Article Text

Download PDFPDF

Original research article
Calcification and extracellular matrix dysregulation in human postmortem and surgical aortic valves
  1. M Victoria Gomez-Stallons1,
  2. Justin T Tretter2,
  3. Keira Hassel1,
  4. Osniel Gonzalez-Ramos1,
  5. Dorothy Amofa1,
  6. Nicholas J Ollberding3,
  7. Wojciech Mazur4,
  8. Joseph K Choo4,
  9. J Michael Smith5,
  10. Dean J Kereiakes4,
  11. Katherine E Yutzey1
  1. 1 Molecular Cardiovascular Biology, Cincinnati Children’s Hospital Medical Center, Cincinnati, Ohio, USA
  2. 2 Division of Cardiology, Cincinnati Children’s Hospital Medical Center, Cincinnati, Ohio, USA
  3. 3 Division of Biostatistics and Epidemiology, Cincinnati Children’s Hospital Medical Center, Cincinnati, Ohio, USA
  4. 4 The Ohio Heart and Vascular Center, The Christ Hospital, Cincinnati, Ohio, USA
  5. 5 Cardiac Surgery, TriHealth, Cincinnati, Ohio, USA
  1. Correspondence to Dr Katherine E Yutzey, Division of Molecular Cardiovascular Biology, Cincinnati Children’s Medical Center, Cincinnati, OH 45229, USA; katherine.yutzey{at}cchmc.org

Abstract

Objectives Calcific aortic valve disease (CAVD) is a progressive disease ranging from aortic valve (AoV) sclerosis to AoV stenosis (AS), characterised by severe calcification with impaired leaflet function. Due to the lack of early symptoms, the pathological progression towards valve dysfunction is poorly understood. The early patterns of AoV calcification and altered extracellular matrix (ECM) organisation were analysed in individuals postmortem without clinical AS compared with clinical AS.

Methods Histological patterns of calcification and ECM organisation in postmortem AoV leaflets without clinical AS obtained from a tissue repository and surgical specimens obtained from individuals with clinical AS were compared with in vivo imaging prior to transcatheter AoV implantation.

Results AoV calcification was detected in all samples from individuals >50 years old, with severity increasing with age, independent of known CAVD risk factors. Two distinct types of calcification were identified: ‘Intrinsic’, primarily found at the leaflet hinge of postmortem leaflets, accompanied by abnormal collagen and proteoglycan deposition; and ‘Nodular’, extending from the middle to the tip regions in more severely affected postmortem leaflets and surgical specimens, associated with increased elastin fragmentation and loss of elastin integrity. Even in the absence of increased thickening, abnormalities in ECM composition were observed in postmortem leaflets without clinical AS and worsen in clinical AS.

Conclusions Two distinct phenotypes of AoV calcification are apparent. While the ‘nodular’ form is recognised on in vivo imaging and is present with CAVD and valve dysfunction, it is unclear if the ‘intrinsic’ form is pathological or detected on in vivo imaging.

  • valvular heart disease
  • aortic stenosis
  • transcatheter valve interventions
  • valve disease surgery
  • cardiac computer tomographic (ct) imaging

Statistics from Altmetric.com

Introduction

Calcific aortic valve disease (CAVD) is a progressive disease that ranges from aortic valve (AoV) sclerosis, characterised by mild AoV thickening and/or calcification without obstruction of blood flow, to AoV stenosis (AS) marked by more severe calcification with impaired leaflet function.1–3 Clinically, these stages are defined by valve anatomy, left ventricular function and patient symptoms.4 However, due to the slow progression of the disease,5 6 symptoms are not usually recognised until severe valve obstruction, resulting in late diagnosis. Consequently, inaccessibility of human samples prior to diagnosis has limited the studies of early disease development. We hypothesise that AoV calcification and associated changes to the extracellular matrix (ECM) occur in human AoV leaflets prior to overt symptoms and diagnosis.

CAVD progression is an active process with multiple pathogenic mechanisms.2 7 8 However, there are no pharmacological-based therapies that prevent or inhibit CAVD progression. Currently, improvement of patient outcomes relies on early assessment of risk factors, severity of AS and timing of AoV replacement surgery.9 10 Understanding how early changes to AoV leaflets and associated ECM relate to disease progression is critical to identifying patients who will progress to valve dysfunction. Histological analyses were performed in postmortem whole AoV from individuals without a clinical diagnosis of AS (‘No Clinical AS’). These were compared with histology and in vivo CT imaging of calcification in stenotic AoV at the time of replacement (‘Clinical AS’) to identify critical AoV changes prior to the development of dysfunction and disease.

Methods

Additional information on materials and methods can be found in online supplementary file.

