%CI 1.8 - 5.6) and the median OS was 12.1 mo (95%CI 8.7 - 17.1). The ORR was 24.1% (2 complete and 12 partial responses), DCR was 53.4% and median DoR was 14.5 mo (95%CI 6.4 - 24.9). Drug related G3-G4 adverse events were: 2 hyponatremia, 2 lipase increased, 1 pneumonitis. Out of 7 CIP clusters identified, 2 were statistically significant at multivariable analysis on 57 patients. Higher baseline counts of 8 subsets within these two clusters were associated with better PFS (HR values range: 0.88-0.98; p values range: 0.001- 0.016). The significant subsets included lymphocytes (cells expressing CD3, or CD19 or CD56, but lacking granulocyte and monocyte markers) and main NK subsets (including CD56dim CD16+, CD56dim CD16- and HLA-DR+ NK cells). Conclusions: This trial confirmed that pembrolizumab as first-line single agent is safe and active also in a subgroup of aNSCLC patients with PD-L1 < 50%. CIP biomarkers can be useful to identify patients with a favourable outcome, thus avoiding the adding toxicity of chemotherapy. Clinical trial information: NTC03447678.

9051 Background: Chemo-immunotherapy is the standard of care for patients with advanced NSCLC and PD-L1 < 50%. Efficacy has been reported in this setting with single agent pembrolizumab, but no reliable biomarkers yet exist for selecting patients likely to respond to single agent immunotherapy. The aim of this trial was to identify potential new immune biomarkers associated with PFS in NSCLC patients with PD-L1 < 50% treated with first line pembrolizumab. Methods: Advanced EGFR and ALK wild type treatment-naïve NSCLC patients with PD-L1 < 50% were enrolled. Gene expression profile was performed using nCounter PanCancer IO 360 Panel (Nanostring) on baseline tissue. Circulating immune profiling was performed by determination of absolute cell counts with multiparametric flow cytometry on freshly isolated whole blood samples at baseline and at first radiological evaluation. Gut bacterial taxonomic abundance was obtained by shotgun metagenomic sequencing of stool sample at baseline. Pembrolizumab was administered at 200 mg flat dose every 3 weeks until 35 cycles, disease progression or unacceptable toxicity. Omics data was analyzed with sequential univariate Cox Proportional Hazards regression predicting PFS, with Benjamini-Hochberg multiple comparisons correction. Biological features significant with univariate analysis were analyzed with multivariate Least Absolute Shrinkage and Selection Operator (LASSO). Results: From May 2018 to October 2020 65 patients were enrolled. Main characteristics were: male 67.6%, smokers 87.7%, PD-L1 positive 70.8%. Median follow up and PFS were 26.4 and 2.9 months, respectively. Gene expression profile was performed in 48 patients, microbiome in 54 patients and circulating immune profiling in 56 patients at baseline and in 46 patients at first radiologic evaluation. LASSO multivariate analysis with optimal lambda of 0.28 showed high expression levels of IRF9 and COMP genes was associated with an unfavorable PFS (HR 3.03, 1.52 – 6.02, p = 0.08 and HR 1.22, 1.08 – 1.37, corrected p = 0.06, respectively). High expression levels of CD244 (HR 0.74, 0.62- 0.67, p = 0.05), PTPRC (HR 0.55, 0.38 – 0.81, p = 0.098), KLRB1 (HR 0.76, 0.66 – 0.89, p = 0.05) in tissue samples and NK cells/CD56dimCD16+ (HR 0.56, 0.41 – 0.76, p = 0.006) in peripheral blood at baseline and non classical CD14dim CD16+ monocytes (HR 0.52, 0.36 – 0.75, p = 0.004), eosinophils (CD 15+CD16-) (HR 0.62, 0.44 – 0.89, p = 0.003) lymphocytes (HR 0.28, 0.15 – 0.5, p < 0.001) in peripheral blood after first radiologic evaluation were associated to a favorable PFS. No microbiome features were selected by LASSO. Conclusions: To the best of our knowledge, this is the first prospective trial in NSCLC with PD-L1 < 50% performed with a multi-omic approach able to identify immune cell subsets and expression levels of genes associated to PFS under first line treatment with pembrolizumab. Clinical trial information: NTC03447678.

