2509 Overall survival analysis from a randomised phase III trial of axitini vs sorafenib as first-line therapy in patients with metastatic renal cell carcinoma

S476 Methods: The association of factors measured at trial entry including age, gender, performance status, stage, grade, and size of tumour, extent of resection, presence of residual mass (RM), haemoglobin (Hb), whole blood count (WBC) and glomerular filtration rate (GFR), plus RT fractionation regimen (55Gray(Gy)/20 fractions(f) or 64Gy/32f) and use of neoadjuvant chemotherapy (neoCT) with the study outcomes were explored by Kaplan– Meier curves and Cox proportional hazards models. Full and reduced models (using backward stepwise selection) were fitted to identify the strongest prognostic factors. Results: Complete data for all factors was available on 332 of 360 patients recruited to the BC2001 chemotherapy randomisation. Median follow up was 69.9 months. For any locoregional recurrence (LRD) and invasive local recurrence (ILRD) the strongest predictive factor for recurrence was presence of RM prior to RT (LRD HR: 1.52 95% confidence interval (CI) 1.01–2.29, p = 0.04; ILRD 2.57 CI: 1.56–4.24, p < 0.0001). Presence of RM was highly correlated with extent of resection (96% complete resections without RM, 66% incomplete resections with RM). The use of cRT conferred similar benefit in the full multivariable model (LRD HR: 0.68 CI: 0.47– 0.99, p = 0.05)to to that previously reported (James et al N Engl J Med 2012;366:1477−88). There was a trend to better outcome with 55Gy/20f v 64Gy/32f (LRD: HR 1.36 CI: 0.92–2.02, p = 0.12; ILRD: 1.67 CI: 1.0– 2.79, p = 0.05). This result may be confounded by greater use of neoCT in sites using 55Gy/20f (41%) than in those 64Gy/32f (27%). In a reduced model prognostic factors for ILRD were RM after resection, cRT and use of neoCT; planned radiotherapy fractionation schedule was not statistically significant.. For overall survival (OS) key prognostic factors were presence of RM (HR: 2.09 CI: 1.55–2.82, p < 0.0001) and WHO performance status (HR for PS1/2: 1.79 CI: 1.32–2.41, p < 0.0001) and age (HR: 1.02 p = 0.02). There was a non statistically significant trend for cRT improving OS (HR 0.79 CI 0.579–1.045 p = 0.091]. Conclusions: The benefit for the addition of MMC/5FU to radiotherapy was maintained in an extended multivariate analysis of prognostic factors in patients from the BC2001 study. The main prognostic factors for outcome were related to tumour bulk as indicated by the presence of a RM post TURBT both for both LRD and OS. No conflict of interest. 2508 ORAL Three-arm phase II randomized trial of docetaxel monotherapy or combined with ramucirumab or icrucumab in second-line locally advanced or metastatic urothelial carcinoma D.P. Petrylak1 , S.T. Tagawa2 , M. Kohli3 , A. Eisen4 , C. Canil5 , S.S. Sridhar6 , A. Spira7,8 , E.Y. Yu9,10 , J.M. Burke11,12 , D. Shaffer13,14 , C.X. Pan15 , J.J. Kim16,17 , J.B. Aragon-Ching18 , S. Tang19 , H. Zhang19 , C.T. Cavanaugh20 , J.S. Kauh21 , R.A. Walgren22 , K.N. Chi23 . 1 Yale Univ Cancer Center, Genitourinary Oncology Research Program, New Haven, USA; 2 Weill Cornell Medical College, Division of Hematology and Medical Oncology, New York, USA; 3 Mayo Clinic, Department of Medical Oncology, Rochester, USA; 4 Sunnybrook Health Sciences Center, Medical Oncology/Hematology, Toronto, Canada; 5 The Ottawa Hospital Cancer Centre, Division of Medical Oncology, Ottawa, Canada; 6 Princess Margaret Hospital, GU Medical Oncology, Toronto, Canada; 7 Virginia Cancer Specialists, Medical Oncology Hematology Stem Cell Transplant, Fairfax, USA; 8 US Oncology Research, Lung Program, The Woodlands, USA; 9 University of Washington Medical Center, Department of Medical Oncology, Seattle, USA; 10 Fred Hutchinson Cancer Research Center, Departmeny of Medicine, Seattle, USA; 11 Rocky Mountain Cancer Centers, Medical Oncology, Aurora, USA; 12 US Oncology Research, McKesson Specialty Health Affiliated Investigators, The Woodlands, USA; 13 New York Oncology Hematology, P.C., USON Affilitate, New York, USA; 14 US Oncology Research, New York Oncology Hematology, The Woodlands, USA; 15 University of California, Davis Medical Center, Department of Internal Medicine, Sacramento, USA; 16 Johns Hopkins University, Geriatrics Center, Baltimore, USA; 17 Sinai Hospital, Division of Hematology/Medical Oncology, Baltimore, USA; 18 George Washington University Medical Center, Department of Medicine, Washington DC, USA; 19 Eli Lilly and Company − Bridgewater, Statistics Oncology, Bridgewater, USA; 20 Eli Lilly and Company − Bridgewater, Clinical Trial Management, Bridgewater, USA; 21 Eli Lilly and Company − Bridgewater, Oncology − Early Phase, Bridgewater, USA; 22 Eli Lilly and Company, Oncology − Early Phase, Indianapolis, USA; 23 BC Cancer Agency, Clinical Trials Unit, Vancouver, USA Background: Patients (pts) with locally advanced or metastatic urothelial carcinoma (UC) progressing after first-line platinum chemotherapy have limited treatment options. Ramucirumab (RAM), a human IgG1 VEGFR-2 antibody, has demonstrated clinical efficacy in several solid tumors. Icrucumab (ICR) is a human IgG1 VEGFR-1 antibody. Study JCDC was an open-label, multicenter, randomized phase II trial designed to evaluate

