Portal de Periódicos da CAPES

Lack of effect of short-term treatment with Amlodipine and Lisinopril on retinal autoregulation in normotensive patients with type 1 diabetes and mild diabetic retinopathy

2010; Wiley; Volume: 89; Issue: 8 Linguagem: Inglês

10.1111/j.1755-3768.2009.01847.x

1755-3768

Jesper Mehlsen, Peter Jeppesen, Mogens Erlandsen, Per Løgstrup Poulsen, Toke Bek,

Retinal Imaging and Analysis

Purpose: Diabetic retinopathy is characterized by morphological changes in the retina secondary to disturbances in retinal blood flow. It has been shown that antihypertensive treatment has a protective effect on the development of diabetic retinopathy, and evidence suggests that inhibitors of the renin–angiotensin system have a protective effect beyond the antihypertensive effect. The background for this additional effect is unknown but might be related to an effect on retinal autoregulation. Methods: In a double-blinded, two-way cross-over study, 25 normotensive patients with type 1 diabetes (T1D) aged 20.6–33.9 (mean 27.9) with mild retinopathy were randomized to receive either 5 mg of the calcium channel blocker (CCB) amlodipine for 14 days followed by a washout period and treatment with 10 mg of the angiotensin converting enzyme (ACE) inhibitor lisinopril for another 14 days or the two treatments in the reverse order. Using a Dynamic Vessel Analyzer (DVA), the diameter response of retinal arterioles during an acute increase in the blood pressure induced by isometric exercise, during flicker stimulation and during both stimulus conditions simultaneously was studied before and during the two treatments periods. Results: Amlodipine and lisinopril induced a similar non-significant decrease in the arterial blood pressure. At baseline, the arterial diameter decreased by 2.4 ± 0.9% (p = 0.004) during isometric exercise, increased by 2.2 ± 0.9% (p = 0.019) during flicker stimulation and increased by 1.8 ± 0.9% (p = 0.03) during the combined stimulus conditions. Neither of the antihypertensive drugs amlodipine (p = 0.76) or lisinopril (p = 0.11) changed the diameter response of retinal vessels significantly; however, the two treatments induced a different response in the veins during combined exercise and flicker (p = 0.021). Conclusions: Short-term treatment with amlodipine and lisinopril had no significant effect on retinal autoregulation in young normotensive patients with T1D and mild retinopathy, and this lack of effect was similar for the two drugs. A possible normalizing effect of antihypertensive treatment on retinal autoregulation was not observed; however, it might take longer time to improve autoregulation than to reduce the arterial blood pressure. Diabetic retinopathy is characterized by morphological lesions in the ocular fundus secondary to disturbances in retinal perfusion. These perfusion disturbances are assumed to involve impaired autoregulation so that changes in the arterial blood pressure are transmitted to the capillary bed to initiate the formation of microaneurysms, haemorrhages, exudates and oedema (Grunwald et al. 1996; Frank 2004; Jeppesen et al. 2004; Flammer & Mozaffarieh 2008). In accordance with this, it has been shown in patients with diabetes that a reduction in the arterial blood pressure, even in normotensive patients, may reduce the risk of developing diabetic retinopathy (Marshall et al. 1993; Estacio et al. 2000; Matthews et al. 2004; Gallego et al. 2008). Furthermore, evidence suggests that treatment with angiotensin converting enzyme (ACE) inhibitors has a protective effect on the retina, which is additional to the blood pressure lowering effect (Larsen et al. 1990; Chase et al. 1993; Sleight 2000; Neroev & Riabina 2006). This effect might potentially be because of a normalizing effect on retinal autoregulation mediated by blockage of ACE in the retina (Wilkinson-Berka 2006). However, the effects of ACE inhibitor and other antihypertensive drugs on retinal autoregulation have not been studied in detail. The purposes of the present study were as follows: (i) To study the effect of antihypertensive treatment on retinal autoregulation in normotensive patients with type 1 diabetes mellitus (T1D) with mild diabetic retinopathy and (ii) To compare the effect of antagonists to ACE and calcium channels on retinal autoregulation in this patient group. In a randomized double-blinded cross-over study, 25 patients with T1D and mild diabetic retinopathy were allocated to