Background The treatment of metastatic non–small-cell lung cancer (mNSCLC) patients with EGFR/HER2 exon 20 insertion mutation (i-mut) remains an unmet clinical need. Poziotinib, a new generation tyrosine kinase inhibitor, is currently under investigation as a potential targeted therapy. This compassionate study of its use aims to describe the activity/toxicity of poziotinib in mNSCLC with EGFR/HER2-exon-20-i-mut. Patients and methods NSCLC patients who were treated either with EGFR or HER2 exon 20-i-mut within an expanded access program were included in this study. Poziotinib (16 mg or less) was administrated orally quaque die (QD). The primary end-point was the overall response rate (ORR) assessed by central review using RECIST v1.1, and secondary end-points were median progression free survival (PFS), disease control rate (DCR), median overall survival (OS) and toxicity. Results The median age of all the 30 patients was 58 years (25–80 years), most of them were females (73%); ECOG 0–1 (83%), EGFR i-mut (73%) and pre-treated (83%). 73% started with poziotinib at a dose of 16 mg. At data cut-off, 22 of 33 patients (73%) experienced a progress in the disease and 12 of 30 (40%) died. Median PFS was 5.6 months (95% CI: 3.6–6.7 months) and the mOS 9.5 months (95% CI: 5.3 – not-reached months). The ORR was 30% (EGFR/HER2: 23/50%) and DCR 80%. G3 AEs were reported in 66% of the patients and were found with skin rash (50%), diarrhoea (17.6%), mucositis (7%) and paronychia (3%). G5, possibly associated with pneumonitis might also have occurred. Conclusions Poziotinib exhibited effects in mNSCLC patients with EGFR/HER2-exon 20-i-mut. The toxicity rate was high leading to frequent dose interruption and reduction, thereby reducing mPFS in patients with good ORR/DCR. ZENITH20 trial is now being used to evaluate the low dose and new scheduled dose (e.g. bis in die (BID)).

Hyperprogressive disease (HPD) is a novel pattern of response during immunotherapy treatment. Several retrospective studies have evaluated its prevalence among various cancer types and, in particular, in nonsmall cell lung cancer patients, based on different definition criteria. If HPD is a just a typical phenomenon of immunotherapy is still an unsolved concern. This paper summarized the available data about HPD in other cancer treatments. Hyperprogressive disease (HPD) is a novel pattern of response during immunotherapy treatment. Several retrospective studies have evaluated its prevalence among various cancer types and, in particular, in non-small cell lung cancer patients, based on different definition criteria. If HPD is a just a typical phenomenon of immunotherapy is still an unsolved concern. This paper summarized the available data about HPD in other cancer treatments.

Background The PI3K/AKT/mTORC1 axis is implicated in hormone receptor-positive HER2-negative metastatic breast cancer (HR+ HER2− mBC) resistance to anti-estrogen treatments. Based on results of the BOLERO-2 trial, the mTORC1 inhibitor everolimus in combination with the steroidal aromatase inhibitor (AI) exemestane has become a standard treatment for patients with HR+ HER2− mBC resistant to prior non-steroidal AI therapy. In the recent SOLAR-1 trial, the inhibitor of the PI3K alpha subunit (p110α) alpelisib in combination with fulvestrant prolonged progression-free survival (PFS) when compared to fulvestrant alone in patients with PIK3CA-mutated HR+ HER2− mBC that progressed after/on previous AI treatment. Therefore, two different molecules targeting the PI3K/AKT/mTORC1 axis, namely everolimus and alpelisib, are available for patients progressing on/after previous AI treatment, but it is unclear how to optimize their use in the clinical practice. Main body of the abstract Here, we reviewed the available clinical evidence deriving from the BOLERO-2 and SOLAR-1 trials to compare efficacy and safety profiles of everolimus and alpelisib in advanced HR+ HER2− BC treatment. Adding either compound to standard endocrine therapy provided similar absolute and relative PFS advantage. In the SOLAR-1 trial, a 76% incidence of grade (G) 3 or 4 (G3/G4) adverse events was reported, while G3/G4 toxicities occurred in 42% of patients in the BOLERO-2 trial. While alpelisib was only effective in patients with PIK3CA-mutated neoplasms, retrospective analyses indicate that everolimus improves exemestane efficacy independently of PIK3CA mutational status. Conclusions Based on the available efficacy and safety data, the “new” alpelisib may be burdened by higher incidence of severe adverse events, higher costs, and anticancer efficacy that is limited to PIK3CA-mutated tumors when compared to the “old” everolimus. Therefore, the everolimus-exemestane combination remains an effective and reasonably well-tolerated therapeutic option for HR+ HER2− mBC patients progressing after/on previous AI treatment, independently of PIK3CA mutational status.

Background: Single-agent capecitabine (C) is a moderately effective chemotherapeutic compound in the treatment of patients with HER2-negative metastatic breast cancer (mBC). The capecitabine-vinorelbine (CV) combination is also used due to a good tolerability profile, but no studies have demonstrated its superiority over single-agent C. Methods: We conducted a retrospective analysis to compare overall response rate (ORR), progression-free survival (PFS), overall survival (OS) and incidence of adverse events (AEs) in patients with HER2-negative mBC treated with CV vs. single-agent C. Results: Out of 290 patients included in this study, 127 (43.8%) received single-agent C, while 163 (56.2%) patients were treated with CV. Median PFS was similar in patients treated with single-agent C or CV, while CV was associated with significantly longer OS in patients with hormone receptor-positive (HR+) BC. This OS advantage was confirmed at multivariable analysis also after propensity score-based matching of patients according to relevant clinical or tumor characteristics. When compared with single-agent C, CV was associated with higher incidence of G3/G4 and any-grade nausea/vomiting, diarrhea and increased transaminases. Conclusions: While prospective studies are needed to confirm our findings, the potential OS advantage of CV over single-agent C in HR+ mBC patients must be weighed against a significantly higher incidence of AEs.

