This analysis is based on 4457 patients (2365 males, 2092 females; mean age 44 years, range 18-83) with severe GH deficiency (GHD) enrolled in KIMS (pharmaco-epidemiological survey of GH replacement in adult GH-deficient patients, sponsored by Pharmacia Corporation). The mean duration of adult GH replacement was 2.8 years (range 0.1-11.0) giving 10,779 patient years of follow up. The baseline prevalence of diabetes mellitus prior to GH replacement was 3.7 % (compared with population prevalences of 2.3 % in the Wickham, UK survey and 2.8 % in Skaraborg County, Sweden). The standardised incidence ratio (SIR) for diabetes mellitus during GH replacement was <1.0 in all patients with a body mass index (BMI) of <30 kg/m2, 3.6 for patients with BMI 30-35 and 4.8 with BMI >35. The SIR for de novo neoplasia derived from country adjusted background rates was not significantly increased for colorectal, breast, bladder, lung, prostate and uterine cancer or for lymphoma and leukaemia. The SIR for de novo intracranial neoplasia was increased to 9.5 (95 % confidence intervals 4.2-12). 73 deaths occurred giving a standardised mortality ratio of 1.2 (95 % CI 0.92-1.97) and 0.8 when deaths in patients with a history of Cushing's disease, acromegaly and malignant disease were excluded. Prevalence of diabetes mellitus is increased in this population prior to GH replacement and subsequent incidence is influenced by BMI but not by GH therapy. The apparent increased incidence of intracranial neoplasia may be an artefact of comparing a surveillance population with general population data and may also reflect the increased incidence of second intracranial tumours in patients with a history of pituitary tumour. Mortality rates are, thus far, similar to those predicted for country adjusted normal populations.
03 - 04 Dec 2001
Society for Endocrinology