Although early detection and diagnosis are indispensable for improving the prognosis of patients with pancreatic cancer, both have yet to be achieved. Except for pancreatic cancer, other cancers have already been screened through scent tests using animals or microorganisms, including
Caenorhabditis elegans
. While such a method may greatly improve the prognosis of pancreatic cancer, no studies have investigated the same, mainly given the difficulty of collecting suitable samples from patients with early-stage pancreatic cancer. In this study, we organized a nationwide study group comprising high-volume centers throughout Japan to collect patients with very-early-stage pancreatic cancer (stage 0 or IA). We initially performed an open-label study involving 83 cases (stage 0–IV), with subsequent results showing significant differences after surgical removal in stage 0–IA (×10 dilution:
p
< 0.001; ×100 dilution:
p
< 0.001). Thereafter, a blinded study on 28 cases (11 patients with stage 0 or IA disease and 17 healthy volunteers) was conducted by comparing very-early-stage pancreatic cancer patients with healthy volunteers to determine whether
C. elegans
could detect the scent of cancer for the diagnosis of early-stage pancreatic cancer. Preoperative urine samples had a significantly higher chemotaxis index compared to postoperative samples in patients with pancreatic cancer [×10 dilution:
p
< 0.001, area under the receiver operating characteristic curve (AUC) = 0.845; ×100 dilution:
p
< 0.001, AUC = 0.820] and healthy volunteers (×10 dilution:
p
= 0.034; ×100 dilution:
p
= 0.088). Moreover, using the changes in preoperative and postoperative chemotaxis index, this method had a higher sensitivity for detecting early pancreatic cancer compared to existing diagnostic markers. The clinical application
C. elegans
for the early diagnosis of cancer can certainly be expected in the near future.