Are aortic biomechanical properties early markers of dilatation in patients with Marfan Syndrome? – A Systematic Review and Meta-Analysis

Although the stiffness of tissue is known to play an important role in aortic dilatation, the current15 guidelines for offering a preventative aortic surgery in patients with Marfan syndrome rely solely on the aortic16 diameter. In this systematic review and meta-analysis, we analyze and compare literature on in-vivo aortic stiffness17 measurements in Marfan patients. Our aim is to assess the potential of these measurements as early indicators of18 aortic dilatation.19

Following the PRISMA guidelines, we collected literature on diameter and three in-vivo stiffness mea-20 sures: Pulse Wave Velocity (PWV), β-stiffness index and Distensibility, at five different aortic locations in patients21 with Marfan syndrome. Reported results were reviewed and compared against each other. For meta-analysis, an22 augmented dataset was created by combining extracted data from the reviewed literature. Regression with respect23 to age and statistical comparison were performed on the augmented dataset for all three measures at five different24 locations.25

30 articles reporting data from 1925 patients with Marfan and 836 patients without Marfan were reviewed.26 PWV was found to be statistically higher in Marfan at most aortic locations, but only when the aorta is already27 dilated. Distensibility was found to be lower at all aortic locations even in non-dilated aortas, and its decrease has28 been associated with higher chances of developing aortic dilatation. β-stiffness index was higher in Marfan patients29 and was positively correlated with the rate of aortic dilatation, emphasizing its role as a valuable indicator. In our30 meta-analysis based on a total 1197 datapoints, diameter was found to be higher only at the root (p < 0.001). All31 stiffness measures showed a significant variation with age. PWV at the root and carotid-femoral region was not32 statistically different (p = 0.62 and p = 0.14 respectively), but was positively correlated with age at all locations.33 Distensibility and β-stiffness index were different in Marfan patients at all locations, and the difference was more34 pronounced after accounting for age-related variation.35

Based on the results in the literature, β-stiffness index and distensibility emerge as the best predictors36 of future aortic dilatation. Our meta-analysis quantifies age-related changes in aortic stiffness and highlights the37 importance of accounting for age in comparing these measurements. Missing diameter values in the literature limited38 our analysis. Further analysis based on combined aortic stiffness and diameter criteria is recommended to evaluate39 aortic disease in a comprehensive way and assist clinical decisions for prophylactic surgery.40

Marfan syndrome; In-vivo stiffness; Aorta; Pulse wave velocity (PWV); Distensibility, β-stiffness index41 1 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted January 7, 2024. ; https://doi.org/10.1101/2024.01.05.24300824doi: medRxiv preprint NOTE: This preprint reports new research that has not been certified by peer review and should not be used to guide clinical practice. 1 Introduction42 Marfan Syndrome (MFS) is a heritable connective tissue disorder caused by a mutation of the fibrillin-1 gene (FBN1).43 The FBN1 mutation increases the fragmentation of elastic fibers in the aortic media, leading to compromised strength44 and structure of the tissue [1]. As a consequence, MFS exposes patients to higher risks of aortic diseases such as45 dilatation, dissection and rupture due to stiffening of the vessel’s walls. 50% of undiagnosed and untreated MFS46 patients die by the age of 40 due to cardiovascular complications [2]. Early diagnosis via cardiovascular imaging is47 thus crucial for suspected Marfan patients, and close monitoring is essential for confirmed patients. Echocardiograms48 and Magnetic resonance imaging (MRI) are examples of imaging procedures conducted to evaluate changes in aortic49 size and expansion rate. The current 2022 ACC/AHA clinical guidelines for preventive surgical intervention is50 largely based on the aortic root diameter with a threshold at 50mm, or 45mm in patients with increased risks of51 aortic dissection [3]. Surgery is also recommended when the cross-sectional aortic root area to patient height ratio52 is greater than 10 cm 2/m. However, dissection and rupture are known to occur below these thresholds [4, 5, 6], and53 diameter alone may not fully account for the biomechanical properties of aortic tissue, which are expected to play54 an important role in aneurysm progression and adverse events. While only diameter is currently used to predict55 risks of dissection and rupture, aortic stiffness emerges as a predictor of aortic dilatation, offering insights into the56 probability of adverse events.57 Considerable research has focused on investigating how biomechanical properties of the aorta can be early predic-58 tors of dilatation using in-vivo measures of aortic stiffness, namely 1) pulse wave velocity (PWV), 2) distensibility,59 and 3) β-stiffness index. Although some results have shown that aortic stiffness measured in-vivo may perform60 better than diameter assessment to predict aneurysmal growth, no consensus has been established on what stiffness61 indicator to use, which aortic segment to consider, how it differs from healthy patients, and how the differences evolve62 with age.63 This systematic review of the literature therefore aims at gathering published research studies that report aortic64 stiffness using PWV, distensibility and β-stiffness index in patients with Marfan syndrome and healthy patients.65 A comprehensive analysis of their correlation with age, and their potential as early indicators of aortic dilatation66 is conducted. In addition, to overcome variations in individual studies, we aim to create a larger, consolidated67 dataset from the selected studies and perform a meta-analysis to determine age- and disease-related variations and68 differences.69 2 Methods70 The following systematic review adheres to the Preferred Reporting Items for Systematic Reviews and Meta-analyses71 (PRISMA) recommendations and guidance [7]. In this section, the eligibility and search criteria necessary for the72 identification and selection of relevant published studies are defined. The data extraction and augmentation process,73 as well as the statistical techniques used to conduct the meta-analysis are also described.74 2.1 Inclusion criteria, information sources and search strategy75 Studies focusing on one or more of the three clinically established aortic stiffness measures - PWV, distensibility and76 β-stiffness index - were selected for patients diagnosed with MFS. Regarding the diagnosis of MFS, the revised Ghent77 criteria is the most widely accepted since its proposition in 1996 [8]. Therefore, only the studies published between78 1996 and September 2022 were included. Cohort studies, cross-sectional studies, case-control studies, and case series79 were considered, whereas conference abstracts, book chapters, case reports, reviews, editorials, expert opinions and80 letters were excluded. The review focused on early signs of dilatation, and thus excluded papers investigating severe81 complications such as dissection and rupture. Articles focusing on the following aspects were also excluded: effect of82 medication on aortic stiffness, ex-vivo mechanical characterization, cellular scale investigations, and effect of aortic83 curvature on its mechanical behavior. Additionally, the review was restricted to publications in the English language.84 Two electronic databases, PubMed and ScienceDirect, were screened to find publications based on the inclusion85 criteria. A time filter was applied to encompass research published between 1996 and September 2022. Between May86 2022 and September 2022, the databases were searched using the following MeSH terms: ‘Aortic’ AND ‘Stiffness’87 AND ‘Marfan’,88 2 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted January 7, 2024. ; https://doi.org/10.1101/2024.01.05.24300824doi: medRxiv preprint 2.2 Clinical metrics of aortic stiffness89 Stiffness refers to the ability of a material to withstand deformation under an applied force. The in-vivo stiffness of90 arteries is influenced by their geometry and the biomechanical properties of the tissue. In this section, we define the91 three in-vivo parameters that are used in this review to quantify aortic stiffness.92 ■ PWV is defined as the speed of the pressure waveform over a designated portion of a vessel. Higher PWV values indicate a wave that travels faster along the arterial segment, generally a consequence of stiffer tissue. The wave is detected using pressure transducers or Doppler echocardiography, and its travel time, called transit time, is measured by estimating the time of travel of the foot of the wave over a known arterial distance. The pulse wave velocity is therefore calculated as the distance between two chosen points divided by the transit time, PWV = Distance Transit time. (1) ■ Distensibility (Dist) is directly calculated on in-vivo images. It is defined as the relative change in a vessel’s lu- minal area for a unit pressure increment. Thus, a lower distensibility value indicates stiffer tissue. From in-vivo images, the arterial luminal area at systole and diastole, denoted as As and Ad respectively, are measured at a chosen location. Pressure measurements are commonly taken at the brachial artery using a sphygmomanome- ter cuff, where the systolic pressure (denoted as Ps) and diastolic pressure (denoted as Pd) are measured. Distensibility is thus calculated as Dist = As − Ad Ad(Ps − Pd) . (2) ■ The β-stiffness index ( β-SI) is also derived from in-vivo image-based measurements. Higher values indicate a stiffer tissue, i.e., β-SI is inversely correlated to the distensibility. β-SI is defined as the logarithm of the pressure ratio to relative change in diameter, and is dimensionless. The diameter in systole and diastole is measured on cross-sectional views of the aorta, and pressure using sphygmomanometer cuff. β-SI is calculated as β-SI = ln  Ps Pd  Dd Dd − Ds , (3) where Ps, Pd, Ds and Dd are the systolic pressure, diastolic pressure, systolic diameter and diastolic diameter,93 respectively.94 2.3 Data collection process95 The details of the selected papers and their full-text manuscripts were stored in a reference management software96 (Zotero). Information from these manuscripts was extracted, capturing the following data: publication details, overall97 aim of the study, study design (prospective, retrospective, multicenter, or longitudinal), cohorts’ size, cohorts’ mean98 age, imaging modality (Echocardiography (Echo) or Magnetic Resonance Imaging (MRI)), presence of diagnosed99 aneurysms in the cohort, and the aortic stiffness measure reported at five aortic locations: aortic root, ascending100 aorta (Aao), aortic arch (Arch), descending aorta (Dao), and carotid-femoral (only for PWV).101 Discrepancies in results from the literature may be attributed to the specificity of the cohorts included in each102 individual paper, particularly factors like age and population size. Additionally, statistical tests conducted on their103 sampled cohorts might not be fully representative of the broader population. A meta-analysis was therefore sought104 on the following quantities of interest: PWV, distensibility, β-stiffness index, diameter, and age. The aim was105 to perform a comprehensive statistical analysis by treating the collective results from the literature as one unified106 dataset, and offering deeper insights than what can be derived from individual papers. The unified dataset creation107 required extraction of data points from selected manuscripts, but with the following exclusion criteria. For studies108 investigating several connective tissue disorders, measurements were excluded from the meta-analysis if Marfan data109 points could not be separated from others. Since the focus of this analysis is on the native biomechanical properties110 without any effects of surgical intervention, measurements were also rejected if results from patients who underwent111 an aortic surgical procedure (e.g., PEARS, Bentall procedure) were not separable from the rest of the Marfan cohort.112 However, data points with unknown surgical status were included under the assumption that if surgery was performed,113 it would be explicitly mentioned in the respective articles. Patients under medication were included since that they114 represent a large portion of the diagnosed Marfan population. When studies presented data separately for aneurysmal115 3 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted January 7, 2024. ; https://doi.org/10.1101/2024.01.05.24300824doi: medRxiv preprint PWV Distance Transit time 1√ρDist r (Ps−Pd)β-SI 2ρ ln  Ps Pd  Distensibility (Dist) 1 ρP W V 2 As−Ad Ad(Ps−Pd) 2 ln  Ps Pd  (Ps−Pd)β-SI β-stiffness index (β-SI) 2ρ ln  Ps Pd  P W V 2 (Ps−Pd) 2 ln  Ps Pd  (Ps−Pd)Dist ln  Ps Pd  Dd Dd−Ds Table 1: Calculation equations (in gray cells) and conversion equations (in white cells) for the three aortic stiffness measures and non-aneurysmal Marfan aortas, only the non-aneurysmal measurement was selected. For longitudinal studies116 with several time points reported, only the baseline measurement was extracted. Finally, in papers where results117 were reported per age range, the mean was calculated and collected. For articles where individual participant data118 points were available, they were digitized directly from plots in the manuscript using WebPlotDigitizer [9]. In cases119 where such individual data points were not provided, mean and standard deviation (SD) or median and interquartile120 ranges (IQR) were manually extracted for each quantity.121 2.4 Data augmentation process122 In the literature, various image-based aortic stiffness measures are employed to characterize the mechanics of vascular123 walls. However, the measures often use different units, leading to inconsistency and lack of standardization, as124 previously pointed out by Alhalimi et al. [10]. To address this issue, our data augmentation process involved125 employing standardized formulae and units, as well as conversion equations to transform one aortic stiffness index126 into another. Specifically, distensibility can be converted into β-SI using the relative change in area ( As − Ad)/Ad =127 (D2 s − D2 d)/D2 d ≈ 2(Ds − Dd)/Dd where As, Ad, Ds, Dd are systolic area, diastolic area, diameter in systole and128 diameter in diastole respectively. The conversion of PWV to image-based aortic stiffness measures can be achieved129 using the Bramwell-Hill equation [11]. The conversion formulae are summarized in Table 1, where ρ is the blood130 density approximated to be 1059 kg/m3. Mean values of systolic ( Ps) and diastolic pressure (P d) reported in the131 articles were used for the calculations. It is worth noting that in cases where only the pulse pressure ( Ps − Pd) was132 reported instead of systolic and diastolic pressures individually, the β-stiffness index could not be calculated.133 In instances where individual data points were not reported, mean and standard deviation of the quantity were used. However, as noted by Weir et al. [12], when results exhibit a skewed distribution, researchers often report the median and interquartile ranges instead of the mean and variance information. To ensure that such cases were not excluded from the analysis, missing mean and standard deviation values were calculated from the provided median and interquartiles using Wan et al.’s method [13], described as follows. The mean ˜ x can be estimated from median and interquartiles as ˜x ≈ q1 + m + q3 3 , (4) where q1 and q3 represent the first and third interquartiles, and m denotes the median. The standard deviation SD is estimated as SD = q3 − q1 2Φ−1  0.75n−0.125 n+0.25  , (5) where Φ is the cumulative distribution function of a standard normal distribution and n is the size of the cohort.134 To summarise, the augmented dataset on which the meta-analysis is conducted consisted of three types of data:135 (i) Individual patients’ data points that were directly digitized and collected from plots in the manuscripts.136 (ii) Mean and standard deviation values collected from the manuscripts.137 4 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted January 7, 2024. ; https://doi.org/10.1101/2024.01.05.24300824doi: medRxiv preprint (iii) Mean and standard deviation values calculated from the median and interquartiles reported in manuscripts138 using Wan et al.’s method [13].139 (iv) Calculated values, which were generated using the conversion equations from Table 1 to transform one reported140 aortic stiffness measure into another.141 2.5 Statistical analysis142 In the meta-analysis, comparison tests between Marfan and control and linear regressions with age were performed on the augmented dataset. As stated previously, the dataset is composed of individual datapoints as well as mean and standard deviation values. In order to run statistical tests, all values were expressed as mean and SD, such that the overall means for the Marfan ( i = M) and control ( i = C) cohorts are calculated as Xi = P(µjnj)P nj (6) where µj and nj are the mean and size of the cohort in paper j, and the corrected sample standard deviations are calculated as si = 1P nj − 1 X ((nj − 1)σ2 j + nj(µj − Xi)2)  (7) where σj is the standard deviation in paper j.143 Marfan and control were compared using the Welch test, which examines the null hypothesis that two populations144 have equivalent means. This test is favored over the Student’s t-test when the two samples have unequal variances.145 Bonferroni correction was applied to account for multiple comparisons type two error.146 To better understand the independent effect of Marfan syndrome on distensibility, PWV or β-SI and to account147 for age-related effects, linear regressions and projections at age-zero were conducted. To perform the regressions, the148 normality of the dataset was tested using Shapiro–Wilk test. Due to the lack of consistent normality in the initial149 data distribution, a logarithmic transformation was applied to the dataset. The association between variables and age150 were evaluated using the coefficient of determination (R2). The slope of the linear regression enables us to determine151 if age-related changes occur at a faster rate in Marfan patients. To investigate whether Marfan patients are born with152 altered stiffness or if it changes over time, log-values of the three aortic stiffness measures were projected at age-zero153 using the linear regression. Welch comparison tests were then run on the projected values. The projection allowed154 us to discern whether statistical differences between the two groups can be found once the age-related variations had155 been factored out. In all statistical tests, significance was considered at a p-value less than 0.05. All analyses were156 performed using Python and the Scipy library.157 3 Results158 3.1 Search results159 The flowchart in Figure 1 illustrates the paper selection process following PRISMA guidelines. Initially, 639 published160 articles were identified, comprising 76 papers from PubMed, 553 from Science Direct, and 10 from reference list hand-161 searching. After removing duplicates, 616 papers remained, which were then assessed against exclusion/inclusion162 criteria by examining the title and abstract only. Among them, 572 articles were excluded, primarily for being163 unrelated to in-vivo measures of aortic stiffness in Marfan diagnosed patients, or for focusing on blood flow patterns.164 Subsequently, 44 texts were read fully, and 14 were rejected for either not using the revised Ghent criteria for MFS165 diagnosis [8] or for merging Marfan patients with other tissue disorders. Finally, a total of 30 papers were selected166 based on the eligibility criteria and search strategy mentioned in Methods section. For the meta analysis, 6 of the167 30 articles were excluded for not reporting Marfan data separately from others.168 3.2 Characteristics of included studies169 The review of the literature offered valuable insights into the assessment of stiffness in patients with Marfan Syndrome,170 as it encompassed studies from diverse sources published over a span of 20 years in ten different countries. Out of171 the 30 selected studies, 16 were longitudinal, providing crucial information on how stiffness evolves with age. The172 5 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted January 7, 2024. ; https://doi.org/10.1101/2024.01.05.24300824doi: medRxiv preprint Studies identified through electronic database search PubMed (n = 76) Science Direct (n = 553) Identification

