JC: Anterior pituitary disorder following head injury. St.Emlyn’s


For reasons which I can’t discuss here I’m quite interested in the incidence of pituitary axis dysfunction following head injury, and indeed in the outcomes following all severities of head injury that we see  in the ED. Over the years my perceptions have changed considerably. Thinking back to when I first started in medicine head injuries were divided into two groups.

Group 1. Serious stuff that needs to be dealt with ICU/Neurosurgery. These patients can die and or remain neurologically impaired.

Group 2. Everyone else who feels a bit unwell at the time but who then get completely better.

Looking back on this I do feel rather silly. We now know that the adverse consequences of head injury are not simply confined to those with the most serious initial injury. I think this paper by van der Naalt et al  first made me question those early beliefs (back in 1999) and there have been many more since. We now know that mild, moderate and severe head injuries can lead to short medium and long term morbidity. Common features include headaches, lethargy, anxiety, poor concentration, the list is long and in some studies also include sexual dysfunction. Such features are also found in anterior pituitary disorders (and lets face it the pituitary is in the head), so it would seem wise to look at pituitary function post head injury.

This month Critical Care Medicine publishes a meta-analysis on pituitary function post head injury. The abstract is below, but as always you should read the full paper if you have access.

Screen Shot 2014-04-27 at 15.57.16

This paper from Canada sought and then analysed 66 papers looking at pituitary function after head injury. Inclusion criteria for papers included cohort, RCT, observational studies that enrolled more than 5 patients critical care patients that assessed at least one element of pituitary function (axis). That’s quite a broad group of trials but this reflects the relative paucity of data in this area. Quite a range of studies were found and included with the majority being of a cross sectional and/or observational type. In total just over 5800 patients were included across the 66 trials.

What’s the incidence of pituitary dysfunction amongst critical care head injured patients?

This is the key question for me. Is this a rare event or something common enough that we should be actively looking for, and the simple answer is that it’s far more coming than I had thought. 66 of the studies looked at the incidence of pituitary disorder and the numbers are really very high indeed. In the short term just under half of patients have identifiable pituitary dysfunction, with an even higher reported incidence in higher quality papers. This was far higher than I had imagined and so we should know about this. This dysfunction extends beyond ICU with roughly a third of patients in these trials demonstrating dysfunction beyond 12 months.

Does it make a difference?

The ICU is a collection of patients with physiological dysfunction to there really should be no surprise to find endocrine dysfunction as a feature of ICU stay. What matters to clinicians is whether this dysfunction affects outcome. Sadly relatively few studies give mortality  or Glasgow Outcome Score data and the findings are not statistically significant, but the trend is for worse outcomes for those with identifiable pituitary disorder. So at a basic level an association between dysfunction and prognosis should make us thing, but we must also remember that association does not equate to causality and it may well be that something else affects both the pituitary axis and prognosis.

Any patients that we should be particularly concerned about?

27 papers looked at factors which might predict pituitary dysfunction and there were some associations with injury severity, presence of skull fracture and age, but if the incidence is as high as the data suggests I’m not sure that identifying subgroups for testing is that valuable. As a general principle if something is important (is it?) and prevalent (seems to be) then universal testing is usually the answer.

So should we look for a treat pituitary dysfunction?

This is an interesting paper, but it cannot answer this important question. Identifying the presence of disorder is a long way away from demonstrating the importance of surveillance or the effectiveness of intervention. Despite the large number of trials and the many patients with significant head injury these questions remained unanswered.

This is critical care though. What about EM patients?

Apart from stating the obvious that nearly all these patients will go through the ED on the way to ICU it’s important to note that the vast majority of head injuries are not admitted to the ICU. Whether pituitary dysfunction is a feature of moderate of mild brain injury remains unanswered by this paper although this is increasingly an area of research (and perhaps of a future blog post).

In summary this paper reinforces the increasing concerns about the endocrine axis post head injury, but it is a little disappointing that the answers I require as a clinician remain unanswered. In discussions with colleagues from intensive care it’s interesting to hear a variety of approaches to this within the ICU and the lack of formalised follow up testing of patients post discharge. We do know that many of out head injured patients develop clinical features of that are similar to those experienced by hypopituitary patients. The importance of this, the incidence amongst moderate and mildly injured patients and the effectiveness of therapy remains uncertain.

I’d love to hear more from around the world on whether anyone is already screening and treating, and on what evidence their practice is based.



You might also consider the following FCEM  style questions after looking at this paper. Answers and comments below please 🙂

1. What’s the difference between a Systematic Review and a Meta-analysis?

2. In Figure 4  the incidence of pituitary disorder at 12 months amongst 27 papers is given as 31.6%. Comment on how the analysis helps us understand how this results findings to be expected in a wider population.

Translate »