Background On the basis of retrospective analyses and phase 2 studies, metastatic colorectal cancer patients whose disease responded to a first-line regimen containing an anti–epidermal growth factor receptor (EGFR) agent may experience benefit from anti-EGFR readministration in later therapy lines. While the analysis of circulating tumor DNA seems a promising tool to select the best candidates for this strategy, identifying clinical predictors of anti-EGFR sensitivity would be useful to drive treatment choices in daily practice. Patients and Methods A real-life database of 5530 patients treated at 6 institutions was queried. Included were patients who were retreated with anti-EGFRs, who had RAS/BRAF wild-type–status tissue samples, who had received a first-line anti-EGFR–based regimen with at least stable disease as best response, and who had received at least one further line of therapy before anti-EGFR retreatment. The association with overall response rate (ORR), progression-free survival (PFS), and overall survival (OS) of variables potentially related to anti-EGFR sensitivity was investigated. Results A total of 86 patients were identified. The ORR during anti-EGFR retreatment was 19.8%; median PFS and OS were 3.8 and 10.2 months, respectively. No significant association of clinical features of anti-EGFR sensitivity with ORR, PFS, and OS was observed. Among the 56 patients with a time from the last anti-EGFR administration to first-line progressive disease of 2 prior therapy lines and a longer anti-EGFR–free interval were associated with higher ORR, but not with longer PFS or OS. Conclusion Clinical features we deemed surrogates of anti-EGFR sensitivity were not reliable predictors of benefit from anti-EGFR retreatment.

Background: HPD has been described in 14-26% of NSCLC pts upon single agent ICI and correlates with poor survival. Currently, HPD occurrence has not been explored upon ICI-PCT and biomarkers of HPD are lacking. Although high circulating neutrophils (Ns) correlated with HPD in NSCLC pts, the role of specific Ns subsets is unknown. Methods: NSCLC pts treated with 1st line ICI as single agent or in combination with PCT were assessed for HPD and circulating Ns' phenotype. Tumor response was evaluated by RECIST 1.1. HPD required 3 assessment (2 before ICI, 1 upon ICI) and was defined as delta tumor growth rate (TGR) (TGR upon ICI - TGR before ICI) >50% and/or TGR ratio (TGR upon ICI/TGR before ICI) ≥2. Circulating low density Ns (LDNs) subtypes were assessed by flow cytometry (FC) on peripheral blood mononuclear cells (PBMCs). LDNs were defined as CD66b+CD15+ cells among CD11b+ PBMCs. Immature subtypes were defined as CD10- LDNs. Fatty acids (FA) (bodipy-palmitate) uptake was assessed by FC. T-cells of healthy donors were isolated and activated in vitro with CD3/CD28 beads in presence or not of pts' Ns to characterize their interplay. Results: Of 63 NSCLC pts, 22% had PD-L1 on tumor cells ≥50%. In the ICI single agent cohort (N=46), PD and HPD occurred in 21 (41%) and 4 (9%) pts, respectively. Before ICI start, HPD pts had significantly higher median percentage (%) of circulating immature CD10- LDNs [43.5 (min 29.5; max 82.6) vs 10.3 (min 0.1; max 79.4), p=0.01] compared to PD pts. In the ICI-PCT cohort (N=17), PD occurred in 4 (23%) pts, no HPD was reported. 5 pts had baseline CD10- LDNs ≥ 43.5% (median % of CD10- LDNs in HPD pts upon single agent ICI), 2 of them had SD and 3 PD upon ICI-PCT. In these 5 pts, CD10- LDNs significantly decreased during ICI-PCT compared to what observed in HPD pts during single agent ICI [median variation -43.4 (min -67.6, max -31.6) vs +6.9 (min -33, max +44), p= 0.03]. CD10- LDNs did not impair T-cells proliferation and IFNγ production, measured after 7 days of coculture but significantly reduced T-cell survival compared to CD10+LDNs. Baseline metabolic profile assessment showed that immature CD10- LDNs had a predominant lipid metabolism with a higher FA uptake [bodipy-palmitate mean florescence intensity (MFI) ± standard deviation (SD): 46754±11562] compared to lymphocytes [bodipy-palmitate MFI ±SD: 11542±405]. Conclusion: Higher baseline immature CD10- LDNs is a circulating biomarker of HPD upon single agent ICI. The addition of PCT prevents HPD reducing the immature Ns subset. The FA metabolism of immature LDNs does not affect T-cells proliferation/activation but reduces T-cells survival, suggesting that these two processes are differently regulated and that immature Ns reduce T-cells life span by metabolic competition. Citation Format: Roberto Ferrara, Elena Jachetti, Giuseppina Calareso, Marta Brambilla, Giuseppe Lo Russo, Claudia Proto, Arsela Prelaj, Diego Signorelli, Giulia Galli, Alessandro De Toma, Mario Occhipinti, Sara Manglaviti, Alice Labianca, Monica Ganzinelli, Sestina Maria Spanò, Giuliano Molino, Antonia Martinetti, Francesca Gabriella Greco, Marta Bini, Teresa Beninato, Filippo de Braud, Mario Paolo Colombo, Marina Chiara Garassino, Sabrina Sangaletti. Immunometabolism of circulating neutrophils in hyperprogressive disease (HPD) upon first-line PD-1/PD-L1 inhibitors (ICI) alone or in combination with platinum-based chemotherapy (PCT) in non-small cell lung cancer (NSCLC) patients (pts) [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2021; 2021 Apr 10-15 and May 17-21. Philadelphia (PA): AACR; Cancer Res 2021;81(13_Suppl):Abstract nr 1669.