identified through hand-search of reference lists (n = 10) Screening Duplicate records removed (n = 23) Records screened (title ± abstract) (n = 616) Included Records excluded (n = 572) ▪Unrelated to in-vivo measures of aortic stiffness ▪Focus on blood flow patterns Full-text assessed for eligibility (n = 44) Reports not retrieved (n = 14) ▪Do not use the revised Ghent criteria for Marfan diagnosis ▪ Marfan syndrome is not separated from other connective tissue disorders Reports excluded from meta- analysis (n = 6) ▪Quantitative data does not separate Marfan from other cohorts Studies included in the literature review (n = 30) Studies included in the meta- analysis (n = 24) Figure 1: The flow diagram presents the process of inclusion given the eligibility criteria using the PRISMA 2020 guidelines 6 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted January 7, 2024. ; https://doi.org/10.1101/2024.01.05.24300824doi: medRxiv preprint remaining 14 were case-control studies, allowing for a quantification of the differences in stiffness properties between173 individuals with and without Marfan Syndrome.174 Of the selected papers, 17 used echocardiography to monitor the size of the aorta and to calculate stiffness175 measures at various aortic locations, while 13 employed MRI. Only one study by Prakash et al. [14] utilized both176 imaging modalities: MRI for distensibility and β-SI calculation and echocardiography for monitoring the aortic size.177 Table 2 provides a comprehensive overview of the included papers in the qualitative synthesis, along with their178 respective main characteristics.179 Table 2: Main characteristics of the 30 studies included in the systematic review and meta-analysis Author Country Marfan cohort: size (mean age) Control cohort: size (mean age) Imaging modality Included in the meta- analysis? Quantity reported Groenink et al. [15] Netherlands 78 (31) 23 (28) MRI Yes PWV, Distensibil- ity Sandor et al. [16] Canada 14 (15.7) 6 (12.3) Echo Yes PWV, β-SI Nollen et al. [17] Netherlands 78 (31) - MRI Yes PWV, Distensibil- ity Baumgartner et al. [18] Austria 19 (17.7) 19 (17.7) Echo Yes Distensibility, β-SI Oosterhof et al. [19] Netherlands 78 (31) 17 (44) MRI Yes PWV Bradley et al. [20] Canada 26 (13.14) 69 (13.14) Echo Yes PWV, β-SI Vitarelli et al. [21] Italy 31 (26) 21 (26) Echo Yes PWV, Distensibil- ity, β-SI Baumgartner et al. [22] Austria 46 (17.4) 46 (17.6) Echo Yes Distensbility, β Fattori et al. [23] Italy 20 (27.8) 14 (29) MRI Yes Distensibility Mortensen et al. [24] Germany 50 (32) - Echo Yes PWV Kiotsekoglou et al. [25] U.K 31 (31) 31 (33) Echo Yes PWV, β-SI Westenberg et al. [26] Netherlands 25 (36) 25 (36) MRI Yes PWV Wit et al. [27] Australia 55 (40.5) 69 (41.35) Echo Yes PWV, distensibil- ity, β-SI Kr¨ oner et al. [28] Netherlands 21 (36) 26 (30) MRI Yes PWV Teixido-Tura et al. [29] Spain 80 (32) 36 (35.2) MRI Yes PWV, Distensibil- ity Prakash et al. [14] U.S 45 (27) - Echo, MRI No Distensibility, β-SI Akazawa et al. [30] Japan 26 (15) - Echo No Distensibility, β-SI Singh et al. [31] U.S 15 (36.9) 10 (42.9) MRI No β-SI Merlocco et al. [32] U.S 26 (25.11) - MRI No Distensibility, β-SI 7 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted January 7, 2024. ; https://doi.org/10.1101/2024.01.05.24300824doi: medRxiv preprint Grillo et al. [33] Italy 51 (12) 80 (11.9) Echo Yes PWV Salvi et al. [34] Italy 116 (33.7) - Echo Yes PWV Selamet Tierney et al. [35] U.S 608 (11.2) - Echo No Distensibility, β-SI Sch¨ afer et al. [36] U.S 20 (18) 22 (15) MRI Yes PWV, Distensbility Guala et al. [37] Spain 117 (25.3) - MRI Yes Distensibility Yan et al. [38] Germany 69 (34.43) 90 (67.27) Echo Yes Distensibility, β-SI Guala et al. [39] Spain 44 (36.95) 36 (39.40) MRI Yes PWV, Distensibil- ity Cui et al. [40] Canada 49 (17.9) 87 (18.20) Echo Yes PWV, β-SI Andel et al. [41] Netherlands 35 (28) - MRI Yes Distensibility Weismann et al. [42] Sweden 20 (22) 67 (25) Echo Yes PWV, Distensibil- ity, β-SI Cox et al. [43] U.S 32 (21.1) - Echo Yes Distensibility, β-SI In total, the review included data from 1925 patients with MFS and 836 patients without MFS, treated as180 controls, with mean age of participants ranging from 2 to 90 years old. A considerable fraction of the papers (10181 out of 30) exclusively reported data for the Marfan cohort and did not include control data. As shown on Figure 2,182 diameter was primarily reported at the root and ascending aorta, with only about a third of the studies comparing183 Marfan and control groups. In contrast, PWV was predominantly reported in the ascending part of the aorta, with184 only one study that considered controls in the descending aorta (Dao) and another in the abdominal aorta (Abao).185 Distensibility was mostly reported in the ascending aorta, but only about a third of the papers included control data186 for comparison. The β-stiffness index was mainly reported in the root and ascending aorta.187 In Figure 3, the size of the Marfan cohort is depicted in relation to their age. The horizontal bars displayed188 represent mean ± SD for each paper. Most of the articles encompassed cohorts with less than 60 patients, with age189 ranges spanning from pediatrics to 50 years old. Only three studies exceeded 100 participants and none of them190 included controls in their analysis.191 3.3 Main findings from the selected papers192 3.3.1 Aortic diameter is the standard measure to compare Marfan and control193 Among the 30 selected papers, 14 compared the diameter at the root between Marfan and control cohorts, 14 at194 the ascending aorta, and 3 at the descending aorta. Table 3 summarizes the results of these comparisons. The195 findings from these studies indicated significant differences in aortic size between the two groups. Notably, all papers196 reported a larger aortic root in Marfan patients, irrespective of the age of the cohort or correction for Body-Surface-197 Area (BSA). 10 studies found that the ascending aorta (Aao) was larger in Marfan patients, and 5 studies found no198 statistical difference in the diameters of the descending aorta (Dao).199 3.3.2 PWV is higher in Marfan patients, but only with dilated aortas200 Among the selected studies, 7 compared PWV at the Aao between Marfan and control cohorts, 8 examined PWV201 at the arch, 7 in the descending aorta, and 5 from carotid to femoral (Table 3). Consistently, a significantly higher202 PWV was observed in all aortic regions from proximal to distal, as well as from carotid to femoral, in MFS patients.203 These differences remained valid even after adjusting for age and diameter. This finding highlights the presence of204 increased aortic stiffness in MFS patients across various regions of the aorta. However, in Oosterhof et al. [19] study,205 after correcting for age and diameter for the ascending aorta, no significant difference in PWV was observed between206 adult patients with and without Marfan Syndrome.207 8 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted January 7, 2024. ; https://doi.org/10.1101/2024.01.05.24300824doi: medRxiv preprint Number of articles 23 20 2 12 7 1 4 7 9 6 8 9 15 2 9 3 9 10 1 5 2 7 4 0 2 1 0 0 0 1 1 3 4 4 1 2 2 3 2 0 0 1 0 5 10 15 20 25 30 Root Ascending Arch Descending Abdominal Root Aao Arch Dao Abao Carotid-femoral Root Ascending Arch Descending Abdominal Root Ascending Arch Descending Abdominal Diameter PWV Distensibility βstiffness index Only Marfan values are reported Marfan and control values are reported and compared Figure 2: Number of articles reporting diameter, PWV, distensibility and β stiffness index at various locations. Black bars represent papers reporting only Marfan data, and gray bars represent papers including a control cohort and comparing results to Marfan. 9 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted January 7, 2024. ; https://doi.org/10.1101/2024.01.05.24300824doi: medRxiv preprint Figure 3: Mean age and standard deviation of the Marfan cohort in selected articles, organized by cohort size. The histogram on the top illustrates the number of articles containing patients within specific age ranges. The histogram on the right depicts the number of articles containing specific cohort sizes. One study (Selamet Tierney et al. [35]) is not included in this plot and consists of 608 MFS patients with a mean age of 11.2 and a standard deviation of 6.3. 10 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted January 7, 2024. ; https://doi.org/10.1101/2024.01.05.24300824doi: medRxiv preprint Diameter PWV Distensibilit y β stiffness index Comparison b etween Marfan and control Ro ot Aao Dao Aao Arc h Dao Carotid- femoral Ro ot Aao Dao Ro ot Aao Dao Statistically differen t [15]∗ [20]∗ [16] [18]∗ [22]∗ [19] [25]∗ [29] [21]∗ [27]∗ [14]∗ [38]∗ [40] [42] [16] [18]∗ [22]∗ [19] [25] [29] [21]∗ [36] [38]∗ [42] [29] [21]∗ [36] [15] [26] [15] [19] [16] [20] [26] [28] [29] [36] [15] [19] [26] [28] [15] [26] [28] [34] [30] [23] [36] [32] [15] [29]∗ [21] [32] [21] [29]∗ [27] [16] [20]⋄ [18] [22] [21] [25] [40] [18] [22] [21] [31] Only statistically differen t when the aorta is already dilated [21] [39] [21] [29] [30] [30] [30] Not statistically differen t [15]∗ [20] [26]∗ [23]∗ [15]∗ [18]∗ [22]∗ [19] [31] [19]• [28] [36] [29] [36] [32] [30] [27] Table 3: Studies comparing diameter, PWV, distensibility and β-SI between Marfan and control, and

of the statistical tests at different aortic locations. The symbol ⋆ indicates quantities corrected for age, pulse pressure and diastolic area, ⋄ indicates that β-SI is corrected for sex, height and age, and • indicates that PWV is corrected for age and diameter. 11 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted January 7, 2024. ; https://doi.org/10.1101/2024.01.05.24300824doi: medRxiv preprint Additionally, in the studies conducted by Vitarelli et al. [21], Teixido-Tura et al. [29], and Guala et al. [39], it208 was observed that when separating the cohorts into dilated aortas and normal diameters, Marfan patients exhibited209 increased PWV compared to controls. However, this difference was significant only for already dilated aortas.210 According to Teixido-Tura et al. [29], compared with distensibility, PWV demonstrated a slower decrease at the211 early stage of aortic dilatation. This suggests that PWV might not be as sensitive to changes in aortic biomechanics212 during the initial stages of aortic dilatation compared to distensibility. However, as the aortic dilatation progresses,213 PWV gradually increases and eventually becomes significantly different between Marfan and control cohorts, but214 only when the aorta is already dilated.215 3.3.3 Aortic distensibility is lower in Marfan patients216 Among the selected papers, 2 studies compared the distensibility at the root between Marfan and control cohorts,217 7 studies examined the distensibility at the Aao, and 3 at the Dao (Table 3). Statistically lower distensibility was218 consistently reported in the aortic root, Aao, and Dao for Marfan patients. Furthermore, distensibility showed a219 significant decrease with age: Groenink et al. [15] highlighted that, compared to juvenile Marfan cohorts in the220 literature, the mean distensibility was nearly half in adult patients. Distensibility appears to be an early marker of221 biomechanical changes in Marfan Syndrome. In Akazawa et al. [30]’s study on children, a difference in distensibility222 of the root was observed between Marfan and control groups, even in non-dilated roots, at an early stage of life. In223 the Dao, distensibility did not differ whether the aorta