ErbB2 (HER2)-CAR-NK-92 cells for enhanced immunotherapy of metastatic fusion-driven alveolar rhabdomyosarcoma

Abstract

Introduction: Metastatic rhabdomyosarcoma (RMS) is a challenging tumor entity that evades conventional treatments and endogenous antitumor immune responses, highlighting the need for novel therapeutic strategies. Applying chimeric antigen receptor (CAR) technology to natural killer (NK) cells may offer safe, effective, and affordable therapies that enhance cancer immune surveillance.

Methods: Here, we assess the efficacy of clinically usable CAR-engineered NK cell line NK-92/5.28.z against ErbB2-positive RMS in vitro and in a metastatic xenograft mouse model.

Results: Our results show that NK-92/5.28.z cells effectively kill RMS cells in vitro and significantly prolong survival and inhibit tumor progression in mice. The persistence of NK-92/5.28.z cells at tumor sites demonstrates efficient antitumor response, which could help overcome current obstacles in the treatment of solid tumors.

Discussion: These findings encourage further development of NK-92/5.28.z cells as off-the-shelf immunotherapy for the treatment of metastatic RMS.

Keywords: ERBB2 (HER2/neu); cancer immunotherapy; chimeric antigen receptor; rhabdomyosarcoma; xenograft.

Figure 1

Lysis of ErbB2-positive, 2D patient-derived tumor organoid aRMS cells by parental NK-92 or NK-92/5.28.z cells. (A) ErbB2 surface expression on RMS102, RMS127 and RMS335 cells was confirmed by flow cytometry. Antigen density per cell quantified by flow cytometry is shown. (B) Cytotoxicity data of NK-92 or NK-92/5.28.z cells during short-term (3 hour) coculture with 2D patient-derived tumor organoid aRMS cells from europium release assays are given. Data of three independent experiments are shown as mean ± SD and differences were analyzed with a two-tailed Student´s t-test and were considered significant for p< 0.05 (*), p< 0.01 (**), p< 0.001 (****), p<0.0001 or ns.

Figure 2

Cytotoxic capacity against ErbB2-positive 3D tumor spheroids. (A) ErbB2 surface expression on RH30 cells was confirmed by flow cytometry. (B) Expression of the anti-ErbB2-targeted CAR on NK-92/5.28.z cells was confirmed by flow cytometry. (C) Exemplary coculture images of NK-92 or NK-92/5.28.z cells added to established 3D RH30GFP/luc+ tumor spheroids on day 4 are shown. Imaging was performed on days 4, 5, 6, 8 and 10. (D) The areas with green fluorescent signals were quantified using Fiji. Data of three independent experiments are shown as mean ± SD. Differences were analyzed with a one-way ANOVA using the Bonferroni method and were considered significant for p< 0.05 (*), p< 0.01 (**) or ns.

Figure 3

Therapeutic potential of NK-92 and NK-92/5.28.z cells in a metastatic xenograft NSG mouse model. (A) An experimental metastasis model in NSG mice carrying low tumor burden of a human aRMS (RH30GFP/luc+) xenograft was established to evaluate the therapeutic potential of NK-92 and NK-92/5.28.z cells in vivo. (B–D) Tumor development and growth was monitored and quantified by BLI combining both dorsal and ventral signals described by the luminoscore for each animal, shown by five representative examples. The mean ± SD is shown and differences were analyzed with a one-sided ANOVA using the Bonferroni-Dunn (nonparametric) method. Differences with p< 0.05 (*) were considered statistically significant. (E) The overall survival of untreated control (n=5), NK-92 cell-treated (n=12) and NK-92/5.28.z cell-treated (n=10) mice is shown. Differences between the survival of different treatment groups were analyzed by the log-rank (Mantel-Cox) test. Differences with p< 0.05 (*) were considered statistically significant. Black arrows indicate animals selected for exemplary images in Figure 4A .

Figure 4

Macroscopic tumor lesions and immunofluorescence staining of metastatic tumors. (A) Tumor cells (Alexa 488, green) and immune cells (Alexa 647, red, in liver and lung additionally indicated by arrows) are shown; cell nuclei were stained with DAPI. Exemplary animals with tumor engraftment were selected from BLI analysis and were sacrificed on day 80 (control), day 72 (NK-92) and day 73 (NK-92/5.28.z), respectively (black arrows in Figure 3D ). Immunofluorescence staining showed that NK-92 cells, and to a more pronounced extent NK-92/5.28.z cells, preferentially homed to spleens as well as to metastatic tumor sites in lung and liver, resulting in disseminated tumors (control), localized tumors (NK-92), and small tumor nodules (NK-92/5.28.z). Low level NK-92 and NK-92/5.28.z cell infiltration was observed within metastatic tumor sites. (B) The size of the spleen in each treatment group is shown. Differences were considered significant for p< 0.01 (**) or ns.

Figure 5

Localization of effector cells within the tumor tissue. Immunofluorescence staining of blood vessels (green) and NK-92/5.28.z cells or NK-92 cells (right image) (red, indicated by arrows) in liver tumors. Animals were sacrificed on day 72 or day 73.

Figure 6

Homing and persistence of NK-92 and NK-92/5.28.z cells in vivo. Animals were sacrificed on day 50, 58, 66, 74 and 80 (control), day 58, 61, 65, 66, 72, 73, 74, 78, 89 and 90 (NK-92) and day 62, 68, 73, 78, 84, 85, 92 and 105 (NK-92/5.28.z). (A) Exemplary plot of human CD45 positive NK-92 and NK-92/5.28.z cells, detectable in all analyzed treatment tissue samples by flow cytometry. (B–D) Human immune cells (B), NK-92/5.28.z cells (C) and aRMS cells (D) were detected by qPCR and are shown as boxplots for different organs of the control (n=5), NK-92 cells (n=12), and NK-92/5.28.5 cells (n=10). (E) The overall tumor burden was calculated for each animal. Differences were analyzed with a one-way ANOVA using the Bonferroni method and were considered significant for p< 0.05 (*), p< 0.01 (**) or ns (not labled).