Oral Steroid Use and Abdominal Aortic Aneurysm Expansion　― Positive Association ―

Abstract

Background: The maximum axial diameter (MAD) of a fusiform abdominal aortic aneurysm (AAA) is an indicator of the risk of expansion or rupture. Apart from smoking and MAD itself, few expansion risk factors have been reported. In this study, we investigated expansion risk factors for AAA.

Methods and Results: This retrospective cohort study included 176 patients who attended Tohoku University Hospital with infrarenal fusiform AAA. AAA expansion rate was determined on multidetector computed tomography, and the correlations between expansion rate and the clinical data were analyzed. The median expansion rate was 2.405 mm/year. On univariate analysis, a significant positive correlation with expansion rate was observed for the initial MAD (P<0.001) and significant negative correlations for oral angiotensin receptor blocker usage (P=0.025), height (P=0.005), body weight (P=0.017), total cholesterol (P=0.007), low-density lipoprotein cholesterol (P=0.004), and HbA1c (P=0.037). On logistic regression analysis, significant positive associations with expansion rate were observed for initial MAD (P<0.001) and oral steroid usage (P=0.029) and a negative association for height (P=0.041).

Conclusions: Oral steroid usage is an important risk factor for AAA expansion, independent of other risk factors of atherosclerosis and MAD.

Abdominal aortic aneurysm (AAA) is a pathological condition in which the abdominal aorta has a diameter >30 mm, or 1.5-fold greater than normal at the level of the renal artery.¹ Most AAA are asymptomatic, but once rupture occurs, the condition has a mortality rate of 65–85% and is ranked as the 16th most prominent cause of death in the USA in individuals aged ≥65 years.^2,3 In Europe and the USA, the prevalence of the condition is 2% in men aged ≥65 years, with major risk factors for morbidity including male sex, age, history of smoking, and Caucasian race.^3,4 The etiology in the majority of AAA is considered to be atherosclerosis, but a causal relationship with the typical risk factors for atherosclerosis, such as hypertension, dyslipidemia, and diabetes mellitus, remains unclear.⁵ Indeed, a negative correlation with diabetes mellitus has been shown, suggesting that AAA is a pathological state distinct from atherothrombosis.^4–6

The only treatment option for this condition is surgery, such as vascular prosthesis implantation or endovascular aneurysm repair (EVAR); so far, no medical therapy has been established.^7–9 Surgical treatment is highly invasive, and even elective surgery is associated with a 30-day mortality rate of 5.4%.⁷ Surgical treatment is therefore indicated only for cases in which the risk of rupture is higher than the operative mortality rate.

The maximum axial diameter (MAD) of the aneurysm is an indicator of rupture risk in fusiform AAA; the greater the MAD, the higher the risk of rupture.^7,8,10 In Europe and the USA, comparisons of the rate of rupture and postoperative mortality rate have led to the surgical indication of MAD ≥55 mm; in Japan, the common surgical indication is ≥50 mm, which takes body size into consideration.^7,8,11 For women, the surgical indication is sometimes taken as MAD ≥45 mm because the rupture risk in women has been reported to be 3-fold higher than in men.¹¹ Patients with smaller AAA (MAD <50 mm), for whom surgery is not indicated, are periodically followed up. A problem during follow-up, however, is that expansion of the AAA cannot be predicted because the expansion rate differs between patients.¹² Various factors for expansion have been investigated, such as hypertension, dyslipidemia, diabetes mellitus (and the medication for these conditions), smoking, sex, coronary artery disease (CAD), cerebrovascular disease (CVD), calcification of the arterial aneurysmal wall, and mural thrombosis, but no clear causal relationships with any of these factors have been demonstrated.^12–16 Currently, the only obvious risk factors for expansion are MAD itself and continued smoking.^9,10 The increase in MAD during a given observation period is greater in cases of larger initial MAD.¹⁰ Confirmation of the risk factors for expansion would be helpful when deciding on surgical indication and ultimately could contribute to the development of non-surgical therapeutic means to prevent AAA expansion and rupture in a minimally invasive manner.

The aim of this study was therefore to measure AAA expansion rate (mm/year) on multidetector computed tomography (CT) in order to identify risk factors for AAA expansion.

Methods

Ethics

The study protocol was approved by the Tohoku University School of Medicine ethics committee (receipt number: 2016-1-481) and registered in the University hospital Medical Information Network (UMIN) Clinical Trials Registry (ID: UMIN000026309).

Subjects

Of the patients who had either undergone surgery for AAA or who were being followed up for small-diameter AAA at Tohoku University Hospital from April 2012 to March 2016, we included in this retrospective study 176 patients with infrarenal and fusiform AAA who had undergone CT at least twice at an interval >100 days. We excluded patients with inflammatory, infectious, or saccular AAA, patients with inherited vascular disorders, such as Marfan syndrome, and patients on dialysis. All subjects were Japanese.

