J-P Machiels

Afatinib versus methotrexate in older patients with second-line recurrent and/or metastatic head and neck squamous cell carcinoma: subgroup analysis of the LUX-Head & Neck 1 trial

In the LUX-Head & Neck 1 study, older age (≥65 years) did not adversely affect the benefit in patient-reported outcomes and antitumor activity observed with afatinib over methotrexate, which was consistent with findings from the overall population. Safety in older patients was also consistent with the overall population, favoring afatinib in terms of fewer dose reductions and discontinuations.

989PDPHASE 1B STUDY OF MEHD7945A (MEHD) PLUS CISPLATIN/FLUOROURACIL (CIS/5FU) OR CARBOPLATIN/PACLITAXEL (CARBO/PAC) FOR 1ST-LINE TREATMENT OF RECURRENT/METASTATIC SQUAMOUS CELL CARCINOMA OF HEAD AND NECK (RMSCCHN)

ABSTRACT Aim: MEHD, a novel dual-action humanized IgG1 antibody that blocks ligand binding to HER3 and EGFR, inhibits signaling from ligand-dependent HER dimers. MEHD is active in multiple tumor models, including models resistant to anti-EGFR or anti-HER3, and enhances activity of chemotherapeutic agents. Single-agent activity in Phase 1a included confirmed PR in 2 RMSCCHN patients (pts) who had high levels of the HER3 ligand NRG1. Methods: This open-label, multicenter, Phase 1b study with a modified 3 + 3 + 3 design assesses safety, PK and preliminary anti-tumor activity (RECIST v1.1) of MEHD plus platinum-based chemotherapy in 1st-line RMSCCHN pts. MEHD 1650 mg IV every 3 wks is combined with cis 100 mg/m2 / 5FU 1000 mg/m2/d on Days 1-4 (Arm A) or carbo (AUC 6 mg/mL·min) / pac 200 mg/m2 (Arm B) on Day 1 of 21-d cycles (up to 6), followed by MEHD maintenance until disease progression / intolerable toxicity. Mandatory tumor samples are assayed by qRT-PCR for biomarkers related to mechanism of action and SCCHN. Results: As of 28MAR14, 18 pts were treated and remain active: 6 ARM A pts have received 2-7 cycles (median 5.5) of MEHD; 13 Arm B pts have received 1-8 cycles (median 3.5). DLTs occurred in 2 pts in Arm A (1 G3 diarrhea, 1 G3 acute renal failure & G3 febrile neutropenia) and 1 pt in Arm B (G3 dehydration, anorexia). G ≥ 3 treatment-related AEs in ≥2 pts were hypokalemia (3), neutropenia (3), dehydration (3), fatigue (2), and hyponatremia (2) in Arm A and neutropenia (6), febrile neutropenia (3), and hyponatremia (2) in Arm B. Chemo dose was reduced in 9/18 pts. Preliminary MEHD PK in both arms was similar to single-agent MEHD profile. In 12 pts with on-treatment tumor assessments, best responses were 9 (75%) PR (4 confirmed) [Arm A: 4; Arm B: 5; HPV: 2 + , 5-, 2 unknown], and 3 (25%) stable disease [Arm A: 1; Arm B: 2]; after data cutoff 2 pts were reported with CR. Further biomarker data are pending. Conclusions: MEHD plus cis/5FU or carbo/pac has been reasonably well tolerated with no new safety signals. G ≥ 3 AEs were manageable and less frequent post-Cycle 1. Both combinations had promising anti-tumor activity. Updated results will be presented. Disclosure: T. Seiwert: Research funding; X. Wang, A. Kapp, S. Royer-Joo, E. Penuel, B. McCall and A. Pirzkall: Genentech employee. All other authors have declared no conflicts of interest.

