Early Successes and Remaining Challenges of Harnessing Natural Killer Cells for Cancer Therapy

Veronika Bachanova, MD, PhD

Natural killer (NK) cells can offer several advantages over T cells for CAR therapy in that the former uses both a CAR dependent and independent mechanism for tumor eradication, has better safety, and off-the-shelf feasibility—all at a potentially lower cost, according to Veronika Bachanova, MD, PhD. Early data on the approach have shown promise, but questions remain.

CAR T-cell therapy has produced impressive responses in patients with relapsed/refractory B-cell malignancies, but challenges with this modality remain significant. The toxicities associated with the approach, such as cytokine release syndrome (CRS) and neurotoxicity, “are often severe,” noted Bachanova, who added that each product is custom made and requires a complex infrastructure for administration. As such, patients with rapidly progressing disease may not be able to wait for the product to be manufactured and delivered.

“Harnessing NK cells for cancer therapy is appealing because these cells are naturally cytotoxic. They kill without prior recognition of virally infected or malignantly transformed cells,” explained Bachanova. “They predominantly reside in lymphoid organs and they circulate in the blood in relatively low frequencies of 5% to 10%. Killing is governed by the concept of ‘missing self.’”

In her talk during the 25th Annual International Congress on Hematologic Malignancies, Bachanova, a professor of medicine of the Division of Hematology, Oncology, and Transplantation; lead of the Lymphoma Interdisciplinary Team; clinical director of Cell Therapies; and section head of Malignant Hematology at the University of Minnesota Medical School, walked through the latest developments made with CAR NK-cell therapy in cancer care and pressing areas of ongoing exploration.¹

NK Cells Vs T Cells: Weighing the Differences

First off, T cells have a single antigen-specific T-cell receptor, while NK cells have multiple antigens; notably, antigen priming is required for the former, while it is not for the latter. NK cells have several mechanisms of activation, said Bachanova. Additionally, while T cells differentiate in the thymus, NK cells do so in the bone marrow.

“Allogeneic T cells induce graft-versus-host [GVHD] disease and they product inflammatory cytokines such as interleukin (IL)-1, IL-2, IL-6, and others, which are, in fact, the reason for the CRS,” said Bachanova. “These cytokines are not produced by NK cells. NK cells produce interferon (INF)-gamma and tumor necrosis factor–alpha; these cells are also not known to cause GVHD.”

Understanding NK Cells

NK cells naturally kill the target through the use of 3 main mechanisms, according to Bachanova. The cells are able to mediate antibody-dependent cellular cytotoxicity by engaging the Fc region of the antibody; they eliminate targets through the use of several inhibitory and activating signals; and cytokine stimulation can activate NK cells and trigger their full killing capacity and function.

The activation of these cells is controlled by integrating signals from both activating and inhibitory NK cell receptors. “NK CAR not only use the CAR for antigen recognition, but they also use naturally existing NK cell function to promote an anticancer effect,” explained Bachanova. “The allogeneic NK cells mediate antitumor response without inducing GVHD.”

There are several main sources of NK CAR cells. Autologous NK CAR cells have been used, but with limited success due to manufacturing challenges, according to Bachanova. As such, there has been a shift toward allogeneic products that are derived from the peripheral blood, cord blood, induced pluripotent stem cells (iPSCs), and NK cell lines like NK-92.

“The main advantage of allogeneic NK CAR is the timeline with an iPSC-derived product, which are immediately available, without a need for apheresis and 4 weeks waiting for the manufacturing,” said Bachanova. “This is an off-the-shelf product that can potentially be available for multiple infusions.”

Several CAR constructs are being evaluated in CAR NK cells in several clinical trials.² NK CAR products are targeting similar antigens to CAR T-cell therapies, predominantly CD19. Other targets include EGFR, HER2, CS1, CD5, CD123, and mesothelin, among others, are also being developed for NK CAR products. The vectors utilized for these products include retrovirus, lentivirus, and transponson transfection.

Transformative Research Shakes Up the Space

A phase 1/2 trial (NCT033056339) conducted by investigators at The University of Texas MD Anderson Cancer Center evaluated the use of human leukocyte antigen-mismatched, anti-CD19 CAR NK cells derived from cord blood in 11 patients with relapsed/refractory CD19-positive cancers, including non-Hodgkin lymphoma and chronic lymphocytic leukemia (CLL).³

Notably, the NK cells were transduced with a retroviral vector that expressed genes that encoded anti-CD19 CAR, IL-15, and inducible caspase 9 as a safety switch. The cells were given to patients through a single infusion at 1 of 3 doses following lymphodepleting chemotherapy.