Measurement

We defined MAD as the maximum short axis diameter of AAA. MAD was measured using axial sections on CT in all subjects. Measurements were rounded to the nearest 0.1 mm. All measurements were performed at Tohoku University Hospital by the same vascular surgeon using a radiographic film viewer. When multiple CT had been done, the earliest and latest scans were taken as the initial CT and the follow-up CT, respectively. In all patients, follow-up CT was obtained between April 2012 and March 2016, but some of the initial CT were obtained prior to this. To exclude interobserver variability, we performed correlation analysis and calculated the intraclass correlation coefficient (ICC) between the MAD measured by the present authors and that measured by the radiologists, with both blinded to each other’s measurements (Figure S1; Table S1). The resulting correlation coefficient was 0.982 (P<0.001), and the ICC was 0.966 (95% CI: 0.798–0.987). To test the intra-observer reliability, we performed correlation analysis and calculated ICC between the present MAD and that measured by the same investigator after a 6-month interval (Figure S2; Table S2). The resulting correlation coefficient was 0.997 (P<0.001), and the ICC was 0.997 (95% CI: 0.996–0.997). Thus, the present MAD measurements had sufficient reliability.

The expansion rate was defined as the difference in MAD between the follow-up and initial CT divided by the interval between them, and is expressed as mm/year. The parameters evaluated were age, sex, history of smoking, medication history (anti-platelet agents, antihypertensives, and cholesterol-lowering agents), history of CAD and CVD, physiological tests (ankle brachial pressure index [ABI] and spirometry), and laboratory data. Age was defined as that at the time of the initial CT. The other data were obtained within 2 months of the follow-up CT. All examinations were conducted at Tohoku University Hospital. Some of the data were missing because this study was retrospective in nature, hence the total number of subjects analyzed for each parameter is stated in the tables.

Statistical Analysis

We performed univariate analysis using expansion rate as the objective variable. Spearman’s rank correlation coefficients were used for continuous variables and Wilcoxon rank sum test for 2-category variables.

We divided the subjects into 2 groups, rapid expansion and slow expansion, according to whether the expansion rate was greater or smaller than the median of 2.405 mm/year (median-split procedure). We then performed multivariate analysis by means of logistic regression analysis (complete-case analysis). Parameters with P<0.20 on univariate analysis were further screened using the stepwise method (input probability, 0.2; removal probability, 0.2), and the remaining parameters were used as explanatory variables: these were initial diameter, height, percent predicted forced expiratory volume in 1.0 s, total cholesterol, glycated hemoglobin (HbA1c), and oral steroid usage. Other parameters excluded in this process were age, weight, creatinine, low-density lipoprotein cholesterol (LDL-C), angiotensin-converting enzyme inhibitor, and angiotensin receptor blockers (ARB). We confirmed the absence of multicollinearity and complete isolation among the explanatory variables.

We compared the oral steroid usage group with the other patients on univariate analysis. Wilcoxon rank sum test was used for continuous variables, and the Fisher exact test was used for 2-category variables.

All tests were non-parametric and 2-sided with a significance level of P<0.05. JMP 12.2 Pro (SAS Institute Japan) was used for analysis.

Results

MAD and Expansion Rates

The median initial MAD was 38.4 mm (IQR, 32.5–44.9 mm; Figure 1A), median follow-up MAD was 49.3 mm (IQR, 42.7–52.7 mm; Figure 1B), median expansion rate was 2.405 mm/year (IQR, 1.53–3.77 mm/year), and the mean expansion rate was 2.84±1.88 mm/year (Figure 1C). The median interval between scans was 925 days (IQR, 489–2,030 days), the minimum interval was 129 days, and the maximum interval was 5,743 days.

Figure 1.

Distribution of (A,D) initial maximum axial diameter (MAD), (B,E) follow-up MAD, and (C,F) expansion rate in patients with abdominal aortic aneurysm according to (A–C) expansion rate (, slow expansion rate [<2.405 mm/year]; ■, rapid expansion rate [>2.405 mm/year]) and (D–F) () oral steroid use (■, no steroid use). Box plots: vertical line, median expansion rate; left and right borders of the box, 25th and 75th percentiles; left whisker, minimum; right whisker, 3rd quartile+1.5 IQR; diamond, mean and 95% CI.

Clinical Characteristics and Univariate Analysis

Clinical patient characteristics are listed in Tables 1,2. A large proportion of the patients were male (84.1%), had smoked at some time (86.8%), and used antihypertensives (84.3%). Median age at initial CT was 72.3 years, and >75% of the patients were aged >65 years. The median blood pressure was slightly high (median systolic blood pressure, 135 mmHg; median diastolic blood pressure, 85 mmHg; mean blood pressure, 106 mmHg). In many patients, however, body mass index, ABI, total cholesterol, and HbA1c were within normal limits.

