Risk factors for postgastric endoscopic submucosal dissection bleeding in direct oral anticoagulant users.

OBJECTIVES: Bleeding after endoscopic submucosal dissection (ESD) for gastric tumors in patients taking antithrombotic drugs, in particular direct oral anticoagulants (DOACs), remains unresolved; therefore, we evaluated the risk factors for post-ESD bleeding and drug differences in patients taking DOACs. METHODS: We included 278 patients taking antithrombotic drugs who underwent gastric ESD between January 2017 and March 2022. Antithrombotic drugs were withdrawn following the 2017 guidelines (Appendix on anticoagulants including DOACs). To further clarify differences in antithrombotic agents' effects, the peri-cancerous mucosa in the resected specimen was pathologically evaluated according to the Updated Sydney System. Multivariate analysis was performed to assess the risk of post-ESD bleeding. RESULTS: The incidence of post-ESD bleeding in patients taking DOACs was 19.6% (10/51). Among patients taking antithrombotic drugs, DOACs were identified as a possible factor involved in post-ESD bleeding (odds ratio [OR] 4.92). Among patients taking DOACs, possible factors included resection length diameter ≥30 mm (OR 3.72), presence of neutrophil infiltration (OR 2.71), lesions occurring in the lower third of stomach (OR 2.34), and preoperative antiplatelet use (OR 2.22). Post-ESD bleeding by DOAC type was 25.0% of patients (4/16) receiving apixaban, in 20.0% (3/15) receiving edoxaban, in 21.4% (3/14) receiving rivaroxaban, and in none of those receiving dabigatran. CONCLUSIONS: The administration of DOACs was shown to be a possible factor involved in post-ESD bleeding, and risk factors for patients taking DOACs included neutrophil infiltration. The pharmacological differences in the effects of DOACs contributing to bleeding in gastric ulcers suggest comparatively less bleeding with dabigatran after ESD.

[Cystic duct carcinoma in a patient with a history of cholecystectomy: a case report].

An 83-year-old Japanese man who underwent cholecystectomy for cholecystolithiasis 17 years ago visited our hospital owing to epigastric pain. He was initially diagnosed with choledocholithiasis and acute cholangitis following white blood cell, C-reactive protein, total bilirubin, alkaline phosphatase, and γ-glutamyltranspeptidase level elevations along with common bile duct stones on computed tomography (CT). Moreover, CT, magnetic resonance imaging, endoscopic retrograde cholangiography (ERC), and endoscopic ultrasonography (EUS) also revealed a 2-cm-diameter mass arising from the remnant cystic duct. The cytology of the bile at the time of ERC was not conclusive. However, EUS-assisted fine needle aspiration (EUS-FNA) of the mass confirmed the diagnosis of adenocarcinoma of the remnant cystic duct. The patient underwent extrahepatic bile duct resection. Cystic duct carcinoma following cholecystectomy is rare. We report a case diagnosed by EUS-FNA.

The risk of metastatic recurrence after non-curative endoscopic resection with negative deep margins for early colorectal cancer: Two-center retrospective cohort study.

INTRODUCTION: Non-curative endoscopic resection of T1 colorectal cancers (CRC) carries a substantial risk of recurrence. However, previous studies have reported a significant proportion of cases in which the deep margin of endoscopic resection was positive for cancer due to the technical difficulties of colorectal endoscopic submucosal dissection (ESD). With the advancement of endoscopic technology and techniques resulting in the reduction of positive resection margins, it is important to reassess the long-term prognosis and major risk factors for recurrence in cases of negative deep margins. METHODS: We conducted a retrospective cohort study of consecutive patients with T1 CRC who underwent endoscopic resection between January 2006 and December 2021 with negative deep margins. Histological findings of the resected specimens were analyzed to determine the risk factors associated with the primary outcomes of this study, including recurrence and cancer-related deaths. RESULTS: The median age of the 190 patients was 70 years, of which 63% were male, and endoscopic treatment was performed in 64% by endoscopic mucosal resection and 36% by ESD. Eighty two patients were in the curative resection (CR) group and 108 were in the non-curative resection (NCR) group, wherein the latter comprised 79 patients who underwent additional surgery (AS) and 29 patients who did not receive AS. Five-year recurrence-free survival rates were 98.4% (95% CI 89.3-99.8) for CR, 98.3% (95% CI 88.8-99.8) for NCR with AS, and 73.7% (95% CI 46.5-88.5) for NCR without AS. Lymphatic invasion and budding grade 2/3 were the major risk factors for recurrence, with hazard ratios of 40.7 (p<0.001) and 23.1 (p=0.007), respectively. Of the patients in the NCR group without AS, the five-year recurrence-free rate was 85.6% (95% CI 52.5-96.3) if there were no major risk factors (i.e. no lymphatic invasion or budding grade 2/3) (n=21), whereas the prognosis was poor in the presence of one or more of the major risk factors, with a median recurrence-free survival and disease-specific survival of 2.5 and 3.1 years, respectively (n=8). DISCUSSION/CONCLUSION: In endoscopically resected T1 CRC with negative deep margins, lymphatic invasion or budding grade 2/3 may indicate a higher risk of recurrence when followed up without additional surgery.