(1) Background. The onset of a drug–drug interaction (DDI) may affect treatment efficacy and toxicity of advanced non-small-cell lung cancer (aNSCLC) patients during epidermal growth factor receptor (EGFR) tyrosine-kinase inhibitor (TKI) use. Here we present the use of Drug-PIN® (Personalized Interactions Network) software to detect DDIs in aNSCLC patients undergoing EGFR-TKIs. (2) Methods. We enrolled patients with Stage IV aNSCLC already treated with or candidates to receive EGFR-TKIs, in any line, ECOG PS 0–2, taking at least one concomitant drug. Cancer treatments, concomitant drugs, and clinical and laboratory data were collected and inserted in Drug-PIN®. (3) Results. Ninety-two patients, median age of 68.5 years (range 43–89), were included. In total, 20 clinically relevant DDIs needing medical intervention in a total of 14 patients were identified, the 14 major DDIs were related to a high-grade interaction between TKIs and SSRIs, antipsychotics, antiepileptics, H2-receptor antagonist and calcium antagonists. A negative association between statin intake and PFS was identified (p = 0.02, HR 0.281, 95% CI 0.096–0.825). (4) Conclusions. This is the first retrospective study assessing the prevalence of DDIs, the clinical need for medical intervention and the impact of concomitant drugs on EGFR-TKIs survival in aNSCLC.

Background Metastatic pancreatic cancer (mPAC) is a devastating disease with few therapeutic options. O6-methylguanine-DNA methyltransferase (MGMT) is a key DNA repair gene and there are reports of MGMT alterations in various gastrointestinal cancers, including colorectal cancer and mPAC. Low MGMT expression by immunohistochemistry (IHC) and MGMT promoter methylation ultimately result in diminished DNA-repair of O6-alkylguanine adducts and enhanced sensitivity to alkylating agents, such as temozolomide (TMZ). This report presents data on MGMT testing mPAC pts admitted at our center. Methods Formalin-fixed paraffin-embedded (FFPE) tissue samples were examined via methyl specific PCR (EZ DNA Methylation-Gold™ KIT), to assess MGMT promoter methylation, and IHC, to assess protein expression. Furthermore, Next Generation Sequencing (50 genes “Hotspot Cancer Panel, Ion Torrent®” and “Oncomine BRCA Research Assay”) and PCR analysis of microsatellite instability (MSI) were performed. Results Archived FFPE tissue sections obtained from 90 pts admitted at Fondazione IRCCS Istituto Nazionale dei Tumori of Milan from October 2017 to April 2019 were analyzed. At the time of this report, 60 samples (66%) had adequate tissue for extended analyses. As expected, 55 (92%) pts had KRAS mutations, while ATM, CDKN2A mutations and microsatellite instability (MSI) were found in 3 (5%), 4( 6%) and 2 (3%) pts, respectively. MGMT promoter methylation was found in 21 pts (35%), of which 13 (61%) had low/negative MGMT protein expression. Of note, amongst MGMT methylated pts, there were 5 (23%) BRCA1/2 somatic mutant and 2 (9%) MSI, suggesting possible genomic instability. Conclusions MGMT is a prognostic and predictive marker in glioblastomas and there is an increasing evidence in its role in metastatic CRC, with phase II studies showing a response rate of 10% in chemorefractory pts with MGMT methylation treated with TMZ. In our single center experience, MGMT methylation was found in 35% of patients with mPAC. This data warrant further prospective confirmation; nevertheless, considering the growing interest in the role of DNA-damage response genes in mPAC, there is definitely a rationale in investigating MGMT methylation as a predictive and prognostic biomarker in mPAC. Legal entity responsible for the study The authors. Funding Has not received any funding. Disclosure F. Morano: Honoraria (self): Servier. F. Pietrantonio: Advisory / Consultancy: Roche; Advisory / Consultancy: Amgen; Advisory / Consultancy: Lilly; Advisory / Consultancy: Sanofi; Advisory / Consultancy: Bayer; Advisory / Consultancy: Servier; Research grant / Funding (institution): BMS. F.G.M. De Braud: Advisory / Consultancy: TizianaLife Sciences; Advisory / Consultancy, Speaker Bureau / Expert testimony: BMS; Advisory / Consultancy: Celgene; Advisory / Consultancy: Novartis; Advisory / Consultancy: Servier; Advisory / Consultancy: Pharm Research Associated; Advisory / Consultancy: Daiichi Sankyo; Advisory / Consultancy, Speaker Bureau / Expert testimony: Ignyta; Advisory / Consultancy: Amgen; Advisory / Consultancy: Pfizer; Advisory / Consultancy: Octimet Oncology; Advisory / Consultancy: Incyte; Advisory / Consultancy: Teofarma; Advisory / Consultancy: Pierre Fabre; Advisory / Consultancy, Speaker Bureau / Expert testimony: Roche; Advisory / Consultancy: EMD Serono; Speaker Bureau / Expert testimony: Eli Lilly; Speaker Bureau / Expert testimony: Novartis; Speaker Bureau / Expert testimony: Amgen; Speaker Bureau / Expert testimony: AstraZeneca. M. Di Bartolomeo: Advisory / Consultancy: Lilly; Advisory / Consultancy: Servier; Travel / Accommodation / Expenses: Roche. All other authors have declared no conflicts of interest.