Table 1. Baseline Characteristics and Univariate Indicators of AAA Expansion: Categorical Data

Variable	No. patients		n (%)	Expansion rate (mm/year)	P-value†
Sex	176	Male	148 (84.1)	2.35 (1.44–3.77)	0.331
		Female	28 (15.9)	2.51 (1.95–3.80)
Smoking	174	Ever	151 (86.8)	2.31 (1.46–3.78)	0.337‡
		Current smoker	27 (15.5)	3.15 (1.91–4.14)	0.080§
		Ex-smoker	124 (71.3)	2.20 (1.35–3.60)
		Never	23 (13.2)	2.57 (1.98–3.63)
Family history	176	Yes	8 (4.5)	1.77 (1.30–3.56)	0.495
		No	168 (95.5)	2.42 (1.55–3.77)
Oral steroids	168	Yes	15 (8.9)	3.72 (1.95–5.56)	0.069
		No	153 (91.1)	2.39 (1.55–3.61)
Antiplatelet agents	174	Yes	71 (40.8)	2.56 (1.43–4.15)	0.692
		No	103 (59.2)	2.41 (1.54–3.63)
Antihypertensives	172	Yes	145 (84.3)	2.41 (1.54–3.88)	0.680
		No	27 (15.7)	2.48 (1.63–3.78)
Ca blockers	171	Yes	115 (67.3)	2.57 (1.58–4.06)	0.299
		No	56 (32.7)	2.30 (1.49–3.21)
β-blockers	169	Yes	42 (24.9)	2.14 (1.49–4.07)	0.561
		No	127 (75.1)	2.51 (1.56–3.72)
ACEI	169	Yes	20 (11.8)	3.21 (1.43–4.27)	0.167
		No	149 (88.2)	2.36 (1.57–3.62)
ARB	170	Yes	79 (46.5)	2.09 (1.51–3.51)	0.025
		No	91 (53.5)	2.56 (1.63–4.29)
α-blockers	169	Yes	11 (6.5)	2.21 (1.58–4.31)	0.959
		No	158 (93.5)	2.47 (1.56–3.80)
Antidiabetic agents	169	Yes	37 (21.9)	1.95 (1.24–3.47)	0.065
		No	132 (78.1)	2.57 (1.61–3.89)
Statins	169	Yes	75 (44.4)	2.41 (1.62–4.29)	0.394
		No	94 (55.6)	2.47 (1.50–3.61)
CAD	169	Yes	53 (31.4)	2.55 (1.61–4.03)	0.789
		No	116 (68.6)	2.42 (1.55–3.70)
CVD	169	Yes	29 (17.2)	2.87 (1.63–4.72)	0.333
		No	140 (82.8)	2.41 (1.57–3.71)

^†Wilcoxon rank sum test. ^‡Ever-smokers vs. never smokers. ^§Current smokers vs. ex-smokers. AAA, abdominal aortic aneurysm; ACEI, angiotensin-converting enzyme inhibitor; ARB, angiotensin II receptor blocker; CAD, coronary artery disease; CVD, cerebrovascular disease.

Table 2. Baseline Characteristics and Univariate Indicators of AAA Expansion: Continuous Data

Variable	No. patients	Median (IQR)	Correlation coefficient	P-value†
Initial MAD (mm)	176	38.4 (32.5–44.9)	0.442	<0.001
Age (years)	176	72.3 (65.9–78.0)	0.156	0.039
Height (cm)	172	164.3 (158.0–169.0)	−0.212	0.005
Weight (kg)	172	63.0 (55.0–70.0)	−0.181	0.017
BMI (kg/m2)	172	23.5 (21.6–25.2)	−0.094	0.222
SBP (mmHg)	167	135 (126–149)	−0.001	0.994
DBP (mmHg)	167	85 (79–92)	0.072	0.357
Mean BP (mmHg)	167	106 (98–114)	−0.020	0.794
ABI‡	167	1.04 (0.96–1.14)	0.074	0.339
%VC	168	94.40 (84.45–107.23)	0.076	0.326
FEV1.0% (%)	168	71.95 (65.45–77.87)	0.106	0.171
Serum BUN (mg/dL)	175	17 (13–21)	−0.056	0.459
Creatinine (mg/dL)	175	0.94 (0.82–1.19)	0.103	0.174
Triglyceride (mg/dL)	169	114 (83–160)	−0.088	0.253
TC (mg/dL)	174	172 (156–198)	−0.202	0.007
LDL-C (mg/dL)	168	105 (82–123)	−0.224	0.004
HDL-C (mg/dL)	168	45 (37–56)	−0.028	0.722
HbA1c (NGSP) (%)	169	5.8 (5.5–6.1)	−0.160	0.037

^†Spearman’s rank correlation coefficient. ^‡Mean of left and right. %VC, percent predicted vital capacity; AAA, abdominal aortic aneurysm; ABI, ankle brachial pressure index; BMI, body mass index; BP, blood pressure; BUN, blood urea nitrogen; DBP, diastolic blood pressure; FEV1.0%, percent predicted forced expiratory volume in 1.0 s; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; MAD, maximum axial diameter; SBP, systolic blood pressure; TC, total cholesterol.

Results of the univariate analysis are also given in Tables 1,2. A significant positive correlation with expansion rate was observed for the initial MAD (P<0.001), and age (P=0.039). The expansion rate tended to be greater in the oral steroids group (P=0.069). Significant negative correlations were observed for use of oral ARB (P=0.028), height (P=0.005), body weight (P=0.017), total cholesterol (P=0.007), LDL-C (P=0.004), and HbA1c (P=0.037).

No significant correlation was observed between the ever smokers and the never smokers, but the expansion rate in the current smokers tended to be greater than in the ex-smokers (P=0.080).

Clinical Characteristics According to Expansion Rate

The clinical characteristics according to expansion rate (rapid vs. slow) are listed in Table 3. The median expansion rate in the rapid expansion group was 3.75 mm/year (IQR, 3.03–4.74 mm/year), and that in the slow expansion group was 1.51 mm/year (IQR, 1.10–1.91 mm/year).

