Abdominal aortic aneurysms in women

Abstract

Abdominal aortic aneurysm (AAA) has long been recognized as a condition predominantly afflicting males, with sex-associated differences described for almost every aspect of the disease from pathophysiology and epidemiology to morbidity and mortality. Women are generally spared from AAA formation by the immunomodulating effects of estrogen but once they develop, the natural history of AAAs in women appears to be more aggressive, with more rapid expansion, a higher tendency to rupture at smaller diameters, and higher mortality following rupture. However, simply repairing AAA at smaller diameters in women is a debatable solution, as even elective endovascular AAA repair (EVAR) is fraught with higher morbidity and mortality in women compared to men. The goal of this review is to summarize what is currently known about the effect of gender on AAA presentation, treatment, and outcomes. Additionally, we aim to review current controversies over screening recommendations and threshold for repair in women.

Epidemiology

The prevalence of AAA in persons age over 65 years is 3–4 times higher in men than women: estimates range between 1.7–4.5 percent in men¹ versus 0.5–1.3 percent in women². Recent data indicates AAA incidence is declining³, attributed to falling rates of cigarette smoking, but declines in AAA-related mortality have been more prominent for men than women^{4, 5}. According to the Centers for Disease Control and Prevention (CDC), in 2013, 0.07% of the 1,290,959 deaths in women and 0.13% of the 1,306,034 deaths in men were attributable to ruptured AAA⁶. However, reports of incidence and prevalence vary widely depending on the definition of AAA used. The standard definition of AAA is an aortic diameter ≥ 30mm a definition derived by Steinberg and Stein, who analyzed a series of aortographs of men and women and determined that aortas measuring 30mm at any of four locations (at the 11^th rib, the region of entrance of the thoracic aorta into the abdomen, above and below the renal arteries, and at the aortic bifurcation) were two standard deviations above the mean and thus considered aneurysmal⁷. AAA prevalence in women may be as high as 6.2%–9.8% if using a definition of ratio of infrarenal to suprarenal diameter of ≥1.2⁸ or a definition of ≥1.5 times the normal infrarenal aortic diameter⁹, with the ramification that studies using a cutoff of 30mm may be under-enrolling women.

Risk factors for developing AAA are the same for women as for men, namely age, smoking, hypertension, and family history¹⁰. Some studies suggest that the association between smoking and AAA is stronger in women than men¹¹, which is consistent with findings that compared to male smokers, women who smoke seem to be more susceptible to lung cancer¹², myocardial infarction¹³, and stroke¹⁴ too. Additionally, a large body of evidence suggests that women have a harder time quitting smoking than men. Women are just as likely to desire and attempt to quit, but less likely to remain abstinent¹⁵. Various treatment methods for smoking cessation, including behavioral counseling, nicotine replacement, and antidepressants have been found to be less effective in women¹⁶.

Screening

Ultrasound screening has been shown to be an efficacious¹⁷ and cost-effective¹⁸ way to prevent AAA rupture and reduce mortality in men and there is consensus among the Society of Vascular Surgery (SVS), the European Society for Vascular Surgery, and the United States Preventative Services Task Force (USPST) that one-time screening for AAA by ultrasonography should be offered to men age 65 years or older, particularly those who have ever smoked and/or have a family history of AAA^19–21. In contrast, there is no consensus on screening recommendations for women. The USPST recommends against screening, citing low disease prevalence in women and insufficient evidence²². Indeed, to date, there has only been one randomized trial to evaluate screening in an unselected population of women, which did not show a significant 5- or 10-year mortality benefit²³. Even in a selected population of smokers, the USPST recommends against screening women, citing low prevalence largely based on a Swedish study of 5140 women over the age of 70, with a AAA prevalence of only 2.1% among active smokers².