was dilated or not.224 Finally, studies conducted by Teixido-Tura et al. [29] and Vitarelli et al. [21] on older cohorts also reported lower225 distensibility in adult Marfan patients, whether their aortic root, ascending, or descending aortas were dilated or226 non-dilated. These results indicate a decrease in distensibility starting from the proximal aorta.227 3.3.4 β-stiffness index is higher in Marfan patients228 Among the selected studies, 2 compared the β-SI at the root between Marfan and control cohorts, 8 examined the229 β-SI at the Aao, and 4 at the Dao (Table 3). The findings consistently revealed that the β-SI was significantly higher230 in the root, Aao, and Dao in the MFS group when compared to control patients, even after adjusting for factors231 such as sex, age, and height. Notably, unlike PWV, MFS patients demonstrated higher β-SI values than controls in232 both cases of aortic dilatation and normal aortic diameters. However, it is essential to consider the observation by233 Wit et al. [27] that, after 40 years of age, the β-SI did not show significant differences between Marfan and control234 cohorts. This may have implications in understanding the progression of stiffness changes associated with age in235 Marfan Syndrome patients.236 3.3.5 Biomechanical stiffness measures can be early predictors of aortic dilatation237 In the 17 longitudinal studies, Marfan patients were followed at various time points throughout their lives, and the238 evolution of stiffness parameters was measured. Statistical tests could identify a potential trend in the parameters’239 evolution with age. The predictive power of each parameters was thus assessed with regards to aortic dilatation.240 PWV241 In the study conducted by Groenink et al. [15], PWV was strongly correlated with age in control subjects at all levels242 of the aorta. However, in MFS patients, the increase in PWV with age was significantly higher in the proximal aorta243 compared to healthy subjects, supporting the hypothesis of media degradation starting at the root [34]. Despite244 its correlation with age, PWV was not found to be associated with progressive aortic dilatation at any level in the245 longitudinal study by Nollen et al. [17]. This suggests that PWV may not be a reliable candidate for predicting246 future aortic dilatation in MFS patients. Furthermore, in patients with aortic root replacement, even though the247 distensibility of the graft was significantly lower than the distensibility of the native aorta, the PWV showed no248 differences [17]. This indicates that PWV, as a regional measure, may not adequately differentiate diseased tissue249 locally and may be insensitive to differences in tissue composition.250 A noteworthy exception is that PWV demonstrated high specificity and low sensitivity for predicting the absence251 of regional dilatation in MFS patients in the longitudinal study by Kr¨ oner et al. [28]. Specifically, at least 78% of MFS252 patients who showed no aortic growth at follow-up did not have increased regional PWV at baseline. Conversely,253 less than 33% of patients who presented with increased PWV at baseline had increased aortic growth at follow-up.254 12 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted January 7, 2024. ; https://doi.org/10.1101/2024.01.05.24300824doi: medRxiv preprint Distensibility255 Several studies, including Baumgartner et al. [18], Akazawa et al. [30], and Sch¨ afer et al. [36], have demonstrated256 that distensibility can serve as a diagnostic parameter in addition to the current diameter measurements. The257 main reason is that distensibility was found to be lower even in patients with normal diameters at the root and258 ascending aorta. In a longitudinal study by Baumgartner et al. [22], the probability of developing an aneurysm259 was calculated based on ascending aortic distensibility. The findings revealed that higher distensibility measured at260 baseline was associated with a lower probability of developing aortic dilatation at follow-up. Similarly, Nollen et al.261 [17] demonstrated that distensibility was predictive of progressive descending thoracic aortic dilatation. A reduction262 of one unit in distensibility was associated with a 4-fold increase in the risk of dilatation, independent of aortic263 diameter. However, distensibility was not found to be a significant predictor of dilatation at other aortic locations,264 as noted by Teixido-Tura et al. [29]. This might be attributed to the relatively advanced stage of aortic disease in265 that particular study group. Additionally, Merlocco et al. [32] found a linear correlation between distensibility and266 age, with a slightly higher decline with age compared to normal subjects.267 β-stiffness index268 The study by Cox et al. [43] provided important insights into the relationship between theβ-stiffness index and aortic269 dilatation in Marfan patients. Their findings revealed that the β-stiffness index in the aortic root was positively270 correlated with the dilatation rate, indicating that higher β stiffness values were associated with a faster rate of271 aortic dilatation. Interestingly, the baseline aortic root dimension alone did not show a significant correlation with272 the dilatation rate. This highlights the potential of the β-stiffness index as an independent and predictive measure273 for assessing aortic dilatation in Marfan patients.274 Additionally, the β-stiffness index was the least dependant on blood pressure variation, making it a robust275 indicator of aortic stiffness, in comparison to distensibility [44]. Indeed, in Wada et al. [45] study on 7 subjects, no276 correlation was found between β-stiffness index and mean blood pressure. In Sugawara et al. [46] study, β-stiffness277 index did not change significantly after decreasing the blood pressure using α-adrenergic blockade.278 3.4 Meta analysis279 3.4.1 Analysis of the augmented dataset280 Through the data extraction and augmentation process, the dataset included 286 data points for diameter, 1063281 for PWV, 1063 for distensibility, and 733 for the β-stiffness index. Out of these, 1278 data points were associated282 with a corresponding age value: 1027 age points had corresponding PWV, distensibility and β-SI, 194 only had283 corresponding PWV and distensibility, and 57 only had corresponding diameter. Original data points constituted284 36% of the entire dataset, with the remaining being part of the augmentation process. Specifically for diameter, 62285 data points were mean values with 11 converted using Eqs. 4 and 5, and 222 individual points were extracted from286 manuscripts using WebPlotDigitizer. For PWV, 464 points were calculated using conversion formulas, 39 points were287 mean values with 8 using Eqs. 4 and 5, and 544 individual data points were from the manuscripts. As for distensibility,288 839 points were calculated using conversion formulas, 20 points were mean values with 9 using Eqs. 4 and 5, and 188289 were individual data points. Lastly, for the β-stiffness index, 579 points were calculated using conversion formulas,290 24 were mean values with 6 using Eqs. 4 and 5, and 238 were individual data points.291 Out of the 1172 data points in the augmented dataset, the majority (1082) were associated with patients who292 did not undergo surgery, 90 data points were unknown (not specified in the article). Regarding medication, 139 data293 points were from patients under medication such as beta-blockers, 866 data points were from patients not under294 medication, and 167 data points were unknown (not specified in the article). Table 4 summarizes the number of data295 points in the augmented dataset obtained from each article.296 3.4.2 Comparison tests between Marfan and control on the augmented dataset297 Consistent with previous literature findings, diameter was indeed statistically larger in Marfan patients at the Root298 (Figure 4), but no difference in the Aao and Dao was detected. Specifically, the mean diameter in the Marfan cohort299 was 3.91cm at the root, 2.98cm at the Aao, and 2.10cm in the Dao, compared to 3.05cm, 3.00cm, and 1.86cm,300 respectively, for the control cohort. Distensibility was significantly lower in Marfan patients except at the root after301 Bonferroni correction with mean values (in 10−3mmHg−1) of 2.46 in the Root, 3.57 in the Aao, 6.18 in the Arch, and302 4.23 in the Dao, compared to 2.99 6.11 , 8.17 , and 6.56, respectively, for controls (Fig. 5). The β-SI was also higher303 13 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted January 7, 2024. ; https://doi.org/10.1101/2024.01.05.24300824doi: medRxiv preprint Papers Number of datapoints Wit et al. [27] 244 Westenberg et al. [26] 149 Cui et al. [40] 134 Teixido-Tura et al. [29] 127 Groenink et al. [15] 109 Oosterhof et al. [19] 101 Baumgartner et al. [18] 95 Weismann et al. [42] 67 Sch¨ afer et al. [36] 27 Fattori et al. [23] 19 Baumgartner et al. [22] 14 Kiotsekoglou et al. [25] 8 Vitarelli et al. [21] 8 Bradley et al. [20] 8 Guala et al. [47] 8 Yan et al. [38] 8 Sandor et al. [16] 6 Kr¨ oner et al. [28] 5 Andel et al. [41] 5 Nollen et al. [17] 5 Salvi et al. [34] 4 Mortensen et al. [24] 4 Grillo et al. [33] 4 Guala et al. [37] 2 Cox et al. [43] 1 Table 4: Papers included the meta-analysis and the number of datapoints in the augmented dataset 14 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted January 7, 2024. ; https://doi.org/10.1101/2024.01.05.24300824doi: medRxiv preprint Figure 4: Mean and SD of aortic diameters in Marfan and control cohorts at three different locations show a signficant difference only at the root. in Marfan patients at all locations, with mean values of 13.93 in the root, 5.97 in the Aao, 4.42 in the Arch, and304 6.53 in the Dao, compared to 8.25, 3.64, 3.36, and 4.90, respectively, for the control cohort (Fig. 5). Interestingly,305 no statistically different PWV was found in the Root and from carotid to femoral (Fig. 5). Table 5 summarizes the306

