Researchers are mapping a molecular puzzle with global implications, proving that Taiwan is not just catching up on oncology research but defining its direction.
The incidence of lung cancer in Taiwan no longer tracks neatly with smoking rates. Clinicians now confront a profile in which non-smokers, especially women, make up the majority of new cases, with adenocarcinoma — a form of the disease that develops in the lung’s outer tissue — emerging as the predominant type.
Dr. Yu Chong-jen, superintendent of National Taiwan University Hospital (NTUH), explains that adenocarcinoma accounts for about 70% of all lung cancer cases in Taiwan. Among never-smokers, the pattern is even more pronounced: about 90% are diagnosed with the condition.
Lung adenocarcinoma’s demographics also reveal a striking gender tilt, with women making up roughly two-thirds of never-smoker cases. In Taiwan, the overall male-to-female incidence ratio has now narrowed to about 1:1, as male rates have leveled off or declined with reduced smoking, says Dr. Yu. The trend contrasts sharply with the persistent gender gap in Western countries, where men are between four and nine times more likely than women to develop the disease.
“We have about 18,000 new lung cancer patients” per year, says Dr. Yu. “Among them, about 60% are never-smokers, with the major histological type being adenocarcinoma.”
The long-held explanations — cooking fumes, incense exposure, or secondhand smoke — pose too small a risk to account for such a sweeping epidemiological shift. Moreover, they no longer reflect the modern social patterns of Taiwanese women, who smoke less, cook less, and spend less time in temples, yet see their lung cancer risk continuing to rise rapidly. According to Dr. Yu, environmental and genetic factors are now the primary drivers.
As patient profiles shift, the screening process becomes more challenging. Standard low-dose CT scans, which use X-rays to detect early lung abnormalities, often yield results that require cautious interpretation.
“About 30% to 40% of people tend to have abnormal findings,” says Dr. Yu. “For example, nodules or fibrosis, many of which reflect prior inflammation rather than cancer.” Because imaging cannot always distinguish indolent or pre-cancerous lesions from invasive disease, Taiwan’s physicians now rely on structured follow-up protocols — scheduling checks at three-, six-, or twelve-month intervals and escalating only when concerning changes appear.
Even so, anxiety and uncertainty often drive care toward intervention. Dr. Yu describes the trade-offs candidly: ground-glass-type nodules — faint, hazy spots on CT images — usually don’t become invasive, meaning surgery could be avoided. Yet both patients and physicians frequently opt for surgical resection rather than years of watchful waiting.
“That’s why we consider this kind of early surgical approach to be overtreatment or overdiagnosis,” he says. About 30% of surgically removed solitary nodules ultimately prove benign, making “case-by-case shared decision-making” essential to balance the small radiation risk of repeated CT scans against unneeded procedures.
Research driving precision
To understand why so many lung-cancer patients in Taiwan develop the disease despite not smoking, Chen Yu-ju, president of the Chemical Society Located in Taipei and professor at Academia Sinica’s Institute of Chemistry, has turned inward, mapping the molecular signatures that define each case.
The Taiwan Cancer Moonshot Project, led by Chen in collaboration with the U.S. National Cancer Institute, applies proteogenomics to map how DNA, RNA, and proteins interact within individual patients. Her team began with lung adenocarcinoma after clinicians observed that never-smoker tumors behave differently from smoking-related ones and sought molecular confirmation of those differences.
Chen’s team found that some patients diagnosed at an early stage exhibit proteomic profiles — patterns of protein expression that reveal how a tumor behaves — strikingly similar to those seen in advanced disease. These “late-like” molecular signatures can emerge even in tumors thought to be curable through surgery alone, yet roughly half of such patients later relapse. “Traditional staging can underestimate risk when the tumor’s molecular behavior already mirrors advanced cancer,” she says.
The team is now validating a four-protein blood-based assay to identify these late-like patterns non-invasively. “This blood test is the first of its kind in the world — nobody has ever done this before,” says Chen.
Her research also reveals two distinct carcinogenic signatures among Taiwanese women. One stems from exposure to nitrosamines — compounds found in food preservatives, cosmetics, and polluted air. The other involves APOBEC-driven mutations, an internal DNA-editing process more common among younger women. These findings, showing how environmental and genetic mechanisms intersect, are but one example of Taiwan’s potential to serve as a model for studying never-smoker lung cancer in East Asian women worldwide.
