一、Challenges of Using Blood NK Cells as Cellular Effectors in Cancer Treatment（使用血液NK细胞作为癌症治疗的效应细胞的挑战）
Over the last few decades, there has been significant advancement of our knowledge concerning the biology of NK cells, which comprise about 10% of the circulating blood lymphocyte pool. A more comprehensive characterization, including gene and surface molecule expression, as well as proteomics, recently has become available [2 4]. It is now well accepted that NK cells are an important component of the rapid cellular immune response that can inhibit the cell-to-cell spread of malignant or infected cells. Not only can they execute immediate (“spontaneous”) killing upon contact with those cells (without the need to be “primed” like T-cells), but they also produce a host of cytokines that support the expansion and function of other immune cells in the blood circulation and the tumor microenvironment.
在过去的几十年里，我们对NK细胞生物学的认识取得了显著进展，NK细胞约占循环血液淋巴细胞池的10%。最近，包括基因和表面分子表达以及蛋白质组学在内的更全面的表征已经可用。现在普遍认为，NK细胞是快速细胞免疫反应的重要组成部分，能够抑制恶性或感染细胞的细胞间传播。它们不仅可以在与这些细胞接触时立即（“自发地”）进行杀伤（无需像T细胞那样“启动”），还可以产生一系列细胞因子，支持血液循环和肿瘤微环境中其他免疫细胞的扩增和功能。
Through their FcgIII (CD16) receptor, NK cells are major effector cells of ADCC, which is the main effector arm for immunoglobulin G (IgG) 1 and IgG3 monoclonal antibodies (mAbs). Notably, NK cells that express the high-affinity (158V) variant of the CD16 receptor are more effective for mAb-mediated targeted cell killing, although only approximately 10% of the human population is homozygous for this variant [5].
通过其FcγIII（CD16）受体，NK细胞是ADCC的主要效应细胞，这是免疫球蛋白G（IgG）1和IgG3单克隆抗体（mAbs）的主要效应臂。值得注意的是，表达高亲和力（158V）CD16受体变异体的NK细胞在mAb介导的靶向细胞杀伤中更为有效，尽管人类人群中只有大约10%的人是这种变异体的纯合子。
The effector molecules that execute rapid NK cell mediated target cell killing are perforin and granzymes and to a lesser extent FasL and tumor necrosis factor related apoptosis inducing ligand (TRAIL) [6]. The former are released on contact with transformed cells and act within minutes, with perforin “punching holes” into the membrane of target cells so that granzymes can enter the cell and destroy it by degrading their genetic material. FasL and tumor necrosis factor related apoptosis inducing ligand seem to have a more delayed cytotoxic effect through activation of the death receptor pathways on target cells [6].
执行快速NK细胞介导的靶细胞杀伤的效应分子是穿孔素和颗粒酶，以及在较小程度上的FasL和肿瘤坏死因子相关凋亡诱导配体（TRAIL）。前者在与目标细胞接触时释放，并在几分钟内起作用，穿孔素在靶细胞膜上“打孔”，使颗粒酶能够进入细胞并通过降解其遗传物质来破坏细胞。FasL和肿瘤坏死因子相关凋亡诱导配体似乎通过激活靶细胞上的死亡受体途径产生更延迟的细胞毒性效应。
Enriching NK cells from blood (peripheral or cord) either for research or clinical applications usually is done by negative selection employing a number of monoclonal antibodies that can deplete and eliminate non-NK cells (reviewed in Kundo et al. [7]). NK cells also may be enriched from blood, based on their characteristic CD56 (neuronal cell adhesion molecule) surface expression. One of the concerns is, however, that the bound antibody may affect the NK cell function and/or can interfere with any further manipulation (engineering) of the cells. Another limitation of such enrichment is the unpredictable variability of yield, particularly for NK cells from patients with cancer whose NK cells often are also dysfunctional due to circulating suppressive serum factors or following chemotherapy. A further challenge is that the NK-cell donors must be connected to a leukapheresis device for several hours to collect their white blood cells, of which only about 10% are NK cells. To achieve sufficient NK cells for therapeutic infusion, NK cells need to be expanded on a cell line feeder layer, which may be genetically altered. In addition, cytokines need to be included to support expansion (reviewed in Saito et al. [8]).
从血液（外周或脐带）中富集NK细胞用于研究或临床应用，通常通过负选择进行，使用多种单克隆抗体耗尽和消除非NK细胞。NK细胞也可以根据其特征性的CD56（神经细胞黏附分子）表面表达从血液中富集。然而，其中一个问题是，结合的抗体可能会影响NK细胞的功能和/或干扰细胞的任何进一步操作（工程改造）。这种富集的另一个限制是产量的不可预测性，特别是对于癌症患者的NK细胞，因为循环抑制血清因子或化疗后，其NK细胞通常也会功能失调。进一步的挑战是，NK细胞供体必须连接到白细胞分离装置数小时，以收集其白细胞，其中只有大约10%是NK细胞。为了获得足够数量的NK细胞用于治疗输注，需要在细胞系饲养层上扩增NK细胞，该饲养层可能经过基因改造。此外，还需要加入细胞因子以支持扩增。
As an alternative to using a patient’s own (autologous) NK cells, allogeneic NK cells can be collected from the blood of a healthy related or unrelated person. Those allogeneic NK-cell collections need to be further treated to remove T lymphocytes, as those can cause the potentially life-threatening graft-versus-host reactions after infusion. This additional purification step adds to the cost of the cell preparation and also results in a further loss of NK cells. Regardless of the source, blood NK cells need to be expanded to have suffi cient numbers of cells available for therapeutic infusion [9]. Another challenge with blood-derived NK cells is the fact that those NK cells are more difficult to manipulate or engineer, requiring more complex, often virus-based gene transduction methods, which come with unpredictable and inconsistent efficacy [8].
