CIRCULATING ENDOTHELIAL CELLS (CEC), CIRCULATING ENDOTHELIAL PROGENITORS (CEP) AND CIRCULATING TUMOUR CELLS (CTC) AS MARKERS FOR RESPONSE TO BEVACIZUMAB/CARBOPLATIN/PACLITAXEL (B plus CP) OR BEVACIZUMAB/ERLOTINIB (B plus E) IN ADVANCED NON-SQUAMOUS NON-SMALL-CELL LUNG CANCER (NS-NSCLC)

ABSTRACT Background The phase II, open-label BRAIN study (ML21823; NCT00800202) is the first to assess the efficacy/safety of bevacizumab combined with chemotherapy in patients (pts) with ns-NSCLC and untreated CNS metastases. Investigation of CEC, CEP or CTC levels as potential markers of response to bevacizumab was undertaken. Methods In the B + CP arm, blood samples were taken before treatment on day 1 (d1) of cycle 1 (C1) (and 6 hrs after [C1 + 6h] treatment at one centre only), and on d1 of C2 and C3, i.e. 4 samples in total. Samples were taken for the B + E arm on d1 of C1, C2 and C3 (3 samples total). CEC and CTC levels were measured with CellSearch (Immunicon, Veridex). CEP were measured using four-colour flow cytometry after enrichment of progenitor cells using Miltenyi Biotec EPC enumeration and enrichment kit. Cut-offs were 15 cells/4mL for CEC, 2 cells/7.5mL for CTC and median for CEP. Relation between marker levels and 6-month progression-free rate (PFR) and best overall response (BOR) was examined. Results CEC and CTC were analysed in 69/91 (76%) pts, and CEP in 32/91 (35%) pts; baseline (BL) characteristics were similar to the overall population. CEC, CTC and CEP were detected at BL in 94%, 94% and 100% of these pts. In the B + CP group, CEC levels were increased at C1 + 6h, possibly reflecting the antivascular effect of chemotherapy. Furthermore, in the B + CP group, pts with higher BL CTC and no peak in CEC levels at C1 + 6h progressed at 6 months. Analysis of association of markers with BOR and PFR for the two regimens is ongoing. Conclusions Measuring CEC, CTC and CEP is feasible in NSCLC. Detailed analyses will be presented. The clinical value of these markers for predicting response to bevacizumab in advanced ns-NSCLC patients warrants further investigation. B + CP B + E Progression/death at 6 months Progression/death at 6 months CEC (cells/4mL) Median (range) No (n = 28) Yes (n = 19) No (n = 13) Yes (n = 8) C1 cells/4mL (n = 27) 24 (4–464) (n = 18) 32 (0–184) (n = 12) 28 (8–108) (n = 8) 44 (4–96) C1 + 6h cells/4mL change from BL (n = 9) 104 (12–632) 36 (-336–72) (n = 7) 28 (0–120) -16 (-172–100) – – C2 cells/4mL change from BL (n = 24) 24 (8–308) (n = 23) 0 (-376–268) (n = 15) 20 (0–216) (n = 14) -4 (-176–196) (n = 12) 34 (4–180) (n = 11) -4 (-72–160) (n = 8) 34 (0–356) (n = 8) -4 (-76–316) C3 cells/4mL change from BL (n = 26) 34 (0–228) (n = 25) 4 (-440–208) (n = 10) 30 (0–464) (n = 10) 10 (-172–444) (n = 11) 28 (0–76) (n = 10) 0 (-100–52) (n = 7) 76 (0–244) (n = 7) 28 (-20–192) CTC (cells/7.5mL) n (%) No (n = 28) Yes (n = 19) No (n = 13) Yes (n = 8) C1 ≥ 2 cells/7.5mL (n = 28) 24 (85.7%) 4 (14.3%) (n = 17) 8 (47.1%) 9 (52.9%) (n = 12) 10 (83.3%) 2 (16.7%) (n = 8) 7 (87.5%) 1 (12.5%) C2 ≥ 2 cells/7.5mL (n = 24) 21 (87.5%) 3 (12.5%) (n = 16) 10 (62.5%) 6 (37.5%) (n = 11) 9 (81.8%) 2 (18.2%) (n = 8) 5 (62.5%) 3 (37.5%) C3 ≥ 2 cells/7.5mL (n = 25) 22 (88.0%) 3 (12.0%) (n = 11) 7 (63.6%) 4 (36.4%) (n = 12) 12 (100%) 0 (0%) (n = 6) 4 (66.7%) 2 (33.3%) CEP (133+ CEP/1000 CD34+) Median (range) No (n = 12) Yes (n = 9) No (n = 7) Yes (n = 4) C1 (n = 12) 1.8 (0.0–4.3) (n = 9) 1.6 (0.0–5.9) (n = 7) 0.0 (0.0–2.1) (n = 4) 5.0 (0.2–7.1) C1 + 6h (n = 8) 1.5 (0.0–7.8) (n = 6) 0.0 (0.0–0.5) – – C2 (n = 12) 0.3 (0.0–5.8) (n = 7) 0.0 (0.0–3.3) (n = 6) 0.4 (0.0–3.2) (n = 4) 0.3 (0.0–1.1) C3 (n = 12) 1.2 (0.0–8.7) (n = 4) 0.0 (0.0–1.5) (n = 7) 0.7 (0.0–3.7) (n = 3) 1.8 (0.0–3.4) Disclosure H. Senellart: Financial Interest: Advisory Board. E. Dansin: Attended advisory boards for Roche, Lilly and Boehringer-Ingelheim. B. Besse: Received research grants from Roche. All other authors have declared no conflicts of interest.