receive 14 days treatment with the ACE inhibitor lisinopril (10 mg) followed by the calcium channel blocker (CCB) amlodipine (5 mg) or to receive these treatments in the reverse order. The diameter response of retinal arterioles secondary to an acute increase in the arterial blood pressure, during flicker stimulation and during both stimulus conditions simultaneously, was examined using the Dynamic Vessel Analyzer (DVA) before and at the end of the two treatment phases. Twenty-five patients with T1D (11 men and 14 women) with no current or previous history of arterial hypertension or other known cardiovascular or eye diseases were included. The patients were recruited successively from the screening clinic for diabetic retinopathy at the Department of Ophthalmology, Aarhus University Hospital, among non-pregnant and non-lactating patients who did not receive any antihypertensive treatment. None of the patients had consumed alcohol for at least 12 hr prior to an examination. Additionally, eligible patients had mild diabetic retinopathy defined as <10 microaneurysms/dot haemorrhages in the worst eye at a screening examination not more than 3 months preceding the inclusion. The study was approved by the local ethics committee and followed the guidelines of the Declaration of Helsinki. Baseline data of the included patients are shown in Table 1. The patients were randomized to receive either antihypertensive treatment with amlodipine 5 mg/day for 14 days, followed by a washout period of 14–21 days and 14 days treatment with lisinopril 10 mg/day (group A-L) or to receive the two medications in the reverse order (group L-A). The hospital pharmacy controlled encapsulation of the drugs and randomization of the patients to ensure the double-blinded design. To document compliance, the patients returned the containers of leftover medication after each treatment period. The antihypertensive drugs were used in accordance with the manufacturer's guidelines, which was the criterion for obtaining permission for the study from the national medical authorities. Each examination consisting of a general physical part and an ophthalmological part was performed four times, i.e. on the two first days of treatment, but immediately preceding the commencement of treatment, and on the two last days of treatment. During the day preceding each of the four examinations, 24-hr ambulatory blood pressure was monitored using SpaceLabs model 90207, Redmond, WA, USA (O'Brien et al. 2000). The monitors were programmed to perform blood pressure measurements every 20 min, using an oscillometric technique. Prior to each examination, the blood glucose was measured on a finger blood droplet (Ascensia Lite, Basel, Switzerland), and a 10-ml blood sample was collected from an antecubital vein for the analysis of HbA1c and plasma creatinine, sodium, potassium and albumin. During the examination, the systolic blood pressure (BPsys) and the diastolic blood pressure (BPdia) were measured on the upper left arm in sitting position using an automated oscillometric technique (Omron M4; Omron, Kyoto, Japan). All patients underwent a routine ophthalmological examination including measurement of best-corrected visual acuity (VA) using ETDRS charts. Intra-ocular pressure (IOP) was measured using pneumotonometry (Nidek NT-3000; Nidek, Gamagori, Japan). Pupil dilatation was induced with phenylephrine 10% (SAD, Copenhagen, Denmark) and tropicamide 1% (Alcon, Fort Worth, TX, USA) eye drops followed by slit lamp examination, 60 degrees fundus photography using a Canon CF 60Z fundus camera, and central macular thickness was measured by optical coherence tomography (OCT) scanning (Stratus, Zeiss, Germany) using the central star scanning mode which consists of six radial scans with a length of 6 mm, each centred on the point of fixation and separated by 30 degrees. Finally, the diameter response of retinal arterioles was assessed using the Dynamic Vessel Analyzer (DVA, Imedos, Germany). The DVA is a further development of a Retinal Vessel Analyzer which is a commercially available system that consists of a Zeiss FF450 fundus camera, a digital band recorder and a real-time monitor connected to a personal computer with image-analysis software which is able to determine the arterial and venous diameters (Vilser et al. 1997). During an examination, a digital video recording of the retinal fundus is grabbed, analysed real time and stored on videotape for later