Abstracts progression-free survival (PFS) in pts with locally advanced or metastatic platinum-resistant UC treated with docetaxel (DOC) monotherapy or DOC in combination with either RAM or ICR. Materials and Methods: Eligible pts with locally advanced or metastatic UC, disease progression 1 year after a platinum regimen, and ECOG PS 0−1 were randomized to receive DOC (75 mg/m2 IV on Day 1 repeated every 21-days) monotherapy or in combination with RAM (10 mg/kg IV on Day 1) or ICR (12 mg/kg IV on Days 1 and 8). Treatment continued until disease progression or unacceptable toxicity interrupted the use of one or more agents. The primary endpoint was investigator-assessed PFS with RECIST v1.1, analyzed by K-M methods with stratified log-rank test and Cox Proportional Hazards model. A sample size of 46 pts per arm enabled 71% power to demonstrate statistical significance of 4.5 mo (RAM or ICR and DOC) vs 3 mo (DOC) at a 1-sided a= 0.1. Results: Study arms were well matched across arms; median age was 68 yrs (29−85), 78.6% were male, 60.7% had ECOG PS 1, and 70.7% had visceral metastases. The median number of infusions was 1.5x more for DOC+RAM vs DOC (4.5 vs 3.0). PFS, overall survival (OS), objective response rate (ORR), and disease control rate (DCR) results are shown in the table below. 101 pts had Grade 3 adverse events (AEs) related to study drug, 55.6% DOC, 78.3% RAM+DOC, 81.6% ICR+DOC. Drugrelated Grade 3 AEs (DOC, RAM+DOC, ICR+DOC) reported 15% in any study arm were: fatigue (13.3, 26.1, 20.4%), neutropenia (26.7, 21.7, 26.5%) and febrile neutropenia (13.3, 17.4, 4.1%). Table: Final efficacy results

PFS Median, mo (95% CI) Stratified HR (95% CI) P-value OS Median, mo (95% CI) Stratified HR (95% CI) P-value ORR n, (%) P-value† DCR n, (%) P-value† † Derived

DOC (N = 45)

RAM + DOC (N = 46) ICR + DOC (N = 49)

2.8 (1.9, 3.6)

5.4 (3.1, 6.9) 0.389 (0.235, 0.643) <0.001

1.6 (1.4, 2.9) 0.863 (0.550, 1.357) 0.505

9.2 (5.7, 11.7)

10.4 (7.0, 15.1) 0.733 (0.454, 1.183) 0.201

6.7 (4.5, 8.5) 1.001 (0.630, 1.589) 0.998

4 (9)

11 (24) 0.088

6 (12) 0.742

26 (58)

36 (78) 0.044

22 (45) 0.224

from two-sided Fisher’s exact test.

