This decline in rAAA incidence caused much debate and is now thought to be a result of multiple factors. In discussing the trend for rAAA, it is useful to talk about the similar trends seen in AAA disease overall. Historically, there was an increase in AAA incidence similar to that for rAAA [26]. In a countrywide analysis of all admissions in Denmark, from 1977 to 1990, a fourfold increase in diagnosis of AAA was found [27]. Scotland showed similar results with a large administrative study showing a threefold increase in hospital admission for AAA, both elective and emergent, from 1981 to 2000, with no change in their elective repair rate of approximately 80 % [28]. This same finding was seen in the USA, where from 1951 to 1980, there was a sevenfold increase in AAA diagnosis in a Minnesota population studied [29]. Multiple reasons for this have been cited, including increased incidence, likely related to smoking trends, increased survival of high-risk populations, increased utilization of advanced imaging, and changing diagnostic criteria, but no one reason has been widely accepted. It is logical that the incidence of rAAA should follow that of AAA, unless dramatic changes in screening occurred over the same time interval. Increasing the number of elective operations for AAA as a prophylactic measure against rupture can also be another confounder to such an assumption. AAA and rAAA are now both decreasing in the twenty-first century [30, 31]. Reduction in risk factors, especially smoking, is likely a major reason for both decreases in AAA prevalence and subsequent rAAA incidence, but as important for rAAA is the increasing elective repair of high-risk patients with the wide acceptance of EVAR, to be discussed further below. This downward trend has been seen in much of the developed world. The data are so consistent that most accept the decrease in hospital admissions and mortality reported for rAAA as acceptable measures of a decrease in overall rupture rates of AAA, even without autopsy studies to confirm that the pre-hospital mortality rate is also declining or at least staying the same.

In order to accurately measure pre-hospital rates of rAAA, it is necessary to have autopsy studies, but few areas have the systems set up to study such events retrospectively. There has been a dramatic decline in the rate of clinical autopsy, making replication of more complete historical studies difficult [32]. Population-based studies have attempted to develop surrogates for the clinical autopsy, but these efforts have not been satisfactory in giving an accurate incidence rate of pre-hospital rAAA. This leaves us with only historical data when trying to determine incidence of current pre-hospital rAAA rates. Such a dependence on outdated estimates is problematic, given the changing prevalence of AAA and incidence of rAAA that present to the hospital. Nevertheless, it is still important to understand the findings of these studies. Caution should be used when comparing studies that report on trends of incidence or total mortality for rAAA as the definition for what is included in these measures may differ.

Starting in 1966, Szilagyi et al. showed larger aneurysms (>6 cm) were more likely to rupture than smaller aneurysms (<6 cm) (Fig. 3.5) [52]. This size relationship with risk of rupture was further supported by autopsy studies [53, 54]. The UK Small Aneurysm Trial (UKSAT) gave a comprehensive estimate of AAA rupture risk for small aneurysms, size 4.0–5.5 cm, and found no difference in survival for early operation versus surveillance [55]. Brown et al. used the randomized UKSAT population prior to any surgery and added in the 1167 patients ineligible for randomization in this study, and who were also followed, and did find a difference in rupture risk for smaller aneurysms; rupture risks per 100 patient-years were 0.3 % for AAA <4.0 cm, 1.5 % for AAA 4.0–4.9 cm, and 6.5 % for AAAs 5.0–5.9 cm [56]. A population-based study from Minnesota followed 176 patients selected for nonoperative management and found an annual rupture risk of 0 % for AAAs <4 cm, 1 % for AAAs 4.0–4.9 cm, and 11 % for AAAs 5.0–5.9 cm [57].

For larger aneurysms, Parkinson et al. performed a meta-analysis of 11 studies, including 1514 patients who were deemed unfit for elective operation but had aneurysms larger than 5.5 cm [58]. Within this study, the rate of rupture was found to be 3.5 % per year (95 % CI 1.6–8.7 %) for aneurysm size 5.5–6.0 cm, 4.1 % (0.7–9.0 %) for 6.1–7.0 cm, and 6.3 % (1.8–14.3 %) for aneurysms >7.0 cm. However, although initially deemed unfit, some patients in these series underwent elective repair. Thus these rates likely underestimate the true rupture risk. Another consideration is that this populations’ increased risk is not entirely related to diameter as patients deemed unfit for elective repair have additional risk factors for rupture, such as gender, smoking status, additional wall stress factors, and comorbidities (e.g., COPD and hypertension), which will be discussed below. As a result, caution should be used when generalizing these rupture risks to the general population, who presumably have fewer comorbidities than those deemed unfit for repair. All-cause mortality in this unfit for repair group is known to be high, with a 2-year survival rate as low as 35 % [59].

