3500 Background: AtezoTRIBE (NCT03721653) is a phase II randomized trial in which unresectable mCRC pts were randomized 1:2 to 1 st -line FOLFOXIRI/bev [arm A] or FOLFOXIRI/bev/atezo [arm B]. Adding atezo to FOLFOXIRI/bev was safe and improved PFS (primary endpoint), with a modest benefit also among pts with pMMR tumors. Subgroup analyses suggest that TMB and Immunoscore IC (IS IC) -an IHC biomarker measuring CD8 and PD-L1 cell densities and their proximity- may identify pts with pMMR tumors deriving benefit from adding atezo to FOLFOXIRI/bev. Methods: The study had 85% power to detect a HR for PFS (time from randomization to 1 st PD or death [PD1]) of .66 in favor of arm B with 1-sided α error of .10. Secondary endpoints included PFS2 (time from randomization to PD on any treatment given after PD1 or death [PD2]), 2 nd PFS (time from PD1 to PD2), and OS. MMR, TMB, IS IC were correlated to clinical outcome. Results: 218 pts (arm A/B:73/145) were enrolled. Main pts’ characteristics were right-sided 44%/45%, RAS mut 71%/74%, BRAF mut 14%/8%, dMMR 7%/6%, high TMB 10%/12%, high IS IC 32%/32%. At a median follow-up of 37.0 mos, 175 (80%, arm A/B: 64/111) PD1, 150 (69%, arm A/B: 53/97) PD2, and 118 (54%, arm A/B: 43/75) OS events were collected. Out of 175 pts with a PD1 event, 135 (77%, arm A/B:50/85) received a subsequent treatment; among them, 121 pts (arm A/B: 43/78) had a PD2 event. PFS, PFS2, 2 nd PFS and OS results in the intention-to-treat (ITT) population and the pMMR group are listed in the Table. In the ITT population, significant interactions between treatment and MMR status (P int .011), TMB (P int .008), and IS IC (P int .037) were reported in terms of PFS. Only IS IC was associated with a differential OS benefit (P int .065), with pts bearing IS IC-high tumors deriving benefit from adding atezo (HR 0.43, 95%CI 0.19-1.00), differently than those with IS IC-low tumors (HR 1.09, 95%CI 0.65-1.83). In the pMMR group, significant interactions between treatment and TMB and IS IC were reported in terms of PFS (P int .016 and .051, respectively) and OS (P int .043 and .063, respectively). Pts bearing IS IC-high tumors derived higher OS benefit from adding atezo (HR 0.44, 95%CI 0.19-1.03), than those with IS IC-low tumors (HR 1.15, 95% CI 0.67-1.97). Conclusions: Pts with IS IC-high and/or TMB high pMMR mCRC seem to derive a survival benefit from adding atezo to FOLFOXIRI/bev as upfront treatment. These findings deserve confirmation in a properly designed phase III trial. Clinical trial information: NCT03721653 . [Table: see text]

Doublet chemotherapy plus anti-EGFR antibodies is notoriously more toxic in older than in younger patients with metastatic colorectal cancer (mCRC). Recent studies showed that a reduced-dose combination chemotherapy decreases the toxicity without compromising the effectiveness in old, vulnerable patients. A retrospective cohort study of the hospital medical records of RAS and BRAF wild-type, vulnerable patients aged ≥ 70 years with previously untreated mCRC was performed to assess the safety and efficacy profile of a doublet chemotherapy administered at reduced doses of 20% (FOLFOX or FOLFIRI) when combined with anti-EGFR antibodies (cetuximab or panitumumab). The primary endpoint was safety, and secondary endpoints were overall response rate (ORR), progression-free survival (PFS), and overall survival (OS). One hundred and eighteen patients were collected from 14 selected Italian centres. The median age was 75 (range, 70-85). Geriatric assessment by G8 tool gave a score < 14 in all patients; about a quarter of them (26%) had own tumor originally located on the right. In total, 75 and 43 patients received FOLFOX or FOLFIRI, respectively, in combination with panitumumab (53%) or cetuximab (47%). Median treatment duration was 5.5 months (1-13.5 months). The overall incidence of grade (G) 3-4 neutropenia was 11.8%, and for skin rash 11%; treatment was discontinued due to toxicity in 5% of patients. The most frequent adverse events (AE) were G1-2 skin rash (49.1%), G1-2 diarrhea (21.1%) and G1-2 nausea (17.7%). FOLFIRI-based regimens were associated with a high incidence of neutropenia, diarrhea, and asthenia, while there was a high incidence of anemia and peripheral sensory neuropathy in patients who received FOLFOX-based regimens. Five (4.2%) patients were hospitalized due to toxicity and 4 of them did so only once. There was no treatment-related death. The ORR was 57.3%. Stable disease was observed in 29.1% of patients, with a disease control rate of 86.4%. With a median follow-up of 18 months, the median PFS was 10.0 months (confidence interval of 95% [CI] 8.5-11.4), while the median OS was 18.0 months (95% CI: 16.0-19.9). No statistically significant difference was observed between the regimens in terms of ORR, PFS (p = 0.908), and OS (p = 0.832). This study shows that with an appropriate design, including reduced doses, vulnerable older patients best tolerate chemotherapy when combined with anti-EGFR antibodies. Larger confirmatory prospective studies are expected in the future.

