In 2024 it is estimated that ovarian cancer will be the second deadliest gynecologic malignancy and the sixth most common cause of cancer related deaths among women [1]. Most ovarian cancer tumors respond to initial chemotherapy and newer maintenance therapy regimens, such as bevacizumab or PARP inhibitors as indicated. Despite therapeutic advancements, the majority of ovarian cancer tumors recur, ultimately leading to death in nearly 50% of patients within 5 years [1]. Thus, new therapeutic strategies are urgently needed for women diagnosed with ovarian cancer.

Ovarian cancer has immunogenic potential, as supported by its expression of a number of known tumor antigens (TAs) and the recognized association of tumor infiltrating lymphocyte (TIL) accumulation and improved patient outcomes [2,3]. However, ovarian cancer has generally been weakly responsive to endogenous immune modulation using immune checkpoint inhibitors (ICIs) [[4], [5], [6]]. This may be due in part to the proportion of ovarian cancers that are immunologically ‘cold’ as well as the relatively small population of tumor-resident tumor-reactive effector T cells essential for anti-tumor function and response to therapy [[7], [8], [9]]. Identifying these tumor-reactive T cells provides an avenue to study, develop, and optimize cellular therapy products with enhanced tumor-sensing capability and cytolytic activity, and the capacity for intratumoral penetrance.

ACT is a promising therapeutic approach that harnesses the intrinsic capabilities of the immune system to recognize and eliminate tumor cells. ACT involves harvesting immune cells from the patient or a donor to develop a living cancer immunotherapy. Immune cells can be activated and expanded then infused into the patient or the cell product can be modified through cell selection or genetic modification to enhance their tumor killing capabilities prior to infusion (Fig. 1). ACT modalities have shown clinical efficacy in leukemia, as evidenced by FDA approvals of CD19- and BCMA-specific chimeric antigen receptor (CAR) T cell therapies, and encouraging ACT clinical trial results have been reported in solid tumors including melanoma, lung cancer, renal cell carcinoma, and various gynecological malignancies, suggesting a path towards ovarian cancer treatment. Here, we review the current landscape of ACT strategies focused on bolstering T cell effector function and targeting ovarian cancer.

TILs are a heterogeneous population of immune cells comprised in part of T cells that migrate to the tumor and are essential for tumor immune surveillance and control. CD8+ TILs are a cytotoxic subset of T cells, some of which can recognize tumor-associated peptides presented on MHC class I molecules on cancer cells, allowing them to kill tumor cells. Importantly, CD8+ TILs are comprised of multiple T cell clones each with its own TCR allowing for diverse antigen recognition and the capacity to target tumors with heterogeneous antigen expression.

TILs, or certain TIL subsets are a predictive biomarker for outcomes in patients with epithelial ovarian cancer (EOC). Numerous studies found a positive correlation between intraepithelial CD3+ or CD8+ T cell infiltration and patient outcome, with CD8+ TILs typically having a stronger association [10]. In 2003, Zhang et al. first reported that the presence of intraepithelial CD3+ TILs correlated with delayed EOC recurrence and increased overall survival to tumors lacking infiltration, highlighting the importance of endogenous anti-tumor immunity [3]. A landmark study of 500 patients by Clarke et al. later reported that EOC had the most CD8+ TIL infiltration compared to other ovarian cancer subtypes and that intraepithelial CD3+ and CD8+ TILs correlated to survival outcomes in disease-specific analyses but were not predictive in endometrial or clear cell carcinomas [11]. More recently, a multi-center observational trial of over 5500 patients with ovarian cancer, confirmed that high grade serous ovarian cancer (HGSOC) has more CD8+ T cell infiltration than other ovarian cancer subtypes, and reported a dose-response relation to CD8+ TIL infiltration in these subjects with a median survival of 5.1 years in patients with high CD8+ TIL infiltrates, compared to 2.8, 3.0, 3.8 years with no, low, or moderate TIL levels, respectively [12]. Survival benefits were also observed among women with endometrioid and mucinous carcinomas but not in other subtypes. Of note, both studies also identified that CD8+ TILs correlated to survival in BRCA1 mutant EOC but not in BRCA2 mutant carriers [11,12]. While CD4+ helper T cells promote CD8+ T cell activity, most studies do not find a correlation with their infiltration and survival, while immunosuppressive regulatory T cells (Tregs) are generally a negative predictor of ovarian cancer survival [13,14].