作为使用患者自身（自体）NK细胞的替代方法，可以从健康相关或无关人员的血液中收集异体NK细胞。这些异体NK细胞收集后需要进一步处理以去除T淋巴细胞，因为这些细胞在输注后可能会导致潜在致命的移植物抗宿主反应。这个额外的纯化步骤增加了细胞制备的成本，并且也导致NK细胞的进一步损失。无论来源如何，血液NK细胞都需要扩增，以获得足够数量的细胞用于治疗输注。血液来源NK细胞的另一个挑战是，这些NK细胞更难以操作或工程改造，需要更复杂、通常基于病毒的基因转导方法，这些方法的效果难以预测且不一致。
Considering those drawbacks of blood-derived NK cells, including the costs of obtaining a consistent NK-cell population from blood for research and clinical applications, a continuously growing NK cell line is an appealing alternative. Although several human NK cell lines have been established [10], only the NK-92 cell line has shown consistently high cytotoxicity against a broad spectrum of cancer cells, and can easily be expanded ex vivo with a short doubling time of 24 36 h [11,12]. Importantly, NK-92 can be genetically engineered by plasmid electroporation to express a high-affinity Fc-receptor, CAR, or molecules that can modulate the tumor microenvironment [13,14].
鉴于血液来源NK细胞的这些缺点，包括从血液中获得一致的NK细胞群用于研究和临床应用的成本，持续生长的NK细胞系是一个有吸引力的替代选择。尽管已经建立了几种人类NK细胞系，但只有NK-92细胞系显示出对广泛癌症细胞的一致高细胞毒性，并且可以在体外轻松扩增，其倍增时间为24-36小时。重要的是，NK-92可以通过质粒电穿孔进行基因工程改造，以表达高亲和力Fc受体、CAR或可以调节肿瘤微环境的分子。
二、Early Clinical Studies with NK-92 （NK-92的早期临床研究）
In 1997, the license and rights for NK-92 were transferred from the University of British Columbia to Rush Medical Center in Chicago, where in 2001 a phase I trial with NK-92 in patients with advanced cancer was conducted [31]. Due to the responses seen with LAK cell infusions in patients with renal cell cancer and melanoma [32], those two diseases were selected as initial indications for NK-92 infusions.
1997年，NK-92的许可和权利从不列颠哥伦比亚大学转移到芝加哥的拉什医学中心，2001年在那里进行了NK-92在晚期癌症患者中的I期试验。由于在肾细胞癌和黑色素瘤患者中使用LAK细胞输注看到了反应，这两种疾病被选为NK-92输注的最初适应症。
In that study, NK-92 cells were infused on days 1, 3 and 5 with the rationale that it usually takes a few days to trigger a T-cell mediated allogeneic immune response that could potentially lead to rejection of the infused NK-92 cells. In this phase I dose-escalation study, 12 patients were enrolled and received a starting dose of 1 x109 NK-92 cells/m2 . The greatest dose administered was 5 x 109 NK-92 cells/m2 cells. None of the patients experienced grade 3 or 4 side effects, and only a few patients developed a mild fever or a rash. Although phase I studies are not designed to assess efficacy, it is noteworthy that the majority of the patients with renal cell cancer experienced a prolonged disease-free survival and overall survival compared with historical controls. The single patient with melanoma in that study had a partial response with significant tumor reduction after the third infusion. In contrast to CAR/T-cell therapy that generally includes some form of immunosuppressive chemotherapy before infusion, NK-92 cells in those and all other studies (including with CAR modified NK-92) were given without any preparative chemotherapy.
在该研究中，NK-92细胞在第1、3和5天进行输注，理由是通常需要几天时间才能触发T细胞介导的异体免疫反应，这可能会导致输注的NK-92细胞被排斥。在这项I期剂量递增研究中，共纳入了12名患者，起始剂量为每平方米1×10^9个NK-92细胞。给予的最大剂量为每平方米5×10^9个NK-92细胞。没有患者出现3级或4级副作用，只有少数患者出现轻度发热或皮疹。尽管I期研究并非旨在评估疗效，但值得注意的是，大多数肾细胞癌患者的无病生存期和总生存期比历史对照组有所延长。该研究中唯一的黑色素瘤患者在第三次输注后出现了部分反应，肿瘤显著缩小。与通常在输注前进行某种形式免疫抑制化疗的CAR/T细胞疗法不同，NK-92细胞在这些以及所有其他研究（包括CAR修饰的NK-92）中都是在没有任何准备性化疗的情况下给予的。
Contemporaneous with the aforementioned study, Dr. TorstenTonn, who had previously performed post-doctoral work in the author’s laboratory, initiated a phase I trial in Frankfurt, Germany, in pediatric and adult patients with advanced, mostly solid cancers [26]. The schedule consisted of two infusions, given on day 1 and 3. The dose of 1 x 1010 cells/m2 was considered dose-limiting, but not because of side effects, but rather because of logistical challenges of expanding larger numbers of NK-92 cells at that time.
与此同时，托尔斯滕·托恩博士（Torsten Tonn）在德国法兰克福启动了一项I期试验，研究对象为患有晚期癌症（主要是实体瘤）的儿童和成人患者。该试验的安排是在第1天和第3天进行两次输注。每平方米1×10^10个细胞的剂量被认为是剂量限制性的，但这并非由于副作用，而是因为当时扩大NK-92细胞数量存在物流挑战。
Another phase I trial was conducted under the guidance of Dr. Armand Keating, at Ontario Cancer Center in Toronto [15]. Among the 12 patients with advanced hematologic malignancies enrolled in the study, two patients (with myeloma and Hodgkin disease) experienced long-lasting clinical remissions. Patients received multiple infusions over time (planned six monthly cycles). Only one-half of the patients developed HLA antibodies, and none had a positive mixed lymphocyte culture when NK-92 cells were used as stimulators.
在多伦多安大略癌症中心，阿曼德·基廷博士（Armand Keating）指导进行了另一项I期试验。在该研究中，共有12名患有晚期血液系统恶性肿瘤的患者接受治疗，其中两名患者（分别患有骨髓瘤和霍奇金病）经历了长期的临床缓解。患者接受了多次输注（计划进行六次每月）。只有一半的患者产生了HLA抗体，没有任何患者的混合淋巴细胞培养物呈阳性反应。
In another phase I trial conducted at the University of Pittsburgh, seven patients with treatment-resistant acute myeloid leukemia were enrolled and received NK-92 [33]. Although none of thepatients experienced serious adverse effects, no significant clinical responses were noted, prompting the group to investigate possible reasons for the lack of response. In a subsequent publication, they reported that acute myeloid leukemia cell derived exosomes collected pre-therapy from all seven patients had the ability to inhibit the cytotoxic anti-leukemia effects of NK-92 in co-incubation assays [34].