Laminar flow is easily disturbed in the blood vessels, especially with aneurysmal degeneration, and over time forces from this disturbance can lead to adverse effects on the vessel wall. Laplace’s law states that the wall tension of a symmetric shape is directly proportional to the radius and intraluminal pressure and inversely proportional to wall thickness. Aneurysms are not symmetric shapes, and logic would tell us that eccentric or saccular aneurysms present a greater risk for rupture than more diffuse and fusiform ones. This has been difficult to quantify, but computer modeling, such as that by Vorp et al., has shown that aneurysm shape is almost as important for wall stress, and likely rupture risk, as is diameter [60]. Furthermore, commonly seen intraoperatively and on preoperative CT scans, small blebs on aneurysm sacs are postulated to pose an added risk for rupture and occur in equal frequency on small and large aneurysms. Histologically, these blebs often show an imbalance of matrix degradation and repair, leaving them especially vulnerable to wall stress [61].

Work is underway to improve our assessment of risk for rupture using models that factor in shape and asymmetry, as well as diameter, to determine overall wall stress. Initial studies have suggested using finite element analysis of wall stress, aided by advances in CT imaging, to advise patients on risk of rupture [62, 63]. This finite element model is basically a stress analysis model for AAA and includes the geometry of the AAA, the mechanical behavior of the AAA tissue, and the boundary conditions (e.g., blood pressure) [64]. As this idea gains momentum, the clinical applicability of such a complex algorithm has been called into question, and studies to simplify it by identifying the essential components of this stress analysis are in progress. Fillinger et al. attempted to show important anatomic details for risk of rupture after matching a group of 259 elective and rAAA by age, gender, and diameter [65]. They found that ruptured AAAs tend to be less tortuous but have more diameter asymmetry than their size matched intact counterparts. What is becoming clear is that to use AAA diameter as the only anatomic measure for risk of rupture and indication for elective repair is likely too simplistic. Fillinger et al. showed a model using wall stress to be superior to aneurysm diameter in predicting rupture and that wall stress was predictive in small aneurysms that rupture as well (Fig. 3.6) [62]. Newer models of wall stress are not ready for general use as of yet but are predicted to become part of common practice as methods improve.

It makes sense that more rapid expansion would cause a higher risk for rupture, but it has been hard to distinguish this risk from that of increased aneurysm size alone. In 2011 Powell et al. published a meta-analysis evaluating expansion rates and showed that larger aneurysms tend to increase in size at a faster rate; specifically a 10-mm increase in diameter size was associated with a mean of 1.62 (SEM 0.20) mm/year increase in growth rate (Fig. 3.7) [66]. In addition to diameter cutoffs, rapid AAA expansion is often used as an indication for elective repair, with expansion rate of >1 cm/year being the most commonly used rate for repair. In addition to diameter, another factor that increases the growth rate of aneurysms is presence of thrombus [67]. Thrombus is thought to induce hypoxia-driven inflammation that weakens the wall of the aneurysm [68]. Please see Chaps. 4 and 5 for more specific details on the role of the thrombus. Rate of expansion is a marker of aneurysm instability and should continue to be used in the identification of high-risk patients who warrant repair.

One of the most modifiable risk factors that could continue to have a large impact on reduction of ruptures from AAA is smoking cessation. Many credit smoking to part of the decline already seen in the incidence of rAAA. Early studies established a clear link between cigarette smoking and aneurysm development, dating back to 1958 [69]. This link with aneurysm development and mortality from AAA has been supported by multiple subsequent studies from both Europe and the USA [5, 30, 31]. A large screening study of US veterans attributed >70 % of all AAAs in the veteran population to smoking [70]. Sweeting et al. performed a meta-analysis using individual patient data from 18 studies analyzing factors that affected growth and rupture of small AAAs [71]. After adjusting for aneurysm diameter, there was a strong association between smoking and both growth rate and rupture, growth mean was 0.33 mm/year (SEM .07) faster, and risk of rupture was twofold higher in current smokers compared to ex- and never smokers (Fig. 3.8). Despite limitations relevant to any meta-analysis, such as heterogeneity of definitions and self-reporting reliability, these results are convincing.