Tumor-reactive TILs account for a small subset of TILs, while the remainder are considered bystander in nature. Upon stimulation, T cells upregulate activation markers including CD25, CD69, CD38, CD103, and CD137, and chronic antigen stimulation can lead to upregulation of inhibitory/exhaustion markers including PD-1, TIGIT, CTLA4, LAG3, and TIM3. Studies have shown added benefit to assessing infiltration of TILs that express these markers as a surrogate for immunoreactive TILs. Increased PD-1+ TIL populations correlate to EOC patient outcomes [15,16]. PDL1 upregulation on tumor-associated macrophages (TAMs) also correlated to increased survival and associated with cytolytic and activated and suppressed CD3 and CD8+ TILs [15,16]. Infiltration of CD103+ TILs is also a prognostic marker [17,18]. A combined immunoscore of CD103+ and CD3+ TILs better predicted patient outcomes than when either marker was assessed independently [18]. Most recently, CD137+ TIL accumulation was found to be associated with overall survival [19], suggesting that activated T cells suppress tumor progression.

In recent trials, TIL infiltration was also found to be an indicator of response to immunotherapies such as ICIs. CD8+ TILs were a positive predictor of response in the phase III JAVELIN Ovarian 200 trial evaluating response to avelumab (anti-PD-L1) and pegylated liposomal doxorubicin combination [20]. Elsewhere immune score, as a surrogate for IFN-primed exhausted CD8+ T cells, showed an association with response to pembrolizumab (anti-PD1) and niraparib (PARP inhibitor) combination therapy [21]. Similarly, markers of immune activation in the peripheral blood can also be prognostic [22,23]. At tumor sites, emerging evidence indicates that not only the overall tumor microenvironment (TME) composition, but also intercellular relationships, may function as indicators of biological interactions that may serve as predictors of response to immunotherapy in other cancers [24,25], providing rationale for exploration in ovarian cancer.

Adoptive TIL therapy is a promising alternative therapy for cancer that harnesses the antitumor properties of endogenous tumor-resident T cells. Autologous TILs are expanded ex vivo from patient tumors and then reinfused back into the patient. TILs are typically expanded from tumor fragments in the presence of high-dose IL-2 for 3–5 weeks followed by a rapid expansion of TILs in the presence of irradiated human peripheral blood mononuclear cells (PBMCs) until sufficient numbers are achieved for patient infusion, with total expansion time ranging from three weeks to three months. TIL engraftment in the patient relies in part on lymphodepleting chemotherapy prior to TIL infusion, and IL-2 administration is provided to support TIL proliferation and survival. The majority of toxicities seen in patients receiving TIL therapy are related to these supportive regimens rather than TILs, which have limited adverse effects.

Dr. Steven Rosenburg and colleagues pioneered this approach in the 1980s when they demonstrated that expanded immune cells from tumors had patient-specific activity and efficacy in advanced primary solid tumors and metastatic disease [26,27]. This work led to a series of clinical trials in metastatic melanoma that showed reproducible efficacy (ORRs, 34%–56%; PFS, 3.7–7.5 months; OS, 15.9–21.8 months), irrespective of studies being performed at different institutions with different manufacturing processes, regimens, and patient selection criteria [28]. In March 2024, lifileucel (Iovance Biotheraputics) became the first TIL therapy to be granted FDA approval. Lifileucel is approved for advanced melanoma previously treated with a PD-1 blocking antibody, and if BRAF V600 mutation positive, a BRAF inhibitor. TIL therapy has also shown promise in other solid tumors with high mutational burden [29], including non-small cell lung carcinoma (NSCLC) and received breakthrough therapy designation for cervical cancer treatment in 2019.

