Somatostatin receptor 2 expression in nasopharyngeal cancer is induced by Epstein Barr virus infection: impact on prognosis, imaging and therapy

Abstract

Nasopharyngeal cancer (NPC), endemic in Southeast Asia, lacks effective diagnostic and therapeutic strategies. Even in high-income countries the 5-year survival rate for stage IV NPC is less than 40%. Here we report high somatostatin receptor 2 (SSTR2) expression in multiple clinical cohorts comprising 402 primary, locally recurrent and metastatic NPCs. We show that SSTR2 expression is induced by the Epstein-Barr virus (EBV) latent membrane protein 1 (LMP1) via the NF-κB pathway. Using cell-based and preclinical rodent models, we demonstrate the therapeutic potential of SSTR2 targeting using a cytotoxic drug conjugate, PEN-221, which is found to be superior to FDA-approved SSTR2-binding cytostatic agents. Furthermore, we reveal significant correlation of SSTR expression with increased rates of survival and report in vivo uptake of the SSTR2-binding 68Ga-DOTA-peptide radioconjugate in PET-CT scanning in a clinical trial of NPC patients (NCT03670342). These findings reveal a key role in EBV-associated NPC for SSTR2 in infection, imaging, targeted therapy and survival.

Conflict of interest statement

B.V. is a consultant for Venthera (Palo Alto, US), iOnctura (Geneva, Switzerland) and Karus Therapeutics, Oxford, UK and has received speaker’s fees from Gilead (Foster City, US). T.Me. has acted as a consultant for Tarveda Therapeutics (Watertown, MA, US). G.W. serves as a scientific advisor to ViraTherapeutics and Boehringer Ingelheim. S.M. receives research grants from Pfizer, Bayer, MSD, AstraZeneca, Nextcure, Roche, BMS, Amgen and Lily. M.C. reports personal fees from Astellas, personal fees from Janssen, grants and personal fees from Ferring, personal fees from MSD Oncology, personal fees from Illumina, nonfinancial support from AstraZeneca, personal fees and nonfinancial support from Varian, nonfinancial support from GenomeDx Biosciences, nonfinancial support from Medlever, outside the submitted work. All other authors declare no competing interests.

Figures

Fig. 1. Evaluation of SSTR2 expression in a multi-institution clinical cohort of NPC.

a Anatomical localization and representative images of hematoxylin and eosin (H&E) stained histology, somatostatin receptor 2 (SSTR2) expression assessed by immunohistochemistry (IHC) and Epstein-Barr Virus (EBV)-encoded small RNAs (EBER) assessed by in situ histochemistry). b Beanplot of the SSTR2 IHC score in EBV-positive (n = 278) and EBV negative (n = 60) cases of NPC (W = 3085.5, p = 4.0e-14, Wilcoxon two-sided test). c Heatmap representation of the clinical annotations in relation to SSTR2 expression levels (asterisks indicate significant associations with SSTR2 expression using multivariate analysis). d, e SSTR2 status and EBV status were not statistically different in the primary (n = 37), local recurrent (n = 47), and metastatic (n = 20) tumor tissue (images of rNPC42; ×400; scale bar 25 µm and summary of data; two-sided Fisher’s test on proportion of SSTR2 positivity in primary, metastatic and local recurrence samples; p = 0.32). Representative tumor samples from a single patient are shown in d. Source Data are provided as a Source data file.

Fig. 2. EBV infection is associated with SSTR2 upregulation in NPC—in vitro experiments and external validation dataset.

a Immunofluorescence analysis of EBNA1 (red) and SSTR2 (green) expression in two representative examples of cultured primary respiratory epithelial cells before and after infection with the epitheliotropic M81 EBV strain. Nuclei are counterstained with DAPI (blue); Replication (n = 5). b Percentages of cells positive for EBNA1 or co-expression of SSTR2 and EBNA1 c LMP1 induces SSTR2 expression in NP69, an immortalized normal nasopharyngeal epithelial cell line (relative quantities of SSTR2 were calculated using the comparative threshold cycle method and normalized using human beta-actin as endogenous control. Data is presented as a ratio relative to vector control). d LMP1-mediated SSTR2 induction in NP69 cells is inhibited by ectopic expression of TRAF3, a negative regulator of NF-κB. e In NP69 cells, LMP1-induced SSTR2 expression is suppressed by the NF-κB inhibitor BAY 11-7085 and MEK inhibitor U0126. f Both CTAR1 and CTAR2 regions of LMP1 (Supplementary Fig. 2) are essential for LMP1-mediated SSTR2 induction. LMP1 mutant constructs 3 A, Delta 8 C and 3 A + Delta 8 C target CTAR1, CTAR2, and both regions respectively. g SSTR2 expression by LMP1 is dose dependent. h In C666-1 cells, LMP1-induced SSTR2 expression is suppressed by the NF-κB inhibitor BAY 11-7085 and MEK inhibitor U0126. i siRNAs mediated knockdown of the subunits of activated NF-κB signal complexes, NFκB1 (p105/p50), RELB, NFκB2 (p100/p52), or c-Jun in C666-1, both resulted in significant SSTR2 suppression. j Left panel, PCA on independent RNA-seq data of NPC (n = 113) identifies two groups of NPC tumors. The color of the samples is based on unsupervised hierarchical clustering. Middle panel, Differential gene expression analysis between Group 1 and 2 tumors shows that SSTR2 is highly-expressed in Group 1 tumors (log2 fold change = 2.3, adjusted p = 2.7e-32). Right panel, Pathway analysis demonstrates that Group 1 tumor show significant enrichment of viral biogenesis pathways. k Left panel, Heatmap of gene expression of EBV genes in an independent cohort of NPC (tumor n = 31, normal n = 10). Middle panel, Microarray SSTR2 expression is positively correlated with viral LMP1 expression (W = 129, p = 0.041, Wilcoxon two-sided test). Center line displays the median, boxes display the interquartile range. Whiskers display 1.5× the interquartile range. Outliers lie beyond the whiskers. Right panel, Pathway analysis demonstrates that LMP1-expressing tumor samples are enriched in viral biogenesis pathways. *, **, ***, and **** denote a significant difference between groups of P < 0.05, P < 0.01, P < 0.0001, and P < 0.0001, respectively. Source Data are provided as a Source data file.

