Systematic Review and Meta‐Analysis of Peak Wall Stress and Peak Wall Rupture Index in Ruptured and Asymptomatic Intact Abdominal Aortic Aneurysms

Abstract

Background

Prior studies have suggested aortic peak wall stress (PWS) and peak wall rupture index (PWRI) can estimate the rupture risk of an abdominal aortic aneurysm (AAA), but whether these measurements have independent predictive ability over assessing AAA diameter alone is unclear. The aim of this systematic review was to compare PWS and PWRI in participants with ruptured and asymptomatic intact AAAs of similar diameter.

Methods and Results

Web of Science, Scopus, Medline, and The Cochrane Library were systematically searched to identify studies assessing PWS and PWRI in ruptured and asymptomatic intact AAAs of similar diameter. Random‐effects meta‐analyses were performed using inverse variance‐weighted methods. Leave‐one‐out sensitivity analyses were conducted to assess the robustness of findings. Risk of bias was assessed using a modification of the Newcastle‐Ottawa scale and standard quality assessment criteria for evaluating primary research papers. Seven case‐control studies involving 309 participants were included. Meta‐analyses suggested that PWRI (standardized mean difference, 0.42; 95% CI, 0.14–0.70; P=0.004) but not PWS (standardized mean difference, 0.13; 95% CI, −0.18 to 0.44; P=0.418) was greater in ruptured than intact AAAs. Sensitivity analyses suggested that the findings were not dependent on the inclusion of any single study. The included studies were assessed to have a medium to high risk of bias.

Conclusions

Based on limited evidence, this study suggested that PWRI, but not PWS, is greater in ruptured than asymptomatic intact AAAs of similar maximum aortic diameter.

Nonstandard Abbreviations and Acronyms

FEA

finite element analysis

PWRI

peak wall rupture index

PWS

peak wall stress

SMD

standardized mean differences

Clinical Perspective

What Is New?

• Prior studies have suggested that aortic peak wall stress and peak wall rupture index can predict abdominal aortic aneurysm rupture. However, the value of measuring peak wall stress and peak wall rupture index over simply measuring maximum abdominal aortic aneurysm diameter is unclear.

• This systematic review and meta‐analysis suggests that peak wall rupture index, but not peak wall stress, is greater in ruptured than asymptomatic intact abdominal aortic aneurysms of similar diameter.

What Are the Clinical Implications?

• The measurement of peak wall rupture index may add to that of maximum aortic diameter in assessing the risk of abdominal aortic aneurysm rupture.

Abdominal aortic aneurysm (AAA) rupture is estimated to be responsible for 200 000 deaths annually worldwide. ¹ , ² AAA rupture is thought to occur when the hemodynamic forces exceed the aortic wall strength. ¹ , ³ In clinical practice, maximum AAA diameter is the main measure used to estimate rupture risk and select patients for elective repair. ¹ Current guidelines recommend elective repair of asymptomatic large AAAs (maximum aortic diameter ≥50 mm in women and ≥55 mm in men). ¹ , ⁴ , ⁵ Approximately 1% to 2% of small asymptomatic AAAs rupture each year ⁴ and some large AAAs remain stable during a patient's lifetime, ⁶ suggesting that diameter is an imperfect measure of rupture risk.

Biomechanical imaging may provide a more precise means to estimate AAA rupture risk and select patients for repair. Finite element analysis (FEA) can noninvasively estimate the maximum tensile stress within the AAA wall (peak wall stress; PWS) and the maximum ratio between wall stress and the estimated local wall strength (peak wall rupture index; PWRI). ³ Semiautomated systems have been developed to enable clinicians without engineering backgrounds to perform FEA using computed tomography (CT) scans that are routinely performed to assess people with AAA (Figure 1). ³ , ⁷ Thus, it would be feasible to use PWS and/or PWRI in clinical practice if these measures were shown to be independent predictors of AAA rupture. Currently, however, the value of measuring PWS and PWRI over simply measuring maximum AAA diameter is unclear.

Figure 1. Examples of 3‐dimensional segmentation produced using finite element analysis from computed tomography images of patients with AAA.

