- Email: [email protected]
Semi-automatic measuring of arteriovenous relation as a possible silent brain infarction risk index in hypertensive patients
a r c h s o c e s p o f t a l m o l . 2 0 1 6;9 1(1 1):513–519
ARCHIVOS DE LA SOCIEDAD ESPAÑOLA DE OFTALMOLOGÍA www.elsevier.es/oftalmologia
Original article
Semi-automatic measuring of arteriovenous relation as a possible silent brain infarction risk index in hypertensive patients夽,夽夽 X.M. Vázquez Dorrego a,∗ , J.M. Manresa Domínguez b,c , A. Heras Tebar b,d , R. Forés b,d , A. Girona Marcé e , M.T. Alzamora Sas b,d , P. Delgado Martínez f,g , I. Riba-Llena f,g , J. Ugarte Anduaga h , A. Beristain Iraola i , I. Barandiaran Martirena i , S.M. Ruiz Bilbao j , P. Torán Monserrat b,k a
Secció d’Oftalmologia, Hospital Municipal de Badalona, Badalona Serveis Assistencials, Badalona (Barcelona), Spain Unitat de Suport a la Recerca Metropolitana Nord, IDIAP Jordi Gol, Santa Coloma de Gramenet, Barcelona, Spain c Departament d’Infermeria, Universitat Autònoma de Barcelona, Cerdanyola del Vallès (Barcelona), Spain d ABS Riu Nord-Riu Sud, Institut Català de la Salut, Santa Coloma de Gramenet (Barcelona), Spain e Servei de Urgències, Hospital de Barcelona, Barcelona, Spain f Laboratori de Malalties Neurovasculars, Institut de Recerca Vall d’Hebron, Institut Català de la Salut, Barcelona, Spain g Universitat Autònoma de Barcelona (UAB), Barcelona, Spain h Promoción e I+D+I, ULMA Innovación SL, Onati, ˜ Gipúzcoa, Spain i Vicomtech-IK4 Visual Interaction Communication Technologies, San Sebastián, Guipúzcoa, Spain j Hospital Universitari Germans Trias i Pujol, Badalona (Barcelona), Spain k Departament de Medicina, Universitat de Girona, Gerona, Spain b
a r t i c l e
i n f o
a b s t r a c t
Article history:
Objective: To evaluate the usefulness of a semiautomatic measuring system of arteriove-
Received 9 February 2016
nous relation (RAV) from retinographic images of hypertensive patients in assessing their
Accepted 3 May 2016
cardiovascular risk and silent brain ischemia (ICS) detection.
Available online 17 August 2016
Methods: Semi-automatic measurement of arterial and venous width were performed with
Keywords:
images belonging to the 976 patients integrated in the cohort Investigating Silent Strokes in
Eye fundus
Hypertensives: a magnetic resonance imaging study (ISSYS), group of hypertensive patients.
the aid of Imedos software and conventional fundus examination from the analysis of retinal
Retinopathy
All patients have been subjected to a cranial magnetic resonance imaging (RMN) to assess
Arteriole to venule ratio
the presence or absence of brain silent infarct.
夽 Please cite this article as: Vázquez Dorrego XM, Manresa Domínguez JM, Heras Tebar A, Forés R, Girona Marcé A, Alzamora Sas MT, et al. Medición semiautomática de la relación arteriovenosa retiniana como posible marcador de riesgo de infarto cerebral silente en pacientes hipertensos. Arch Soc Esp Oftalmol. 2016;91:513–519. 夽夽 Other results derived from this project have also been presented at: XVIII Jornada de Recerca Sanitària al Maresme. Mataró (Spain), ˜ June 18, 2015; 35 Congreso de la Sociedad Espanola de Medicina Familiar y Comunitaria. Gijón (Spain), June 11–13, 2015; 20th WONCA Europe Conference. Istanbul (Turkey), October 22–25, 2015. ∗ Corresponding author. E-mail addresses: [email protected], [email protected] (X.M. Vázquez Dorrego). ˜ ˜ S.L.U. All rights reserved. 2173-5794/© 2016 Sociedad Espanola de Oftalmolog´ıa. Published by Elsevier Espana,
514
a r c h s o c e s p o f t a l m o l . 2 0 1 6;9 1(1 1):513–519
Arterial hypertension
Results: Retinal images of 768 patients were studied. Among the clinical findings observed,
Retinography
association with ICS was only detected in patients with microaneurysms (OR 2.50; 95%
Silent brain infarct
CI: 1.05–5.98) or altered RAV (<0.666) (OR: 4.22; 95% CI: 2.56–6.96). In multivariate logistic
Stroke
regression analysis adjusted by age and sex, only altered RAV continued demonstrating as
Semi-automatic retinal vessel
a risk factor (OR: 3.70; 95% CI: 2.21–6.18).
quantification
Conclusions: The results show that the semiautomatic analysis of the retinal vasculature from retinal images has the potential to be considered as an important vascular risk factor in hypertensive population. ˜ ˜ S.L.U. All rights © 2016 Sociedad Espanola de Oftalmolog´ıa. Published by Elsevier Espana, reserved.
Medición semiautomática de la relación arteriovenosa retiniana como posible marcador de riesgo de infarto cerebral silente en pacientes hipertensos r e s u m e n Palabras clave:
Propósito: Evaluar la utilidad de un sistema semiautomático de medición de relación arte-
Fondo de ojo
riovenosa (RAV) retiniana sobre imágenes retinográficas de pacientes hipertensos en la
Retinopatía
valoración del riesgo cardiovascular y la detección de isquemia cerebral silente (ICS).
Ratio arteria vena
Método: Un total de 976 pacientes de la cohorte Investigating Silent Strokes in Hypertensives:
Hipertensión arterial
a magnetic resonance imaging study (ISSYS) estudiados mediante resonancia magnética
Retinografía
craneal para valorar la presencia o no de ICS fueron invitados a realizar una retinografía para
Isquemia cerebral silente
un examen convencional de fondo de ojo y una medición semitautomática del promedio de
Ictus
los calibers vasculares para el cálculo de la relación arteriovenosa (RAV).
Medición semiautomática de la
Resultados: Se analizaron las retinografías de 768 pacientes. Entre las lesiones observadas,
microcirculación retiniana
solamente se encontró una asociación con la detección de ICS en aquellos pacientes con microaneurismas (OR: 2,50; IC 95%: 1,05–5,98) o una RAV alterada (<0,666) (OR: 4,22; IC 95%: 2,56–6,96). En el análisis de regresión logística multivariante ajustado por edad y sexo, solamente la RAV alterada continuó manifestándose como un factor de riesgo (OR: 3,70; IC 95%: 2,21-6,18). Conclusiones: Los resultados muestran que el análisis semiautomático de la vasculatura retiniana a partir de retinografías tiene el potencial de ser considerado como un factor de riesgo vascular importante en la población hipertensa. ˜ ˜ S.L.U. Todos de Oftalmolog´ıa. Publicado por Elsevier Espana, © 2016 Sociedad Espanola los derechos reservados.
