Not Tonight, Darling, I Might Get a Headache
See related article, pages 1878–1882
Advances in neuroimaging and an increase in the number of neuroimaging procedures performed have created a new management dilemma. Unruptured intracranial aneurysms (UIA) are found in ≈2% of people undergoing digital subtraction angiography1 and in 0.4% of people undergoing an MRI.2 The best risk management strategy in these patients is not known.
Such patients are at increased risk for aneurysmal subarachnoid hemorrhage (SAH), but to know how they might best be managed, we need to understand more about the natural history of UIA, about factors that might provoke rupture, and about the risks and potential successes of intervention. In recent years, the Utrecht group,3 whose most recent work appears in this issue of Stroke, has contributed greatly to our understanding of these issues.
The best estimate of the risk of rupture probably comes from the prospective arm of International Study of Unruptured Intracranial Aneurysms (ISUIA),4 which suggests an annual rupture risk of ≈0.1% for aneurysms <7 mm in diameter and ≈2% for larger aneurysms, with this risk being mostly concentrated in those with large aneurysms in the posterior circulation. Modeling suggests that risk increases in proportion to the third power of aneurysm diameter. Importantly, <1 in 5 asymptomatic aneurysms have a diameter >10 mm and most are in the anterior circulation.
Paradoxically, despite this apparently very low rupture risk, most aneurysmal SAH appear to be attributable to rupture of small anterior circulation aneurysms. The most rational explanation, proposed by Wiebers et al5 and strongly supported by detailed mathematical modeling,6 is that most SAH arises from recently formed rather than long-standing aneurysms, although it is long-standing aneurysms that are included in studies of UIA rupture risk. Mitchell and Jakubowski6 suggested that for small (<10 mm) aneurysms in patients without previous SAH, the risk of rupture is maximal in the first 42 weeks after formation and very low thereafter. For larger aneurysms, they found that rupture risk was constant over time.6 This is clearly important when considering the possible benefits of treatment and of modifying behavior in people with UIA.
What, then, are the risks and potential benefits of treatment? Endovascular treatment is probably safer than coiling, and a recent systematic review7 found a procedural risk of death of 1.2% and of unfavorable outcome of 4.8%. A satisfactory angiographic result was obtained in 86.1%, but 24.2% of patients had recurrence and 9.1% underwent a repeat procedure; 75% of the aneurysms included in this systematic review were <10 mm in diameter.
Although of course these largely observational studies may have been enriched with patients considered to be at high risk for bleeding, the limited follow-up is likely to have led to under-ascertainment of complications. More worrying is that Naggara et al7 were concerned at the methodological quality of many included coiling studies, including the short duration of follow-up and the paucity of randomized evidence. With these caveats, they estimated the annual risk of bleeding after UIA coiling was still 0.2%.7 That is, the risks of hemorrhage after an endovascular procedure were not substantially different than that suggested by longitudinal studies for untreated aneurysms. Unfortunately, a large randomized control trial testing the efficacy of endovascular treatment of unruptured aneurysms has been suspended because of slow recruitment.
It therefore seems reasonable to suggest that for the nearly 2% of the adult population with small unruptured anterior circulation aneurysms, the best advice is to avoid treatment unless there is a good alternative reason to intervene. However, in these circumstances, patients often become somewhat anxious about their health and the prospects that they might have a “time bomb” in their head, ready to explode at a moment's notice. Although few studies exist, it is a common clinical observation that such patients are as much at risk from the effects of health-related anxiety and depression attributable to the knowledge of their potential risk as they are from the effects of SAH if it were to occur. In this context, patients often ask whether they should change their lifestyle or behaviors to minimize the risk of aneurysm rupture. Should we perhaps advise them, considering the findings of Vlak et al,3 to avoid many routine activities, including caffeine consumption, nose-blowing, sexual intercourse, or other vigorous physical exercise?
First, are their findings valid? The case-crossover design is appropriate, but there remain potential risks of incorporation bias and recall bias. Their study describes only 54% of potentially eligible patients, and women were substantially over-represented. The authors acknowledge this risk of incorporation bias, but it is not clear that their efforts to address this (review of routine medical records for details of activities at time of onset of SAH in included and excluded patients) were sufficiently robust to have overcome this risk of bias.
There are also some hints that recall bias may be present. For instance, in an urban Northern European population of whom >60% are current smokers, it seems unlikely that 20% had not coughed, sneezed, or blown their nose in the year preceding their SAH. It is always a great pleasure to collaborate with the Dutch, who seem to be able to take everything in stride with equanimity; however, it is slightly implausible, even for the Netherlands, that ≈60% of patients had not experienced anger over the course of 1 year. These observations are important because if, for instance, anger in the preceding year was differentially recalled by those experiencing anger immediately before their SAH, then this would have the effect of magnifying the relationship between anger and SAH onset and would inflate the observed risk.
A further problem relates to the difficulty in demonstrating causality rather than simple association. Caffeine is a common component of many over-the-counter headache preparations, and many of us reach for a strong coffee at the onset of “fuzzy headedness” or minor headache. The association between coffee intake and SAH therefore may be one of reverse causality, in which an unidentified nonspecific prodromal symptomatology invokes caffeine intake and therefore the appearance of association with SAH.
Second, even if these findings are valid, are they of sufficient clinical significance for us to recommend that people with known incidental asymptomatic aneurysms desist from certain behaviors? The answer to this question lies in the authors' own reluctance to recommend that people avoid strenuous physical activity or sexual intercourse, even though they report a combined population-attributable risk of 12.2% from these activities. When looked at from the perspective of those who have experienced SAH, the risk appears high; however, because the absolute risk of SAH is so low, the clinical benefits in those with unruptured aneurysms are at best marginal and are probably offset by the reduction in quality and quantity of life attributable to the avoidance of physical exercise and sexual intercourse.
In fact, it is possible to combine data for population-attributable risk and the frequency of various activities provided by the authors with the estimates of the risk of UIA rupture outlined to calculate the number of exposures an at-risk individual needs to avoid to prevent 1 SAH. People at high risk (aneurysms of >7 mm) need to avoid ≈670 000 cups of coffee or 64 000 episodes of sexual intercourse to prevent 1 SAH. People at low risk (aneurysms of ≤7 mm) need to avoid 13.4 million cups of coffee or 1.3 million episodes of sexual intercourse. On the basis of their prevalence figures, this represents the avoidance of 20 000 years of sexual activity to avoid 1 SAH.
So, even if the data presented do accurately describe the attributable risk of certain factors in provoking SAH in patients with intracranial aneurysm, the absolute attributable risks in the at-risk population are trivial and for most patients are far outweighed by the lifestyle consequences of avoiding such activity. If the absolute risk of rupture is sufficiently high to warrant such action, then the medical imperative should be definitive treatment of the aneurysm. It is perhaps reasonable that in the few hours before such intervention patients should avoid sexual intercourse, strong coffee, heavy lifting, and straining at stool, but this advice probably already is within the compass of good clinical practice.
The opinions expressed in this editorial are not necessarily those of the editors or of the American Heart Association.
- © 2011 American Heart Association, Inc.
- de Rooij NK,
- Linn FH,
- van der Plas JA,
- Algra A,
- Rinkel GJ
- Morris Z,
- Whiteley WN,
- Longstreth WT Jr.,
- Weber F,
- Lee YC,
- Tsushima Y,
- et al
- Vlak MHM,
- Rinkel GJE,
- Greebe P,
- van der Bom JG,
- Algra A
- Mitchell P,
- Jakubowski J