- Email: [email protected]
Aortic growth rates in chronic aortic dissection
Clinical Radiology (2007) 62, 866e875
Aortic growth rates in chronic aortic dissection A.M. Kellya,*, L.E. Quinta, B. Nanb, J. Zhengb, P. Cronina, G.M. Deebc, D.M. Williamsd a
Department of Radiology, Division of Thoracic Radiology, cDivision of Cardiac Surgery, Division of Vascular Interventional Imaging, University of Michigan Medical Center, and b School of Public Health, University of Michigan, Ann Arbor, Michigan, USA d
Received 27 September 2006; received in revised form 1 March 2007; accepted 26 April 2007
AIM: To determine and compare rates of descending aortic enlargement and complications in chronic aortic dissection with and without a proximal aortic graft. METHODS AND MATERIALS: Fifty-two patients with dissection involving the descending aorta and who had undergone at least two computed tomography (CT) examinations at our institution between November, 1993 and February, 2004 were identified, including 24 non-operated patients (four type A, 20 type B) and 28 operated patients (type A). CT examinations per patient ranged from two to 10, and follow-up ranged from 1e123 months (mean 49 months, median 38.5 months). On each CT image, the aortic short axis (SA), false lumen (FL), and true lumen (TL) diameters were measured at the longitudinal midpoint of the dissection and at the point of maximum aortic diameter. Complications were tabulated, including aortic rupture and aortic enlargement requiring surgery. RESULTS: For non-operated patients, the midpoint and maximum point SA, TL, and FL diameters increased significantly over time. For operated patients, the midpoint and maximum point SA and FL diameters increased significantly over time. In both groups, aortic enlargement was predominantly due to FL expansion. Diameter increases in nonoperated patients were significantly larger than those in operated patients. The rate of change in aortic diameter was constant, regardless of aortic size. Four non-operated and six operated patients developed aortic complications. CONCLUSIONS: In patients with a dissection involving the descending thoracic aorta, the FL increased in diameter over time, at a constant rate, and to a greater degree in non-operated patients (mostly type B) compared with operated patients (all type A). ª 2007 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved.
Introduction There are few data in the medical literature regarding the natural history of chronic type A or type B dissections involving the descending aorta.1,2 This applies to both non-operated dissections, as well as residual dissections in the descending aorta of patients with a proximal aortic graft. It is known that the false lumen (FL) tends to dilate over time, occasionally leading to rupture and/or the need for surgical repair. However, the
* Guarantor and correspondent: A.M. Kelly, Department of Radiology, Division of Thoracic Radiology, University of Michigan Hospitals, B1 132K Taubman Center/0302, 1500 East Medical Center Drive, Ann Arbor, Michigan 48109-0030, USA. Tel.: þ1 734 936 0024; fax: þ1 734 936 0013. E-mail address: [email protected] (A.M. Kelly).
rate of FL enlargement has not been well documented. Furthermore it is unknown whether non-operated type A or B dissections behave differently compared with dissections involving the descending aorta in a patient with previous, proximal aortic graft placement. The purpose of the present study was to determine and compare the rates of aortic enlargement and frequency of complications in these two types of patient.
Materials and methods Study population Institutional internal review board approval was obtained for this retrospective review and a waiver of informed consent was approved. All patients
0009-9260/$ - see front matter ª 2007 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.crad.2007.04.006
Aortic growth rates in chronic aortic dissection
seen at our institution between November, 1993 and February, 2004 with descending thoracic aortic dissection and who had undergone an inhouse computed tomography (CT) examination were identified using electronic databases from the Department of Radiology and the Division of Cardiothoracic Surgery. Of the 118 patients identified, 53 patients with only one CT study at our institution or with less than 1 month between the initial and the first follow-up CT exam were excluded. Also excluded were 13 patients with a completely thrombosed FL on the initial CT, based on review of the study. The reason to exclude these patients was that the focus of the study was on the natural history of the perfused FL. The remaining 52 patients were divided into two groups. The first group (non-operated group) comprised 24 patients with a non-operated type A or type B dissection (four type A and 20 type B). The second group (operated group) was composed of 28 patients with a type A dissection, proximal aortic repair, and a dissection flap in the descending aorta. The average age of patients was 59 years (range 34e81 years) in the operated group (22 males, six females) and 64 years (range 43e83 years) in the non-operated group (14 males, 10 females). The patients entered the study at different time points in the course of their disease. The first CT examination was considered the baseline or time zero. The number of CT examinations per patient ranged from two to 10 (with a mean of four examinations). The length of follow-up ranged from 1 month to 10 years 3 months, with the average length of follow-up being 49 months in the operated group and 34 months in the non-operated group. The median length of follow-up was 45.5 months in the operated group and 24 months in the non-operated group.
CT imaging Because of the long time frame of the study, extending from 1993 to 2004, the CT technique used has evolved. All studies were performed using helical technique, with a single row, four row, eight row or 16 row CT machine. Imaging was performed with the patient at end inspiration. Collimation ranged from 3 mm on the early images to 1.25 mm on the more recent images, with a pitch of 1e2. Patients were imaged from above the aortic arch down to at least the level of the aortic bifurcation using a bolus of intravenous contrast material administered via pump injector. CT was generally performed on an annual basis. However, if a patient
867
experienced new signs or symptoms of complications between scheduled examinations, interval examinations were performed.
Image interpretation Images were interpreted from hard-copy film photographed on a 20 to 1 format; soft-copy images were not available for the older studies. The hard-copy format was consistent for all studies with the same degree of image minification. The display field of view was consistent for all aorta studies at our institution. Patients with outside studies whose images did not have a reference measurement scale were excluded. All patients were reviewed using hard-copy images to improve consistency between measurements. The same author measured all images, using callipers. The craniocaudal extent of dissection was recorded in centimetres. The short axis (SA) diameter of the aorta was measured at the midpoint along the craniocaudal length; the location of the midpoint was recorded by using adjacent anatomic landmarks such as the carina, pulmonary veins, diaphragm, adrenal glands, and abdominal aortic branch vessels. Aortic SA diameter measurements were chosen rather than long axis diameters, because the former are less influenced by the orientation of the aorta with regard to the plane of section. True lumen (TL) and FL diameters perpendicular to the intimal flap were also measured at the midpoint along the craniocaudal length of the dissection; the sum of the TL and FL diameters often exceeded the SA measurement, because the flap was not necessarily perpendicular to the SA of the aorta (Fig. 1). The location of the maximum SA aortic diameter was determined by visual inspection of the images, and was recorded by using adjacent anatomic landmarks. The maximum SA aortic diameter, as well as TL and FL diameters perpendicular to the intimal flap at this level, were also measured and recorded. Periaortic haematoma was considered evidence for rupture.
Chart review For each patient, the medical record was searched for the following information: aortic complications, including aortic rupture and aortic enlargement requiring surgery; sex; age; history of smoking; history of chronic obstructive pulmonary disease (COPD); and hypertension (defined as a mean arterial pressure of greater than 110, with mean arterial pressure being calculated as diastolic pressure plus one third of the difference
868
A.M. Kelly et al.
