Abstract Background Non-small cell lung cancer (NSCLC) is one of the commonest disease worldwide and the leading cause of cancer-related death. Incidence increases with age and reaches a peak in senility, when patients’ (pts) comorbidities may limit the efficacy of treatments. At this time, no homogeneous indications are available for elderly NSCLC pts and the optimization of treatment, with the lowest side effects, is still an unmet need, also after the introduction of innovative drugs. The ribonucleotide reductase catalytic subunit M1(RRM1), the DNA-excision repair protein ERCC1 and the thymidylate synthetase (TS) are cancer molecular markers able to predict response to gemcitabine, platinum compounds and pemetrexd, respectively. Originally, aim of the EPIC study (active from July 2012) was to optimize survival of elderly NSCLC pts with the use of a genomic driven chemotherapy, comparing it to standard treatment. After the proven efficacy of first-line immunotherapy, on June 2018 the protocol was amended with the introduction of a new study arm, aiming to clarify the best strategy in this setting. Trial design EPIC is an Italian multicenter, randomized phase III trial, comparing Avelumab (A) vs genomic driven chemotherapy (B) and genomic driven chemotherapy vs standard chemotherapy according to investigator’s choice (C) in elderly untreated NSCLC pts, randomized in a 2:2:1 fashion manner. Age over 70 years, ECOG performance-status 0-1, stage IV non-oncogene addicted NSCLC and tissue availability for gene expression analysis, are key inclusion criteria. Before study entry, pharmacogenomic evaluations of RRM1, ERCC1 and TS is performed in the entire study population, but results are disclosed only to pts randomized in arm B to not influence subsequent treatments in the other arms. Primary endpoint of the EPIC study is overall survival (OS), while secondary endpoints include progression-free survival (PFS) and treatment response rate (by investigator’s assessment). Further objectives are the evaluation of feasibility of treatment selection based on pharmacogenomic parameters, and treatment-related toxicities in every study arm. End of enrollment is expected by the end of 2022. Legal entity responsible for the study Department of Oncology - University of Turin, Italy. Funding Italian Pharmacology Agency (AIFA). Disclosure E. Capelletto: Advisory / Consultancy: AstraZeneca; Advisory / Consultancy: Boehringer Ingelheim. F. Grossi: Advisory / Consultancy: Eli Lilly; Advisory / Consultancy: Roche; Advisory / Consultancy: Boehringer Ingelheim; Advisory / Consultancy: AstraZeneca; Advisory / Consultancy: Pierre Fabre; Advisory / Consultancy: BMS; Advisory / Consultancy: MSD; Advisory / Consultancy: Novartis. P. Bidoli: Advisory / Consultancy: Eli Lilly; Advisory / Consultancy: Boheringher; Advisory / Consultancy: BMS. O. Caffo: Honoraria (self): Pfizer ; Honoraria (institution): AstraZeneca. S. Novello: Advisory / Consultancy: Boehringer Ingelheim; Advisory / Consultancy: AstraZeneca; Advisory / Consultancy, Speaker Bureau / Expert testimony: Eli Lilly; Advisory / Consultancy: BMS; Advisory / Consultancy: MSD; Advisory / Consultancy: Roche ; Advisory / Consultancy: Takeda; Advisory / Consultancy: Pfizer; Advisory / Consultancy: AbbVie; Advisory / Consultancy: Celgene. G. Scagliotti: Advisory / Consultancy, Speaker Bureau / Expert testimony: Eli Lilly; Advisory / Consultancy: AstraZeneca; Advisory / Consultancy: Roche; Advisory / Consultancy: MSD; Advisory / Consultancy: Pfizer. All other authors have declared no conflicts of interest.