Introduction The existence of ischemic events in the central nervous system without apparent clinic repercussion has been known under different denominations for a number of decades1 up to the most recent denomination of silent cerebral ischemia (SCI). However, the advent of modern imaging diagnostic methods, mainly nuclear magnetic resonance (NMR), have enabled in-depth studies of SCI characteristics and incidence.2 A recent systematic review by Fanning et al.3 provided epidemiological data that evidence a prevalence in the general population between 5% and 62%, although the majority of reviewed studies report a prevalence between 10 and 20%. Longitudinal studies report annual incidence rates between 2% and 4%. Factors exhibiting closer association with SCI are age, arterial hypertension (AHT), carotid stenosis, chronic kidney disease and metabolic syndrome.3
The study of the clinic repercussion of AHT on retinal circulation has been the object of numerous studies since Keith and Wagener published their classification.4 However, in contrast with diabetic retinopathy, the usefulness of systematically studying changes in retinal circulation of patients with AHT is controversial.5 This accounts for the absence of an explicit recommendation studying eye fundus in all hypertensive patients. The guide of the European cardiology and hypertension societies (ESH/ESC) takes into account retinal examination of patients with resistant or difficult to control AHT for detecting advanced retinal lesions, but discards the systematic study of the entire hypertense population due to variability in observations and poor reproducibility.6,7 Some authors have proposed an initial retinographic assessment of all hypertensive patients as this would enable the discovery of some advanced retinal lesions.8 Recent studies confirm the importance of asymptomatic cerebrovascular disease9,10 as well as the main associated risk factors and the need of finding useful markers for the early detection of this condition in order
a r c h s o c e s p o f t a l m o l . 2 0 1 6;9 1(1 1):513–519
515
to design diagnostic strategies applicable to broad population groups.11 On the other hand, well-known research is being carried out associating retinal microcirculation alterations with increased stroke and cardiac disease risk,12–17 to the extent of indicating associations with greater cognitive decline risk.18,19 Several analyses carried out by the Atherosclerosis risk in communities study related retinal microvascular alterations with subclinical or asymptomatic lesions found in NMR.20,21 However, said study does not include some microvasculature parameters that can be observed with the aid of specifically designed computer applications. Other large epidemiological studies reported similar findings in several sub-populations and ethnic groups.22–24 The objective of the present study is to analyze the value of eye fundus study with non-mydriatic retinography assessed with clinical criteria, supplemented by the semiautomatic determination of the arteriole-venule ratio applying a specific software, for detecting silent cerebral ischemia in hypertensive patients.
Subjects, material and methods Participants belonged to the cohort known as Investigating Silent Strokes in Hypertensives: a magnetic resonance imaging study (ISSYS), comprised by a randomized selection of 1037 hypertensive patients with ages between 50 and 70, without previous history of stroke or dementia, who underwent a cerebral NMR study with the aim of detecting cerebral infarcts and other asymptomatic brain injuries. In addition, various clinic and analytic assessments were made in order to determine the possible repercussion in target organs.10 The study protocol has been published.25 After obtaining informed consent, 976 of the 1037 patients of the cohort (those who had informed magnetic resonance) were invited to participate in the present study, which consisted in obtaining a retinography of both eyes centered on the optic disk/macula area at 45◦ without previous pharmacologically induced midriasis and in comparable lighting conditions carried out by 2 specifically trained technicians. The Topcon TRC-NW65 retinograph (Topcon SA, Barcelona, Spain) was utilized for obtaining the digital photographs. Said device included the following supplements: Nikon D7000 camera for Topcon TRC-NW6 retinograph, Nikon Corporation, Tokyo, Japan; Ibase DigiCaptura (Topcon SA, Barcelona, Spain); 19 TFT screen; ordenador computer with Windows 7 OS (Microsoft Corporation, Redmond WA, United States) type I, ATE-600 and Eye-Viewer software (Topcon SA, Barcelona, Spain) for visualizing and measuring retinal vasculature images. The images were stored in JPG format in a Microsoft Access database in which the data were completely anonymized without any clinical patient data. The retinographies were assessed by an ophthalmologist experienced in retinal pathology. All ophthalmoscopic findings were studied and adequately documented, particularly findings that determined the Keith and Wagener classification, i.e. arteriole-venule ratio (AVR), focal arteriole narrowing, presence of pathological arterial venous crossings,
Fig. 1 – Image showing the selection process of arteriole and venule vessels for obtaining the average of vascular diameters for calculating the arteriovenous ratio using the computer application. Source: http://www.imedos.de/Vesselmap2.98.0.html?&L=1.
hemorrhages or microaneurysms, hard exudates, cotton-like nodules and papilla edema. As it was established, supported by evidence, that the simple observation of the arteriole and vein diameter was very difficult to interpret due to observer subjectivity, the ophthalmoscopic AVR was substituted by that obtained with the semiautomatic measurement of arterial and venous inside diameters with available market software. After locating the optic nerve and establishing the region of interest, said software enabled the manual location of arterioles and venules (Fig. 1) in order to subsequently provide automatically the measurements of both vessels and calculation of the AVR. According with existing references, it was decided to consider AVR under 0.66 as pathological.26,27
Statistical analysis Quantitative variables are described with mean and standard deviation, while qualitative variables with absolute frequency and percentage. Quantitative variables were compared utilizing the T for student test for independent data, Pearson’s Chi square or Fishers exact test for qualitative variables, depending on the application conditions. Logistic regression analysis was performed to study the role of different independent variables as SCI risk factors. Bivariate analysis was performed first, obtaining the odds ratio (OR) for each independent variable as well as its 95% confidence interval (CI 95%). The variables having a Wald statistical equivalent to a significance of p ≤ 0.100 were initially included in a complete multivariate model adjusted for age and sex, eliminating those exhibiting a significance value of p > 0.050 and did not modify the ˇ coefficient values over 20%.
516
a r c h s o c e s p o f t a l m o l . 2 0 1 6;9 1(1 1):513–519
Table 1 – Participant characteristics summary based on gender.
Age Silent cerebral infarct n (%)
Males (n = 385)
Females (n = 383)
Total (n = 768)
62.6 (5.6) 59 (15.3)
62.9 (5.4) 22 (5.7)
62.7 (5.5) 81 (10.5)
Keith–Wagener criteria n (%) Altered AVRa ,b Crossings Microaneurysms Flame hemorrhage Soft exudate Hard exudate Papillary edema
74 (19.2) 278 (72.2) 19 (4.9) 7 (1.8) 5 (1.3) 7 (1.8) 0
45 (11.7) 267 (69.9) 13 (3.4) 1 (0.3) 2 (0.5) 2 (0.5) 0
119 (15.5) 545 (71.1) 32 (4.2) 8 (1.0) 7 (0.9) 9 (1.2) 0
Keith-Wagener Classificationb n (%) Without disease Grade I Grade II Grade III Grade IV
84 (21.8) 21 (5.5) 258 (67.0) 22 (5.7) 0
99 (25.8) 13 (3.4) 257 (67.1) 14 (3.7) 0
183 (23.8) 34 (4.4) 515 (67.1) 36 (4.7) 0
a b
p 0.407 <0.001
0.004 0.480 0.289 0.069 0.259 0.177 – 0.180
The arteriovenous ratio is considered altered or pathological when exhibiting a value <0.666. Arteriovenous ratio obtained by reading with the Imedos software.