However, in a recent study using data from the National Health Service Abdominal Aortic Aneurysm Screening Program (NAAASP) in England²⁴, AAA screening among men 65 years of age or older was shown to be cost effective despite AAA prevalence of 1.5%. Using a Markov model, the authors showed that screening would continue to be cost effective down to a threshold prevalence of 0.35%. Wanhainen et al. similarly used a Markov model they had previously used for men²⁵ to show that the incremental cost-effectiveness ratio for women was similar to that found for screening men, largely because the higher rupture rate in women balanced out the lower prevalence²⁶.

The incidence of AAA among women age 65 or older who have ever smoked may be higher than that in their non-smoking male counterparts¹¹, who may nevertheless be offered screening based on the current USPST recommendations. In line with these considerations, the SVS recommends screening women age 65 or older who have smoked or have a family history¹⁹. Currently, Medicare Part B covers only a one-time screening for women with a family history of AAA²⁷.

Pathophysiology

AAAs are thought to result from a degenerative process, characterized histologically by smooth muscle cell loss and structural deterioration of the aortic wall, particularly within the elastic media and adventitia²⁸. The destruction of the elastin and collagen is via matrix metalloproteinases (MMPs), proteolytic enzymes released by T and B lymphocytes, macrophages, and other chronic inflammatory cells²⁹. Experimental rodent models have demonstrated that like in humans, males have a higher tendency to develop AAAs than females and higher levels of MMP production and activity are observed in males³⁰.

It is hypothesized that the difference in AAA prevalence rates between males and females is mediated at least in part by the protective effects of estrogen. Female rat aortas transplanted into male rats develop AAAs at the same rate as male rats, suggesting loss of the protective female sex hormone milieu³⁰. Female rats that undergo oophorectomy have higher AAA growth rates when compared to control animals but rats injected with estradiol have significantly lower AAA growth rates³¹. In vitro, smooth muscle cells of male rats demonstrate decreased MMP activity when treated with estradiol³² and in vivo, male rats in which estrogen pellets are implanted have significantly smaller aortic aneurysms than male rats in which sham pellets are implanted³⁰.

Estrogen is believed to mediate its protective effect via immunomodulation. Estrogens reduce macrophage MMP production and thus diminish collagen destruction and remodeling³¹. Estrogen also attenuates immune cell migration, cytokine production, growth factor expression, and chronic inflammation, processes implicated in AAA development³³.

Not only do estrogens appear to be protective, male androgens may directly have an adverse effect on AAA formation. Henriques et al. performed orchidectomy in an AAA mouse model and found that this significantly reduced the AAA incidence in male mice to the level observed in female mice³⁴. In the same study, administration of dihydrotestosterone to castrated mice increased the incidence of AAA in both males and females and these mice tended to have AAAs of greater weight (used as a proxy for AAA severity) that had a significantly higher incidence of rupture.

Expansion and Rupture

Despite having a lower prevalence of AAA than men, the growth rate in women has been shown to be faster³⁵, rupture is four times as likely during surveillance, and the likelihood of fatal rupture is three times as high, even after adjustment for age³⁶. When rupture occurs, it is at aneurysm diameters 5–10mm smaller in women than in men³⁷, which is perhaps intuitive given that women have smaller aortas to begin with³⁸.

Some authors believe variability in geometric and biomechanical properties may be the reason why women with AAAs have an increased growth rate and have a higher risk of rupture. Vande Geest et al. found a non-significant trend toward lower tensile strength in the freshly excised anterior aortic walls of 10 females compared to 24 males undergoing open AAA repair³⁹. Using finite element analysis, Larsson et al. compared the computed tomography scans of 15 men and 15 women matched for AAA diameter and found no differences in peak wall stress but an almost significantly higher peak wall rupture risk (a parameter which includes both peak wall stress and strength) in women (P=0.06)⁴⁰. Similarly, using finite element analysis and three-dimensional reconstruction, Fillinger et al. showed that women have a higher proportion of aneurysms with high peak wall stress (termed “high stress” aneurysms, defined as ≥44 N/cm²) but also noted that female gender was an independent predictor of rupture after controlling for wall stress.⁴¹

Although lower socioeconomic status has been shown to contribute to lower rates of AAA surveillance and thus increased risk of rupture in men, the same finding has not been shown in women⁴². It is unclear whether this difference is due to sex-related differences in attitudes towards utilization of medical care, screening and preventative care, family structure and gender roles, or whether studies to date have just been underpowered to detect differences in rupture rate based on insurance status among women.