Growth rate calculations and statistical analysis Individual aortic growth rates were calculated by dividing the change in aortic diameter (measured in millimetres) by time (measured in years) for each patient. Time points for evaluating individual growth rates were selected to be at least 1 year apart; this was done to minimize spurious shortterm changes. The group level growth rates for both non-operated and operated patient groups were estimated by using linear mixed effects models. Distributions of the anatomic levels of midpoints and maximum points in the non-operated and operated groups were analysed using the KolmogoroveSmirnov test. (This test is analogous to a t-test but makes no assumptions about the normality of the data.) The predictors of outcome were age, sex, history of smoking, history of COPD, and presence of hypertension. The outcome variables were craniocaudal extent of dissection and aortic diameters at midpoint (SA, TL, FL) and at maximum point (SA, TL, FL). Statistical analysis was performed on clinical and CT measurement variables, using a linear mixed effects model. Statistical significance was set at p < 0.05.
Results Midpoint diameters
Figure 1 (a) Aortic SA measurement (double arrow). (b) True and FL measurements (double arrows) were made perpendicular to the dissection flap.
between systolic and diastolic). The medical records were also searched for follow-up data on patients, including hypertensive status, control of hypertension, continued cigarette smoking, and cardiovascular risk factors.
The midpoint SA aortic diameters in both the non-operated and operated groups increased significantly over time compared with the baseline (p < 0.001), showing average rates of 1.5 and 0.8 mm/year, respectively. The mean growth rate in the non-operated patients was significantly higher than the rate in the operated patients (p ¼ 0.013; Table 1). The range in rates in individual patients in both groups combined was from 3 to 9 mm/year. Two patients had a decrease in midpoint diameter and in both of these cases; this was related to thrombosis of the FL. Increase over time in the aortic midpoint SA diameter appeared to be largely accounted for by an increase in the FL diameter (Table 1). The diameters of the TL in the non-operated group increased significantly over time compared with the baseline (p < 0.001), showing an average growth rate of 1 mm/year. However, the TL diameters in the operated group did not increase significantly over time (p ¼ 0.077) (Table 1). The range in rates in individual patients in both groups combined was from 1 to 11 mm/year. Negative rates could have been due to movement of the intimal
Aortic growth rates in chronic aortic dissection
Table 1 patients
869
The rate of change in aortic midpoint short axis, true lumen, and false lumen diameters in non-operated and operated True lumen diameter at midpoint
Short axis diameter at midpoint mm/year Non-operated patient group Operated patient group Difference between groups
1.5 0.8 0.7
SE 0.2 0.16 0.26
p-Value a
<0.001 <0.001a 0.013b
mm/year 1 0.3 0.7
SE 0.19 0.15 0.25
False lumen diameter at midpoint p-Value a
<0.001 0.077a 0.003b
mm/year
SE
p-Value
1.4 0.7 0.7
0.23 0.18 0.29
<0.001a <0.001a 0.030b
SE, standard error. a p-Value indicates the level of significance for change in diameter compared with the baseline. b p-Value indicates the level of significance for difference in growth rates between operated and non-operated groups.
flap between systole and diastole or measurement error. The diameters of the FL in both the nonoperated and operated groups increased significantly over time compared with the baseline (p < 0.001), showing average rates of 1.4 and 0.7 mm/year, respectively. The growth rate in the non-operated patients was significantly higher than the rate in the operated patients (p ¼ 0.03; Table 1). All patients maintained at least partial patency of the FL, and the growth rates in individual patients in both groups combined ranged from 4 to 6 mm/year. Negative rates could have been due to movement of the intimal flap between systole and diastole or actual decrease in size of FL or measurement error.
Maximum point diameters The maximum point SA aortic diameters in both the non-operated and operated groups increased significantly over time compared with the baseline (p ¼ 0.001 and 0.004, respectively), showing average rates of 1.2 and 0.5 mm/year, respectively. The mean growth rate in the non-operated patients was significantly higher than the rate in the operated patients (p ¼ 0.007; Table 2). The range in rates in individual patients in both groups combined was from 0.2 to 9.5 mm/year. Five patients had a decrease in maximum point diameter and in four of these patients; this was related to
thrombosis of the FL. In the fifth patient, diameter decrease may have been due to the dynamic nature of the aorta between systole and diastole or possibly measurement error. Increase over time in the aortic maximum point SA diameter in operated patients appeared to be largely accounted for by an increase in the TL diameter, and in non-operated patients, was largely due to FL diameter increase (Table 2). The diameters of the TL in both the nonoperated and operated groups increased significantly over time compared with the baseline (p ¼ 0.004 and 0.001, respectively), showing average growth rates of 0.8 and 0.7 mm/year (Table 2). The rate of change in TL diameter was not significantly greater for the non-operated patients compared with the operated patients (p ¼ 0.83). The range in rates in individual patients in both groups combined was from 2.5 to 12.4 mm/ year. Negative rates could have been due to movement of the intimal flap between systole and diastole or measurement error. The diameter of the FL in the non-operated group increased significantly over time compared with the baseline (p ¼ 0.001), showing an average rate of 0.9 mm/year. However, the FL diameter in the operated group did not increase significantly over time (p ¼ 0.296). All patients maintained at least partial patency of the FL, and the growth rates in individual patients in both groups combined ranged from 15 to 10 mm/year (Table 2).
Table 2 The rate of change in aortic maximum point short axis, true lumen, and false lumen diameters in non-operated and operated patients Short axis diameter at maximum point
Non-operated patient group Operated patient group Difference between groups
True lumen diameter at maximum point
False lumen diameter at maximum point
mm/year
SE
p-Value
mm/year
SE
p-Value
mm/year
SE
p-Value
1.2 0.5 0.7
0.2 0.17 0.27
<0.001a 0.004a 0.007b
0.8 0.7 0.1
0.25 0.2 0.32
0.004a 0.001a 0.832b
0.9 0.2 0.7
0.28 0.22 0.35
0.001a 0.296a 0.047b
SE, standard error. a p-Value indicates the level of significance for change in diameter compared with the baseline. b p-Value indicates the level of significance for difference in growth rates between operated and non-operated groups.
870
Negative rates could have been due to movement of the intimal flap between systole and diastole, actual decrease in size of the FL or measurement error.
Growth rate constancy The rate of change in aortic midpoint (SA, TL, and FL) and maximum point (SA, TL and FL) diameters was constant in both non-operated and operated patients, regardless of the aortic diameter. In other words, increased aortic diameter did not correlate with an increased growth rate (p > 0.05).
Distribution of initial midpoint and maximum point locations and change in craniocaudal extent over time In almost all patients, the anatomic level of the midpoint was in the descending aorta at, or close to, the diaphragmatic hiatus. The distributions of craniocaudal midpoint locations did not differ significantly between the non-operated and operated groups (p ¼ 0.593, KolmogoroveSmirnov test). In two thirds of patients, the anatomic level of the maximum point was at the distal aortic arch or proximal descending thoracic aorta (at or close to the level of carina). The distributions of maximum point locations did not differ significantly between the non-operated and operated groups (p ¼ 0.892, KolmogoroveSmirnov test). The craniocaudal extent (length) of the dissection flap did not change significantly over time in either the non-operated or the operated groups.
Effect of hypertension After the initial clinical and imaging evaluation, most patients only return periodically for surveillance CT examinations because our institution is a tertiary care centre; these patients are then seen in the clinics only if the CT examination suggests a new or progressing abnormality. Generally, routine clinical follow-up is performed by the patients’ primary-care physicians outside of our hospital system. Because of this, the serial blood pressure measurements of many of the patients in this study were not available, and therefore, there were too few measurements to correlate hypertension with rate of aortic growth.