Anjum et al. analyzed health statistics for England and Wales from 1979 to 2009 and estimated that the decrease in prevalence of smoking in England and Wales led to an avoidance of 8–11 deaths from rAAA per 100,000 persons [23]. This study also suggested treatment of hyperlipidemia, and coronary artery disease played a role in the decline rAAA deaths (Fig. 3.9). There are multiple reasons that seem plausible for the reduction in rAAA mortality, but it seems likely that smoking cessation has contributed greatly.

It has been clearly established that the incidence of AAA increases with age [26, 72].

The UKSAT did not find age to be associated with rupture after adjusting for known risk factors, including diameter; however multiple other studies found age strongly predictive [56, 71, 73, 74]. A possible explanation for the lack of predictive ability found in the UKSAT data is the collinear (overlapping) effects of covariates included in the model, such as aneurysm diameter and declining FEV1, which are closely related to increasing age.

Studies are now showing that the decline in the incidence of rAAA is also being felt across most age groups, although at variable rates. In England and Wales, from 1997 to 2009, there was a decrease in aneurysm rupture across all age ranges, significant in all except the >85-year-old group [23]. This oldest group was also found to have the highest incidence of hospitalization for rAAA at 94.7 per 100,000 people.

In general, older patients with rAAA have a higher in-hospital mortality rate, but a large part of this may be explained by disproportionate intervention rates favoring younger patients [23]. These in-hospital-mortality and rAAA-intervention-rate differences by age are consistent across countries, as highlighted by Karthikesalingam et al. who compared mortality from rAAA between England and the USA from 2005 to 2010 (Fig. 3.10) [40]. However, this difference between age groups may be closing thanks to the incorporation of rEVAR and presumably a lower turndown rate for those older patients getting rEVAR who would have been less likely to get open repair compared to their younger counterparts. Using National Inpatient Sample data from 2000 to 2005, our group showed the mortality benefit of rEVAR compared to open repair in those over 70 years old: in-hospital mortality rate of 36.3 % after rEVAR compared to 47 % after open repair (p < .001) for this age group [73]. Using Medicare data from 1995 to 2008, our group also showed that all ages had a decline in short-term AAA-related death, but this was most evident in the >80-year-old age group, and most of this mortality benefit was related to a steep decline in rupture deaths (Fig. 3.11) [4]. This age group had the largest increase in elective AAA repairs over the same time period suggesting that more aggressive management of intact AAAs in this high-risk group was preventing subsequent ruptures and related deaths.

Symptoms of abdominal, back, groin, or buttock pain related to an AAA have long been considered an indicator of impending rupture. Tenderness to palpation is a particularly ominous sign when associated with symptoms, and many clinicians have anecdotally used tenderness to palpation in the absence of pain symptoms as an indication for semi-urgent repair given the suspected rupture risk [56]. In the pre-EVAR era, multiple studies consistently showed worse 30-day and in-hospital mortality rates for repair of the symptomatic but non-ruptured AAA, from 5 to 26 %, compared to elective repair [75–77]. While there is general agreement on the need for urgent repair in this population, Cambria et al. helped clarify the value in a delay of surgery to medically optimize and have a full operating team available for repair of symptomatic patients who can wait [75]. More recent data have suggested that the gap between mortality in the elective versus symptomatic patients may be closing, perhaps due to increasing EVAR use and optimization of comorbid conditions (Table 3.1) [78].

Multiple clinical trials and epidemiologic studies have identified a lower prevalence of AAA in females, with the male to female ratio of 5–1 [79–81]. Despite having a lower prevalence of AAA disease, women with AAA had a fourfold increased risk for rupture in the UKSAT trial, after adjustment for age, AAA diameter, smoking status, and mean blood pressure compared to men [56]. In this study the mean diameter at time of rupture was 5 cm for women and 6 cm for men. Further meta-analysis has supported an increased rupture risk in females (Fig. 3.12) [71]. A possible explanation is that female aortas are smaller and more compliant than males; therefore a smaller aneurysm is of greater risk in the female population [82]. Another possible reason for worse outcomes in females is that they undergo elective repair at relatively larger aneurysm sizes compared to males. By indexing aortic size to body surface area, the Aortic Size Index (ASI), our group showed that women in New England are undergoing repair at larger ASI measurements compared to males for elective AAA repair [83]. ASI has already been shown to be more reliable than aneurysm diameter in predicting rupture, death, and dissection in patients with thoracic aortic aneurysms and has been incorporated into a nomogram used for prediction of rupture risk by both the Society for Thoracic Surgeons and the American College of Cardiology [85].