The red areas indicate areas of high aortic wall stress. AAA indicates abdominal aortic aneurysm.

Previous meta‐analyses ³ , ⁸ have suggested that PWS is greater in patients with ruptured than intact AAAs; however, the generalizability of this finding is unclear owing to a number of limitations. These included lack of adjustment or matching for aortic diameter, ³ inclusion of symptomatic AAAs mixed with ruptured AAAs, ⁸ and small sample sizes. ⁹ These limitations have been addressed in more recent studies that have been reported after the publication of the most recent meta‐analysis, ⁷ , ⁸ suggesting that higher quality data are now available for an updated meta‐analysis. Furthermore, PWRI has been suggested by one, ¹⁰ but not another study, ¹¹ to be a superior measure of rupture risk than PWS. No meta‐analysis comparing PWRI in ruptured and intact AAA has been reported. ³ , ⁸ The aim of this systematic review and meta‐analysis was to provide an up‐to‐date pooled analysis of prior studies that compared PWS and PWRI in patients with ruptured and asymptomatic intact AAA of similar diameter.

Methods

Literature Search and Inclusion Criteria

The data that support the findings of this study are available from the corresponding author upon reasonable request. This systematic review and meta‐analysis was performed according to the Preferred Reporting Items for Systematic Reviews and Meta‐Analyses statement. ¹² A study protocol for this systematic review was designed (Data S1) and was not preregistered with any database. A literature search was performed using the following databases: Web of Science (via Institute for Scientific Information Web of Knowledge; 1965), Scopus (1966), Medline (via OvidSP, 1966), and The Cochrane Library to identify case‐control studies investigating PWS in patients with ruptured and diameter matched asymptomatic intact AAAs. The following search terms were applied: “peak wall stress” OR “peak wall rupture index” OR “rupture potential index” AND “abdominal aortic aneurysm.” The search was performed in October 2020 without language restrictions by one author (T.P.S.). Reference lists of primary articles and reviews were searched to increase the yield of relevant publications. Titles and abstracts were screened to identify relevant studies. If the suitability of an article was uncertain, the full text was reviewed. For inclusion in the meta‐analysis studies needed to have compared PWS or PWRI in asymptomatic intact AAAs and ruptured AAAs of similar diameter (within 3 mm mean difference between groups). Studies in which it was not possible to separate symptomatic from ruptured AAAs were excluded.

Data Extraction and Risk of Bias of the Included Studies

Data were extracted from included studies independently by 3 authors (T.S., J.M., and J.G.). The following data were collected: sample sizes for the ruptured and intact AAA group; study design; software used to perform FEA, PWS, and PWRI estimates; AAA diameter; risk factors (including age, sex, smoking history, hypertension, diabetes mellitus, ischemic heart disease; stroke; chronic obstructive pulmonary disease); and systolic blood pressure. If relevant data were not reported in the publication, the corresponding author was contacted for this information. The risk of bias was assessed independently by 3 authors (T.S., J.M., and J.G.). A quality assessment tool was designed to assess the risk of bias of the included studies adapted from 2 previously reported tools (Newcastle‐Ottawa scale and Standard Quality Assessment Criteria for Evaluating Primary Research Papers). ¹³ , ¹⁴ A number of additional aspects of the included studies relevant to this systematic review were also assessed including criteria used to define AAA rupture, method used to estimate PWS and PWRI and reproducibility reported, use of a standardized blood pressure in PWS and PWRI calculations (ie, use of a single blood pressure measurement for all participants or omission of blood pressure in calculations), inclusion of CT scan before or after rupture (for ruptured cases), matching for AAA diameter between asymptomatic intact and ruptured cases, and matching for other confounding variables. The overall risk of bias assessed within each study was assessed as low, medium, or high based on predefined criteria. Details regarding the quality assessment criteria can be found in Table S1.

