Use of SD-OCT to Quantitatively Evaluate Peripheral Congestion in Diastolic Dysfunction with a Role Model in Hemodialysis Patients with End-Stage Renal Disease.

Purpose: There is no practical quantitative tool to assess peripheral congestion in patients with congestive heart failure. In this study, we aimed to evaluate peripheral congestion by measuring choroidal thickness with SD-OCTin patients with end stage renal disease(ESRD). In addition, volume management by hemodialysis (HD) is often dicult in patients with ESRD. Volume overload or hypovolemia may increase morbidity and mortality in long-term HD patients. Therefore, it is important to determine the ideal ultraltration rate (UFR) for each patient. Also, we aimed to nd a data to help determine the ideal UFR. Methods: Fifty HD patients with ESRD were divided into 3 groups according to diastolic functions. Patients with mild diastolic dysfunction were enrolled in Group1, those with moderate diastolic dysfunction were enrolled in Group 2, and those with severe diastolic dysfunction were enrolled in Group 3. Choroid was used to evaluate peripheral congestion and choroidal thickness (CT) was measured by optical coherence tomography (OCT) before and after HD. Results: The average CT before HD in Group 3 (259.3 ± 7.5 μm ) was signicantly higher than in Group 1 (249.6 ± 9.4 μm) (p=0,015). The average CT after HD in Group 3 (224.1 ± 5.2 μm ) was signicantly higher than in Group 1 (232.4 ± 9.3 μm) (p=0,033). The mean of DCT was signicantly higher in Group 3 than in Group 1 (p< 0.001) Conclusion:DCT was correlated with diastolic function in ESRD patients undergoing HD, and further evaluation of diastolic function before HD may play a role in nding the ideal UFR. = 5); chronic glomerulonephritis (n = 4); and polycystic kidney disease (n = 1).


Introduction
Until now, peripheral congestion-edema is evaluated by pressing-pushing the pretibial area and graded between 0 and 4 according to the physician's experience. Moreover, this method is subjective.
End-stage renal disease (ESRD) is permanent irreversible loss of renal function. Patients with ESRD have a de ciency in removing uremic substances and water from body. Thus, the hemodialysis (HD) is used to control the body volume and composition by removal of water and uremic substances [1]. But, volume management is often di cult in HD patients with ESRD. It remains a mystery to assess the volume status and decide how much uid to take during HD, called the ultra ltration rate (UFR) [2].
Choroid is a highly vascularized cavernous tissue and choroid vascular resistance is low. Thus, the choroidal vascular circulation ow is very high. In addition, perfusion pressure affects the blood ow directly because the choroid is poorly autoregulated [3,4]. Therefore, the choroid can be easily affected by systemic diseases [5,6]. In Inter-dialytic period, volume overload negatively affects diastolic function [7]and causes a decrease in venous return. And increased volume load with reduced venous return leads to the accumulation of uid in tissues such as choroid. Previous studies showed that choroid thickness changes before and after HD [4,8,9].
In this study, ESRD patients who underwent HD were grouped according to their diastolic functions. Choroidal thickness (CT) measured by SD-OCT was aimed toevaluate the peripheral congestion. We also aimed to evaluate the importance of the diastolic function in calculating the UFR. Therefore, this study is different from previous similar studies.

