Tracking of [¹⁸F]FDG-labeled natural killer cells to HER2/neu-positive tumors

Abstract

Introduction

The objective of this study was to label the human natural killer (NK) cell line NK-92 with [¹⁸F]fluoro-deoxy-glucose (FDG) for subsequent in vivo tracking to HER2/neu-positive tumors.

Methods

NK-92 cells were genetically modified to NK-92-scFv(FRP5)-zeta cells, which express a chimeric antigen receptor that is specific to the tumor-associated ErbB2 (HER2/neu) antigen. NK-92 and NK-92-scFv(FRP5)-zeta cells were labeled with [¹⁸F]FDG by simple incubation at different settings. Labeling efficiency was evaluated by a gamma counter. Subsequently, [¹⁸F]FDG-labeled parental NK-92 or NK-92-scFv(FRP5)-zeta cells were intravenously injected into mice with implanted HER2/neu-positive NIH/3T3 tumors. Radioactivity in tumors was quantified by digital autoradiography and correlated with histopathology.

Results

The NK-92 and NK-92-scFv(FRP5)-zeta cells could be efficiently labeled with [¹⁸F]FDG by simple incubation. Optimal labeling efficiencies (80%) were achieved using an incubation period of 60 min and additional insulin (10 IU/ml). After injection of 5×10⁶ [¹⁸F]FDG-labeled NK-92-scFv(FRP5)-zeta cells into tumor-bearing mice, digital autoradiography showed an increased uptake of radioactivity in HER2/neu-positive tumors at 60 min postinjection. Conversely, injection of 5×10⁶ NK-92 cells not directed against HER2/neu receptors did not result in increased uptake of radioactivity in the tumors. Histopathology confirmed an accumulation of the NK-92-scFv(FRP5)-zeta cells, but not the parental NK cells, in tumor tissues.

Conclusion

The human NK cell line NK-92 can be directed against HER2/neu antigens by genetic modification. The genetically modified NK cells can be efficiently labeled with [¹⁸F]FDG, and the accumulation of these labeled NK cells in HER2/neu-positive tumors can be monitored with autoradiography.

Introduction

New approaches to breast cancer immunotherapies are based on natural killer (NK) cells [1], [2]. The NK-cell-based therapies provide a high and selective cytotoxicity against tumor cells without toxicity against nonmalignant cells. The antitumoral activity of NK cells has been extensively investigated [3], [4], [5]. Breast cancer immunotherapy with NK cells can be achieved by adoptive transfer of ex vivo expanded and activated autologous or donor-derived NK cells, or by activation of endogenous NK cells through systemic application of cytokines [6], [7]. The former approach includes the use of cytotoxic NK cell lines that are suitable for clinical use [8].

Based on the high intrinsic cytotoxic activity of the human NK cell line NK-92 against a wide range of malignancies and on lack of toxicity and tumorigenic potential, Phase I/II clinical trials have been initiated [3], [8], [9], [10]. However, NK-92 cells did not show specific cytotoxicity to HER2/neu-positive tumors. Thus, the parental NK-92 cell line has been genetically modified by retroviral transduction with an pL-scFv(FRP5)-zeta-SN vector. The resulting NK-92-scFv(FRP5)-zeta cell line expresses a chimeric antigen receptor that enhances the antitumoral activity of NK-92 cells, specifically to HER2/neu-expressing cells [11]. HER2/neu (ErbB2) is an antigen that is overexpressed in many human tumors of epithelial origin and is associated with increased disease recurrence and worse prognosis [12], [13].

The effectiveness of NK cell therapy in patients is evaluated by indirect measures such as reduced tumor markers and improved survival of patients [8]. Polymerase chain reaction (PCR)-based monitoring of NK-92 kinetics in the peripheral blood of patients revealed that apparently a large portion of NK-92-cells leave the circulation within minutes of transfusion. However, as the adoptively transfused NK-92 cells were not labeled, homing of NK-92 cells to the tumor site could not be confirmed in clinical studies. Consequently, one could argue that an apparent tumor response may have occurred due to other cytotoxic drugs that were given in conjunction with the NK cells.

New molecular imaging techniques that provide NK cell tracking could directly confirm or disprove the accumulation of the NK cells in tumor tissues [14], [15], [16], [17], [18], [19], [20], [21], [22]; thus, these techniques might be able to early identify responders and nonresponders after NK cell treatment. In addition, a direct NK-cell-tracking technique could help to further evaluate the activating and inhibitory mechanisms of the function and cytotoxicity of NK cells. Various imaging techniques, such as single photon emission computed tomography, positron emission tomography (PET), optical imaging (OI) and MRI, are available for cell tracking. OI is a new inexpensive, fast and noninvasive imaging technique, based on the detection of fluorescence, with a high sensitivity. However, this technique has limited anatomical resolution and is currently not available for clinical applications [19], [20]. MRI provides three-dimensional data with high anatomical resolution and high soft tissue contrast, but with limited sensitivity for the detection of small cell populations [14], [18], [21], [22], [23], [24], [25], [26], [27]. Labeling techniques with radioactive tracers provide a very high sensitivity [15], [16], [17], [28]. [¹⁸F]Fluoro-deoxy-glucose (FDG) PET has a distinctive advantage over other conventional nuclear medicine techniques — it provides superior spatial resolution in conjunction with a Food and Drug Administration (FDA)-approved PET radiopharmaceutical; thus, the technique would be readily applicable in patients.

In this study, [¹⁸F]FDG was chosen as a clinically applicable and widely available cell-labeling tracer with documented high specificity and sensitivity for cell detection [16], [29], [30], [31], [32], [33]. In addition, after phosphorylation to [¹⁸F]FDG-6-phosphate, it is trapped in cells without further metabolism. Since the rate of dephosphorylation is low, confounding signals from free [¹⁸F]FDG are minimal within the observed time period of 2 h postinjection (pi). Cell labeling with [¹⁸F]FDG for subsequent cell tracking with PET has been previously performed in our laboratory and in other laboratories [16], [29], [30], [31], [32], [33]. Most previous cell-labeling studies have been conducted using monocytes, white blood cells and bone marrow stem cells. Forstrom et al. [29] indicated the feasibility of leukocyte labeling with [¹⁸F]FDG and its high diagnostic potential in the PET imaging of inflammation and infection. Rini et al. [32] compared the labeling of leukocytes with [¹⁸F]FDG versus the labeling of leukocytes with ¹¹¹In oxine and found comparable sensitivities and specificities of [¹⁸F]FDG–PET and ¹¹¹In scintigraphy for the detection and characterization of inflammation. Few studies have applied [¹⁸F]FDG-labeling techniques to the labeling and in vivo tracking of cytotoxic T lymphocytes and NK cells. Melder et al. [28] and Ritchie at al. [33] described different approaches to the stable incorporation of the positron-emitting substance [¹⁸F]FDG into lymphocytes. However, they described some problems, such as limited tracer uptake into the cells and loss of [¹⁸F]FDG from labeled cells.

Thus, the purpose of this study was to optimize the labeling of genetically engineered NK-92-scFv(FRP5)-zeta (targeted against HER2/neu receptors) and parental NK-92 cells (not targeted against HER2/neu receptors) with [¹⁸F]FDG for subsequent in vivo tracking to HER2/neu-positive tumors.