The number of TIL clinical trials for ovarian cancer have been more limited. In 1991, Aoki et al. published trial results of lymphodepletion followed by TIL therapy for ovarian cancer treatment. A 71% objective response rate was reported with one complete response and 4 partial responses (n = 7). In a second cohort, TIL plus a cisplatin regimen showed a 90% response rate (n = 10), with 4 patients having no recurrence for at least 15 months [30]. Follow up trials from Freedmen et al. in 1994 and Penderson et al. in 2018 showed no objective response in 6 and 8 patients, respectively [31,32]. Conflicting study results could be due to a number of factors. The former trial included TIL therapy as front-line therapy with only two enrolled patient tumors that were refractory to chemotherapy, while the latter two trials predominantly enrolled patients with chemotherapy refractory or resistant metastatic ovarian cancer. TIL products were also variable between trials. Aoki et al. reported predominantly CD8 + CD3+ TIL products, whereas Freedman et al. reported predominantly CD4+ CD8- TILs expanded from pleural effusion, ascites, or tumor digests and the TIL product infused in Penderson's study had a high number of TILs expressing immune inhibitory ligands including LAG3 and PD-1. At this time, the identification of 1) patients who are more likely to respond to TIL therapy and 2) expanded TIL products that will have the greatest efficacy in ovarian cancer are most critical to improving the effectiveness of this approach.

To enhance the efficacy of the TIL product, tumor-specific TIL subpopulations can be selected before the rapid expansion step in the manufacturing process. In early TIL studies for melanoma, individual TIL fragment cultures were selected for reactivity against HLA-matched or autologous tumor cell cultures by measuring IFNγ secretion. Since then, other selection strategies for enriching tumor-reactive T cells have been developed including methods for CD103+, CD39+, PD-1+, CD137+ TILs or neoantigen recognizing TILs, with PD-1 selected TILs under clinical evaluation for ovarian cancer treatment (NCT03449108). Enriching tumor-reactive TILs addresses antigen specificity, but expanded TILs may still possess properties of terminal differentiation, and poor persistence and proliferative capacity in vivo, limiting their effectiveness. In melanoma trials, TIL protocols have been adapted to expand bulk TILs with less differentiated features and sustained proliferative potential [33]. Still, the ovarian cancer TIL trials conducted to date have not deployed selection beyond PD-1 selection or adapted expansion methods strategies, which represents a missed opportunity.

The addition of antagonistic CD3 antibodies is common practice during TIL production, however, the addition of other antagonistic antibodies may provide additional co-stimulatory signals that enhance TIL fitness, proliferation, and efficacy. CD137 (4-1BB) is a T cell co-stimulatory receptor that promotes proliferation, survival, and drives T cell memory [34]. Addition of agonistic 4-1BB monoclonal antibody (i.e., Urelumab) to TIL culture can preferentially stimulate proliferation of activated CD8+ TILs that display HLA-restricted tumor recognition, improving the total yield [35]. A recent study reported outcomes from 17 patients infused with TILs expanded with CD3 and CD137 agonist [36]. At infusion, TILs had high expression of CD39 and CD69 with low expression of checkpoint markers PD1, CTLA4, LAG3, TIM3, and TIGIT indicating strongly reactive TIL product. 63% demonstrated stable disease with one pancreatic cancer patient who had tumor burden reduction across metastatic lesions lasting over 1 year. Of three treated patients with ovarian cancer, response included two stable disease and one progressive disease.

