Introduction
In 1991, an ad hoc committee appointed by the Society for Vascular Surgery and the
International Society for Cardiovascular Surgery defined an aneurysm as a localized
dilation of an artery that increases the diameter of the blood vessel by ≥ 50%[1]. Ruptured aortic aneurysms are the 13th-leading cause of death in the United States,
estimated to cause ≈15,000 American deaths per year[2]. Furthermore, because ruptures carry a 90% mortality rate[3], treatments to reduce the occurrence of aortic aneurysms are highly valued in the
medical community.
People with Marfan syndrome, a genetic disorder caused by mutations in the gene that
encodes fibrillin-1, are particularly susceptible to aortic aneurysms. The disorder
is estimated to affect 1 in 5,000 people and is clinically recognized by tall stature
and long appendages, which are the results of bone overgrowth[4]. Because Marfan syndrome may also stretch and weaken the aorta, 70% of Marfan syndrome
deaths arise as a result of cardiovascular complications[5]. Furthermore, people with Marfan syndrome have an annual death risk between 1% and
2%[5]. In an effort to improve the lifespan of those with this genetic disorder, studies
on aortic aneurysms have often centered on Marfan syndrome.
Fortunately, the mortality rates from aortic aneurysms have declined steadily since
1990 because of the introduction of various therapies[6]
[7]. β-Blockers in particular have frequently been used to treat aortic aneurysms since
a small 70-patient study published in 1994 investigated their effects on patients
with Marfan syndrome[8]. The results of that study and several others, all of which were performed on patients
with Marfan syndrome, suggested that β-blockers could lower rates of aortic dilation
and mortality[5]
[8]
[9]. Intuitively, the theoretical reasoning behind β-blocker usage seems logical: The
drugs slow the heart rate and reduce arterial pressure, thereby decreasing stress
on the aorta that could rupture an aneurysm[10].
Soon after the 1994 study by Shores et al.[8], β-blockers began to be prescribed to treat aortic aneurysms in the general populace,
not just patients with Marfan syndrome. Today, β-blockers have become the standard
treatment for small aortic aneurysms, despite an inadequate number of published studies
to support β-blocker use for aortic aneurysms in patients without Marfan syndrome.
These drugs, now considered the “gold standard” for aortic aneurysm treatment[11], carry the heavy responsibility of preventing approximately 15,000 annual deaths
in the United States, yet they remain unproven in clinical trials—and, as we shall
see, the supporting evidence is unconvincing at best.
This review analyzes the scientific basis for the use of β-blockers to prevent aortic
aneurysms and questions whether this treatment is truly acceptable without conclusive
evidence to substantiate its use. Just as other medical therapies closely vetted by
the U.S. Food and Drug Administration require rigorous laboratory studies and clinical
trials to ensure the safety and efficacy of medicinal products[12], β-blocker treatment and its role in aortic expansion need to be further delineated
by prospective, randomized trials. This review seeks to draw attention to the limited
evidence for β-blocker use for aortic aneurysms and encourages more critical investigations
to prove that the therapy truly benefits this cohort of patients.
Literature Review
The study most frequently cited to demonstrate that β-blockers reduce the risk of
aortic aneurysms was published in 1994[8]. In that landmark study, Shores et al. divided 70 patients with Marfan syndrome
into a control group of 38 patients who received no treatment and a treatment group
of 32 patients who received propranolol. The authors defined an aortic ratio: the slope of the regression line for the increase in aortic dimensions over time.
The aortic ratio of the control group was 0.084 per year, whereas the aortic ratio
of the treatment group was only 0.023 per year. Five patients in the treatment group,
two of whom did not follow the propranolol regimen, and nine patients in the control
group reached a composite clinical end point, which was defined as heart failure,
aortic dissection, cardiovascular surgery, or death[8]. The authors contended that their results, summarized in [Table 1], supported the use of β-blockers, propranolol specifically, in patients with Marfan
syndrome to treat aortic aneurysms on two grounds: first, aortic dilation was faster
for patients in the control group than for the treatment group, and second, more patients
in the control group reached the composite clinical end point than in the treatment
group[8].
Table 1.
Comparison of the Effects of β-Blockers on Patients with Marfan Syndrome in the 1994
Experiment by Shores et al.[8]
|
Control group
|
Treatment group
|
|
Patients
|
38
|
32
|
|
Aortic ratio
|
0.084
|
0.023
|
|
Clinical end points
|
9
|
5
|
n=70 patients.