Materials And Method
This study was planned retrospectively. Sixty seven patients with ESRD undergoing HD in hemodialysis unit of the Ağrı State Hospital were included to this study between 2014-2017.
The study conformed the tenets of the Declaration of Helsinki and also it was approved by local ethical committee (ID:71522473/050.01.04/197). The inclusion criteria was availability of OCT images with highquality.
Based on transthoracic echocardiogram, patients having diastolic dysfunction with preserved systolic function (EF>50%) were included. On the basis of the hemodialysis method, patients who underwent hemodialysis sessions three times a week for more than ve years were included. The exclusion criteria were as follows: the presence of any glaucomatous and retinal diseases except diabetic retinopathy, such as macular hole, epiretinal membrane, degenerative myopia, age-related macular degeneration, or macular edema of any origin.
Sixty seven patients met the study criteria. Seventeen patients were excluded because of normal diastolic function (9 patients), unavailability of OCT images (5 patients), macular edema (2 patients) and degenerative myopia (2 patient). The patients were divided into three groups according to diastolic dysfunction. Diastolic function grading was assigned based on current guidelines [10]. Group 1 was composed of patients with mild diastolic dysfunction (n = 20), Group 2 was composed of patients with moderate diastolic dysfunction (n = 20) and Group 3 was composed of patients with severe diastolic dysfunction (n = 10).
The patients underwent 3-h to 4-h hemodialysis sessions three times per week, using a high-performance dialyzer at a blood ow rate of 250 mL/min. The standard dialysate ow rate was 500 mL/min. A detailed ophthalmological examination was performed before and two to three hours after HD, including slit-lamb bio-microscopy, fundus examination, and OCT imaging. In selected cases, fundus angiography was used. All patients were examined by the same ophthalmologist (IO).
OCT was performed using the 3D OCT 2000 FA Plus Spectral Domain (Topcon Medical Systems, Inc. Tokyo, Japan), which has a 840 -nm wavelength light source, 5-µm axial image resolution, and speed of 27,000 Ascans per second. CT was measured using a single 9 -mm horizontal line scan through the center of the fovea with 1024 A-scan/B-scan captured 50 times in the same position and overlapped with 50 B-scan images by the OCT device software (9 mm × 0,15 mm, 1024 × 50 voxels). The enhanced choroidal mode was used to get a ne focus of the choroidal structure. B-scan images with low quality and indistinctive choroidal-scleral boundary were rejected. The eye that had a higher quality B-scan image was used for measurement. B-scan scale was adjusted to 1 : 1 mm and image size was doubled up. CT was described as the vertical distance between the outer edge of the hyperre ective retinal pigment epithelium and the scleral boundary and was manually measured at the fovea centralis by the same ophthalmologist using '"built-in caliper" in the linear measurement tools of the OCT device software.
Transthoracic echocardiograms were completed according to echocardiography laboratory protocol by same cardiologist (SCE). Images were analyzed investigators with remeasurement of all relevant parameters.
These included left atrial volume measurements, peak early (e) and atrial (a) velocities of mitral in ow, early mitral in ow deceleration time (DT), septal and lateral mitral annular e' velocities. Each measurement was averaged over multiple cardiac cycles. Diastolic function grading was assigned based on current guidelines [10].
Data was statistically analyzed of using the SPSS statistical package version 25.0 (SPSS Inc., Chicago, IL, USA). The Kolmogorov-Smirnov test was used to assess the normal distribution of range of CT, e, a, e', e/a and e/e'. Levene's test was used to assess variance homogeneity of the variables. Chi Square test and Kruskal Wallis test were used to evaluate the signi cance of differences between the groups. The choroidal change (ΔCT) was de ned by the formula: ΔCT = [CT after hemodialysis] −[CT before hemodialysis].
Additionally, the relationship between the range of CT and e/a was assessed using the Spearman rank correlation test. The value of statistical signi cance was set at P < 0.05.