Table 3. Subject Characteristics vs. AAA Expansion Speed

Variable	Rapid expansion (Growth rate >2.405 mm/year)		Slow expansion (Growth rate <2.405 mm/year)
Categorical variables	n (%)	Total	n (%)	Total
Sex (M/F)	73 (83.0)/15 (17.0)	88	75 (85.2)/13 (14.8)	88
Smoking (Ever/Never)	71 (82.6)/15 (17.4)	86	80 (90.9)/8 (9.1)	88
Family history (Yes/No)	3 (3.4)/85 (96.6)	88	5 (5.7)/83 (94.3)	88
Oral steroids (Yes/No)	10 (11.8)/75 (88.2)	85	5 (6.0)/78 (94.0)	83
Antiplatelet agents (Yes/No)	36 (40.9)/52 (59.1)	88	35 (40.7)/51 (59.3)	86
Antihypertensives (Yes/No)	73 (83.0)/13 (17.0)	86	72 (85.7)/11 (14.3)	83
Ca blockers (Yes/No)	60 (69.0)/27 (31.0)	87	55 (65.5)/29 (34.5)	84
β-blockers (Yes/No)	18 (20.9)/68 (79.1)	86	24 (28.9)/59 (71.1)	83
ACEI (Yes/No)	13 (15.1)/73 (84.9)	86	7 (8.4)/76 (91.6)	83
ARB (Yes/No)	36 (41.9)/50 (58.1)	86	43 (51.2)/41 (48.8)	84
α-blockers (Yes/No)	4 (4.7)/82 (95.3)	86	7 (8.4)/76 (91.6)	83
Anti-diabetic agents (Yes/No)	15 (17.4)/71 (82.6)	86	22 (26.5)/61 (73.5)	83
Statins (Yes/No)	38 (44.2)/48 (55.8)	86	37 (44.6)/46 (55.4)	83
CAD (Yes/No)	27 (31.4)/59 (68.6)	86	26 (31.3)/57 (68.7)	83
CVD (Yes/No)	17 (19.5)/70 (80.5)	87	12 (14.6)/70 (85.4)	82
Continuous variables	Median (IQR)	Total	Median (IQR)	Total
Expansion rate (mm/year)	3.75 (3.03–4.74)	88	1.54 (1.10–1.91)	88
Initial MAD (mm)	42.3 (37.2–46.7)	88	35.7 (31.1–40.6)	88
Age (years)	74.9 (68.6–78.5)	88	69.8 (63.8–77.0)	88
Height (cm)	163.0 (157.0–167.0)	87	165.0 (160.5–170.0)	85
Weight (kg)	60.9 (53.5–68.0)	87	64.0 (58.8–72.0)	85
BMI (kg/m2)	23.3 (21.4–24.9)	87	23.7 (21.9–25.3)	85
SBP (mmHg)	136 (126–150)	85	135 (126–149)	82
DBP (mmHg)	85 (80–92)	85	85 (79–90)	82
Mean BP (mmHg)	106 (97–115)	85	106 (98–114)	82
ABI†	1.06 (0.96–1.14)	85	1.02 (0.94–1.12)	82
%VC	97.45 (85.95–108.05)	86	92.7 (83.78–105.70)	82
FEV1.0% (%)	72.78 (65.57–78.65)	86	71.72 (63.56–76.54)	82
Serum BUN (mg/dL)	17 (13–21)	88	17 (13–21)	87
Creatinine (mg/dL)	0.97 (0.86–1.24)	88	0.91 (0.79–1.14)	87
Triglyceride (mg/dL)	113 (84–148)	85	119 (82–183)	84
TC (mg/dL)	171 (155–190)	88	181 (157–207)	86
LDL-C (mg/dL)	99 (81–122)	84	108 (85–126)	84
HDL-C (mg/dL)	43 (37–52)	84	46 (36–58)	84
HbA1c (NGSP) (%)	5.8 (5.5–6.0)	84	5.8 (5.6–6.2)	85

^†Mean of left and right. FEV1.0%, percent predicted forced expiratory volume in 1.0 s; HbA1c, glycated hemoglobin. Other abbreviations as in Tables 1,2.

Multivariate Analysis

On multivariate analysis, parameters with a significant positive association with expansion rate were initial MAD (P<0.001) and oral steroid usage (P=0.029; Table 4); height was observed to have a significant negative association (P=0.041; Figure 2). HbA1c and total cholesterol showed a tendency toward negative association, as on univariate analysis.

Table 4. Multivariate Indicators of AAA Expansion Speed (n=156)

Variable	Rapid expansion (Growth rate >2.405 mm/year) (n=80)	Slow expansion (Growth rate <2.405 mm/year) (n=76)	OR (95% CI)	P-value
Initial MAD (mm)	42.5 (37.2–46.7)	35.8 (31.1–41.5)	1.14 (1.08–1.22)	<0.001
Oral steroids usage	10 (12.5)	5 (6.6)	4.12 (1.15–17.0)	0.029
Height (cm)	163.3 (157.3–167.0)	165.9 (161.3–170.0)	0.96 (0.91–0.99)	0.041
HbA1c (%)	5.8 (5.5–6.0)	5.9 (5.6–6.2)	0.51 (0.23–1.05)	0.067
TC (mg/dL)	171 (155–190)	183 (158–209)	0.99 (0.98–1.00)	0.088
FEV1.0% (%)	73.61 (65.45–78.62)	71.72 (65.06–76.18)	1.02 (0.99–1.06)	0.143

Data given as n (%) or median (IQR). Abbreviations as in Tables 2,3.