Results, which were published in the New England Journal of Medicine, showed that 73% of patients (n = 8) achieved a response with the approach; 7 of these patients, 4 of whom had lymphoma and 3 had CLL, experienced a complete remission, and 1 patient experienced remission of the Richter’s transformation component, although they still had persistent CLL. Notably, responses were observed within 30 days after the time of infusion at all dose levels examined.

“[The NK cells were detectable in] the peripheral blood up to 1 year, and they are most likely expanding within the 14 days; this expansion appears to be dose dependent,” noted Bachanova. “The study all suggested that the patients with a higher copy vector at the peak had better responses than those who did not.”

Off-the-Shelf Products Under Exploration

In another study, investigators utilized induced pluripotent stem cells (iPSCs) to produce NK CAR cells that would specifically target cancer cells in an antigen-specific manner in an ovarian cancer xenograft model.⁴ Here, the cells were found to significantly inhibit tumor growth and to prolong survival. The cells were found to induce in vivo activity that proved comparable to that of CAR T cells. Notably, however, less toxicity was observed with the NK cells.

“iPSCs are differentiated into the NK cells and subsequently transduced with a CAR of interest and subsequent gene modification leads to iPSC-derived NK CAR which have now been evaluated in a number of phase 1 clinical studies,” noted Bachanova.

FT596 is another investigational, off-the-shelf, multi-antigen targeting, CAR NK cell therapy that has been derived from a human clonal master iPSC line, that has emerged in the landscape.⁵ The product was engineered with 3 antitumor modalities: a CD19-targeting CAR; a high-affinity, non-cleavable CD16 Fc receptor that allows for tumor targeting and stronger antibody-dependent cell cytotoxicity in combination with a monoclonal antibody; and cytokine-autonomous persistence that promotes IL-15/IL-15 receptor fusion.

“[The product’s] potent cytokine complex is secreted at the site of engaging with the tumor and this promotes survival and proliferation of the NK cells and the CD8 T cells,” explained Bachanova.

Advantages of iPSC-Derived NK CAR Approaches

The fact that iPSCs can be multiplexed engineered with a single iPSC selected to generate a clonal master engineered iPSC bank (MCB) offers a distinct advantage over other approaches under exploration, according to Bachanova. Moreover, the MCB can be “renewably used” as a starting cell source to encourage the mass production of uniformly engineered immune effector cells in a cost-effective way.

Additionally, a single manufacturing run of these products can produce several doses of these immune effector cells; the cells that are not immediately being utilized can then be cryopreserved, stored, and distributed for off-the-shelf availability, added Bachanova.

“Patients can be administered multiple doses in the outpatient setting, including in combination with other anticancer therapies, to promote deep, durable responses,” said Bachanova.

Dual-Targeting of NK CAR

In preclinical models, dual-antigen targeting approaches are being explored, specifically with CD19 and CD20. This is being done by engaging the Fc region of the monoclonal antibody, which can then be co-infused with the NK CAR, explained Bachanova.

“In an in vivo model⁶ [investigators explored] NK CAR, [specifically FT596], with or without rituximab [Rituxan] and rituximab coadministration with the NK CAR resulted in an improved tumor killing and significantly better animal survival,” said Bachanova. “This suggests that a combinatory approach is more effective.”

In the FT596-101 phase 1 trial (NCT04245722), investigators are evaluating several doses of the NK CAR product across several cohorts of patients with relapsed/refractory B-cell lymphoma.⁷ To be eligible for inclusion, patients must have histologically documented disease expected to express CD19 and CD20.

The trial will utilize a 3+3 dose-escalation design followed by disease- and regimen-specific expansion cohorts. Those who receive regimen A will receive FT596 as a monotherapy, while those given regimen B1 will receive FT596 plus rituximab, and those who receive regimen B2 will be given FT596 in combination with obinutuzumab (Gazyva).

The primary goal of the trial is to determine the recommended phase 2 dose of the product, while secondary objectives include gaining a better understanding on the safety, tolerability, preliminary activity, and pharmacokinetics of the agent.