在匹兹堡大学进行的另一项I期试验中，共有7名治疗抵抗性的急性髓系白血病患者接受NK-92治疗。尽管没有患者出现严重的不良反应，但也没有观察到显著的临床反应，这促使该团队调查缺乏反应的可能原因。在随后的出版物中，他们报告称，所有7名患者在治疗前收集的细胞来源的外泌体在共培养试验中能够抑制NK-92对白血病的细胞毒性效应。
Pooled response data from all 4 phase I studies reveal that approximately 36% of the treated patients showed some tumor reduction, ranging from tumor shrinkage to partial remission. Most importantly, repeated infusions with NK-92 cells, even at high cell numbers, did not induce any side effects greater than grade 2. In all trials the NK-92 cells were expanded in flasks, bags or G-Rex bioreactors using X-Vivo 10 medium with 5% human serum. For currently ongoing trials with engineered NK-92, the same medium is used but production is scaled up by use of large bioreactors.
所有4项I期研究的汇总反应数据显示，大约36%的接受治疗的患者显示出某种程度的肿瘤缩小，范围从肿瘤缩小到部分缓解。最重要的是，即使在高细胞数量下，重复输注NK-92细胞也没有引起超过2级的副作用。在所有试验中，NK-92细胞都是在含有5%人血清的X-Vivo 10培养基中，使用培养瓶、袋子或G-Rex生物反应器进行扩增。对于目前正在开展的基于改造后的NK-92细胞的临床试验，使用了相同的培养基，但通过使用大型生物反应器扩大了生产规模。
三、The Next Generation of Engineered NK-92: haNK, taNK, t-haNK, qt-haNK （下一代工程化NK-92：haNK、taNK、t-haNK、qt-haNK）
(1) High-affinity fc-receptor expressing NK-92 (haNK) （高亲和力Fc受体表达NK-92（haNK）
Largely as the result of predictable rapid proliferation (doubling time of 24 36 h) and ease of expansion, several genetically engineered variants of NK-92 cells have been generated, including a line expressing a high-affinity Fc-receptor, haNK. NK cells are the main effector cells for mAbs of IgG1 or IgG3 type such as trastuzumab, rituximab or avelumab that engage the CD16 Fc-receptor on NK cells for ADCC. The Fc-receptor on NK cells can have low, intermediate or high affinity for IgG. However, only approximately 10% of the general population expresses the high-affinity Fc-receptor for mAbs on their NK cells, with the majority of people expressing a low- or intermediate-affinity Fc-receptor [5]. This also implies that the majority of the population lacks the most relevant effector mechanism for mAb mediated cytotoxicity. Moreover, the Fc-receptor on blood NK cells is sensitive to the enzymatic cleavage by the ADAM17 enzyme, which can result in reduced efficacy of mAbs [35].
由于NK-92可预测的快速增殖（倍增时间为24-36小时）和易于扩增，已经产生了几种基因工程改造的NK-92细胞变体，包括表达高亲和力Fc受体的haNK。NK细胞是IgG1或IgG3型单克隆抗体（如曲妥珠单抗、利妥昔单抗或阿维鲁单抗）的主要效应细胞，这些抗体通过与NK细胞上的CD16 Fc受体结合来介导ADCC。NK细胞上的Fc受体对IgG的亲和力可以是低、中或高。然而，在一般人群中，只有大约10%的人在其NK细胞上表达高亲和力Fc受体，大多数人表达低或中等亲和力的Fc受体。这也意味着大多数人缺乏mAb介导细胞毒性效应的最相关效应机制。此外，血液NK细胞上的Fc受体对ADAM17酶的酶切敏感，这可能导致mAb的疗效降低。
With these considerations in mind, haNK cells were generated from NK-92 that express a high-affinity Fc-receptor (CD16A, 158V) genetically linked to endoplasmatic reticulum IL-2 [35]. The endoplasmatic reticulum linked IL-2 guarantees that only low amounts of IL-2 are secreted by haNK cells, but the intra-cellular concentration is sufficient enough to maintain cell viability, expansion and cytotoxicity. Importantly, the Fc-receptor on haNK cells is resistant to ADAM17-mediated degradation [35]. Moreover, studies from the National Cancer Institute have shown that haNK cells do not lose cytotoxicity under low oxygen (hypoxic) conditions, which is in contrast to blood derived NK cells [18]. Hypoxia is one of the significant immunosuppressive factors in the tumor microenvironment.
鉴于这些考虑，从NK-92中产生了haNK细胞，它们表达高亲和力Fc受体（CD16A，158V），并且该受体与内质网IL-2基因连接。内质网连接的IL-2确保haNK细胞只分泌少量IL-2，但细胞内浓度足以维持细胞的活性、扩增和细胞毒性。重要的是，haNK细胞上的Fc受体对ADAM17介导的降解具有抵抗力。此外，美国国家癌症研究所的研究表明，haNK细胞在低氧（缺氧）条件下不会失去细胞毒性，这与血液来源的NK细胞形成对比。缺氧是肿瘤微环境中一个重要的免疫抑制因素。
In a phase II clinical trial (NCT03853317) at the University of Washington, haNK cells were given in combination with Avelumab (anti-programmed death-ligand 1 [PD-L1] mAb) to patients with refractory Merkel cell cancer [36]. The haNK infusions were well tolerated, and despite the very advanced disease state of these patients, objective responses were seen in two of seven patients, including reversal of programmed cell death protein 1 (PD-1) refractoriness in one patient.
在美国华盛顿大学进行的一项II期临床试验（NCT03853317）中，haNK细胞与阿维鲁单抗（抗程序性死亡配体1 [PD-L1]单克隆抗体）联合用于治疗耐药的默克尔细胞癌患者。haNK输注耐受性良好，尽管这些患者的疾病状态非常晚期，但在7名患者中有2名患者观察到客观反应，其中1名患者逆转了对PD-1的耐药性。
NK-92 cells have been further engineered to express a recombinant receptor containing the extracellular portion of the high-affinity Fc-receptor CD64 with the transmembrane and intracellular region of CD16A (referred to as CD64/16A). According to preliminary in vitro studies, ADCC is further improved over NK cells that express the CD16A variant [37].