To increase their efficacy, next-generation TILs are being developed using synthetic biology. In ovarian cancer, two early phase clinical trials are underway to evaluate TILs engineered with a co-stimulatory antigen receptor [37] (NCT04389229) and TILs engineered to target the tumor antigen NY-ESO-1 (NCT02650986). Other strategies are being explored to improve TIL efficacy including engineering TILs to secrete supportive cytokines (e.g. IL-2) [38] or express chemokines receptors to promote tumor homing [39]. Alternatively, knocking down immune inhibitory receptors, e.g. PD-1 or more recently SOCS1, may improve TIL effector function and is being evaluated clinically [40] (NCT05361174, NCT06237881).

T cells recognize peptides of tumor associated or specific antigens (TAA or TSAs) presented by major histocompatibility complex (MHC) complex of antigen presenting cells (APCs), including dendritic cells (DCs), macrophages, and tumor cells. Thus, therapeutic cancer vaccines can stimulate the endogenous immune response and educate the adaptive immune cells in a target specific manner against cancer cells. Importantly, they also have the capacity to immunologically reverse “cold” tumors lacking TAA expression or low TIL infiltration to “hot” tumors that express T cell targets and subsequently improve TIL penetration [41].

Monocytes isolated from patient PBMCs can be differentiated and matured ex vivo into DCs, antigen loaded, and then administered back to the patient as a vaccine or can be utilized to prime or boost antigen-specific T cells from PBMCs or TILs for use in ACT trials. Most ovarian cancer vaccine development uses whole tumor cell lysate for priming against patient-specific TAAs or enriched for specific antigen targets, including NY-ESO-1, HER2, MUC1, WT1, CA125, mesothelin, and folate receptor alpha (FRα) [42], which have limited expression in healthy tissue. However, targeting a single antigen can hinder therapeutic potential as tumors are heterogeneous in antigen expression. One new approach for developing T cells targeting a collection of TAAs has recently been evaluated in a first in human trial of metastatic carcinomas [43]. Ex vivo primed T cells specific for up to four target antigens were infused and most patients had stable disease and all patients were alive for >100 days post-therapy. Persisting T cells were detected up to 79 weeks, and subjects with high persistence of T cells had significantly improved PFS compared to those with low persistence, albeit there was no objective response likely due to persistence of T cell clones with low TCR avidity.

Alternatively, antigen-loaded DCs can be infused for developing T cells for use in ACT. DC vaccines have been FDA-approved for prostate cancer and recently gained fast-track designation for the treatment in glioblastoma multiforme highlighting their safety and efficacy. In ovarian cancer, early pilot and phase I clinical trials showed DC cell vaccination is well tolerated with some early signs of response in some patients prompting larger trials [44]. Unfortunately, phase II studies reported limited efficacy and did not show a survival advantage of DC vaccine over standard of care treatment [45]. As such, DC vaccines are primarily in development in the adjuvant setting in combination with standard of care therapies. In the SOV02 randomized phase II trial, DC vaccine pulsed with human OV-90 and SKOV-3 cells in combination with chemotherapy significantly increased OS, but not PFS, compared to chemotherapy alone [46]. Due to promising trial results, the study team evaluated concomitant versus sequential chemotherapy then DC vaccine infusion [47]. Sequential dosing significantly delayed PFS, suggesting scheduling of DC vaccines may be critical to optimizing therapy response in ovarian cancer. Biomarker studies from peripheral blood identified tumor mutational burden (TMB) and CD8+ cells abundance as a positive predictor of response to chemotherapy and was negatively associated with patients who received a DC vaccine, suggesting that DC vaccines may have a benefit in immunologically cold tumors [48].