Effect of smoking, COPD, sex, and age There was no significant difference between smokers and non-smokers, between patients with
A.M. Kelly et al.
and without COPD, or between males and females with respect to the rate of aortic midpoint enlargement (SA, TL, and FL). With respect to change in the TL and FL diameters at the midpoint, there was no statistically significant difference in rate of change with different ages. Although the aortic midpoint SA diameter increased at a statistically significantly greater rate in younger patients compared with older patients (p ¼ 0.025), the difference was small (0.03 mm/year for each year younger) and not felt to be clinically significant. There was no significant difference between smokers and non-smokers or between males and females with respect to rate of aortic maximum point enlargement (SA, TL and FL; p > 0.05). Age did not have a significant effect on rate of maximum point enlargement (p > 0.05). With respect to change in the FL diameters at the aortic maximum point, there was a borderline statistically significant difference in the rate of change with different ages, but the rate of change was minimal (0.02 mm/year for each year younger) and felt not to be clinically significant (p ¼ 0.04). The aortic maximum point SA and FL diameters increased at a statistically significantly greater rate in patients without COPD compared with COPD patients (p ¼ 0.045 and 0.020, respectively), which is in contrast to the findings seen with respect to the midpoint. None of the above findings regarding statistical significance changed when the four patients with type A dissections in the non-operated group were excluded from analysis.
Complications In the non-operated group, four of 24 (17%) patients developed aortic complications during the follow-up period of this study, and all required subsequent aortic surgery. Two patients developed aortic rupture 3.5 and 8 years (Fig. 2), respectively, after the initial dissection event, showing maximum SA diameters of 6.3 and 5.8 cm, respectively, at the time of rupture. Two other patients developed marked aortic dilatation requiring surgery. In the first patient, the aorta SA maximum diameter increased in size from 4.2 to 5.3 cm over a 5.5 year interval. In the second patient, the aorta SA maximum diameter increased in size from 4.9 to 6.0 cm over a 4.75 year interval (Fig. 3). The level at which aortic enlargement or rupture occurred in these four patients varied from the distal arch to the mid descending thoracic aorta, close to the level of the pulmonary arteries. The SA midpoint and maximum point aortic growth rates in these four patients ranged from 1e3.4 and 0.1e4.2 mm/year, respectively.
Aortic growth rates in chronic aortic dissection
871
In the operated group, six of the 28 (21%) patients developed aortic complications during the follow-up period of this study, and all required further aortic surgery. One patient showed aortic rupture 3 years after surgery, showing a maximum SA diameter of 8.6 cm at the time of rupture. Five patients either had or developed pronounced aortic dilatation (maximum SA diameters 4e7.1 cm) an average of 2 years and 10 months after surgery (range 1e6 years and 5 months). The area of aortic enlargement or rupture was in the distal arch or proximal descending thoracic aorta, close to the tracheal carina, in four of the six patients, at the level of the adrenals in one patient, and there was diffuse enlargement throughout the descending thoracic aorta in one patient (Fig. 4). The SA midpoint and maximum point aortic growth rates in this group of six patients ranged from 0e6 and 0e10 mm/year, respectively.
Discussion
Figure 2 Increase in FL measurement and interval rupture of the FL between 1993 (a) and 1994 (b) in a 66-year-old non-operated man, necessitating urgent surgery.
Thoracic aortic aneurysms have been reported to grow at a rate of approximately 0.7e1.9 mm/ year.3,4 However, the rates of aortic enlargement in aortic dissections have been scantily documented in the medical literature.5,6 In the present study, the finding of 1.5 mm/year SA aortic diameter mean growth rate for non-operated patients is somewhat less than the mean SA growth rate of 3.3 mm/year in 62 non-operated type B patients reported by Sueyoshi et al.5 In the present study a patient by patient analysis was used, whereas in the study of Sueyoshi and colleagues5 the data were analysed based on aortic segments; therefore, it is unclear if the results are directly comparable. The latter study was potentially limited by relatively thick (5 mm) sections and a large proportion of non-helically acquired images (approximately 37%), thus leading to somewhat limited spatial resolution, by current standards. The report of Onitsuka et al.6 found a mean growth rate of 3.8 mm/year in non-operated type B patients; however, that study used long axis aortic measurements and such measurements may not reflect true diameters, particularly in tortuous aortas. In addition, although the study of Onitsuka et al. included a substantial number of patients (76), it is not specified how many patients had follow-up CT examinations, and therefore, it is not clear how many patients contributed to the mean growth rate reported. For operated patients, the data from the present study showed a mean SA midpoint growth rate of 0.8 mm/year and a SA maximum point growth
872
A.M. Kelly et al.
rate of 0.5 mm/year. To the authors’ knowledge, there are no published reports documenting aortic growth rates in operated aortic dissection patients. In vitro aortic dissection models have demonstrated that increase in total cross-sectional area with increasing hydrostatic pressure occurs primarily due to enlargement of the FL.7 The present study confirmed that FL expansion was indeed the major contributor to aortic enlargement at the midpoint in patients with either operated or nonoperated aortic dissection, although the degree of enlargement was significantly greater in nonoperated patients. The data for maximum point were somewhat contradictory: although FL expansion appeared to account for the majority of aortic enlargements in non-operated patients, TL expansion appeared to be the dominant force in the operated group. It is possible that the data for maximum point measurements were less robust than the data for midpoint measurements; the maximum point often occurred in the region where the distal aortic arch curved into the proximal descending aorta, and the curvature may have led to inaccurate measurements. Using the midpoint and maximum point measurements, the FL expanded at a higher rate in non-operated patients compared with operated patients. The operated group was composed of patients with type A dissections, whereas the nonoperated group contained mostly patients with type B dissections. These two diseases are likely due to different types of pathophysiology,8 and it is possible that the non-operated patients had intrinsically more disease in the descending aorta compared with the operated group, leading to the increased rate of growth. Alternatively, surgical repair of the proximal aorta may have lessened the degree to which the distal aorta dilated over time, due to alteration in haemodynamics. In addition, patients who underwent surgery may have had better blood pressure control compared with non-surgical patients, thus lessening the degree of subsequent aortic dilatation. Several investigators have reported that thrombosis of the FL has a better prognosis compared with a patent FL, with regard to subsequent aortic enlargement and/or dissection related death.5,9,10e13 This was not verified in the present study, because all of the patients had flow in the
Figure 3 Increase in FL measurement between October 1998 (a) and July 2003 (b) in a 59-year-old man. This nonoperated patient required surgery because of marked FL enlargement, with a 1.9 cm interval increase in diameter.
Aortic growth rates in chronic aortic dissection
873
Figure 4 Increase in FL measurements between June 2002 (a and b) and September 2004 (c and d), in a 57-year-old man. This previously operated patient required further surgery because of marked FL enlargement, with a 2.7 cm interval increase in diameter.