Results Out of the 970 patients with informed NMR, retinal graphic images were obtained in 775. Seven cases could not be assessed due to poor quality image. Excluded patients were not included due to the inability to establish contact with them or who declined participation. Finally, retinographies of 768 patients were analyzed, diagnosing the presence of CSI by means of NMR. The mean age was 62.7 (±5.5) years, 50.1% being males. Table 1 shows the baseline characteristics of the studied population. Observed SCI prevalence was 10.5%, higher in males (15.3 versus 5.7%; p < 0.001). Most frequently observed lesions were pathological AVR (15.5%) and the presence of crossings (71.1%). Prevalence of males was higher only in AVR (19.2 versus 11.7%; p = 0.004). No eye fundus anomalies were found in 23.8% of patients. The logistic regression analysis for exploring candidate variables for SCI risk factors demonstrated that males and age over 60 increased OR significantly. Patients with Keith–Wagener classification grades i and iii exhibited greater risk (OR = 3.39 and 3.74 respectively; p < 0.005). In this classification, the significant criteria were altered AVR (p < 0.001) and the presence of microaneurysms (p = 0.039), as shown in Table 2. Subsequently, multivariate logistic regression analysis was carried out containing initially the Keith–Wagener classification or some of its criteria (AVR, microaneurysm, flame hemorrhage and hard exudate), in all cases adjusting for sex and age above 60. In the subsequent model, only the presence of altered AVR performed as risk factor with OR = 3.70 (2.21–6.18) (Table 3).
Discussion The results of the present study demonstrate that the alteration of retinal microvasculature is associated to an increased
risk of SCI in a range between 3.4 times greater risk for patients with Keith–Wagener grade I, and 3.7 times greater risk for patients with retinal involvement grade III. The alterations which, in isolation exhibit a statistically significant association with silent cerebral ischemia are diminished AVR under 0.66 assessed by means of semiautomatic software OR = 4.22 (2.56–6.96) and observation of microaneurysms OR = 2.50 (1.05–5.98), although the latter disappears when adjusting a multivariate model (Table 3). The baseline characteristics of patients who underwent retinography do not differ from those of the parent study population.10 Being a randomly obtained population-based study, it can be regarded as representative of the general hypertense population between 50 and 70 years of age in our country. Both populations (with and without retinography) reproduce SCI prevalence in the area of 10% and are located in the lower range of prevalences found in other large population-based epidemiological studies.3 However, said SCI prevalence is lower to that described in other hypertense cohorts. The difference could be attributed to specific sociodemographic characteristics (younger age) and with better arterial blood pressure control in the present cohort.10 Said differential characteristics should also be considered when assessing the repercussion of AHT on the retina of the present study population. In what concerns the prevalence of retinal lesions, it is noteworthy that the present study produced a high number of detected arteriovenous crossings (71% of retinographies), whereas similar studies in hypertense populations12 detected said alteration in 17.5% of cases and, in the general population,22 in 7.5%. For the rest of lesions, this discrepancy was not found when comparing the results of the present study with those of other series. The authors assume that this lesion could be overdiagnosed in the present series due to the fact that the observer was aware that the patients were hypertensive, whereas in the referenced studies retinography readings kept the observer masked in relation to the clinic condition of patients.
517
a r c h s o c e s p o f t a l m o l . 2 0 1 6;9 1(1 1):513–519
Table 2 – Bivariate logistic regression analysis of all candidates to be considered as SCI risk factors. SCI No. = 81
Beta
OR (CI 95%)a
p (Wald)
Sex n (%) Female Male
22 (5.7) 59 (15.3)
1.088
2.97 (1.78–4.96)
Age n (%) 50–55 56–60 61–65 > 65
3 (3.0) 14 (8.6) 24 (10.8) 40 (14.1)
– 1.118 1.361 1.672
– 3.06 (0.86–10.9) 3.90 (1.146–13.3) 5.32 (1.61–17.6)
– 0.085 0.029 0.006
Keith-Wagener Classificationc n (%) Without pathology Grade I Grade II Grade III Grade IV
13 (7.1) 7 (20.6) 53 (10.3) 8 (22.2) 0
– 1.121 0.406 1.318 –
– 3.39 (1.24–9.53) 1.50 (0.80–2.82) 3.74 (1.42–9.83) –
– 0.017 0.208 0.008 –
50 (7.7) 31 (26.1)
1.440
4.22 (2.56–6.96)
<0.001
22 (9.9) 59 (10.8)
0.099
1.10 (0.66–1.85)
0.708
74 (10.1) 7 (21.9)
0.917
2.50 (1.05–5.98)
0.039
79 (10.4) 2 (25.0)
1.504
2.87 (0.57–14.5)
0.201
80 (10.5) 1 (14.3)
0.348
1.42 (0.17–11.9)
0.749
79 (10.4) 2 (22.2) 0
0.898 –
2.46 (0.50–12.0) –
0.268 –
Keith-Wagener criteriac n (%) Altered AVRb No Yes Crossings No Yes Microaneurysms No Yes Flame hemorrhage No Yes Soft exudate No Yes Hard exudate No Yes Papillary edema
<0.001
CI 95%, confidence interval 95%. a b c
Dependent variable: silent cerebral infarct (diagnosed with magnetic resonance). Arteriovenous ratio is considered altered or pathological when exhibiting a value <0.666. Arteriovenous ratio obtained by reading processed by VesselMap2 software version 3.10 (Imedos).
Table 3 – Model obtained from the multivariate logistic regression analysis for the risk of presenting silent cerebral infarct. ˇ coefficient Constant Altered arteriovenous ratiob, c Male sex Age >60 years
OR (CI 95%)a
p (Wald)
−3.611 1.308
3.70 (2.21–6.18)
<0.001
1.031 0.743
2.80 (1.66–4.74) 2.10 (1.19–3.73)
<0.001 0.011
CI 95%, confidence interval 95%. a
b
c
Dependent variable: silent cerebral infarct (diagnosed with magnetic resonance). Arteriovenous ratio is considered altered or pathological when exhibiting a value <0.666. Arteriovenous ratio obtained by reading processed by VesselMap2 software version 3.10 (Imedos).
The results match those described by Kwon et al. in a group of 550 Korean hypertensive patients without cerebral ischemia antecedents, with risk levels similar to those found in the present study.26 Other transversal studies exhibit the same association between retinopathy and subclinic cerebral lesions.22,28 With the present data, hemorrhages (n = 2), soft exudates (n = 1) and hard exudates (n = 2) also appear as SCI risk factors without reaching statistical significance, which is attributed to the small number of cases. The case of arteriovenous crossings is also noteworthy because, in the present case, do not exhibit either association with SCI although said association is described in the literature,20–23 even in longitudinal studies which describe said crossings as a predictor of changes in white substance lesions. The authors believe that this could be due to the overdiagnosis of arteriovenous crossings in the present series, which has the effect of diluting the statistical results.
518
a r c h s o c e s p o f t a l m o l . 2 0 1 6;9 1(1 1):513–519
The article by Ikram et al.23 on a general population over 55 years of age describes the association between vein enlargement and the progression of microvascular lesions at the cerebral level. However, this association is not maintained when utilizing AVR as a parameter for measuring retinal circulation. In the present case, with currently available reading data as provided by the software, it is not possible to analyze separately the arterial and small vein diameters to determine which AVR component has greater influence in the variation thereof. As conclusion, the authors believe that despite the limitations described above, the results demonstrate that semiautomatic analysis of retinal vasculature on the basis of retinographies has the potential of being considered as a significant vascular risk factor in hypertensive populations.
