Aims:This study: a)revisits the effectiveness of preoperative embolization alone for hypervascular lesions of the thoracolumbar spine and b) compares its action with intraoperative cryotherapy alone or in conjunction with embolization. Methods:14 patients underwent 15 surgeries for hypervascular spinal tumors. Ten of the surgeries were augmented by preoperative embolization alone. Four surgeries involved intraoperative cryocoagulation, and one surgery used a combination of preoperative embolization and intraoperative cryocoagulation for tumor resection. When cryocoagulation was used, its extend was controlled by intraoperative ultrasound or by establishing physical separation of the spinal cord from the tumor. Results:Among cases treated with embolization alone, 50% still had intraoperative blood-loss in excess of 3 liters. Mean blood-loss was of 2.8 liters/patient. One patient bled excessively (over 8000ml) terminating the surgery prematurely and resulted in suboptimal tumor resection. All procedures using cryo-coagulation achieved adequate hemostasis with average blood-loss of only 500 ml/patient by far better than embolization (P< 000.1). Conclusions:Preoperative embolization alone may not always be satisfactory for surgery of hypervascular tumors of the thoracolumbar spine. Although experience with cryocoagulation is limited, its use, with or without embolization, suggests its effectiveness in limiting blood-loss. Cryocoagulation may also assist resection by preventing tumor spillage, facilitating more radical excision of the tumor and enabling spinal reconstruction that eventually may contribute to improve survival. The extent of cryocoagulation could be controlled adequately with ultrasound. Somatosensory evoked potentials may provide early warning of cord cooling. No new neurological deþcits were attributable to the use of cryocoagulation.

Purpose: To describe the percutaneous transpedicular biopsy technique as a novel way of approaching lesion of the thoracic and lumbar spine, to determine the amount of bone retrievable through the pedicle and its diagnostic yield. Material and Methods: Seventy-nine patients underwent 84 biopsies. Seventy-seven procedures were performed with ßuoroscopic guidance arid seven with CT guidance. Seventy-one biopsies underwent under local anesthesia and ten under general anesthesia. Age range of patient was from 3 to 81 years. Results: Adequate specimens for correct diagnosis were obtained in 80 of the 84 patients with the following diagnoses. Pyogenic spondylodiscitis 31, tuberculosis 4, coccidiomycosis 2, echinococcus cyst 1, blastomycosis 1, brucella 4, primary neoplasm 7, metastatic neoplasms 16, osteoporotic fractures 8, osseous repair for insufþciency fractures 5, Pagetñs disease 1. The 4 negative biopsies subsequently proven to be Ç false negative È and were related to faulty biopsy techniques. Conclusion: Pitfalls can be avoided when adhering to the details of our technique. These pitfalls can occur while retrieving the instrumentation without simultaneous withdrawal of the guiding pin; crushing pathological soft tissue against sclerotic or normal bone; or when encountering a sclerotic lesion distal to normal bone without using a sequential type of biopsy specimen-retrieval technique. Any type of bleeding is controllable. The approach is a safe, efþcacious and cost effective and avoids so the problems such low diagnostic yield nerve root injury, pneumothorax and hematoma encountered with conventional needle technique