Background: Nivolumab has been approved by different regulatory agencies worldwide for the treatment of nSq-NSCLC based on its superiority in Overall Survival (OS) versus docetaxel in CheckMate 057 trial. In a pre-specified subgroup analysis of the same trial, this advantage was confirmed also in KRAS-mutation+ pts but the small number precluded any definitive conclusion. The Italian nivolumab EAP for non-sq-NSCLC might represent an important source of information in that respect. The current analysis describes results of the use of nivolumab in the group of EAP pts with KRAS mutations. Methods: Nivolumab was provided upon physicians' request for pts aged ≥18 years who had relapsed after a minimum of one prior systemic treatment for stage IIIB/stage IV non-Sq-NSCLC. Nivolumab 3 mg/kg was administered intravenously every 2 weeks for ≤24 months. Pts included in the analysis received ≥1 dose of nivolumab and were monitored for adverse events (AEs) using Common Terminology Criteria for Adverse Events. Results: In total, 1588 Italian pts with advanced nSq-NSCLC received at least one dose of nivolumab in the EAP across 168 sites. Among pts evaluated for KRAS mutation, 206 (39%) resulted positive. In this subgroup of pts, with a median follow-up of 7.7 months (0.1-21.2) and a median number of 8 doses (1-45), the best overall response rate (BORR) was 20%, including 2 pts with complete response and 39 pts with partial response, and the median OS was 10.7 months (8.6-12.8). These results were in line with those ones showed in the overall population (18% BORR and 11 months median OS, respectively). Conclusions: This analysis confirms, in a real word setting and in a much larger number of pts, the efficacy of nivolumab in KRAS-positive pts in CheckMate 057. Nivolumab represents a potentially effective therapeutic option for KRAS mutation, a molecular alteration for which there is currently no direct targeted therapy. Clinical trial identification: CA209-966 Legal entity responsible for the study: Lucio Crinò Funding: None Disclosure: F. Cappuzzo: Consultant and participation in advisory boards for BMS, Roche, Pfizer, AZ. All other authors have declared no conflicts of interest.

Background: The risk of developing renal cell carcinoma (RCC) increases with age, and given the constant gain in life expectancy of the general population, RCC is frequently observed in the elderly. More than 80% of cancer pts aged ≥ 70 years have at least one comorbidity requiring treatment, leaving them exposed to drug interactions. Due to high frequency of comorbidities, these pts are often under-represented in clinical trials. The purpose of this analysis is to evaluate the feasibility of treatment with nivolumab in the elderly (≥ 70 years) and very elderly (≥ 75 years) in the EAP in Italy, given a more realistic picture of real world setting. Methods: Nivolumab was available upon physician request for pts aged ≥18 years who had relapsed after at least one prior systemic treatment in the advanced or metastatic setting. Nivolumab 3 mg/kg was administered intravenously every 2 weeks to a maximum of 24 months. Pts included in the analysis had received ≥ 1 dose of nivolumab and were monitored for adverse events (AEs) using Common Terminology Criteria for Adverse Events. Results: Of 389 Italian pts with mRCC enrolled in the EAP in Italy 125 pts (32%) had ≥70 years and 70 (18%) had ≥75 years. With a median follow-up of 9.8 months (0.1-16.2) in the elderly population (≥70 years), the disease control rate (DCR) was 58% including 1 patient in complete response (CR), 32 pts in partial response (PR) and 40 patients in stable disease (SD). Regarding the very elderly population (≥75 years), with a median follow-up of 9.8 months (0.1 -14.9), the DCR was 60% including 1 patient with CR, 19 pts with PR and 22 with SD. As of May 2017, 6 and 12 months overall survival (OS) rate were 87.2% and 77.8% respectively in the elderly population. Regarding the very elderly, the 6 and 12 months OS rate was 83.6% and 77.7%, respectively. The safety profile was consistent to what already observed in the general population. Conclusions: These results suggest that elderly population can benefit from nivolumab treatment with safety results consistent to what previously reported, supporting the use of nivolumab in this subpopulation. Clinical trial identification: CA209-99M Legal entity responsible for the study: Sergio Bracarda coordinator Italian RCC EAP Group Funding: None Disclosure: All authors have declared no conflicts of interest.