Funding The present project has been financed by the Basque Government and the Regional Development European Fund of the European Union under the following programs: 1) NETs 2012–2014 (support for projects launching scientific and technology enterprises): IN-2012/0000061. IN2013/0000040. IN-2014/0000124. 2) Ekintzaile (00321/2013)/Txekintek (0154/2013). The project has also received a funding provided by ULMA Innovation SL.
Conflict of interests Co-author Jurgi Ugarte is linked to the technical staff of ULMA Innovation SL as a third-party subcontractor. ULMA Innovation is developing projects on image-based noninvasive diagnostics. Retinal images are one of its fields of research and development. Jurgi Ugarte has no other type of engagement or association, joint property or shareholdings in ULMA Innovation SL. The rest of co-authors have not declared any conflict of interests.
references
1. Fisher CM. Lacunes: small, deep cerebral infarcts. Neurology. 1965;15:774–84. 2. Norrving B. Evolving concept of small vessel disease through advanced brain imaging. J Stroke. 2015;17:94. 3. Fanning JP, Wong AA, Fraser JF. The epidemiology of silent brain infarction: a systematic review of population-based cohorts. BMC Med. 2014;12:119. 4. Keith N, Wagener H, Barker N. Some different types of essential hypertension: their course and their prognosis. Am J Med Sci. 1939:332–43. 5. Van den Born B-JH, Hulsman CA, Hoekstra JB, Schlingemann RO, van Montfrans GA. Value of routine fondocopy in patients with hypertension: systematic review. BMJ. 2005; 331:73. 6. Mancia G, Fagard R, Narkiewicz K, Redon J, Zanchetti A, Böhm M, et al. Guía de práctica clínica de la ESH/ESC para el manejo
7.
8.
9.
10.
11.
12.
13.
14.
15.
16. 17.
18.
19.
20.
21.
22.
23.
de la hipertensión arterial (2013). Rev Esp Cardiol. 2013;66:880.e1–64. Mancia G, Fagard R, Narkiewicz K, Redon J, Zanchetti A, Böhm M, et al. 2013 ESH/ESC guidelines for the management of arterial hypertension: the Task Force for the management of arterial hypertension of the European Society of Hypertension (ESH) and of the European Society of Cardiology (ESC). Eur Heart J. 2013;34:2159–219. Coll-de-Tuero G, Rodríguez Poncelas A, Faixedas Brunsoms D, Cadena Calderó A. Aplicación práctica de la retinografía en los hipertensos. Hipertens Riesgo Vasc. 2007;24:61–9. ˜ A, López-Cancio E, Dorado L, Millán M, Reverté S, Sunol Massuet A, et al. The Barcelona-Asymptomatic Intracranial Atherosclerosis (AsIA) study: prevalence and risk factors. Atherosclerosis. 2012;221:221–5. Delgado P, Riba-Llena I, Tovar JL, Jarca CI, Mundet X, López-Rueda A, et al. Prevalence and associated factors of silent brain infarcts in a mediterranean cohort of hypertensives. Hypertension. 2014. Jiménez M, Dorado L, Hernández-Pérez M, Alzamora MT, Pera G, Torán P, et al. Ankle-brachial index in screening for asymptomatic carotid and intracranial atherosclerosis. Atherosclerosis. 2014;233:72–5. Ong Y-TT, Wong TY, Klein R, Klein BE, Mitchell P, Sharrett AR, et al. Hypertensive retinopathy and risk of stroke. Hypertension. 2013;62:706–11. Yatsuya H, Folsom AR, Wong TY, Klein R, Klein BE, Sharrett AR. Retinal microvascular abnormalities and risk of lacunar stroke: atherosclerosis risk in communities study. Stroke. 2010;41:1349–55. McGeechan K, Liew G, Macaskill P, Irwig L, Klein R, Klein BE, et al. Prediction of incident stroke events based on retinal vessel caliber: a systematic review and individual-participant meta-analysis. Am J Epidemiol. 2009;170:1323–32. Liew G, Wong TY, Mitchell P, Cheung N, Wang JJ. Retinopathy predicts coronary heart disease mortality. Heart. 2009;95:391–4. McGeechan K. Meta-analysis. Retinal vessel caliber and risk for coronary heart disease. Ann Intern Med. 2009;151:404. Cheung N, Bluemke DA, Klein R, Sharrett AR, Islam FM, Cotch MF, et al. Retinal arteriolar narrowing and left ventricular remodeling: the multi-ethnic study of atherosclerosis. J Am Coll Cardiol. 2007;50:48–55. Ding J, Patton N, Deary IJ, Strachan MW, Fowkes FG, Mitchell RJ, et al. Retinal microvascular abnormalities and cognitive dysfunction: a systematic review. Br J Ophthalmol. 2008;92:1017–25. Gatto NM, Varma R, Torres M, Wong TY, Johnson PL, Segal-Gidan F, et al. Retinal microvascular abnormalities and cognitive function in Latino adults in Los Angeles. Ophthalmic Epidemiol. 2012;19:127–36. Cheung N, Mosley T, Islam A, Kawasaki R, Sharrett AR, Klein R, et al. Retinal microvascular abnormalities and subclinical magnetic resonance imaging brain infarct: a prospective study. Brain. 2010;133:1987–93. Hanff TC, Sharrett AR, Mosley TH, Shibata D, Knopman DS, Klein R, et al. Retinal microvascular abnormalities predict progression of brain microvascular disease: an atherosclerosis risk in communities magnetic resonance imaging study. Stroke. 2014;45:1012–7. Longstreth W, Larsen EK, Klein R, Wong TY, Sharrett AR, Lefkowitz D, et al. Associations between findings on cranial magnetic resonance imaging and retinal photography in the elderly: the cardiovascular health study. Am J Epidemiol. 2007;165:78–84. Ikram MK, de Jong FJ, van Dijk EJ, Prins ND, Hofman A, Breteler MM, et al. Retinal vessel diameters and cerebral small vessel disease: the Rotterdam Scan study. Brain. 2006;129: 182–8.
a r c h s o c e s p o f t a l m o l . 2 0 1 6;9 1(1 1):513–519
24. Cheung CY, Tay WT, Ikram MK, Ong YT, De Silva DA, Chow KY, et al. Retinal microvascular changes and risk of stroke: the Singapore Malay eye study. Stroke. 2013;44:2402–8. 25. Riba-Llena I, Jarca CI, Mundet X, Tovar JL, Orfila F, López-Rueda A, et al. Investigating silent strokes in hypertensives: a magnetic resonance imaging study (ISSYS): rationale and protocol design. BMC Neurol. 2013;13: 130. 26. Kwon HM, Kim BJ, Oh JY, Kim SJ, Lee SH, Oh BH, Yoon BW. Retinopathy as an indicator of silent brain infarction in
519
asymptomatic hypertensive subjects. J Neurol Sci. 2007;252:159–62. 27. Fischer MD, Huber G, Feng Y, Tanimoto N, Mühlfriedel R, Beck SC, et al. In vivo assessment of retinal vascular wall dimensions. Invest Opthalmol Vis Sci. 2010;51:5254. 28. Cooper LS, Wong TY, Klein R, Sharrett AR, Bryan RN, Hubbard LD, et al. Retinal microvascular abnormalities and MRI-defined subclinical cerebral infarction: the atherosclerosis risk in communities study. Stroke. 2006;37:82–6.