Statistical Analysis

Meta‐analysis was performed using inverse variance‐weighted methods ¹⁵ in order to calculate standardized mean differences (SMD) with 95% (CI. PWS outcome data were converted from newton per square centimeter (N/cm²) to kilopascal (kPa) where required. ¹⁶ Because of anticipated interstudy heterogeneity in methods and biomechanical analyses, SMDs were calculated using random‐effects models. ¹⁷ Interstudy heterogeneity was assessed using the I² index and values <25%, between 25% to 75%, and >75% were considered to represent low, moderate, and high heterogeneity, respectively. ¹⁸ If PWS and PWRI were computed at a standardized blood pressure (ie, same blood pressure for all participants) this value was used in the meta‐analysis. One study calculated PWS and PWRI using a standardized blood pressure of 140/80 mm Hg for the main analysis and sensitivity analyses were performed using a lower (120/70 mm Hg) and higher (160/90 mm Hg) blood pressure. ⁷ For that study, results from the main analysis were used in the meta‐analysis. If studies did not use a standardized blood pressure, PWS and PWRI values computed with patient‐specific blood pressures were used. ¹¹ , ¹⁹ , ²⁰ In one study the SD of PWS was not reported and this was derived from the SE using Review Manager version 5.4 (The Cochrane Collaboration) as previously described. ³ , ¹⁷ To identify sources of heterogeneity a leave‐one‐out‐sensitivity analysis was performed by excluding individual studies one at a time and recalculating the pooled estimates for the remaining studies. Publication bias was assessed by funnel plots comparing the summary estimate of each study to its precision (1/SE) for outcomes that were reported in ≥5 studies. ²¹ Analyses were conducted using Stata version 16.1 (StataCorp LP, College Station, TX). All statistical tests were 2‐sided and a P value of <0.05 was considered significant.

Results

Study Identification

The initial database searches identified 836 studies after removal of duplicates. After title and abstract screening, the full texts of 20 studies were assessed against the inclusion criteria. Thirteen articles were excluded after full text review. Common reasons for exclusion included mismatch in AAA diameter between ruptured and intact AAAs, ²² inclusion of symptomatic but not ruptured AAAs ²² , ²³ , ²⁴ and lack of comparison of PWS or PWRI between ruptured and intact AAAs. ²⁵ Ultimately 7 studies were included (Figure 2).

Figure 2. PRISMA diagram describing the literature search.

AAA indicates abdominal aortic aneurysm; PRISMA, Preferred Reporting Items for Systematic Reviews and Meta‐Analyses; PWRI, peak wall rupture index; and PWS, peak wall stress.

Study Characteristics

A total of 309 participants with ruptured (n=139) and asymptomatic intact (n=170) AAAs of similar aortic diameter were investigated in the 7 included studies. ⁷ , ¹⁰ , ¹¹ , ¹⁶ , ¹⁹ , ²⁰ , ²⁶ All studies were of case‐control design and sample sizes ranged between 14 and 75 (see Table 1). ⁷ , ¹⁰ , ¹¹ , ¹⁶ , ¹⁹ , ²⁰ , ²⁶ Three studies were performed in Sweden ¹⁰ , ¹⁹ , ²⁶ and the remaining studies were conducted in Australia, ⁷ Spain, ¹¹ Czechia, ²⁰ and The Netherlands. ¹⁶ Six studies used the A4 Clinics 5.0 (VASCOPS GmbH, Graz, Austria) platform ⁷ , ¹⁰ , ¹¹ , ¹⁹ , ²⁰ , ²⁶ and 1 study used ABAQUS v.6.5 (Hibbit, Karlsson and Sorensen, Inc, Pawtucket, RI) for FEA. ¹⁶ One study used a combination of the A4 Clinics 5.0 and the ANSYS (Ansys Inc.) platforms. ²⁰ The inclusion criteria varied between studies. In 4 studies AAA cases were included only if the available CT scan satisfied specific imaging criteria, ⁷ , ¹¹ , ¹⁹ , ²⁶ whereas other studies did not report this as a requirement for inclusion. ¹⁰ , ¹⁶ , ²⁰ The imaging criteria used to select CT scans differed between studies. One study specifically reported excluding patients with juxtarenal or thoracoabdominal aneurysms and patients with ruptured AAAs that had massive contrast extravasation. ⁷ Another study included only participants with CT scans in which the aorta was visible from the renal arteries to the iliac bifurcation and the lumen was distinguishable from intraluminal thrombus. ¹¹ One study required CT scans to have a sufficiently high out‐of‐plane image resolution with good visibility of the exterior aneurysm surface. ²⁶ In another study, only participants with good quality CT scans were included; however, the criteria used to determine this were not reported. ¹⁹ All studies either matched cases and controls for aortic diameter or included cases and controls with similar mean aortic diameter (within 3 mm difference; see Tables 1 and 2). Three studies used a standardized blood pressure to compute PWS or PWRI in all participants, ⁷ or matched cases and controls for blood pressures, ²⁶ or omitted blood pressure from calculations. ¹⁰ The remaining studies used patient‐specific blood pressures although the relationship between their measurement and the timing of CT scan varied across studies (Table 1). For ruptured AAA cases, blood pressure readings before rupture were frequently used. ¹¹ , ¹⁶ , ²⁶ For participants with asymptomatic intact AAAs, measurements were either taken from the same hospital visit ²⁶ in which the CT scan was performed or from a prior visit. ¹¹ , ¹⁶ The timing of blood pressure measurements in relation to CT scans were not reported in 2 studies. ¹⁹ , ²⁰ Three studies reported the reproducibility of their FEA estimates (Table 1). ⁷ , ¹⁹ , ²⁰

