Multistate oral carcinogenesis-A prospective cohort study and a parallel case-control study in Taiwan

Abstract

Background: To characterize multistate oral carcinogenesis, we conducted a cohort study of patients with oral precancer and a parallel case-control study of oral cancers and controls in Taiwan.

Methods: During 2013-2019, we recruited patients with oral precancer (n = 1998) or invasive oral cancer (n = 768) and hospital-based controls (n = 717). Precancer patients were followed up biannually for up to five years; questionnaire data and biospecimens were collected at multiple timepoints. Precancer natural history (regression/persistence, incidence, progression) was evaluated through follow-up visits and linkages with Taiwan's Cancer Registry.

Cohort updates: Cross-sectionally, 71 % of oral precancers and 62 % of cancers were attributable to betel-quid chewing, smoking, and alcohol. Precancer patients had substantially elevated risk of oral cancer (standardized-incidence-ratio vs. Taiwan general population = 14.1; 95 %CI = 12.0-16.6). Among precancer patients, 156 incident invasive oral cancers occurred (median follow-up = 6.4 years; incidence rate = 1,221/100,000 person-years; annual incidence = 1.2 %; 1-year cumulative-incidence = 1.8 %; 5-year cumulative-incidence = 6.9 %; 10-year cumulative-incidence = 9.5 %). Baseline precancer histopathology strongly predicted risk of progression to oral cancer (5-year cumulative-incidence: no-dysplasia = 5.2 %, mild-dysplasia = 7.1 %, moderate-dysplasia = 32.8 %, severe-dysplasia = 45.9 %). Most oral cancers (88.5 %) were preceded by precancers identified during the study. The study has established a resource of >63,500 biospecimens, including biopsies (n = 6,012), oral cytology (n = 18,422), oral rinses (n = 15,054), saliva (n = 15,066), and blood (n = 8,990). Ongoing investigations are characterizing oral carcinogenesis at the epidemiologic, macroscopic, microscopic, microbiomic, and genomic levels.

Conclusions: A majority of oral precancers/cancers in Taiwan are caused by betel-quid chewing, smoking, and alcohol. Patients with oral precancer have substantially elevated risk of site-concordant oral cancer. We highlight our study as a resource to collaboratively address questions regarding oral precancer/cancer natural history and clinical management.

Keywords: Alcohol; Betel quid; Biomarkers; Multistate carcinogenesis; Natural history; Oral cavity cancer; Oral leukoplakia; Oral precancer; Prevention; Screening; Smoking.