re-analysis. The examiner places one or two markers on the vessel(s) to be analysed. In the present study, the markers were placed on a temporal arteriole and its adjoining venule within three disc diameters from the optic disc. The upper or the lower temporal vascular arcade with fewest branchings in this area was selected. During a recording, the computer software corrects for displacements because of eye movements at video speed (25 times per second) and identifies the edges of contrast defining the vessel borders to calculate vessel diameter in arbitrary units (aU), approximately corresponding to microns at the retinal plane. Additionally, the DVA allows the presentation of flickering light to the retina during the recording of vessel diameters (Bek et al. 2008). A low-pass filter with a cut-off limit at 550 nm is entered into the light path in order for the flickering light not to interfere with the imaging of the retina and is conveyed to the pathway that illuminates the retina. The light reflected to produce the image of the retina passes through an interference filter with a transmission maximum of 577 nm and a bandwidth of 10 nm. For the present examinations, the green light flicker frequency was set to 8 Hz, which has previously been shown to produce the most pronounced dilation of both arterioles and venules (Polak et al. 2002). Examinations were performed on the eye with best-corrected VA, and in case of equal VA on the two eyes, the left eye was examined. The patient was placed in front of the fundus camera and was asked to fixate on a bar, which was moved in the viewing system until the vessels of interest were within three disc diameters from the optic disc. Prior to the examination, the patient was allowed to rest for at least 5 min. An examination was continuous and consisted of seven phases, each with a duration of 120 seconds (Fig. 1). The experimental protocol used with the DVA. The initial resting phase (1) was used to determine the baseline blood pressure and diameters of retinal vessels, and the following resting phases (3,5,7) were used as reference for the interventions. During the lifting phases (2 & 6), the patient lifted a 2-kg hand weight by a straight right arm. The blood pressure was measured on the upper left arm during the last 30 seconds of each phase. The intraocular pressure has previously been shown not to change during exercise in the present experimental set-up in both normal persons (Jeppesen et al. 2004) and patients with diabetes (Jeppesen P unpublished data) and was therefore not measured during the examination. The mean arteriolar diameter (dmean) was calculated as the sum of all measured diameters divided by the number of measurements on the basis of the data obtained during the last 45 seconds of each examination phase. The mean arterial blood pressure (MAP) was calculated as the diastolic blood pressure (BPdia) plus 1/3 of the difference between the systolic and diastolic blood pressure (BPsys−BPdia). Central retinal thickness was calculated as the average (in μm) of the centre point of the six radial scans. Changes in the resting values for vessel diameters and blood pressures were tested using two-way anova with repeated measurements, with the following four factors: (i) the resting phases 1,3,5,7 during each examination, (ii) the treatment type (amlodipine and lisinopril), (iii) treatment period and (iv) the randomization group (A-L or L-A). Similarly, changes in the vessel diameter and blood pressures were tested using two-way anova with repeated measurements, with the following four factors: (i) the resting phase (1,3 or 5) before each examination, (ii) interventions 1 and 2 (exercise), (iii) interventions 3 and 4 (flicker) and (iv) inter-ventions 5 and 6 (exercise + flicker). Central retinal thickness and 24-hr ambulatory blood pressure were analysed similarly, however excluding the first factor as only one measurement was made at each examination. Probability plots of the residuals in the anova models verified that all data were normally distributed. A p-value of <0.05 was considered statistically significant. All values are represented as mean ± SD. The 24-hr ambulatory blood pressure at baseline was BPsys = 116.2 ± 2.9 mmHg and BPdia = 77.9 ± 2.1 mmHg. Antihypertensive treatment lowered the blood pressure non-significantly by lisinopril to BPsys = 114.1 ± 3.4 mmHg (p = 0.21) and BPdia = 76.7 ± 2.6 mmHg (p = 0.41) and by amlodipine to BPsys = 115.1 ± 