Objective: To study the role of 125 I and 125 I plus gemcitabine (GEM) in treatment of unresectable carcinoma of pancreas. Methods: From April 2000 to April 2003, 38 untreated patients with locally advanced pancreatic cancer (LAPC) were collected and randomized into two groups: Arm A 125 I (18 patients) and Arm B 125 I+GEM (20 patients). Eligibility criteria were: cytologically and pathologically proven pancreatic carcinoma, Karnofsky performance status (kps) 60 80, age 18 75 years, adequate hematological, renal and liver function, and controllable pain. Arm A patients were treated with 125 I implants. Arm B patients started chemotherapy within 10 14 d post operatively following the implant procedure. Chemotherapy doses were as follows: GEM 1 000 mg/m 2 weekly × 3 followed by 1 week of rest for 3 cycles. In addition, all patients underwent laparotomy and surgical staging. The surgical procedures performed were biopsy, gastric bypass and biliary bypass. The total activity and number of seeds used were as recommended by Anderson. The mean activity, minimal peripheral dose (MPD), and volume of implants were 20 mCi, 14 000 cGy, and 53 cm 3, respectively. Results: Overall response rate (CR+PR) in Arm A was 37.6% and in Arm B it was 44.5% ( P >0.05). PR median duration in Arm A was 6.7 months and in Arm B it was 4.8 months ( P <0.05). Clinical benefit response was experienced by 11.7 % of Arm A compared with 42.1% of Arm B ( P <0.05). The incidences of hematological toxicity (such as neutropenia) between Arm A and Arm B were 5.8% and 21.1%, respectively ( P >0.05). The survival rates of 12 and 24 month were 32.5%, 16.3% for Arm A and 61%, 38.7% for Arm B ( P =0.04). The rate of complication of Arm A was lower than that of Arm B without statistical significance. Conclusion: To some extent, 125 I or 125 I plus GEM is able to lead to a moderate objective response for LAPC with obstructive jaundice on the base of biliary bypass or/and gastric bypass, but 125 I plus GEM is mor