FL on the initial CT examination, and none had complete thrombosis of the FL on follow-up. Previous studies have suggested that an increase in aortic diameter leads to an increase in the rate of aortic diameter growth in aneurysms and dissections.11,14,15 However, Sueyoshi and colleagues5 did not find such an association: patients with initial aortic diameters greater than 40 mm actually showed a lower mean growth rate compared with those with an initial diameter less than 40 mm, although the difference was not statistically significant. These authors hypothesized that the perpendicular stress on the aortic wall is lessened in the presence of two lumens, and therefore dissections behave differently compared with aneurysms in this respect. In the present study
there were no data regarding aortic diameter at the time of the initial dissection event, because patients entered the study at various time points in their disease. However, by using multiple time points in each patient, rather than just a beginning and an ending point, growth rates could be compared over time; the data showed that aortic growth rates at midpoint and maximum point were constant, supporting the premise that aortic diameter does not affect growth rate in this patient population. Published studies have reported that a history of smoking,4 COPD,5,16,17 female sex,5 and increasing age16,17 correlate with an increased chance of rupture and increased aortic growth rate in aneurysms and dissections. However, other studies have
874
disputed some of these findings and, in fact, found contradictory results.5,16 In the present study no significant correlation between aortic midpoint growth rates and smoking, sex, age or COPD was found. There was no correlation between aortic maximum point growth rates and smoking, sex, or age, although there was an inverse relationship between COPD and growth rate of borderline statistical significance. In non-operated patients, chronic dissectionrelated complications, such as aortic rupture, increase in aortic diameter >50 mm, mean aortic enlargement rate >5 mm/year, surgery and/or death, have been reported at rates of 12e33%.6,18,19 The rate of 17% from the present study falls at lower end of this range. Six of 28 (21%) patients in the operated group experienced dissection-related complications, including aortic dilatation or rupture requiring subsequent surgery downstream from the original repair. This is comparable with published rates of 11e30% from other investigations.1,18e20 The limitations of the present study included the use of axial images rather than images perpendicular to the long axis of the aorta; such images were used because only hard-copy, conventional axial images were available for the older examinations. This technique could potentially have led to inaccuracies in aortic diameter measurement; however, this problem may have been alleviated by the use of SA measurements, which are less dependent on the orientation of the aorta with regard to the image plane compared with long axis measurements. However, aortic dissections tend to be oval in cross-section due to greater expansion of the FL relative to the TL; thus, SA measurements might underestimate the degree of FL expansion. In the present study hard copy rather than soft copy images were used. Many of the patients had a long follow-up and in the earlier years, images were on hard copy. All images were reviewed as hard copy to maintain consistency to reduce errors. Review of soft-copy images would have been preferable. Two-dimensional (2D) length measurements rather than three-dimensional (3D) volumes were measured in the present study, because currently available 3D tools could not be employed on the older images; volume measurements would likely be more accurate than diameters in assessing size. A limitation in the measurement of TL and FL diameters was the static nature of the images: although, in reality, the dissection flap tends to move back and forth during systole and diastole, the images capture only a moment in time, and
A.M. Kelly et al.
measurements were made at a single, flap position. Finally, in the few patients with aortic rupture, the presence of periaortic haematoma may have led to inaccuracies in measuring aortic diameters. The present study is also limited by the lack of follow-up data with respect to clinical parameters, which can affect dissection outcomes such as hypertensive status, medication control of hypertension, continued cigarette smoking, and the presence of other cardiovascular disease. It was difficult to obtain this clinical information from the patients’ records because many of the patients received primary care elsewhere. The data that were obtained on a few patients were insufficient to draw statistical inferences. A large prospective study of a similar group of patients would be of benefit to assess the contribution of the clinical factors above to aortic dissection outcomes. In addition, patients entered the study at different time points, both chronologically and with respect to their aortic dissection disease. Some patients will have entered the study much later in the course of their aortic dissection relative to others. Some patients will have entered the study in the immediate post-surgical period and others much later after surgery. This may contribute to differences in disease course and therefore mask or exaggerate differences the two groups. In conclusion, in patients with a dissection involving the descending thoracic aorta, the FL increased in diameter over time, and the rate of increase in diameter was constant, regardless of aortic diameter. FL dilatation occurred to a significantly greater degree in non-operated patients (mostly type B dissections) compared with patients with a type A dissection and a proximal aortic graft. This greater increase in aortic diameter in type B patients compared with type A patients may reflect different pathophysiologies, rather than the effect of a proximal aortic graft. Complication rates in the present study were 17% for non-operated patients and 21% for operated patients.
Acknowledgements This study was funded in part by a GE-AUR Radiology Research Academic Fellowship.
References 1. Heinemann M, Las J, Crack M, et al. Thoracic aortic aneurysms after acute type A aortic dissection: necessity for follow-up. Ann Thorac Surg 1990;49:580e4.
Aortic growth rates in chronic aortic dissection
2. Wheat Jr MW, Palmer RF. Dissecting aneurysms of the aorta. Curr Probl Surg 1971; 1e43. 3. Davies RR, Goldstein LJ, Coady MA, et al. Yearly rupture or dissection rates for thoracic aortic aneurysms: simple prediction based on size. Ann Thorac Surg 2002;73:17e27 [discussion 27e8]. 4. Coady MA, Rizzo JA, Hammond GL, et al. What is the appropriate size criterion for resection of thoracic aortic aneurysms? J Thorac Cardiovasc Surg 1997;113:476e91. 5. Sueyoshi E, Sakamoto I, Hayashi K, et al. Growth rate of aortic diameter in patients with type B aortic dissection during the chronic phase. Circulation 2004;110(Suppl. 1): II256e61. 6. Onitsuka S, Akashi H, Tayama K, et al. Long-term outcome and prognostic predictors of medically treated acute type B aortic dissections. Ann Thorac Surg 2004;78:1268e73. 7. Williams DM, LePage MA, Lee DY. The dissected aorta: part I. Early anatomic changes in an in vitro model. Radiology 1997;203:23e31. 8. Coady MA, Rizzo JA, Goldstein LJ, et al. Diseases of the aorta. Natural history, pathogenesis and etiology of thoracic aortic aneurysms and dissections. Cardiol Clin 1999; 17:615e35. 9. Bernard Y, Zimmermann H, Chocron S, et al. False lumen patency as a predictor of late outcome in aortic dissection. Am J Cardiol 2001;87:1378e82. 10. Akutsu K, Nejima J, Kiuchi K, et al. Effects of the patent false lumen on the long-term outcome of type B acute aortic dissection. Eur J Cardiothorac Surg 2004;26:359e66. 11. Marui A, Mochizuki T, Mitsui N, et al. Toward the best treatment for uncomplicated patients with type B acute aortic
875
12.
13.
14.
15.
16.
17.
18.
19.
20.
dissection: a consideration for sound surgical indication. Circulation 1999;100(Suppl.19):II275e80. Kozai Y, Watanabe S, Yonezawa M, et al. Long-term prognosis of acute aortic dissection with medical treatment: a survey of 263 unoperated patients. Jpn Circ J 2001;65:359e63. Fattori R, Bacchi-Reggiani L, Bertaccini P, et al. Evolution of aortic dissection after surgical repair. Am J Cardiol 2000; 86:868e72. Dapunt OE, Galla JD, Sadeghi AM, et al. The natural history of thoracic aortic aneurysms. J Thorac Cardiovasc Surg 1994;107:1323e32. discussion 1332e3. Kato M, Bai H, Sato K, et al. Determining surgical indications for acute type B dissection based on enlargement of aortic diameter during the chronic phase. Circulation 1995; 92(Suppl.9):II107e12. Juvonen T, Ergin MA, Galla JD, et al. Prospective study of the natural history of thoracic aortic aneurysms. Ann Thorac Surg 1997;63:1533e45. Juvonen T, Ergin MA, Galla JD, et al. Risk factors for rupture of chronic type B dissections. J Thorac Cardiovasc Surg 1999;117:776e86. Glower DD, Speier RH, White WD, et al. Management and long-term outcome of aortic dissection. Ann Surg 1991; 214:31e41. Gysi J, Schaffner T, Mohacsi P, et al. Early and late outcome of operated and nonoperated acute dissection of the descending aorta. Eur J Cardiothorac Surg 1997;11:1163e9 [discussion 1169e70]. Haverich A, Miller DC, Scott WC, et al. Acute and chronic aortic dissectionsddeterminants of long-term outcome for operative survivors. Circulation 1985; Sep;72(3 Pt 2): II 22e34.