ARCHIVOS DE LA SOCIEDAD ESPAÑOLA DE OFTALMOLOGÍA www.elsevier.es/oftalmologia
Original article
Semi-automatic measuring of arteriovenous relation as a possible silent brain infarction risk index in hypertensive patients夽,夽夽 X.M. Vázquez Dorrego a,∗ , J.M. Manresa Domínguez b,c , A. Heras Tebar b,d , R. Forés b,d , A. Girona Marcé e , M.T. Alzamora Sas b,d , P. Delgado Martínez f,g , I. Riba-Llena f,g , J. Ugarte Anduaga h , A. Beristain Iraola i , I. Barandiaran Martirena i , S.M. Ruiz Bilbao j , P. Torán Monserrat b,k a
Secció d’Oftalmologia, Hospital Municipal de Badalona, Badalona Serveis Assistencials, Badalona (Barcelona), Spain Unitat de Suport a la Recerca Metropolitana Nord, IDIAP Jordi Gol, Santa Coloma de Gramenet, Barcelona, Spain c Departament d’Infermeria, Universitat Autònoma de Barcelona, Cerdanyola del Vallès (Barcelona), Spain d ABS Riu Nord-Riu Sud, Institut Català de la Salut, Santa Coloma de Gramenet (Barcelona), Spain e Servei de Urgències, Hospital de Barcelona, Barcelona, Spain f Laboratori de Malalties Neurovasculars, Institut de Recerca Vall d’Hebron, Institut Català de la Salut, Barcelona, Spain g Universitat Autònoma de Barcelona (UAB), Barcelona, Spain h Promoción e I+D+I, ULMA Innovación SL, Onati, ˜ Gipúzcoa, Spain i Vicomtech-IK4 Visual Interaction Communication Technologies, San Sebastián, Guipúzcoa, Spain j Hospital Universitari Germans Trias i Pujol, Badalona (Barcelona), Spain k Departament de Medicina, Universitat de Girona, Gerona, Spain b
a r t i c l e
i n f o
a b s t r a c t
Article history:
Objective: To evaluate the usefulness of a semiautomatic measuring system of arteriove-
Received 9 February 2016
nous relation (RAV) from retinographic images of hypertensive patients in assessing their
Accepted 3 May 2016
cardiovascular risk and silent brain ischemia (ICS) detection.
Available online 17 August 2016
Methods: Semi-automatic measurement of arterial and venous width were performed with
Keywords:
images belonging to the 976 patients integrated in the cohort Investigating Silent Strokes in
Eye fundus
Hypertensives: a magnetic resonance imaging study (ISSYS), group of hypertensive patients.
the aid of Imedos software and conventional fundus examination from the analysis of retinal
Retinopathy
All patients have been subjected to a cranial magnetic resonance imaging (RMN) to assess
Arteriole to venule ratio
the presence or absence of brain silent infarct.
夽 Please cite this article as: Vázquez Dorrego XM, Manresa Domínguez JM, Heras Tebar A, Forés R, Girona Marcé A, Alzamora Sas MT, et al. Medición semiautomática de la relación arteriovenosa retiniana como posible marcador de riesgo de infarto cerebral silente en pacientes hipertensos. Arch Soc Esp Oftalmol. 2016;91:513–519. 夽夽 Other results derived from this project have also been presented at: XVIII Jornada de Recerca Sanitària al Maresme. Mataró (Spain), ˜ June 18, 2015; 35 Congreso de la Sociedad Espanola de Medicina Familiar y Comunitaria. Gijón (Spain), June 11–13, 2015; 20th WONCA Europe Conference. Istanbul (Turkey), October 22–25, 2015. ∗ Corresponding author. E-mail addresses: [email protected], [email protected] (X.M. Vázquez Dorrego). ˜ ˜ S.L.U. All rights reserved. 2173-5794/© 2016 Sociedad Espanola de Oftalmolog´ıa. Published by Elsevier Espana,
514
a r c h s o c e s p o f t a l m o l . 2 0 1 6;9 1(1 1):513–519
Arterial hypertension
Results: Retinal images of 768 patients were studied. Among the clinical findings observed,
Retinography
association with ICS was only detected in patients with microaneurysms (OR 2.50; 95%
Silent brain infarct
CI: 1.05–5.98) or altered RAV (<0.666) (OR: 4.22; 95% CI: 2.56–6.96). In multivariate logistic
Stroke
regression analysis adjusted by age and sex, only altered RAV continued demonstrating as
Semi-automatic retinal vessel
a risk factor (OR: 3.70; 95% CI: 2.21–6.18).
quantification
Conclusions: The results show that the semiautomatic analysis of the retinal vasculature from retinal images has the potential to be considered as an important vascular risk factor in hypertensive population. ˜ ˜ S.L.U. All rights © 2016 Sociedad Espanola de Oftalmolog´ıa. Published by Elsevier Espana, reserved.
Medición semiautomática de la relación arteriovenosa retiniana como posible marcador de riesgo de infarto cerebral silente en pacientes hipertensos r e s u m e n Palabras clave:
Propósito: Evaluar la utilidad de un sistema semiautomático de medición de relación arte-
Fondo de ojo
riovenosa (RAV) retiniana sobre imágenes retinográficas de pacientes hipertensos en la
Retinopatía
valoración del riesgo cardiovascular y la detección de isquemia cerebral silente (ICS).
Ratio arteria vena
Método: Un total de 976 pacientes de la cohorte Investigating Silent Strokes in Hypertensives:
Hipertensión arterial
a magnetic resonance imaging study (ISSYS) estudiados mediante resonancia magnética
Retinografía
craneal para valorar la presencia o no de ICS fueron invitados a realizar una retinografía para
Isquemia cerebral silente
un examen convencional de fondo de ojo y una medición semitautomática del promedio de
Ictus
los calibers vasculares para el cálculo de la relación arteriovenosa (RAV).
Medición semiautomática de la
Resultados: Se analizaron las retinografías de 768 pacientes. Entre las lesiones observadas,
microcirculación retiniana
solamente se encontró una asociación con la detección de ICS en aquellos pacientes con microaneurismas (OR: 2,50; IC 95%: 1,05–5,98) o una RAV alterada (<0,666) (OR: 4,22; IC 95%: 2,56–6,96). En el análisis de regresión logística multivariante ajustado por edad y sexo, solamente la RAV alterada continuó manifestándose como un factor de riesgo (OR: 3,70; IC 95%: 2,21-6,18). Conclusiones: Los resultados muestran que el análisis semiautomático de la vasculatura retiniana a partir de retinografías tiene el potencial de ser considerado como un factor de riesgo vascular importante en la población hipertensa. ˜ ˜ S.L.U. Todos de Oftalmolog´ıa. Publicado por Elsevier Espana, © 2016 Sociedad Espanola los derechos reservados.