Abstract P1-12-06: UCBG intergroup: 3-years efficacy results of the Unicancer-PACS08 trial including poor prognosis patients treated with docetaxel or ixabepilone in adjuvant setting

Background: Ixabepilone, an epothilone B analog, has demonstrated single-agent activity in metastatic and neoadjuvant settings. The PACS08 trial aimed to compare adjuvant FEC100-Docetaxel regimen to FEC100-Ixabepilone in poor prognosis early breast cancer (BC) composed of patients presenting with triple-negative (TN) [i.e. estrogen receptor (ER)-/progesterone receptor (PR)-/HER2-] or ER+/PR-/HER2- tumor, which are subgroups significantly associated with worse prognosis. Patients and methods: Between 2007 and 2010, 762 patients with unilateral TNBC (n=592, 78%) or node-positive ER+/PR-/HER2- BC (n=170, 22%) were enrolled. Recruitment was interrupted due to BMS decision to stop ixabepilone development in adjuvant setting. Main inclusion criteria were: age Results: Main pts characteristics were well balanced between the 2 arms. As of September 2014, the median follow-up was 36 months. The safety profile indicates that Docetaxel is more often associated to significant haematological toxicities whereas both neurotoxicities and haematological toxicities are reported in Ixabepilone arm. Log-Rank tests indicate no difference between two arms in terms of both DFS and OS (HR=1.2, 95%CI (0.864-1.728), p=0.256 and HR=1.8, 95%CI (0.751-1.855), p=0.473, respectively). Pathological analysis of the PACS08 collection showed that TNBC displayed significantly higher proliferative activity as shown by mitotic count and Ki67 index (p Conclusions: Our results indicate that Ixabepilone doesn9t show higher efficiency compared to Docetaxel in adjuvant setting in poor prognosis early breast cancer. We have an unusual biological collection associated to our clinical data which will allow us to correlate efficacy data to breast cancer subgroups. Other several translational researches are still ongoing. Citation Format: Campone M, Lacroix-Triki M, Roca L, Spielmann M, Wildiers H, Cottu P, Kerbrat P, Levy C, Mayer F, Bachelot T, Wiston T, Eymard J-C, Uwer L, Machiels J-P, Verhoeven D, Jaubert D, Facchini T, Orfeuvre H, Canon J-L, Asselain B, Lemonnier J, Roche H. UCBG intergroup: 3-years efficacy results of the Unicancer-PACS08 trial including poor prognosis patients treated with docetaxel or ixabepilone in adjuvant setting. [abstract]. In: Proceedings of the Thirty-Eighth Annual CTRC-AACR San Antonio Breast Cancer Symposium: 2015 Dec 8-12; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2016;76(4 Suppl):Abstract nr P1-12-06.

784PPHARMACOKINETIC (PK) ACTIVITY OF CABAZITAXEL (CBZ) IN PATIENTS (PTS) WITH RENAL IMPAIRMENT (RI)

ABSTRACT Aim: Limited data are available on Cbz PK in pts with RI (14 with moderate RI, 1 with severe RI [Ferron, Cancer Chemother Pharmacol 2013]). This open-label, multicentre study assessed Cbz PK in pts with advanced solid tumours and normal or impaired renal function. Methods: Pts enrolled into cohorts A (normal control, creatinine clearance [CrCL] >80 ml/min/1.73 m2), B (moderate RI, CrCL 30– Results: Pts (n = 25) received a median of 3 Cbz cycles (range 1–20: cohort A, 5 [2–13]; cohort B, 3 [1–15]; cohort C, 5 [1–20]), and 24 were eligible for PK analysis (8/cohort). For moderate and severe RI vs controls, GMR estimates were: CL/BSA 0.95 (90% CI 0.80–1.13) and 0.89 (0.61–1.32); AUC/dose 1.06 (0.88–1.27) and 1.14 (0.76–1.71); FU 0.99 (0.94–1.04) and 0.97 (0.87–1.09), respectively. Estimated slope of linear regression of log parameters vs log CrCL was: CL/BSA 0.06 (90% CI -0.15–0.28); AUC/dose -0.07 (-0.30–0.16); Cbz FU 0.02 (-0.05–0.08). Cbz safety was consistent with previous reports. Conclusions: RI had no clinically meaningful effect on Cbz PK. Non-significant trends of increasing AUC and decreasing Cbz CL with greater RI were seen. PK parameter estimates for pts with moderate and severe renal impairment and controls GM estimate (90% CI) Control CrCL 90 ml/min/1.73 m2 Moderate renal impairment CrCL 40 ml/min/1.73 m2 Severe renal impairment CrCL 15 ml/min/1.73 m2 CL/BSA, l/h/m2 29.81 (24.18–36.75) 28.34 (24.44–32.86) 26.66 (20.15–35.27) AUC/dose, ng*h/ml/mg/m2 33.23 (26.65–41.44) 35.21 (30.24–40.99) 37.75 (28.29–50.39) Cbz FU, % 5.51 (5.08–5.96) 5.44 (5.13–5.76) 5.36 (4.95–5.80) Disclosure: J. Machiels: has been a member of advisory boards for Boehringer Ingelheim and Novartis and his institution has received research funding from Sanofi; S. Rottey: has attended advisory boards and received research funding from Sanofi; R. Baird: has received research funding from Sanofi; P. Clot and C. Wack: is an employee of Sanofi; L. Shen and D. Bobilev: is an employee and stock holder of Sanofi. All other authors have declared no conflicts of interest.