Diagnostic accuracy of frozen section biopsy for early gastric cancer extent during endoscopic submucosal dissection: a prospective study.

BACKGROUND: Accurate diagnosis of the lateral extent of early gastric cancer during endoscopic submucosal dissection (ESD) is crucial to achieve negative resection margins. Similar to intraoperative consultation with a frozen section in surgery, rapid frozen section diagnosis with endoscopic forceps biopsy may be useful in assessing tumor margins during ESD. This study aimed to evaluate the diagnostic accuracy of frozen section biopsy. METHODS: We prospectively enrolled 32 patients undergoing ESD for early gastric cancer. Biopsy samples for the frozen sections were randomly collected from fresh resected ESD specimens before formalin fixation. Two different pathologists independently diagnosed 130 frozen sections as "neoplasia," "negative for neoplasia," or "indefinite for neoplasia," and the frozen section diagnosis was compared with the final pathological results of the ESD specimens. RESULTS: Among the 130 frozen sections, 35 were from cancerous areas, and 95 were from non-cancerous areas. The diagnostic accuracies of the frozen section biopsies by the two pathologists were 98.5 and 94.6%, respectively. Cohen's kappa coefficient of diagnoses by the two pathologists was 0.851 (95% confidence interval: 0.837-0.864). Incorrect diagnoses resulted from freezing artifacts, a small volume of tissue, inflammation, the presence of well-differentiated adenocarcinoma with mild nuclear atypia, and/or tissue damage during ESD. CONCLUSIONS: Pathological diagnosis of frozen section biopsy is reliable and can be applied as a rapid frozen section diagnosis for evaluating the lateral margins of early gastric cancer during ESD.

Effectiveness of second-look endoscopy after gastric endoscopic submucosal dissection in patients taking antithrombotic agents: a multicenter propensity score matching analysis.

BACKGROUND: The risk of bleeding after gastric endoscopic submucosal dissection (ESD) in antithrombotic agent users has increased, and its management remains a problem. Second-look endoscopy (SLE) following gastric ESD in antithrombotic agent users may be effective in preventing delayed bleeding, but this requires elucidation. Therefore, this study aimed to investigate the efficacy of SLE in reducing bleeding after gastric ESD in patients receiving antithrombotic agents. METHODS: This retrospective cohort study was conducted at 19 referral hospitals in Japan. A total of 1,245 patients who were receiving antithrombotic agents underwent gastric ESD between January 2013 and July 2018. The incidence of delayed bleeding was compared between SLE and non-SLE groups using propensity score matching analysis. RESULTS: Overall, 858 patients (SLE group, 657 patients; non-SLE group, 201 patients) were analyzed. After matching, 198 pairs were created. Delayed bleeding occurred in 10 patients (5.1%) in the SLE group and 16 patients (8.1%) in the non-SLE group [odds ratio (OR) 0.605, 95% confidence interval (CI) 0.23-1.46, p = 0.310]. In the subgroup analysis, SLE reduced the incidence of delayed bleeding in patients receiving heparin bridging therapy (6.3% and 40.0%, respectively; p = 0.004). In the SLE group, prophylactic coagulation did not significantly reduce delayed bleeding compared to the no treatment group (14.6% and 8.6%, respectively; p = 0.140). CONCLUSIONS: SLE was ineffective in reducing bleeding after gastric ESD in antithrombotic agent users, overall. A prospective comparative study is warranted to definitively evaluate the effectiveness of SLE in reducing bleeding in high-risk patients.

