Thromboembolism (TE) is a well-known complication in neurosurgical patients and in patients with brain tumors. Estimates of postoperative TE range from 25 to 33% in general surgical patients and from 45 to 70% in patients who have undergone total hip replacement.1 The incidence of deep venous thrombosis (DVT) and pulmonary embolism in neurosurgical patients is similar to that in high-risk surgical patients. The incidence of DVT ranges from 9 to 50% and of fatal pulmonary embolism from 1.5 to 3%.2 The incidence of TE in patients with glioma varies from 36.7% in patients who receive no antithrombotic prophylaxis to 10% in those who receive intermittent pneumatic compression to the calves.3 The current study was undertaken to explore, in patients with high-grade glioma, the associations between TE and factors potentially related to an increased or decreased risk of occurrence of TE.

MATERIAL AND METHODS

Study Subjects.-In this retrospective study, we reviewed the records of 64 patients from the Mayo Clinic who participated in two prospective trials of single-agent nitrosourea chemotherapy and radiation therapy for newly diagnosed high-grade glioma. The first trial, a pilot study to assess the toxicity of carmustine [BCNU; l,3-bis(2-chloroethyl)-lnitrosourea] and hyperfractionated radiation therapy (180 cGy three times daily to a total of 6,000 cGy), enrolled 18 patients between April and August 1987. The second trial, a phase ÉÐ study of radiation therapy (6,000 cGy in 30 fractions) in combination with either BCNU or another nitrosourea derivative [PCNU; l-(2-chloroethyl)-3-(2,6- dioxo-3-piperidyl)-l-nitrosourea] accrued 46 Mayo patients from July 1985 to April 1988. For both studies, patients were 18 years of age or older and had histologically confirmed supratentorial grade 3 or 4 astrocytoma, mixed astrocytoma-oligodendroglioma, or gliosarcoma. The performance status was categorized from 0 to 3 (Eastern Cooperative Oncology Group). None of the patients had prior chemotherapy or other coexistent malignant disease. Surgical procedures included biopsy only, partial resection, or gross total resection.

Prophylaxis against TE was implemented at the discretion of the physician in charge of the patient. An episode of TE was diagnosed clinically by the physician responsible for the care of the patient and usually was confirmed by any of the following methods: impedance plethysmography, venography, duplex ultrasonography, ventilation-perfusion lung scanning, or pulmonary angiography.

Compilation of Data.-Data collected prospectively for each clinical trial included age, sex, performance status, histologie features and grade of the tumor, randomization date, and survival. Data obtained retrospectively were surgical dates, duration and extent of the surgical procedure, any operation within the 3 months before randomization, duration of anesthesia, duration of hospitalization, duration of bed rest, ambulatory status, history of smoking, use of aspirin, use of glucocorticoids, type of TE prophylaxis, prior TE, and presence or absence of diabetes, obesity, hypertension, chronic lung disease, paretic arm, or paretic leg. The medical records were reviewed for the absence or presence of TE, the number of days the symptoms were present before TE was diagnosed, the type of TE documented, the diagnostic method used, the type of treatment and the related side effects, and the cause of death.

Statistical Analysis.-In a limited, carefully monitored group of patients, we searched for clues about possible risk factors for development of TE. The distributions of all potential prognostic factors were compared between the patients with and those without TE by using ÷2 and Fisher's exact tests for nominal discrete variables, Wilcoxon tests for ordinal discrete and continuous variables, and log-rank tests for censored survival data. Survival distributions were estimated with Kaplan-Meier curves. Inasmuch as false-positive findings were highly likely because of multiple testing, P values between 0.01 and 0.05 were considered "interesting" but inconclusive. With only 18 TEs noted in 64 patients, the study had little power to detect distributional differences of clinical interest.

RESULTS

. . .Continue to read rest of article (PDF).