NK-92细胞还被进一步工程化改造，以表达一个重组受体，该受体包含高亲和力Fc受体CD64的胞外部分以及CD16A的跨膜和细胞内区域（称为CD64/16A）。根据初步体外研究，与表达CD16A变体的NK细胞相比，ADCC得到了进一步改善。
The main mechanism of action for most clinically effective mAbs is through ADCC with direct cytotoxicity and complement-mediated cytotoxicity playing a less important role. For that reason, the high-affinity FcR expressing NK-92 cells continue to be used by numerous biotech companies and research laboratories for development and testing of mAbs (available from www.BrinkBiologics.com).
大多数临床上有效的单克隆抗体的主要作用机制是通过ADCC的直接细胞毒性，补体介导的细胞毒性作用较小。因此，高亲和力FcR表达的NK-92细胞继续被众多生物技术公司和研究实验室用于开发和测试单克隆抗体（可通过www.BrinkBiologics.com获得）。
(2) Targeted NK-92 (taNK) cells expressing CARs （表达CAR的靶向NK-92（taNK）细胞）
The features of NK-92 cells have stipulated the generation of CARexpressing variants. Table 2 lists the variants of NK-92 cells, including NK-92ci, NK-92mi, haNK, taNK, t-haNK and qt-haNK. Much of the research with CAR-engineered NK-92 cells has been summarized in review papers [38 40]. Investigators have used first and second generation CAR constructs generally delivered by lentiviral or retroviral constructs. Although several studies confirmed the efficacy of CAR-modified NK-92, ImmunityBio, Inc. (which now holds worldwide rights to NK-92 variants) developed CAR-modified NK-92 cells by use of plasmid-based CAR gene constructs and transfected by simple electroporation to generate the NK-92 variants currently used in clinical studies [38,41].
NK-92细胞的特性促使产生了表达CAR的变体。表2列出了NK-92细胞的变体，包括NK-92ci、NK-92mi、haNK、taNK、t-haNK和qt-haNK。关于CAR工程化NK-92细胞的许多研究已在综述论文中进行了总结。研究人员使用第一代和第二代CAR构建体，这些构建体通常通过慢病毒或逆转录病毒载体传递。尽管几项研究证实了CAR修饰的NK-92的疗效，但ImmunityBio公司（现持有NK-92变体的全球权利）使用基于质粒的CAR基因构建体，并通过简单的电穿孔转染来产生目前用于临床研究的NK-92变体。
Over the last few decades, there has been significant advancement of our knowledge concerning the biology of NK cells, which comprise about 10% of the circulating blood lymphocyte pool. A more comprehensive characterization, including gene and surface molecule expression, as well as proteomics, recently has become available [2 4]. It is now well accepted that NK cells are an important component of the rapid cellular immune response that can inhibit the cell-to-cell spread of malignant or infected cells. Not only can they execute immediate (“spontaneous”) killing upon contact with those cells (without the need to be “primed” like T-cells), but they also produce a host of cytokines that support the expansion and function of other immune cells in the blood circulation and the tumor microenvironment.
在过去的几十年里，我们对NK细胞生物学的认识取得了显著进展，NK细胞约占循环血液淋巴细胞池的10%。最近，包括基因和表面分子表达以及蛋白质组学在内的更全面的表征已经可用。现在普遍认为，NK细胞是快速细胞免疫反应的重要组成部分，能够抑制恶性或感染细胞的细胞间传播。它们不仅可以在与这些细胞接触时立即（“自发地”）进行杀伤（无需像T细胞那样“启动”），还可以产生一系列细胞因子，支持血液循环和肿瘤微环境中其他免疫细胞的扩增和功能。
Through their FcgIII (CD16) receptor, NK cells are major effector cells of ADCC, which is the main effector arm for immunoglobulin G (IgG) 1 and IgG3 monoclonal antibodies (mAbs). Notably, NK cells that express the high-affinity (158V) variant of the CD16 receptor are more effective for mAb-mediated targeted cell killing, although only approximately 10% of the human population is homozygous for this variant [5].
通过其FcγIII（CD16）受体，NK细胞是ADCC的主要效应细胞，这是免疫球蛋白G（IgG）1和IgG3单克隆抗体（mAbs）的主要效应臂。值得注意的是，表达高亲和力（158V）CD16受体变异体的NK细胞在mAb介导的靶向细胞杀伤中更为有效，尽管人类人群中只有大约10%的人是这种变异体的纯合子。
The effector molecules that execute rapid NK cell mediated target cell killing are perforin and granzymes and to a lesser extent FasL and tumor necrosis factor related apoptosis inducing ligand (TRAIL) [6]. The former are released on contact with transformed cells and act within minutes, with perforin “punching holes” into the membrane of target cells so that granzymes can enter the cell and destroy it by degrading their genetic material. FasL and tumor necrosis factor related apoptosis inducing ligand seem to have a more delayed cytotoxic effect through activation of the death receptor pathways on target cells [6].
执行快速NK细胞介导的靶细胞杀伤的效应分子是穿孔素和颗粒酶，以及在较小程度上的FasL和肿瘤坏死因子相关凋亡诱导配体（TRAIL）。前者在与目标细胞接触时释放，并在几分钟内起作用，穿孔素在靶细胞膜上“打孔”，使颗粒酶能够进入细胞并通过降解其遗传物质来破坏细胞。FasL和肿瘤坏死因子相关凋亡诱导配体似乎通过激活靶细胞上的死亡受体途径产生更延迟的细胞毒性效应。
Enriching NK cells from blood (peripheral or cord) either for research or clinical applications usually is done by negative selection employing a number of monoclonal antibodies that can deplete and eliminate non-NK cells (reviewed in Kundo et al. [7]). NK cells also may be enriched from blood, based on their characteristic CD56 (neuronal cell adhesion molecule) surface expression. One of the concerns is, however, that the bound antibody may affect the NK cell function and/or can interfere with any further manipulation (engineering) of the cells. Another limitation of such enrichment is the unpredictable variability of yield, particularly for NK cells from patients with cancer whose NK cells often are also dysfunctional due to circulating suppressive serum factors or following chemotherapy. A further challenge is that the NK-cell donors must be connected to a leukapheresis device for several hours to collect their white blood cells, of which only about 10% are NK cells. To achieve sufficient NK cells for therapeutic infusion, NK cells need to be expanded on a cell line feeder layer, which may be genetically altered. In addition, cytokines need to be included to support expansion (reviewed in Saito et al. [8]).