Dasatinib-induced colitis: clinical, endoscopic and histological findings.

BACKGROUND: Dasatinib, a second-generation tyrosine kinase inhibitor, is widely used in patients with haematological malignancies. The main side effects of dasatinib are myelosuppression and pleural effusion; however, colitis, such as haemorrhagic colitis and cytomegalovirus (CMV) colitis, have been reported as rare side effects. There are only a few studies conducted on dasatinib-induced colitis. AIMS: This study aimed to clarify the clinical, endoscopic and pathological features of dasatinib-induced colitis. METHODS: This retrospective study included 51 consecutive patients who received dasatinib therapy between June 2009 and July 2020. Dasatinib-induced colitis was defined as the presence of colitis symptoms, exclusion of other diseases that could cause colitis, and improvement in symptoms after dasatinib withdrawal or dose reduction. CMV positivity was determined based on the positive result of CMV immunostaining. RESULTS: Dasatinib-induced colitis was diagnosed in nine of 51 patients (17.6%), and most of the symptoms were mild diarrhoea and bloody stools. The endoscopic findings were characterised by loss of vascular pattern (100%) and multiple small erosions (83.3%) which were mainly found in the transverse and descending colon. In a patient who underwent follow-up colonoscopy once a year while taking dasatinib, endoscopic findings changed from initial erythematous spots to multiple erosions, and finally to multiple small round elevations with erosion on the top that disappeared after discontinuation of dasatinib. Anti-CMV therapy was administered to one patient, but the treatment failed. All patients with dasatinib-induced colitis were cured after the discontinuation of dasatinib. CONCLUSION: Physicians should consider CMV reactivation to manage dasatinib-induced colitis.

[A case report of Legionella pneumonia in a patient with ulcerative colitis undergoing remission maintenance using azathioprine].

A 67-year-old man was diagnosed with ulcerative colitis one year ago. Remission was induced via the oral administration of prednisolone and azathioprine;prednisolone was gradually reduced and discontinued. He maintained remission with azathioprine but developed fever and general malaise and visited the Kagawa Prefectural Central Hospital. Chest radiography and a urinary antigen test revealed Legionella pneumonia. His symptoms reduced immediately after the initiation of levofloxacin. Azathioprine suppresses cellular immunity and may increase the risk of Legionella pneumonia.

Neuroendocrine carcinoma arising from Barrett's esophageal adenocarcinoma: a case report.

Neuroendocrine carcinoma in Barrett's esophagus is rare and its developmental mechanisms remain unclear. Neuroendocrine carcinoma arising in Barrett's esophagus with adenocarcinoma was detected at an early stage and resected by endoscopic submucosal dissection. Detailed pathological examination revealed that the neuroendocrine carcinoma originated via differentiation of the preexisting adenocarcinoma. A 79-year-old man presented with a flat protruding lesion in the esophagogastric junction. Esophagogastroduodenoscopy revealed a red flat 10-mm protruding lesion in the Barrett's epithelium and a shallow depression at the distal end. Narrow band imaging with magnification showed that the blood vessels in the protrusion were dilated and meandered irregularly, while those in the depression were small and did not form a network; the blood vessels were missing in some parts of the depression. Well-differentiated adenocarcinoma was diagnosed after analysis of the biopsy specimen of the protrusion, and endoscopic submucosal dissection was performed. The pathological diagnosis was neuroendocrine carcinoma with an adenocarcinoma component.

[A case of metastatic prostate cancer with gastric metastases and high serum CEA].

A 74-year-old male who was receiving endocrine therapy for prostate cancer, with multiple bone and lymph node metastases (T2bN1M1 Stage D2), underwent follow-up computed tomography (CT). The CT revealed multiple liver metastases, a high serum CEA level, and an unchanged PSA level. Upper gastrointestinal endoscopy showed an elevated lesion with mucosal erosion on the lesser curvature of the middle gastric corpus, revealed to be a metastatic prostate cancer lesion following immunohistochemical confirmation. This case demonstrates the potential for gastric metastases in patients with advanced prostate cancer and high serum CEA levels.

Clinicopathological examination of ESD as salvage therapy for esophageal cancer after definitive chemo-radiation therapy.

