Patients after coarctation repair still have an increased risk of cardiovascular or cerebrovascular events. This has been explained by the persisting hypertension and alterations in the peripheral vessels. However, involvement of the central vessels such as the retinal arteries is virtually unknown. A total of 34 patients after coarctation repair (22 men and 12 women; 23 to 58 years old, age range 0 to 32 years at surgical repair) and 34 nonhypertensive controls underwent structural and functional retinal vessel analysis. Using structural analysis, the vessel diameters were measured. Using functional analysis, the endothelium-dependent vessel dilation in response to flicker light stimulation was assessed. In the patients after coarctation repair, the retinal arteriolar diameter was significantly reduced compared to that of the controls (median 182 μm, first to third quartile 171 to 197; vs 197 μm, first to third quartile 193 to 206; p <0.001). These findings were independent of the peripheral blood pressure and age at intervention. No differences were found for venules. The functional analysis findings were not different between the patients and controls (maximum dilation 3.5%, first to third quartile 2.1% to 4.5% vs 3.6%, first to third quartile 2.2% to 4.3%; p = 0.81), indicating preserved autoregulative mechanisms. In conclusion, the retinal artery diameter is reduced in patients after coarctation repair, independent of their current blood pressure level and age at intervention. As a structural marker of chronic vessel damage associated with past, current, or future hypertension, retinal arteriolar narrowing has been linked to stroke incidence. These results indicate an involvement of cerebral microcirculation in aortic coarctation, despite timely repair, and might contribute to explain the increased rate of cerebrovascular events in such patients.
A valid method to assess structural or functional disorders of cerebral microcirculation is static (structural) and dynamic (functional) retinal vessel analysis. As measured by static analysis, one finds significantly altered vessel diameters in the presence of arterial hypertension or diabetes mellitus. The arteriolar/venular ratio of the retinal vessels (AVR) has proved to be an independent predictor of cardiovascular and cerebrovascular risk, at least in some populations. Dynamic analysis allows the application of flicker light, inducing a dilation of the retinal vessels that is partly mediated by nitric oxide and indicative of endothelial dysfunction if reduced. Altered reactions have been shown for patients with diabetes and hypertension. The main goal of the present study was to investigate the involvement of the central vessels in patients with aortic coarctation by structural and functional retinal vessel analysis to find possible explanations for the increased incidence of cerebrovascular accidents in such patients.
Methods
We included 34 patients (22 men and 12 women, mean age 38 ± 10 years, range 23 to 58) with successfully repaired aortic coarctation and without recoarctation, as defined by a peak pressure gradient of >20 mm Hg between the upper and lower extremities. None of these patients had additional congenital or acquired heart disease. The patients were selected sequentially during their annual coarctation follow-up examination. The mean age at intervention was 11 ± 9 years (range 0 to 32). The participants were included in the study 27 ± 5 years (range 13 to 33) after surgery. Of the 34 patients, 28 had undergone resection and end-to-end-anastomosis, 5 had received Dacron prostheses (interposition graft), and 1 patient had received a Gore-Tex prosthesis (on-lay patch). The blood pressure was measured with the patient ambulatory. A mean value of >135/85 mm Hg during the day and >120/75 mm Hg during the night (n = 8) and/or antihypertensive treatment (n = 11) was considered to indicate hypertension. Those with hypertension were mainly treated with combinations of angiotensin-converting enzyme inhibitors, AT 1 -receptor antagonists, and β blockers. The body mass index was 24.4 ± 3.6 kg/m 2 . The patients were compared with an age- and gender-matched healthy control group (age 38 ± 10 years, body mass index 23.2 ± 2.4 kg/m 2 ) without past or present hypertension. The control group consisted of those who had undergone examination in our clinic for preventive purposes, including detailed medical history, physical examination, electrocardiography, bicycle or treadmill ergometry, and echocardiography. Those with an elevated blood pressure at rest or during exercise were not included in the present study. Additional risk factors such as smoking (yes, exsmoker, and no, 5, 5, and 24 patients and 1, 8, and 25 controls) and a family history of cardiovascular events (myocardial infarction or stroke at age ≤60 years in 4 patients and 1 control) were recorded; none had diabetes. All participants gave their written and informed consent, and the university ethical board approved the study.
Retinal vessel analysis was performed using the Dynamic Vessel Analyzer (Imedos, Jena, Germany), which includes a fundus camera allowing normal retina examination and photography, and real-time measurement of retinal vessel diameter changes in reaction to external (flicker light) or internal (blood pressure) stimuli. For valid results, the application of tropicamide 5 mg/ml was necessary. For structural analysis, photographs were taken from either the left or right retina with an angle of 30° and the optic disk in the center. Retinal arterioles and venules coursing through an area of 0.5 to 1 disk diameter from the margin of the optic disk were identified using analyzing software (Imedos, Jena, Germany), and their diameters were summarized as the average central arteriolar and venular equivalents (CRAEs and CRVEs), using formulas that have been previously published. The relative caliber of these vessels was then calculated as the AVR.