Introduction In advanced Non-Small Cell Lung Cancer (aNSCLC), Programmed Death Ligand 1 (PD-L1) remains the only used biomarker to candidate patients (pts) to immunotherapy (IO) even if its predictive accuracy is not satisfactory. Indeed, given the complex dynamics underlying the cross-talk between the tumor and its microenvironment, it is unlikely that a single biomarker could be able to profile prediction with high precision. Artificial Intelligence (AI) and machine learning (ML) are techniques able to analyze and interpret big data, which cope with this complexity. The present study aims at using AI tools to improve response and efficacy prediction in aNSCLC pts treated with IO. Methods A classification task to determine if a pt is likely to benefit from IO was formulated using complete clinical data, PD-L1, histology, molecular data, and the blood microRNA signature classifier (MSC), which include 24 different microRNAs. Pts were divided into responders (R), who obtained a partial response or stable disease as best response, and non-R, who experienced progressive disease. A forward feature selection technique based on the Akaike Information Criterion was used to extract a specific subset of the pts data, being the most informative ones for the task. To develop the final predictive model, different ML methods have been tested: K-nearest neighbors, Logistic Regression, Kernel Support Vector Machines, Feedforward Neural Network, and Random Forest. Results Of 164 enrolled pts, 73 (44.5%) were R and 91 (55.5%) non-R. At data cut-off (Nov 2020), median Overall Survival (mOS) was 10.1 (95%IC 7.0 - 13.2) months (m). mOS for R pts was 38.5 m (95%IC 23.9 - 53.1) vs 3.8 m (95%IC 2.8 - 4.7) of non-R, p Citation Format: Arsela Prelaj, Mattia Boeri, Alessandro Robuschi, Claudia Proto, Giuseppe Lo Russo, Roberto Ferrara, Giulia Galli, Alessandro De Toma, Marta Brambilla, Mario Occhipinti, Sara Manglaviti, Alice Labianca, Teresa Beninato, Marta Bini, Mavis Mensah, Monica Ganzinelli, Nicoletta Zilembo, Filippo de Braud, Gabriella Sozzi, Marcello Restelli, Alessandra Pedrocchi, Marina Chiara Garassino, Francesco Trovo. Artificial intelligence to improve selection for NSCLC patients treated with immunotherapy [abstract]. In: Proceedings of the AACR Virtual Special Conference on Artificial Intelligence, Diagnosis, and Imaging; 2021 Jan 13-14. Philadelphia (PA): AACR; Clin Cancer Res 2021;27(5_Suppl):Abstract nr PO-065.