To date, no biomarkers for second-line anti-angiogenic treatment in RAS wild type (wt) metastatic colorectal cancer (mCRC) patients (pts) with progression after first-line anti-epidermal growth factor receptor (EGFR) agents have been validated. Data on dynamic change of circulating pro-angiogenic factors levels during treatment from the pre-planned interim analysis of DISTINCTIVE trial (NCT04252456) are presented.

The approval trials have established sunitinib, multi-targeted TKI, a standard treatment in patients with metastatic RCC. RCTs may fail to show relevant clinical benefit in wider population and routine use, as certain patient subtypes are under-represented, notably those with poorer prognostic factors and pretreated. The data were collected through national registry Onco-AIFA as part of the mandatory surveillance. RCC patients receiving sunitinib once daily, on schedule 4/2, between 2007 and 2011, had been checked prospectively for toxicity, clinical outcomes and length of treatment. Based on the difference in effectiveness in PFS between approval RCT and clinical practice, a new price adjusted for effectiveness has been calculated. A total of 106 patients (64 years, M 70/F 36) were reviewed, 77% had prior nephrectomy and 74% were treatment-naïve. At first evaluation we found 28% partial responses, 25% stable diseases, 45% progressions and one discontinuation due to toxicity. The median PFS and OS were 7 and 11.3 months, respectively. Among 79 RCC patients (pts) with metastases (mets) at first diagnosis, 63% had lung mets, 21% pts had liver mets, 21% had bone mets and 9% had brain mets. Cox regression model showed in subgroups analysis significantly worse survival prognosis in males, brain and liver mets, ECOG PS > 1, non-clear cell histology and non prior nephrectomy. Sorafenib treatment after progression on sunitinib (33% pts) did not improved OS. Grade 3 thrombocytopenia, neutropenia, mucositis and HFS caused dosage reductions. Effectiveness adjusted price was 3,87 euro/mg as opposed to the ex-factory price of 2,46 euro/mg proportionally to the difference of PFS form RCTs (11 months) and that form clinical practice oncology (7 months). The results show that sunitinib clinical benefit in RCC patients was gained in selected responders, but in overall the effectiveness was nearly half of that reported in the approval RCTs. Evaluating cost-effectiveness ratio, price should be at least 36% lower than actual ex-factory price based on net clinical benefit achieved in real life practice. Post-marketing studies in real life practice are required in order to verify both effectiveness and safety in general population, testing for external validity of randomized trials.

ABSTRACT Background Study results demonstrated that IFN augments BEV activity and improves median PFS in pts with mRCC. Thus, combination BEV + IFN is a standard first-line treatment option for mRCC. Combining BEV with the mTOR inhibitor EVE may be an efficacious and well-tolerated treatment option. The open-label, phase II RECORD-2 trial compared first-line EVE + BEV and IFN + BEV in mRCC. Patients and methods: Therapy-naive pts with clear cell mRCC and prior nephrectomy were randomized 1:1 to BEV 10 mg/kg IV every 2 weeks with either EVE 10 mg oral daily or IFN (9 MIU SC 3 times/week, if tolerated). Tumour assessments were every 12 weeks. Primary objective was treatment effect on progression-free survival (PFS) per central review based on an estimate of the chance of a subsequent phase III trial success (50% threshold for phase II success). Results In EVE + BEV (n = 182) and IFN + BEV (n = 183) arms, median age was 60/60 years, 76/72% of pts were men, MSKCC risk was favourable/intermediate/poor in 36/57/7% and 36/57/7% of pts, and 43/46% of pts had >2 organs involved, respectively. For EVE + BEV and IFN + BEV, median treatment duration was 8.5/8.3 months, respectively; 23/26% of pts discontinued due to AEs. In EVE + BEV and IFN + BEV arms, median PFS by central review was 9.3/10.0 months (HRIFN/EVE, 0.91; 95% CI, 0.69-1.19; P =0.485), respectively; probability of subsequent phase III success was 5.1%. Results of central and local PFS analysis were consistent. Objective response rate was 27/28% in EVE + BEV and IFN + BEV arms, respectively. Median overall survival (OS) was not reached in the EVE + BEV arm and was 25.9 months (95% CI: 21.1, 30.2) in the IFN + BEV arm. Most frequent AEs (%) were stomatitis (63), proteinuria (49), diarrhoea (39), hypertension (38), and epistaxis (35) in EVE + BEV arm and decreased appetite (45), fatigue (41), proteinuria (37), and pyrexia (35) in IFN + BEV arm. Conclusions In RECORD-2, PFS and tolerability were similar for first-line EVE + BEV and IFN + BEV. Final OS analysis will occur after 2-year follow-up. Disclosure A. Ravaud: Alain Ravaud is a member of global, European, and/or French boards on urological tumors for Pfizer, Novartis, GlaxoSmithKline, Bayer-Schering, and Dendreon, and has received institutional grant support from Pfizer, Novartis, and Roche. O. Anak: Ozlem Anak is an employee of Novartis Pharma AG. D. Pelov: Diana Pelov is an employee of Novartis Pharmaceuticals Corporation. A. Louveau: Anne-Laure Louveau is an employee of Novartis Pharma S.A.S. T. M-H: Tay M-H is a speaker for an advisory board for Novartis Pharmaceuticals Corporation. B. Melichar: Bohuslav Melichar has received honoraria from Novartis and Roche and served on an advisory board for Roche. All other authors have declared no conflicts of interest.