Introduction The existence of ischemic events in the central nervous system without apparent clinic repercussion has been known under different denominations for a number of decades1 up to the most recent denomination of silent cerebral ischemia (SCI). However, the advent of modern imaging diagnostic methods, mainly nuclear magnetic resonance (NMR), have enabled in-depth studies of SCI characteristics and incidence.2 A recent systematic review by Fanning et al.3 provided epidemiological data that evidence a prevalence in the general population between 5% and 62%, although the majority of reviewed studies report a prevalence between 10 and 20%. Longitudinal studies report annual incidence rates between 2% and 4%. Factors exhibiting closer association with SCI are age, arterial hypertension (AHT), carotid stenosis, chronic kidney disease and metabolic syndrome.3
The study of the clinic repercussion of AHT on retinal circulation has been the object of numerous studies since Keith and Wagener published their classification.4 However, in contrast with diabetic retinopathy, the usefulness of systematically studying changes in retinal circulation of patients with AHT is controversial.5 This accounts for the absence of an explicit recommendation studying eye fundus in all hypertensive patients. The guide of the European cardiology and hypertension societies (ESH/ESC) takes into account retinal examination of patients with resistant or difficult to control AHT for detecting advanced retinal lesions, but discards the systematic study of the entire hypertense population due to variability in observations and poor reproducibility.6,7 Some authors have proposed an initial retinographic assessment of all hypertensive patients as this would enable the discovery of some advanced retinal lesions.8 Recent studies confirm the importance of asymptomatic cerebrovascular disease9,10 as well as the main associated risk factors and the need of finding useful markers for the early detection of this condition in order
a r c h s o c e s p o f t a l m o l . 2 0 1 6;9 1(1 1):513–519
515
to design diagnostic strategies applicable to broad population groups.11 On the other hand, well-known research is being carried out associating retinal microcirculation alterations with increased stroke and cardiac disease risk,12–17 to the extent of indicating associations with greater cognitive decline risk.18,19 Several analyses carried out by the Atherosclerosis risk in communities study related retinal microvascular alterations with subclinical or asymptomatic lesions found in NMR.20,21 However, said study does not include some microvasculature parameters that can be observed with the aid of specifically designed computer applications. Other large epidemiological studies reported similar findings in several sub-populations and ethnic groups.22–24 The objective of the present study is to analyze the value of eye fundus study with non-mydriatic retinography assessed with clinical criteria, supplemented by the semiautomatic determination of the arteriole-venule ratio applying a specific software, for detecting silent cerebral ischemia in hypertensive patients.
Subjects, material and methods Participants belonged to the cohort known as Investigating Silent Strokes in Hypertensives: a magnetic resonance imaging study (ISSYS), comprised by a randomized selection of 1037 hypertensive patients with ages between 50 and 70, without previous history of stroke or dementia, who underwent a cerebral NMR study with the aim of detecting cerebral infarcts and other asymptomatic brain injuries. In addition, various clinic and analytic assessments were made in order to determine the possible repercussion in target organs.10 The study protocol has been published.25 After obtaining informed consent, 976 of the 1037 patients of the cohort (those who had informed magnetic resonance) were invited to participate in the present study, which consisted in obtaining a retinography of both eyes centered on the optic disk/macula area at 45◦ without previous pharmacologically induced midriasis and in comparable lighting conditions carried out by 2 specifically trained technicians. The Topcon TRC-NW65 retinograph (Topcon SA, Barcelona, Spain) was utilized for obtaining the digital photographs. Said device included the following supplements: Nikon D7000 camera for Topcon TRC-NW6 retinograph, Nikon Corporation, Tokyo, Japan; Ibase DigiCaptura (Topcon SA, Barcelona, Spain); 19 TFT screen; ordenador computer with Windows 7 OS (Microsoft Corporation, Redmond WA, United States) type I, ATE-600 and Eye-Viewer software (Topcon SA, Barcelona, Spain) for visualizing and measuring retinal vasculature images. The images were stored in JPG format in a Microsoft Access database in which the data were completely anonymized without any clinical patient data. The retinographies were assessed by an ophthalmologist experienced in retinal pathology. All ophthalmoscopic findings were studied and adequately documented, particularly findings that determined the Keith and Wagener classification, i.e. arteriole-venule ratio (AVR), focal arteriole narrowing, presence of pathological arterial venous crossings,
Fig. 1 – Image showing the selection process of arteriole and venule vessels for obtaining the average of vascular diameters for calculating the arteriovenous ratio using the computer application. Source: http://www.imedos.de/Vesselmap2.98.0.html?&L=1.
hemorrhages or microaneurysms, hard exudates, cotton-like nodules and papilla edema. As it was established, supported by evidence, that the simple observation of the arteriole and vein diameter was very difficult to interpret due to observer subjectivity, the ophthalmoscopic AVR was substituted by that obtained with the semiautomatic measurement of arterial and venous inside diameters with available market software. After locating the optic nerve and establishing the region of interest, said software enabled the manual location of arterioles and venules (Fig. 1) in order to subsequently provide automatically the measurements of both vessels and calculation of the AVR. According with existing references, it was decided to consider AVR under 0.66 as pathological.26,27
Statistical analysis Quantitative variables are described with mean and standard deviation, while qualitative variables with absolute frequency and percentage. Quantitative variables were compared utilizing the T for student test for independent data, Pearson’s Chi square or Fishers exact test for qualitative variables, depending on the application conditions. Logistic regression analysis was performed to study the role of different independent variables as SCI risk factors. Bivariate analysis was performed first, obtaining the odds ratio (OR) for each independent variable as well as its 95% confidence interval (CI 95%). The variables having a Wald statistical equivalent to a significance of p ≤ 0.100 were initially included in a complete multivariate model adjusted for age and sex, eliminating those exhibiting a significance value of p > 0.050 and did not modify the ˇ coefficient values over 20%.
516
a r c h s o c e s p o f t a l m o l . 2 0 1 6;9 1(1 1):513–519
Table 1 – Participant characteristics summary based on gender.
Age Silent cerebral infarct n (%)
Males (n = 385)
Females (n = 383)
Total (n = 768)
62.6 (5.6) 59 (15.3)
62.9 (5.4) 22 (5.7)
62.7 (5.5) 81 (10.5)
Keith–Wagener criteria n (%) Altered AVRa ,b Crossings Microaneurysms Flame hemorrhage Soft exudate Hard exudate Papillary edema
74 (19.2) 278 (72.2) 19 (4.9) 7 (1.8) 5 (1.3) 7 (1.8) 0
45 (11.7) 267 (69.9) 13 (3.4) 1 (0.3) 2 (0.5) 2 (0.5) 0
119 (15.5) 545 (71.1) 32 (4.2) 8 (1.0) 7 (0.9) 9 (1.2) 0
Keith-Wagener Classificationb n (%) Without disease Grade I Grade II Grade III Grade IV
84 (21.8) 21 (5.5) 258 (67.0) 22 (5.7) 0
99 (25.8) 13 (3.4) 257 (67.1) 14 (3.7) 0
183 (23.8) 34 (4.4) 515 (67.1) 36 (4.7) 0
a b
p 0.407 <0.001
0.004 0.480 0.289 0.069 0.259 0.177 – 0.180
The arteriovenous ratio is considered altered or pathological when exhibiting a value <0.666. Arteriovenous ratio obtained by reading with the Imedos software.