从血液（外周或脐带）中富集NK细胞用于研究或临床应用，通常通过负选择进行，使用多种单克隆抗体耗尽和消除非NK细胞。NK细胞也可以根据其特征性的CD56（神经细胞黏附分子）表面表达从血液中富集。然而，其中一个问题是，结合的抗体可能会影响NK细胞的功能和/或干扰细胞的任何进一步操作（工程改造）。这种富集的另一个限制是产量的不可预测性，特别是对于癌症患者的NK细胞，因为循环抑制血清因子或化疗后，其NK细胞通常也会功能失调。进一步的挑战是，NK细胞供体必须连接到白细胞分离装置数小时，以收集其白细胞，其中只有大约10%是NK细胞。为了获得足够数量的NK细胞用于治疗输注，需要在细胞系饲养层上扩增NK细胞，该饲养层可能经过基因改造。此外，还需要加入细胞因子以支持扩增。
As an alternative to using a patient’s own (autologous) NK cells, allogeneic NK cells can be collected from the blood of a healthy related or unrelated person. Those allogeneic NK-cell collections need to be further treated to remove T lymphocytes, as those can cause the potentially life-threatening graft-versus-host reactions after infusion. This additional purification step adds to the cost of the cell preparation and also results in a further loss of NK cells. Regardless of the source, blood NK cells need to be expanded to have suffi cient numbers of cells available for therapeutic infusion [9]. Another challenge with blood-derived NK cells is the fact that those NK cells are more difficult to manipulate or engineer, requiring more complex, often virus-based gene transduction methods, which come with unpredictable and inconsistent efficacy [8].
作为使用患者自身（自体）NK细胞的替代方法，可以从健康相关或无关人员的血液中收集异体NK细胞。这些异体NK细胞收集后需要进一步处理以去除T淋巴细胞，因为这些细胞在输注后可能会导致潜在致命的移植物抗宿主反应。这个额外的纯化步骤增加了细胞制备的成本，并且也导致NK细胞的进一步损失。无论来源如何，血液NK细胞都需要扩增，以获得足够数量的细胞用于治疗输注。血液来源NK细胞的另一个挑战是，这些NK细胞更难以操作或工程改造，需要更复杂、通常基于病毒的基因转导方法，这些方法的效果难以预测且不一致。
Considering those drawbacks of blood-derived NK cells, including the costs of obtaining a consistent NK-cell population from blood for research and clinical applications, a continuously growing NK cell line is an appealing alternative. Although several human NK cell lines have been established [10], only the NK-92 cell line has shown consistently high cytotoxicity against a broad spectrum of cancer cells, and can easily be expanded ex vivo with a short doubling time of 24 36 h [11,12]. Importantly, NK-92 can be genetically engineered by plasmid electroporation to express a high-affinity Fc-receptor, CAR, or molecules that can modulate the tumor microenvironment [13,14].
鉴于血液来源NK细胞的这些缺点，包括从血液中获得一致的NK细胞群用于研究和临床应用的成本，持续生长的NK细胞系是一个有吸引力的替代选择。尽管已经建立了几种人类NK细胞系，但只有NK-92细胞系显示出对广泛癌症细胞的一致高细胞毒性，并且可以在体外轻松扩增，其倍增时间为24-36小时。重要的是，NK-92可以通过质粒电穿孔进行基因工程改造，以表达高亲和力Fc受体、CAR或可以调节肿瘤微环境的分子。
二、Early Clinical Studies with NK-92 （NK-92的早期临床研究）
In 1997, the license and rights for NK-92 were transferred from the University of British Columbia to Rush Medical Center in Chicago, where in 2001 a phase I trial with NK-92 in patients with advanced cancer was conducted [31]. Due to the responses seen with LAK cell infusions in patients with renal cell cancer and melanoma [32], those two diseases were selected as initial indications for NK-92 infusions.
1997年，NK-92的许可和权利从不列颠哥伦比亚大学转移到芝加哥的拉什医学中心，2001年在那里进行了NK-92在晚期癌症患者中的I期试验。由于在肾细胞癌和黑色素瘤患者中使用LAK细胞输注看到了反应，这两种疾病被选为NK-92输注的最初适应症。
In that study, NK-92 cells were infused on days 1, 3 and 5 with the rationale that it usually takes a few days to trigger a T-cell mediated allogeneic immune response that could potentially lead to rejection of the infused NK-92 cells. In this phase I dose-escalation study, 12 patients were enrolled and received a starting dose of 1 x109 NK-92 cells/m2 . The greatest dose administered was 5 x 109 NK-92 cells/m2 cells. None of the patients experienced grade 3 or 4 side effects, and only a few patients developed a mild fever or a rash. Although phase I studies are not designed to assess efficacy, it is noteworthy that the majority of the patients with renal cell cancer experienced a prolonged disease-free survival and overall survival compared with historical controls. The single patient with melanoma in that study had a partial response with significant tumor reduction after the third infusion. In contrast to CAR/T-cell therapy that generally includes some form of immunosuppressive chemotherapy before infusion, NK-92 cells in those and all other studies (including with CAR modified NK-92) were given without any preparative chemotherapy.
在该研究中，NK-92细胞在第1、3和5天进行输注，理由是通常需要几天时间才能触发T细胞介导的异体免疫反应，这可能会导致输注的NK-92细胞被排斥。在这项I期剂量递增研究中，共纳入了12名患者，起始剂量为每平方米1×10^9个NK-92细胞。给予的最大剂量为每平方米5×10^9个NK-92细胞。没有患者出现3级或4级副作用，只有少数患者出现轻度发热或皮疹。尽管I期研究并非旨在评估疗效，但值得注意的是，大多数肾细胞癌患者的无病生存期和总生存期比历史对照组有所延长。该研究中唯一的黑色素瘤患者在第三次输注后出现了部分反应，肿瘤显著缩小。与通常在输注前进行某种形式免疫抑制化疗的CAR/T细胞疗法不同，NK-92细胞在这些以及所有其他研究（包括CAR修饰的NK-92）中都是在没有任何准备性化疗的情况下给予的。
Contemporaneous with the aforementioned study, Dr. TorstenTonn, who had previously performed post-doctoral work in the author’s laboratory, initiated a phase I trial in Frankfurt, Germany, in pediatric and adult patients with advanced, mostly solid cancers [26]. The schedule consisted of two infusions, given on day 1 and 3. The dose of 1 x 1010 cells/m2 was considered dose-limiting, but not because of side effects, but rather because of logistical challenges of expanding larger numbers of NK-92 cells at that time.