ABSTRACT Background: Phase III trial aiming to compare first line pharmacogenomic-driven chemotherapy based on Excision-Repair-Cross-Complementing-1 (ERCC1, E), Ribonucleotide Reductase subunit M1 (RRM1, R) and Thymidylate Synthase (TS, T) gene expression versus standard treatment in elderly patients (pts) with advanced NSCLC. Trial design: pts aged ≥70 years, ECOG Performance Status 0 or 1, EGFR negative mutational status, chemonaive stage IV NSCLC will be evaluated. Pts will be randomized (2:1) to experimental arm (A) or standard arm (B). In arm A, treatment will be based on histology, E, R and T mRNA expression. The cut-off for high or low expression have been previously defined. Treatments for pts with squamous NSCLC: carboplatin for E low/R high, gemcitabine for E high/R low, carboplatin and gemcitabine for E low/R low, docetaxel or vinorelbine for E high/R high. Treatments for non-squamous NSCLC pts: carboplatin for E low/T high, pemetrexed for E high/T low, carboplatin and pemetrexed for E low/T low, gemcitabine for E high/T high/R low, docetaxel or vinorelbine for E high/T high/R high. In arm B treatment will be at the investigator discretion. The primary endpoint is overall survival. The secondary endpoints are progression-free survival, response rate (RECIST 1.1) and tolerability (CTCAE version 4.0). Feasibility of treatment selection based on pharmacogenomic parameters will be also assessed. Treatment will continue to a maximum of 6 cycles if tolerated or until disease progression. Assuming an exponential survival distribution for both arms and a median survival time of 8 months in the control arm, we anticipate to detect an improvement of 3 months in the median survival time in the experimental arm. To have 90% power to detect a three-month improvement in median survival at a significance level of 5% (2-sided) and assuming a 10% failure rate in gene analyses or loss to follow-up rate, a sample size of 567 patients is planned. Conclusion: To our knowledge, this is the first prospective pharmacogenomic-driven trial designed in elderly advanced NSCLC population. Disclosure: D. Cortinovis: Consultant for AZ, BI, Roche and Lilly; M. Papotti: Honoraria have been received from Eli Lilly; G. Scagliotti: Honoraria have been received from Eli Lilly, Roche, Pfizer and Astrazeneca. All other authors have declared no conflicts of interest.

e19037 Background: No difference in efficacy has been observed between 3 or 4 vs 6 cycles of 1st line CT in patients (pts) with advanced NSCLC. Although longer treatment may improve progression-free survival, this must be weighed against increased toxicity and potential for worse quality of life. There are no prospective, randomized data on the optimal duration of 2nd line CT. The aims of this retrospective study were to (i) confirm the lack of impact of treatment duration on response rate in 1st line NSCLC and (ii) investigate the impact of treatment duration in the 2nd line setting. Methods: Eligible pts had received at least 4 cycles of 1st line (n=95) or 2nd line (n=55) CT for stage IIIB/IV NSCLC in clinical trials of the ATOM group, Udine, Italy, or the National Cancer Institute, Genoa, Italy and achieved stable disease or responded to treatment; trials included one in the elderly, and the addition of biological agents to CT. Response was assessed according to RECIST after alternate CT cycles and tumor dimensions recorded. The proportion of pts achieving a complete or partial response, as well as its timing and any subsequent tumor shrinkage, were analyzed by treatment line. Results: In the 1st line setting, 48 of 95 pts responded; they did so after 2 (29 pts, 61%), 4 (15 pts, 31%) and 6 cycles (4 pts, 8%). In these pts, 16 (55%) and 9 (41%) pts who had responded by the 2nd and 4th cycle respectively and continued on treatment showed further tumor shrinkage; the median further reduction in tumor size was 16% (range 1–52%) and 6% (range 1–11%), respectively. In the 2nd line setting, 12 of 55 pts responded; they did so after 2 (3 pts, 25%), 4 (7 pts, 58%), and 6 cycles (2 pts, 17%). One and two pts who had responded by the 2nd and 4th cycle respectively and continued on treatment showed further tumor shrinkage of 25%, 5% and 24%, respectively. Conclusions: Approximately 90% of CT-naive pts who responded did so by the 4th cycle; most such pts achieved only minor tumor reduction with further CT and did so at the likely expense of increased toxicity. In the 2nd line setting the achievement of response appeared slower. These results support the discontinuation of 1st CT after 4 cycles and suggest the same is true for 2nd line CT. No significant financial relationships to disclose.