Discussion
Peripheral congestion in ESRD patients with preserved ejection fraction could be affected by diastolic function.Therefore, unlike other studies, we grouped patients according to their diastolic function using ECO.
In this study, although pre-HD CT was signi cantly high in patients with severe diastolic dysfunction (p=0,015) and the same amount of uid was taken by the HD, these patients with severe diastolic dysfunction had signi cantly more reduction in post-HD CT than patients with mild diastolic dysfunction (p=0,033). In other words,although the same amount of body uid was extracted by HD (same UFR), the ΔCT was signi cantly higher in severe diastolic dysfunction group than in mild diastolic dysfunction group (p< 0.001). Moreover, the ΔCT was different in each group according to diastolic function (r = 0.570, p = 0.0001;p< 0.001). The ΔCT in group 2 was signi cantly higher than group 1 (p=0.023) and the ΔCT in group 3 was signi cantly higher than group 2 (p=0.021). In patients with preserved left ventricular ejection fraction, diastolic dysfunction causes an increase in end diastolic pressure, resulting in systemic congestion or insu ciency. The severity of this congestion can be determined by the severity of diastolic dysfunction. In present study, the diastolic function before HD was signi cantly correlated with ΔCT (r = 0.570, p = 0.0001).
This indicates that peripheral volume overload can be misleading in the presence of diastolic dysfunction. It also suggests that low UFR may be administered to get similar treatment response between patients with severe diastolic dysfunction and patients with mild diastolic dysfunction.
The choroid is vascularized and cavernous tissue with a low vascular resistance [3,11], so it can be easily affected by various ocular and systemic conditions [4,5,6,12,13]. For example, it was reported that the water drink test (1000 mL water intake in a short time) caused a signi cant increase in choroid thickness in healthy individuals [13]. This situation is similar to the patient with ESRD in the inter-dialytic period. In the inter-dialysis period, increased vascular hydrostatic pressure due to volume overload and decreased osmotic pressure due to volume overload may cause uid to accumulate in the choroid interstitial space. HD provides the opposite of this situation, as a result, choroidal vascular volume and interstitial uid volume decrease. Recent studies have shown that CT decreases after HD in patients with ESRD by using OCT [4,8,9]. Akihiro et al. [9]grouped patients with ESRD diabetic and non-diabetic, and found that CT decreased in all eyes after HD. The mean CT values before and after hemodialysis were 268 µm and 234 µm,similar to our results [9]. The ΔCT was also evaluated, and the ΔCT in the diabetic group was higher than in the non-diabetic group. In this study, the ΔCT in all patients was similar with diabetic group. Because the majority of our patients were ESRD patients due to diabetes. In the light of the above data [4,5,6,12,13], the choroid may be considered as a region of congestion under volume overload in patients with ESRD. Moreover, the choroid can be quantitatively evaluated by OCT [14].
In patients with ESRD, diastolic dysfunction is associated with high-left ventricular hypertrophy (LVH) as well as volume overload [15,16]. It is known that LVH is very common in ESRD patients and left ventricular mass index gradually increases after the start of HD [16,17]. Therefore, diastolic dysfunction is common in ESRD patients and is a more common mechanism of heart failure than reduction of the ejection fraction in patients with ESRD [15,16]. But the number of studies associated with diastolic dysfunctionin patients with ESRD is low. Mostly, the effect of dialysis on left ventricular function or left ventricular hypertrophy was investigated. In present study, left ventricular function was not evaluated.
Volume management in patients with ESRAD involves monitoring weight gains between HD sessions. In general, the weight gain between two HD sessions should be less than 1.5 kg (or < 20 ml / kg). High IDWG, more food intake and a strong appetite will often lead to increased UFR in the next HD.Or, low inter-dialytic weight gain, poor intake, diminished appetite and diarrhea require less uid removal in the next HD session [2]. When adjusting UFR, the patient's symptoms or morbid conditions and physical examination ndings should also be considered. For example; high pre-HD systolic blood pressure (e.g., > 160 mmHg), pulmonary edema, lower extremity and sacral edema may suggest volume overload or low UFR. Conversely, low pre-HD systolic blood pressure (e.g., < 120 mmHg), tachycardia, palpitation, chest pain, lightheadedness, cramps, sweating towards after HD session may suggest hypovolemia or high UFR [2,18]. Based on the above information, a patient with severe diastolic dysfunction giving symptoms and signs of volume overload may be administered high UFR for a long time. High body uid removal in patients with severe diastolic dysfunction may increase morbidity and mortality.
This study is limited by its retrospective nature and shares all the limitations of a retrospective study and represents a relatively small sample size in groups. Our SD-OCT device cannot give you the chance to create a 3D virtual image and measure the entire choroid volume. Furthermore, we aimed to associate the ΔCT with diastolic dysfunction change, but after HD it was not possible to evaluate diastolic function due to poor visibility of ECO. If we could associate the ΔCT with the diastolic function change, the correlation rate could be higher than 0.570. A prospective, randomized longitudinal study would allow to get more precise result.
Choroidal thickness measured by OCT can be used as a quantitative tool to assess peripheral congestion. Additionally, in ESRD patients with diastolic dysfunction, assessing the diastolic function periodically may help to estimate the each patient's volume status and may help to nd ideal UFR for each patients.

Declarations
Ethical Rules: The author has taken these ethical rules by Sakarya Universityat 2020 (ID: 71522473 / 050.01.04 / 197) to prepare, write and publish.
Author contribution: Both authors collected the data together. IO did the statistics and writing of the article.