CORRESPONDENCE 677 FIGURE 4. Venous phase of celiac angiogram shows collaterals at the esophagogastric junction and greater omentum. The splenic vein is not opacified. The opaque plaque at the esophagogastric junction is caused by previous sclerotherapy (arrow). Occlusion of the splenic vein as a consequence of chronic pan- creatitis has been recognized with increasing frequency. 1-1o Its prev- alence was either 2.2% or 24% in two surgical series, depending on whether splenic vein obstruction had been systematically searched for with splenoportography. 11'12 Extrahepatic portal hypertension should be considered in any patient with chronic pancreatitis who has splenomegaly and gastrointestinal bleeding. Diagnosis can be made with abdominal ultrasonography, computed tomography, and angiography. Sonographic features of splenic vein thrombosis may be an echogenic thrombus within the lumen of the vein, demonstra- tion of splenic vein collateral circulation, and enlargement of the thrombosed segment of the vein. 13 On computed tomography, hy- podense thrombi will be found in the splenic vein while the portal vein is normal.i4 Angiography is the key to diagnosis and usually shows esophagogastric varices, an obliterated splenic vein, and splenomegaly. The frequency of variceal bleeding in patients with splenic vein thrombosis secondary to chronic pancreatitis is not high. 3'6'8'15 The most common location of varices is the fundus of the stomach, and gastric varices are easily missed endoscopically unless they have been suspected. Treatment for patients with splenic vein thrombosis and gastroesophageal varices depends on the presence or absence of bleeding and the magnitude of bleeding. In the past, surgery was quickly performed on patients with active bleeding unresponsive to medical management. At present, sclerotherapy seems applicable. Since gastric varices caused by segmental portal hypertension are largely located at the fundus, sclerotherapy for active bleeding usu- ally has had poor results. 16 Although bleeding can be stopped tem- porarily, rebleeding soon occurs. Therefore, it has been postulated that diagnosed splenic vein thrombosis caused by chronic pancre- atitis with massive bleeding should be treated with splenectomy, If the obstruction is confined to the splenic vein, splenectomy is the treatment of choice, a7 If a pancreatic pseudocyst is also found, cys- tojejunostomy or resection of pseudocyst is favored. 8 As a result of severe adhesions, splenectomy may be difficult and the risk of brisk bleeding during operation is high, as occurred in this case. To reduce perioperative bleeding, ligation of the splenic artery is recom- mended before mobilizing the spleen with a great number of collat- erals in the ligaments. In conclusion, in patients with variceal bleeding and chronic pan- creatitis, splenic vein thrombosis caused by inflammatory disease of the pancreas should be actively sought because this complication is one of the rare instances when the problem of portal hypertension can be treated satisfactorily by splenectomy. YEONG-L[ANa LXN, MD PEI-MING YANG, MD GUAN-TARN HUANG, MD TZONG-HsI LEE, MD SEH-HUANG CHAU, MD YvK-MIN6 TSANG, MD SHIH-MING WANG, MD National Taiwan University Hospital Taipei, Taiwan, ROC References 1. Rignault D, Mine J, Moire D: Splenoportographic changes in chronic pancreatitis. Surgery 1968;63:571-575 2. McDermott WV Jr: Portal hypertension secondary to pan- creatic disease. Ann Surg 1960;152:147-150 3. Little AG, Moossa AR: Gastrointestinal hemorrhage from left-sided portal hypertension: An unappreciated complication of pancreatitis. Am J Surg 1981;141:153-158 4. Harnar T, Johansen K, Haskey R, et al: Left-sided portal hypertension from pancreatic pseudotumor. Am J Gastroenterol 1982;77:639-641 5. Stone RT, Wilson SE, Passaro E: Gastric portal hyperten- sion. Am J Surg 1978;136:73-79 6. Warshaw AL, Jin G, Ottinger LW: Recognition and clinical implications of mesenteric and portal vein obstruction in chronic pancreatitis. Arch Surg 1987;122:410-417 7. McEIroy R, Christiansen PA: Hereditary pancreatitis in a kinship associated with portal vein thrombosis. Am J Med 1972; 52:228-241 8. Salam AA, Warren WD, Tyras DH: Splenic vein thrombo- sis: A diagnosable and curable form of portal hypertension. Sur- gery 1973;74:961-972 9. Mossa AR, Gadd MA: Isolated splenic vein thrombosis. World J Surg 1985;9:384-390 10. Madsen MS, Petersen TH, Sommer H: Segmental portal hypertension. Ann Surg 1986;204:72-77 11. Bradley EL II1: The natural history of splenic vein throm- bosis due to chronic pancreatitis: Indications for surgery. Int J Pancreatol 1987;2:87-92 12. Leger L, Lenriot JP, Lemaigre G: L'hypertension at la stase portales segmentaires dans les pancreatites chroniques. A propos de 126 cas examines par splenoportographie et sple- nomanometrie. J Chir 1968;95:599-608 13. Van Gansbeke D, Avni EF, Delcour C, et al: Sonographic features of portal vein thrombosis. AJR Am J Roentgenol 1985; 144:749-752 14. Vogelzang RL, Gore RM, Anschuetz SL, et al: Thrombosis of the splanchnic veins: CT diagnosis. A JR Am J Roentgenol 1988;150:93-96 15. Bernades P, Baetz A, Levy P, et al: Splenic and portal venous obstruction in chronic pancreatitis. A prospective longi- tudinal study of a medical-surgical series of 266 patients. Dig Dis Sci 1992;37:340-346 16. Gimson AES, Westaby D, Williams R: Endoscopic sclero- therapy and management of gastric variceal haemorrhage. Gut 1989;30:A1497 (abstr) 17. Longstreth GF, Newcomer AD, Green PA: Extrahepatic portal hypertension caused by chronic pancreatitis. Ann Intern Med 1971;75:903-908 NEBULIZED FENOTEROLVERSUS IV AMINOPHYLLINE TREATMENT OF ACUTE SEVEREASTHMA To the Editor.'--From July 1991 to June 1993, 30 patients visited the emergency department (ED) at National Taiwan University Hospital, where acute severe asthmatic attacks were investigated. Informed consent was obtained from all patients included in this clinical trial. All of these patients had forced expiratory volume in one second (FEV1) <30% of predicted value and peak expiratory