Reminder of important clinical lesson

Aneurysmal ‘pepper-pot’ atrial septal defect in an older
gentleman with multiple cerebrovascular attacks
Ewan J Mckay,1 Reza Ashrafi,2 Victoria Mckay,3 Aham Amadi2
1Cardiology Department, Royal Oldham Hospital, Manchester, UK;
2Cardiology Department, Aintree Cardiac Centre, Liverpool, UK;
3Cardiology Department, Alder Hey Children’s Hospital, Liverpool, UK

Correspondence to Dr Ewan J Mckay, ewan_mckay@hotmail.com

Summary
Our patient presented to a large university teaching hospital with a history of light-headedness, falls and multiple cerebrovascular ischaemic
events. This caused a right sided hemiplegia and the patient experienced significant functional limitation. Extensive investigations were
carried out to exclude any causative factors such as carotid artery disease and the patient had all identifiable cardiovascular risk factors
identified and modified. No significant pathology was found and a referral was made to the cardiology service. Transthoracic echocardiography
revealed a complex type 3 perforate ‘pepper pot’ atrial septal aneurysm with associated thrombus. The patient was commenced on warfarin
and appropriate rate limiting medication. After discussion of all interventional modalities, the patient opted for conservative management.

BACKGROUND
Atrial septal defects (ASD) are an important cause of
thromboembolic cerebrovascular attacks (CVA). In cases
of CVA where the patient has a simple ASD, the two
pathologies are likely to be unrelated. But when there are
further congenital abnormalities in addition to an ASD,
the prevalence of CVA is significantly increased. This has
been reflected in studies outlining the association of atrial
septal aneurysm (ASA) with cerebral ischaemia in patients
with normal carotid vasculature. This association is seen
in approximately 27% of patients and is much higher than
previously thought. In one such study of over 200 patients
under the age of 45 presenting with acute CVA, an ASA
was found in 86%.
We discuss here the different morphologies of aneurysmal ASD, the formation of the aneurysm and the pathophysiology of CVA. We look specifically at aneurysmal
ASD and the significant increase in the incidence of CVA
in these patients.

CASE PRESENTATION
An 83-year-old man presented to a large university teaching hospital in Liverpool. The patient’s initial presentation
was due to ‘constriction of his chest’ on exertion associated with
dizziness and collapse. These symptoms were mild but
causing significant effects on everyday activities. Medical
history included paroxysmal atrial fibrillation, dyslipidaemia, hypothyroidism and multiple previous transient ischaemic attacks (TIA). The calculated CHA2D2-VASc score of 2
indicated a risk of cerebral thromboembolic events in the
next 12 months of 4% (95% CI).1 He was commenced on
rate limiting treatment and formally anticoagulated with
warfarin. Following extensive investigation a complicated
type 3 ASD with aneurysm was identified.
Unfortunately, the patient had further TIAs despite having a therapeutic INR value and it was felt highly likely that
BMJ Case Reports 2012; doi:10.1136/bcr.12.2011.5402

the ASA was responsible for his numerous previous cerebral
ischaemic events. It is possible that some of the previous
cerebral events were secondary to thrombembolic foci from
the left atrium secondary to underlying atrial fibrillation but
there was significant distortion to the normal cardiac anatomy; the calculated annual thromboembolic risk is likely to
be significantly underestimated in this demographic given
this degree of distortion. The possible treatment options
available were discussed including remaining on current
therapy or attempting interventional closure. The patient
decided he would prefer conservative management.