e19010 Background: No combination regimen has proven superior to single agent chemotherapy as 2nd-line treatment for NSCLC. The absence of cross-resistance with cisplatin/carboplatin, favorable toxicity profile, along with both pre-clinical and clinical evidence of activity make O a good candidate for combination with D as 2nd-line therapy of NSCLC. We evaluated the activity of DO in this setting using a novel phase II trial design. Methods: This multicenter, non-comparative randomized phase II trial evaluated the activity of D (75 mg/m2 d1) and O (70 mg/m2 d2) every 3 weeks in previously treated NSCLC pts; the comparator arm was D (75 mg/m2 d1 every 3 weeks). This one-stage, three-outcome phase II trial design (Sargent, Control Clin Trials 2001) had 21 evaluable pts/arm. All had histologically confirmed NSCLC that progressed during/after platinum-based chemotherapy. Primary endpoint was response rate; secondary endpoints were toxicity, time to progression (TTP), 1-yr survival. Results: Fifty pts were enrolled. Pts characteristics: M/F, 76/24%; median age 62 yrs (range 43–69); ECOG PS 0/1, 36/64%; adenocarcinoma/other, 36/64%. With 48 pts evaluable, partial response was seen in 20% and 8% of pts; stable disease in 52% and 32% and progressive disease in 24% and 56% for DO and D, respectively; 1 pt was inevaluable due to early death (D arm). Main grade 3–4 toxicities were: neutropenia 56% and 64%; febrile neutropenia 4% and 8%; diarrhea 12% and 4% for DO and D, respectively. Median TTP was 4.9 and 1.8 months, median survival 10.9 and 6.9 months, and 1-yr survival 41% and 16% for DO and D, respectively. Conclusions: This study shows how novel phase II trial designs enrolling a limited number of pts may help identify promising regimens for subsequent study in phase III trials. The level of activity for DO we observed satisfied the pre-defined study primary endpoint and warrants further evaluation of this combination as 2nd-line therapy for NSCLC. Protocol developed at the 6th FECS/AACR/ASCO Workshop on Methods in Clinical Cancer Research, Flims 2004, with Professors Marc Buyse and Chris Twelves. [Table: see text]

Though chemotherapy remains a mainstay of non-small cell lung cancer (NSCLC) treatment, its efficacy has probably reached a plateau. The management of advanced NSCLC has evolved considerably in recent years due to a progressive understanding of tumour biology and the identification of promising molecular targets. Bevacizumab is a monoclonal antibody targeting vascular endothelial growth factor (VEGF), a key signalling protein in tumour neoangiogenesis, growth and dissemination. The finding in study E4599 of a survival benefit for carboplatin-paclitaxel plus bevacizumab over chemotherapy alone led the U.S. FDA to approve the novel combination for the first-line treatment of patients with advanced, non-squamous NSCLC. This study is the first to show a survival advantage with the addition of a targeted agent to chemotherapy in this setting: in particular, for the first time the survival of NSCLC patients has been extended beyond one year. Recently, in a randomised phase III trial, patients receiving cisplatin-gemcitabine plus bevacizumab experienced a significantly longer progression-free survival compared to the standard arm. Based on these data, the EMEA has just issued a positive opinion to extend the drug's indication to include first-line treatment – in combination with any platinum-based chemotherapy – of advanced, non-squamous NSCLC. The aim of this review is to provide an overview of the evidence supporting the emergence of this new treatment. Key questions – including the optimal dose of bevacizumab, safety of the drug in special populations, the selection of patients most likely to benefit from the treatment, the role of maintenance – are addressed.