Results Out of the 970 patients with informed NMR, retinal graphic images were obtained in 775. Seven cases could not be assessed due to poor quality image. Excluded patients were not included due to the inability to establish contact with them or who declined participation. Finally, retinographies of 768 patients were analyzed, diagnosing the presence of CSI by means of NMR. The mean age was 62.7 (±5.5) years, 50.1% being males. Table 1 shows the baseline characteristics of the studied population. Observed SCI prevalence was 10.5%, higher in males (15.3 versus 5.7%; p < 0.001). Most frequently observed lesions were pathological AVR (15.5%) and the presence of crossings (71.1%). Prevalence of males was higher only in AVR (19.2 versus 11.7%; p = 0.004). No eye fundus anomalies were found in 23.8% of patients. The logistic regression analysis for exploring candidate variables for SCI risk factors demonstrated that males and age over 60 increased OR significantly. Patients with Keith–Wagener classification grades i and iii exhibited greater risk (OR = 3.39 and 3.74 respectively; p < 0.005). In this classification, the significant criteria were altered AVR (p < 0.001) and the presence of microaneurysms (p = 0.039), as shown in Table 2. Subsequently, multivariate logistic regression analysis was carried out containing initially the Keith–Wagener classification or some of its criteria (AVR, microaneurysm, flame hemorrhage and hard exudate), in all cases adjusting for sex and age above 60. In the subsequent model, only the presence of altered AVR performed as risk factor with OR = 3.70 (2.21–6.18) (Table 3).
Discussion The results of the present study demonstrate that the alteration of retinal microvasculature is associated to an increased
risk of SCI in a range between 3.4 times greater risk for patients with Keith–Wagener grade I, and 3.7 times greater risk for patients with retinal involvement grade III. The alterations which, in isolation exhibit a statistically significant association with silent cerebral ischemia are diminished AVR under 0.66 assessed by means of semiautomatic software OR = 4.22 (2.56–6.96) and observation of microaneurysms OR = 2.50 (1.05–5.98), although the latter disappears when adjusting a multivariate model (Table 3). The baseline characteristics of patients who underwent retinography do not differ from those of the parent study population.10 Being a randomly obtained population-based study, it can be regarded as representative of the general hypertense population between 50 and 70 years of age in our country. Both populations (with and without retinography) reproduce SCI prevalence in the area of 10% and are located in the lower range of prevalences found in other large population-based epidemiological studies.3 However, said SCI prevalence is lower to that described in other hypertense cohorts. The difference could be attributed to specific sociodemographic characteristics (younger age) and with better arterial blood pressure control in the present cohort.10 Said differential characteristics should also be considered when assessing the repercussion of AHT on the retina of the present study population. In what concerns the prevalence of retinal lesions, it is noteworthy that the present study produced a high number of detected arteriovenous crossings (71% of retinographies), whereas similar studies in hypertense populations12 detected said alteration in 17.5% of cases and, in the general population,22 in 7.5%. For the rest of lesions, this discrepancy was not found when comparing the results of the present study with those of other series. The authors assume that this lesion could be overdiagnosed in the present series due to the fact that the observer was aware that the patients were hypertensive, whereas in the referenced studies retinography readings kept the observer masked in relation to the clinic condition of patients.
517
a r c h s o c e s p o f t a l m o l . 2 0 1 6;9 1(1 1):513–519
Table 2 – Bivariate logistic regression analysis of all candidates to be considered as SCI risk factors. SCI No. = 81
Beta
OR (CI 95%)a
p (Wald)
Sex n (%) Female Male
22 (5.7) 59 (15.3)
1.088
2.97 (1.78–4.96)
Age n (%) 50–55 56–60 61–65 > 65
3 (3.0) 14 (8.6) 24 (10.8) 40 (14.1)
– 1.118 1.361 1.672
– 3.06 (0.86–10.9) 3.90 (1.146–13.3) 5.32 (1.61–17.6)
– 0.085 0.029 0.006
Keith-Wagener Classificationc n (%) Without pathology Grade I Grade II Grade III Grade IV
13 (7.1) 7 (20.6) 53 (10.3) 8 (22.2) 0
– 1.121 0.406 1.318 –
– 3.39 (1.24–9.53) 1.50 (0.80–2.82) 3.74 (1.42–9.83) –
– 0.017 0.208 0.008 –
50 (7.7) 31 (26.1)
1.440
4.22 (2.56–6.96)
<0.001
22 (9.9) 59 (10.8)
0.099
1.10 (0.66–1.85)
0.708
74 (10.1) 7 (21.9)
0.917
2.50 (1.05–5.98)
0.039
79 (10.4) 2 (25.0)
1.504
2.87 (0.57–14.5)
0.201
80 (10.5) 1 (14.3)
0.348
1.42 (0.17–11.9)
0.749
79 (10.4) 2 (22.2) 0
0.898 –
2.46 (0.50–12.0) –
0.268 –
Keith-Wagener criteriac n (%) Altered AVRb No Yes Crossings No Yes Microaneurysms No Yes Flame hemorrhage No Yes Soft exudate No Yes Hard exudate No Yes Papillary edema
<0.001
CI 95%, confidence interval 95%. a b c
Dependent variable: silent cerebral infarct (diagnosed with magnetic resonance). Arteriovenous ratio is considered altered or pathological when exhibiting a value <0.666. Arteriovenous ratio obtained by reading processed by VesselMap2 software version 3.10 (Imedos).
Table 3 – Model obtained from the multivariate logistic regression analysis for the risk of presenting silent cerebral infarct. ˇ coefficient Constant Altered arteriovenous ratiob, c Male sex Age >60 years
OR (CI 95%)a
p (Wald)
−3.611 1.308
3.70 (2.21–6.18)
<0.001
1.031 0.743
2.80 (1.66–4.74) 2.10 (1.19–3.73)
<0.001 0.011
CI 95%, confidence interval 95%. a
b
c
Dependent variable: silent cerebral infarct (diagnosed with magnetic resonance). Arteriovenous ratio is considered altered or pathological when exhibiting a value <0.666. Arteriovenous ratio obtained by reading processed by VesselMap2 software version 3.10 (Imedos).
The results match those described by Kwon et al. in a group of 550 Korean hypertensive patients without cerebral ischemia antecedents, with risk levels similar to those found in the present study.26 Other transversal studies exhibit the same association between retinopathy and subclinic cerebral lesions.22,28 With the present data, hemorrhages (n = 2), soft exudates (n = 1) and hard exudates (n = 2) also appear as SCI risk factors without reaching statistical significance, which is attributed to the small number of cases. The case of arteriovenous crossings is also noteworthy because, in the present case, do not exhibit either association with SCI although said association is described in the literature,20–23 even in longitudinal studies which describe said crossings as a predictor of changes in white substance lesions. The authors believe that this could be due to the overdiagnosis of arteriovenous crossings in the present series, which has the effect of diluting the statistical results.
518
a r c h s o c e s p o f t a l m o l . 2 0 1 6;9 1(1 1):513–519
The article by Ikram et al.23 on a general population over 55 years of age describes the association between vein enlargement and the progression of microvascular lesions at the cerebral level. However, this association is not maintained when utilizing AVR as a parameter for measuring retinal circulation. In the present case, with currently available reading data as provided by the software, it is not possible to analyze separately the arterial and small vein diameters to determine which AVR component has greater influence in the variation thereof. As conclusion, the authors believe that despite the limitations described above, the results demonstrate that semiautomatic analysis of retinal vasculature on the basis of retinographies has the potential of being considered as a significant vascular risk factor in hypertensive populations.
