One of the important stages during splenectomy is the search for accessory spleens. The average frequency of accessory spleens is an estimated 11%. In our own material in 611 patients who had splenectomy in hematological disorders, accessory spleens were found in 110 patients (18%). The surgical anatomy, embryology, and clinical significance of the accessory spleens, with particular reference to the recurrence of idiopathic thrombocytopenic purpura (ITP) are discussed. The group of 177 patients with ITP treated by splenectomy 2–25 years ago is analyzed. Recurrence of thrombocytopenia after a good immediate response to splenectomy developed in 52 patients. In 13 patients the recurrence was transient, but in 19 it was longlasting. In these patients the presence of the accessory spleens was suspected by the absence of Howell‐Jolly bodies in the peripheral blood. The accessory spleen was demonstrated by splenic scan in 16 cases and confirmed in 13 of the 15 patients who underwent reoperation. Removal of the accessory resulted in permanent disappearance of thrombocytopenia in 12 patients.

1. A left-sided, paravertebral, hour-glass tumour causing destruction of the neural arches of the third and fourth thoracic vertebrae with evidence of spinal cord compression, is described. The tumour presented the typical histological appearance of a chondroblastoma. 2. The intraspinal part of the tumour was excised and the mediastinal part curetted. Post-operative radiotherapy was given. The patient was symptomless two years after operation. 3. No example of Codman's tumour with similar features and in such a situation has been described before in the literature. Pathological, clinical and radiological aspects of chondroblastomata are briefly discussed and some remarks concerning their treatment are added.

The aims of this prospective study were to determine the patterns of gastrointestinal (GI) bleeding in hemophiliacs and to assess the hemostatic effect of injection therapy with alcohol. During a 5-year period (1990-1994) 89 hemophiliacs were admitted to our department with acute GI bleeding. Among these patients duodenal ulcer was found endoscopically to be the most common (42.7%) cause of hemorrhage; gastric ulcer was the source of the bleeding in only three patients (3.4%). A group of 46 patients met the criteria of active or recent bleeding and underwent injection therapy with alcohol. The injected bleeding lesions were duodenal ulcer in 32 patients, duodenal erosion in 2, gastric ulcer in 3, and other gastric lesions (Mallory-Weiss tear, Dieulafoy lesion, stomal ulcer, erosions) in 9 patients. Initial hemostasis was achieved in 100% and permanent hemostasis in 82.6%. Rebleeding was observed in eight patients (17.4%), with five of them successfully treated by reinjections. Three patients (6.5%) required emergency surgery. The mortality rate in the group of injected patients was 2.2%. One patient died of stroke on day 10 after partial gastrectomy. All injected patients were given replacement therapy with factor VIII or IX for 2 days (29 patients) or 7 to 14 days (17 patients). Analysis of the hemostatic effect achieved in these two subgroups indicate that short-term replacement therapy (2 days) may be sufficient to ensure adequate hemostasis in hemophiliacs. The results of the present study indicate that injection therapy with alcohol is an effective, safe, proved method to control GI bleeding in hemophiliacs.

In the last 20 years, surgery in patients with hemophilia has been carried out with progressively less mortality and morbidity. Even extensive and major surgical operations can be carried out provided that the underlying coagulopathy is corrected and normal hemostasis restored. We report the results of 196 operations performed in 147 hemophiliacs during the period 1961–1985 in the Department of Surgery, Warsaw Institute of Hematology. Our patients experienced, in this 25‐year period, 23% of hemorrhagic complications in hemophilia A and 12.5% in hemophilia B. In the last 5 years, however, the incidence of hemorrhagic complications was reduced to 5.2% in hemophilia A and 8.3% in hemophilia B. This correlates with the discovery of cryoprecipitate and with the development of concentrates of factor VIII and IX, respectively. In all patients, deficient factor levels were determined before the operation. The replacement therapy was based on the policy of raising the level of factor VIII or IX above 60% of the normal value and maintaining this level above 30% in the postoperative period until final wound healing. Problems discussed in this article include the principles of replacement therapy, its advantages and disadvantages, and the organization and degree of cooperation among surgeon, hematologist, and blood bank staff.

From 1965 to 1985, 64 deep vein thrombosis (DVT) patients were treated with streptokinase (SK). In 26 cases 'high-dose SK' (IV 100,000 units/h for 4 days) was used and in 38 patients a 'low-dose SK' regime (IV 250,000 units every 12 h for 4 days) was employed. The clinical signs of DVT subsided in 78 per cent of treated patients within 30 days of completing SK treatment. A repeat phlebography was performed immediately after SK therapy in 29 patients (45 per cent) and a total recanalization or partial thrombolysis was achieved in 80 per cent of the studied cases. In 15 patients minor and major haemorrhagic complications occurred. There were five fatalities, all in the high-dose SK group (three intracranial haemorrhages and two major bleeds). Three patients developed pulmonary embolism and none of them died. The post-treatment clinical and phlebographic evaluation did not reveal any significant difference between the two methods of SK administration, but more haemorrhagic complications (P less than 0.02, chi=5.50825) occurred in the high-dose SK patients. This report emphasizes the risk of bleeding complications during thrombolytic therapy. If SK is to be used, therefore, careful selection of patients and meticulous monitoring are mandatory.