2.9 mmHg (p = 0.45) and BPdia = 78.2 ± 2.1 mmHg (p = 0.81). There were no significant differences between the blood pressure lowering effects of lisinopril and amlodipine for either BPsys (p = 0.67) or BPdia (p = 0.45). During exercise, the blood pressure increased significantly by BPsys = 23.0 ± 2.6 mmHg and BPdia = 20.3 ± 1.7 mmHg (p = 0.001 for both comparisons). There was no significant difference between the blood pressure responses obtained during the two treatments (p = 0.21 for both). During combined exercise and flicker stimulation, the blood pressure increased significantly by BPsys = 26.9 ± 2.8 mmHg and BPdia = 21.7 ± 1.9 mmHg (p = 0.001 for both comparisons). There was no difference between the blood pressure responses obtained during the two treatments (p = 0.55 for both). The baseline diameter of the studied arterioles was 114.4 ± 3.5 aU and of the venules 147.8 ± 6.4 aU. There was no significant difference between the arteriolar diameter during the four resting phases (p = 0.21) regardless of treatment (p = 0.97) and the sequence of which it was administered (p = 0.48), whereas the venules remained dilated after phases 4 and 6 (p = 0.002 for both). The diameter changes during the interventions are shown in Table 2. During exercise, the mean arteriolar diameter decreased significantly by 2.4 ± 0.9% (p = 0.004), whereas there was no significant change in the venular diameter (p = 0.08). During flicker stimulation, the mean arteriolar diameter increased significantly by 2.2 ± 0.9% (p = 0.02), and the diameter of the venule incre-ased significantly by 7.4% ± 0.9% (p = 0.001). During combined stimulation with flicker and exercise, the mean arteriolar diameter increased significantly by 1.8 ± 0.9% (p = 0.03) and the mean venular diameter significantly by 4.4 ± 0.9% (p = 0.001). No significant differences were found in the diameter responses of retinal venules between the two treatments, except for a significantly reduced dilation during the combined flicker and exercise phase for patients treated with Lisionopril when compared to amlodipine (p = 0.021). However, the diameter of retinal venules did not return to baseline during the 120 seconds following the two flicker phases 5 and 7 (p = 0.02). The baseline mean central retinal thickness was 208.9 ± 6.2 microns, and neither of the treatments induced any change in this diameter (p = 0.47 for both). The present study has shown a lack of effect of short-term antihypertensive treatment with either lisinopril or amlodipine on retinal autoregulation as studied with the DVA in patients with type 1 diabetes mellitus and mild retinopathy. The antihypertensive treatment did not affect the blood pressure significantly in the studied normotensive patients which is similar to the effect obtained in other intervention studies on normotensive patients (Chaturvedi et al. 1998, 2008), indicating that effect of the treatment would be expected to be other than a change in the baseline blood pressure. During each examination, the increase in the blood pressure induced by exercise was similar to that achieved in previous studies (Frederiksen et al. 2006; Bek et al. 2008), indicating that the arteriolar smooth muscle cells had been sufficiently stretched by the blood pressure to respond with a contraction. The smallest diameter change that might be detected in this study including 25 patients can be calculated to approximately 1%, assuming risks of making a type I error of 5%, a type II error of 10% and assuming a measurement precision of 1%. Previous studies have shown that in younger non-diabetic persons, the diameter of the retinal arterioles is reduced by approximately 4% during exercise (Jeppesen et al. 2004) and increased by approximately 3.8% during flicker stimulation (Garhofer et al. 2004) This implies that the examined patients had severely impaired autoregulation and that a possible normalizing effect of the antihypertensive treatment on autoregulation would have been detected. An effect on retinal vascular diameters by the eye drops used to dilate the pupils cannot be excluded, but in preliminary experiments, we were unable to obtain video recordings of the fundus with a sufficient quality omitting one or both of the eye drops. Consequently, it was decided to use the paired design, which eliminates an influence of this factor on the conclusions regarding differences between the two treatment groups. In previous studies, retinal autoregulation has been