Aortic growth rates in chronic aortic dissection A.M. Kellya,*, L.E. Quinta, B. Nanb, J. Zhengb, P. Cronina, G.M. Deebc, D.M. Williamsd a
Department of Radiology, Division of Thoracic Radiology, cDivision of Cardiac Surgery, Division of Vascular Interventional Imaging, University of Michigan Medical Center, and b School of Public Health, University of Michigan, Ann Arbor, Michigan, USA d
Received 27 September 2006; received in revised form 1 March 2007; accepted 26 April 2007
AIM: To determine and compare rates of descending aortic enlargement and complications in chronic aortic dissection with and without a proximal aortic graft. METHODS AND MATERIALS: Fifty-two patients with dissection involving the descending aorta and who had undergone at least two computed tomography (CT) examinations at our institution between November, 1993 and February, 2004 were identified, including 24 non-operated patients (four type A, 20 type B) and 28 operated patients (type A). CT examinations per patient ranged from two to 10, and follow-up ranged from 1e123 months (mean 49 months, median 38.5 months). On each CT image, the aortic short axis (SA), false lumen (FL), and true lumen (TL) diameters were measured at the longitudinal midpoint of the dissection and at the point of maximum aortic diameter. Complications were tabulated, including aortic rupture and aortic enlargement requiring surgery. RESULTS: For non-operated patients, the midpoint and maximum point SA, TL, and FL diameters increased significantly over time. For operated patients, the midpoint and maximum point SA and FL diameters increased significantly over time. In both groups, aortic enlargement was predominantly due to FL expansion. Diameter increases in nonoperated patients were significantly larger than those in operated patients. The rate of change in aortic diameter was constant, regardless of aortic size. Four non-operated and six operated patients developed aortic complications. CONCLUSIONS: In patients with a dissection involving the descending thoracic aorta, the FL increased in diameter over time, at a constant rate, and to a greater degree in non-operated patients (mostly type B) compared with operated patients (all type A). ª 2007 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved.
Introduction There are few data in the medical literature regarding the natural history of chronic type A or type B dissections involving the descending aorta.1,2 This applies to both non-operated dissections, as well as residual dissections in the descending aorta of patients with a proximal aortic graft. It is known that the false lumen (FL) tends to dilate over time, occasionally leading to rupture and/or the need for surgical repair. However, the
* Guarantor and correspondent: A.M. Kelly, Department of Radiology, Division of Thoracic Radiology, University of Michigan Hospitals, B1 132K Taubman Center/0302, 1500 East Medical Center Drive, Ann Arbor, Michigan 48109-0030, USA. Tel.: þ1 734 936 0024; fax: þ1 734 936 0013. E-mail address: [email protected] (A.M. Kelly).
rate of FL enlargement has not been well documented. Furthermore it is unknown whether non-operated type A or B dissections behave differently compared with dissections involving the descending aorta in a patient with previous, proximal aortic graft placement. The purpose of the present study was to determine and compare the rates of aortic enlargement and frequency of complications in these two types of patient.
Materials and methods Study population Institutional internal review board approval was obtained for this retrospective review and a waiver of informed consent was approved. All patients
0009-9260/$ - see front matter ª 2007 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.crad.2007.04.006
Aortic growth rates in chronic aortic dissection
seen at our institution between November, 1993 and February, 2004 with descending thoracic aortic dissection and who had undergone an inhouse computed tomography (CT) examination were identified using electronic databases from the Department of Radiology and the Division of Cardiothoracic Surgery. Of the 118 patients identified, 53 patients with only one CT study at our institution or with less than 1 month between the initial and the first follow-up CT exam were excluded. Also excluded were 13 patients with a completely thrombosed FL on the initial CT, based on review of the study. The reason to exclude these patients was that the focus of the study was on the natural history of the perfused FL. The remaining 52 patients were divided into two groups. The first group (non-operated group) comprised 24 patients with a non-operated type A or type B dissection (four type A and 20 type B). The second group (operated group) was composed of 28 patients with a type A dissection, proximal aortic repair, and a dissection flap in the descending aorta. The average age of patients was 59 years (range 34e81 years) in the operated group (22 males, six females) and 64 years (range 43e83 years) in the non-operated group (14 males, 10 females). The patients entered the study at different time points in the course of their disease. The first CT examination was considered the baseline or time zero. The number of CT examinations per patient ranged from two to 10 (with a mean of four examinations). The length of follow-up ranged from 1 month to 10 years 3 months, with the average length of follow-up being 49 months in the operated group and 34 months in the non-operated group. The median length of follow-up was 45.5 months in the operated group and 24 months in the non-operated group.
CT imaging Because of the long time frame of the study, extending from 1993 to 2004, the CT technique used has evolved. All studies were performed using helical technique, with a single row, four row, eight row or 16 row CT machine. Imaging was performed with the patient at end inspiration. Collimation ranged from 3 mm on the early images to 1.25 mm on the more recent images, with a pitch of 1e2. Patients were imaged from above the aortic arch down to at least the level of the aortic bifurcation using a bolus of intravenous contrast material administered via pump injector. CT was generally performed on an annual basis. However, if a patient
867
experienced new signs or symptoms of complications between scheduled examinations, interval examinations were performed.
Image interpretation Images were interpreted from hard-copy film photographed on a 20 to 1 format; soft-copy images were not available for the older studies. The hard-copy format was consistent for all studies with the same degree of image minification. The display field of view was consistent for all aorta studies at our institution. Patients with outside studies whose images did not have a reference measurement scale were excluded. All patients were reviewed using hard-copy images to improve consistency between measurements. The same author measured all images, using callipers. The craniocaudal extent of dissection was recorded in centimetres. The short axis (SA) diameter of the aorta was measured at the midpoint along the craniocaudal length; the location of the midpoint was recorded by using adjacent anatomic landmarks such as the carina, pulmonary veins, diaphragm, adrenal glands, and abdominal aortic branch vessels. Aortic SA diameter measurements were chosen rather than long axis diameters, because the former are less influenced by the orientation of the aorta with regard to the plane of section. True lumen (TL) and FL diameters perpendicular to the intimal flap were also measured at the midpoint along the craniocaudal length of the dissection; the sum of the TL and FL diameters often exceeded the SA measurement, because the flap was not necessarily perpendicular to the SA of the aorta (Fig. 1). The location of the maximum SA aortic diameter was determined by visual inspection of the images, and was recorded by using adjacent anatomic landmarks. The maximum SA aortic diameter, as well as TL and FL diameters perpendicular to the intimal flap at this level, were also measured and recorded. Periaortic haematoma was considered evidence for rupture.
Chart review For each patient, the medical record was searched for the following information: aortic complications, including aortic rupture and aortic enlargement requiring surgery; sex; age; history of smoking; history of chronic obstructive pulmonary disease (COPD); and hypertension (defined as a mean arterial pressure of greater than 110, with mean arterial pressure being calculated as diastolic pressure plus one third of the difference
868
A.M. Kelly et al.
Growth rate calculations and statistical analysis Individual aortic growth rates were calculated by dividing the change in aortic diameter (measured in millimetres) by time (measured in years) for each patient. Time points for evaluating individual growth rates were selected to be at least 1 year apart; this was done to minimize spurious shortterm changes. The group level growth rates for both non-operated and operated patient groups were estimated by using linear mixed effects models. Distributions of the anatomic levels of midpoints and maximum points in the non-operated and operated groups were analysed using the KolmogoroveSmirnov test. (This test is analogous to a t-test but makes no assumptions about the normality of the data.) The predictors of outcome were age, sex, history of smoking, history of COPD, and presence of hypertension. The outcome variables were craniocaudal extent of dissection and aortic diameters at midpoint (SA, TL, FL) and at maximum point (SA, TL, FL). Statistical analysis was performed on clinical and CT measurement variables, using a linear mixed effects model. Statistical significance was set at p < 0.05.