与此同时，托尔斯滕·托恩博士（Torsten Tonn）在德国法兰克福启动了一项I期试验，研究对象为患有晚期癌症（主要是实体瘤）的儿童和成人患者。该试验的安排是在第1天和第3天进行两次输注。每平方米1×10^10个细胞的剂量被认为是剂量限制性的，但这并非由于副作用，而是因为当时扩大NK-92细胞数量存在物流挑战。
Another phase I trial was conducted under the guidance of Dr. Armand Keating, at Ontario Cancer Center in Toronto [15]. Among the 12 patients with advanced hematologic malignancies enrolled in the study, two patients (with myeloma and Hodgkin disease) experienced long-lasting clinical remissions. Patients received multiple infusions over time (planned six monthly cycles). Only one-half of the patients developed HLA antibodies, and none had a positive mixed lymphocyte culture when NK-92 cells were used as stimulators.
在多伦多安大略癌症中心，阿曼德·基廷博士（Armand Keating）指导进行了另一项I期试验。在该研究中，共有12名患有晚期血液系统恶性肿瘤的患者接受治疗，其中两名患者（分别患有骨髓瘤和霍奇金病）经历了长期的临床缓解。患者接受了多次输注（计划进行六次每月）。只有一半的患者产生了HLA抗体，没有任何患者的混合淋巴细胞培养物呈阳性反应。
In another phase I trial conducted at the University of Pittsburgh, seven patients with treatment-resistant acute myeloid leukemia were enrolled and received NK-92 [33]. Although none of thepatients experienced serious adverse effects, no significant clinical responses were noted, prompting the group to investigate possible reasons for the lack of response. In a subsequent publication, they reported that acute myeloid leukemia cell derived exosomes collected pre-therapy from all seven patients had the ability to inhibit the cytotoxic anti-leukemia effects of NK-92 in co-incubation assays [34].
在匹兹堡大学进行的另一项I期试验中，共有7名治疗抵抗性的急性髓系白血病患者接受NK-92治疗。尽管没有患者出现严重的不良反应，但也没有观察到显著的临床反应，这促使该团队调查缺乏反应的可能原因。在随后的出版物中，他们报告称，所有7名患者在治疗前收集的细胞来源的外泌体在共培养试验中能够抑制NK-92对白血病的细胞毒性效应。
Pooled response data from all 4 phase I studies reveal that approximately 36% of the treated patients showed some tumor reduction, ranging from tumor shrinkage to partial remission. Most importantly, repeated infusions with NK-92 cells, even at high cell numbers, did not induce any side effects greater than grade 2. In all trials the NK-92 cells were expanded in flasks, bags or G-Rex bioreactors using X-Vivo 10 medium with 5% human serum. For currently ongoing trials with engineered NK-92, the same medium is used but production is scaled up by use of large bioreactors.
所有4项I期研究的汇总反应数据显示，大约36%的接受治疗的患者显示出某种程度的肿瘤缩小，范围从肿瘤缩小到部分缓解。最重要的是，即使在高细胞数量下，重复输注NK-92细胞也没有引起超过2级的副作用。在所有试验中，NK-92细胞都是在含有5%人血清的X-Vivo 10培养基中，使用培养瓶、袋子或G-Rex生物反应器进行扩增。对于目前正在开展的基于改造后的NK-92细胞的临床试验，使用了相同的培养基，但通过使用大型生物反应器扩大了生产规模。
三、The Next Generation of Engineered NK-92: haNK, taNK, t-haNK, qt-haNK （下一代工程化NK-92：haNK、taNK、t-haNK、qt-haNK）
(1) High-affinity fc-receptor expressing NK-92 (haNK) （高亲和力Fc受体表达NK-92（haNK）
Largely as the result of predictable rapid proliferation (doubling time of 24 36 h) and ease of expansion, several genetically engineered variants of NK-92 cells have been generated, including a line expressing a high-affinity Fc-receptor, haNK. NK cells are the main effector cells for mAbs of IgG1 or IgG3 type such as trastuzumab, rituximab or avelumab that engage the CD16 Fc-receptor on NK cells for ADCC. The Fc-receptor on NK cells can have low, intermediate or high affinity for IgG. However, only approximately 10% of the general population expresses the high-affinity Fc-receptor for mAbs on their NK cells, with the majority of people expressing a low- or intermediate-affinity Fc-receptor [5]. This also implies that the majority of the population lacks the most relevant effector mechanism for mAb mediated cytotoxicity. Moreover, the Fc-receptor on blood NK cells is sensitive to the enzymatic cleavage by the ADAM17 enzyme, which can result in reduced efficacy of mAbs [35].
由于NK-92可预测的快速增殖（倍增时间为24-36小时）和易于扩增，已经产生了几种基因工程改造的NK-92细胞变体，包括表达高亲和力Fc受体的haNK。NK细胞是IgG1或IgG3型单克隆抗体（如曲妥珠单抗、利妥昔单抗或阿维鲁单抗）的主要效应细胞，这些抗体通过与NK细胞上的CD16 Fc受体结合来介导ADCC。NK细胞上的Fc受体对IgG的亲和力可以是低、中或高。然而，在一般人群中，只有大约10%的人在其NK细胞上表达高亲和力Fc受体，大多数人表达低或中等亲和力的Fc受体。这也意味着大多数人缺乏mAb介导细胞毒性效应的最相关效应机制。此外，血液NK细胞上的Fc受体对ADAM17酶的酶切敏感，这可能导致mAb的疗效降低。
With these considerations in mind, haNK cells were generated from NK-92 that express a high-affinity Fc-receptor (CD16A, 158V) genetically linked to endoplasmatic reticulum IL-2 [35]. The endoplasmatic reticulum linked IL-2 guarantees that only low amounts of IL-2 are secreted by haNK cells, but the intra-cellular concentration is sufficient enough to maintain cell viability, expansion and cytotoxicity. Importantly, the Fc-receptor on haNK cells is resistant to ADAM17-mediated degradation [35]. Moreover, studies from the National Cancer Institute have shown that haNK cells do not lose cytotoxicity under low oxygen (hypoxic) conditions, which is in contrast to blood derived NK cells [18]. Hypoxia is one of the significant immunosuppressive factors in the tumor microenvironment.