The construction of a composite end point was necessary because no single clinical
end point reached statistical significance on its own merit. Although the results
were certainly promising, the authors concede that the study was neither placebo-controlled
nor blind, with every patient and investigator aware of the patient's group. Thus,
although the results did show potential for β-blockers in aneurysm treatment, it is
highly possible that the study's results were subject to bias and a placebo effect.
Furthermore, although heart failure, dissection, and death are hard end points, the
decision for surgery is a softer call and might have been influenced. The study also
did not have a definitive means of ensuring patient compliance; the patients in the
treatment group may not have followed the correct propranolol dosage, and patients
in the control group may have taken other medications.
The largest limitation of the study, however, was the small sample size. By the end
of the trial, the already minimal population decreased by 20% because of clinical
end points. Although the authors appropriately believed the presence of more end points
in the control group supported their conclusions, a mere four-person difference between
the control group and treatment group seems unconvincing, even more so when one takes
into account that two of the deaths in the control group were unrelated to nonaortic
complications (mitral valve prolapse and Wolff-Parkinson-White syndrome). Additionally,
the control group was larger than the treatment group by more than 15%, so more clinical
end points in the control group should have been expected. Lastly, because patient
compliance could not be monitored effectively, the interpretation of the results would
need to be tempered.
Expanded Clinical Trials on Patients with Marfan Syndrome
Further studies examining the clinical benefit of β-blockers on aortic aneurysms in
patients with Marfan syndrome have done little to clarify the picture. Although supporting
data certainly exist in the literature, the lack of definitive evidence has been demonstrated
through a meta-analysis of six studies conducted by Gersony et al.[13], which concluded that no significant improvement in β-blocker treatment groups could
be found compared with control groups.
One study in 1995, however, did concur with the findings of Shores et al. Over the
course of 25 years, Silverman et al.[5] compared the health of 191 patients taking β-blockers with 142 patients who had
never taken β-blockers and 84 patients whose β-blocker usage was unknown. The treatment
group took various types of β-blockers, with atenolol being the most commonly used,
followed by nadolol, propranolol, and metoprolol. By the end of the study, there were
8 deaths and 58 operations in the treatment group, whereas there were 39 deaths and
54 operations among the rest of the observed patients. Although the number of operations
was comparable between the groups, the low number of deaths in the β-blocker treatment
group was significant. Furthermore, the study noted that the life expectancy of those
taking β-blockers was 72 years, whereas the life expectancy of those who had never
taken β-blockers was 70 years (P = 0.01)[5].
Another study by Salim et al.[9] also agreed with the results of the study by Shores et al.[8] and concluded that β-blockers should be used at young ages to slow aortic root dilation.
Between 1979 and 1992, 113 patients < 21 years of age were divided into a treatment
group of 100 and a control group of 13. The study found that patients in the treatment
group had an aortic root growth rate of 1.0 mm per year, whereas patients in the control
group had an aortic root growth rate of 2.1 mm per year[9]. The limited number of patients in the control group compared with the treatment
group, however, makes it difficult to lend credence to the comparison.
The results of the studies by Silverman et al.[5] and Salim et al.[9] are promising; however, these studies do not provide enough data to promote the
use of β-blockers for aortic aneurysm treatment, particularly in light of other studies
that provide conflicting findings. For example, in a study of 113 patients, Roman
et al.[14] found that patients taking β-blockers and patients not taking β-blockers had similar
aortic complication rates, with 33% of the treatment group and 30% of the control
group having complications[14]. This study is difficult to analyze, however, because it was not specifically designed
to address β-blocker treatment in patients with Marfan syndrome.
A paper published by Legget et al.[15] in 1996 concluded that no significant difference existed between the β-blocker treatment
group of 28 patients and a control group of 55 patients. In fact, with clinical end
points defined as death or surgery for ascending aortic aneurysms, the treatment group
reached 9 negative end points, whereas the control group achieved only 8 negative
end points over 5 years[15]. Unfortunately, this study also had a small sample size and did not focus solely
on the effects of β-blocker treatment on aortic aneurysms.
For the reasons indicated above, these extended clinical studies, detailed in [Table 2], have limitations and are simply not conclusive or convincing in their support of
β-blocker treatment for aneurysm disease. A call for a large, multicenter, prospective
placebo-controlled trial is needed.