For functional analysis, we performed a 50-second baseline analysis, followed by the application of flicker light with a frequency of 12.5 Hz and a bright/dark ratio of 25:1 for 20 seconds. This period was followed by 80 seconds of baseline observation. The whole procedure was repeated twice for averaged values, resulting in a total examination time of about 6 minutes per eye. The criteria for the selection of the vessel segments were a location within a circular area of 2 disk diameters, no crossing or bifurcation in the segment, a minimum length 1 mm, a distance to neighboring vessels of ≥1 diameter, and sufficient contrast with the surrounding fundus. Slight ocular movements were corrected automatically. All examinations were performed by a single, experienced observer who was unaware of the patients’ medical history.
Statistical analysis was performed using the Statistical Package for Social Sciences, version 16.0 (SPSS, Chicago, Illinois). The Kolmogorov-Smirnov test was applied to determine normal distribution. The data are presented as the mean ± SD or median, ranges, and interquartile ranges, as appropriate. Because the data from retinal vessel analysis were not normally distributed, we used Wilcoxon’s rank sum test for comparisons between the patients and controls and the Mann-Whitney U test for comparisons between patient subgroups. For bivariate relations, we performed Spearman-rho correlation tests, and for multivariate analysis, analysis of variance and linear regression models with AVR, CRAE, and CRVE as dependent variables and blood pressure, gender, age, and body mass index as explanatory variables. A value of p <0.05 was considered statistically significant.
Results
We found a significant reduction of the AVR in patients after coarctation repair compared to the healthy controls (median 0.80, first to third quartile 0.75 to 0.86 vs median 0.89, first to third quartile 0.84 to 0.90; p <0.001; Figure 1 ). To determine whether this alteration was caused by smaller arteries or, instead, by larger veins, we analyzed both the summarized arteriolar and venular diameters expressed as the CRAEs or CRVEs. Only the arterioles showed a significant reduction between the patients and controls (182 μm, first to third quartile 171 to 197 vs 197 μm, first to third quartile 193 to 206; p <0.001). In contrast, the venules had the same size (232 μm, first to third quartile 221 to 246 vs 229 μm, first to third quartile 219 to 236; p = 0.49; Figure 1 ).
Because it is known that arterial hypertension can lead to a reduction of the AVR, we divided the patients into those with persisting hypertension despite surgical repair and those with normal blood pressure ( Table 1 ). Both patient subgroups still showed significantly reduced retinal arteriolar diameters compared to their corresponding age- and gender-matched controls, indicating that the alterations occurred independently of the current blood pressure level.
| Variable | Group I (CoA HT; n = 19) | Group II (Controls; n = 19) | Group III (CoA NHT; n = 15) | Group IV (Controls; n = 15) | p Value | ||
|---|---|---|---|---|---|---|---|
| I–III | I–II | III–IV | |||||
| Arteriolar/venular ratio | 0.81 (0.08) | 0.86 (0.07) | 0.81 (0.11) | 0.89 (0.05) | 0.67 | <0.01 ⁎ | 0.02 ⁎ |
| Central arteriolar equivalent (μm) | 180 (27) | 197 (12) | 182 (25) | 199 (21) | 0.84 | 0.01 ⁎ | 0.04 ⁎ |
| Central venular equivalent (μm) | 230 (19) | 229 (21) | 227 (34) | 225 (12) | 0.81 | 0.68 | 0.55 |
Another important question in investigating patients with coarctation repair is the influence of early surgery on the clinical outcome, because particularly long-standing hypertension before surgery might have damaged the vascular bed. We compared patients with early and late repair ( Table 2 ). Again, no significant differences were found between the 2 groups for AVR, CRAEs, and CRVEs.
| Variable | Group I (Early Repair; n = 6) | Group II (Controls; n = 6) | Group III (Late Repair; n = 28) | Group IV (Controls; n = 28) | p Value | ||
|---|---|---|---|---|---|---|---|
| I–III | I–II | III–IV | |||||
| Arteriolar/venular ratio | 0.79 (0.07) | 0.90 (0.06) | 0.80 (0.11) | 0.87 (0.06) | 0.72 | 0.02 ⁎ | <0.01 ⁎ |
| Central arteriolar equivalent (μm) | 180 (27) | 206 (14) | 183 (27) | 197 (14) | 0.70 | 0.02 ⁎ | <0.01 ⁎ |
| Central venular equivalent (μm) | 230 (19) | 230 (25) | 232 (26) | 229 (17) | 0.67 | 0.91 | 0.34 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
