The relations of nasopharyngeal carcinoma with vitamin D and its indicators : a multicentre case-control study in Hong Kong

Abstract

The aetiology of nasopharyngeal carcinoma (NPC) remains controversial. Classical risk factors cannot fully explain the recent decline in NPC incidence rates in most parts of high-risk regions where sunlight is usually plentiful. Vitamin D deficiency is common even in the sunniest areas, and is associated with smoking and Epstein-Barr virus (EBV) activation. Given the close relation of NPC with smoking and EBV activation, vitamin D may also have a role in NPC pathogenesis. However, no report has shown the relation of NPC with indicators of vitamin D exposure. Therefore, I studied: (1) the prevalence of vitamin D deficiency, and the factors associated with circulating 25-hydroxyvitamin D (25(OH)D), (2) the relation of EBV IgA against viral capsid antigen (VCA-IgA) with putative risk factors of NPC, and the factors associated with circulating 25(OH)D, and (3) the relation of NPC with indicators of vitamin D exposure, and whether the relation is modified by putative risk factors of NPC.

In a multicentre case-control study in Hong Kong (a high-risk region of NPC) in 2014-2017, 821 incident NPC cases and 1,512 hospital controls recruited from five major regional hospitals, and 356 community controls recruited from two communities were used for the cross-sectional, case-control and genetic analyses. Information on demographic, socioeconomic and lifestyle factors was collected. I also measured EBV VCA-IgA serostatus, and circulating 25(OH)D and its related genetic variants. My analysis showed that the questionnaire data of most putative risk factors of NPC had acceptable reliability.

A large proportion of the subjects were vitamin D deficient (serum 25(OH)D <50 nmol/L: 52.4% for NPC cases, 55.8% for hospital controls and 63.5% for community controls). Associations of serum 25(OH)D level with exposure to any occupational hazards, season of blood collection (summer), duration of sunlight exposure and hand skin tone were positive, but negative with the composite genetic score (based on four variants: rs3829251, rs12794714, rs4588 and rs1155563). NPC putative risk factors and factors of circulating 25(OH)D were not associated with EBV VCA-IgA serostatus. Case-control analyses showed no association of NPC with circulating 25(OH)D. However, after adjusting for potential negative confounders (socioeconomic status score, occupational exposure and EBV VCA-IgA), circulating 25(OH)D was negatively associated with NPC risk (Odds ratio (OR), 95% CI for the categories of 37.5-<50 and ≥50 nmol/L versus <37.5 nmol/L was, respectively 0.54, 0.33-0.90 and 0.57, 0.35-0.91; P for linear trend = 0.041). Genetically predicted lowest level of 25(OH)D as denoted by highest composite genetic score (‘4-8’) was associated with higher NPC risk in subjects with EBV VCA-IgA seropositivity (OR: 1.99, 95% CI: 1.14-3.47, versus ‘0-1’ indicating highest 25(OH)D).

To conclude, I first found a negative association of high NPC risk with circulating 25(OH)D, and genetically predicted lower 25(OH)D exposure, particularly in subjects with EBV VCA-IgA seropositivity. My findings suggest that lower 25(OH)D exposure might be a cause of NPC. Therefore, drivers of 25(OH)D and other bio-active form of vitamin D should be investigated to confirm whether the relation is causal, and if so, to identify new targets for NPC prevention, early detection and treatment.