Table 1.

Characteristics of Case Control Studies Comparing PWS and PWRI Between Ruptured and Asymptomatic Intact AAAs of Similar Aortic Diameter

Study	Total Sample Size	Sample Size (Asymptomatic Intact: Ruptured)*	Biomechanical Measurements	Software Used	Variables That Were Balanced Between Asymptomatic iAAA and rAAAs	Blood Pressure Used in PWS or PWRI Analysis	Reproducibility of Biomechanical Measurements
Singh et al (2020) 7	75	75 (50:25)	PWS, PWRI	A4 Clinics 5.0 (VASCOPS GmbH, Graz, Austria).	Diameter and blood pressure	Standardized blood pressure (140/80 mm Hg) used and sensitivity analysis with lower and higher blood pressures (120/70 mm Hg and 160/90 mm Hg) for all participants	Intraobserver reproducibility; CV 2.7% and 4.7% for PWS in iAAA and rAAAs respectively
Siika et al (2019) 10	283	60 (40:20)	PWS, PWRI	A4 Clinics 5.0 (VASCOPS GmbH, Graz, Austria)	Diameter, sex, age and blood pressure	Blood pressure omitted from analyses	NR
Siika et al (2018) 19	90	43 (15:28)	PWS, PWRI	A4 Clinics 5.0 (VASCOPS GmbH, Graz, Austria)	Diameter	Patient‐specific blood pressure; timing of blood pressure measurement in relation to CT scan not reported	Intraobserver reproducibility; Mean % difference of 6.86±6.46 and 7.70±6.26 for PWS and PWRI respectively. Interobserver reproducibility; mean % difference 7.09±6.16 and 9.47±8.18 †
Leemans et al (2018) 11	175	62 (31:31)	PWS, PWRI, rupture risk equivalent diameter	A4 Clinics 5.0 (VASCOPS GmbH, Graz, Austria).	Diameter	Patient‐specific blood pressure obtained from the last measurement in a noncritical setting within 1 year before presentation	NR
Polzer et al (2015) 20	14	14 (7:7)	PWRI and probabilistic rupture risk index	A4 Clinics 5.0 (VASCOPS GmbH, Graz, Austria) and ANSYS (Ansys Inc.)	Diameter	Patient‐specific blood pressure; timing of blood pressure measurement not reported	Intraobserver reproducibility; CV <5.5% for PWS and PWRI in participants with iAAAs. Interobserver reproducibility; intraclass correlation coefficient 0.98 (range 0.97–0.99) for PWS and PWRI †
Gasser et al (2010) 26	50	35 (17:18)	PWS, PWRI	A4 Clinics 5.0 (VASCOPS GmbH, Graz, Austria)	Diameter and blood pressure	Participants were matched for blood pressure between groups. Blood pressure measurements obtained in the same admission in which CT scan was performed or earlier hospital or healthcare center visit (for ruptured cases)	NR
Truijers et al (2007) 16	30	20 (10:10)	PWS	ABAQUS v.6.5 (Hibbit, Karlsson and Sorensen, Inc, Pawtucket, RI)	Diameter and blood pressure	Patient‐specific blood pressure obtained from a year before CT scan; Sensitivity analysis also performed in which a standardized blood pressure (120 mm Hg systolic) was used	NR