10053 Background: In pts with metastatic melanoma bearing BRAF V600E/K mutations BRAF V600±MEK inhibitors are administered until PD/unacceptable toxicity. In patients achieving durable responses, outcomes following discontinuation for reasons other than PD are largely unknown. Methods: We identified all patients who interrupted BRAF±MEK inhibitors for reasons other than PD after complete (CR) or partial response (PR) from a clinical dataset of patients with BRAF mutated metastatic/unresectable melanoma treated with targeted therapy at a single Institution. Results: We included 24 pts. Fifteen (62.5%) and 9 (37.5%) pts were treated respectively with BRAF inhibitor monotherapy and BRAF+MEK inhibitor combination. All pts had normal baseline LDH and ECOG PS0, 2 (8%) pts had brain metastases and 15 (62.5%) had multi-organ metastatic involvement. Dose reduction was required for 12 (50%) pts. Median treatment duration was 59 (12-88) months. Causes of discontinuation were unacceptable toxicity (19 pts-79%) and consent withdrawal (5 pts-21%). At the time of discontinuation, 17 (71%) and 7 (29%) pts had achieved respectively CR and PR. At a median follow up of 31 (8-59) months after treatment discontinuation, 9 (37.5%) pts had experienced PD. Median time to PD after treatment discontinuation was 9 (3-16) months. At time of PD, 2 (22%) pts displayed involvement of new organ sites. Risk of PD following discontinuation was respectively 31% and 45% at 12 and 24 months. Neither baseline characteristics nor treatment duration and time to best response influenced risk of PD; we found a non-significant trend towards higher risk of relapse for patients interrupting treatment with residual disease compared to those who interrupted treatment after achieving CR [HR 3.3; 95%CI (0.8–14.1); log-rank p = 0.081]. After PD, 6 pts received BRAF+MEK inhibitors with a response rate of 100% and 3/6 pts achieving CR. Conclusions: In a subset of patients with favorable prognostic characteristics and retained sensitivity to BRAF±MEK inhibitors, treatment discontinuation was associated with relevant risk of relapse with about one third of pts experiencing PD within one year. Biomarker studies are needed to identify pts who might safely discontinue therapy due to sustained toxicity, especially after achieving CR.

e23518 Background: In patients with GIST, literature reports a risk of second primary tumors between 4.5% and 33% with different distribution in the worldwide. The network of 7 italian EURACAN centers has collected clinical and molecular features of GIST patients with second primary tumors. Methods: We reviewed the clinical characteristics of 201 patients with GIST and second primary tumors in order to evaluate association between risk of dead and each possible factor, using Kaplan meier curve, log-rank test and Cox model for Hazard Ratio and it’s interval Confidence 95% estimation. Furthermore, NGS analysis ( 56 gene onco panel) in 72 patients with GIST was performed. Results: On the basis of the significant correlation previously observed between the Miettinen risk criteria of GIST (low/very low) and the incidence of second primary tumors (gastrointestinal tumors),P < 0.001 (Abstr 11032 ASCO 2019), we observed in these patients a median age of diagnosis of GIST of 68, with prevalent gastric site localization ( KIT exon 11 mutation), NF1 and Lynch (kit/pdgfra WT) syndromes in the low risk subgroup. The more frequent site of second epithelial tumor in gastrointestinal tract was the colon followed by gastric, pancreatic and biliary tract. In patients with GIST with low-very low risk according Miettinen classification, after a median follow-up period of 25 years, we have observed that the gastroenteric site of second tumors occurrence is significantly related to the survival (p < .0003). In the NGS analysis of the GIST we observed the pathogenetic somatic mutations in the following genes: BRCA 2 (p.Thr2125fs), but germline test negative, TP53 (p.Arg192*,p.Gly244Ser,p.His168Leu), RET (p.Lys120Asn, p.His168Leu, p.Thr930Met ), NRAS (p.Gly134Asp), CTNNB1 (p.Ser45Phe), MSH6 (P.Ala164Val), SMARCB1 (p.Arg192), VHL (p.Gly93Val), PTEN(p.Val158Ile, p.Asn323fs), STK1I (p.Arg40Cys), SMO (p.Glu194Lys), EGFR (p.Gly721Asp), ATM (p.Asp2708Asn), ERBB4 (p.Asn181Ile). Conclusions: In our analysis patients with GIST (low-very low classes according to Miettinen) have significant risk to death because of second primary tumors in gastrointestinal tract. Specific attention to gastrointestinal screening during the follow-up of GIST (low and very low risk) is required.