Funding The present project has been financed by the Basque Government and the Regional Development European Fund of the European Union under the following programs: 1) NETs 2012–2014 (support for projects launching scientific and technology enterprises): IN-2012/0000061. IN2013/0000040. IN-2014/0000124. 2) Ekintzaile (00321/2013)/Txekintek (0154/2013). The project has also received a funding provided by ULMA Innovation SL.
Conflict of interests Co-author Jurgi Ugarte is linked to the technical staff of ULMA Innovation SL as a third-party subcontractor. ULMA Innovation is developing projects on image-based noninvasive diagnostics. Retinal images are one of its fields of research and development. Jurgi Ugarte has no other type of engagement or association, joint property or shareholdings in ULMA Innovation SL. The rest of co-authors have not declared any conflict of interests.
references
1. Fisher CM. Lacunes: small, deep cerebral infarcts. Neurology. 1965;15:774–84. 2. Norrving B. Evolving concept of small vessel disease through advanced brain imaging. J Stroke. 2015;17:94. 3. Fanning JP, Wong AA, Fraser JF. The epidemiology of silent brain infarction: a systematic review of population-based cohorts. BMC Med. 2014;12:119. 4. Keith N, Wagener H, Barker N. Some different types of essential hypertension: their course and their prognosis. Am J Med Sci. 1939:332–43. 5. Van den Born B-JH, Hulsman CA, Hoekstra JB, Schlingemann RO, van Montfrans GA. Value of routine fondocopy in patients with hypertension: systematic review. BMJ. 2005; 331:73. 6. Mancia G, Fagard R, Narkiewicz K, Redon J, Zanchetti A, Böhm M, et al. Guía de práctica clínica de la ESH/ESC para el manejo
7.
8.
9.
10.
11.
12.
13.
14.
15.
16. 17.
18.
19.
20.
21.
22.
23.
de la hipertensión arterial (2013). Rev Esp Cardiol. 2013;66:880.e1–64. Mancia G, Fagard R, Narkiewicz K, Redon J, Zanchetti A, Böhm M, et al. 2013 ESH/ESC guidelines for the management of arterial hypertension: the Task Force for the management of arterial hypertension of the European Society of Hypertension (ESH) and of the European Society of Cardiology (ESC). Eur Heart J. 2013;34:2159–219. Coll-de-Tuero G, Rodríguez Poncelas A, Faixedas Brunsoms D, Cadena Calderó A. Aplicación práctica de la retinografía en los hipertensos. Hipertens Riesgo Vasc. 2007;24:61–9. ˜ A, López-Cancio E, Dorado L, Millán M, Reverté S, Sunol Massuet A, et al. The Barcelona-Asymptomatic Intracranial Atherosclerosis (AsIA) study: prevalence and risk factors. Atherosclerosis. 2012;221:221–5. Delgado P, Riba-Llena I, Tovar JL, Jarca CI, Mundet X, López-Rueda A, et al. Prevalence and associated factors of silent brain infarcts in a mediterranean cohort of hypertensives. Hypertension. 2014. Jiménez M, Dorado L, Hernández-Pérez M, Alzamora MT, Pera G, Torán P, et al. Ankle-brachial index in screening for asymptomatic carotid and intracranial atherosclerosis. Atherosclerosis. 2014;233:72–5. Ong Y-TT, Wong TY, Klein R, Klein BE, Mitchell P, Sharrett AR, et al. Hypertensive retinopathy and risk of stroke. Hypertension. 2013;62:706–11. Yatsuya H, Folsom AR, Wong TY, Klein R, Klein BE, Sharrett AR. Retinal microvascular abnormalities and risk of lacunar stroke: atherosclerosis risk in communities study. Stroke. 2010;41:1349–55. McGeechan K, Liew G, Macaskill P, Irwig L, Klein R, Klein BE, et al. Prediction of incident stroke events based on retinal vessel caliber: a systematic review and individual-participant meta-analysis. Am J Epidemiol. 2009;170:1323–32. Liew G, Wong TY, Mitchell P, Cheung N, Wang JJ. Retinopathy predicts coronary heart disease mortality. Heart. 2009;95:391–4. McGeechan K. Meta-analysis. Retinal vessel caliber and risk for coronary heart disease. Ann Intern Med. 2009;151:404. Cheung N, Bluemke DA, Klein R, Sharrett AR, Islam FM, Cotch MF, et al. Retinal arteriolar narrowing and left ventricular remodeling: the multi-ethnic study of atherosclerosis. J Am Coll Cardiol. 2007;50:48–55. Ding J, Patton N, Deary IJ, Strachan MW, Fowkes FG, Mitchell RJ, et al. Retinal microvascular abnormalities and cognitive dysfunction: a systematic review. Br J Ophthalmol. 2008;92:1017–25. Gatto NM, Varma R, Torres M, Wong TY, Johnson PL, Segal-Gidan F, et al. Retinal microvascular abnormalities and cognitive function in Latino adults in Los Angeles. Ophthalmic Epidemiol. 2012;19:127–36. Cheung N, Mosley T, Islam A, Kawasaki R, Sharrett AR, Klein R, et al. Retinal microvascular abnormalities and subclinical magnetic resonance imaging brain infarct: a prospective study. Brain. 2010;133:1987–93. Hanff TC, Sharrett AR, Mosley TH, Shibata D, Knopman DS, Klein R, et al. Retinal microvascular abnormalities predict progression of brain microvascular disease: an atherosclerosis risk in communities magnetic resonance imaging study. Stroke. 2014;45:1012–7. Longstreth W, Larsen EK, Klein R, Wong TY, Sharrett AR, Lefkowitz D, et al. Associations between findings on cranial magnetic resonance imaging and retinal photography in the elderly: the cardiovascular health study. Am J Epidemiol. 2007;165:78–84. Ikram MK, de Jong FJ, van Dijk EJ, Prins ND, Hofman A, Breteler MM, et al. Retinal vessel diameters and cerebral small vessel disease: the Rotterdam Scan study. Brain. 2006;129: 182–8.
a r c h s o c e s p o f t a l m o l . 2 0 1 6;9 1(1 1):513–519
24. Cheung CY, Tay WT, Ikram MK, Ong YT, De Silva DA, Chow KY, et al. Retinal microvascular changes and risk of stroke: the Singapore Malay eye study. Stroke. 2013;44:2402–8. 25. Riba-Llena I, Jarca CI, Mundet X, Tovar JL, Orfila F, López-Rueda A, et al. Investigating silent strokes in hypertensives: a magnetic resonance imaging study (ISSYS): rationale and protocol design. BMC Neurol. 2013;13: 130. 26. Kwon HM, Kim BJ, Oh JY, Kim SJ, Lee SH, Oh BH, Yoon BW. Retinopathy as an indicator of silent brain infarction in
519
asymptomatic hypertensive subjects. J Neurol Sci. 2007;252:159–62. 27. Fischer MD, Huber G, Feng Y, Tanimoto N, Mühlfriedel R, Beck SC, et al. In vivo assessment of retinal vascular wall dimensions. Invest Opthalmol Vis Sci. 2010;51:5254. 28. Cooper LS, Wong TY, Klein R, Sharrett AR, Bryan RN, Hubbard LD, et al. Retinal microvascular abnormalities and MRI-defined subclinical cerebral infarction: the atherosclerosis risk in communities study. Stroke. 2006;37:82–6.