Results Midpoint diameters
Figure 1 (a) Aortic SA measurement (double arrow). (b) True and FL measurements (double arrows) were made perpendicular to the dissection flap.
between systolic and diastolic). The medical records were also searched for follow-up data on patients, including hypertensive status, control of hypertension, continued cigarette smoking, and cardiovascular risk factors.
The midpoint SA aortic diameters in both the non-operated and operated groups increased significantly over time compared with the baseline (p < 0.001), showing average rates of 1.5 and 0.8 mm/year, respectively. The mean growth rate in the non-operated patients was significantly higher than the rate in the operated patients (p ¼ 0.013; Table 1). The range in rates in individual patients in both groups combined was from 3 to 9 mm/year. Two patients had a decrease in midpoint diameter and in both of these cases; this was related to thrombosis of the FL. Increase over time in the aortic midpoint SA diameter appeared to be largely accounted for by an increase in the FL diameter (Table 1). The diameters of the TL in the non-operated group increased significantly over time compared with the baseline (p < 0.001), showing an average growth rate of 1 mm/year. However, the TL diameters in the operated group did not increase significantly over time (p ¼ 0.077) (Table 1). The range in rates in individual patients in both groups combined was from 1 to 11 mm/year. Negative rates could have been due to movement of the intimal
Aortic growth rates in chronic aortic dissection
Table 1 patients
869
The rate of change in aortic midpoint short axis, true lumen, and false lumen diameters in non-operated and operated True lumen diameter at midpoint
Short axis diameter at midpoint mm/year Non-operated patient group Operated patient group Difference between groups
1.5 0.8 0.7
SE 0.2 0.16 0.26
p-Value a
<0.001 <0.001a 0.013b
mm/year 1 0.3 0.7
SE 0.19 0.15 0.25
False lumen diameter at midpoint p-Value a
<0.001 0.077a 0.003b
mm/year
SE
p-Value
1.4 0.7 0.7
0.23 0.18 0.29
<0.001a <0.001a 0.030b
SE, standard error. a p-Value indicates the level of significance for change in diameter compared with the baseline. b p-Value indicates the level of significance for difference in growth rates between operated and non-operated groups.
flap between systole and diastole or measurement error. The diameters of the FL in both the nonoperated and operated groups increased significantly over time compared with the baseline (p < 0.001), showing average rates of 1.4 and 0.7 mm/year, respectively. The growth rate in the non-operated patients was significantly higher than the rate in the operated patients (p ¼ 0.03; Table 1). All patients maintained at least partial patency of the FL, and the growth rates in individual patients in both groups combined ranged from 4 to 6 mm/year. Negative rates could have been due to movement of the intimal flap between systole and diastole or actual decrease in size of FL or measurement error.
Maximum point diameters The maximum point SA aortic diameters in both the non-operated and operated groups increased significantly over time compared with the baseline (p ¼ 0.001 and 0.004, respectively), showing average rates of 1.2 and 0.5 mm/year, respectively. The mean growth rate in the non-operated patients was significantly higher than the rate in the operated patients (p ¼ 0.007; Table 2). The range in rates in individual patients in both groups combined was from 0.2 to 9.5 mm/year. Five patients had a decrease in maximum point diameter and in four of these patients; this was related to
thrombosis of the FL. In the fifth patient, diameter decrease may have been due to the dynamic nature of the aorta between systole and diastole or possibly measurement error. Increase over time in the aortic maximum point SA diameter in operated patients appeared to be largely accounted for by an increase in the TL diameter, and in non-operated patients, was largely due to FL diameter increase (Table 2). The diameters of the TL in both the nonoperated and operated groups increased significantly over time compared with the baseline (p ¼ 0.004 and 0.001, respectively), showing average growth rates of 0.8 and 0.7 mm/year (Table 2). The rate of change in TL diameter was not significantly greater for the non-operated patients compared with the operated patients (p ¼ 0.83). The range in rates in individual patients in both groups combined was from 2.5 to 12.4 mm/ year. Negative rates could have been due to movement of the intimal flap between systole and diastole or measurement error. The diameter of the FL in the non-operated group increased significantly over time compared with the baseline (p ¼ 0.001), showing an average rate of 0.9 mm/year. However, the FL diameter in the operated group did not increase significantly over time (p ¼ 0.296). All patients maintained at least partial patency of the FL, and the growth rates in individual patients in both groups combined ranged from 15 to 10 mm/year (Table 2).
Table 2 The rate of change in aortic maximum point short axis, true lumen, and false lumen diameters in non-operated and operated patients Short axis diameter at maximum point
Non-operated patient group Operated patient group Difference between groups
True lumen diameter at maximum point
False lumen diameter at maximum point
mm/year
SE
p-Value
mm/year
SE
p-Value
mm/year
SE
p-Value
1.2 0.5 0.7
0.2 0.17 0.27
<0.001a 0.004a 0.007b
0.8 0.7 0.1
0.25 0.2 0.32
0.004a 0.001a 0.832b
0.9 0.2 0.7
0.28 0.22 0.35
0.001a 0.296a 0.047b
SE, standard error. a p-Value indicates the level of significance for change in diameter compared with the baseline. b p-Value indicates the level of significance for difference in growth rates between operated and non-operated groups.
870
Negative rates could have been due to movement of the intimal flap between systole and diastole, actual decrease in size of the FL or measurement error.
Growth rate constancy The rate of change in aortic midpoint (SA, TL, and FL) and maximum point (SA, TL and FL) diameters was constant in both non-operated and operated patients, regardless of the aortic diameter. In other words, increased aortic diameter did not correlate with an increased growth rate (p > 0.05).
Distribution of initial midpoint and maximum point locations and change in craniocaudal extent over time In almost all patients, the anatomic level of the midpoint was in the descending aorta at, or close to, the diaphragmatic hiatus. The distributions of craniocaudal midpoint locations did not differ significantly between the non-operated and operated groups (p ¼ 0.593, KolmogoroveSmirnov test). In two thirds of patients, the anatomic level of the maximum point was at the distal aortic arch or proximal descending thoracic aorta (at or close to the level of carina). The distributions of maximum point locations did not differ significantly between the non-operated and operated groups (p ¼ 0.892, KolmogoroveSmirnov test). The craniocaudal extent (length) of the dissection flap did not change significantly over time in either the non-operated or the operated groups.
Effect of hypertension After the initial clinical and imaging evaluation, most patients only return periodically for surveillance CT examinations because our institution is a tertiary care centre; these patients are then seen in the clinics only if the CT examination suggests a new or progressing abnormality. Generally, routine clinical follow-up is performed by the patients’ primary-care physicians outside of our hospital system. Because of this, the serial blood pressure measurements of many of the patients in this study were not available, and therefore, there were too few measurements to correlate hypertension with rate of aortic growth.
