Trajectories and Influencing Factors of Psychological Distress in Nasopharyngeal Carcinoma Patients Receiving Radiotherapy (Incorporating Genetic Factors): A Multicenter Longitudinal Study

Background: This multicenter longitudinal study seeks to investigate the dynamic changes of psychological distress (PD) in nasopharyngeal carcinoma (NPC) patients during radiotherapy, and reveal the expression profiles and regulatory networks of circRNAs in these NPC patients. Methods: 282 newly diagnosed NPC patients from three hospitals in China were included. Participants completed questionnaires and provided blood samples. PD trajectories were identified via a latent class growth model (LCGM). Moreover, the factors that influence the PD trajectories were explored. Whole transcriptome sequencing was performed to investigate genetic factors. The real-time quantitative PCR was applied to validate circRNAs. We predicted the target mirnas, target Mrnas and target RNA-binding proteins (RBPs) of the top 10 malregulated circrnas. Subsequently, circRNA-miRNA-mRNA (ceRNA) and circRNA-RBP networks were constructed. In addition, the role of circRNA and target mRNA parent genes was predicted by KEGG and GO analysis. Results: LCGM identified two of the most important PD trajectories during radiotherapy in NPC patients: Class 1 “decline distress group” (11.0%) and Class 4 “rise distress group” (20.3%). Household monthly per capita income, coping strategies, and perceived social support emerged as important predictors of PD trajectories. Regarding genetic factors, 600 circRNAs and 123 miRNAs were identified as being significantly differentially expressed. Notably, hsa_circ_0004277 demonstrated significant differences between patients in the rise and decline distress groups (P < 0.01). ceRNA and RBP networks may influence the pathophysiology of PD in NPC patients undergoing radiotherapy. Conclusion: This study unraveled that PD trajectories in NPC patients during radiotherapy were heterogeneous, indicating the need for screening and timely interventions within this population. Furthermore, the expression patterns of ceRNA and circRNA–RBP networks and pathways related to these networks suggested a potential role of circRNAs in developing PD among NPC patients receiving radiotherapy. Our experimental design involves collecting questionnaire data and blood samples from patients at admission and after radiotherapy through a multi-center longitudinal approach. Then, based on the latent categorization model (a statistical method), they were divided into the rise distress group and the decline distress group. Then, randomly select the blood of 3 patients from each of the rise and decline distress groups for whole transcriptome sequencing (a total of 12 blood samples at two time points, with D1-3 being the blood of patients in the rise psychological distress group at admission; TD1-3 is the blood of patients in the group with a rise distress after radiotherapy. R1-3 represents the blood of patients in the group with decline distress upon admission. TR1-3 is the blood of patients in the group with decline distress after radiotherapy. Among them, the single-end (e.g. D1.fastq.gz) is the raw data of miRNA, and the paired end (e.g. D1_R1.fastq.gz) is the raw data of IncRNA, mRNA, and circRNA. Peripheral blood (5 mL) was collected from participants who provided informed consent at T1 and T4. Using the PAXgene Blood RNA Kit (BD, USA), the total RNAs were extracted from samples of the rise distress group (Class 4) and decline distress group (Class 1) divided by LCGM following the manufacturer’s instructions. A Nanodrop ND-1000 (Thermo Fisher, USA) was applied to measure the concentration and quality of total RNAs. The sequencing library was built using around 1–2 μg of total RNAs per sample. RNA sequencing services were provided by CloudSeq Inc. (Shanghai, China). Trizol (Invitrogen, USA) was utilized to extract Total RNA, and rRNA was removed via the GenSeq® rRNA Removal Kit (GenSeq, Inc., Shanghai, China). Directional RNA libraries were built via the GenSeq® Directional RNA Library Prep Kit. RNA was fragmented to ~300 nt in length. Reverse transcriptase and random hexamers were utilized to synthesize the first-strand cDNA. Moreover, the dUTP Mix was applied to synthesize the second-strand cDNA. The double-stranded cDNA fragments underwent end repair, adapter ligation, and dA-tailing. The cDNA, ligated with adapters, was PCR-amplified and purified to generate the final sequencing libraries. Furthermore, the libraries were sequenced in paired-end 150 bp mode.