Figure 2.

Abdominal aortic aneurysm expansion rate vs. (A) initial maximum axial diameter (MAD; n=176); (B) height (n=172); and (C) oral steroid use (yes, n=15; no, n=153). Box plots: horizontal line, median expansion rate; top and bottom borders, 75th and 25th percentiles (the top whisker marks the 3rd quartile+1.5 IQR; the bottom whisker, the minimum value).

Oral Steroid Use

Details of oral steroid usage are listed in Table 5. Of 15 patients taking oral steroids, 14 had been diagnosed with underlying diseases with no obvious causal relationship with aneurysm formation in the aorta (i.e., mesangial proliferative glomerulonephritis, blue toe syndrome, myasthenia gravis, systemic lupus erythematosus (SLE), rheumatoid arthritis, chronic glomerulonephritis, hypereosinophilic syndrome, and interstitial pneumonia). The duration of oral steroid use accounted for a large proportion of the duration of AAA follow-up. Subject characteristics were compared between the oral steroid and non-oral steroid groups (Table 6). The proportion of men was significantly lower in the oral steroid group than in the non-oral steroid group (P=0.015).

Table 5. Oral Steroid Use: Subject Characteristics

	Primary disease	Sex	Age (years)†	Steroid	Dose (mg/day)‡	Duration of treatment (months)	AAA follow-up (months)	Expansion rate (mm/year)
1	MPG	M	62.7	Prednisolone	2.5	29	131	1.08
2	BTS	M	76.2	Prednisolone	15	15	31	1.09
3	MG	M	57.0	Prednisolone	10	132	140	1.28
4	SLE	M	70.7	Prednisolone	5	78	17	1.95
5	SLE	F	62.3	Prednisolone	5	264	123	2.27
6	SLE	F	66.0	Prednisolone	10	444	157	2.46
7	RA	F	75.8	Methyl-prednisolone	2	312	19	2.75
8	CGN	M	81.1	Prednisolone	10	84	31	3.72
9	Unknown	F	83.4	Prednisolone	5	>5	5	4.68
10	Lumbar radiculopathy	F	81.3	Prednisolone	2	24	25	4.69
11	HES	M	65.5	Prednisolone	5	28	26	5.35
12	IP	M	70.1	Prednisolone	17.5	33	46	5.56
13	Bullous pemphigoid	F	78.5	Prednisolone	5	24	18	5.77
14	BTS	M	72.6	Prednisolone	5	4	7	6.37
15	Renal cholesterol embolism	M	77.9	Prednisolone	7.5	9	9	11.76

^†Age at initial CT. ^‡Dose at follow-up CT. AAA, abdominal aortic aneurysm; BTS, blue toe syndrome; CGN, chronic glomerulonephritis; CT, computed tomography; HES, hypereosinophilic syndrome; IP, interstitial pneumonia; MG, myasthenia gravis; MPG, mesangial proliferative glomerulonephritis; RA, rheumatoid arthritis; SLE, systemic lupus erythematosus.

Table 6. Subject Characteristics vs. Oral Steroid Use

Variable	Oral steroids		No oral steroids		P-value
	n (%)	Total	n (%)	Total
Sex (M/F)	9 (60.0)/6 (40.0)	15	133 (86.9)/20 (13.1)	153	0.015†
Antihypertensives (Yes/No)	14 (93.3)/1 (6.7)	15	127 (83.0)/26 (17.0)	153	0.470†
ARB (Yes/No)	6 (40.0)/9 (60.0)	15	71 (46.4)/82 (53.6)	153	0.788†
Statins (Yes/No)	8 (53.3)/7 (46.7)	15	67 (43.8)/86 (56.2)	153	0.589†
CAD (Yes/No)	7 (46.7)/8 (53.3)	15	46 (30.3)/106 (69.7)	152	0.245†
CVD (Yes/No)	3 (20.0)/12 (80.0)	15	25 (16.7)/125 (83.3)	150	0.722†
	Median (IQR)	Total	Median (IQR)	Total
Expansion rate (mm/year)	3.72 (1.95–5.56)	15	2.39 (1.55–3.61)	153	0.069‡
Age (years)	72.6 (65.5–78.5)	15	72.1 (65.7–77.8)	153	0.868‡
Initial MAD (mm)	37.0 (31.7–42.3)	15	38.6 (32.6–45.3)	153	0.281‡
Height (cm)	162.3 (154.0–167.0)	15	165.0 (159.0–169.1)	150	0.226‡
HbA1c (%)	6.1 (5.8–6.6)	15	5.8 (5.5–6.1)	146	0.052‡
TC (mg/dL)	179 (158–204)	15	171 (156–198)	151	0.408‡
FEV1.0% (%)	72.32 (68.04–76.29)	15	71.81 (64.74–78.57)	147	0.931‡

^†Fisher exact test; ^‡Wilcoxon rank sum test. Abbreviations as in Tables 1–3.