of the statistical tests.307 Ro ot Aao Arc h Dao Carotid-femoral Diameter All ages combined: p = 7.89 × 10−9 All ages combined: p = 0.86 All ages combined: p = 0.15 PWV All ages combined: p = 0.62 Age-zero projection: p = 0.30 All ages combined: p = 0.0025 Age-zero projection:: p = 0.0020 All ages combined: p = 3.30 × 10−5 Age-zero projection: p = 3.25 × 10−9 All ages combined: p = 1.57 × 10−6 Age-zero projection: 2.92 × 10−14 All ages combined: p = 0.14 Age-zero projection: p = 0.65 Distensibilit y All ages combined: p = 0.015 Age-zero projection: p = 1.15 × 10−17 All ages combined: p = 8.36 × 10−13 Age-zero projection: p = 0.0020 All ages combined: p = 0.00011 Age-zero projection: p = 3.25 × 10−9 All ages combined: p = 2.67 × 10−7 Age-zero projection: p = 2.92 × 10−14 β - SI All ages combined: p = 1.97 × 10−5 Age-zero projection: p = 1.14 × 10−18 All ages combined: p = 5.16 × 10−11 Age-zero projection: p = 1.60 × 10−5 All ages combined: p = 0.00086 Age-zero projection: p = 3.01 × 10−9 All ages combined: p = 0.0027 Age-zero projection: p = 1.97 × 10−18 Table 5: p-values for the Welch comparison tests between Marfan and control patients, for the three stiffness measures at different locations 3.4.3 Linear regressions with age and projection at age-zero comparison308 Figure 6 and Table 6 present the outcomes of the linear regressions at various aortic locations. Because the quantities309 were not normally distributed, the linear regressions were performed after log transformation. Figure 7 and Table 5310 provides the results from the Welch comparison test after projecting the data to age-zero. The correlation analyses311 15 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted January 7, 2024. ; https://doi.org/10.1101/2024.01.05.24300824doi: medRxiv preprint Figure 5: Mean and SD of Distensibility, PWV and β-stiffness index for the augmented dataset, without age consideration. 16 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted January 7, 2024. ; https://doi.org/10.1101/2024.01.05.24300824doi: medRxiv preprint and age-zero projection tests revealed important findings regarding the relationship between PWV, distensibility,312 β-stiffness index, age, and Marfan syndrome.313 Notably, a positive correlation was observed between PWV and age in both Marfan and control patients at all314 aortic locations (p < 0.05). However, the age projection suggested that there was no statistically significant difference315 in PWV at the root (p = 0.30) and from carotid to femoral ( p = 0.65) between Marfan and control patients at age-316 zero. Additionally, both MFS and control patients exhibited a negative correlation between distensibility and age317 at all aortic locations. The age-zero projection demonstrated that distensibility was lower, indicating higher aortic318 stiffness in MFS patients already at birth. However, the slopes of the linear regressions revealed that distensibility319 decreased more rapidly for the control cohort compared to Marfan patients (-0.027 vs. -0.022 for MFS in the root,320 -0.016 vs. -0.010 in the Arch, and -0.026 vs. -0.021 in the Dao, all values in 10 −3mmHg−1 per year). This result321 could also be influenced by the fact that the meta-analysis does not include patients who have had surgery. Since322 patients who have had surgery tend to be older, this may introduce a bias in the dataset, potentially affecting the323 observed rate of distensibility decline in the Marfan cohort. The correlation analyses showed that β-SI was positively324 correlated with age in both Marfan and control patients at all aortic locations ( p < 0.05). After conducting age-zero325 projection, the results confirmed that there is a statistically significant difference in β-stiffness index between Marfan326 and control patients at all locations. Marfan patients had higher values after age-zero projection, indicating stiffer327 tissues at an early stage of life compared to controls.328 17 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted January 7, 2024. ; https://doi.org/10.1101/2024.01.05.24300824doi: medRxiv preprint ROOT AAO ARCH DAO PWV (cm/s) Distensibility (10 -3 mmHg) β-stiffness index CAROTID FEMORAL Figure 6: Linear regressions of PWV, distensibility and β stiffness index with respect to age at the different locations. Note the log scale on the vertical axis, since the regressions were performed on log-transformed values for each quantity to ensure normality. 18 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted January 7, 2024. ; https://doi.org/10.1101/2024.01.05.24300824doi: medRxiv preprint Ro ot Aao Arc h Dao Carotid-femoral PWV Marfan: y = 0.015x + 1.92 R2 = 0.14 ; p = 1.11 × 10−5 Control: y = 0.015x + 1.84 R2 = 0.15 ; p = 6.65 × 10−5 Marfan: y = 0.007x + 1.51 R2 = 0.24 ; p = 1.41 × 10−3 Control:: y = 0.0057x + 1.36 R2 = 0.19 ; p = 2.77 × 10−2 Marfan: y = 0.0049x + 1.43 R2 = 0.13 ; p = 1.70 × 10−2 Control: y = 0.0082x + 1.21 R2 = 0.35 ; p = 2.81 × 10−9 Marfan: y = 0.011x + 1.41 R2 = 0.20 ; p = 2.04 × 10−7 Control: y = 0.013x + 1.11 R2 = 0.56 ; p = 3.12 × 10−8 Marfan: y = 0.0095x + 1.36 R2 = 0.26 ; p = 5.17 × 10−8 Control: y = 0.0098x + 1.38 R2 = 0.18 ; p = 4.78 × 10−6 Distensibilit y Marfan: y = −0.022x + 1.57 R2 = 0.17 ; p = 9.25 × 10−7 Control: y = −0.027x + 2.30 R2 = 0.42 ; p = 1.26e − 13 Marfan: y = −0.014x + 1.81 R2 = 0.24 ; p = 1.41 × 10−3 Control: y = −0.011x + 2.11 R2 = 0.19 ; p = 2.77 × 10−2 Marfan: y = −0.0098x + 1.98 R2 = 0.13 ; p = 1.70 × 10−2 Control: y = −0.016x + 2.42 R2 = 0.34 ; p = 2.81 × 10−9 Marfan: y = −0.021x + 2.023 R2 = 0.20 ; p = 2.04 × 10−7 Control: y = −0.026x + 2.62 R2 = 0.56 ; p = 3.12 × 10−8 β - SI Marfan: y = 0.022x + 1.54 R2 = 0.19 ; p = 1.81 × 10−7 Control: y = 0.026x + 0.81 R2 = 0.44 ; p = 3.58 × 10−14 Marfan: y = 0.011x + 1.37 R2 = 0.15 ; p = 1.35 × 10−2 Control: y = 0.011x + 0.93 R2 = 0.19 ; p = 2.77 × 10−2 Marfan: y = 0.0098x + 1.18 R2 = 0.13 ; p = 1.70 × 10−2 Control: y = 0.016x + 0.74 R2 = 0.34 ; p = 2.81 × 10−9 Marfan: y = 0.020x + 1.14 R2 = 0.24 ; p = 5.64 × 10−4 Control: y = 0.037x − 0.027 R2 = 0.89 ; p = 3.10 × 10−10 Table 6: Results of the linear regression equations 4 Discussion329 4.1 Key findings in the literature330 The diameter of blood vessels is a commonly used indicator for detecting biomechanical changes, and abnormal331 diameters at specific locations in the aorta are considered signs of disease. Diameter was reported to be larger in332 Marfan patients at the root and Aao. These results align with previously observed patterns in pediatric Marfan333 patients, where dilatation typically begins at the sinus of Valsalva, followed by the sino-tubular junction (STJ), and334 is less frequent in the descending aorta [48]. The reported higher diameters at the root in MFS patients also provide335 support for the current guideline of measuring the aortic root to plan prophylactic surgery.336 Across most aortic regions, studies reported a significantly higher PWV in MFS when compared to control cohorts,337 but only in the already dilated aortas. Indeed, PWV was not as sensitive to changes in aortic biomechanics in the338 initial phases of aortic dilatation, e.g., compared to distensibility. This observation underscores the importance of339 considering the stage of aortic dilatation when interpreting PWV values and highlights the complexity of using PWV340 as a sole predictor for aortic dilatation in MFS patients. While it may provide valuable information about aortic341 stiffness, its ability to precisely predict aortic growth is limited. Combining PWV with other relevant parameters,342 such as distensibility, may yield more comprehensive insights into the dynamics of aortic changes in MFS and aid in343 making accurate clinical assessments.344 The results derived from the studies included in this review provided evidence that change in distensibility values345 can serve as an effective mean to detect alterations in the biomechanical properties of the aorta. Specifically, a346 decrease in distensibility was indicative of increased tissue stiffness within the aortic wall. Distensibility was lower at347 all aortic locations even in juvenile non-dilated aortas, with values half of those in the healthy patients. A significant348 correlation between distensibility and age was also put forward. Finally, longitudinal studies showed that a higher349 distensibility is associated with higher chances of developing an aneurysm in the Aao and Dao. Although sensitive to350 location, these findings underscore the potential utility of distensibility as an early indicator of aortic biomechanical351 changes in Marfan patients.352 19 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted January 7, 2024. ; https://doi.org/10.1101/2024.01.05.24300824doi: medRxiv preprint Figure 7: Mean and SD in log-scale of distensibility, PWV and β-stiffness index projected at zero age. The statistical tests were run on log-transformed values for each quantity. 20 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted January 7, 2024. ; https://doi.org/10.1101/2024.01.05.24300824doi: medRxiv preprint The results across the studies indicated that β-SI values were notably higher in the root, Aao, and Dao among353 patients with Marfan Syndrome in comparison to control subjects, even in non-dilated aortas. It was positively354 correlated with the rate of aortic dilatation, emphasizing its role as a valuable indicator. Unlike baseline aortic355 dimensions, β-SI appeared to be a better predictor of the progression of aortic dilatation. However, it is worth noting356 that, after 40 years of age, the control and MFS cohorts could not be distinguished using the β-SI. Interestingly,357 younger MFS patients had higher β-SI values compared to their age-matched controls, but this distinction diminished358 in older populations. This particular observation offers insights on progression of aortic stiffness in MFS patients as359 they age. It may suggest that altered aortic stiffness eventually leads to convergence between the Marfan and control360 cohorts after a certain age threshold.361 4.2 Meta-analysis results362 In the meta-analysis, in consensus with the literature, diameter was found to be significantly higher in MFS at the363 root, which align with the current guideline of measuring the aortic root to plan prophylactic surgery. Interestingly,364 no difference was found in the Aao and Dao. In this case, however, no age consideration was possible to due lack365 of data reporting diameter with respect to age. Combining all patients regardless of their age includes bias when366 comparing Marfan and control since diameter depends highly on age, weight and height of the patient. Diameter367 assessment adjusted for body surface area, which has been found to be more useful than age, height, or weight alone368 for the measuring the size of the aorta [49], would have possibly led to different statistical results.369 The meta-analysis results also provide valuable insights into the relationship between aortic stiffness measures,370 age, and Marfan syndrome. Firstly, it was observed that without considering age, Marfan patients exhibited higher371 PWV values at all locations, except for the root and the carotid-femoral region. It is particularly interesting to note372 that distinctions between cohorts were challenging when assessing PWV at the carotid-femoral region, which is the373 most commonly used PWV measurement in clinical practice. This observation can be explained by the fact that374 carotid-femoral PWV covers a substantial portion of the aortic tract and may not be sensitive to local variations in375 tissue stiffness. On the other hand, PWV was found to be approximately