Currently the US Preventive Services Task Force recommends AAA screening for men aged 65–75 years with a history of smoking but recommends against screening for women who have never smoked and also states there is insufficient evidence to support screening for women who have smoked [86]. A recent Markov model, which considered the higher rupture rate in women, higher prevalence in the over-75-year-old female population, and increased lifespan of females versus males, found screening for AAA in females older than 75 years old to be cost effective [87]. Furthermore, a Medicare analysis from 1994 to 2003 found that 30–34 % of ruptured AAAs that result in death in the USA occur in women, while only 26 % of elective AAA repairs are performed in women [81]. The in-hospital mortality associated with rupture in this same analysis was 52.8 % for women and 44.2 % for men (p < .001). Given these disproportionate age-adjusted mortality figures, a reevaluation of current screening guidelines and incorporation of an adjustment for female patients, such as ASI, should be considered.

For those women who receive intervention for rAAA, further administrative studies have shown that rEVAR is less likely to be offered to females, accounting for 28–32.4 % of all repairs for females versus 44.3–46.7 % for men [81, 88]. The aortic anatomy of women that adds difficulty to performing EVAR, such as shorter aortic neck and smaller iliac vessels, may contribute to this difference [89]. As a result of these sex differences, female sex has been found to be an independent predictor of mortality during repair of both elective AAA and rAAA repair when not adjusting for ASI [81, 84]. This highlights the challenge that lies ahead in ensuring equal access and benefit from the gains that have been made with regard to the treatment of AAA disease and subsequent rupture rates.

In 1985 Cronenwett et al. was first to note the impact of blood pressure on increased rupture risk [90]. Later, the UKSAT identified higher mean blood pressure as a risk factor for rupture with a HR (95 % CI) of 1.04 (1.02–1.07) for each 1 mmHg increase in mean blood pressure [56]. Now it is accepted that the higher the pressure in the aorta the higher the wall stress and thus increased risk of rupture. The shape of the aorta also plays an important role in determining the vector of force from blood flow on the aortic wall and the surface area this energy is spread over. Interestingly, in a large meta-analysis, mean arterial pressure was found to have no effect on rate of aneurysm growth but did increase the risk of rupture for small aneurysms, HR (95 % CI) for each 10 mmHg increase was 1.32 (1.11–1.56), suggesting the mechanism is not through growth [71]. The association between a decrease in rAAA mortality and blood pressure control has been supported in population studies as well (see Fig. 3.9) [23].

A family history of AAA disease increases the chance of a first-degree relative having an AAA. In a study evaluating the family pedigree of 542 consecutive patients with AAA, 86 individuals were found to have a first-degree relative with AAA, and 40.7 % of the 86 had a history of aneurysm rupture in their family [91]. This was higher than the general population rupture rate at the time. This study also showed that the frequency of rupture increased with the number of first-degree relatives one had with AAA: 15 % with two first-degree relatives, 29 % with three, and 36 % with four or more. Another pedigree study published 6 years later found rupture rates again higher in the familial AAAs compared to sporadic, 32 % vs. 9 %, respectively, and that these ruptures tend to happen 10 years earlier in the familial group [92]. A population-based cross-sectional study from Denmark in 2008 to 2011 found similar results with a higher prevalence of AAA in individuals with a family history of AAA (6.7 % vs. 3.0 %) [93]. In addition, this study found a larger mean maximum aortic diameter in those with positive family history versus those with no family history, 20.50–19.07 mm (p < .0001). This diameter difference brings up a noteworthy limitation regarding possible confounding not adjusted for in pedigree analysis, such as diameter. That said, it is likely that family history plays an important role, but further research is needed to elucidate the reason, whether it be genetic (e.g., collagen maintenance or alpha 1-antitrypsin), lifestyle, or environment related. The reader should continue to take a thorough family history and pay attention to mention of sudden death or aneurysmal disease for any patient being evaluated for vascular disease.