In another phase I trial for ovarian cancer, DC vaccine in combination with bevacizumab and cyclophosphamide showed expansion and reactivity of neoantigen-specific T cells and significantly increased OS compared to DC vaccine alone or in combination with bevacizumab [49]. Adding the nonsteroidal anti-inflammatory drug (NSAID) acetylsalicylic acid (ASA) and IL-2 showed a 100% survival at three years compared to 40% with the addition of only ASA [50]. However, it is difficult to determine if ASA + IL-2 is necessary for enhanced efficacy as the original trial data only included 2-year follow-up data. A third study from this group performed ACT therapy using vaccine-primed peripheral blood T cells following the DC vaccine-cyclophosphamide-bevacizumab regimen. Although response was limited, those who responded did show patient-specific, neoantigen-specific T cell responses [51].

Like classical therapeutic vaccines, cancer vaccines can also be administered in the form of DNA, RNA, or protein which are internalized, processed, and presented by APCs to induce cancer immunogenicity. Unlike ACT, these can be used as off-the-shelf therapies. Immunization with FRα peptides showed durable immunogenicity with all patients alive at the last follow up, at least 2 years post immunization [52], and high response rates were reported when patients were immunized against NY-ESO-1 [53]. Nucleic acid-based vaccines deliver transcribed/translated DNA or translated RNA that encodes antigenic protein that eventually leads to presentation of TAA fragments. Adding sequences encoding tumor antigens together with genetic information encoding proteins such as chemokines or T cell co-stimulatory molecules aiming to facilitate the immune response has shown clinical value [54]. Since the FDA-approval of COVID-19 mRNA vaccines, momentum has gained for mRNA vaccine research for solid tumors including ovarian cancer. For example, triplexed mRNA vaccine mRNA-2752 encoding OX40L, IL-23, and IL-36 is currently in phase II development in combination with immune checkpoint blockade (NCT03739931).

Other cancer vaccines can be administered in situ to induce immunogenic cell death (ICD) of tumor cells in vivo [55], enhancing the immune response from dead-cell-associated antigens derived from solid tumors. Oncolytic viruses, like Talimogene laherparepvec (T-Vec), are viruses modified to increase immunogenicity while reducing pathogenicity. Combination of oncolytic viruses and ICIs has been clinically shown to improve T cell infiltration and proliferation of CD8+ T cells and responses in metastatic melanoma [56], providing rationale for testing this approach in other solid tumors. ICD can also be induced by low dose chemo- and radiotherapy, making combination with ACT an attractive approach.

T cells recognize peptide-MHC in an HLA-dependent manner through interaction with their TCR comprised of heterodimeric alpha and beta chains specific for tumor antigen targets and the CD3 complex. To enhance TAA recognition, T cells can be transduced with genes that encode for shared TAA-specific TCRs or can be developed in a personalized manner where TCRs are sequenced from tumor-reactive TIL subsets and engineered into non-reactive T cells for recognition of TAAs or patient-specific neoantigens. A key benefit to TCR approaches is the ability to generate “off the shelf” tumor-reactive T cells for ACT and to harness natural signaling networks activated by TCR detection of extracellular and intracellular tumor antigens [41].

TCR-T cell therapy in ovarian cancer remains in the early stages of development. Most active trials are evaluating gene engineered TCR-T cells targeting melanoma-associated antigen (MAGE-A4) and New York esophageal-1 (NY-ESO-1) in basket trials for multiple target-expressing solid and or metastatic cancers, with only one trial specifically enrolling ovarian cancer (NCT03691376). Immantics US Inc., is clinically evaluating a TCR-T cell targeting PRAME (PReferentially expressed Antigen in MElanoma), and a second-generation therapy with co-transduction of CD8 to leverage activity of both CD8 and CD4 lymphocyte populations both as a monotherapy or in combination with nivolumab (NCT03686124). Most genetic modifications are performed through viral transduction, although the non-viral sleeping beauty transposon/transposase system to generate driver mutation-reactive TCR-T cells is feasible for clinical use [57].

