Patient characteristics
Out of 198 cases diagnosed with PC histologically, 192 were considered, wherein contrast-enhanced or dynamic CT was performed and reported. These cases consisted of 124 men (64.6%) and 68 women (35.4%) and the median age was 69 years (33–87 years).
Diagnostic rate of PC in initial CT reports
In total, 192 cases were considered in which contrast-enhanced or dynamic CT was performed and reported. In total, 174 cases (90.6%) were directly diagnosed with PC in initial CT reports and 18 (9.4%) were not directly diagnosed with PC in initial CT reports (Figure 1). Among these 18 cases, indirect findings, such as bile duct/pancreatic duct stenosis or dilation, were detected in 5 (2.6%) and intrapancreatic mass lesions were detected in 3 (1.6%). Specialists for pancreatic diseases were able to suspect PC in 15 cases initially among these 18 cases. The remaining three cases were not suspected of having PC as the main diagnosis from the initial CT even by specialists, however, they were diagnosed by endoscopic ultrasound (EUS)-guided fine needle aspiration (FNA).
Clinical features of cases not diagnosed with PC in initial CT reports
The 18 cases not directly diagnosed with PC in initial CT reports were reviewed in detail. The results are shown in Table 1. These 18 cases comprised of 11 men and 7 women, with a median age of 69 years (45–84 years). Their clinical features included the following, 13 patients had symptoms, including five patients with abdominal pain, four patients with jaundice, three patients with vomiting, two patients with fever, and one patient with back pain (with duplication). The remaining five cases were asymptomatic, among which two were detected during follow-up of pancreatic diseases, two were detected during follow-up of other diseases, and one was detected owing to a pancreatic tumor in magnetic resonance cholangiopancreatography.
The CT imaging methods were contrast-enhanced in 10 cases and dynamic in 8 cases. The tumor size was: 0–10 mm in 2 cases, 10–20 mm in 12 cases, and 20–30 mm in 4 cases, the median tumor size in these cases was 18 mm (9.2–28.9 mm). According to the general rules for the study of PC by the Japan Pancreatic Society published in September 2020, the stage of the disease was: IA in 4 cases, ⅠB in 4 cases, ⅡA in 4 cases, ⅡB in 1 case, Ⅲ in 1 case, and Ⅳ in 4 cases (Table 1) [10].
The tumor occupied the head, body, and tail in 12, 3, and 3 cases, respectively. The treatments administered were surgery in 13 cases, chemotherapy in 1 case, conversion surgery in 1 case, and best supportive care in 3 cases (Table 1).
Details of initial CT reports in cases not diagnosed with PC
Out of 18 cases, 10 (55.6%) did not indicate any findings associated with PC (Table 1). Figure 2 shows a representative case of those without any findings associated with PC. In this case, no finding associated with PC was identified in the report, even though the specialist noted a tumor extending outside the pancreas.
In five cases (2.6%), dilation or stricture of the bile duct or dilation or stricture of the caudal pancreatic duct was detected. Among these cases, the initial CT reports suspected benign pancreatic lesion in one case, bile duct cancer in three cases, and malignant lymphoma in one case. Figure 3 shows a representative case in which only indirect findings were noted. In this case, even though the dilation of the bile duct and pancreatic duct was reported in the CT report, PC was not detected directly. However, the specialists suspected PC from the beginning and performed EUS, leading to PC diagnosis.
Three cases (1.6%) were found to have pancreatic mass lesions. Among these cases, the initial CT reports suspected accessory spleen and bile duct cancers in one and two patients, respectively. A representative case in which a pancreatic mass was noted but not diagnosed was suspected as accessory spleen because of the regularity of the surface and internal homogeneity of the lesion (Figure 4).
How missed cases in the CT reports lead to the diagnosis
Among five cases with detected dilation or stricture of the bile duct or dilation or stricture of the caudal pancreatic duct, specialists found the pancreatic tumor and suspected PC first in four, and one was diagnosed with acute pancreatitis. Among the three cases with detected pancreatic mass lesions, specialists observed the pancreatic tumor and suspected PC first in two cases, one was diagnosed with accessory spleen. However, EUS showed a pancreatic mass in these two cases, which were diagnosed with PC by EUS-FNA.
Among the other 10 cases with neither direct nor indirect findings, the specialists suspected PC in 9 directly. In the remaining case, the specialist could not detect PC, even from when examined retrospectively. In this case, magnetic resonance cholangiopancreatography, which was performed simultaneously to follow intraductal papillary mucinous neoplasms (IPMN), detected pancreatic duct stenosis, which led to the diagnosis of PC.
To summarize, the specialists initially diagnosed PC in 15 out of the 18 cases. The three cases where PC was not reported were diagnosed as pancreatitis, accessary spleen, and IPMN, respectively. In cases not diagnosed by CT reports, the final diagnostic modality of PC was EUS-FNA in 17 cases and through upper gastrointestinal endoscopy in 1 case. In the case in which gastrointestinal endoscopy was performed, PC had invaded the stomach, and a biopsy was performed on the stomach. Twelve out of these 18 cases, which were diagnosed using EUS-FNA, underwent surgery. The median time from CT reports to diagnosis was 10 days (2–73 days).