554PB CELLS (CD20+) ASSOCIATED TO TUMOR INFILTRATING CYTOTOXIC T-CELLS OBSERVED ON RESECTED LIVER COLORECTAL METASTASES (LCM) ARE PROGNOSTIC

ABSTRACT Aim: Colorectal cancer infiltrating cytotoxic T-cells (CD8+ cells) are a strong prognostic factor for survival after primary tumor and metastases resection. The impact of B cells for prognosis is less characterized. Methods: Metastatic colorectal patients (pts) engaged for curative liver surgery with available FFPE blocks for all resected LCM, were included. The density of CD8+ and CD20+ cells in the metastasis (CT) and the invasive margin (IM) for all LCM was determined by whole-slide immunohistochemistry and quantified with dedicated image analysis software. The mean value of the 3 most infiltrated areas (0.8 mm2) was calculated. The densities of CD8 and CD20 (CT and IM regions) were classified into Hi or Lo according to cut-off values (minimal p-value approach). The total number of Hi densities was calculated to determine the immunoscore (IS) 0-2 Hi (low IS) or 3-4 Hi (high IS). For pts with multiple LCM, all LCM were quantified. The mean value of all densities, the least and the most infiltrated LCM/pt were analyzed. Cumulative DFS/OS analyses were performed using the Kaplan-Meier estimator. The hazard ratio (HR) for OS/DFS comparing (IS0-2 vs 3-4) was determined using univariate Cox regression and the significance by log-rank tests. Results: 294 LCM from 88 patients (M/F 1.1, mean 3.3/pt, synchr/metachr 5.8) were included. For the least infiltrated metastasis: a high IS is prognostic for DFS and OS. Independently of CD8 cells, a high CD20 density associated concomitantly with both regions (CT/IM) is prognostic for OS (HR: 0.36; p = 0.00004) but not for DFS (HR 0.58, p = 0.1). LCM/patient IS Survival HR Logrank (IS0-2 vs 3-4; 95%IC) p-value Mean of all CD8-CD20 DFS 0.44 (0.22-0.91) 0.5 OS 0.29 (0.09-1.00) 0.03 Less of all CD8-CD20 DFS 0.31 (0.18-0.53) 0.0005 OS 0.25 (0.12-0.51) 0.00003 Most of all CD8-CD20 DFS 0.92 (0.53-1.56) 0.7 OS 0.95 (0.42-2.12) 0.9 Conclusions: B cells (CD20+) associated with cytotoxic T cells (CD8+) in both tumor regions (CT/IM) of the least infiltrated LCM/pt are highly prognostic after curative resection. These results confirm the important role of tumor regions and of B-cells (CD20+) for the tumor immune response. Disclosure: All authors have declared no conflicts of interest.

Abstract P3-04-04: Identification of a “BRCAness” signature in triple negative breast cancer by Comparative Genomic Hybridization