two times higher at the root. However, the376 root’s relatively small length may lead to difficulties in tracking the foot of the pulse wave, resulting in the considerably377 large standard deviation observed in patients. PWV at the root is rarely utilized in clinical practice and failed to378 distinguish between the two cohorts effectively. Furthermore, a positive correlation was identified between PWV and379 age in both Marfan and control patients across all aortic locations. However, the age-zero projection suggested that380 a statistically significant difference in PWV exists at all locations, except at the root and the carotid-femoral region.381 This notable result implies that age plays a significant role in the evolution of PWV. More importantly, it indicates382 that the difference between Marfan and control patients does not develop with age. Instead, Marfan patients are383 born with higher PWV, indicating stiffer aortic tissues. It is worth highlighting that PWV measurements taken at384 the carotid-femoral region and the root fail to capture these inherent differences.385 Both Marfan and control patients were found to exhibit a negative correlation between distensibility and age386 at all aortic locations. The age-zero projection further emphasized that distensibility is lower, signifying increased387 aortic stiffness in MFS patients, even at a young age. Notably, the slopes of the linear regressions indicated that388 distensibility decreases at a faster rate for the control cohort compared to Marfan patients. It is essential to note389 that this observation might be influenced by the absence of patients past 40 years old in the dataset, introducing a390 potential bias.391 Additionally, our analysis demonstrated that β-stiffness index was positively correlated with age in both Marfan392 and control patients at all aortic locations. Before and after conducting age-zero projection, the results affirmed a393 statistically significant difference in β-stiffness index between Marfan and control patients at all locations. These394 findings suggest that Marfan patients exhibit higher β-SI values even at an early stage of life, indicating stiffer aortic395 tissues compared to controls.396 The results at age-zero projection provide a more reliable assessment of aortic stiffness in Marfan patients and397 underscore the importance of considering age as a confounding factor in such studies. Overall, the data indicates398 that distensibility and β-index are consistently altered in Marfan patients compared to controls, while PWV still399 shows location-specific differences between the two groups before and after age-zero projection.400 21 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted January 7, 2024. ; https://doi.org/10.1101/2024.01.05.24300824doi: medRxiv preprint 4.3 Recommendations for future studies401 Handling missing data poses a significant challenge in systematic quantitative reviews. This review underscores402 the importance of reporting data to facilitate statistically robust and comprehensive meta-analyses. Rather than403 reporting meand and SD, we recommend that studies report individual data points. This could be achieved through404 supplementary dataset if needed. The incorporation of a substantial number of participants and the diverse range of405 study designs greatly contributes to the robustness and depth of the findings presented in this review. However, if406 each patient had reported diameter, age and stiffness measure, it would have allowed for a more thorough statistical407 analysis (such as multivariate regression).408 It is noticeable that there is a lack of studies focusing on specific age ranges and cohorts with more than 40409 patients. For pediatrics, only one multicenter study comprises 608 patients [35]. A plausible explantation could be410 that MFS is rarely diagnosed in paediatric populations since patients do not appear phenotypically different. Not411 all paediatric patients have family history eihter, which makes the diagnosis at an early stage of life challenging.412 Similarly, no studies focused on patients past 40 years old, which represents a turning point in stiffness increase413 according to Wit et al. [27]. This absence could potentially be attributed to the fact that aortic surgeries are414 typically performed before patients reach this age, resulting in limited accessible data for older patients.415 4.4 Limitations416 Regarding limitations, we acknowledge that all pertinent studies may not have been captured in our search. Our417 search criteria might have missed studies that were not explicitly categorized under, or did not explicitly reference a418 Marfan syndrome diagnosis. Since Marfan syndrome can be misdiagnosed for other connective tissue disorders caused419 by pathogenic variants in genes other than FBN1, studies published before the revised Ghent criteria were excluded.420 However, they likely include true Marfan patients and provide valuable insights. Similarly, mild Marfan syndrome421 cases and cases of suspected but not yet verified Marfan syndrome, were possibly overlooked in this present work.422 Despite the revised Ghent criteria, Marfan syndrome clinically overlaps with other connective tissue disorders, such as423 Loeys-Dietz syndrome, and distinguishing them is challenging in the absence of a molecular diagnosis. Consequently,424 individuals with mutations in proteins related to the TGF- β pathway might receive a Marfan syndrome diagnosis425 against the Ghent nosology and be included in this study.426 While this work does touch upon aortic diameter, it is important to note that a significant portion of the existing427 literature primarily focuses on aortic diameter in Marfan patients, but these studies were not encompassed in this428 review. Instead, discussions concerning diameter in this review are derived exclusively from data within the selected429 papers that primarily address aortic stiffness measures.430 The aortic sites and regions were not defined identically between the selected papers. To facilitate the reporting431 of results and minimize potential inconsistencies, we made efforts to categorize them into five main regions: Root,432 Aao, Arch, Dao and Carotid-femoral, although minor variations may remain. The data gap in specific age brackets433 is an essential consideration when interpreting the findings of this review. The review’s strength lies in its ability to434 compile and analyze aortic stiffness in different populations. However, the absence of age-specific studies, especially435 in the pediatric and older age groups, highlights a potential area for future research.436 Concerning the meta-analysis, despite our efforts to address missing data, several factors introduced bias into the437 statistical tests. This bias stems from theoretical conversion equations, which cannot precisely mimic physiological438 behaviour, as well as the absence of individual data points and the lack of age-associated values for each data point.439 Similarly, conducting an age projection involves making inferences beyond the data range, which may not align440 with in-vivo behaviour. The regression analysis with age was conducted solely on data that included individual data441 points with corresponding age. Additionally, the predictive power of aortic stiffness measures could not be thoroughly442 investigated in the meta-analysis due to the absence of articles reporting individual values for patients followed at443 various ages. Lastly, examining diameter as a potential confounding variable was unfeasible because of the scarcity444 of studies reporting individual data points for stiffness measures, alongside corresponding age and diameter values.445 4.5 Conclusion446 To the best of the authors’ knowledge, this is the first systematic review and meta analysis investigating how in-447 vivo aortic stiffness measures can be early markers of aortic disease in Marfan syndrome, and their prediction of448 aortic dilatation. Our study emphasizes the importance of using a combination of parameters, including diameter449 22 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted January 7, 2024. ; https://doi.org/10.1101/2024.01.05.24300824doi: medRxiv preprint measurements and stiffness indices, to obtain a more comprehensive evaluation of aortic disease in MFS patients.450 This approach can provide a deeper understanding of disease progression and assist clinical decision making.451 Funding452 This study was partially funded by the Engineering and Physical Sciences Research Council (EP/N02124X/1) and453 University of Glasgow’s College of Science and Engineering via PhD studentship.454 Data availability statement455 All data produced in the present study are available upon reasonable request to the authors456 23 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted January 7, 2024. ; https://doi.org/10.1101/2024.01.05.24300824doi: medRxiv preprint References457 [1] Stephen C Mockrin. Molecular Genetics & Gene Therapy of Cardiovascular Diseases . CRC Press, 1996. doi:458 https://doi.org/10.1201/9781482273243.459 [2] Goetz C. Mueller, Veronika Stark, Kristoffer Steiner, Yskert von Kodolitsch, Meike Rybczynski, Jochen Weil,460 and Thomas S. Mir. Impact of Age and Gender on Cardiac Pathology in Children and Adolescents With Marfan461 Syndrome. Pediatric Cardiology, 34(4):991–998, April 2013. ISSN 1432-1971. doi: 10.1007/s00246-012-0593-0.462 URL https://doi.org/10.1007/s00246-012-0593-0.463 [3] Eric M. Isselbacher, Ourania Preventza, James Hamilton Black, John G. Augoustides, Adam W. Beck,464 Michael A. Bolen, Alan C. Braverman, Bruce E. Bray, Maya M. Brown-Zimmerman, Edward P. Chen, Ty-465 rone J. Collins, Abe DeAnda, Christina L. Fanola, Leonard N. Girardi, Caitlin W. Hicks, Dawn S. Hui, William466 Schuyler Jones, Vidyasagar Kalahasti, Karen M. Kim, Dianna M. Milewicz, Gustavo S. Oderich, Laura Og-467 bechie, Susan B. Promes, Elsie Gyang Ross, Marc L. Schermerhorn, Sabrina Singleton Times, Elaine E. Tseng,468 Grace J. Wang, and Y. Joseph Woo. 2022 acc/aha guideline for the diagnosis and management of aortic469 disease: A report of the american heart association/american college of cardiology joint committee on clin-470 ical practice guidelines. Circulation, 146(24):e334–e482, 2022. doi: 10.1161/CIR.0000000000001106. URL471 https://www.ahajournals.org/doi/abs/10.1161/CIR.0000000000001106.472 [4] M. Groenink, T. A. Lohuis, J. G. Tijssen, M. S. Naeff, R. C. Hennekam, E. E. van der Wall, and B. J. Mulder.473 Survival and complication free survival in Marfan’s syndrome: implications of current guidelines. Heart (British474 Cardiac Society), 82(4):499–504, October 1999. ISSN 1468-201X. doi: 10.1136/hrt.82.4.499.475 [5] Vincent L. Gott, Peter S. Greene, Diane E. Alejo, Duke E. Cameron, David C. Naftel, D. Craig