AI-Powered IHC Digital Pathology Software for PD-L1 Scoring in Non-Small Cell Lung Cancer Receives First Approval in China

Shanghai, October 22, 2020 /PRNewswire/ -- AstraZeneca announced today that the "Immunohistochemistry Digital Pathology Image Processing Software" for AI-assisted PD-L1 interpretation in non-small cell lung cancer (NSCLC), developed in collaboration with Hangzhou Dipath Technology Co., Ltd., has recently received Class II medical device registration certification from the Zhejiang Provincial Medical Products Administration. This makes it the first AI-assisted software for PD-L1 interpretation in NSCLC approved for clinical application in China.

Pathological Diagnosis: “The Doctor’s Doctor,” The First Step in Precision Medicine

In 2019, the number of new lung cancer cases in China reached 784,000, with the incidence and mortality rates accounting for 20.03% and 26.99% of all malignant tumors, respectively, ranking first among all malignancies.^[1]With the emergence of innovative therapeutic approaches such as targeted therapy and immunotherapy, the long-term survival rates and quality of life for lung cancer patients have been significantly improved. However, tailoring the most appropriate treatment regimen to each patient’s individual condition is of critical importance for clinical management and prognosis, all of which hinges on high-quality precision diagnosis.

In the field of immunotherapy, multiple anti-PD-1/PD-L1 inhibitors have been approved for clinical use in China. The expression level of PD-L1 protein on tumor cells is an important biomarker for immunotherapy in advanced non-small cell lung cancer. PD-L1 testing can help identify the population most likely to benefit from immuno-oncology treatment.

Current Status of Pathological Diagnosis: Talent Shortage and Uneven Resource Distribution

The prerequisite for precision medicine is precise diagnosis; however, China faces a severe shortage of registered practicing pathologists, far failing to meet clinical demands. According to the latest data, by the end of 2018, there were approximately 18,000 registered pathologists in China. Based on the standard staffing ratio of 1–2 pathologists per 100 hospital beds, the shortfall amounts to approximately 90,000 pathologists.^[2，3]. Furthermore, the distribution of pathologist resources across hospitals at all levels in China is uneven. As a foundational discipline, pathology directly impacts the diagnostic and therapeutic standards for various diseases at primary care hospitals.

Meanwhile, under the traditional model, pathologists determine disease types by observing cellular and tissue lesions on pathological slides under a microscope. However, manual slide review is time-consuming and labor-intensive, making it difficult to readily identify challenging tumor cells, resulting in lower efficiency. This poses a particularly stringent test for the diagnostic capabilities of primary-care hospitals.

AI-Powered Digital Pathology for Assisted Precision Diagnosis

In recent years, artificial intelligence has increasingly become a hot topic of discussion within the medical community and society at large. AI holds significant advantages in pathological diagnosis; for instance, after learning from big data, it can more accurately and rapidly differentiate and quantify tumors, thereby serving as an effective aid to pathologists.

The newly approved software leverages artificial intelligence to assist pathologists in interpreting PD-L1 expression in non-small cell lung cancer (NSCLC). Annotated by leading Chinese pathology experts, the algorithmic model effectively addresses technical challenges such as inaccurate localization caused by damaged cell membranes. Through integrated analysis, the system generates whole-slide analysis results in an unattended manner. Pathologists need only to review the results, enabling rapid and accurate whole-slide PD-L1 interpretation. This significantly reduces labor and time costs for hospital pathology departments while enhancing the quality and efficiency of pathological diagnosis.

Furthermore, by leveraging internet technology, primary care hospitals can transmit challenging PD-L1 tissue slides directly to the central hospital of the medical consortium via teleconsultation. This approach effectively addresses pain points such as the lack of access to testing at the primary level, establishing a "cloud pathology department" that breaks geographical barriers and thereby enhancing overall diagnostic capabilities.

Ms. Yang Haiying, Vice President of AstraZeneca China and Head of Medical Affairs, stated: “AstraZeneca has been deeply committed to the field of lung cancer. We aim to collaborate with partners across various sectors, leveraging advanced technologies such as machine vision, predictive algorithms, and natural language understanding to integrate into the entire diagnostic and therapeutic pathway—including screening, diagnosis, treatment, and patient follow-up. Our goal is to promote the widespread application of artificial intelligence in disease areas with the most urgent patient needs in China, comprehensively enhance healthcare efficiency, improve the five-year survival rate for lung cancer, and contribute to the realization of ‘Healthy China 2030.’”

Dr. Yang Lin, Chairman and CEO of Dipath, stated, “Dipath is currently the digital pathology company with the most comprehensive coverage of disease types. For fifteen years, we have focused on computer vision, deep learning, and digital pathology to address critical pain points in the field of pathology. Through this strategic partnership between Dipath and AstraZeneca China, we aim to resolve challenges faced by pathologists in PD-L1 testing, thereby enabling precision oncology treatment. The integration of AI and pathology represents the future trend of pathological diagnosis. We are committed to providing superior and precise diagnostic services to a broad population of cancer patients, thereby driving the development and innovation of the traditional healthcare industry.”

"2019 National Cancer Report of China"

2 "National Report on Medical Services and Quality Safety 2018"

3. "2019 Edition of China Health and Family Planning Statistical Yearbook"