Another potential explanation for why women rupture at smaller diameters is that women generally have smaller aortas because of their smaller body sizes, leading to the suggestion that an aneurysm of a given diameter in a woman represents a greater relative dilatation and thus more advanced disease of the aorta compared with an aneurysm of the same diameter in a man. Multiple methods have been undertaken to index the aortic diameter to some measure of body size, including suprarenal aortic diameter³⁸, the third lumbar vertebral body⁴³, and body surface area (BSA)^{37, 44}. Although these indices have been shown to be more predictive of rupture than maximal aortic diameter alone, none have yet been incorporated into clinical practice for decision-making. In contrast, the Society of Thoracic Surgeons and the American College of Cardiology have formally integrated BSA into a nomogram⁴⁴ for the prediction of rupture risk of thoracic aortic aneurysms.

Once rupture occurs, it is well documented that women fare worse than men. A Canadian study showed that far fewer women than men undergo repair when they present with rupture, despite having similar comorbidity profiles⁴⁵ and those who did not undergo repair had a 30-day mortality of 90% compared to 41% for those who did. These findings were confirmed in an American study using data from the Nationwide Inpatient Sample⁴⁶, which in addition to showing lower repair rates in women, also showed that in-hospital mortality was higher in women after rupture repair. Similarly, in a study of Medicare beneficiaries from 1995 to 2006, thirty-day mortality after repair of ruptured AAAs was observed to be 8.9% higher after open repair and 7.1% higher after endovascular aneurysm repair (EVAR) in women compared to men⁴⁷. In this study, female gender was found to be an independent risk factor for death even after controlling for age, year, and type of procedure. Using the same database, this research group later evaluated long-term (6-year) outcomes and reported that survival was similarly superior in men for both open and EVAR repair of ruptured AAA and women had higher rates of readmission⁴⁸. Using propensity score matching to compare EVAR and open surgical repair for ruptured AAA among Medicare beneficiaries, Edwards et al. similarly found that female sex was a significant predictor of perioperative mortality⁴⁹.

Threshold for Repair

Due to the higher proclivity of aneurysms to rupture in women, the smaller diameters at which they rupture, and their inferior outcomes following rupture, significant controversy exists over whether a lower size threshold should be used for women in consideration of repair. In 2003, the Joint Council of the American Association of Vascular Surgery and the Society of Vascular Surgery (SVS) suggested considering repair for women at diameters of 4.5–5.5cm⁵⁰ and the updated 2009 SVS Practice Guidelines modified this recommendation¹⁹, saying “young, healthy patients, and especially women, with AAA between 5.0cm and 5.4cm may benefit from early repair”. However, several arguments against earlier repair in women have been raised.

Four randomized controlled trials, including the UK Small Aneurysm Trial (UKSAT), the Aneurysms Detection and Management Trial (ADAM), the Comparison of Surveillance vs. Aortic Endografting for Small Aneurysm Repair Trial (CESAR), and the Positive Impact of Endovascular Options for Treating Aneurysms Early Trial (PIVOTAL) all separately concluded that there was no short- or long-term mortality benefit to early repair of small aneurysms (4.0–5.5cm)⁵¹. However, women were underrepresented in all of these trials. The UKSAT had the highest proportion of women (17%) with just 188 women. Whether the results of these trials can be generalized to women is therefore uncertain.