Miller,476 A. Marc Gillinov, John C. Laschinger, Hans G. Borst, Christian E.A. Cabrol, Denton A. Cooley, Joseph S.477 Coselli, Tirone E. David, Randall B. Griepp, Nicholas T. Kouchoukos, Marko I. Turina, and Reed E.478 Pyeritz. Replacement of the Aortic Root in Patients with Marfan’s Syndrome. New England Jour-479 nal of Medicine , 340(17):1307–1313, April 1999. ISSN 0028-4793. doi: 10.1056/NEJM199904293401702.480 URL https://doi.org/10.1056/NEJM199904293401702. Publisher: Massachusetts Medical Society eprint:481 https://doi.org/10.1056/NEJM199904293401702.482 [6] Lars G. Svensson and Lev Khitin. Aortic cross-sectional area/height ratio timing of aortic surgery in asymp-483 tomatic patients with Marfan syndrome. The Journal of Thoracic and Cardiovascular Surgery , 123(2):360–361,484 February 2002. ISSN 0022-5223, 1097-685X. doi: 10.1067/mtc.2002.118497. URL https://www.jtcvs.org/485 article/S0022-5223(02)13008-X/fulltext. Publisher: Elsevier.486 [7] Matthew J. Page, Joanne E. McKenzie, Patrick M. Bossuyt, Isabelle Boutron, Tammy C. Hoffmann, Cynthia D.487 Mulrow, Larissa Shamseer, Jennifer M. Tetzlaff, Elie A. Akl, Sue E. Brennan, Roger Chou, Julie Glanville,488 Jeremy M. Grimshaw, Asbjørn Hr´ objartsson, Manoj M. Lalu, Tianjing Li, Elizabeth W. Loder, Evan Mayo-489 Wilson, Steve McDonald, Luke A. McGuinness, Lesley A. Stewart, James Thomas, Andrea C. Tricco, Vivian A.490 Welch, Penny Whiting, and David Moher. The PRISMA 2020 statement: an updated guideline for reporting491 systematic reviews. BMJ, 372:n71, March 2021. ISSN 1756-1833. doi: 10.1136/bmj.n71. URL https://492 www.bmj.com/content/372/bmj.n71. Publisher: British Medical Journal Publishing Group Section: Research493

& Reporting.494 [8] A. De Paepe, R. B. Devereux, H. C. Dietz, R. C. Hennekam, and R. E. Pyeritz. Revised diagnostic criteria for495 the Marfan syndrome. American Journal of Medical Genetics, 62(4):417–426, April 1996. ISSN 0148-7299. doi:496 10.1002/(SICI)1096-8628(19960424)62:4⟨417::AID-AJMG15⟩3.0.CO;2-R.497 [9] Ankit Rohatgi. Webplotdigitizer: Version 4.6, 2022. URL https://automeris.io/WebPlotDigitizer.498 [10] Taha Alhalimi, Jisok Lim, Drew Gourley, and Hirofumi Tanaka. Converting and Standardizing Various Measures499 of Arterial Stiffness to Pulse Wave Velocity. Pulse, 9(3-4):72–82, 2021. ISSN 2235-8676, 2235-8668. doi: 10.500 1159/000517872. URL https://www.karger.com/Article/FullText/517872. Publisher: Karger Publishers.501 [11] J. Crighton Bramwell and Archibald Vivian Hill. The velocity of pulse wave in man. Proceedings of the Royal502 Society of London. Series B, Containing Papers of a Biological Character , 93(652):298–306, February 1922.503 24 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted January 7, 2024. ; https://doi.org/10.1101/2024.01.05.24300824doi: medRxiv preprint doi: 10.1098/rspb.1922.0022. URL https://royalsocietypublishing.org/doi/10.1098/rspb.1922.0022.504 Publisher: Royal Society.505 [12] Christopher J. Weir, Isabella Butcher, Valentina Assi, Stephanie C. Lewis, Gordon D. Murray, Peter Langhorne,506 and Marian C. Brady. Dealing with missing standard deviation and mean values in meta-analysis of continuous507 outcomes: a systematic review. BMC Medical Research Methodology , 18(1):25, March 2018. ISSN 1471-2288.508 doi: 10.1186/s12874-018-0483-0. URL https://doi.org/10.1186/s12874-018-0483-0.509 [13] Xiang Wan, Wenqian Wang, Jiming Liu, and Tiejun Tong. Estimating the sample mean and standard deviation510 from the sample size, median, range and/or interquartile range. BMC Medical Research Methodology , 14(1):511 135, December 2014. ISSN 1471-2288. doi: 10.1186/1471-2288-14-135. URL https://doi.org/10.1186/512 1471-2288-14-135.513 [14] Ashwin Prakash, Himanshu Adlakha, Nicole Rabideau, Cara J. Hass, Shaine A. Morris, Tal Geva, Kimberlee514 Gauvreau, Michael N. Singh, and Ronald V. Lacro. Segmental Aortic Stiffness in Children and Young Adults515 With Connective Tissue Disorders: Relationships With Age, Aortic Size, Rate of Dilation, and Surgical Root516 Replacement. Circulation, 132(7):595–602, August 2015. ISSN 1524-4539. doi: 10.1161/CIRCULATIONAHA.517 114.014934.518 [15] Maarten Groenink, Albert de Roos, Barbara J. M. Mulder, Ben Verbeeten, Janneke Timmermans, Aeilko H.519 Zwinderman, Jos A. E. Spaan, and Ernst E. van der Wall. Biophysical Properties of the Normal-sized Aorta520 in Patients with Marfan Syndrome: Evaluation with MR Flow Mapping. Radiology, 219(2):535–540, May521 2001. ISSN 0033-8419. doi: 10.1148/radiology.219.2.r01ma01535. URL https://pubs.rsna.org/doi/10.522 1148/radiology.219.2.r01ma01535. Publisher: Radiological Society of North America.523 [16] George G. S Sandor, Takashi Hishitani, Ross E Petty, Mary T Potts, Astrid DeSouza, Eustace DeSouza, and524 James E Potts. A novel doppler echocardiographic method of measuring the biophysical properties of the aorta525 in pediatric patients. Journal of the American Society of Echocardiography , 16(7):745–750, July 2003. ISSN526 0894-7317. doi: 10.1016/S0894-7317(03)00407-3. URL https://www.sciencedirect.com/science/article/527 pii/S0894731703004073.528 [17] Gijs J. Nollen, Maarten Groenink, Jan G.P. Tijssen, Ernst E. van der Wall, and Barbara J.M. Mulder. Aortic529 stiffness and diameter predict progressive aortic dilatation in patients with Marfan syndrome. European Heart530 Journal, 25(13):1146–1152, July 2004. ISSN 0195-668X. doi: 10.1016/j.ehj.2004.04.033. URL https://doi.531 org/10.1016/j.ehj.2004.04.033.532 [18] Daniela Baumgartner, Christian Baumgartner, Gabor M´ aty´ as, Beat Steinmann, Judith L¨ offler-Ragg, Elisabeth533 Schermer, Ulrich Schweigmann, Ivo Baldissera, Bernhard Frischhut, John Hess, and Ignaz Hammerer. Diagnostic534 power of aortic elastic properties in young patients with Marfan syndrome. The Journal of Thoracic and535 Cardiovascular Surgery, 129(4):730–739, April 2005. ISSN 0022-5223. doi: 10.1016/j.jtcvs.2004.07.019. URL536 https://www.sciencedirect.com/science/article/pii/S0022522304009997.537 [19] Thomas Oosterhof, Gijs J. Nollen, Ernst E. van der Wall, Anje M. Spijkerboer, Jana Hrudova, Berto J.538 Bouma, Marcel G. W. Dijkgraaf, and Barbara J. M. Mulder. Comparison of aortic stiffness in patients with539 juvenile forms of ascending aortic dilatation with versus without Marfan’s syndrome. The American Jour-540 nal of Cardiology , 95(8):996–998, April 2005. ISSN 0002-9149. doi: 10.1016/j.amjcard.2004.12.047. URL541 https://www.sciencedirect.com/science/article/pii/S000291490500130X.542 [20] Timothy J. Bradley, James E. Potts, Mary T. Potts, Astrid M. DeSouza, and George G. S. Sandor. Echocar-543 diographic Doppler Assessment of the Biophysical Properties of the Aorta in Pediatric Patients With the544 Marfan Syndrome. The American Journal of Cardiology , 96(9):1317–1321, November 2005. ISSN 0002-545 9149. doi: 10.1016/j.amjcard.2005.06.080. URL https://www.sciencedirect.com/science/article/pii/546 S0002914905012798.547 [21] Antonio Vitarelli, Ysabel Conde, Ester Cimino, Ilaria D’Angeli, Simona D’Orazio, Simona Stellato, Viviana548 Padella, and Fiorella Caranci. Aortic Wall Mechanics in the Marfan Syndrome Assessed by Transesophageal549 Tissue Doppler Echocardiography. The American Journal of Cardiology, 97(4):571–577, February 2006. ISSN550 0002-9149. doi: 10.1016/j.amjcard.2005.09.089. URL https://www.sciencedirect.com/science/article/551 pii/S0002914905019120.552 25 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted January 7, 2024. ; https://doi.org/10.1101/2024.01.05.24300824doi: medRxiv preprint [22] Daniela Baumgartner, Christian Baumgartner, Elisabeth Schermer, Georg Engl, Ulrich Schweigmann, Gabor553 M´ aty´ as, Beat Steinmann, and J¨ org Ingolf Stein. Different patterns of aortic wall elasticity in patients with554 Marfan syndrome: A noninvasive follow-up study. The Journal of Thoracic and Cardiovascular Surgery , 132(4):555 811–819, October 2006. ISSN 0022-5223. doi: 10.1016/j.jtcvs.2006.07.001. URL https://www.sciencedirect.556 com/science/article/pii/S0022522306011810.557 [23] Rossella Fattori, Letizia Bacchi Reggiani, Guglielmina Pepe, Gabriella Napoli, Claudio Bna, Francesca Celletti,558 Luigi Lovato, and Giampaolo Gavelli. Magnetic Resonance Imaging Evaluation of Aortic Elastic Properties559 as Early Expression of Marfan Syndrome. Journal of Cardiovascular Magnetic Resonance , July 2009. URL560 https://www.tandfonline.com/doi/abs/10.3109/10976640009148688. Publisher: Taylor & Francis.561 [24] Kai Mortensen, Muhammet A. Aydin, Meike Rybczynski, Johannes Baulmann, Nazila Abdul Schahidi, Georgina562 Kean, Kristine K¨ uhne, Alexander M.J. Bernhardt, Olaf Franzen, Thomas Mir, Christian Habermann, Dietmar563 Koschyk, Rodolfo Ventura, Stephan Willems, Peter N. Robinson, J¨ urgen Berger, Hermann Reichenspurner,564 Thomas Meinertz, and Yskert von Kodolitsch. Augmentation Index Relates to Progression of Aortic Disease565 in Adults With Marfan Syndrome. American Journal of Hypertension , 22(9):971–979, September 2009. ISSN566 0895-7061. doi: 10.1038/ajh.2009.115. URL https://doi.org/10.1038/ajh.2009.115.567 [25] Anatoli Kiotsekoglou, James C. Moggridge, Samir K. Saha, Venediktos Kapetanakis, Malini Govindan,568 Francisco Alpendurada, Michael J. Mullen, John Camm, George R. Sutherland, Bart H. Bijnens, and569 Anne H. Child. Assessment of Aortic Stiffness in Marfan Syndrome Using Two-Dimensional and Doppler570 Echocardiography. Echocardiography, 28(1):29–37, 2011. ISSN 1540-8175. doi: 10.1111/j.1540-8175.2010.571 01241.x. URL https://onlinelibrary.wiley.com/doi/abs/10.1111/j.1540-8175.2010.01241.x. eprint:572 https://onlinelibrary.wiley.com/doi/pdf/10.1111/j.1540-8175.2010.01241.x.573 [26] Jos J.M. Westenberg, Arthur J.H.A. Scholte, Zuzana Vaskova, Rob J. van der Geest, Maarten Groenink, Gerda574 Labadie, Pieter J. van den Boogaard, Teodora Radonic, Yvonne Hilhorst-Hofstee, Barbara J.M. Mulder, Lu-575 cia J.M. Kroft, Johan H.C. Reiber, and Albert de Roos. Age-related and regional changes of aortic stiffness576 in the marfan syndrome: Assessment with velocity-encoded MRI. Journal of Magnetic Resonance Imaging, 34577 (3):526–531, 2011. ISSN 1522-2586. doi: 10.1002/jmri.22646. URL https://onlinelibrary.wiley.com/doi/578 abs/10.1002/jmri.22646. eprint: https://onlinelibrary.wiley.com/doi/pdf/10.1002/jmri.22646.579 [27] Aletta de Wit, Kelly Vis, and Richmond W. Jeremy. Aortic Stiffness in Heritable Aortopathies: Relationship to580 Aneurysm Growth Rate. Heart, Lung and Circulation, 22(1):3–11, January 2013. ISSN 1443-9506, 1444-2892.581 doi: 10.1016/j.hlc.2012.08.049. URL https://www.heartlungcirc.org/article/S1443-9506(12)01256-5/582 fulltext. Publisher: Elsevier.583 [28] Eleanore S. J. Kr¨ oner, Arthur J. H. A. Scholte, Patrick J. H. de Koning, Pieter J. van den Boogaard, Lucia584 J. M. Kroft, Rob J. van der Geest, Yvonne Hilhorst-Hofstee, Hildo J. Lamb, Hans-Marc