AAA indicates abdominal aortic aneurysm; CT, computed tomography; CV, coefficient of variation; iAAA, asymptomatic intact AAA; NR, not reported; PWRI, peak wall rupture index; PWS, peak wall stress; and rAAA, ruptured AAA.

^*Sample sizes reported are reflective of the cases and control that were similar in AAA diameter and excluded symptomatic AAA cases.

^†Reported in an external publication.

Table 2.

Clinical Characteristics of Participants With Ruptured and Asymptomatic Intact AAAs of Similar Aortic Diameter

Study	Group	Number	Age, y	Male	Diabetes Mellitus	Ischemic Heart Disease	Stroke	Chronic Obstructive Pulmonary Disease	Smoking	Systolic Blood Pressure (mm Hg)
Singh et al (2020) 7	iAAA	50	72±7	94	18	52	10	24	94	140*
	rAAA	25	73±7	72	13	38	0	13	88	140*
Siika et al (2019) 10	iAAA	40	78±7	60	NR	NR	NR	NR	NR	NR †
	rAAA	20	79±7	55	NR	NR	NR	NR	NR	NR †
Siika et al (2018) 19	iAAA	15	75±8	87	NR	NR	NR	NR	NR	NR
	rAAA	28	76±10	75	NR	NR	NR	NR	NR	NR
Leemans et al (2018) 11	iAAA	31	NR	NR	NR	NR	NR	NR	NR	NR
	rAAA	31	NR	NR	NR	NR	NR	NR	NR	NR
Polzer et al (2015) 20	iAAA	7	NR	NR	NR	NR	NR	NR	NR	132±8
	rAAA	7	NR	NR	NR	NR	NR	NR	NR	152±26
Gasser et al (2010) 26	iAAA	17	75±8	78	NR	NR	NR	NR	NR	NR ‡
	rAAA	18	76±11	78	NR	NR	NR	NR	NR	NR ‡
Truijers et al (2007) 16	iAAA	10	72±2	90	10	70	20	30	40	120*
	rAAA	10	70±2	70	10	30	20	20	40	120*

Values are expressed as mean±SD, median [interquartile range] or n (%). AAA indicates abdominal aortic aneurysm; iAAA, asymptomatic intact AAA; NR, not reported; and rAAA, ruptured AAA.

^*A standardized blood pressure was used for biomechanical analyses.

^†Blood pressure was omitted from patient‐specific parameters.

^‡iAAAs and rAAAs were matched for blood pressure.

Participant Characteristics

The participant characteristics are summarized in Table 2. The average age of participants ranged between 70 and 79 years. ⁷ , ¹⁰ , ¹⁶ , ¹⁹ , ²⁶ There were no significant differences in the average age of participants between asymptomatic intact and ruptured AAA groups in the 3 studies that statistically assessed this. ⁷ , ¹⁰ , ¹⁹ The proportions of men in the asymptomatic intact and ruptured groups were 60% to 94% and 55% to 78% respectively. One study included a significantly larger proportion of women in the ruptured AAA group, ⁷ whereas 2 studies reported no significant differences in sex between groups. ¹⁰ , ¹⁹ The remaining studies either did not report sex ¹¹ , ²⁰ or did not statistically compare this. ¹⁶ , ²⁶ Details regarding diabetes mellitus, ischemic heart disease, stroke, chronic obstructive pulmonary disease, smoking, and blood pressure were reported in only 2 studies (Table 2).