Conclusions: The primary endpoint was met with a statistically significant median PFS improvement of >2.5 months in pts treated with RAM+DOC vs. DOC. Significant improvements were also observed in DCR in the RAM+DOC arm, but the difference in OS was not significant. ICR+DOC arm did not meet the primary endpoint. The safety profile was acceptable and in line with expectations. Conflict of interest: Ownership: S. Tang − Lilly. R. A. Walgren − Lilly. C. T. Cavanaugh − Lilly. Advisory Board: D. P. Petrylak − Sanofi, Genentech, Eli Lilly and Company. C. Canil − Janssen, Amgen. J. M. Burke − Dr. Reddy’s Laboratories, Gilead, Millenium, Incyte, and Janssen. K. N. Chi − Lilly. Corporate-sponsored Research: D. P. Petrylak − Sanofi, Genentech, Eli Lilly and Company. E. Y. Yu − Eli Lilly and Company. Other Substantive Relationships: C. Canil − Speakers Fees: Sanofi, Janssen, Astellas, Amgen,. Educational Award Grant: Sanofi, Novartis. S. Tang − Employment: Lilly. C. T. Cavanaugh − Lilly. J. S. Kauh − Employment: Lilly. R. A. Walgren − Employment: Lilly. 2509 ORAL Overall survival analysis from a randomised phase III trial of axitinib vs sorafenib as first-line therapy in patients with metastatic renal cell carcinoma T.E. Hutson1 , S. Al-Shukri2 , V.P. Stus3 , O.N. Lipatov4 , Y. Shparyk5 , A.H. Bair6 , B. Rosbrook7 , G.I. Andrews6 , N.J. Vogelzang8 . 1 Baylor Sammons Cancer Center, GU Oncology Program, Dallas, USA; 2 First Saint-Petersburg State Pavlov Medical University, Urology, Saint Petersburg, Russian Federation; 3 Municipal Institution “Dnipropetrovs’k Regional Clinical Hospital n.a. I.I. Mechnikov”, Urology, Dnipropetrovs’k, Ukraine; 4 Republican Clinical Oncology Dispensary, Oncology, Ufa, Russian Federation; 5 Lviv State Oncologic Regional Treatment and Diagnostic Center, Department of Chemotherapy, Lviv, Ukraine; 6 Pfizer Oncology, Clinical Development, San Diego, USA; 7 Pfizer Oncology, Clinical Statistics, San Diego, USA; 8 Comprehensive Cancer Centers of Nevada, US Oncology Research, Las Vegas, USA Background: Axitinib, a potent and selective inhibitor of vascular endothelial growth factor receptors, is approved to treat advanced renal cell

Abstracts carcinoma (RCC) after 1 prior systemic therapy. In a randomised phase III trial (NCT00920816; sponsor: Pfizer) in previously untreated patients with metastatic RCC (mRCC), axitinib vs sorafenib yielded numerically longer progression-free survival (PFS; median 10.1 vs 6.5 mo; stratified hazard ratio [HR] 0.77; 95% confidence interval [CI] 0.56, 1.05; 1-sided P = 0.038) and significantly higher objective response rate (ORR; 32% vs 15%, 1-sided P = 0.0006). Overall survival (OS) data, as of Dec 18, 2014, are reported. Material and Methods: A total of 288 treatment-na¨ıve patients with measurable (per RECIST v1.0), clear-cell mRCC and Eastern Cooperative Oncology Group performance status (ECOG PS) 0 or 1 were enrolled from Eastern Europe (51%), Asia (25%), North America (14%), and South America (10%). Patients were stratified by ECOG PS and randomised 2:1 to receive open-label axitinib 5 mg twice daily (n = 192) or sorafenib 400 mg twice daily (n = 96). Primary endpoint was PFS per blinded independent review committee; ORR and OS were secondary endpoints. Results: Median OS (95% CI) was 21.7 mo (18.0, 31.7) with axitinib vs 23.3 mo (18.1, 33.2) with sorafenib (stratified HR 0.995; 95% CI 0.731, 1.356; 1-sided P = 0.4883). In patients with ECOG PS 0, median OS (95% CI) was 41.2 mo (29.2, not estimable [NE]) with axitinib vs 31.9 mo (18.1, NE) with sorafenib (HR 0.811; 95% CI 0.522, 1.259; 1-sided P = 0.1748). In patients with ECOG PS 1, median OS (95% CI) was 14.2 mo (9.4, 18.1) with axitinib vs 19.8 mo (12.3, 25.8) with sorafenib (HR 1.203; 95% CI 0.778, 1.859; 1-sided P = 0.7973). Incidence and severity of common adverse events were consistent with previous reports. Of 164 and 83 patients who discontinued axitinib and sorafenib treatment, respectively, 29 (18%) and 19 (23%) received follow-up systemic therapy. Conclusions: Median OS was similar with axitinib and sorafenib in the overall population of previously untreated patients with mRCC, and no new safety signals emerged after long-term follow-up. Median OS was numerically longer with axitinib vs sorafenib in patients with ECOG PS 0 and numerically longer with sorafenib vs axitinib in patients with ECOG PS 1. The noteworthy difference in median OS in patients with ECOG PS 0 vs 1 in the axitinib arm (41.2 vs 14.2 mo) may have been influenced by practice standards in resource-limited geographic regions where the majority of patients were enrolled, as well as subjectivity in classifying patients to ECOG PS categories. Conflict of interest: Ownership: A.H. Bair, B. Rosbrook, and G.I. Andrews own stock in Pfizer. Advisory Board: TE Hutson has served as an adviser for Pfizer, Novartis, GlaxoSmithKline, Bayer, Genentech, and AVEO. Corporate-sponsored Research: TE Hutson has received research funding from Pfizer, Novartis, GlaxoSmithKline, Bayer, Genentech, and AVEO. Other Substantive Relationships: TE Hutson has received honoraria from Pfizer, Novartis, GlaxoSmithKline, Bayer, Genentech, and AVEO. A.H. Bair, B. Rosbrook, and G.I. Andrews are employees of Pfizer. N.J. Vogelzang is on a speakers bureau for Pfizer.