Discussion

Initial MAD

Initial MAD was an independent risk factor for AAA expansion, in agreement with previous reports. On univariate analysis of the continuous variables and expansion rate, the correlation coefficient for initial MAD (0.442) was the highest; thus, initial MAD was certainly a strong risk factor for AAA expansion. Conversely, some of the large MAD patients had low expansion rate. The larger the MAD of an AAA, the wider the range of growth rate distribution (Figure 2). Thus, it was expected that other risk factors of AAA expansion may exist, particularly in patients with large MAD.

Steroid Use

A positive association was observed between oral steroid use and AAA expansion on multivariate analysis. The effect of steroids on AAA expansion has not been clear in past studies. Reportedly, the AAA expansion rate in patients who used oral steroids was higher than that in those who did not, only on univariate analysis.¹⁷ Thus, the present result may be a new important finding.

A question that arises when considering steroids as a risk factor for AAA expansion is whether it is the underlying diseases themselves that require steroid treatment, that cause AAA expansion. In patients taking oral steroids for vasculitis-related diseases such as Takayasu’s arteritis and Behcet’s disease, which may cause aneurysm of the arteries,^18–20 it is often difficult to exclude the effect of underlying diseases on expansion of the aneurysms. In the present study, however, none of the patients had Takayasu’s arteritis or Behcet’s disease. Three patients used oral steroids because of SLE, a vasculitis-related disease. Large AAA expansion rates or high AAA rupture rates have been noted in SLE patients,^21–23 but there were no histopathological findings of SLE in the aneurysmal wall.²¹ There is some possibility of a causal link between oral steroid usage and AAA expansion in SLE patients also; therefore, we did not exclude SLE patients in the present study.

In a previous small cohort study, infrarenal aortic thrombus volume was associated with AAA progression.¹⁶ In the present steroid use group, the primary disease in patients 14 and 15 was blue toe syndrome and renal cholesterol embolism, included under shaggy aorta diseases. Therefore, these 2 patients probably had rich thrombus in the infrarenal aorta and predominant AAA expansion. We evaluated AAA only in terms of MAD and not thrombus volume because some patients had undergone only plain CT and hence thrombus volume could not be evaluated. In those who had undergone enhanced CT, some patients had rich thrombus in the infrarenal aorta regardless of steroid use. Therefore, we did not exclude patients who had shaggy aorta disease.

Treatment with oral steroids has been known to promote atherosclerosis via impaired glucose tolerance and dyslipidemia.^22,24,25 In the present study, however, HbA1c and total cholesterol did not differ significantly between the oral steroid and non-oral steroid groups, and impaired glucose tolerance and dyslipidemia were not identified as expansion-promoting factors. Steroids are also known to cause tissue fragility, and mainly affect small- to medium-sized arteries, but there have also been reports indicating that they promote AAA rupture by causing tissue fragility in the aorta.^22,23 Thus, oral steroids may promote AAA expansion by causing aorta wall fragility via an unknown mechanism.

Height

In this study, height had a negative association with expansion rate, similar to previous reports on the relationship between physique and AAA.^3,13,26 Normal aorta diameter in a smaller patient may be lower than that in a bigger patient; thus, the wall fragility of AAA in a smaller patient may be advanced compared with that in a bigger patient for the same MAD. Thus, shorter height may be a risk factor for AAA expansion.

Smoking

In the present study, dividing the patients into 2 groups based on smoking history (ever smoker or never smoker) did not have any effect on AAA expansion. When the patients were divided into 3 groups (continuing smoker, ex-smoker, or never smoker), however, the expansion rate in the continuing smokers tended to be larger than that in the ex-smokers on univariate analysis (P=0.080). This tendency is in accordance with previous studies that have reported a relationship between AAA expansion and continued smoking.⁹ It would therefore be very useful to have more detailed investigations that include Brinkman index, smoking period, age of stopping smoking, and amount of toxic substances.

Other Factors

Small randomized studies have suggested that macrolides, tetracycline, and statins may inhibit AAA expansion.^27–30 In addition, ARB and hyperglycemia have inhibited the expansion in animal studies.^31–34 In the present study, ARB (P=0.025 on univariate analysis) and hyperglycemia (P=0.037 on univariate analysis, P=0.067 on multivariate analysis) tended to be negatively associated with AAA expansion, consistent with previous reports.^31–34 In addition, we did not observe any suppressive effects on AAA expansion of decreased cholesterol, oral statin use, or lower blood pressure. This suggests that the mechanism of AAA expansion differs from that of progression of atherothrombosis.

Study Limitations

This was a retrospective observational study; there was inter-subject variability in the length of follow-up and in the parameters measured at the latest CT, except for age and initial MAD. In addition, medication and laboratory data may have changed between the initial and latest CT. We investigated duration of oral steroid treatment, but other medications were not investigated. Thus, a further limitation was that it was not possible to determine whether a given factor promoted expansion or if the data represented the results of the expansion, aging, and progression of other diseases.