BACKGROUND AND STUDY AIMS: Salvage therapy for esophageal cancer following chemo-radiation therapy (CRT) has not been established. We aimed to evaluate endoscopic submucosal dissection (ESD) as a salvage therapy based on histopathological features of lesions. PATIENTS AND METHODS: We compared 10 lesions in eight patients with local residual, recurrent, or metachronous esophageal squamous cell carcinoma treated by ESD after CRT (CRT group) and 59 lesions treated by ESD without CRT (non-CRT group) during the same period. RESULTS: The en bloc resection rate was 100 % while the complete resection rate was 80.0 % in the lesions after CRT, indicating no difference between the CRT and non-CRT groups. Pathological examination showed that fibrosis was more intense in the lamina propria mucosa, muscularis mucosa, and submucosa. The muscularis mucosa was thicker in both non-tumor and tumor sites in the CRT group compared to the non-CRT group. However, severe submucosal fibrosis was observed only in one lesion in the CRT group. The maximum diameter of the submucosal artery was significantly larger in the CRT group ( P  < 0.001). CONCLUSIONS: Compared to the non-CRT group, the lesions in the CRT group were accompanied by fibrosis while the muscularis mucosa were thicker; however, severe fibrosis of the submucosa was rare. It is important to dissect the muscularis mucosa appropriately during ESD, which makes successful dissection of the submucosa possible. Attention should be paid to bleeding from large arteries.

Bleeding in patients who underwent scheduled second-look endoscopy 5 days after endoscopic submucosal dissection for gastric lesions.

BACKGROUND: Bleeding after endoscopic submucosal dissection (ESD) in antithrombotic drug users is still one of the important issues to be solved. We performed scheduled second-look endoscopy (SLE) 5 days after ESD, when the resumption of antithrombotic agents is assumed to have achieved a steady state, rather than on the day after ESD. We investigated bleeding incidence and the status of ulcers. METHODS: A total of 299 lesions in 299 patients subjected to ESD for gastric neoplasms were enrolled. A double dose of proton pump inhibitors was administered after ESD. SLE was planned 5 days after ESD. Post-ESD bleeding occurring before SLE was defined as early phase post-ESD bleeding, whereas bleeding after SLE was defined as later phase post-ESD bleeding. Forrest IIa and IIb ulcers are defined as high-risk ulcers requiring prophylactic hemostasis. We investigated risk factors for post-ESD bleeding, particularly focusing on the use of antithrombotic agents and the presence of high-risk ulcers requiring prophylactic hemostasis during SLE. RESULTS: Under a double dose of proton pump inhibitors, early phase post-ESD bleeding occurred in 2.3% of non-users (5/218) and 6.2% of users of antithrombotic agents (5/81). High-risk ulcers were found in 19.0% of the cases during scheduled SLE (55/289). Later phase bleeding occurred in 5.5% of cases [2.8% of non-users (6/213) and 13.2% of users of antithrombotic agents (10/76)]. Cox regression analysis revealed that the risk factor for post-ESD bleeding was antithrombotic treatment (HR: 3.56; 95% CI: 1.63-8.02, p = 0.002) alone. Among patients with high-risk ulcers, a statistically significant increase in bleeding was observed in the later phase in patients under antithrombotic therapy, compared to those not receiving any antithrombotic agents (p = 0.001). CONCLUSIONS: Antithrombotic treatment is a risk factor for post-ESD bleeding despite SLE being scheduled 5 days after ESD. Later phase post-ESD bleeding was observed in 13.2% of the patients under antithrombotic treatment even after prophylactic hemostasis for high-risk ulcers. TRIAL REGISTRATION: This study was registered in the UMIN Clinical Trials Registry System ( 000023306 ). Retrospectively registered on 23rd July 2016.

Risk factors for non-curative resection of early gastric neoplasms with endoscopic submucosal dissection: Analysis of 1,123 lesions.

