Reduced cardiovascular mortality in hemophilia despite normal atherosclerotic load


Mike Makris, Sheffield Haemophilia and Thrombosis Centre, Royal Hallamshire Hospital, Glossop Road, Sheffield S10 2JF, UK.
Tel.: +44 114 271 2760; fax: +44 114 275 6126.


See also Zwiers M, Lefrandt JD, Mulder DJ, Smit AJ, Gans ROB, Vliegenthart R, Brands-Nijenhuis AVM, Kluin-Nelemans JC, Meijer K. Coronary artery calcification score and carotid intima–media thickness in patients with hemophilia. This issue, pp 23–9; Biere-Rafi S, Tuinenburg A, Haak BW, Peters M, Huijgen R, de Groot E, Verhamme P, Peerlinck K, Visseren FLJ, Kruip MJHA, Laros-van Gorkom BAP, Gerdes VEA, Buller HR, Schutgens REG, Kamphuisen PW. Factor VIII deficiency does not protect against atherosclerosis. This issue, pp 30–7.

It is a testament to the success of hemophilia treatment that this editorial is being written, as 100 years ago the median life expectancy of a person with severe hemophilia (PWSH) was less than 12 years of age [1]. The wide availability of safe clotting factor concentrates led to a significant increase in life expectancy and the latest UK data showed PWSH to have a median survival of 63 years whereas patients with mild hemophilia had a median survival of 75 years [2]. As persons with hemophilia are getting older, we are now seeing the comorbidities that affect the general population also occurring in hemophilia. The issue of cardiovascular disease in hemophilia has been increasing in prominence in the last 5–10 years.

A number of studies have examined cardiovascular mortality in PWH or hemophilia carriers [3] and the general conclusion is that there is reduced cardiovascular mortality. Recently, Biere-Rafi et al. performed a systematic review and reported a reduced but not significant overall standardized mortality ratio (SMR) in the published studies of 0.51 (95% confidence interval [CI] of 0.24–1.09). When analysis was restricted to studies with complete follow-up for a period of 10 years or more the SMR was 0.59 with 95% CI of 0.48–0.72 [4]. Thus, whilst the literature supports the view that there is reduced mortality from cardiovascular causes in patients with reduced levels of factor (F) VIII and FIX, most studies reported a relatively small number of events and did not specifically exclude patients with HIV infection, a well-known risk factor for cardiovascular disease.

Only a single study assessed the frequency of non-fatal cardiovascular events. Kulkarni et al. used hospital discharge diagnosis to compare 3422 hemophiliacs with the normal US male population. They reported 79 episodes of ischemic heart disease in 48 patients, which were strongly age related, with 0.05% of those aged < 18 years experiencing an event, 1.7% in the 40–49-year age group and 15.2% in those aged over 60 years [5]. The traditional risk factors for ischemic heart disease in the general population, such as hypertension, diabetes mellitus and hyperlipidemia, were also found to be risk factors for IHD in hemophilia [5].

Although the reduced cardiovascular mortality and morbidity in hemophilia is generally accepted, the reasons behind it are unclear. Possible factors include:

  • 1 reduced risk factors for cardiovascular disease;
  • 2 reduced atherosclerosis;
  • 3 more stable atherosclerotic plaques; and
  • 4 reduced occlusive thrombus formation at sites of plaque rupture.

At least five studies investigated the association between risk factors and the development of atherosclerotic cardiac disease in hemophilia and all concluded that these are present to the same degree as in non-hemophilic individuals and cannot be the explanation for the reduced mortality rate [5–9].

The frequency of asymptomatic atheroma was investigated using a number of techniques, including carotid and femoral artery intima measurements, endothelial vascular dilation and plague density in larger vessels. With the exception of three studies from the same group in Italy [10–12], the other studies have suggested that the risk of atherosclerosis is the same in patients with hemophilia and control subjects [4].