Von Hippel-Lindau disease is a rare autosomal dominant inherited disorder that predisposes the occurrence of cysts and various types of cancers such as hemangioblastoma, pheochromocytoma, renal cell carcinoma and more rarely pancreatic tumors. In this review, we analyze the characteristics and management of pancreatic lesions, in particular cysts and neuroendocrine tumors, in Von Hippel-Lindau disease.

1084 Background: BRCA1-associated breast cancers have pathologic features and biologic behavior that are distinct from sporadic breast tumors while BRCA2 tumors are indistinguishable. Comparative survival studies have reported conflicting results and recent studies suggest relative resistance to taxane-based chemotherapy in patients with BRCA1 mutations. To determine if there are differences in long-term outcomes for BRCA1, BRCA2 and high-risk non-carriers with stage I, II, and III breast cancer based on adjuvant therapy received, we conducted a retrospective analysis in US and Italian patients. Methods: Patients diagnosed with breast cancer between January 1, 1990 and December 31, 2008 were identified at the University of Chicago Cancer Risk Clinic and the Veneto Institute of Oncology in Padua. There were 80 BRCA1carriers, 78 BRCA2 carriers and 177 high-risk non-carriers ascertained within the same time period. End points were recurrence-free survival (RFS) and overall survival (OS) by type of adjuvant chemotherapy received. Hazard Ratios (HR) and 95% Confidence Intervals (95% CI) were calculated from Cox proportional hazard models. A multivariate analysis was used to adjust for year at diagnosis, stage, grade and Estrogen Receptor (ER) status. Results: The median follow up was 4.98 (0.03-18.9). The RFS-adjusted hazard ratios were not significantly different among mutation carriers and non-carriers (HR among BRCA1 carriers, 0.52; 95% CI, 0.21 to 1.31; HR among BRCA2 carriers, 0.68; 95% CI, 0.31 to 1.48; P=0.326). The OS-adjusted hazard ratios were not significantly different (HR among BRCA1 carriers, 1.24; 95% CI, 0.45 to 3.45; HR among BRCA2 carriers, 0.89; 95% CI, 0.33 to 2.38; P=0.837). Ten year survival was 78% for BRCA1, 83% for BRCA2 and 92% for non-carriers, respectively. The risk of recurrence and death was also not statistically significant among patients treated with different chemotherapy regimens. Conclusions: RFS and OS survival are similar for carriers of a BRCA mutation and non-carriers.

About 10% of all ovarian cancers are due to BRCA 1 and/or BRCA 2 mutations. Some studies have shown that patients belonging to this group have a better survival compared to sporadic groups but data are still inconclusive. The aim of this study was to investigate overall survival in patients with ovarian cancer and germ-line mutations in the BRCA1/2 genes in comparison to high-risk patients, defined as patients with ovarian cancer and a strong family history of breast and ovarian cancer, but who tested negative for the BRCA mutation. We collected all the clinical features and did follow-up. The two groups showed similar characteristics concerning age at diagnosis, histological type and stage. Grade 3 was more frequent in the BRCA group. Survival data did not show any advantage for the BRCA mutated group.