鉴于这些考虑，从NK-92中产生了haNK细胞，它们表达高亲和力Fc受体（CD16A，158V），并且该受体与内质网IL-2基因连接。内质网连接的IL-2确保haNK细胞只分泌少量IL-2，但细胞内浓度足以维持细胞的活性、扩增和细胞毒性。重要的是，haNK细胞上的Fc受体对ADAM17介导的降解具有抵抗力。此外，美国国家癌症研究所的研究表明，haNK细胞在低氧（缺氧）条件下不会失去细胞毒性，这与血液来源的NK细胞形成对比。缺氧是肿瘤微环境中一个重要的免疫抑制因素。
In a phase II clinical trial (NCT03853317) at the University of Washington, haNK cells were given in combination with Avelumab (anti-programmed death-ligand 1 [PD-L1] mAb) to patients with refractory Merkel cell cancer [36]. The haNK infusions were well tolerated, and despite the very advanced disease state of these patients, objective responses were seen in two of seven patients, including reversal of programmed cell death protein 1 (PD-1) refractoriness in one patient.
在美国华盛顿大学进行的一项II期临床试验（NCT03853317）中，haNK细胞与阿维鲁单抗（抗程序性死亡配体1 [PD-L1]单克隆抗体）联合用于治疗耐药的默克尔细胞癌患者。haNK输注耐受性良好，尽管这些患者的疾病状态非常晚期，但在7名患者中有2名患者观察到客观反应，其中1名患者逆转了对PD-1的耐药性。
NK-92 cells have been further engineered to express a recombinant receptor containing the extracellular portion of the high-affinity Fc-receptor CD64 with the transmembrane and intracellular region of CD16A (referred to as CD64/16A). According to preliminary in vitro studies, ADCC is further improved over NK cells that express the CD16A variant [37].
NK-92细胞还被进一步工程化改造，以表达一个重组受体，该受体包含高亲和力Fc受体CD64的胞外部分以及CD16A的跨膜和细胞内区域（称为CD64/16A）。根据初步体外研究，与表达CD16A变体的NK细胞相比，ADCC得到了进一步改善。
The main mechanism of action for most clinically effective mAbs is through ADCC with direct cytotoxicity and complement-mediated cytotoxicity playing a less important role. For that reason, the high-affinity FcR expressing NK-92 cells continue to be used by numerous biotech companies and research laboratories for development and testing of mAbs (available from www.BrinkBiologics.com).
大多数临床上有效的单克隆抗体的主要作用机制是通过ADCC的直接细胞毒性，补体介导的细胞毒性作用较小。因此，高亲和力FcR表达的NK-92细胞继续被众多生物技术公司和研究实验室用于开发和测试单克隆抗体（可通过www.BrinkBiologics.com获得）。
(2) Targeted NK-92 (taNK) cells expressing CARs （表达CAR的靶向NK-92（taNK）细胞）
The features of NK-92 cells have stipulated the generation of CARexpressing variants. Table 2 lists the variants of NK-92 cells, including NK-92ci, NK-92mi, haNK, taNK, t-haNK and qt-haNK. Much of the research with CAR-engineered NK-92 cells has been summarized in review papers [38 40]. Investigators have used first and second generation CAR constructs generally delivered by lentiviral or retroviral constructs. Although several studies confirmed the efficacy of CAR-modified NK-92, ImmunityBio, Inc. (which now holds worldwide rights to NK-92 variants) developed CAR-modified NK-92 cells by use of plasmid-based CAR gene constructs and transfected by simple electroporation to generate the NK-92 variants currently used in clinical studies [38,41].
NK-92细胞的特性促使产生了表达CAR的变体。表2列出了NK-92细胞的变体，包括NK-92ci、NK-92mi、haNK、taNK、t-haNK和qt-haNK。关于CAR工程化NK-92细胞的许多研究已在综述论文中进行了总结。研究人员使用第一代和第二代CAR构建体，这些构建体通常通过慢病毒或逆转录病毒载体传递。尽管几项研究证实了CAR修饰的NK-92的疗效，但ImmunityBio公司（现持有NK-92变体的全球权利）使用基于质粒的CAR基因构建体，并通过简单的电穿孔转染来产生目前用于临床研究的NK-92变体。
When CAR/T-cell therapies were being developed and the first clinical trial results were made available, it became clear that the high costs of producing them as well as the CRS and ICANS were sufficient motivation to explore NK cells and particularly NK-92 cells as an alternative source of CAR engineered cytotoxic immune cells.
当CAR/T细胞疗法正在开发并且首次临床试验结果公布时，生产CAR/T的高成本以及CRS和ICANS因素促使人们探索将NK细胞，特别是NK-92细胞，作为CAR工程化细胞毒性免疫细胞的替代来源。
In collaboration with the German group in Frankfurt, lentivirus based transduction was used to generate a clinical grade NK-92 variant that expresses a CAR for HER2 (ErbB2) [42]. The group went on to start a clinical trial (NCT03383978) in patients with HER2-positive glioblastoma. Patients are receiving intracranial HER2 CAR expressing NK-92 cells injected into the resection margin during relapse surgery. The single-dose escalation part of the phase I study in nine patients has been completed, and no significant side effects were observed [43]. In the current extension of the study, a more permanent micro catheter is placed into the post-surgical cavity, and patients receive repeated injections of HER2 CAR-engineered NK-92 cells. Since approximately 40% of glioblastoma tumors are positive for HER2 expression, it will be important to see whether this treatment can prolong remissions and maintain quality of life for patients with this devastating brain cancer.