“Patients are given low-dose chemotherapy followed by the FT596 infusion. Patients who experience a clinical benefit can receive a second infusion,” noted Bachanova. “The first patient who had received monotherapy experienced a partial response [PR] and was able to get a second infusion with another PR and a 67% reduction in tumor size” Notably, no dose-limiting toxicities, CRS, neurotoxicity, or GVHD of any grade were reported.⁷

Ongoing Efforts Gain Ground

In the phase 1 CAR2BRAIN trial (EudraCT 2016-00025-39), investigators are evaluating the use of the ErbB2-specific cell clone NK-92/5.28z as a potential treatment for patients with recurrent HER2-positive glioblastoma.⁸ The product carries a codon-optimized CAR based on ErbB2-specific antibody FRP5 and CD28 and CD33-zeta signaling domains.

In the trial, patients are injected with NK-92/5.28z cells. For the dose-escalation portion of the trial, the maximum-tolerated (MTD) dose of the product will be established, with a maximum dose of up to 1 x 10⁸ cells planned. After the MTD is identified, investigators seek to understand safety of prolonged treatment with the product in an expansion cohort.

Investigators will also examine the pharmacokinetics and pharmacodynamics of the product, immune responses to NK-92 and/or CAR, as well as signs of efficacy in the form of objective response rate, progression-free survival, and overall survival.

“Also, CRISPR technology opens up a number of new avenues for the NK CAR field [in that it] can [help to] further genetically manipulate these cells by either adding the genes for the checkpoint inhibition or by allowing the cell destruction, if needed,” said Bachanova.

Remaining Challenges to Address

The optimal CAR design for optimal NK cell activation and cytotoxicity remains unknown, according to Bachanova. Efforts will need to be focused on identifying what the best costimulatory molecule will be.

“The main issue with NK CAR therapy is the persistence. How do we improve the expansion and how do we overcome the tumor suppression and escape?” questioned Bachanova. “Another very interesting and important question has to do with NK CAR combinatory approaches. Which strategies can improve the clinical impact of these efforts?”

Another challenge is to improve the efficiency of cryopreservation of these approaches and the recovery of off-the-shelf NK CAR T-cell products. “This is a work in progress,” concluded Bachanova.

References

1. Bachanova V. The potential of CAR NK cells in oncology. Presented at: 25th Annual International Congress on Hematologic Malignancies; February 25-28, 2021; virtual. Accessed February 25, 2021.
2. Xie G, Dong H, Liang Y, et al. CAR-NK cells: a promising cellular immunotherapy for cancer. EBioMedicine. 2020;59:102975. doi:10.1016/j.ebiom.2020.102975
3. Liu E, Marin D, Banerjee P, et al. Use of CAR-transduced natural killer cells in CD19-positive lymphoid tumors. N Engl J Med. 2020;382(6):545-553. doi:10.1056/NEJMoa1910607
4. Li Y, Hermanson DL, Moriarty BS, et al. Human iPSC-derived natural killer cells engineered with chimeric antigen receptors enhance anti-tumor activity. Cell Stem Cell. 2018;23(2):181-192.e5. doi:10.1016/j.stem.2018.06.002
5. Bachanova V, Cayci Z, Lewis D, et al. Initial clinical activity of FT596, a first-in-class, multi-antigen targeted, off-the-shelf, iPSC-derived CD19 CAR NK cell therapy in relapsed/refractory B-cell lymphoma. Blood. 2020;136(suppl 1):8. doi:10.1182/blood-2020-141606
6. Goodridge JP, Jacobs B, Saetersmoen ML, et al. Remodeling of secretory lysosomes during education tunes functional potential in NK cells. Nat Commun. 2019;10(1):514. doi:10.1038/s41467-019-08384-x
7. FT596 as a monotherapy and in combination with anti-CD20 monoclonal antibodies. ClinicalTrials.gov. Updated June 17, 2020. Accessed February 26, 2021. https://clinicaltrials.gov/ct2/show/NCT04245722
8. Burger MC, Mildenberger IC, Cieplik HC, et al. The CAR2BRAIN study: a monocentric phase I trial with ErbB2-specific NK-92/5.28.z cells in recurrent glioblastoma. Neuro Oncol. 2017;19(suppl 3):iii51-iii52. doi:10.1093/neuonc/nox036.186