Risk of Bias Assessment

The methodological quality assessment and overall risk of bias of the included studies are reported in Figure 3. Six studies were assessed to have a high risk of bias, ¹⁰ , ¹¹ , ¹⁶ , ¹⁹ , ²⁰ , ²⁶ and 1 study was assessed to have a medium risk of bias. ⁷ Six studies were of retrospective design, and the design of 1 study was unclear. ²⁰ Only 1 study used an objective definition of AAA rupture, which was defined as the presence of blood in the retroperitoneum or peritoneum identified on CT by a consultant vascular specialist. ⁷ The method of estimating PWS and PWRI was well described in 3 out of the 7 studies that included the reporting of the reproducibility of the method ⁷ within the same or a previous publication (see Figure 3). ¹⁹ , ²⁰ Only 2 of these studies assessed intra‐ and interobserver reproducibility (see Table 1). ¹⁹ , ²⁰ One study assessed the reproducibility of methods in both asymptomatic and ruptured AAAs (coefficients of variation 2.7% and 4.7% for PWS in asymptomatic intact AAA and ruptured AAAs respectively ⁷ ), whereas in the other 2 studies reproducibility was assessed in asymptomatic intact AAAs only. Six studies matched ruptured and asymptomatic intact AAA cases by AAA diameter, ⁷ , ¹⁰ , ¹¹ , ¹⁹ , ²⁰ , ²⁶ whereas in one study participants were not matched; however, the mean diameter between intact and ruptured cases was similar (51±2 versus 53±2 mm respectively). ¹⁶ Three studies used a standardized blood pressure to calculate PWS and PWRI whereas the other 3 studies ¹¹ , ¹⁹ , ²⁰ used patient‐specific blood pressures. One study matched participants for age and sex ¹⁰ in addition to AAA diameter. An a priori sample size calculation was performed in only 1 study. ⁷

Figure 3. Strengths and weaknesses of studies included in this systematic review.

The yellow and red colored cells represent criteria, which were and not met in each study respectively. For the sample size criterion, red colored cells represent studies that had a sample size <100 and yellow‐colored cells represent studies that had sample sizes >100. The green colored cells represent criteria that were partially met in each study. A blue colored cell was used if it was unclear whether a criterion was met by a study. AAA indicates abdominal aortic aneurysm; CT, computed tomography; PWRI, peak wall rupture index; and PWS, peak wall stress. *AAA rupture cases were identified using an International Classification; **cases and controls matched for age and sex; ^†reproducibility reported in an external publication; ^††cases and controls matched for blood pressure; ^‡cases and controls were not matched by study design although AAA diameter was similar between groups; ^‡‡in this study blood pressure was omitted from biomechanical calculations; ^#PWS not assessed in this study.

Reported Association of PWS and PWRI With AAA Rupture

The mean aortic diameter of included patients ranged between 51 to 82 mm and 53 to 82 mm in included asymptomatic intact and ruptured AAAs respectively (see Table 3). No significant differences in PWS between groups were reported, although in one study PWS was not assessed ²⁰ and another study did not statistically compare PWS between groups in the matched participants. ²⁶ PWRI was significantly higher in ruptured AAAs than asymptomatic intact AAAs in 2 studies. ¹⁰ , ¹⁹ PWRI was higher in the remaining studies that assessed this ⁷ ; however, differences were not statistically significant (see Table 3). ⁷ , ¹¹

Table 3.

Comparison of PWS and PWRI of Participants With Ruptured and Asymptomatic Intact AAAs of Similar Aortic Diameter