18187 Background: Cutaneous toxicities due to E or G are common and reversible adverse effects. In the two large randomized trials with E (Shepherd F et al, NEJM 2005) and G (Thatcher N et al, Lancet 2005) skin rash toxicity incidence seems different between the two drugs: all grades (gr) were observed in 76% and 37% of patients (pts), gr 3–4 in 9% and 2% respectively. The aim of this study was to assess the incidence of the most common toxicities occurred in a cohort of pts treated with E or G in the same period of time and evaluated by the same medical staff according to the Common Toxicity Criteria version 3.0. Methods: From May 2005 to July 2006, 47 pts suitable for a treatment with an EGFR inhibitor were evaluated for their eligibility in an open label phase II trial with E 150 mg/day. Pts ineligible for the latter (16) were treated with G 250 mg/day in a compassionate-use program. Fifteen pts were treated with G at progression after E. Results: Toxicities (all gr, gr 3/3–4) for E vs G consisted in: acneiform rash (64 vs 25%, 13 vs 0%), rash/desquamation (83 vs 19%, 12 vs 0%), pruritus (38 vs 16%, 6 vs 0%), dry skin (48 vs 32%, 3 vs 0%), nail changes (15 vs 6%, 3 vs 0%), diarrhea (54 vs 19%). When pts were treated with both E and G the toxicities (all gr, gr 3/3–4) observed were: acneiform rash (93 vs 20%, 27 vs 0%), rash/desquamation (100 vs 20%, 20 vs 0%), pruritus (41 vs 13%, 7 vs 0%), dry skin (54 vs 27%, 7 vs 0%), nail changes (27 vs 7%, 7 vs 0%), diarrhea (53 vs 7%). Conclusions: Our data are not conclusive being a non-randomized trial. However, these results suggest that there are less cutaneous adverse effects with G. This difference may be probably attributed either to the dose or the bioavailability of E. No significant financial relationships to disclose.

18138 Background: Two case reports describe a response to E after failure of G (Garfield DH, J Clin Oncol 2005) or to G after failure of E (Choong NW et al, Nat Clin Pract Oncol 2006) in patients (pts) with advanced NSCLC. Otherwise, a limited experience in 5 pts suggests that E is not effective in pts progressing on G (Viswanathan A et al, Lung Cancer 2005). Aim of this study was the evaluation of response and time to progression (TTP) in advanced NSCLC pts treated with G after failure of E. Methods: Pts received G 250 mg/day after disease progression (PD) with E 150 mg/day. Pts accrual was stopped on August 2006 after the approval of E for use in Italy and the consequent closure of the G compassionate-use program. Results: From May 2005 to August 2006, 15 pts were enrolled. Median age 65 years (50–85); males= 14 pts (93%); never/former smokers= 4/10 pts (26/67%); adenocarcinoma= 10 pts (67%); PS 0/1= 5/10 pts (33/67%); in 2 pts (13%) E was administered as first-line therapy, 8 pts (53%) received 2 prior lines of chemotherapy (CT) and 3 pts (20%) received CT between E and G. One patient (7%) had a partial response (PR) and 5 pts (33%) had disease stabilization (SD) with E; with G no PR and 6 SD (40%) were obtained. Five out of 6 RP/SD pts with E, had SD with G; 8 out of 9 PD pts with E, had PD with G; 1 SD patient with E, progressed with G and 1 vice versa. TTP in RP/SD pts was 7.2 and 3.4 months for E and G respectively; in PD pts TTP was 1.7 and 1.6 for E and G respectively. Conclusions: Our data suggest that there is a benefit with G in pts who had RP/SD with E and that is associated with a good TTP. Conversely G is not recommended in pts that immediately progressed after E. Moreover these results support the rationale of treating PD pts with an EGFR TKI with another one; it may be worthwhile to collect more data on E. No significant financial relationships to disclose.