TCR binding specificity and avidity to TSAs are essential for efficacy and safety. Ideally, TCRs should have high avidity for TSAs, with no cross-reactivity to other peptide-HLA complexes, or HLA-cross reactivity [58]. A MAGEA3/A12 TCR study led to unexpected neurotoxicity and death in two patients. Autopsy revealed previously unknown expression of MAGE-A12 in the brain which led to TCR-mediated inflammatory response in select patients [59]. Neoantigens arising from somatic mutations not observed in normal tissue have become desirable targets for cellular immunotherapy. Whole exome sequencing and enrichment of T cells expressing activation markers can help identify neoantigens and TCRs for specific targeting of neoantigens [60]. As a driver of cancer [61], mutated RAS antigens represent good candidates for TCR therapy, and mutation in RAS proteins has been found in ovarian cancers [62]. While promising, a clinical trial performing autologous neoantigen-specific TCR T cells (NCT03970382) has been suspended due to a business decision, highlighting the tremendous cost and labor required to identify patient-specific neoantigens.

Excitement over CAR-T cell therapy in ovarian cancer has been bolstered by its success in the treatment of hematological malignancies using CD19- and BMCA-specific CAR-T cells. The main components of synthetic CARs include the extracellular domain, hinge region, transmembrane domain and the signaling domain [63]. The extracellular domain is the single chain variable fragment region (scFv), that targets TAAs in an MHC-independent manner, overcoming downregulation of MHC molecules on tumor cells, which occurs in ovarian cancer [64]. The hinge region adjusts the steric distance between the CAR-T cells and target antigen, whilst the transmembrane domain functions to transduce extracellular antigen recognition signals to the intracellular signaling domain. Currently, there are four distinct generation CARs differentiated by their intracellular signaling domains. The first-generation CAR constructs, pioneered by Eshhar et al., were comprised of scFv coupled to intracellular CD3-z or Fc receptor gamma signaling domains that succumbed to exhaustion and inadequate costimulatory signals [65,66]. Second-generation CARs introduced 4-1BB or CD28 as a costimulatory domain [67]. Treatment with CD19-CAR-T cells with a 4-1BB costimulatory domain achieved complete remission in patients with B cell malignancies and CAR-T cell persistence >10 years post infusion [68]. Third-generation CARs introduced dual costimulatory domains, such as CD28/OX40 and CD28/4-1BB. However, third-generation CAR-T cells against hematologic malignancies did not provide better anti-tumor activity in patients with non-Hodgkin's lymphoma, B-cell lymphoma, or leukemia, but then showed better proliferation and persistence [69,70]. Fourth-generation CARs (or TRUCKs) and other possible next generation CARs are being developed which have a third costimulatory domain or a transgenic payload to deliver proinflammatory cytokines such as IL-12 or other factors to the TME, thereby initiating innate cell recruitment, re-educating the TME, broadening the immune response and addressing possible antigen escape.

The first phase I clinical study of CAR T cell therapy in ovarian cancer evaluated first-generation CAR-T cells targeted FRα [71]. Patients received a dose escalation of FRα-CAR-T cells with high-dose IL-2 or received dual-specific T cells (reactive with both FRα and allogeneic cells) followed by immunization using allogeneic PBMCs in cohort two. No clinical response was observed likely due to low CAR expression and poor persistence of infused T cells. There is currently an ongoing dose escalation trial evaluating FRα CAR with 4-1BB costimulatory domain in ovarian cancer, which showed promising preclinical efficacy [72]. HER-2, mesothelin, MUC 16 (CA125), and epithelial cell adhesion molecules (EpCAM) are other CAR targets that have been widely validated in vitro and are now under clinical investigation for ovarian cancer.

CAR-T cell therapy efficacy has generally been limited in solid tumors due to a myriad of factors, leaving room for CAR-T optimization. Unlike TCR therapy, CARs are restricted to surface antigen targets and require a higher tumor antigen density per cell and larger number of effector cells to initiate response. To increase the sensitivity of CARs towards antigens expressed at low levels, the scFV binding affinity can be increased, however this may increase off-target effects if the antigen is also expressed on healthy tissue suggesting tuning recognition may be needed on a per target basis [73].