e14618 Background: The aim of this study was to identify factors that could predict TTP and OS for patients with uHCC treated by combination G and PLD chemotherapy. G was administered before PLD because G induces an increase of topoisomerase II expression, and a subsequent anti-topoisomerase II drug (PLD) causes a more relevant cytotoxic effect. Methods: We examined 41 consecutive patients with uHCC treated from November 2003 to May 2008, in a prospective study. All patients underwent gemcitabine 1,000 mg/m2 on day 1 and 8, PLD 30 mg/m2 on day 1, once every 4 weeks up to a maximum of eigth cycles. Univariate and multivariate analyses of patient and disease characteristics were used to identify factors predicting TTP and OS. Results: The median TTP was 5.8 months (95% CI 2.6-8.9), the median OS was 22.5 months (95% CI 4.6-40.3). According to univariate analysis, the following variables were related to a longer TTP: good performance status (PS) (p = 0.004), Child-Pugh A (p < 0.001), earlier stage disease (Okuda stage, p < 0.001), AFP < 400 ng/mL (p = 0.04); PS (p = 0.04) and Okuda staging (p < 0.001) were found to be independent predictors for TTP on multivariate analysis. On univariate analysis: good PS (p = 0.003), Child-Pugh A (p = 0.03), earlier stage disease (Okuda stage, p < 0.001), previous hepatic resection (p = 0.04), AFP < 400 ng/mL were related to OS; good PS, earlier stage disease and AFP < 400 ng/mL were associated with a longer OS on multivariate analysis. Among patients with elevated serum AFP (> 400 ng/mL) before the treatment, a decrease in AFP of more than 20% after two cycles of chemotherapy was independent favourable predictor for TTP (p < 0.001) and OS (p = 0.02). Conclusions: Patients with uHCC who have a good PS and earlier stage disease have a better chance of prolonged TTP; because of the greater probability of survival, patients with good PS, earlier stage disease and AFP<400ng/ml might be good candidates for G and PLD chemotherapy. AFP response is an indipendent predictor for TTP and OS. This treatment should be tested in combination with targeted agents such as sorafenib, erlotinib and bevacizumab. No significant financial relationships to disclose.

The major symptom at diagnosis of endometrial cancer is post-menopausal bleeding; it is present in around 90% of cases. Singular bone metastasis is described as an uncommon site for endometrial cancer at diagnosis, showing in just 5-6% of cases. In this report we describe a rare presentation of a singular bone metastasis because of endometrial cancer of a woman with previous diagnosis of early breast cancer. A review of literature uncovered some cases of bone metastasis at presentation of endometrial cancer and that it can occur as first symptom of cancer before vaginal bleeding. This rare presentation of uterine cancer needs to be studied and described because it may be seen and needs a homogeneous treatment to improve survival.

606 Background: Neoadjuvant chemotherapy is currently the standard of care for the management of locally advanced breast cancer, as it improves both disease-free and overall survival. Methods: Patients with histologic confirmation of locally advanced breast cancer, age >18 years, left ventricular ejection fraction >50%, good performance status and adequate bone marrow, negative chest and abdominal CT scan, renal and hepatic function were included in the study. Prior systemic therapy and radiotherapy and the presence of metastases were not allowed. The treatment schedule was as follows: GC: 1,250 mg/m2 (day 1); GC 1,000 mg/m2 plus TT 75 mg/m2 plus DX 30 mg/m2 (day 8). Pegfilgrastim was routinely administered as primary prophylaxis against febrile neutropenia. Courses were repeated each 21 days for 4–6 cycles. This study was aimed to assess the activity of GC, DX and TT in locally advanced breast cancer. Results: Fifty patients were enrolled and treated. Median age was 51 years (range: 34–73), 27 (54%) being postmenopausal patients. Infiltrating ductal carcinoma was diagnosed in 84% of cases. Tumor size was as follows: T1 (n=2, 4%), T2 (n= 31, 62%), T3 (n=14, 28%), T4 (n=2, 4%). G3 grading was found in 28% of cases. Median tumor size was 4 cm (range: 2–10). Tumors were ER+, PR+ and c-Erb-B2+ in 34 (68%), 31(62%) and 27 (54%) cases, respectively. Surgery was performed in 47 (94%) patients: 12 (26%) pathologic complete response, 24 (51%) partial response and 9 (19%) stable disease were observed resulting in a pathologic overall response rate of 77%. Logistic regression analysis revealed that larger tumor size, Mib1 higher expression and higher number of cycles were independently associated with better tumor response rates. Conclusions: This combination chemotherapy can yield a high tumor response rate. Correlation analysis suggests that this neoadjuvant regimen might be more effective in patients with larger tumors expressing Mib1, particularly if full treatment can be administered. No significant financial relationships to disclose.