INVESTIGATIONS
All of the usual investigations for such a presentation
were performed with no pathology identified. Following
thorough cardiovascular investigations a perforated ASA
was identified on transthoracic echocardiography and
confirmed on transoesophageal echocardiography (TOE).
Most recent echocardiography in February 2010 demonstrated an appearance consistent with an ASA with
lipomatous hypertrophy of the basal segment of the septum. The septum itself appeared thinned and aneurysmal
with a spherical structure seen adjacent to the right atrial
cavity. Colour flow was detected across the interatrial septum flowing from left to right seen as two adjacent high
velocity jets (figures 1 and 2). In addition, mild to moderate mitral and tricuspid regurgitation with preserved left
atrial size was noted. The pulmonary arterial pressure was
19 mm Hg indicating normal pulmonary arterial pressure.
Using transthroacic and TOE a bubble contrast enhanced
study can be performed. When a defect has already been
visualised or suspected this additional manoeuvre can be
used to determine if there is a significant volume of blood
crossing the defect. With attention to the subcostal view in
transthoracic echocardiography we used colour doppler to
identify flow but could not fully appreciate the true fluid
dynamic across the defect.
1 of 5

Figure 1

Two jets of high velocity colour flow across the atrial septum in parasternal demonstrated.

Bubble contrast studies are performed by injecting saline
or aggregated blood from the patient into a peripheral vein
and increasing venous return to the right atrium by deep
inspiration. This opacifies the right atrium and ventricle
with microbubbles, which reflect the ultrasound waves
from the echo probe.
Where there is flow from the right to left atrium across
a defect then bubbles will be seen in the left atrium. By
instructing the patient to perform a valsalva manoeuvre
the right atrial pressure is increased, allowing the practitioner to get a true appreciation of flow. Unfortunately,
patients undergoing TOE often receive sedation and the
coordinated movement required to perform adequate valsalva manoeuvre can be suboptimal.

DIFFERENTIAL DIAGNOSIS
Classification of aneurysms of the perforated atrial septum
The diagnostic criteria for ASA can vary greatly as there is
currently no consensus. The common criteria employed to
define an ASA seems to be the identification of a dilatation
of the atrial septum, with a base width and depth in excess
of 15 mm by 15 mm in adults (or 8 mm/m2 body surface
area in children) or if there is an oscillating membrane with
lateral movement of 15 mm or more. Perforated aneurysms
can then broadly be divided into four subgroups, which
then influence the management of the patient.2
Type 1 Aneurysm with a persistent foramen ovale (PFA)
closed with a large PFO occluder for example, star closure
device.

2 of 5

Type 2 Aneurysm with a single ASD, occluded with an
Amplatzer ASD occluder.
Type 3 Aneurysms with two or more perforations
located in no more than two clusters and not treatable with
a single device, treated with an occlusion device appropriate to the size of the defect.
Type 4 Multiple perforations spread irregularly over the
aneurysm leading to surgical closure or conservative treatment with anticoagulants.
There are a multitude of varying morphologies of
interatrial communications. With attention to that of the
secundum-type ASD, the possible management options
with regards to percutaneous closure using Amplatzer
septal occluder (ASO) devices are heavily influenced by
the structural characteristics. One hundred and ninety
patients who underwent a either percutaneous or surgical closure of this kind were studied and the morphology detailed. Centrally placed defects were observed in
24.2%. Morphological variations of secundum-type ASD
were detected in 75.8%. 79.5% underwent percutaneous
closure using ASO. 20.5% underwent surgical closure. It
was deemed that centrally placed defects, defects with
deficient superior anterior rim, multiple defects and perforated aneurysms of the interatrial septum are all morphological variations of secundum-type ASD suitable for
percutaneous closure using ASO.3
In a further study analysing the echocardiographic characteristics of isolated secundum-type ASD (ASD II), 64
patients were evaluated using both transthoracic and TOE.