Table 2.
Comparison of the Effects of β-Blockers on Patients with Marfan Syndrome in Five Separate
Clinical Trials
|
Control group
|
Treatment group
|
|
Patients
|
End points
|
Patients
|
End points
|
|
Shores et al.[8]
|
32
|
5
|
38
|
9
|
|
Silverman et al.[5]
|
226
|
93
|
191
|
64
|
|
Salim et al.[9]
|
13
|
0
|
100
|
5
|
|
Roman et al.[14]
|
—
|
—
|
—
|
—
|
|
Legget et al.[15]
|
55
|
5
|
28
|
9
|
|
Total
|
326
|
103
|
357
|
87
|
Randomized Clinical Trials of Patients without Marfan Syndrome
Interestingly, despite the fact that the mixed literature on whether β-blockers actually
help prevent aortic aneurysms in Marfan syndrome remains decidedly inconclusive[16]
[17], β-blockers appear more commonly used to treat both thoracic aortic aneurysms and
abdominal aortic aneurysms in the general aneurysm population as well. When these
trials have been performed on patients without Marfan syndrome and with abdominal
aortic aneurysms, β-blockers have failed to consistently decrease the growth rate
of the aneurysms[18]
[19].
In a 2002 study published in The Journal of Vascular Surgery, the effects of propranolol on the growth rate of abdominal aortic aneurysms were
studied[18]. In a double-blind, randomized fashion, 272 patients were treated with a placebo,
whereas 276 patients were treated with propranolol, which means that this study, unlike
the landmark 1994 study, was placebo-controlled and had precautions against bias[18]. During the observation period, which averaged 2.5 years, 73 patients in the placebo
group stopped taking the medication, whereas an outstanding 117 patients dropped out
in the propranolol group because of the drug's side effects, which caused patients
taking propranolol to have drastically poorer quality-of-life scores in three dimensions
of the Medical Outcomes Study 36-Item Short-Form Health Survey (SF-36). The groups
were comparable in their number of deaths, with 26 in the control group and 33 in
the propranolol group. Most importantly, the growth rates of the aneurysms were similar,
with the placebo group having a mean annual growth rate of 0.26 cm per year and the
propranolol group having a mean annual growth rate of 0.22 cm per year (P = 0.11). Interestingly, the β-blockers did appear to have some benefits, because
only 35 patients in the propranolol group required aortic resection compared with
55 patients in the control group. Because surgery is typically only performed on relatively
large aneurysms, the drugs may help slow the growth rate of aneurysms past a certain
size. However, the decision to operate is a subjective decision that reflects the
state of mind of the surgeons, as well as physical processes in the patient. Regardless,
the authors concluded that they could find no clinically significant effect of propranolol
on the growth rate of the studied abdominal aortic aneurysms[18].
Wilmink et al.[19] reached the same conclusions in a randomized, blinded study of 477 patients. The
aneurysms in the placebo group (221 patients) experienced a mean growth of 0.25 mm
during the observation period, whereas aneurysms in the propranolol group grew a mean
of 0.24 mm. The investigators also found that compliance with propranolol treatment
was exceedingly low, with 31% of the propranolol group dropping out of the study compared
with 15% of the control group. One benefit that propranolol manifested in the study,
however, was that aneurysms larger than 3.9 cm grew a mean of 0.44 mm in the placebo
group and only 0.13 mm in the propranolol group, which demonstrates that β-blockers
may potentially slow the growth rate in aneurysms larger than 3.9 cm. Nevertheless,
as in the aforementioned study, the authors concluded that β-blocker therapy for aneurysms
should not be recommended, because no statistically significant reduction in aneurysm
growth could be demonstrated[19].
Still, other smaller studies provide data that suggest that β-blocker treatment is
beneficial in the management of abdominal aortic aneurysms[20]
[21]. In 1988, Leach et al.[20] placed 12 patients in a β-blockade treatment group and 15 patients in a control
group. The researchers found that over the course of 3 years, the control group had
an annual aneurysm growth rate of 0.44 cm per year, whereas the treatment group had
an annual aneurysm growth rate of just 0.17 cm per year. With the end points defined
as death, rupture, or surgery, the control group had 6 end points, whereas the treatment
group had 5 end points[20]. In 1994, Gadowski et al.[21] also found β-blocker treatment beneficial in slowing aortic growth rate. The 38
patients receiving β-blockers had a reduced aortic aneurysm growth rate of 0.36 cm
per year compared with 0.68 cm per year for the 83 patients in the control group.