Effect of smoking, COPD, sex, and age There was no significant difference between smokers and non-smokers, between patients with
A.M. Kelly et al.
and without COPD, or between males and females with respect to the rate of aortic midpoint enlargement (SA, TL, and FL). With respect to change in the TL and FL diameters at the midpoint, there was no statistically significant difference in rate of change with different ages. Although the aortic midpoint SA diameter increased at a statistically significantly greater rate in younger patients compared with older patients (p ¼ 0.025), the difference was small (0.03 mm/year for each year younger) and not felt to be clinically significant. There was no significant difference between smokers and non-smokers or between males and females with respect to rate of aortic maximum point enlargement (SA, TL and FL; p > 0.05). Age did not have a significant effect on rate of maximum point enlargement (p > 0.05). With respect to change in the FL diameters at the aortic maximum point, there was a borderline statistically significant difference in the rate of change with different ages, but the rate of change was minimal (0.02 mm/year for each year younger) and felt not to be clinically significant (p ¼ 0.04). The aortic maximum point SA and FL diameters increased at a statistically significantly greater rate in patients without COPD compared with COPD patients (p ¼ 0.045 and 0.020, respectively), which is in contrast to the findings seen with respect to the midpoint. None of the above findings regarding statistical significance changed when the four patients with type A dissections in the non-operated group were excluded from analysis.
Complications In the non-operated group, four of 24 (17%) patients developed aortic complications during the follow-up period of this study, and all required subsequent aortic surgery. Two patients developed aortic rupture 3.5 and 8 years (Fig. 2), respectively, after the initial dissection event, showing maximum SA diameters of 6.3 and 5.8 cm, respectively, at the time of rupture. Two other patients developed marked aortic dilatation requiring surgery. In the first patient, the aorta SA maximum diameter increased in size from 4.2 to 5.3 cm over a 5.5 year interval. In the second patient, the aorta SA maximum diameter increased in size from 4.9 to 6.0 cm over a 4.75 year interval (Fig. 3). The level at which aortic enlargement or rupture occurred in these four patients varied from the distal arch to the mid descending thoracic aorta, close to the level of the pulmonary arteries. The SA midpoint and maximum point aortic growth rates in these four patients ranged from 1e3.4 and 0.1e4.2 mm/year, respectively.
Aortic growth rates in chronic aortic dissection
871
In the operated group, six of the 28 (21%) patients developed aortic complications during the follow-up period of this study, and all required further aortic surgery. One patient showed aortic rupture 3 years after surgery, showing a maximum SA diameter of 8.6 cm at the time of rupture. Five patients either had or developed pronounced aortic dilatation (maximum SA diameters 4e7.1 cm) an average of 2 years and 10 months after surgery (range 1e6 years and 5 months). The area of aortic enlargement or rupture was in the distal arch or proximal descending thoracic aorta, close to the tracheal carina, in four of the six patients, at the level of the adrenals in one patient, and there was diffuse enlargement throughout the descending thoracic aorta in one patient (Fig. 4). The SA midpoint and maximum point aortic growth rates in this group of six patients ranged from 0e6 and 0e10 mm/year, respectively.
Discussion
Figure 2 Increase in FL measurement and interval rupture of the FL between 1993 (a) and 1994 (b) in a 66-year-old non-operated man, necessitating urgent surgery.
Thoracic aortic aneurysms have been reported to grow at a rate of approximately 0.7e1.9 mm/ year.3,4 However, the rates of aortic enlargement in aortic dissections have been scantily documented in the medical literature.5,6 In the present study, the finding of 1.5 mm/year SA aortic diameter mean growth rate for non-operated patients is somewhat less than the mean SA growth rate of 3.3 mm/year in 62 non-operated type B patients reported by Sueyoshi et al.5 In the present study a patient by patient analysis was used, whereas in the study of Sueyoshi and colleagues5 the data were analysed based on aortic segments; therefore, it is unclear if the results are directly comparable. The latter study was potentially limited by relatively thick (5 mm) sections and a large proportion of non-helically acquired images (approximately 37%), thus leading to somewhat limited spatial resolution, by current standards. The report of Onitsuka et al.6 found a mean growth rate of 3.8 mm/year in non-operated type B patients; however, that study used long axis aortic measurements and such measurements may not reflect true diameters, particularly in tortuous aortas. In addition, although the study of Onitsuka et al. included a substantial number of patients (76), it is not specified how many patients had follow-up CT examinations, and therefore, it is not clear how many patients contributed to the mean growth rate reported. For operated patients, the data from the present study showed a mean SA midpoint growth rate of 0.8 mm/year and a SA maximum point growth
872
A.M. Kelly et al.
rate of 0.5 mm/year. To the authors’ knowledge, there are no published reports documenting aortic growth rates in operated aortic dissection patients. In vitro aortic dissection models have demonstrated that increase in total cross-sectional area with increasing hydrostatic pressure occurs primarily due to enlargement of the FL.7 The present study confirmed that FL expansion was indeed the major contributor to aortic enlargement at the midpoint in patients with either operated or nonoperated aortic dissection, although the degree of enlargement was significantly greater in nonoperated patients. The data for maximum point were somewhat contradictory: although FL expansion appeared to account for the majority of aortic enlargements in non-operated patients, TL expansion appeared to be the dominant force in the operated group. It is possible that the data for maximum point measurements were less robust than the data for midpoint measurements; the maximum point often occurred in the region where the distal aortic arch curved into the proximal descending aorta, and the curvature may have led to inaccurate measurements. Using the midpoint and maximum point measurements, the FL expanded at a higher rate in non-operated patients compared with operated patients. The operated group was composed of patients with type A dissections, whereas the nonoperated group contained mostly patients with type B dissections. These two diseases are likely due to different types of pathophysiology,8 and it is possible that the non-operated patients had intrinsically more disease in the descending aorta compared with the operated group, leading to the increased rate of growth. Alternatively, surgical repair of the proximal aorta may have lessened the degree to which the distal aorta dilated over time, due to alteration in haemodynamics. In addition, patients who underwent surgery may have had better blood pressure control compared with non-surgical patients, thus lessening the degree of subsequent aortic dilatation. Several investigators have reported that thrombosis of the FL has a better prognosis compared with a patent FL, with regard to subsequent aortic enlargement and/or dissection related death.5,9,10e13 This was not verified in the present study, because all of the patients had flow in the
Figure 3 Increase in FL measurement between October 1998 (a) and July 2003 (b) in a 59-year-old man. This nonoperated patient required surgery because of marked FL enlargement, with a 1.9 cm interval increase in diameter.
Aortic growth rates in chronic aortic dissection
873
Figure 4 Increase in FL measurements between June 2002 (a and b) and September 2004 (c and d), in a 57-year-old man. This previously operated patient required further surgery because of marked FL enlargement, with a 2.7 cm interval increase in diameter.
