Phase Ib/II trial of expanded and activated autologous natural killer (NK) cells with trastuzumab (Tras) in refractory HER2+ metastatic breast cancer (MBC).
3045 Background: A mechanism of Tras is antibody-dependent cell cytotoxicity (ADCC) exerted by NK cells. ADCC may be impaired in cancer patients. We hypothesize that this can be restored by NK activation, and developed a method to generate large numbers of activated NK cells from MBC patients (pts) and tested the effects of NK infusion with Tras. Methods: Peripheral blood harvested from 11 refractory HER2+ MBC pts aged 32-67 (median 59), was co-cultured for 10 days with irradiated K562-mb15-41BBL cells. Expanded NK cells were infused 24 hours (H) post-Tras, with interleukin-2, 1MU/m2 thrice weekly for 6 doses starting 24H pre-infusion. Tras was continued 3 weekly up to 8 cycles (C); pts with sustained stable disease (SD) at C4 or C6 received a 2nd and 3rdinfusion at C6 and C8 respectively. Blood was taken serially for NK cell counts and function. Results: 16 NK cell products were prepared for the 11 pts; 3 pts had a 2nd, and 2 a 3rd infusion. Median NK cell expansion was 337-fold (range 91-478). Products contained 81±16% NK cells, virtually all CD16+ and highly cytotoxic: ADCC on HER2+ SKBR3 cells at 4H with Tras was 85.8±10.1% at 2:1 effector:target ratio (E:T). Pts had 2-13 (median 6) prior lines of palliative therapy. 3 of 16 infusions were at 1st dose level (106 NK cells/kg), and 13 at target dose (107/kg). NK infusion was well tolerated; worst toxicity was transient G2 thrombocytopenia (n = 1). There was no objective response but 7/10 evaluable pts, including 6 who had 107/kg at C1, had SD. Median progression free survival was 1.8 vs 8.5 months for pts who received 106/kg vs 107/kg cells at C1. NK cell counts increased from day (D) 1 (0.16±0.11x103/µL) to D8 (0.32±0.19x103/µL) or D22 post-infusion (0.36±0.20x103/µL) (p < 0.01), although levels were only slightly higher than baseline (0.31±0.22x103/µL). NK cells were 13±8% of PBMCs on D1 and 17±9% on D8; ADCC on SKBR3 cells with Tras without NK enrichment was 11±4% on D1 and 16±5% on D8 (2:1 E:T). Conclusions: This first-in-human study shows that large numbers of highly potent NK cells can be reliably generated from MBC pts and safely infused with Tras. The disease control seen in heavily pre-treated pts is encouraging and warrants further studies. Clinical trial information: NCT02030561.