BMJ Case Reports 2012; doi:10.1136/bcr.12.2011.5402

Figure 2 Aneurysmal ‘pepper-pot’appearance of the atrial septal defect. Note the lipomatous hypertrophy of the atrial septum seen at the
base of the perforate aneurysm.
Defects in 45.3% fulfilled the echocardiographic criteria
for percutaneous closure. Thirteen separate variations in
morphology of ASD II were identified. A statistically significant correlation between defect size and the number of
deficient defect rims was found. Importantly it has been
demonstrated that ASDs (II) are not static defects with
fixed morphology but in fact can ‘grow’ and evolve. With
this ‘growth’ the clearly defined borders of the defect that
at one time were present meaning that the defect could
be closed with a percutaneous device become eroded or
attenuated. It is postulated that this process is responsible
for the huge variety of morphologies encountered. It may
well be that these observations represent various snapshots in time at different stages of the evolution of this
process. With better understanding of the natural history
of this condition we can better classify the characteristics and therefore understand which patient cohorts will
respond best to intervention.4

OUTCOME AND FOLLOW-UP
Unfortunately, despite medical treatment with appropriate
rate limiting medication and anticoagulation with warfarin, the patient went on to have further TIA and a cerebrovascular event with left-sided hemiplegia. The patient
continues to maintain that he does not wish to be referred
for interventional therapy. Overall, while our patient preferred conservative management his recurrent events and
closely linked defects would have made him an ideal candidate for defect closure.

BMJ Case Reports 2012; doi:10.1136/bcr.12.2011.5402

DISCUSSION
Pathophysiology of CVA in this circumstance
Importantly the pathophysiology and causality of cerebrovascular accident in this cohort is quite different from
that caused by deep venous thrombosis with paradoxical embolus. In this particular group of patients it is not
unheard of for patients to have a number of ‘herald’ TIA
before CVA. This is reflected in studies outlining the prevalence of ASA in those with cerebral ischaemia and normal carotid vasculature to be much higher than previously
thought at 27%.5
While the morphology and position of defects in the
atrial septum are of great interest, the most important
issue is the high thromboembolic potential. While the
exact mechanism of this increased risk is open to question, there have been suggestions that thrombi can form
in the aneurysmal sac that can be transmitted through
a defect such as a PFO or ASD. For the physician caring
for these patients and deciding on management such as
interventional closure there is little information on how to
accurately assess the potential embolic risk. In two studies so far the only factor associated with additional stroke
risk was a thickness of greater than 5 mm of the aneurysmal membrane.6 7 There is no validated protocol for
which this thickness should be reproducibly measured.
The precise constellation of structures from which this
value is established from echocardiography is left entirely
to operator discretion which of course introduces interoperator variability. It would be our recommendation that

3 of 5

the value is established from what appears to be the thickest segment by eye.
It is postulated that microthrombi formation takes
place at the base of the perforated aneurismal defect. This
walled structure will shelter the developing thrombi from
turbulent high velocity flow. It may be that increased aneurismal membrane thickness is better able to shield developing thrombus as it develops in the aneurismal sac. It is
the swinging of this mobile ‘pepper-pot’ ASD through the
cardiac cycle, not to dissimilar to that of a swinging chandelier, which leads to the displacement of the thrombus
across the ASD with the eventual repercussions. It is for
this reason that CVAs are strongly associated with ASA
and this clinical hypothesis has since been confirmed by
autopsy.8 Importantly, it should be acknowledged that
in this circumstance other unusual complications can be
encountered. Additionally, patent foramen ovale (PFO) is
also associated with CVA, mostly when associated with
ASAs. Unusual systemic associations such as cerebellar
abscess have been documented secondary to gemella morbillorum infection, a commensal of the oropharynx and
upper respiratory tract.9