Although the frequency of residual disease and recurrence following endoscopic submucosal dissection (ESD) has markedly decreased, a few cases of residual disease and recurrence following ESD are still observed. The aims of the present study were to clarify the causes of non-curative resection and to investigate the risk factors. A total of 1,123 early gastric neoplasm lesions treated by ESD were investigated. Non-curative resection was defined as histological positivity of the resected margins, vascular invasion or failure of en bloc resection. Cases of non-curative resection were classified as being caused by one of three reasons: Inadequate technique, pre-procedural misdiagnosis or problems in the histological diagnosis. Following classification, the cases of non-curative and curative resection were compared based on a range of patient characteristics: Procedure time, and size, type and location of the lesions. The frequency of non-curative resection was 16% (182 lesions). Non-curative resection occurred due to inadequate technique in 59 cases, pre-procedural misdiagnosis in 88 cases and problems in the histological diagnosis in 35 cases. Multivariate analysis revealed that a large lesion size, long procedure time and inexperienced endoscopist were associated with a significantly higher risk of non-curative resection due to an inadequate technique. Furthermore, it was found that lesions located in the upper area of the stomach and cancer with submucosal invasion were associated with a significantly higher risk of non-curative resection due to pre-procedural misdiagnosis. In conclusion, the present study has shown that the major reasons for non-curative resection are an inadequate technique and pre-procedural misdiagnosis. The risk factors for these problems have been clarified.

A case of pancreatic mucinous cystic neoplasm that enlarged during pregnancy and was resected after childbirth.

A 28-year-old pregnant woman presented with an abdominal mass at 12 weeks' gestation. Magnetic resonance imaging revealed a 7 cm cystic lesion arising extrinsically from the pancreatic head, which was diagnosed as a mucinous cystic neoplasm. Although we recommended surgical excision during the second trimester, the patient refused the procedure and chose to continue her pregnancy. We monitored the lesion and noted that it gradually grew to 13 cm over the course of the pregnancy. Subsequently, we enucleated it after childbirth. Histopathological examination was compatible with high-grade dysplasia and confirmed the diagnosis of mucinous cystadenoma of the pancreas.

[Two cases of dabigatran-induced esophageal ulcer indicating the usefulness of drug administration guidance].

Here we report two cases of dabigatran-induced esophageal ulcer. Case 1 was a 67-year-old man who presented with heartburn that developed a month after dabigatran administration. Case 2 was an 81-year-old woman who presented with epigastralgia that developed within a few days of dabigatran administration. Endoscopic findings were similar in both cases, including shallow esophageal ulcers covered with a thin whitish membrane. The patients were advised to consume the drug with plenty of water during meals and to remain in a sitting position for 30 min after consumption. This method successfully decreased their symptoms and ulcers, indicating that drug administration guidance is extremely effective in managing dabigatran-induced esophageal injury.

[Present status of gastrointestinal damage due to non-steroidal anti-inflammatory drugs (NSAIDs)].

Non-steroidal anti-inflammatory drugs (NSAIDs) are roughly divided into a low-dose aspirin group used for primary and secondary prevention of cardiovascular events and non-aspirin NSAIDs used for treatment of bone and joint diseases. Both cause gastrointestinal damage directly or indirectly. In the present study, we reviewed gastrointestinal damage due to non-aspirin NSAIDs with respect to the esophagus, stomach/duodenum, small intestine and colon. Damage due to NSAIDs occurs in all digestive tracts and since the analgesic effect of NSIADs hides subjective symptoms, the symptoms are often not treated until they are advanced to a serious state. Further, patients receiving NSAIDs are mostly elderly and have complications so that the onset of the conditions is serious and prevention is important. It is necessary to investigate a method that is effective for preventing damage for all digestive tracts and the mechanisms of damage must be understood for this reason.

Exposed blood vessels of more than 2 mm in diameter are a risk factor for rebleeding after endoscopic clipping hemostasis for hemorrhagic gastroduodenal ulcer.

BACKGROUND AND AIM: There are few clinical studies on the risk factors for rebleeding based on the endoscopic hemostatic procedure carried out, including ulcer characteristics such as exposed blood vessels. The present study aims to clarify the risk factors for rebleeding after endoscopic clipping hemostasis for hemorrhagic gastroduodenal ulcers. METHODS: A retrospective study was carried out with data collected during the 10-year period from January 2000 to December 2009 for 312 consecutive patients with hemorrhagic gastroduodenal ulcer. Two hundred and ninety-three patients (216 men and 77 women; mean age, 67.0 ± 15.0 years) who underwent endoscopic clipping as the initial hemostatic treatment were analyzed. The risk factors for rebleeding were determined by comparing 271 patients who did not rebleed after initial treatment with 22 patients who developed rebleeding. RESULTS: The success rate of initial clipping hemostasis was 100%; however, rebleeding occurred in 7.5% (22/293) and a multivariate analysis identified exposed blood vessels of more than 2 mm in diameter as independent risk factors for rebleeding (P = 0.0124, odds ratio 6.25 [95% CI: 1.53-13.62]). CONCLUSIONS: Endoscopic clipping monotherapy is effective for hemorrhagic gastroduodenal ulcers; however, exposed blood vessels of more than 2 mm in diameter in the initial endoscopic procedure are a risk factor for rebleeding.