One of the strongest arguments against early repair in women, aside from the lack of level I data to support it, is that multiple studies have suggested women have notably worse outcomes after elective AAA repair. If this were the case, the risk of intervention could potentially outweigh any added benefit. For open repair of intact aneurysms, women have been consistently reported to have higher in-hospital and 30-day mortality⁴⁶, higher postoperative complication and readmission rates, and inferior long-term survival^{48, 52}.

AAA Repair

In the current era, EVAR has become the primary treatment method for AAA repair due to improved perioperative morbidity and mortality compared to open surgical repair⁵³. While women would be expected to derive the same amount of benefit from EVAR as men, retrospective analyses that have examined gender-based outcomes after elective EVAR have arrived at conflicting conclusions. Studies using large-scale databases such as that from Medicare⁵⁴, data from the American College of Surgeons Quality Improvement Program (ACS NSQIP)⁵⁵, and the Nationwide Inpatient Sample⁵⁶, have tended to demonstrate higher perioperative mortality rates for women after elective EVAR (Table I).

TABLE I.

Gender-specific 30-day mortality rates following endovascular and open abdominal aortic aneurysm repair

				EVAR				OAR
				rAAA		iAAA		rAAA		iAAA
Author	Database	Years	No. of Patients	Men	Women	Men	Women	Men	Women	Men	Women
Abedi et al.48	American College of Surgeons National Surgical Quality Improvement Program (ACS NSQIP)	2005–2007	3,662		1.9 (men)α vs. 3.4 (women)β			-	-	-	-
Egorova et al.41	Medicare Beneficiary Database	1995–2006	371,757	30.7	41.1	1.8	3.0	46.6	52.6	5.0	6.4
McPhee et al.39	Nationwide Inpatient Sample (NIS)	2001–2004	220,403	29.4*	27.7*	0.8*	2.1*	36.3*	44.7*	4.0*	6.1*
Lo, et al.45	Vascular Study Group of New England (VSGNE)	2003–2011	4,026	29.0	27.3	0.9	1.2	34.1	48.4	1.7	3.6
Chung et al.52	single institution	1992–2012	1,380	-	-	2.0	2.3	-	-	-	-
Gloviczki et al.50	single institution	1997–2011	934		0.8 (men)χ vs. 2.8% (women)δ			-	-	-	-
Ouriel et al.51	single institution	1996–2002	704	1.3 (men) vs. 3.1 (women)**				-	-	-	-

EVAR, endovascular aneurysm repair; iAAA, intact/elective abdominal aortic aneurysm; OAR, open aneurysm repair; rAAA, ruptured/emergent abdominal aortic aneurysm

α, 4.3% emergent; β, 6.5% emergent; χ, 2.8% ruptured; δ, 4.2% ruptured

^*number reflects in-hospital mortality rate rather than 30-day mortality rate

^**specific presentation (i.e. elective vs. urgent/emergent, intact vs. ruptured) not reported

In contrast, studies reporting single-center experiences or those using regional, clinical databases tend to show that women have mortality rates comparable to those of men after elective EVAR (Table I). Data from the Vascular Study Group of New England (VSGNE) showed that while women had higher mortality rates after open repair (both elective and emergent), in-hospital, 30-day, and 1-year mortality rates were equivalent among men and women undergoing EVAR for intact aneurysms⁵². However, women did suffer more procedural complications such as longer operating room (OR) times, unintentional hypogastric artery coverage, arterial injury, and return to the OR. They also had longer intensive care unit (ICU) and overall hospital stays and we less likely to be discharged home than men. These findings are consistent with reports by Gloviczki et al.⁵⁷, Ouriel et al⁵⁸, and Chung et al.⁵⁹, who similarly found that women had equivalent perioperative mortality rates but higher complication rates compared to men. These differences in study conclusions suggest either that the smaller institutional/regional studies were underpowered to detect significant differences in perioperative mortality rates or that women in these particular studies achieved better survival because they were treated at referral centers of excellence. Results from larger databases may better reflect outcomes for women in more “real world” settings, while keeping in mind the limitations of administrative databases that were originally intended for billing purposes.

