The multicenter trial, open since April 2024 across 12 Japanese hospitals, has reached the case count needed for performance evaluation. Craif is targeting a Japanese marketing approval application in 2026 and approval in 2027.

TOKYO, September 17, 2026 - Craif Inc., a bio-AI company developing a urine analysis platform for early detection of cancer and other diseases, has completed patient enrollment and sample collection in the multicenter pivotal trial of its software as a medical device (SaMD) for aiding the diagnosis of pancreatic cancer. The trial opened in April 2024 and has now enrolled the number of cases required for performance evaluation.

Craif is targeting a manufacturing and marketing approval application in 2026 at the earliest, and approval in 2027.

This is a Japanese regulatory pathway. The application will be a manufacturing and marketing approval application filed with Japan's Pharmaceuticals and Medical Devices Agency (PMDA), which reviews on behalf of the Ministry of Health, Labour and Welfare. It is not an FDA submission, and the device holds no FDA status.

The device analyzes microRNA, a class of regulatory molecules whose discovery was recognized with the 2024 Nobel Prize in Physiology or Medicine. Craif was founded in 2018 and has published its urinary microRNA work in pancreatic cancer in eClinicalMedicine, a Lancet Discovery Science journal. In Japan, the company also offers miSignal, a urine-based cancer risk test distributed through medical institutions, clinics and pharmacies. miSignal is a risk assessment test, not a medical device, and it does not diagnose cancer.

What the device does

The device is software as a medical device. It calculates a score from the expression levels of microRNA and other markers in urine and returns a positive or negative result, which is used to aid the diagnosis of pancreatic cancer.

Craif plans it for use in patients who are suspected of having pancreatic cancer and who are at high risk. For the purposes of the trial, high risk was defined as having at least one of the following: type 2 diabetes, chronic pancreatitis, a family history of pancreatic cancer, intraductal papillary mucinous neoplasm (IPMN), pancreatic cyst, or pancreatic duct dilation.

Why pancreatic cancer

Pancreatic cancer is difficult to catch early and carries a poor prognosis. Deaths from it in Japan have risen year over year, and in 2023 it passed gastric cancer to become the third leading cause of cancer death in the country, according to the Ministry of Health, Labour and Welfare Vital Statistics 2023.

Five-year relative survival is around 10 percent, which is notably low among malignancies, and it falls further at advanced stages: 6.8 percent at Stage III and 1.4 percent at Stage IV, according to the National Cancer Center Japan Cancer Registry and Statistics, five-year survival for 2013 to 2014.

Early disease produces few symptoms, and those that appear are rarely specific. Tumor markers such as CA19-9 often show no change. No useful screening method for early detection has been established.

Craif is using AI to identify microRNA and related expression profiles characteristic of pancreatic cancer, including early-stage disease, and is developing a software medical device that analyzes those profiles to aid diagnosis. MicroRNA is considered well suited to early detection because tumor cells and surrounding cells are thought to begin secreting microRNA in abnormal patterns, carried by exosomes, while the tumor is still small.

Where it fits in clinical practice

Craif expects the device to be used the way existing tumor markers and abdominal ultrasound are used today: in patients who carry risk factors for pancreatic cancer and in whom pancreatic cancer is suspected, to help the physician decide whether to proceed to more definitive workup.

Trial overview

• Trial name: Study of the diagnostic performance of a machine learning model using urinary biomarkers in pancreatic cancer
• Start: April 2024
• Populations: pancreatic cancer patients, high-risk patients, and healthy adults

Participating institutions: 12 sites across Japan, in no particular order

• Kawasaki Medical School Hospital
• Kumagaya General Hospital
• Kagoshima University Hospital
• Aizu Chuo Hospital
• Ehime Prefectural Central Hospital
• Asahikawa Medical University Hospital
• Wakayama Medical University Hospital
• Tokyo Women's Medical University Hospital
• Juntendo University Hospital
• Yamaguchi University Hospital
• Tottori University Hospital
• Utsukushigaoka Hospital (Hokubukai Medical Corporation)

UMIN trial ID: UMIN000053882

Registry record: https://center6.umin.ac.jp/cgi-open-bin/ctr/ctr_view.cgi?recptno=R000061507

What happens next

With enrollment and sample collection complete, Craif will clean the data and conduct performance evaluation analysis, then file for medical device approval in Japan in 2026.

In parallel, Craif has obtained the Type 1 medical device marketing authorization holder license from the Tokyo Metropolitan Government as part of building out its manufacturing and marketing structure. In Japan, the Type 1 license covers highly controlled devices, meaning Class III and Class IV. The company is targeting regulatory approval in 2027 and is preparing to move to commercial use promptly after approval.

About Craif

Craif is a bio-AI company founded in 2018 that develops non-invasive tests for the early detection of cancer. Its platform combines proprietary biomarker detection from urine and other bodily fluids with AI to assess disease risk. Building on one of the world's largest urine datasets, Craif aims to create a urine analysis platform that detects a wide range of diseases early, with cancer as the first step. The company is based in Tokyo and operates a wholly owned laboratory in San Diego, opened in 2026. Learn more at craif.com.

Media inquiries
press_us@craif.com