Formation of aneurysmal ASD
While there is little known about the precise nature of
how ASA come to be, there are data from a number of
multi-centre observational trials that search for associated
anomalies. The results clearly demonstrate that the most
common associated cardiac malformation with regard to
ASA is the presence of a PFO.5 In one such study of over
200 patients under the age of 45 with stroke, an ASA was
found in 86% but in only 10% of the general population.5
This association has been confirmed in other studies looking at causes of stroke.6–7 10
2D transthoracic echocardiography has become one of
the routine initial investigations performed in the diagnosis and management of TIA or CVA, usually to exclude the
possibility of intraventricular thrombus. As demonstrated
in this patient there are other important cardiac pathologies that may be responsible for acute vascular neurological
pathology. It would be our recommendation that patients
with recurrent TIA with normal carotid Doppler studies
and few identifiable atherosclerotic risk factors, that a 2D
transthoracic echocardiogram with valsalva and bubble
contrast be performed to exclude the possibility of PFO or
ASD. It must be remembered that TOE is a more detailed
investigation with greater spatial resolution of structural
pathology. This investigation should be reserved for those
patients where transthoracic images are of limited diagnostic quality.7
Early recognition and delineation of the precise structural pathology of an ASD with aneurysm may influence
the management of the patient as discussed above. ASDs
may be missed as a possible cause of CVA when not as
obvious as in this case and so we should have a low threshold for investigation.
A major advancement recently was a greater appreciation
of the management of patients with PFO. The CLOSURE-1
trial completed in 2010 was the first randomised controlled
trial with a cohort of 909 patients over three independent
centres. This trial studied PFO closure using the STARFlex
closure device with the patient receiving 6 months of aspirin and clopidogrel plus an additional 18 months of aspirin
4 of 5

versus best medical therapy defined as aspirin or warfarin on the basis of CHADS2 score. After 2 years of study,
the rates of stroke and TIA were very similar between
the two study groups but statistical significance was not
achieved. Major adverse vascular complications and incidence of further atrial fibrillation were statistically higher
in the interventional group. It is essential to emphasise that
this does not negate the merit of intervention but suggests
that further study is required to refine selection criteria.
We would hope that positive criteria rather than a lack of
criteria can be identified and distributed in a format that is
easy to follow without confusion. This will allow patients
to be reassured with confidence. Current indications for
interventional closure include:
▶ Infarction in both cerebral hemispheres
▶ Anterior and posterior circulation involvement
▶ Cerebrovascular event with silent renal or splenic
infarction.
Other indications for percutaneous closure of interatrial
communications of this kind outside of that documented
following CVA include hypoxia due to persistent or intermittent right-to-left shunting,11 dilated right ventricle,12
peripheral embolic event, disabling migraine with aura,
professional scuba diving and severe platypnea-orthodeoxia syndrome.13 Whenever undertaking an interventional
procedure the potential risks of the procedure versus the
benefits in terms of primary and secondary prevention
plus of course any symptom benefit must be weighed.
While symptom benefit is of paramount importance to the
patient, when the potential cost although rare may include
arrhythmias (specifically supraventricular tachycardia),
pericardial effusion, thrombus formation on the left atrial
disc with systemic embolisation and right iliac vein dissection with retroperitoneal and groin haematoma, we must
ensure the patient is fully aware of these potential complications and in the first instance that the circumstances are
such that intervention can be justified.14

Learning points
There are causes of common conditions/presentations
such as TIA/CVA which are unknown to junior doctors.
In the younger demographic with TIA/CVA ASA have
been shown to be more prevalent than previously
thought.
▶ We must be aware of pathologies although previously
thought to be uncommon which research dictates is
more prevent than we perceive. We must use imaging
modalities available (not involving radiation or contrast
material where possible) in aiding diagnosis. By doing
this we minimise risk to our patients while identifying
pathology which may be amenable to intervention.
▶ It would be our recommendation that patients with
recurrent TIA with normal carotid Doppler studies and
little identifiable atherosclerotic risk factors, that a
2D transthoracic echocardiogram with valsalva and
bubble contrast be performed to exclude the possibility
of PFO or ASD.
▶

Competing interests None.
Patient consent Obtained.

