Abstract Body (Do not enter title and authors here): Background: The mechanism of hemodialysis arteriovenous fistula (AVF) vein intima hyperplasia (IH) may be specific to its layers (intima, media, and adventitia) but is poorly understood. To address this gap, we studied layer-specific transcriptomics related to IH in veins of two-stage AVF creation.

Methods: Vein samples were collected during the initial creation of brachial artery-basilic vein AVF (1st stage) and vein transposition surgery 6 weeks later (2nd stage) and were sectioned in OCT. Tissues from each layer were separately collected by laser microdissection. Following Takara’s SMARTer Stranded Total RNA-Seq Pico Input protocol, libraries were constructed and sequenced in NovaSeq. The sequencing reads were processed by fastp, aligned to genome GENCODE 45 by STAR and gene-level counts were estimated by RSEM. Intima was assessed by histology. Layer-specific differentially expressed genes in pre-access or AVF veins associated with IH in AVF veins were detected by edgeR and further identified the enriched pathways and predicted functions using Qiagen IPA.

Results: 44 veins were collected from 22 patients. 13,522 protein-coding genes that had a raw count of at least 10 in more than 70% of samples were included in analysis. 10 AVF veins had severe IH. Among the layers of pre-access vein, intima had the largest number of genes significantly associated with severe AVF IH and most or all these genes were upregulated (35/4, 26/2, 5/0 for the number of up/downregulated genes in intima, media, and adventitia of pre-access vein, respectively); cellular movement and immune cell trafficking were more strongly activated in intima of pre-access vein with severe AVF IH. No genes in any layer of AVF vein were significantly associated with severe AVF IH. From 1^st to 2^nd stage, more genes significantly changed expression in intima with severe AVF IH than intima without severe AVF IH (197/150 vs 144/93 for up/downregulated genes), while media (20/4 vs 149/31) and adventitia (99/27 vs 196/70) were the opposite; TGFB1, angiotensinogen, and TNF were more strongly activated in intima, while cell survival and migration were diminished in media and adventitia with severe AVF IH, respectively.

Conclusion: Intima of pre-access vein has more genes associated with severe AVF IH. The layer-specific gene expression changes during AVF development and predicted functions are associated with AVF IH. These data can be used to develop layer-specific therapies to reduce AVF IH.