Fig. 1. Tc-99m labeled RBC scan depicts a curvilinear accumulation of radionuclide in the left mid abdomen most compatible with hemorrhage into the descending colon.     Case Reference No. CC-0298-07   An 80-year-old Asian-American man presented with a 2-day history of bright red blood per rectum. The patient was managed with supportive care initially but bleeding per rectum continued intermittently over the next 12 hours. The patient's hematocrit had fallen from 33 to 27, and he was receiving a fourth unit of packed red blood cells. A 99m-Tc-labeled RBC radionuclide study had been done 3 hours earlier, and showed focal accumulation of activity in the mid left colon (Fig. 1). With use of a 5-F Simmons catheter, the IMA was selectively engaged and conventional cut-film arteriography performed, centered over the left upper quadrant (Fig. 2). No definite abnormality was appreciated. Given the patient's clinical and radionuclide findings, it was elected to repeat the inferior mesenteric arteriogram before catheterizing the SMA. The repeat inferior mesenteric arteriogram, obtained 10 minutes after the initial injection, showed gross extravasation of contrast material into the descending colon (Fig. 3). Superselective catheterization using a Tracker system was then performed (Fig. 4). Three 2-mm straight coils were deployed. Completion arteriography demonstrated successful occlusion of the bleeding vessel with no further extravasation (Fig. 5). The contrast material seen in the colonic lumen is residual from prior injections. The patient subsequently stopped bleeding (after the passage of a few additional bloody stools). No clinical signs of ischemia developed, and the patient was discharged home 3 days later. The value of a pre-arteriography tagged-RBC study is illustrated in this case, where the point of bleeding was localized to the mid descending colon. This knowledge was extremely useful in the deciding which artery to catheterize. Moreover, when the initial inferior mesenteric arteriogram was negative for bleeding, the appearance of the nuclear medicine study motivated us to repeat the injection. The intermittent nature of GI bleeding is well shown in this patient. The vasodilatory effect of arteriographic contrast material may have contributed to the reappearance of bleeding.		Fig. 2. Inferior mesenteric arteriogram shows no evidence of bleeding. Fig. 3. Repeat inferior mesenteric arteriogram 10 minutes later reveals obvious extravasation into the mid descending colon (arrow). Fig. 4. Selective injection of contrast material through a coaxial catheter into the bleeding vessel (arrow). Fig. 5. Repeat inferior mesenteric arteriogram following embolization shows no further hemorrhage. Residual contrast material from prior injections is present within the lumen of the descending colon.
Questions 1) At what point should patients with lower GI hemorrhage undergo arteriography? Should all patients undergo a nuclear medicine evaluation immediately or should a trial of supportive therapy be given? 2) Should this patient have been treated with a vasopressin infusion rather than embolization? What is the current role of vasopressin in the management of GI bleeding?