While Chen’s research sheds light on the biological roots of Taiwan’s never-smoker lung cancer profile, Dr. James Chih-hsin Yang, director of the National Taiwan University Cancer Center, has translated molecular understanding into treatment. Over the past three decades, he has built Taiwan’s infrastructure for early-phase cancer trials and helped pioneer targeted therapies for the epidermal growth factor receptor (EGFR) mutation — a genetic change found in roughly half of Taiwanese lung adenocarcinoma patients, far higher than in Western populations.
“EGFR mutation is a very special type of lung cancer that’s very prevalent in East Asia,” says Dr. Yang. Such mutations cause cells to keep receiving “grow” signals even when they shouldn’t, leading to unchecked tumor formation, particularly in non-smokers.
About 70% of patients now have identifiable genetic drivers that determine which precision drug will be most effective, he says. Expanded government support for next-generation sequencing (NGS) and advanced genetic testing under the National Health Insurance system has enabled hospitals across Taiwan to match patients with the right therapies more quickly.
One initiative that helped advance this effort was the NHRI Pilot Project, jointly launched by the Ministry of Health and Welfare and the National Health Research Institute (NHRI) in partnership with Roche, a global healthcare company and leader in pharmaceuticals and diagnostics.
The project introduced Taiwan’s first use of an FDA-approved comprehensive genomic profiling panel covering more than 300 genes, linking genomic and clinical data from 17 hospitals and focusing on six major cancers prevalent in Asia, giving patients with actionable gene mutations access to matched therapies at no cost and improving clinical trial enrollment across hospitals. Supported by an NHRI-developed common data model, it established Taiwan’s first clinico-genomic database.
For patients without such mutations, chemotherapy and immunotherapy remain the main options, but as understanding deepens, new targeted drugs are expected to follow. Dr. Yang notes that these treatments have already transformed outcomes — extending median survival from under a year to more than three years and significantly improving quality of life.
The implementation gap
Despite Taiwan’s impressive scientific advances, moving discoveries from the lab to the clinic remains a slow and uneven process. Researchers describe a system capable of breakthrough innovation but constrained by regulation, limited funding, and complex approval procedures.
Academia Sinica’s Chen says that the next step — large-scale clinical validation — is where many projects stall. “Most academic work ends once papers are published,” she notes. “But bringing discoveries into hospitals requires sustained funding, regulatory support, and effort.”
Her team’s findings — among them, a potential blood test designed to detect high-risk lung cancer at early stages — illustrate this challenge. Taiwan lacks a streamlined pathway for new diagnostic technologies, leaving many innovations in regulatory limbo. Without clear evaluation criteria or a fast-track process similar to the U.S. Food and Drug Administration’s (FDA) breakthrough designation, promising tests can take years to reach patients.
Clinicians face similar barriers. NTUH’s Dr. Yu points out that genetic testing costs remain high — around US$3,000 (nearly NT$92,00) per patient, with only part covered by National Health Insurance (NHI). Only large medical centers have the equipment to perform sequencing in-house, forcing regional hospitals to send samples elsewhere and delaying treatment decisions.
Though initiatives such as the NHRI Pilot Project have advanced integration of NGS testing with treatment use, the rapid evolution of the technology continues to outpace regulatory updates, slowing clinical adoption.
Dr. Yang of the NTU Cancer Center adds that while Taiwan’s cancer research and clinical trials now meet international standards, converting scientific progress into routine care remains slow. Taiwan’s NHI provides universal access but requires lengthy cost-effectiveness reviews before approving new tools or treatments for reimbursement — a process that can take years.
The result, researchers say, is a paradox. Taiwan leads the world in cancer data and discovery, yet bureaucracy keeps many of its breakthroughs from reaching patients. Bridging that gap — through stronger public-private partnerships, dedicated translational-research funding, and clearer regulatory pathways — will determine whether Taiwan’s innovations stay in journals or reshape global oncology.
Ultimately, Taiwan’s challenge is not innovation but implementation. The country’s ability to align research, industry, and regulation will decide whether its discoveries become lifelines for patients or remain trapped behind administrative walls.