BMJ Case Reports 2012; doi:10.1136/bcr.12.2011.5402

REFERENCES
1. Gage BF, Waterman AD, Shannon W, et al. Validation of clinical classification
schemes for predicting stroke. Results from the national registry of atrial
fibrillation. JAMA 2001;285:2864–70.
2. Ewert P, Berger F, Vogel M, et al. Morphology of perforated atrial septal
aneurysm suitable for closure by transcatheter device placement. Heart
2000;84:327–31.
3. Podnar T, Martanovic P, Gavora P, et al. Morphological variations of
secundum-type atrial septal defects: feasibility for percutaneous closure using
Amplatzer septal occluders. Catheter Cardiovasc Interv 2001;53:386–91.
4. Prokselj K, Kozelj M, Zadnik V, et al. Echocardiographic characteristics
of secundum-type atrial septal defects in adult patients: implications
for percutaneous closure using Amplatzer septal occluders. J Am Soc
Echocardiogr 2004;17:1167–72.
5. Tchana B, Hagler DJ, Carano N, et al. Device closure of fenestrated atrial
septal aneurysm: difficulties and complications with implantation of two
devices. J Invasive Cardiol 2004;16:532–4.
6. Mattioli AV, Aquilina M, Oldani A, et al. Atrial septal aneurysm as a
cardioembolic source in adult patients with stroke and normal carotid arteries.
A multicentre study. Eur Heart J 2001;22:261–8.

7. Schneider B, Hanrath P, Vogel P, et al. Improved morphologic characterization
of atrial septal aneurysm by transesophageal echocardiography: relation to
cerebrovascular events. J Am Coll Cardiol 1990;16:1000–9.
8. Mattioli AV. [Study of the interatrial septum using transthoracic
echocardiography]. Cardiologia 1997;42:257–64.
9. Lopes A, Providencia R, Pais RP, et al. Cerebellar abscess by Gemella
morbillorum in a patient with inter-atrial communication. Arq Neuropsiquiatr
2007;65:1022–5.
10. Silver MD, Dorsey JS. Aneurysms of the septum primum in adults. Arch
Pathol Lab Med 1978;102:62–5.
11. Ilkanhoff L. Transcatheter device closure of interatrial septal defects in
patients with hypoxia. J Interv Cardiol 2005;18:227–32.
12. Egred M, Andron M, Albouaini K, et al. Percutaneous closure of patent
foramen ovale and atrial septal defect: procedure outcome and medium-term
follow-up. J Interv Cardiol 2007;20:395–401.
13. Lisignoli V, Lanzone AM, Zavalloni D, et al. Closure of patent foramen ovale:
when and how? Curr Vasc Pharmacol 2007;5:322–7.
14. Chessa M, Carminati M, Butera G, et al. Early and late complications
associated with transcatheter occlusion of secundum atrial septal defect.
J Am Coll Cardiol 2002;39:1061–5.

This pdf has been created automatically from the final edited text and images.
Copyright 2012 BMJ Publishing Group. All rights reserved. For permission to reuse any of this content visit
http://group.bmj.com/group/rights-licensing/permissions.
BMJ Case Report Fellows may re-use this article for personal use and teaching without any further permission.
Please cite this article as follows (you will need to access the article online to obtain the date of publication).
Mckay EJ, Ashrafi R, Mckay V, Amadi A. Aneurysmal ‘pepper-pot’ atrial septal defect in an older gentleman with multiple cerebrovascular attacks.
BMJ Case Reports 2012;10.1136/bcr.12.2011.5402, Published XXX
Become a Fellow of BMJ Case Reports today and you can:
▶ Submit as many cases as you like
▶ Enjoy fast sympathetic peer review and rapid publication of accepted articles
▶ Access all the published articles
▶ Re-use any of the published material for personal use and teaching without further permission

For information on Institutional Fellowships contact consortiasales@bmjgroup.com
Visit casereports.bmj.com for more articles like this and to become a Fellow
Keep up to date with all published cases by signing up for an alert (all we need is your email address) http://casereports.bmj.com/cgi/alerts/etoc

BMJ Case Reports 2012; doi:10.1136/bcr.12.2011.5402

5 of 5