found to be impaired in patients with diabetes as studied by several methods including laser Doppler velocimetry, blue field entoptoscopy, fluorescent dye retinal transit time and diameter measurement using the Retinal Vessel Analyzer (Dumskyj et al. 1996, Blum et al. 1999, Schmetterer & Wolzt 1999). Furthermore, the contraction of retinal arterioles after increasing the blood pressure by isometric exercise has been found to be reduced in persons older than 40 years of age, probably because of stiffness of the vessels because of arteriosclerosis (Jeppesen et al. 2004). The antihypertensive effect of amlodipine is attributed to the blocking of L-type calcium channels in vascular smooth muscle cells. The resulting decrease in the intracellular calcium concentration relaxes the tone of arterioles with a resulting reduction in the arterial blood pressure (Wang et al. 2008). In the present study, there was found no vasodilating effect of amlodipine on retinal arterioles. This may be because of lack of effect on the calcium transport in retinal vascular smooth muscle cells or a counter regulatory response that had masked such dilation. A further elucidation of the significance of such a counter regulatory response requires in vitro models where the influence of perivascular factors on the tone of retinal arterioles can be studied (Holmgaard et al. 2008). The effect of lisinopril is because of blocking of the ACE, which reduces the synthesis of the vasoconstrictive peptide angiotensin II (Patchett et al. 1980). The renin–angiotensin system has been shown to be involved in the development of diabetic microangiopathy in general (Itoh et al. 1993; Wilkinson-Berka 2006). However, evidence suggests the existence of a special renin–angiotensin system in the retina (Wagner et al. 1996; Murata et al. 1997; Otani et al. 1998), and a study has shown a normalization of retinal blood flow after treatment with the angiotensin II receptor blocker candesartan (Horio et al. 2004). The reduced dilatation of retinal venules during exercise combined with flicker stimulation in patients treated with lisinopril remains to be explained, but as the diameter of the arterioles were not affected, it can be concluded that the study could not demonstrate a significant effect of lisinopril on retinal autoregulation that might potentially affect retinal blood flow significantly. Consequently, the stabilizing effect on retinopathy, which is additional to the blood pressure lowering effect, might not be attributed to a restoration of normal retinal autoregulation. However, it cannot be excluded that significant effects might have been detected in patients with arterial hypertension where the medication had reduced the blood pressure substantially. Similarly, the treatment could have been administered for a longer period, and dosages of the used antihypertensive drugs might have been higher (Heran et al. 2008). However, with the therapeutic concentrations of the two antihypertensive drugs used on normotensive patients with mild retinopathy, no effect was observed on retinal autoregulation as measured with the DVA. It is generally assumed that blockers of the renin–angiotensin system have a protective effect on the development of diabetic retinopathy, which is additional to the antihypertensive effect. The present study has shown that this effect is not mediated through a significant normalization of retinal autoregulation in younger type 1 patients with diabetes and mild retinopathy. The intervention period was set to 14 days using a medium dose of the antihypertensive drugs, which is sufficient for the effect on the blood pressure to stabilize but may be insufficient for a possible effect on autoregulation to occur. This is in accordance with the experience from the DIRECT study where a longer follow-up period was needed to find a borderline protective effect of the angiotensin II receptor blocker candesartan on the development of diabetic retinopathy (Chaturvedi et al. 2008). A possible effect of antihypertensive treatment on retinal autoregulation in intervention studies of longer duration would suggest that the effects on the blood pressure and autoregulation might be mediated through separate mechanisms. This possible dual effect of blockers of the renin–angiotensin system requires further elucidation. The present study was supported by the VELUX Foundation. The skilful assistance of technician Merete Møller is gratefully acknowledged.