Background: Triple Negative Breast Cancers (TNBC) represent 12% to 20% of total Breast Cancers (BC) and have a worse outcome compared with other breast cancer subtypes. TNBC often show a deficiency in DNA double strand break repair mechanisms. This deficiency is generally related to inactivation of a repair enzymatic complex involving BRCA1 caused either by genetic mutations, epigenetic modifications, or by post-transcriptional regulations. The identification of BC presenting a BRCA1 deficiency could be useful to select patients that could benefit from PARP inhibitors, alkylant agents or platinum-based chemotherapy. In this study, we have identified by Comparative Genomic Hybridization array (CGH-array) a recurrent gain in 17q25.3 characteristic of BRCA1 mutated or methylated TNBC. Methods: 130 formalin-fixed paraffin embedded (FFPE) tumours including TNBC (unknown-BRCA1-status, BRCA1 mutated and non-mutated), Luminal A, Luminal B, Her2-neu amplified (Her2+) BC and BRCA1-mutated non-TNBC (mutated-Luminal A, Luminal B, Her2+) were obtained from our local tumour collection. DNA was extracted and genomic copy-number alterations were analysed by CGH-array (Agilent, SuperPrint G3 Human CGH 8×60K Oligo Microarrays). FISH analyses were performed on section from FFPE samples to validate some chromosomal aberrations belonging to the 17q25.3 amplified region evidenced by CGH-array. The study of BRCA1 promoter methylation status in all tumours was carried out by MDxHealth (Liege, Belgium). Results: In this study, we have identified by CGH-array a genomic region (17q25.3) amplified in 90% of the BRCA1 mutated tumours (29/32). This chromosomal gain was studied in other subtypes of BC by CGH-array and it was only evidenced in 30% (6/20) of BRCA1 non-mutated TNBC, 26.67% (4/15) of unknown-BRCA1-status TNBC, 13.64% (3/22) of Luminal B, 19.05% (4/21) of Her2+ and 0% (0/20) of Luminal A breast cancers. FISH assays confirmed these chromosomal amplifications and evidenced like CGH array analyses a significant difference between BRCA1 mutated and non-mutated BC for the 17q25.3 gain. BRCA1 methylation was found only in TNBC (11/58) and was not found in the BRCA1-mutated BC cohort, nor in the Luninal A, Luminal B and Her2+ samples. In BRCA1 non-mutated TNBC, the methylation was found in 8 cases including 4 with 17q25.3 amplification. In the unknown-BRCA1-status TNBC, 3 methylated samples were found with 1 case with co-amplification of the 17q25.3 region. Recurrence of 17q25.3 amplification in BRCA1 mutated tumours as well as its detection in BC having a methylation in BRCA1 promoter suggests that the 17q25.3 gain could be a marker of the BRCA1 deficiency. Identification of relevant genes whose expression is up-regulated within the recurrently gained region is underway. Conclusions: The CGH signature observed in 17q25.3 chromosomal region and FISH assay developed in this study could allow the identification of “BRCAness” breast tumours, improving the diagnostic performance and orienting the selection of the appropriate therapy. The up-regulated genes themselves might also represent potential therapeutic targets. Acknowledgements: This work was financed through the “Plan National Cancer-Action 29” (Belgium) and supported by the UNICANCER-PACS08 trial (France). Citation Information: Cancer Res 2012;72(24 Suppl):Abstract nr P3-04-04.

P5-18-04: Safety Profile of Ixabepilone as Adjuvant Treatment for Poor Prognosis Early Breast Cancer: First Results of the Unicancer-PACS 08 Trial.

Purpose: PACS 01 trial demonstrated that the sequential adjuvant chemotherapy with FEC100 followed by docetaxel (D) significantly improves disease-free and overall survival in node-positive(N+) early breast cancer (BC). However, Triple negative (TN) and ER+/ PR-/HER2− subgroups are significantly associated to a worse prognosis even after adjunction of D. As Ixabepilone (Ixa) has notable preclinical and clinical activity in these subgroups, the PACS 08 trial aims to compare standard FEC100-D regimen to 3 cycles of FEC100 followed by 3 cycles of Ixa. We report the preliminary results of the toxicity profile. Patients and methods: Patients (pts) had localized resectable unilateral ER-/PR-/HER2− or ER+/PR-/HER2− BC. Main inclusion criteria were: age Conclusion: Our results indicate that D arm is more often associated to significant haematological toxicities, whereas both neurotoxicities and haematological toxicities are reported in the Ixa arm. Although significantly more pts discontinued treatment due to adverse events in Ixa arm compared to D arm, Ixa may still represent a promising therapeutic option for pts in the adjuvant setting especially for poor prognosis BC. Citation Information: Cancer Res 2011;71(24 Suppl):Abstract nr P5-18-04.