It was evident that the expansion rate was faster when the initial MAD was greater during the same observation period, and the relationship was not linear.^10,35 Accordingly, it was important to ensure that the interval between CT scans matched for each patient. It is difficult to decide the best interval, but we would have needed at least 6 months to reduce the error range of expansion rate according to the Guidelines for Diagnosis and Treatment of Aortic Aneurysm and Aortic Dissection (JCS 2011).¹¹ Given that the minimum interval in this study was 129 days, this may have increased the error range.

In the steroid use group, no patient had a disease that is a known cause of aneurysm of the aorta, such as Takayasu’s arteritis or Behcet’s disease. We cannot deny, however, that the presence of underlying diseases, particularly SLE and shaggy aorta-related diseases, affect AAA formation and expansion. Thus, we need to investigate the relationship between AAA and vasculitis-related diseases and the relationship between AAA expansion and thrombus volume on enhanced CT.

It is important to confirm these results in a prospective observational study. Although it would be expected that the morbidity risk in Japan is similar to that in Europe and the USA, this would need to be confirmed on further study.

Conclusions

A positive association was observed between oral steroid use and AAA expansion, independent of the other risk factors of atherosclerosis and MAD. Oral steroid use is therefore an important risk factor for AAA expansion; thus, we should examine AAA patients who take steroids more carefully.

Disclosures

The authors declare no conflict of interest.

Supplementary Files

Supplementary File 1

Figure S1. Correlation between abdominal aortic aneurysm maximum axial diameter (MAD) measured by the present authors and by the radiologists in a blind fashion.

Figure S2. Correlation between abdominal aortic aneurysm maximum axial diameter (MAD) measured by the same investigator after a 6-month interval.

Table S1. MAD: Interobserver variability (number of MAD mesurements=201)

Table S2. MAD: Intra-observer variability (number of MAD mesurements=352)

Please find supplementary file(s);