Tools for the present
While new treatments and diagnostics move at glacial speed through research and regulatory pipelines, a new generation of Taiwan-based MedTech and AI startups is improving the way oncologists and other lung specialists work day to day. Their tools don’t replace scientific breakthroughs; they extend them, making care more efficient, accurate, and sustainable.
At EndoSemio, a Taipei-based medical AI startup affiliated with NTU, engineers are working alongside anesthesiologists to improve precision in lung surgery. Their flagship system, Blocker Eye, uses AI-based image recognition and depth sensing to visualize the position of a bronchial blocker — a device that helps isolate one lung during surgery.
“Before the surgeon can operate, anesthesiologists must isolate one lung — a delicate process that isn’t easy to maintain throughout surgery,” says Dr. Wu Chun-yu, NTUH’s Hsinchu Branch director of Anesthesiology and lead developer of Blocker Eye, which integrates EndoSemio’s AI and imaging modules. “A core issue is that our primary tools are essentially ‘dumb’ devices. We place them with meticulous, fiberoptic-guided precision, but the second we pull that scope out, we are flying blind.”
By tracking the bronchial blocker’s position in real time, physicians are proactively alerted to when the tool has been displaced. By minimizing displacement risk, which could inadvertently cause ventilation in the wrong lung, the system saves valuable operating-room minutes and improves turnover so that patients can be treated sooner.
Dr. Wu explains that conventional double-lumen tubes and bronchial blockers often move mid-surgery, forcing repeated checks and interruptions. Blocker Eye “gives us instant confirmation without needing to reinsert the bronchoscope,” he says. “I no longer worry about displacement while taking care of multiple operating rooms at once.”
In addition, Blocker Eye’s camera is mounted on a mechanical arm that allows anesthesiologists to position and adjust the blocker with precision throughout surgery — making care more efficient, accurate, and sustainable. The Blocker Eye remains in active development, with commercial release targeted for 2027 following both U.S. FDA and Taiwan FDA clearance.
Ruby Hung, EndoSemio’s chief marketing officer, notes that the system’s AI and visualization modules are built to integrate with existing hospital infrastructure and legacy devices, reducing the need for costly replacements.
Hung adds that EndoSemio’s technology is designed to complement, not compete with, larger medtech platforms. Yet smaller companies rarely get invited to collaborate within those full-package ecosystems, even when their innovations could strengthen them. This dynamic is unfortunate, as inclusive collaboration would help sustain Taiwan’s broader innovation pipeline, especially as local startups move from prototype to scale.
While EndoSemio focuses on improving precision at the bedside, another local company, Aesop Technology, tackles a different challenge: information overload. Founded as a research project at Taipei Medical University (TMU) in 2011, Aesop began by detecting prescription errors in Taiwan’s vast NHI database. Today, it has evolved into a comprehensive clinical-decision platform that uses AI to guide doctors through complex treatment pathways in real time.
Aesop Technology cofounder and Chief Product Officer Jeremiah Scholl argues that some of the most transformative uses of AI come from decision support. “Our AI doesn’t tell doctors what to do — it helps them find the information they are looking for to make decisions much faster,” he says.
Embedded directly into hospital electronic records, Aesop’s software, named Medigator, automatically loads the latest U.S. National Comprehensive Cancer Network (NCCN) cancer-treatment guidelines, cross-checks them with patient data, and provides physicians with the most recent published evidence that is relevant for their patient. Of the products Aesop offers, Medigator continuously scans journals and clinical-trial databases, updating recommendations within hours of new findings — a process Scholl describes as “bringing the latest data quietly into the doctor’s workflow.”
The platform was made possible through the national data exchange already established in Taiwan, which has allowed physicians to share patient data securely across hospitals. The foundational research projects at TMU obtained access to this data to develop the AI models it uses, showing how Taiwan’s healthcare infrastructure is shaping the island into an AI-enabled healthcare powerhouse.
At a broader level, Aesop’s analytics evaluate how well hospitals adhere to international standards and where real-world results diverge, giving tumor boards and research partners a clearer picture of where care can improve. By connecting guidelines with real outcomes, Scholl says, Taiwan’s hospitals can turn overwhelming data into practical insight — “making accuracy routine, not exceptional.”
Sustaining this momentum will require more than innovation alone. With the right policies and partnerships, Taiwan’s research strengths could shape the next generation of global cancer care.