Study	Group	Number	Diameter (mm)	P Value	PWS (kPa)	P Value	PWRI	P Value
Singh et al (2020) 7	iAAA	50	82±14	0.906	263.8±69.4	0.192	1.09±0.52	0.982
	rAAA	25	82±13		279.8±90.5		1.20±0.76
Siika et al (2019) 10	iAAA	40	53±5	0.319	197.0±40.3	0.162	0.35±0.08	0.016
	rAAA	20	55±5		216.3±45.3		0.43±0.11
Siika et al (2018) 19	iAAA	15	73±11	0.674	284±53.4*	0.194	0.48±0.11*	<0.001
	rAAA	28	74±12		249±53.9*		0.80±0.54*
Leemans et al (2018) 11	iAAA	31	71±15	0.81	261±89 †	0.99	0.69±0.33	0.61
	rAAA	31	72±18		262±75 †		0.70±0.27
Polzer et al (2015) 20	iAAA	7	73±11	NR	NR		0.48±0.41 ‡	NR
	rAAA	7	76±14		NR		0.69±0.41 ‡
Gasser et al (2010) 26	iAAA	17	75±12	NR	292.0±108.7	NR	0.61±0.26	NR
	rAAA	18	77±15		330.8±114.2		0.74±0.29
Truijers et al (2007) 16	iAAA	10	51±2	0.57	317±73 †	0.30	NR
	rAAA	10	53±2		367±126 †		NR

Values are expressed as mean±SD unless indicated otherwise. AAA indicates abdominal aortic aneurysm; iAAA, asymptomatic intact AAA; kPa, kilopascal; NR, not reported; PWRI, peak wall rupture index; PWS, peak wall stress; and rAAA, ruptured AAA.

^*SDs not available and were imputed from the diameter mismatched analysis reported in the same study.

^†PWS converted from newton per square centimeter to kPa.

^‡Derived PWRI values reported that have been divided by the mean arterial pressure inflation factor used in the study to obtain comparable results.

Data Synthesis

In the meta‐analysis, PWS was not significantly different between ruptured and asymptomatic AAAs (SMD, 0.13; 95% CI, −0.18 to 0.44; P=0.418; Figure 4). Moderate heterogeneity was observed (I²=40.6%). In contrast, PWRI was significantly higher in participants with ruptured compared with asymptomatic intact AAA (SMD, 0.42; 95% CI, 0.14–0.70, P=0.004; Figure 5). Interstudy heterogeneity was low (I²=25.5%). Leave‐one‐out sensitivity analysis suggested that the results of the meta‐analyses were not dependent on the inclusion of any single study (Figures S1 and S2). The funnel plot for PWRI appeared asymmetrical (Figure S3) suggesting potential publication bias.

Figure 4. Differences in peak wall stress in ruptured and asymptomatic intact AAAs.

The SMD is the mean difference between both groups, standardized to 1 SD difference in PWS (kilopascal) within that study. The summary SMD is estimated from inverse variance‐weighted meta‐analysis. Box areas are inversely proportional to the variance of the SMD and horizontal lines illustrate 95%CIs. AAA indicates abdominal aortic aneurysm; PWS, peak wall stress; and SMD, standardized mean difference.

Figure 5. Differences in peak wall rupture index in ruptured and asymptomatic intact AAAs.

The SMD is the mean difference between both groups, standardized to 1 SD difference in PWRI within that study. The summary SMD is estimated from inverse variance‐weighted meta‐analysis. Box areas are inversely proportional to the variance of the SMD and horizontal lines illustrate 95% CIs. AAA indicates abdominal aortic aneurysm; CT, computed tomography; PWRI, peak wall rupture index; and SMD, standardized mean difference.

Discussion

This meta‐analysis suggested that PWRI, but not PWS, is greater in ruptured than asymptomatic intact AAAs of similar diameter. This finding is in contrast with a previous meta‐analysis that reported greater PWS in ruptured than intact AAAs. ³ A major limitation of the previous meta‐analysis was the mismatch in aortic diameter between groups and inclusion of symptomatic patients in the ruptured group. Participants with symptomatic intact AAAs were not included in the current study as their risk of rupture is uncertain. ⁵ , ⁹