Background Biliary Tract Cancers (BTC) are a devastating and molecularly heterogeneous group of diseases, with significant differences between intrahepatic (ICC) and extrahepatic (ECC) cholangiocarcinoma and Gall bladder (GB) cancer. O6-methylguanine-DNA methyltransferase (MGMT) is a key DNA repair gene, responsible of alkyl groups’ elimination from the O6-position of guanine. Reductions in MGMT expression and MGMT promoter methylation result in diminished DNA-repair of O6-alkylguanine adducts and enhanced sensitivity to alkylating agents, such as temozolomide (TMZ). This alterations have been described in BTC in small reports with variable frequencies. Here we present data on MGMT methylation tested in BTC pts admitted at our center. Methods MGMT promoter methylation was studied via methyl specific PCR (EZ DNA Methylation-Gold™ KIT), while protein expression was assessed with immunohistochemistry (IHC). Formalin-fixed paraffin-embedded (FFPE) tissue samples were also examined using Next Generation Sequencing (50 genes “Hotspot Cancer Panel, Ion Torrent®”). Results Archived FFPE tissue sections from 100 pts admitted at Fondazione IRCCS Istituto Nazionale dei Tumori of Milan from October 2017 to February 2019 were analyzed. Amongst the 89 samples with adequate tissue, 73% were ICC, 16% were ECC and 8 % were GC. 34 pts (38%) had MGMT promoter methylation, while low/negative MGMT protein expression was found in 45 pts (50%). MGMT methylation was identified in 38% of ICC, 26% of ECC and 62% of GC. As expected, we identified IDH1/2 mutations in 15%, all affected by ICC. Of note, amongst MGMT methylated pts, there were 4 pts (11%) with concomitant IDH1/2 mutations, which are reported to be associated with CpG Island Methylator Phenotype. Conclusions MGMT is a prognostic and predictive marker in glioblastomas and there is an increasing evidence in its role in gastrointestinal cancer, with phase II studies showing a response rate of 10% in chemorefractory MGMT- methylated colorectal cancer treated with TMZ. In our single center experience, MGMT methylation was found in 38% of patients with BTC. This data warrant further confirmation, but there is a growing interest in novel targeted therapies exploiting the role of MGMT methylation as a predictive and prognostic biomarker in BTC. Legal entity responsible for the study The authors. Funding Has not received any funding. Disclosure F. Morano: Honoraria (institution): Servier. F. Pietrantonio: Honoraria (institution), Advisory / Consultancy: Roche; Honoraria (institution), Advisory / Consultancy: Amgen; Honoraria (institution), Advisory / Consultancy: EMD Serono; Honoraria (institution), Advisory / Consultancy: Sanofi; Honoraria (institution), Advisory / Consultancy: Bayer; Honoraria (institution), Advisory / Consultancy: Servier; Research grant / Funding (self): BMS. M. Di Bartolomeo: Honoraria (institution), Advisory / Consultancy: Eli Lilly; Honoraria (institution), Advisory / Consultancy: Servier; Travel / Accommodation / Expenses: Roche. F.G.M. De Braud: Speaker Bureau / Expert testimony: Bayer; Advisory / Consultancy, Speaker Bureau / Expert testimony: BMS; Advisory / Consultancy: Celgene; Advisory / Consultancy, Speaker Bureau / Expert testimony: Novartis; Advisory / Consultancy: Servier; Advisory / Consultancy: Pharm Research Associated; Advisory / Consultancy: Daiichi Sankyo; Advisory / Consultancy, Speaker Bureau / Expert testimony: Ignyta; Advisory / Consultancy, Speaker Bureau / Expert testimony: Amgen; Advisory / Consultancy, Speaker Bureau / Expert testimony: Pfizer; Speaker Bureau / Expert testimony: ACCMED; Advisory / Consultancy: Incyte; Advisory / Consultancy: Teofarma; Advisory / Consultancy: Pierre Fabre; Advisory / Consultancy, Speaker Bureau / Expert testimony: Roche; Advisory / Consultancy: EMD Serono; Speaker Bureau / Expert testimony: Eli Lilly; Speaker Bureau / Expert testimony: AstraZeneca; Speaker Bureau / Expert testimony: Fondazione Menarini; Speaker Bureau / Expert testimony: MSD. All other authors have declared no conflicts of interest.

3536 Background: FOLFOXIRI/bevacizumab is a valuable upfront option in mCRC based on results of phase III TRIBE and TRIBE2 studies: 1187 pts aged 18–70 years with ECOG performance status (PS) ≤ 2 or between 71–75 years with an ECOG PS of 0 were randomized to receive first-line FOLFOXIRI/bevacizumab or a doublet (FOLFIRI in TRIBE and mFOLFOX6 in TRIBE2)/bevacizumab. Here, we aimed at assessing the effect of the intensification of the upfront chemotherapy (triplet versus doublet) in terms of safety and efficacy in pts aged < 70 versus 70-75. Methods: Subgroup analyses for ORR, PFS, G3/4 overall adverse events (AEs), chemo-related and bevacizumab-related AEs were performed according to baseline age. Results: 182 (15%) out of 1187 pts were 70-75 years old (97 in the FOLFOXIRI/bevacizumab and 85 in the doublets/bevacizumab arms). The benefit provided by the intensification of the upfront chemotherapy was independent of the age subgroup in terms of both ORR (p for interaction = 0.684) and PFS (p for interaction = 0.634). The risk of overall and chemo-related G3/4 AEs was increased with the triplet independently of age (p for interaction = 0.736 and 0.790), while no difference in bevacizumab-related AEs was observed in both subgroups (p for interaction = 0.566). In the overall population, as compared to younger pts, those aged 70-75 were more susceptible to overall G3/4 AEs (70% vs 57%, p = 0.001). In the FOLFOXIRI/bevacizumab arm a higher incidence of G3/4 diarrhea (27% vs 17%, p = 0.016) and febrile neutropenia (16% vs 6% p = 0.001) and a lower incidence of all grade nausea (51% vs 65%, p = 0.009) and vomiting (26% vs 44% p = 0.001) were reported among elderly pts. Conclusions: The activity and efficacy of FOLFOXIRI/bevacizumab are confirmed among selected pts between 70 and 75 years old, with a relative increase in the risk of chemo-related AEs similar to that of younger pts. However, elderly pts are more susceptible to experience AEs independently of the treatment arm. Considering the increased incidence of febrile neutropenia and diarrhea with FOLFOXIRI/bevacizumab, the use of G-CSF as primary prophylaxis or an initial dose reduction of irinotecan and 5-fluorouracil might be considered in this population.