Two new studies examining the incidence of atherosclerosis in hemophilia are reported in this issue of the journal. In the first, Zwiers et al. [13] report the coronary artery calcification scores (CACS) and carotid intima media thickness (CIMT) of 69 patients over the age of 30 years with all severities of hemophilia attending the Groningen hemophilia centre. CACS is currently the only non-invasive method available to measure coronary atherosclerosis and this was performed by electron beam or dual source computed tomography, the methods having been previously shown to be equivalent [14]. This is the first time that use of this technique has been reported in hemophilia. The CACS scores of the hemophilic individuals were no different from those of a large international reference range of ethnically similar individuals, which seems appropriate because CACS is standardized worldwide [15]. The CIMT scores were also within the local reference range, which is based on normal individuals previously studied in Groningen. In contrast to CACS, there is no consensus on how to use the IMT technique so it is important that the results are compared with locally measured control subjects where exactly the same technique is used. The study concludes that hemophiliacs have the same level of atherosclerosis as control individuals. The findings of this study are strengthened by the demonstration that hemophiliacs with previous cardiovascular events had elevated CACS and IMT scores and also by the fact that both of these scores were correlated with the Systematic Coronary Risk Evaluation (SCORE) value adjusted for the Dutch population [16].

In the second study, from the Netherlands and Belgium, Biere-Rafi et al. [17] report the CIMT, femoral intima media thickness (FIMT) and brachial flow mediated dilation (FMD) in 51 obese (body mass index [BMI] ≥ 30 kg m−2) and 47 non-obese (BMI ≤ 25 kg m−2) hemophiliacs aged over 18 years of all severities. Control subjects were recruited appropriately for this study and were matched to the hemophiliacs for BMI, age and gender. Although CIMT and FIMT were increased in obese hemophiliacs and control subjects compared with non-obese individuals, there was no difference between hemophiliacs and control subjects when matched for BMI. The FMD data were no different between hemophiliacs and control subjects or between obese and non-obese individuals. The authors also looked at the incidence of carotid plaques and these were found in 35% of obese hemophiliacs and 29% of the obese control group [17]. Both of these publications support the previously proposed concept of normal levels of atherosclerosis in hemophilia. It must be appreciated, however, that both studies were relatively small and recruited mild as well as severe hemophiliacs and the lower ages for recruitment were relatively low at 18 and 30 years, respectively. A better study would have included only severe hemophiliacs aged over 60 years without HIV or chronic hepatitis C but this would have been very difficult to do due to the rarity of suitable patients.

Although studies of the rate of atherosclerosis are small and can be criticized on several levels, the overall message is convincing: there is no difference in the rate of atheroma between hemophiliacs and control individuals. Animal studies support this observation. In the ApoE and FVIII double knockout mice there is retardation but not prevention of atherosclerosis and, in the model believed to more closely mimic the human situation, the LDLR/FVIII double knockout mice had the same degree of atherosclerosis as the LDLR single knockout mice [18,19].

If cardiovascular risk factors and rates of atheroma are the same in hemophiliacs and control individuals what could be causing the reduced mortality? A difference in atherosclerotic plaque stability is a possibility but this has not been studied. A more attractive alternative is the limitation of thrombus size and thus vessel occlusion in hemophiliacs once a plaque has ruptured. Thrombus size and structure depend on thrombin generation, which in turn is critically dependent on the concentration of FVIII and FIX. Reduced thrombus generation is easily demonstrated by thrombography, thromboelastometry and thromboelastography as well as in flow chamber and animal studies [20].

The importance of identifying the reason for the reduced cardiovascular mortality in hemophilia is far from being simply academic. The age of the hemophilic population is increasing and even with the reduced rate, non-fatal cardiovascular events are likely to increase. There is little published evidence on the management of these events when they occur and better understanding of the situation is likely to lead to more directed and logical therapy.

Disclosure of Conflict of Interests

The authors state that they have no conflict of interest.