在与德国法兰克福小组合作中，使用慢病毒介导的转导产生了表达针对HER2（ErbB2）的CAR的临床级NK-92变体。该小组随后启动了一项临床试验（NCT03383978），针对HER2阳性胶质母细胞瘤患者。患者在接受复发手术期间，将HER2 CAR表达的NK-92细胞注入切除边缘。该I期研究的单剂量递增部分已在9名患者中完成，未观察到显著的副作用。在当前研究的扩展中，将更永久的微导管放置在术后腔内，患者接受重复注射HER2 CAR工程化的NK-92细胞。由于大约40%的胶质母细胞瘤肿瘤表达HER2，因此将重要的是看看这种治疗是否能够延长缓解期并维持这些患有这种毁灭性脑癌患者的生活质量。
To further dissect why HER2 taNK cells kill their tumor target so effectively, the group at the University in Dresden, Germany, used special confocal microscopy for live imaging of NK-92 cells and its HER2 CAR variant. The purpose of the study was to determine the different steps of lytic granule movement in both cell lines [44]. They found that although unmodified NK-92 cells are able to form conjugates with Her-2 expressing cancer cells, their lytic granules do not polarize toward the synapse with cancer targets. In contrast, CAR expressing NK-92 cells or haNK cells, expressing a high-affinity Fc receptor in combination with a HER2 specific antibody, were able to
effectively polarize their granules toward the synapse and release them for effective killing of the cancer target.
为了进一步剖析为什么HER2 taNK细胞能如此有效地杀死其肿瘤靶标，德国德累斯顿大学的小组使用特殊的共聚焦显微镜对NK-92细胞及其HER2 CAR变体进行活体成像。该研究的目的是确定两种细胞系中溶酶体颗粒运动的不同步骤。他们发现，尽管未修饰的NK-92细胞能够与表达Her-2的癌细胞形成共轭物，但它们的溶酶体颗粒不会极化朝向与癌细胞靶标的突触。相比之下，表达CAR的NK-92细胞或表达高亲和力Fc受体与HER2特异性抗体结合的haNK细胞能够有效地将它们的颗粒极化朝向突触并释放它们，以有效地杀死癌细胞靶标。
In addition to systemic infusion of NK-92 cells/variants, intra tumoral injection has been studied. Our group has shown in a murine model that intra-tumor injection of CD19 CAR-modified NK-92 cells can induce regression of subcutaneous lymphoma [45]. Remarkably, upon re-challenge of the mice with the same lymphoma cell line, no tumor re-growth occurred. This strongly suggest that intra-tumor injection of NK-92 cells can induce a systemic memory-like immune response. In a murine glioblastoma model, intra-tumor injection of ErbB2 CAR-NK-92 cells the Frankfurt group confirmed that specific IgG antibodies against the tumor had developed in treated mice that together with T cells provided long-term protection against rechallenge with the same tumor [42].
除了全身输注NK-92细胞/变体外，还研究了瘤内注射。我们的小组在小鼠模型中表明，瘤内注射CD19 CAR修饰的NK-92细胞可以诱导皮下淋巴瘤的消退。值得注意的是，在用相同淋巴瘤细胞系重新挑战小鼠时，没有肿瘤重新生长。这强有力地表明，NK-92细胞的瘤内注射可以诱导系统性记忆样免疫反应。在小鼠胶质母细胞瘤模型中，法兰克福小组证实，在接受ErbB2 CAR-NK-92细胞治疗的小鼠中，针对肿瘤产生了特异性IgG抗体，这些抗体与T细胞一起为小鼠提供了长期的保护，使其免受相同肿瘤的复发。
(3) t-haNK (targeted high-affinity NK-92) cells expressing a high-affinity FcR and a CAR （表达高亲和力FcR和CAR的t-haNK（靶向高亲和力NK-92）细胞）
The t-haNK modification of NK-92 uses a polycistronic plasmid comprising FcR (158V) and erIL-2 as well as a first-generation CAR. In contrast to CAR/T cells, these t-haNK cells not only provide CAR-specific target cell recognition and killing but also have additional target cell killing mechanisms in place including spontaneous NK-mediated cytotoxicity as well as ADCC through their high-affinity FcR [38,41].
t-haNK对NK-92的修饰使用包含FcR（158V）和erIL-2的多顺反子质粒。与CAR/T细胞不同，这些t-haNK细胞不仅提供CAR特异性靶细胞识别和杀伤，还具有其他靶细胞杀伤机制，包括自发的NK介导的细胞毒性以及通过其高亲和力FcR进行的ADCC效应。
A number of clinical-grade t-haNK clones have been generated expressing the CARs for PD-L1, CD19, Her-2 and epidermal growth factor receptor (EGFR) [4,38]. PD-L1 t-haNK cells have been given to a number of patients with solid tumors in a phase I/II study. In some patients with triple-negative breast cancer and pancreatic cancer, some remarkable responses have been observed. Importantly, no significant side effects were noted, in particular no CRS/ICANS (ImmunityBio, internal data from ongoing studies).
已经产生了一系列表达针对PD-L1、CD19、Her-2和表皮生长因子受体（EGFR）的CAR的临床级t-haNK细胞。PD-L1 t-haNK细胞已在一项I/II期研究中给予多名实体瘤患者。在一些患有三阴性乳腺癌和胰腺癌的患者中，观察到了一些显著的反应。重要的是，没有观察到显著的副作用，特别是没有CRS/ICANS（ImmunityBio正在进行的内部研究数据）。
The next generation of t-haNK clones introduces a fourth gene coding for a molecule of choice. In the CD19 t-haNK cell line, the gene for the CCR7 homing receptor for lymphatic tissue was incorporated into such a quadrocistronic construct [46] (qt-haNK). Encouraging preclinical data including a reduction of human lymphoma growth were noted in NOD SCID mice compared with a CD19 t-haNK without CCR7. The quadrocistronic concept also allows to incorporate genes coding for proteins that can positively affect the tumor microenvironment such as a TGF-beta trap or IL-12 (ImmunityBio, internal data).
下一代t-haNK克隆引入了编码选定分子的第四个基因。在CD19 t-haNK细胞系中，将编码CCR7淋巴组织归巢受体的基因纳入了这种四顺反子构建体中（qt-haNK）。在NOD SCID小鼠中，与没有CCR7的CD19 t-haNK相比，观察到了令人鼓舞的临床前数据，如人类淋巴瘤生长的抑制。四顺反子还可以接入编码可以积极影响肿瘤微环境的蛋白的基因，例如TGF-beta陷阱或IL-12（ImmunityBio内部数据）。
销售咨询
技术支持