Fig. 3. In vitro and in vivo effects of SSTR2 agonists on the C666-1 NPC cell line.

a Immunohistochemical characterization of C666-1, NPC43, and C17 cell lines cultured in vitro and from xenografted C666-1, C15, C17, or C18 tumor tissues (scale bar 100 µm); Replication n = 2. b In vitro dose response curves and half-maximal effective concentration (EC50 values) of the indicated SSTR agonists on C666-1, NPC43, and C17 cells. EC50 = half-maximal effective concentration. c Growth curves of C666-1 tumors in nude mice treated with vehicle (n = 9), octreotide (n = 9), or PEN-221 (n = 8). The dotted lines indicate the time points of drug injection. d Kaplan–Meier curves of athymic nude mice with C666-1 tumors, treated with vehicle control (n = 9), octreotide (n = 9), or PEN-221 (n = 8), with dotted lines showing time points of drug or vehicle injection (*p = 0.0368; two-sided Log-rank Mantel-Cox test). e Geneset enrichment analysis reveals upregulation of senescence pathways 24 h post lanreotide treatment (left), and upregulation of apoptosis and mitotic spindle assembly pathways 24 h post-PEN-221 treatment (right) in treated vs untreated cell lines. f mRNA sequencing analysis of C666-1 cells treated in vitro (72 h with PEN-221) reveals downregulation of SSTR2 expression (two-sided Wald test, adjusted p = 3.9e-7). Center line displays the median, boxes display the interquartile range. Whiskers display 1.5× the interquartile range. Source Data are provided as a Source data file.

Fig. 4. Visualization and prognostic value of SSTR2 expression in NPC patients.

a Visualization of SSTR2 expression by 68Ga-DOTA-TATE PET-CT imaging (clinical characteristics and SSTR2 status of NPC patients undergoing 68Ga-DOTA-TATE PET-CT imaging are shown in Supplementary Table 4). b Correlation of SSTR2 expression with in vivo uptake of 68Ga-DOTA-TATE. SSTR2 IHC score shows significant positive correlation to SUVmax of biopsied lesion. (black circles: biopsied lesions, n = 12; Spearman’s correlation coefficient: Rs = 0.65; p = 0.023). d SSTR2 expression status remains prognostic independent of EBV status, age, and primary tumor (T), lymph node (N), and metastasis (M) staging. n.s not significant; *p < 0.05; **p < 0.01; ***p < 0.001. Vertical lines display the hazard ratio estimate, horizontal lines display the 95% confidence interval. e Proposed model of NPC oncogenesis and cancer progression involving EBV and SSTR2 expression. In the multistep carcinogenesis of NPC, inactivation of tumor suppressor genes is believed to occur prior to EBV infection and to be induced by dietary carcinogens and other environmental factors. Infection of nasopharyngeal cells with EBV and establishment of a latent infection probably occurs at a late stage in the acquisition of the malignant phenotype. Genetic alterations identified in premalignant nasopharyngeal epithelium may play crucial roles to support stable EBV infection. Once a premalignant cell has been infected by EBV, it appears to rapidly evolve towards an invasive tumor, with the stage of EBV-positive in situ carcinoma being very transient. SSTR2 expression is acquired following the onset of latent EBV infection by LMP1 expression via NF-ĸB signaling. On the basis of this tentative scenario, pharmacological agonists of SSTR2 are expected to provide the maximal benefit for three types of indications: (1) as part of the initial curative treatment of the primary tumors, (2) as part of adjuvant treatment following clinical remission of the primary tumor; (3) with a prophylactic intent for subjects at risk of NPC manifested by EBV serological changes and/or increasing circulating EBV DNA load. Source Data are provided as a Source data file.