e15770 Background: Metastatic pancreatic cancer (mPAC) has a poor prognosis, with few therapeutic options and an overall survival (OS) at 5 years < 5%. O6-methylguanine-DNA methyltransferase ( MGMT) is a key DNA repair gene, responsible of alkyl groups’ elimination from the O6-position of guanine. Its promoter methylation results in diminished DNA-repair of O6-alkylguanine adducts and enhanced sensitivity to alkylating agents, such as temozolomide (TMZ). Of note, both reductions in MGMT expression and MGMT promoter methylation are described in a variety of gastrointestinal malignancies, including colorectal cancer (CRC). Here we present data on MGMT methylation tested in mPAC pts treated at our center. Methods: Formalin-fixed paraffin-embedded (FFPE) tissue samples were examined using Next Generation Sequencing (50 genes “Hotspot Cancer Panel, Ion Torrent®” and “Oncomine BRCA Research Assay”) and PCR analysis of microsatellite instability (MSI). Furthermore, the exploratory analysis of MGMT status was performed via methyl specific PCR (EZ DNA Methylation-Gold™ KIT) to assess promoter methylation, and immunohistochemistry (IHC) was done to assess protein expression. Results: Archived FFPE tissue sections obtained from 60 pts treated at Fondazione IRCCS Istituto Nazionale dei Tumori of Milan from October 2017 to December 2018 were analyzed. 47 samples (78%) had adequate tissue for extended analyses. As expected, 44 (93%) pts had KRAS mutations, while ATM, CDKN2A mutations and microsatellite instability (MSI) were found in 3 pts (6%), respectively. MGMT promoter methylation was identified in 14 pts (29%), with low/negative MGMT protein expression in 7 (14%). Interestingly, amongst MGMT methylated pts, there were 3 (21%) BRCA1/2 somatic mutant and 1 (7%) MSI, suggesting possible genomic instability. Conclusions: MGMT is a prognostic and predictive marker in glioblastomas and there is an increasing evidence in its role in metastatic CRC, with phase II studies showing a response rate of 10% in chemorefractory pts with MGMT methylation treated with TMZ. In our single center experience, MGMT methylation was found in 29% of patients with mPAC. This data warrant further prospective confirmation, but there is definitely a growing interest in the role of MGMT methylation as a predictive and prognostic biomarker in mPAC.

11032 Background: GISTs are the most common mesenchymal tumors of the digestive tract. As of recent, new links are being made between GISTS and secondary malignancies. However, whether the coexistence of GISTs with other tumors is stochastic, or the result of related pathogenetic mechanisms is still unknown. Methods: We retrospectively reviewed clinical and molecular features from all GIST patients with second tumors treated in seven Italian GIST reference centers. Qualitative variables were compared using the Fisher exact test. Results: Clinical data of 184 patients with diagnosis of GIST were evaluated. Median age at diagnosis was 66 years, KIT exon 11 resulted the most frequent mutation (73%) while seven patients (3.8%) had a genetic syndrome. The most common primary GIST localizations were stomach (54%) and small intestine (33%). Second tumors arose mostly from gastrointestinal and genitourinary tract. Fourtythree patients had two primary tumors other than GIST and five patients had three other primary malignancies. According to Miettinen criteria, 45% of non-metastatic patients at diagnosis belong to low or very low-risk classes. We highlighted a significant correlation (P=0.002) between risk class and second/third tumor localization, with considerably high percentage of GI second malignancies in low/very low risk GISTs (table). Conclusions: The high frequency of second/third tumors reported in low and very low GIST calls for a careful follow-up also in these patients. Furthermore, this population requires further genetic investigation, NGS analysis is ongoing. [Table: see text]