Higher prevalence of clinically relevant pituitary adenomas confirmed

Belgium and the UK are two distinct countries with modern healthcare systems that emphasize primary care (general practice) as the major interface between patients and clinicians. These countries are culturally and genetically diverse, both having mixtures of stable local populations in addition to continued national migrations and extensive influxes of new citizens during the recent past. As such, both represent appropriate settings for undertaking epidemiological studies in collaboration with general practices. In 2006, Daly et al.1 reported a cross-sectional, epidemiological study of the prevalence of clinically relevant pituitary adenomas in three sampling centres (urban, suburban and rural) in the Belgian province of Liège. Daly et al.’s study was performed via local general practices and demonstrated a prevalence of clinically relevant pituitary adenomas of one case per 1,064. This prevalence was 3–5 times higher than previous cross-sectional estimates in the UK and Italy.2,3 Since then, the question has remained as to whether this high prevalence was due to local factors or whether these results reflected a previously unrecognized high prevalence of pituitary adenomas in the community. In the current issue of Clinical Endocrinology, Fernandez et al.4 report the findings of a similar, cross-sectional, epidemiological study of pituitary adenoma prevalence in Banbury, Oxfordshire. They reported that 63 patients had pituitary adenomas among a population of 81,149, giving a prevalence of one case per 1,289. In this study, prolactinomas were most frequent, followed by nonfunctioning tumours, somatotropinomas, Cushing disease and tumours of unknown functional status. These results are generally and specifically in agreement with those of Daly et al. Any prevalence study of a single geographic region, no matter how intrinsically diverse the population, will always raise questions regarding clustering because of the involvement of local genetic or environmental factors. To date there is no evidence of geographic clustering of pituitary adenoma cases, even in sites of high environmental toxicity where an increase in tumour formation might be expected.5 Taken together, these two studies from Belgium and the UK argue in favour of the generalizability of these findings to wider populations. Estimates of the epidemiology of pituitary adenomas have been bedevilled by classification issues in cancer databases. While pituitary adenomas carry a heavy burden in terms of clinical effects and costly treatments, they are almost invariably benign by histology. For this reason, relevant data on incidence and mortality in patients with pituitary adenomas have not been compiled mandatorily in some cancer databases, which are largely focused on malignant tumours. This has led to pituitary adenomas being incompletely reported as a dataset and they may be grouped together with craniopharyngiomas or with benign and completely unrelated tumours of the central nervous system, such as meningiomas. Such incomplete information makes it difficult to estimate pituitary adenoma prevalence accurately. As healthcare and research resources are often allocated on the basis of estimates of population burden or costs of treatment, this study hinders arguments for expanded support of pituitary based research and therapy. It is difficult to make a strong argument for a greater proportion of available resources when one cannot compare the prevalence of pituitary adenomas with that of other endocrine or nonendocrine conditions that have accurate epidemiological backing. In addition, a lack of emphasis on pituitary adenomas may contribute to poorer awareness and late diagnosis, which can have therapeutic repercussions, particularly if tumours have become large or invasive. Extrapolation of the results of Fernandez et al. suggests that there are currently more than 47,000 individuals with clinically relevant pituitary adenomas in the UK. As Fernandez et al. note, the majority of patients are affected during their years of greatest economic activity [mean age at diagnosis: 37 (UK), 40 years (Belgium)]. A long mean time to diagnosis in both studies indicates the importance of improving awareness of pituitary adenomas in order to reduce symptom burden and facilitate early effective treatment. While the higher prevalence of pituitary adenomas demonstrated in Liège, Belgium and Banbury, UK is approximately four times higher than previously estimated, it does, in fact, still represent an underestimate. The true prevalence of clinically relevant pituitary adenomas would also include those that produce clinical effects but have not yet been diagnosed. Despite easier access to modern imaging and biochemical tests in a modern healthcare setting, even clinically important pituitary macroadenomas can remain undetected, particularly in the elderly.6 Therefore, it is correct to state that based on current data, the prevalence of clinically relevant pituitary adenomas is at least 1 in 1,064–1,289 of the population. Initial estimates of pituitary adenoma prevalence came from autopsy studies, and with the advent of modern imaging was followed by computed tomography and magnetic resonance imaging series. These studies were performed in unselected populations and subjects were not recorded as suffering from clinical symptoms. In daily practice, such adenomas are referred to as “incidentalomas”, which can represent a clinical conundrum. As imaging and pathology methods have improved, the size of incidentally found lesions has decreased. In 2006, Buurman and Saeger7 reported 334 pituitary adenomas in 3048 autopsy cases and the mean adenoma diameter was only 1·97 mm. It is important to note that neither the Belgian nor the UK study included such incidental tumours. Despite these concordant results, further work remains to be done on the epidemiology of pituitary adenomas. The Belgian and UK studies were both performed in northern Europe and do not address whether similar results will be seen in other geographically distant regions. Similarly, a more enriched and diverse genetic basis to a study population that includes significant numbers from the India or the Americas would be desirable. Finally, a larger number of sampling sites would significantly strengthen mean estimates that could then be extrapolated worldwide. To this end, a multinational collaborative epidemiological study has been completed this year and preliminary results support the findings from Liège and those in Banbury reported by Fernandez et al. in this issue of Clinical Endocrinology.