Fresh Ex Vivo Expanded Natural Killer Cells Demonstrate Robust Proliferation in Vivo in High-Risk Relapsed Multiple Myeloma (MM) Patients

Abstract Abstract 579 Stimulation of NK cells with the HLA-deficient cell line K562 genetically modified to express 41BB-ligand and interleukin (IL) 15 allows for the production of large numbers of highly activated NK cells for therapy (Fujisaki et al, Cancer Res. 2009;69:4010–4017). We tested the safety, persistence and anti-MM activity of these ex-vivo expanded NK cells (ENK) in patients with high-risk MM relapsing after AUTO-PBSCT. Apheresis products were obtained from MM patients (AUTO-ENK) or from killer immunoglobulin-like receptor-ligand (KIR-L) mismatched haplo-identical family donors (ALLO-ENK). The preparative regimen comprised bortezomib only (Bor, days −9, −6, and −2) or Bor and immunosuppression with cyclophosphamide (Cy, day −7) and fludarabine (Flu, days −5 to −2), see Table 1. ENK cells were produced at the Center for Cell and Gene Therapy (CAGT) in Houston, shipped to UAMS either cryopreserved or fresh (see separate abstract), infused on day 0 and followed by daily administrations of 3×106 units of IL2 (13 doses total) to support ENK cell persistence in vivo. Flow cytometry was used to determine immunophenotype and the absolute number of NK and T cells in blood. Chimerism was assessed by PCR for donor specific short tandem repeats. Standard 4h chromium release assays were used to assess potency of ENK against MM cells. Of 6 patients treated, ENK cell doses of 2–8×107/kg (with T cell doses in ALLO-ENK products limited to <5×105/kg) were safely infused. IL2 therapy-induced rigors and grade II fever were the most common adverse events. No in vivo expansion was observed in the 3 patients receiving cryopreserved ENK products. One patient had stable disease for 3 months followed by slow disease progression. The other two patients experienced early progression, with one succumbing to MM. The 3 subsequent patients received fresh ENK products and transient expansion in vivo was observed in all 3, peaking at day 7 (absolute number of NK cells/μl blood: 512, 547, 153; NK cell percentage of lymphocytes: 49%, 92%, 64%). For the fresh ALLO-ENK cell recipient we confirmed that the in vivo expanded cells (Figure 1A) were derived from the donor (> 92–95% donor chimerism on days 5, 7 and 9). Furthermore, NK cells isolated from the blood on days 5 and 7 avidly killed recipient MM cells (Figure 1B). Unfortunately, this patient died of septic shock 11 days post-infusion and clinical response could not be evaluated. While non-expanded NK cells expressed low levels of CD26 and CD70, these molecules were highly expressed on ENK cells and NK cells in the PB post-ENK cell infusion, suggesting that CD26 and CD70 can be used to track the infused cells. AUTO-ENK cells failed to kill recipient MM in vitro but significant killing of third party MM cells (43%, 63%) and K562 (80%, 78%) was observed (effector: target ratio of 10:1). In keeping with this, the two patients who received fresh AUTO-ENK had progressive disease. Patients receiving Cy/Flu had increased serum levels of the NK cell growth factor IL15 (median 34 vs. 4pg/ml, respectively). High frequencies of CD25+/Foxp3+ T regulatory cells (24%, 44%, 47% of the CD3+/4+ lymphocytes) were observed 14 days after ENK infusion in 3/4 patients studied and we are investigating whether these cells were involved in ENK cell suppression. In conclusion, this study has established the safety and feasibility of infusing up to 8×107/kg ENK cells. We only observed dramatic in vivo expansion of fresh ENK suggesting that infusing fresh rather than freeze-thawed ENK cells is critical. ALLO-ENK cells isolated from the PB post-infusion killed recipient MM cells. In contrast, this was not observed with AUTO-ENK cells suggesting that further optimization of the AUTO-ENK cell approach may be required to overcome KIR-L induced inhibition. In addition, studies are under way to determine whether suppression by T regulatory cells, soluble factors, or suboptimal IL15 surge may prove important in modulating ENK activity and in vivo persistence. Table 1. ENK product type and preparative regimen Bortezomib (Bor), Cyclophosphamide (Cy), Fludarabine (Flu). ENK Patient Donor Preparation Freeze-Thawed ENK 1 AUTO Bor/Cy/Flu 2 ALLO Bor/Cy/Flu 3 AUTO Bor/Cy/Flu Fresh ENK 4 AUTO Bor 5 ALLO Bor/Cy/Flu 6 AUTO Bor Disclosures: No relevant conflicts of interest to declare.