J. Siebelink, Barbara585 J. M. Mulder, Maarten Groenink, Teodora Radonic, Ernst E. van der Wall, Albert de Roos, Johan H. C.586 Reiber, and Jos J. M. Westenberg. MRI-assessed regional pulse wave velocity for predicting absence of regional587 aorta luminal growth in marfan syndrome. International Journal of Cardiology , 167(6):2977–2982, September588 2013. ISSN 0167-5273. doi: 10.1016/j.ijcard.2012.08.057. URL https://www.sciencedirect.com/science/589 article/pii/S0167527312010868.590 [29] Gisela Teixido-Tura, Alban Redheuil, Jose Rodr´ ıguez-Palomares, Laura Guti´ errez, Violeta S´ anchez, Alberto591 Forteza, Joao A. C. Lima, David Garc´ ıa-Dorado, and Artur Evangelista. Aortic biomechanics by magnetic592 resonance: Early markers of aortic disease in Marfan syndrome regardless of aortic dilatation? International593 Journal of Cardiology, 171(1):56–61, January 2014. ISSN 0167-5273. doi: 10.1016/j.ijcard.2013.11.044. URL594 https://www.sciencedirect.com/science/article/pii/S0167527313020627.595 [30] Yohei Akazawa, Noriko Motoki, Akira Tada, Shoko Yamazaki, Akira Hachiya, Satoshi Matsuzaki, Motoko596 Kamiya, Tomohiko Nakamura, Tomoki Kosho, and Yuji Inaba. Decreased Aortic Elasticity in Children With597 Marfan Syndrome or Loeys-Dietz Syndrome. Circulation Journal, 80(11):2369–2375, 2016. doi: 10.1253/circj.598 CJ-16-0739.599 26 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted January 7, 2024. ; https://doi.org/10.1101/2024.01.05.24300824doi: medRxiv preprint [31] S. D. Singh, X. Y. Xu, N. B. Wood, J. R. Pepper, C. Izgi, T. Treasure, and R. H. Mohiaddin. Aortic flow600 patterns before and after personalised external aortic root support implantation in Marfan patients. Journal601 of Biomechanics , 49(1):100–111, January 2016. ISSN 0021-9290. doi: 10.1016/j.jbiomech.2015.11.040. URL602 https://www.sciencedirect.com/science/article/pii/S002192901500682X.603 [32] Anthony Merlocco, Ronald V. Lacro, Kimberlee Gauvreau, Nicole Rabideau, Michael N. Singh, and Ashwin604 Prakash. Longitudinal Changes in Segmental Aortic Stiffness Determined by Cardiac Magnetic Resonance in605 Children and Young Adults With Connective Tissue Disorders (the Marfan, Loeys-Dietz, and Ehlers-Danlos606 Syndromes, and Nonspecific Connective Tissue Disorders). The American Journal of Cardiology , 120(7):1214–607 1219, October 2017. ISSN 0002-9149. doi: 10.1016/j.amjcard.2017.06.064. URL https://www.sciencedirect.608 com/science/article/pii/S0002914917311268.609 [33] Andrea Grillo, Paolo Salvi, Susan Marelli, Lan Gao, Lucia Salvi, Andrea Faini, Giuliana Trifir` o, Renzo Carretta,610 Alessandro Pini, and Gianfranco Parati. Impaired Central Pulsatile Hemodynamics in Children and Adolescents611 With Marfan Syndrome. Journal of the American Heart Association , 6(11):e006815, November 2017. ISSN 2047-612 9980. doi: 10.1161/JAHA.117.006815. URL https://www.ahajournals.org/doi/10.1161/JAHA.117.006815.613 [34] Paolo Salvi, Andrea Grillo, Susan Marelli, Lan Gao, Lucia Salvi, Maurizio Viecca, Anna Maria Di Blasio,614 Renzo Carretta, Alessandro Pini, and Gianfranco Parati. Aortic dilatation in Marfan syndrome: role of arterial615 stiffness and fibrillin-1 variants. Journal of Hypertension, 36(1):77–84, January 2018. ISSN 1473-5598. doi:616 10.1097/HJH.0000000000001512.617 [35] Elif Seda Selamet Tierney, Jami C. Levine, Lynn A. Sleeper, Mary J. Roman, Timothy J. Bradley, Steven D.618 Colan, Shan Chen, M. Jay Campbell, Meryl S. Cohen, Julie De Backer, Haleh Heydarian, Arvind Hoskoppal,619 Wyman W. Lai, Aimee Liou, Edward Marcus, Arni Nutting, Aaron K. Olson, David A. Parra, Gail D. Pearson,620 Mary Ella Pierpont, Beth F. Printz, Reed E. Pyeritz, William Ravekes, Angela M. Sharkey, Shubhika Srivastava,621 Luciana Young, and Ronald V. Lacro. Influence of Aortic Stiffness on Aortic-Root Growth Rate and Outcome in622 Patients With the Marfan Syndrome. The American Journal of Cardiology, 121(9):1094–1101, May 2018. ISSN623 0002-9149. doi: 10.1016/j.amjcard.2018.01.016. URL https://www.sciencedirect.com/science/article/624 pii/S0002914918301589.625 [36] Michal Sch¨ afer, Lorna P Browne, Uyen Truong, James J Jaggers, Max B Mitchell, Ladonna Malone, Gareth626 Morgan, Kathryn Chatfield, Daniel McLennan, Harma Turbendian, Danial Vargas, Brian Fonseca, Michael627 DiMaria, Anar Shah, Margaret P Ivy, Alex J Barker, Kendall S Hunter, Neil Wilson, D Dunbar Ivy, and628 David N Campbell. Aortic stiffness in adolescent Turner and Marfan syndrome patients. European Journal of629 Cardio-Thoracic Surgery, 54(5):926–932, November 2018. ISSN 1010-7940. doi: 10.1093/ejcts/ezy168. URL630 https://doi.org/10.1093/ejcts/ezy168.631 [37] Andrea Guala, Gisele Teixido-Tura, Jose Rodriguez-Palomares, Aroa Ruiz-Munioz, Lysia Dux-Santoy, Nicolas632 Villalva, Chiara Granato, Laura Galian, Laura Gutierrez, Teresa Gonzalez-Alujas, Violeta Sanchez, Alberto633 Forteza, David Garcia-Dorado, and Artur Evangelista. Proximal aorta longitudinal strain predicts aortic root634 dilation rate and aortic events in Marfan syndrome. European heart journal, 2019. URL https://doi.org/10.635 1093/eurheartj/ehz191.636 [38] Junfeng Yan, Ann-Cathrin Lehsau, Benjamin Sauer, Barbara Pieper, Salah A. Mohamed, Bart L. Loeys,637 Harry C. Dietz, Lut Van Laer, Andrew S. McCallion, Per Eriksson, Anders Franco-Cereceda, Luc Mertens,638 Seema Mital, Salah A. Mohamed, and Gregor Andelfinger. Comparison of biomechanical properties in as-639 cending aortic aneurysms of patients with congenital bicuspid aortic valve and Marfan syndrome. Interna-640 tional Journal of Cardiology , 278:65–69, March 2019. ISSN 0167-5273. doi: 10.1016/j.ijcard.2018.11.102. URL641 https://www.sciencedirect.com/science/article/pii/S016752731735074X.642 [39] Andrea Guala, Jose Rodriguez-Palomares, Lydia Dux-Santoy, Gisela Teixido-Tura, Giuliana Maldonado, Laura643 Galian, Marina Huguet, Filipa Valente, Laura Guti´ errez, Teresa Gonz´ alez-Alujas, Kevin M. Johnson, Oliver644 Wieben, Augusto Sao Avil´ es, David Garcia-Dorado, and Arturo Evangelista. Influence of Aortic Dilation on the645 Regional Aortic Stiffness of Bicuspid Aortic Valve Assessed by 4-Dimensional Flow Cardiac Magnetic Resonance:646 Comparison With Marfan Syndrome and Degenerative Aortic Aneurysm. JACC: Cardiovascular Imaging, 12(6):647 1020–1029, June 2019. ISSN 1936-878X. doi: 10.1016/j.jcmg.2018.03.017. URL https://www.sciencedirect.648 com/science/article/pii/S1936878X18303000.649 27 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted January 7, 2024. ; https://doi.org/10.1101/2024.01.05.24300824doi: medRxiv preprint [40] Jason Z. Cui, Kevin C. Harris, Koen Raedschelders, Zsuzsanna Hollander, James E. Potts, Astrid De Souza,650 Marla Kiess, Bruce M. McManus, Pascal Bernatchez, Leslie A. Raffin, Heidi Paine, Cornelis van Breemen,651 George G. S. Sandor, and Mitra Esfandiarei. Aortic Dimensions, Biophysical Properties, and Plasma Biomarkers652 in Children and Adults with Marfan or Loeys-Dietz Syndrome. CJC Open, 3(5):585–594, May 2021. ISSN653 2589-790X. doi: 10.1016/j.cjco.2020.12.018. URL https://www.sciencedirect.com/science/article/pii/654 S2589790X20302389.655 [41] Mitzi M. van Andel, Vivian de Waard, Janneke Timmermans, Arthur J. H. A. Scholte, Maarten P. van den656 Berg, Aeilko H. Zwinderman, Barbara J. M. Mulder, and Maarten Groenink. Aortic distensibility in Marfan657 syndrome: a potential predictor of aortic events? Open Heart, 8(2):e001775, October 2021. ISSN 2053-3624. doi:658 10.1136/openhrt-2021-001775. URL https://openheart.bmj.com/content/8/2/e001775. Publisher: Archives659 of Disease in childhood Section: Aortic and vascular disease.660 [42] Constance G. Weismann, Joanna Hlebowicz, Anna ˚Akesson, Petru Liuba, and Katarina Hanseus. Comprehensive661 Characterization of Arterial and Cardiac Function in Marfan Syndrome—Can Biomarkers Help Improve Out-662 come? Frontiers in Physiology , 13, 2022. ISSN 1664-042X. URL https://www.frontiersin.org/articles/663 10.3389/fphys.2022.873373.664 [43] Kelly Cox, Yousi A. Oquendo, David Liang, and Elif Seda Selamet Tierney. Aortic root dilation in adult665 patients with Marfan syndrome: Does aortic root stiffness matter? JTCVS Open, 10:113–120, June 2022. ISSN666 2666-2736. doi: 10.1016/j.xjon.2022.04.011. URL https://www.sciencedirect.com/science/article/pii/667 S2666273622001632.668 [44] Tomoaki Morioka, Katsuhito Mori, and Masanori Emoto. Is stiffness parameter β useful for the evaluation of669 atherosclerosis?˜ its clinical implications, limitations, and future perspectives ˜. Journal of Atherosclerosis and670 Thrombosis, 28(5):435–453, 2021. ISSN 1340-3478. doi: 10.5551/jat.RV17047. URL https://www.ncbi.nlm.671 nih.gov/pmc/articles/PMC8193788/.672 [45] Takashi Wada, Kentaro Fujishiro, Tsutomu Fukumoto, Sayaka Yamazaki, and Takashi Wada. Relationship673 between ultrasound assessment of arterial wall properties and blood pressure. Angiology, 48(10):893–900, 1997.674 ISSN 0003-3197. doi: 10.1177/000331979704801006. URL https://doi.org/10.1177/000331979704801006.675 Publisher: SAGE Publications Inc.676 [46] Jun Sugawara, Hidehiko Komine, Koichiro Hayashi, Mutsuko Yoshizawa, Takashi Yokoi, Takeshi Otsuki, Nobu-677 take Shimojo, Takashi Miyauchi, Seiji Maeda, and Hirofumi Tanaka. Effect of systemic nitric oxide synthase678 inhibition on arterial stiffness in humans. Hypertension Research, 30(5):411–415, 2007. ISSN 1348-4214. doi:679 10.1291/hypres.30.411. URL https://www.nature.com/articles/hr200755. Number: 5 Publisher: Nature680 Publishing Group.681 [47] Andrea Guala, Gisele Teixido-Tura, L. Dux-Santoy, C. Granato, A. Ruiz-Munioz, F. Valente, L Galian-Gay,682 L Gutierrez, T Gonzalez-Alujas, K. M. Johnson, O. Wieben, A. Sao Aviles, A. Evangelista, and J. Rodriguez-683 Palomares. Decreased rotational flow and circumferential wall shear stress as early markers of descending aorta684 dilation in Marfan syndrome: a 4D flow CMR study. Journal of Cardiovascular Magnetic Resonance, 21:1–11,685 2019. URL https://doi.org/10.1186/s12968-019-0572-1.686 [48] Tirone E. David. Aortic Valve Sparing in Different Aortic Valve and Aortic Root Conditions. Journal687 of the American College of Cardiology, 68(6):654–664, August 2016. doi: 10.1016/j.jacc.2016.04.062. URL688 https://www.jacc.org/doi/abs/10.1016/j.jacc.2016.04.062. Publisher: American College of Cardiology689 Foundation.690 [49] Thierry Sluysmans and Steven D. Colan. Theoretical and empirical derivation of cardiovascular allometric691 relationships in children. Journal of Applied Physiology, 99(2):445–457, 2005. ISSN 8750-7587. doi: 10.692 1152/japplphysiol.01144.2004. URL https://journals.physiology.org/doi/full/10.1152/japplphysiol.693 01144.2004. Publisher: American Physiological Society.694 28 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted January 7, 2024. ; https://doi.org/10.1101/2024.01.05.24300824doi: medRxiv preprint