Maximum aortic diameter is currently the most established measure of AAA rupture risk. ¹ , ⁵ , ²⁷ There are, however, a number of limitations in using aortic diameter in clinical practice, in particular the measurement error, which may be greater than the annual change in diameter. ²⁸ , ²⁹ Additional methods of estimating rupture risk and determining management may therefore be valuable. The findings of this study suggest that measurement of PWRI may add to aortic diameter in assessing the risk of AAA rupture. There are, however, many limitations of this technology that need to be addressed. There is currently no standardized approach to conducting FEA. There was substantial variation in the approach used to incorporate blood pressures in the calculation of PWS and PWRI in the included studies. Some studies used an arbitrary blood pressure for all participants, whereas others used patient‐specific blood pressures. It is currently unclear which approach is most appropriate. Additionally, wall thickness and strength have an important effect on the risk of aortic rupture and prior investigations have suggested that increased aortic wall thickness is associated with reduced aortic wall stress. ²⁶ , ³⁰ Currently there is no accurate and feasible method to estimate wall thickness from imaging. ⁷ Six of the 7 studies used FEA software to estimate PWRI using the same formula that was derived from prior tensile testing of human AAA wall specimens ex vivo, but this may not be representative of the situation in individual patients in vivo. ³⁰ Aortic calcification has previously been suggested to have an important influence on biomechanical forces but there remains no standardized method of including this in estimations of wall stress. ³¹ , ³²

Although the current meta‐analysis suggested that PWRI is likely to be higher in ruptured AAAs compared with asymptomatic intact AAAs of similar diameter, the confidence in this finding is lessened as the included studies were assessed to have either a medium or high risk of bias because of a number of design limitations. First, studies included participants with CT scans performed after rupture and it is likely that the biomechanical forces before rupture were different. Second, some studies used patient‐specific blood pressures to perform biomechanical analyses rather than a standardized blood pressure. ¹¹ , ¹⁹ , ²⁰ This may have contributed to heterogeneity and led to under‐ or overestimation of PWS and PWRI. Although patients with asymptomatic intact and ruptured AAAs had similar aortic diameter, other characteristics were generally poorly reported and confounding owing to an unmeasured factor cannot be ruled out. Additionally, the CT scans of ruptured AAA cases were required to meet certain inclusion criteria in some studies and selection bias cannot be excluded. ⁷ , ¹¹ , ¹⁹ , ²⁶ We were unsuccessful in contacting the corresponding author of 2 studies ¹⁰ , ¹⁹ to clarify whether there was an overlap in participants included in these investigations. Nevertheless, the leave‐one‐sensitivity analysis suggested that the findings of the PWRI meta‐analysis was not materially altered with individual omission of either of these studies. ¹⁰ , ¹⁹ Lastly the relevance of the findings of this meta‐analysis to small AAAs is limited as 5 studies included only patients with large AAAs ⁷ , ¹¹ , ¹⁹ , ²⁰ , ²⁶ (mean±SD aortic diameter [mm] ranged between 71±15 and 82±14 for the asymptomatic intact AAAs; 72±18 and 82±13 for the ruptured AAAs). Furthermore, this meta‐analysis compared PWS and PWRI in individuals with asymptomatic intact and ruptured AAAs but did not examine the predictive ability of these biomechanical measures for AAA rupture. Investigating this would require a large observational study; however, owing to the low rupture rate of small AAAs and the high repair rate of large AAAs, such a study maybe infeasible to perform.

Conclusions

In conclusion the results of this study suggest that PWRI is greater in ruptured than asymptomatic intact AAAs of similar diameter. The findings suggest the potential value of biomechanical measures in estimating AAA rupture risk accepting the medium to high risk of bias of the included studies.

Sources of Funding

The Townsville Hospital and Health Services Study, Education and Research Trust Fund and The Queensland Government supported this work. Jonathan Golledge holds a Practitioner Fellowship from the National Health and Medical Research Council (1117061) and a Senior Clinical Research Fellowship from the Queensland Government, Australia. Joseph V. Moxon holds an Advance Queensland Mid‐Career fellowship from the Queensland Government. Singh holds a Junior Doctor Research Fellowship from the Queensland Government.

Disclosures

T. Christian Gasser is a scientific advisor for VASCOPS GmbH. The remaining authors have no disclosures to report.

Supporting information

Data S1

Table S1

Figures S1–S3

(J Am Heart Assoc. 2021;10:e019772. DOI: 10.1161/JAHA.120.019772.)