http://dx.doi.org/10.1253/circj.CJ-16-0902

References

• 1.   Johnston MD, Rutherford RB, Tilson MD, Shah DM, Hollier L, Stanley JC. Suggested standards for reporting on arterial aneurysms: Subcommittee on Reporting Standards for Arterial Aneurysms, Ad Hoc Committee on Reporting Standards, Society for Vascular Surgery and North American Chapter, International Society for Cardiovascular Surgery. J Vasc Surg 1991; 13: 452–458.
• 2.   Sakalihasan N, Limet R, Defawe OD. Abdominal aortic aneurysm. Lancet 2005; 365: 1577–1589.
• 3.   Jahangir E, Lipworth L, Edwards T, Kabagambe EK, Mumma MT, Mensah GA, et al. Smoking, sex, risk factors and abdominal aortic aneurysms: A prospective study of 18 782 persons aged above 65 years in the Southern Community Cohort Study. J Epidemiol Community Health 2015; 69: 481–488.
• 4.   Golledge J, Muller J, Daugherty A, Norman P. Abdominal aortic aneurysm: Pathogenesis and implications for management. Arterioscler Thromb Vasc Biol 2006; 26: 2605–2613.
• 5.   Golledge J, Norman PE. Atherosclerosis and abdominal aortic aneurysm: Cause, response, or common risk factors? Arterioscler Thromb Vasc Biol 2010; 30: 1075–1077.
• 6.   Lederle FA, Johnson GR, Wilson SE, Chute EP, Littooy FN, Bandyk D, et al. Prevalence and associations of abdominal aortic aneurysm detected through screening. Ann Intern Med 1997; 126: 441–449.
• 7.   The United Kingdom Small Aneurysm Trial Participants. Long-term outcomes of immediate repair compared with surveillance of small abdominal aortic aneurysms. N Engl J Med 2002; 346: 1453–1458.
• 8.   Lederle FA, Wilson SE, Johnson GR, Reinke DB, Littooy FN, Acher CW, et al. Immediate repair compared with surveillance of small abdominal aortic aneurysms. N Engl J Med 2002; 346: 1445–1452.
• 9.   Baxter BT, Terrin MC, Dalman RL. Medical management of small abdominal aortic aneurysms. Circulation 2008; 117: 1883–1889.
• 10.   Vardulaki KA, Prevost TC, Walker NM, Day NE, Wilmink ABM, Quick CRG, et al. Growth rates and risk of rupture of abdominal aortic aneurysms. Br J Surg 1998; 85: 1674–1680.
• 11.   JCS Joint Working Group. Guidelines for diagnosis and treatment of aortic aneurysm and aortic dissection (JCS 2011): Digest version. Circ J 2013; 77: 789–828.
• 12.   Hendy K, Gunnarson R, Golledge J. Growth rates of small abdominal aortic aneurysms assessed by computerized tomography: A systematic literature review. Atherosclerosis 2014; 235: 182–188.
• 13.   Schlosser FJV, Tangelder MJD, Verhagen HJM, Van der Heiiden GJMG, Muhs BE, Van der Graaf Y, et al. Growth predictors and prognosis of small abdominal aortic aneurysms. J Vasc Surg 2008; 47: 1127–1233.
• 14.   Lindholt JS. Aneurysmal wall calcification predicts natural history of small abdominal aortic aneurysms. Atherosclerosis 2008; 197: 673–678.
• 15.   Nakayama A, Morita H, Hayashi N, Nomura Y, Hoshina K, Shigematsu K, et al. Inverse correlation between calcium accumulation and the expansion rate of abdominal aortic aneurysms. Circ J 2016; 80: 332–339.
• 16.   Parr A, McCann M, Bradshaw B, Shahzad A, Buttner P, Golledge J. Thrombus volume is associated with cardiovascular events and aneurysm growth in patients who have abdominal aortic aneurysms. J Vasc Surg 2011; 53: 28–35.
• 17.   Lindholt JS, Heickendorff L, Antonsen S, Fasting H, Henneberg EW. Natural history of abdominal aortic aneurysm with and without coexisting chronic obstructive pulmonary disease. J Vasc Surg 1998; 28: 226–233.
• 18.   JCS Joint Working Group. Guideline for management of vasculitis syndrome (JCS 2008): Digest version. Circ J 2011; 75: 474–503.
• 19.   Kerr GS, Hallahan CW, Giordano J, Leavitt RY, Fauci AS, Rorrem M, et al. Takayasu arteritis. Ann Intern Med 1994; 120: 919–929.
• 20.   Al-Otaibi LM, Porter SR, Poate TWJ. Behcet’s disease: A review. J Dent Res 2005; 84: 209–222.
• 21.   Ohara N, Miyata T, Kurata A, Oshiro H, Sato O, Shigematsu H. Ten years’ experience of aortic aneurysm associated with systemic lupus erythematosus. Eur J Vasc Endovasc Surg 2000; 19: 288–293.
• 22.   Bulkley BH, Roberts WC. The heart in systemic lupus erythematosus and the changes induced in it by corticosteroid therapy: A study of 36 necropsy patients. Am J Med 1975; 58: 243–264.
• 23.   Ohara N, Miyata T, Sato O, Oshiro H, Shigematsu H. Aortic aneurysm in patients with autoimmune disease treated with corticosteroids. Int Angiol 2000; 19: 270–275.
• 24.   Movahedi M, Beauchamp ME, Abrahamowicz M, Ray DW, Michaud K, Pedro S, et al. Risk of incident diabetes mellitus associated with the dosage and duration of oral glucocorticoid therapy in patients with rheumatoid arthritis. Arthritis Rheumatol 2016; 68: 1089–1098.
• 25.   Strohmayer EA, Krakoff LR. Glucocorticoids and cardiovascular risk factors. Endocrinol Metab Clin North Am 2011; 40: 409–417.
• 26.   Lederle FA, Johnson GR, Wilson SE, Chute EP, Hye RJ, Makaroun MS, et al. The aneurysm detection and management study screening program: Validation cohort and final results. Aneurysm Detection and Management Veterans Affairs Cooperative Study Investigators. Arch Intern Med 2000; 160: 1425–1430.
• 27.   Vammen S, Lindholt JS, Ostergaard L, Fasting H, Henneberg W. Randomized double-blind controlled trial of roxithromycin for prevention of abdominal aortic aneurysm expansion. Br J Surg 2001; 88: 1066–1072.
• 28.   Mosorin M, Juvonen J, Biancari F, Satta J, Surcel HM, Leinonen M, et al. Use of doxycycline to decrease the growth rate of abdominal aortic aneurysms: A randomized, double-blind, placebo-controlled pilot study. J Vasc Surg 2001; 34: 606–610.
• 29.   Sukhija R, Aronow WS, Sandhu R, Kakar P, Babu S. Mortality and size of abdominal aortic aneurysm at long-term follow-up of patients not treated surgically and treated with and without statins. Am J Cardiol 2006; 97: 279–280.
• 30.   Schouten O, Van Laanen JHH, Boersma E, Vidakovic R, Feringa HHH, Dunkelgrun M, et al. Statins are associated with a reduced infrarenal abdominal aortic aneurysm growth. Eur J Vasc Endovasc Surg 2006; 32: 21–26.
• 31.   Rizzoni D, Rodella L, Porteri E, Rezzani R, Sleiman I, Paiardi S, et al. Effects of losartan and enalapril at different doses on cardiac and renal interstitial matrix in spontaneously hypertensive rats. Clin Exp Hypertens 2003; 25: 427–441.
• 32.   Daugherty A, Manning MW, Cassis LA. Angiotensin II promotes atherosclerotic lesions and aneurysms in apolipoprotein E-deficient mice. J Clin Invest 2000; 105: 1605–1612.
• 33.   Dua MM, Miyama N, Azuma J, Schultz GM, Sho M, Morser J, et al. Hyperglycemia modulates plasminogen activator inhibitor-1 expression and aortic diameter in experimental aortic aneurysm disease. Surgery 2010; 148: 429–435.
• 34.   Golledge J, Karan M, Moran CS, Muller J, Clancy P, Dear AE, et al. Reduced expansion rate of abdominal aortic aneurysms in patients with diabetes may be related to aberrant monocyte-matrix interactions. Eur Heart J 2008; 29: 665–672.
• 35.   Brady AR, Thompson SG, Fowkes FG, Greenhalgh RM, Powell JT; UK small Aneurysm Trial Participants. Abdominal aortic aneurysm expansion risk factors and time intervals for surveillance. Circulation 2004; 110: 16–21.