Cerebrovascular mortality in patients with pituitary adenoma

OBJECTIVE To assess cerebrovascular mortality in a UK cohort of patients with pituitary adenoma known to have increased incidence of cerebrovascular accidents (CVA). METHODS A total of 334 patients treated at the Royal Marsden Hospital (RMH) between 1962 and 1986 with surgery and postoperative radiotherapy were followed up via the NHS Central Register (NHSCR) to identify deaths and emigrations. The causes of death were assessed by NHSCR-based death certificates and coded according to the 9th revision of ICD. Follow-up was censored at age 85, on emigration or cancellation of NHSCR. Thirteen patients could not be traced. A total of 4982 person-years was accumulated in the cohort. Expected numbers of deaths were computed from the national age-, sex- and period-specific mortality rates for England and Wales. RESULTS In the pituitary adenoma cohort, 128 deaths were observed compared to 80.9 expected [relative risk (RR) of death 1.58 (95% CI: 1.32-1.90)]. There were 33 cerebrovascular deaths compared with 8.04 expected (RR 4.11, 95% CI 2.84-5.75). Three deaths were from subarachnoid haemorrhage compared to 0.54 expected (RR 5.51, 95% CI 1.14-16.09). There was an increased cerebrovascular mortality in women (RR 6.93, 95% CI 4.29-10.60) compared to men (RR 2.4, 95% CI 1.24-4.20; P = 0.002) and in patients having debulking surgery (RR 5.19, 95% CI 3.50-7.42) compared to biopsy/no surgery (RR 1.33, 95% CI 0.27-3.88; P = 0.02). The RR in patients with nonsecretory tumours was 3.65 (95% CI 2.26-5.58), compared with 5.23 (95% CI 2.25-10.30) in secretory tumours ( P = 0.4). The effect of age at radiotherapy was not significant ( P = 0.4). CONCLUSION Patients with pituitary adenoma treated with surgery and radiotherapy have an increased risk of cerebrovascular mortality compared to the general population, which mirrors the increased incidence of CVA. The possible risk factors include hypopituitarism, radiotherapy and extent of surgery but none are at present proven causes. The evaluation of new treatment strategies should not only assess intermediate end-points of tumour and endocrine control but should concentrate on long-term survival with particular emphasis on CVA incidence and mortality.