Abstract
145 Background: Endemic NPC is invariably linked to Epstein-Barr virus infection, and often presents as locoregionally-advanced (LA-) disease. Nonetheless, it is acknowledged that LA-NPC is clinically heterogeneous, with disparate clinical presentations of ascending (A) and descending (D) subtypes that are characterized by T3-4N0-1 and T1-2N2-3 disease, respectively. Here, we investigated the TIME between these distinct NPC phenotypes using whole transcriptome sequencing and digital pathology. Methods: We utilized a cohort of 111 patients with NPC from 2 institutions – NCCS (n=68) and JXCH (n=43) as test and validation datasets, respectively. Whole transcriptome profiling using RNAseq (Illumina, CA) was performed in the NCCS cohort, while spatial analyses of hematoxylin and eosin-stained whole slide images (WSIs) were undertaken in the NCCS cohort (35 A- and 33 D-subtypes), and validated in the JXCH cohort (26 A- and 17 D-subtypes). A customized pipeline was developed to analyze the spatial relationships between tumor-rich and immune-rich regions (tiles) within the WSIs. Mean distances between each tumor-rich and its nearest immune-rich tiles were computed for each WSI. Highest and lowest quartiles of the mean distances in the test cohort were used as thresholds to classify NPC tumors into 3 immune classes of differing immune cell density. Results: Transcriptomic profiling revealed an enrichment of immune-related genesets and differentially expressed genes in D- relative to A-subtype tumors. This corresponded to observations of higher overall immune, T and B cell scores in D- than A-subtypes by deconvolution analyses (ESTIMATE and CIBERSORT). Through spatial analyses of our WSIs, we defined three immune classes – immune-dense, -sparse, and -desert, which corresponded to mean distances of <88 μm, 88-208 μm, and >208 μm between tumour- to immune-rich tiles, respectively. D-subtype NPC had significantly more immune-rich regions compared with A-subtypes in both NCCS and JXCH cohorts ( P=0.02 [NCCS]; P=0.01 [JXCH]). This corresponded to shorter median distances between tumor-rich to immune-rich tiles in D- than A-subtypes (103 μm vs 200 μm, P=0.01 [NCCS]; 123 μm vs250 μm, P=0.01 [JXCH]). Based on our novel spatial-based classification, we observed a higher proportion of immune dense tumors in D- than A-subtypes in the NCCS cohort, with only 3/33 (9%) D-subtypes being classified as immune desert. Using the same distance thresholds, we derived a consistent trend in the JXCH cohort, where a higher proportion of D-subtypes were classified as immune dense and sparse than A-subtypes (59% vs 31%, P=0.01). Conclusions: Our results support a model where the TIME delineates the biology underpinning the disparate clinical presentations of distinct A- and D-subtypes of LA-NPC.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.