FL on the initial CT examination, and none had complete thrombosis of the FL on follow-up. Previous studies have suggested that an increase in aortic diameter leads to an increase in the rate of aortic diameter growth in aneurysms and dissections.11,14,15 However, Sueyoshi and colleagues5 did not find such an association: patients with initial aortic diameters greater than 40 mm actually showed a lower mean growth rate compared with those with an initial diameter less than 40 mm, although the difference was not statistically significant. These authors hypothesized that the perpendicular stress on the aortic wall is lessened in the presence of two lumens, and therefore dissections behave differently compared with aneurysms in this respect. In the present study
there were no data regarding aortic diameter at the time of the initial dissection event, because patients entered the study at various time points in their disease. However, by using multiple time points in each patient, rather than just a beginning and an ending point, growth rates could be compared over time; the data showed that aortic growth rates at midpoint and maximum point were constant, supporting the premise that aortic diameter does not affect growth rate in this patient population. Published studies have reported that a history of smoking,4 COPD,5,16,17 female sex,5 and increasing age16,17 correlate with an increased chance of rupture and increased aortic growth rate in aneurysms and dissections. However, other studies have
874
disputed some of these findings and, in fact, found contradictory results.5,16 In the present study no significant correlation between aortic midpoint growth rates and smoking, sex, age or COPD was found. There was no correlation between aortic maximum point growth rates and smoking, sex, or age, although there was an inverse relationship between COPD and growth rate of borderline statistical significance. In non-operated patients, chronic dissectionrelated complications, such as aortic rupture, increase in aortic diameter >50 mm, mean aortic enlargement rate >5 mm/year, surgery and/or death, have been reported at rates of 12e33%.6,18,19 The rate of 17% from the present study falls at lower end of this range. Six of 28 (21%) patients in the operated group experienced dissection-related complications, including aortic dilatation or rupture requiring subsequent surgery downstream from the original repair. This is comparable with published rates of 11e30% from other investigations.1,18e20 The limitations of the present study included the use of axial images rather than images perpendicular to the long axis of the aorta; such images were used because only hard-copy, conventional axial images were available for the older examinations. This technique could potentially have led to inaccuracies in aortic diameter measurement; however, this problem may have been alleviated by the use of SA measurements, which are less dependent on the orientation of the aorta with regard to the image plane compared with long axis measurements. However, aortic dissections tend to be oval in cross-section due to greater expansion of the FL relative to the TL; thus, SA measurements might underestimate the degree of FL expansion. In the present study hard copy rather than soft copy images were used. Many of the patients had a long follow-up and in the earlier years, images were on hard copy. All images were reviewed as hard copy to maintain consistency to reduce errors. Review of soft-copy images would have been preferable. Two-dimensional (2D) length measurements rather than three-dimensional (3D) volumes were measured in the present study, because currently available 3D tools could not be employed on the older images; volume measurements would likely be more accurate than diameters in assessing size. A limitation in the measurement of TL and FL diameters was the static nature of the images: although, in reality, the dissection flap tends to move back and forth during systole and diastole, the images capture only a moment in time, and
A.M. Kelly et al.
measurements were made at a single, flap position. Finally, in the few patients with aortic rupture, the presence of periaortic haematoma may have led to inaccuracies in measuring aortic diameters. The present study is also limited by the lack of follow-up data with respect to clinical parameters, which can affect dissection outcomes such as hypertensive status, medication control of hypertension, continued cigarette smoking, and the presence of other cardiovascular disease. It was difficult to obtain this clinical information from the patients’ records because many of the patients received primary care elsewhere. The data that were obtained on a few patients were insufficient to draw statistical inferences. A large prospective study of a similar group of patients would be of benefit to assess the contribution of the clinical factors above to aortic dissection outcomes. In addition, patients entered the study at different time points, both chronologically and with respect to their aortic dissection disease. Some patients will have entered the study much later in the course of their aortic dissection relative to others. Some patients will have entered the study in the immediate post-surgical period and others much later after surgery. This may contribute to differences in disease course and therefore mask or exaggerate differences the two groups. In conclusion, in patients with a dissection involving the descending thoracic aorta, the FL increased in diameter over time, and the rate of increase in diameter was constant, regardless of aortic diameter. FL dilatation occurred to a significantly greater degree in non-operated patients (mostly type B dissections) compared with patients with a type A dissection and a proximal aortic graft. This greater increase in aortic diameter in type B patients compared with type A patients may reflect different pathophysiologies, rather than the effect of a proximal aortic graft. Complication rates in the present study were 17% for non-operated patients and 21% for operated patients.
Acknowledgements This study was funded in part by a GE-AUR Radiology Research Academic Fellowship.
References 1. Heinemann M, Las J, Crack M, et al. Thoracic aortic aneurysms after acute type A aortic dissection: necessity for follow-up. Ann Thorac Surg 1990;49:580e4.
Aortic growth rates in chronic aortic dissection
2. Wheat Jr MW, Palmer RF. Dissecting aneurysms of the aorta. Curr Probl Surg 1971; 1e43. 3. Davies RR, Goldstein LJ, Coady MA, et al. Yearly rupture or dissection rates for thoracic aortic aneurysms: simple prediction based on size. Ann Thorac Surg 2002;73:17e27 [discussion 27e8]. 4. Coady MA, Rizzo JA, Hammond GL, et al. What is the appropriate size criterion for resection of thoracic aortic aneurysms? J Thorac Cardiovasc Surg 1997;113:476e91. 5. Sueyoshi E, Sakamoto I, Hayashi K, et al. Growth rate of aortic diameter in patients with type B aortic dissection during the chronic phase. Circulation 2004;110(Suppl. 1): II256e61. 6. Onitsuka S, Akashi H, Tayama K, et al. Long-term outcome and prognostic predictors of medically treated acute type B aortic dissections. Ann Thorac Surg 2004;78:1268e73. 7. Williams DM, LePage MA, Lee DY. The dissected aorta: part I. Early anatomic changes in an in vitro model. Radiology 1997;203:23e31. 8. Coady MA, Rizzo JA, Goldstein LJ, et al. Diseases of the aorta. Natural history, pathogenesis and etiology of thoracic aortic aneurysms and dissections. Cardiol Clin 1999; 17:615e35. 9. Bernard Y, Zimmermann H, Chocron S, et al. False lumen patency as a predictor of late outcome in aortic dissection. Am J Cardiol 2001;87:1378e82. 10. Akutsu K, Nejima J, Kiuchi K, et al. Effects of the patent false lumen on the long-term outcome of type B acute aortic dissection. Eur J Cardiothorac Surg 2004;26:359e66. 11. Marui A, Mochizuki T, Mitsui N, et al. Toward the best treatment for uncomplicated patients with type B acute aortic
875
12.
13.
14.
15.
16.
17.
18.
19.
20.
dissection: a consideration for sound surgical indication. Circulation 1999;100(Suppl.19):II275e80. Kozai Y, Watanabe S, Yonezawa M, et al. Long-term prognosis of acute aortic dissection with medical treatment: a survey of 263 unoperated patients. Jpn Circ J 2001;65:359e63. Fattori R, Bacchi-Reggiani L, Bertaccini P, et al. Evolution of aortic dissection after surgical repair. Am J Cardiol 2000; 86:868e72. Dapunt OE, Galla JD, Sadeghi AM, et al. The natural history of thoracic aortic aneurysms. J Thorac Cardiovasc Surg 1994;107:1323e32. discussion 1332e3. Kato M, Bai H, Sato K, et al. Determining surgical indications for acute type B dissection based on enlargement of aortic diameter during the chronic phase. Circulation 1995; 92(Suppl.9):II107e12. Juvonen T, Ergin MA, Galla JD, et al. Prospective study of the natural history of thoracic aortic aneurysms. Ann Thorac Surg 1997;63:1533e45. Juvonen T, Ergin MA, Galla JD, et al. Risk factors for rupture of chronic type B dissections. J Thorac Cardiovasc Surg 1999;117:776e86. Glower DD, Speier RH, White WD, et al. Management and long-term outcome of aortic dissection. Ann Surg 1991; 214:31e41. Gysi J, Schaffner T, Mohacsi P, et al. Early and late outcome of operated and nonoperated acute dissection of the descending aorta. Eur J Cardiothorac Surg 1997;11:1163e9 [discussion 1169e70]. Haverich A, Miller DC, Scott WC, et al. Acute and chronic aortic dissectionsddeterminants of long-term outcome for operative survivors. Circulation 1985; Sep;72(3 Pt 2): II 22e34.