Poster Discussion Session and Poster Session (Monday, 28 September) Genitourinary Malignancies 2510 POSTER DISCUSSION/POSTER 3-year follow-up of chemotherapy following radium-223 dichloride (Ra-223) in castration-resistant prostate cancer (CRPC) patients (Pts) with symptomatic bone metastases (Mets) from ALSYMPCA O. Sartor1 , R.E. Coleman2 , S. Nilsson3 , N.J. Vogelzang4 , K. Staudacher5 , M. Thuresson6 , C. Parker7 . 1 Tulane Cancer Center, Departments of Medicine and Urology, New Orleans, USA; 2 Weston Park Hospital, University of Sheffield, Academic Unit of Clinical Oncology, Sheffield, United Kingdom; 3 Karolinska University Hospital, Department of Oncology, Stockholm, Sweden; 4 Comprehensive Cancer Centers of Nevada, Department of Medical Oncology, Las Vegas, NV, USA; 5 Bayer AS formerly Algeta ASA, Study Management Europe, Oslo, Norway; 6 Statisticon AB, Biostatistics Department, Uppsala, Sweden; 7 The Royal Marsden NHS Foundation Trust and Institute of Cancer Research, Academic Urology Unit, Sutton, United Kingdom Background: Ra-223 is approved for pts with CRPC, symptomatic bone mets, and no visceral mets based on ALSYMPCA results. Ra-223 prolongs overall survival (OS), whether administered with prior or no prior docetaxel treatment (tx) (Hoskin Lancet Oncol 2014). Chemotherapy (chemo) after Ra-223 tx was safe and well tolerated (Sartor ESMO 2012). These current post hoc analyses evaluated chemo after ALSYMPCA study tx, including an additional 62 pts and a more extensive analysis of 3-year follow-up data. Materials and Methods: The analysis includes ALSYMPCA pts who had chemo following study tx (Ra-223 or placebo [pbo]). Chemo agents were

