patients diagnosed with oral cavity squmous cell carcinoma

Based on clinical and pathophysiological factors, this service develops precision medicine predictive models applied to the treatment strategy of oral cavity squamous cell carcinoma (OSCC). The models encompass four core domains: postoperative recurrence prediction, distant metastasis prediction, prognostic evaluation for salvage surgery, and radiomic biomarker analysis. By integrating structured clinical data — including pathological staging, surgical margins, lymph node involvement, and imaging features — the predictive framework enables clinicians to stratify patient risk with greater accuracy, facilitating early intervention and individualized treatment planning.

A proprietary institutional database of over one thousand OSCC cases has been established in-house, serving as the foundation for model training, validation, and ongoing refinement. This data infrastructure supports robust, evidence-based clinical decision-making without reliance on high-cost genomic testing such as next-generation sequencing (NGS). The approach bridges clinical oncology and artificial intelligence, offering a scalable, cost-effective solution that enhances treatment precision, reduces patient burden, and strengthens institutional capability in cancer care — with continued expansion toward multi-center international collaboration.

• Replaces high-cost genomic sequencing with clinical pathophysiological factors, making precision prediction accessible and affordable for every patient

• Research findings presented at the American Society of Clinical Oncology (ASCO) Annual Meeting for seven consecutive years

• Institutional database of over 1,000 oral cancer cases built entirely in-house

• Designated major oral cancer treatment center for central Taiwan

• National Health Insurance Administration (NHIA) feedback confirms all core performance indicators surpass the national average

• Future development roadmap includes integration of next-generation sequencing (NGS), imaging features, and AI-driven computation to further enhance model predictive accuracy

Since 2011, our institution has diagnosed approximately 200 new cases of oral cancer annually. Over the past decade, this has led to more than 3,000 newly diagnosed patients with oral cancer receiving treatment at our center. Recognizing the growing demand for personalized care, we have progressively advanced our approach toward individualized and precision medicine in the diagnosis and treatment of oral cancer.

The cornerstone of our precision medicine approach is predicting critical clinical events, rather than relying on gene sequencing-guided drug selection. To date, four core predictive models have been established, covering postoperative early recurrence, distant metastasis, salvage surgery prognosis, and radiomic biomarker analysis. For early recurrence prediction, integrating pathologic staging, extracapsular spread, and perineural invasion successfully stratified patients into low, intermediate, and high-risk groups, achieving AUC values of 0.837 and 0.703 — providing clinicians with a reliable basis for early pharmacological intervention. For distant metastasis prediction, separate models were developed for lymph node-positive and lymph node-negative patients, both exceeding AUC 0.75, enabling earlier identification and closer surveillance of high-risk individuals. For salvage surgery prognosis, our institution pioneered the world's first broadly applicable predictive model for salvage surgery necessity, achieving an AUC of 0.734 — significantly outperforming models from international teams in Italy, Belgium, Spain, and France (c-index: 0.562–0.675). Results were published in Oral Oncology, the top-ranked journal in oral cancer globally (Rank: 6/92, 6.5%).

Our institution continuously participates in the Cancer Care Quality Certification Program administered by the Health Promotion Administration (HPA), Ministry of Health and Welfare, undergoing regular data reviews and on-site evaluations. According to HPA feedback, all six core oral cancer quality indicators at our institution have surpassed the national average for three consecutive years. Notably, the 90-day mortality rate following postoperative radiotherapy and concurrent chemoradiotherapy, as well as the rate of pathological lymph node examination achieving 15 or more nodes, have each exceeded the national benchmark for three consecutive years — reflecting the sustained excellence of our treatment quality. Survival rates across all disease stages have consistently outperformed both the national average and the medical center average, demonstrating that our precision medicine strategy translates into measurably better clinical outcomes for patients.

Our institution maintains a long-term inpatient treatment retention rate of over 85% and a treatment completion rate exceeding 90%, reflecting strong patient trust in the care we provide. Since the implementation of our precision medicine strategy, annual new case volume has grown steadily, and treatment retention has risen year over year — from 85.1% in 2018 to 88.7% — demonstrating the positive impact of precision oncology on patient confidence and willingness to commit to treatment.

• 2019(e17536): Predictive score for recurrence after complete resection in locally advanced oral cavity squamous cell carcinoma.

• 2020(e18536): Predictive model for distant metastases in oral cavity squamous cell carcinoma.

• 2021(e18040): Clinical outcomes, prognosis, and predictors of salvage surgery in oral cavity squamous cell carcinoma.

• 2022(#6038): Predictors of the survival for platinum-refractory head and neck squamous cell carcinoma by using contrast-enhanced magnetic resonance imaging.

• 2023(e18052): Positive margin and advanced depth of invasion as the indicators for assessing the necessity of adjuvant chemoradiotherapy in early-stage oral cavity squamous cell carcinoma.

Our precision medicine model replaces high-cost next-generation sequencing (NGS) with clinical pathophysiological features, substantially reducing patient financial burden while remaining applicable to all oral cancer patients — overcoming the limitations of conventional precision medicine restricted to select populations. Multiple predictive models enable clinicians to identify high-risk individuals earlier, improving prognosis and quality of survival. Clinically, all six HPA core oral cancer indicators have exceeded the national average for three consecutive years, with survival rates across all disease stages outperforming both national and medical center benchmarks. Treatment retention and completion rates remain above 85% and 90%, respectively, reflecting the tangible impact of our precision medicine strategy. Academically, findings have been presented at the ASCO Annual Meeting for seven consecutive years, with publications in journals ranked within the top 20% of their fields. These achievements have elevated Taiwan's international profile in oral cancer care, while offering a reproducible, locally validated precision medicine framework with strong potential for adoption across institutions and borders.