Both of these studies concluded that β-blocker treatment significantly reduced abdominal
aortic aneurysm growth rates.
We offer the additional observation that aneurysm growth is usually indolent, and
very long follow-up is required for conclusive evidence regarding differential rates
of aneurysm growth or patient death. Trials involving a mean follow-up of 5–10 years
would be much more convincing. Although the studies of patients with small abdominal
aortic aneurysms have not conclusively demonstrated an effect on growth rate, the
follow-up has been short, and there is evidence to suggest that there may be a beneficial
effect with longer follow-up and with larger aneurysms.
Thus, the medical community has been left without clear-cut evidence for or against
the treatment of abdominal aortic aneurysms with β-blockers. As shown in [Table 3], in the two largest studies, the investigators could not find a definitive correlation
between β-blocker treatment and reduced aneurysm growth rate. Nevertheless, in two
smaller studies, researchers found that aneurysm growth was significantly slowed by
β-blockers. So, medical professionals are left with the following question: Are the
potential benefits of β-blocker treatment worth the severe side effects and toll on
the patient's quality of life?
Table 3.
Comparison of the Effects of β-Blockers on Patients without Marfan Syndrome in Four
Separate Clinical Trials
|
Control group
|
Treatment group
|
|
Patients
|
Growth
|
Patients
|
Growth
|
|
Propranolol Aneurysm Trial Investigators[18]
|
272
|
0.26 cm/y
|
276
|
0.22 cm/y
|
|
Wilmink et al.[19]
|
221
|
0.25 mm
|
256
|
0.24 mm
|
|
Leach et al.[20]
|
15
|
0.44 cm/y
|
12
|
0.17 cm/y
|
|
Gadowski et al.[21]
|
83
|
0.68 cm/y
|
38
|
0.36 cm/y
|
Although the study by Shores et al.[8] and similar studies have suggested β-blockers could possibly be used to treat aortic
aneurysms in patients with Marfan syndrome, it would be inappropriate to presume that
such findings can be extrapolated directly to patients without Marfan syndrome. Furthermore,
the few encouraging results of the abdominal aortic aneurysm treatment studies should
not be extrapolated to thoracic aortic aneurysms because of the inherent differences
between abdominal aortic aneurysms and thoracic aortic aneurysms[22].
Limitations
As with all reviews, the present review is limited by the publications it analyzes,
which vary widely in their design. For instance, the studies varied in the type of
β-blockers used, with some using propranolol, others using atenolol, and still others
using a mix of different β-blockers. The studies also varied greatly in the age of
the patients observed; therefore, it is possible that the age of the patients influenced
how well the patients responded to β-blocker treatment. Most importantly, the studies
varied in the standards used to compare treatment and control groups. Some trials
emphasized an annual aneurysm growth rate, whereas others focused on end points such
as death and rupture.
Conclusions
Our review discusses the evidence (or lack thereof) to support the routine administration
of β-adrenergic blockade in patients with aortic aneurysms. The review seeks to underscore
why a conventional treatment for aortic aneurysms requires more robust scientific
evidence for its use. Although the theoretical reasoning behind β-adrenergic blockade
therapy is logical, that is simply not enough. β-Blockers have not been proven to
consistently reduce the aortic aneurysm growth rate in Marfan syndrome or the general
population.
The current American College of Cardiology Foundation/American Heart Association guidelines
include a class I recommendation for the use of β-adrenergic–blocking drugs for all
patients with Marfan syndrome and aortic aneurysm to reduce the rate of aortic dilation,
but acknowledge a level B evidence for their recommendation[23]. In addition, there is a class IIa recommendation for patients with thoracic aortic
aneurysm for the use of β-blockers (as well as angiotensin-converting enzyme inhibitors
or angiotensin-receptor blocker blockers) to lower blood pressure to the lowest point
patients can tolerate without adverse effects.
Our findings do not dispute these recommendations but reinforce a strong call for
larger, multicenter, randomized clinical trials to test the efficacy of β-blockers
to reduce the rate of dilation and clinical outcomes in individuals with aortic aneurysm.
Finally, our results also do not address or impugn the use of β-adrenergic blockade
in individuals with aortic dissection or its variants or in individuals with fixed
atherosclerotic disease, for whom evidence remains for their use.