18115 Background: One retrospective study (Hotta K, Ann Oncol 2005), investigating the prognosis of patients (pts) obtaining SD as best response with G treatment, has demonstrated that both progression-free survival (PFS) and survival (S) were significantly longer than those in pts with progressive disease (PD). The aim of this retrospective study was to compare the PFS and S outcome in pts with advanced NSCLC who achieved SD or partial response (PR) after treatment with E or G. Methods: Pooled data from 62 pts, entered into an open label phase II program of E (n=31) and a compassionate-use program of G (n=31), were retrospectively analyzed. E and G were given orally at 150 and 250 mg per day respectively and were continued until disease progression, development of unacceptable toxicity or patient’s refusal. Results: Pts characteristics: median age 69 years (42–85); females= 21 pts (34%); never/former smokers= 16/38 pts (26/61%); adenocarcinoma/BAC= 35/10 pts (56/16%); PS 0/1= 18/38 pts (29/61%). In 16 pts (26%) E or G were given as first-line therapy; 21 pts (34%) had received =2 prior lines of chemotherapy. Six pts (10%) achieved a PR and 18 pts (29%) obtained SD. TTP and OS in pts obtaining PR and SD were comparable: 7 vs 5.5 and 9.7 vs 9.1 months respectively. In progressing pts median TTP and OS were 1.7 and 3.7 months. Conclusions: Our findings indicate the importance of achieving disease control with both E and G treatment. Pts obtaining SD had a similar PFS and S compared with those having PR. An analysis of the role of mutational status and other biomarkers in predicting clinical outcome is currently underway. No significant financial relationships to disclose.

introductionThe role of chemotherapy in patients with advanced non-small cell lung cancer (NSCLC) has been established only in the last decade.The cornerstone is represented by the 1995 metaanalysis, which demonstrated the usefulness of platinum-based chemotherapy in this setting [1].In spite of the improved results of the last years using new drugs-combinations, chemotherapy still remains a palliative approach whose efficacy has to be balanced with its acute and cumulative toxicity.For these reasons, the duration of the chemotherapeutic treatment is a field of debate between clinical researchers and new chemotherapy strategies have been investigated, including maintenance chemotherapy.Maintenance chemotherapy is the prolongation of chemotherapy duration with the administration of additional chemotherapy at the end of a defined number of initial symposium article

17014 Background: Topotecan is the only single agent currently approved for the treatment of relapsed or recurrent SCLC, showing activity both in chemotherapy-refractory (RR 2–14%) and in chemotherapy-sensitive patients (RR 14–38%). The role of topotecan in combination with other active agents is still under investigation. Methods: Aim of this phase II study is to assess the activity and safety of topotecan (1mg/sqm iv d1–5) plus gemcitabine (1250 mg/sqm iv d1) in relapsed or recurrent SCLC patients. Treatment is repeated every 4 weeks, up to a maximum of 6 cycles. Eligibility criteria: histologically or cytologically confirmed SCLC; documented progressive disease after ≥ 1 chemotherapy regimen; age ≥ 18 yrs; ECOG PS 0–2; measurable disease (RECIST); no prior treatment with topotecan or gemcitabine; adequate hematologic, hepatic and renal function; brain metastases are allowed. Results: A total of 44 patients have been enrolled. Patient characteristics are as follows: median age, 64 yrs (range 35–77); male/female, 35/9; ECOG PS 0/1/2, 12/21/11 patients; 68% patients had sensitive disease (recurrence > 3 months after first-line chemotherapy) and 32% patients had refractory disease (failure ≤3 months after first-line chemotherapy). One-hundred and seventeen chemotherapy courses have been administered (median 2, range 1–6). The following preliminary results refer to 37 patients. Grade 3–4 toxicities: 54% neutropenia, 16% anemia, 46% thrombocytopenia, 13% neutropenic fever, 27% fatigue. One toxic death was observed. Objective responses have been documented in 9 patients, for an overall response rate of 24% (3% CR, 21% PR); SD was observed in 7 patients (19%), PD in 17 patients (46%). Four early deaths were reported. Median time to progression is 8.9 weeks. Median survival time is 16.3 weeks, and 1-year survival rate is 14%. Conclusions: Based on these preliminary results, the combination of topotecan plus gemcitabine shows moderate activity and an acceptable toxicity profile in previously treated SCLC patients. However, it is unlikely that the addition of gemcitabine improves the outcome compared to single agent topotecan. No significant financial relationships to disclose.