Why do women have worse morbidity after elective repair? For one thing, AAAs develop on average 10–15 years later in women than men and thus women present at an older age for repair⁴⁸. Furthermore, medical management of cardiovascular risk factors is often inferior in women. Women are less likely to be asked about smoking and advised to quit than men⁶⁰, lipid-lowering agents are underutilized in women⁶¹ and among diabetics, women are less likely to reach their targeted blood pressure, hemoglobin A1c, and low density lipoprotein goals than men⁶². Failure to recognize or optimize cardiovascular comorbidities may contribute to the higher postoperative morbidity observed in women.

Another contributing factor may be that women have smaller body sizes and narrower arteries. Women undergo elective EVAR at lower rates than men⁵², presumably because of higher rates of anatomic unsuitability and failure to meet device Instructions For Use (IFU) criteria. In a cohort of 812 men and 251 women undergoing EVAR, Sweet et al. found that disproportionately more women than men failed to meet IFU criteria, even after adjustment for age and AAA size: 63% of women vs. 47% of men had neck length <15mm (p<.001), 26% vs. 12% had neck angulation >60 degrees (p<.001), and 55% vs. 35% had iliac diameters <6mm (p<.001)⁶³.

When women do undergo EVAR, their smaller vessels have been implicated in higher rates of access-related complications. Among patients undergoing elective EVAR, women are more likely than men to sustain arterial injury (5.4% vs. 2.7%, p=0.13)⁵². Additional access maneuvers are more frequently performed in women than men: iliac access angioplasty (31% vs. 10.9%, p=0.007), uni-iliac conversion (13.8% vs. 1.4%, p=0.005), and iliac “chimney” conduit (12.1% vs. 1.2%, p=0.0001)⁶⁴. It is likely for this reason that lower profile devices, such as the Gore Excluder (W.L. Gore & Associates, Flagstaff, AZ, USA), are chosen more frequently for use in women⁶⁵. Newer, lower profile, more flexible devices hold promise in further reducing the gender gap in complication rates.

Finally, as previously mentioned, because the same AAA diameter criteria are typically used for men and women, by the time women present for repair, it is likely that their disease is more advanced. Delayed diagnosis contributes to delayed repair, and women may be particularly susceptible to delayed diagnosis since they do not undergo routine screening for AAA.

Surveillance after Repair

Although EVAR is associated with lower perioperative mortality compared to open surgical repair, patients who undergo EVAR need lifelong yearly surveillance to identify potential endoleaks, graft migration, graft failure, and persistent aneurysm growth¹⁹. However, reported compliance with this surveillance regimen has been unsatisfactory⁶⁶. Although risk factors for poor compliance and loss to follow-up include advanced age and emergent presentation with rupture, factors more prevalent in women, gender has not been found to be an independent predictor of incomplete surveillance⁶⁷.

With respect to long-term graft-related complications and secondary interventions following EVAR, several retrospective studies have demonstrated no gender association with development of endoleak, sac growth, post-repair rupture, graft migration, graft occlusion, and reintervention rates^{58, 64}.

In summary, although women are generally protected from the development of AAAs, the ones that do develop present at a later age and behave more aggressively with higher growth and rupture rates. Screening is thus warranted in women with multiple risk factors and those with a family history. Whether women should undergo repair at smaller aneurysm diameters than men remains controversial but more data are accumulating to suggest that elective EVAR may produce comparable mortality rates among men and women, at least in centers of excellence with higher operative volumes⁶⁸. Women nonetheless continue to suffer higher complication rates, a problem that may be expected to improve with the development of newer, lower-profile devices that can better navigate challenging aortoiliac anatomy. In the meantime, clinicians should maintain a higher index of suspicion for access-related complications in women undergoing EVAR. They should also be alert to the higher possibility of suboptimal medical management of cardiovascular risk factors in women, which may contribute to higher rates of postoperative complications.