S477 identified, and time from last study tx to start of chemo and its duration were calculated. OS was analyzed, and hematologic (heme) safety based on lab values reviewed. Results: 142/614 (23%) Ra-223 and 64/307 (21%) pbo pts had chemo following study tx, including an additional 52 Ra-223 and 10 pbo pts from time of preliminary analysis in 2012. Most common chemo agents were docetaxel (100 [69%] Ra-223, 46 [72%] pbo) and mitoxantrone (23 [16%] Ra-223, 13 [20%] pbo). Median time to chemo was 35 days longer after Ra-223 (115 d) vs pbo (80 d). Median chemo duration was longer with Ra223 (141 d) than with pbo (127 d). Median total daily docetaxel dose was lower with Ra-223 (93 mg) than with pbo (120 mg). Percentages of Ra-223 vs pbo pts with grade 3/4 heme values were 9/116 (8%) vs 2/49 (4%) for hemoglobin; 11/114 (10%) vs 1/48 (2%) for neutrophils; and 1/116 (1%) vs 0/49 (0%) for platelets. Heme lab measurements over time are presented (table). Median OS from start of chemo was 16 months following Ra-223 and 15.8 months following pbo. Similar proportions of Ra-223 pts (41/142 [29%]) and pbo pts (21/64 [33%]) died during and 30 days after chemo; most common cause was prostate cancer–related death, in 37/41(90%) Ra-223 and 20/21(95%) pbo pts. Conclusions: This post hoc analysis showed that pts could receive chemo following Ra-223 with OS and heme safety profiles similar to those for chemo after pbo with up to 3 years of follow-up. Table: Median heme measurements Time after start of chemo, mo

Hemoglobin, g/dL Baseline* 2 4 6 12 18 Neutrophils (absolute), ×109 /L Baseline* 2 4 6 12 18 Platelets, ×109 /L Baseline* 2 4 6 12 18

Ra-223 (n = 142)

Pbo (n = 64)

N

Median (range)

N

Median (range)

142 82 69 50 28 17

11 11 11 11 11 11

64 42 25 22 8 8

11 11 10 11 10 11

142 78 66 46 24 14

4 4 4 4 5 4

64 38 23 19 8 7

5 6 5 5 4 4

142 82 69 50 28 17

215 199 209 216 215 246

64 42 25 22 8 8

236 (48–432) 250 (48–385) 220 (131–411) 227 (99–334) 215 (89–373) 311.0 (143–378)

(6–108) (7−14) (6−86) (9–121) (8−13) (9–117)

(1−26) (1−24) (<1−13) (1−16) (2−8) (1−8) (<1–782) (<1–692) (48–458) (55–494) (50–464) (78–601)

(8−15) (8−14) (7−15) (8−14) (8−13) (9−14)

(2−13) (1−13) (1−13) (2−12) (3−9) (2−9)

*Last nonmissing measurement prior to start of first poststudy chemo.

Conflict of interest: Ownership: Dr Vogelzang has owned stock or held an ownership interest in Caris Life Sciences. Advisory Board: Dr Sartor has had a consulting or advisory role with Bayer AS (formerly Algeta ASA) and Bayer HealthCare. Drs Coleman and Nilsson with Bayer HealthCare. Dr Vogelzang with Janssen Biotech, Amgen, Aveo, and BIND Therapeutics. Dr Thuresson with Bayer. and Dr Parker with Bayer, BNIT, and Janssen. Corporate-sponsored Research: Dr Sartor has received research funding from Bayer AS (formerly Algeta ASA) and Bayer HealthCare. Dr Coleman (to his instutition) from Celgene and Amgen. Dr Vogelzang (to his institution) from Parexel, Progenics, Exelexis, US Oncology, Viamet Pharm, Endocyte, and Glaxo Smith Kline. and Dr Parker (to his institution) from Bayer. Other Substantive Relationships: Ms Staudacher is employed by Bayer AS (formerly Algeta ASA). 2511 POSTER DISCUSSION/POSTER TI-CE for relapsed germ cell tumours: A European BMT centre experience G. O’Kane1 , L. Bacon2 , P. Browne2 , E. Conneally2 , P. Daly1 , M. Kennedy1 , E. Vandenberghe2 , P. Hayden2 , D. O’Donnell1 . 1 St. James’s Hospital, Medical Oncology, Dublin, Ireland; 2 St. James’s Hospital, Haematology, Dublin, Ireland Background: Optimal therapy of relapsed germ cell tumours (GCT) is uncertain. The TI-CE protocol of induction Paclitaxel/Ifosfamide (TI) and high-dose Carboplatin/Etoposide with autologous stem cell (PBSC) support (CE) was developed in Memorial Sloan-Kettering Hospital for relapsed GCT with poor prognostic features for response to conventional-dose chemotherapy (CDCT). The EORTC-sponsored TIGER trial compares TICE to CDCT in a wider range of relapsed GCT patients (pts). We describe TI-CE experience in a European BMT centre. Materials and Methods: We reviewed records of pts who had high-dose chemotherapy for relapsed GCT in St. James’s Hospital, Dublin, Ireland from 1998 to 2014. We included pts planned for TI-CE and with at least