Universal immune receptors or “adaptable CARs” can allow for CAR engineering to be performed in a uniform manner with loading of optimized, off-the-shelf antibodies for redirected CAR specificity [74]. Loading can be performed ex vivo and/or in vivo, against multiple antigens simultaneously or sequentially over time to overcome tumor heterogeneity and relapse due to antigen loss, as observed in CAR trials for hematological malignancies. CARs engineered to induce expression of secretion of accessory molecules for increased efficacy are also in development. For example, CARs that secrete soluble IL-12 have been shown to have enhanced anti-tumor efficacy in preclinical ovarian cancer models [75,76]. However, constitutive secretion of IL-12 can lead to severe toxicities, and novel inducible systems may be needed to transiently deliver cytokines in a localized safe manner [76]. Analogously, IL-18 induces cell-mediated immunity but IL-18 secreting CARs may enhance T cell activity and durable responses without causing fatal toxicities [77].

An orthogonal approach includes disruption of immune inhibitory ligands during CAR engineering. Surprisingly, PD-1 disruption in CAR T cells results in poor expansion and persistence in vivo, faster maturation and upregulation of exhaustion markers, all leading to functional impairment and reduced potency [78,79]. On the contrary, disruption of CTLA4 improves CAR T cell fitness and enhances tumor efficacy through CD28 signaling that is nullified in PD-1 deficient CAR T cells [80]. Alternatively, CAR-T fitness can be enhanced by engineering CAR constructs into stem-like CD8+ memory T cell progenitors which are more fit and amenable to chronic stimulation [81] or engineering T cells to express transcription factors that instill a stem-like program [76].

Of note, CAR-T cells can induce toxicities including cytokine release syndrome (CRS), hemophagocytic lymphohistiocytosis (HLA)/ macrophage activation syndrome (MAS), and immune effector cell-associated neurotoxicity syndrome (ICANS), and in the case of allogeneic transfer, graft-versus-host disease (GVHD) all which have been well documented in patients. Tocilizumab, an IL-6R blocking antibody, is the primary intervention for CRS. Armored CARs secreting tocilizumab is a promising approach to reduce toxicities [76]. However, tocilizumab is ineffective against ICANS as it does not cross the blood brain barrier. Incorporation of inducible safety switches or suicide genes may be essential for limiting these toxicities in the clinic [82].

The majority of ACT trials for ovarian cancer utilize T cells as the effector cell population for infusion. However, therapeutic strategies that exploit the cytolytic features of innate Natural Killer (NK) cells are also in clinical development. Initial clinical studies of ex vivo expanded autologous NK cells were safely tested in a variety of liquid and solid tumors, though limited anti-tumor effect observed due to self-recognition limiting NK cell activation. Later, reports showed that allogeneic and haploidentical NK cells could be safely administered and garnered promising responses in hematopoietic malignancies [83,84]. Since then, allogenic NK cell therapy has been widely investigated as a cancer therapy with varying success. In ovarian cancer, a number of early phase trials are ongoing, but study results are limited to one early phase study of 14 patients reporting 4 patients with partial response and eight patients with stable disease [85]. In a second study with two subjects, one had a partial response [86], and third case study reported clinical symptom improvement [87]. Further, ongoing APOLLO (NCT03213964) and INTRO (NCT03539406) are evaluating the safety and feasibility of intraperitoneal delivery of NK cells in platinum resistant or recurrent ovarian cancer.

Similar to T cells, NK cells can be genetically engineered with a CAR or TCR for antigen specific targeting. Most CAR-NK cells in clinical investigation are in early phase trials and target CD19 for hematopoietic malignancies or BCMA for multiple myeloma. CD19 CAR-NK cells show response rates of 72% and 73% in two trials with 11 and 8 subjects, respectively [88]. There are two active trials currently recruiting patients with ovarian cancer (NCT05776355 and NCT05410717).