Characteristics of serious complicated gastroduodenal ulcers in Japan.

BACKGROUND/AIMS: Even in Japan where the incidence of H. pylori infection is high, patients with gastroduodenal ulcers caused by NSAIDs are on the increase. A prospective study was conducted to elucidate the characteristics of gastroduodenal ulcers complicated with ulcerous hemorrhage and perforation among Japanese. METHODOLOGY: The subjects were 305 consecutive cases with hemorrhaging ulcers and 76 consecutive cases with perforated ulcers basis at Kagawa Prefectural Central Hospital between January 2000 and December 2008. These subjects were divided into 3 groups (lowdose aspirin, non-aspirin NSAIDs and non-NSAIDs) and were further stratified by the presence of an H. pylori infection. RESULTS: The 76 perforating ulcers were composed of 54 non-NSAIDs ulcers and 22 NSAIDs-related ulcers. The 305 hemorrhaging ulcers were composed of 156 non-NSAIDs ulcers, 94 non-aspirin NSAIDs ulcers and 55 on low-dose aspirin. The mortality for the non-aspirin NSAIDs group (12.8%, 12/94) and for the low-dose aspirin group (10.9%, 6/55) was significantly higher (p<0.01 and p<0.05, respectively) than the corresponding figure (2.6%, 4/156) for the non-NSAIDs group. The causes of death were exacerbation of or complications from the background diseases (Charlson Co-morbidity Index 3≤) (Odds ratios (OR) 6.01, 95% CI (1.98-18.89)). CONCLUSIONS: Approximately 50% of the gastroduodenal ulcers with complications found in Japanese are NSAIDs-related and may take a fatal turn. It is necessary to take measures to prevent the complicated ulcers corresponding to risk factors such as the severity of background diseases.

Risk factors for perforation and delayed bleeding associated with endoscopic submucosal dissection for early gastric neoplasms: analysis of 1123 lesions.

BACKGROUND AND AIM: Endoscopic submucosal dissection (ESD) is a useful procedure for the treatment of early gastric neoplasms; however, this advanced technique has also resulted in an increase in serious complications such as perforation and delayed bleeding. This study aimed to elucidate the risk factors for these complications. METHODS: A total of 1123 lesions diagnosed with early gastric neoplasms and treated by ESD at three institutions were investigated. Retrospectively, patients with or without these complications were compared on the basis of the patient characteristics and treatment results. RESULTS: Perforation occurred in 27 lesions (2.4%) and delayed bleeding in 56 lesions (5.0%). Multivariate analysis indicated that lesions located in the upper area of the stomach (odds ratio [OR]: 4.88, 95% confidence interval [CI]: 2.21-10.75) was associated with a significantly higher risk of perforation, and that age ≥ 80 years (OR: 2.15, 95% CI: 1.18-3.90) and a long procedure time (OR: 1.01, 95% CI: 1.001-1.007) were associated with a significantly higher risk of delayed bleeding after ESD. The en bloc resection rate (74% vs 94%) and curative resection rate (48% vs 85%) of lesions with perforation were significantly lower than those without perforation. The rate of residual disease or recurrence after ESD was significantly higher in lesions with delayed bleeding than that without delayed bleeding (5.4% vs 0.84%). CONCLUSIONS: This study demonstrated risk factors for perforation and delayed bleeding associated with ESD. Furthermore, it was clarified that perforation and delayed bleeding influenced post-procedure results and prognosis after ESD.

[A case of hemolytic uremic syndrome after adjuvant chemotherapy with gemcitabine in a patient with pancreatic cancer].

A 63-year-old man with Stage IVa pancreas tail cancer was admitted for a distal pancreatectomy and splenectomy; adjuvant chemotherapy with gemcitabine was also administered. The chemotherapy was terminated after 16 courses due to hemolytic anemia, thrombocytopenia and renal dysfunction. Plasma exchange was performed; however the patient's renal function was diminished, requiring chronic hemodialysis. Physicians should be cautious of hemolytic uremic syndrome as a possible adverse reaction to gemcitabine and be aware that tests are needed for its early detection.