Supplemental material

Human AoV tissue collection

Postmortem whole human hearts were procured from the National Disease Research Interchange and Restore Life USA (Elizabethton, Tennessee). Whole heart specimens (n=23) with intact AoV leaflets and supporting structures were examined. Notably, these donors had no history of AoV disease diagnosis (with one exception), enabling the study of early stages of AoV disease prior to clinical diagnosis. In addition, surgical specimens (n=29) were obtained from patients diagnosed with symptomatic severe AS undergoing AoV replacement surgery at TriHealth, Cincinnati. These specimens were received in pieces after surgery, without identification of anatomical origins, preventing analysis by leaflet and leaflet regions. Additional demographic and comorbidity information is provided in online supplementary tables 1–3.

Histological analysis of human AoV calcification and ECM

Alizarin red staining was used to detect the presence of calcification in AoV leaflet sections, as previously described.11 For postmortem AoV samples, the right coronary (RCL), non-coronary (NCL) and left coronary (LCL) leaflets were analysed separately, and each leaflet was further subdivided into three equal leaflet regions: hinge (including aortic root but excluding the adjacent aortic wall, determined by pentachrome staining), middle and tip.12 Analysis of the extent of AoV calcification was performed based on a previously published quantification system.13 14 Each AoV leaflet sample was assigned an average qualitative calcification grade value (0–3). For calcification analysis of surgical samples, Alizarin red staining was used to detect intrinsic versus nodular calcification patterns. No calcification grade was calculated for these samples. Movat’s pentachrome staining (American MasterTech) of human AoV leaflets was used to determine ECM composition relative to AoV thickening and calcification.

Study population for pre-TAVI CT calcification analysis

CAVD with severe AS was qualitatively evaluated by retrospective analysis of CT angiographic data sets from 104 consecutive patients considered for transcatheter aortic valve implantation (TAVI) from January 2016 to September 2017 at The Christ Hospital. Demographic and comorbidity information is provided in online supplementary table 3.

Statistical analysis

Means with SD and frequencies with per cents were used to describe patient demographic and clinical characteristics. Non-parametric Kruskal-Wallis and Fisher’s exact tests were used to test for differences in patient demographic and clinical characteristics according to heart disease diagnosis, AoV calcification severity and source of valve tissue. Fisher’s exact tests were also used to test for differences in the pattern and distribution of calcification in clinical TAVI cohort. Non-parametric Mann-Whitney tests were used to test for differences in average calcification grade, age and AoV thickness. All analyses were conducted using R V.3.5.0 (R Development Core Team15) and Prism V.7.0 (GraphPad Software, La Jolla, California, USA).

Results

Study population

Based on histological staining of calcification, postmortem leaflets (no clinical AS) were divided into three groups of increasing AoV calcification severity: (1) non-calcified (n=4): average age 22.5, 75% male; (2) mild (n=10): average age 62.3, 50% male; and (3) moderate (n=9): average age 76.8, 56% male. Parallel analyses were performed in surgical specimens (clinical AS), including (1) bicuspid aortic valve (BAV, n=4): average age 57.7, 75% male; and (2) tricuspid aortic valve (TAV, n=25): average age 68.2, 60% male. No differences in calcification severity were observed between BAV and TAV; thus, they were treated as one group (online supplementary table 3). Except for hypertension directly related to clinical AS, comorbid conditions did not correlate with AoV calcification severity (online supplementary tables 1–3).

AoV calcification is present in individuals >50 years old without clinical AS

AoV calcification patterns and severity in the absence of clinical AS were assessed in postmortem AoV leaflets. AoV calcification severity was classified as non-calcified, mild or moderate based on Alizarin red staining (figure 1B), which was consistent for all three leaflets of each valve (online supplementary figure 1A). Notably, the age of individuals with mild (p=0.002) or moderate (p=0.003) calcification was significantly increased compared with non-calcified, and the average age in the moderate group was higher than in the mild group (p=0.051) (figure 1C). Strikingly, all samples from individuals >50 years old had detectable AoV calcification, independent of related cardiovascular disease (online supplementary figure 2). Thus, calcification is increased with advancing age in postmortem AoV from individuals >50 years old.

Figure 1

Calcification severity in postmortem AoV leaflets. Alizarin red staining was used for detection of calcium deposits (A, black arrowheads) in non-calcified (n=4), mild (n=10) and moderate (n=9) whole human AoV leaflets, grouped based on AoV calcification severity (A, only right coronary leaflets shown, fibrosa layer shown on top). The dotted line in A indicates the border of the aortic root and aortic wall as determined by Movat’s pentachrome staining of an adjacent section. Positive Alizarin red staining was used to calculate the average calcification grade for each specimen, including the hinge, middle and tip regions of all three leaflets (B). The average age (years) was calculated for each group (C) (circles: female; squares: male). P values from Mann-Whitney test (B, C; **p<0.005). Ao, aorta; AoV, aortic valve.

Figure 2

Pattern of calcification distribution in postmortem AoV leaflets. Patterns of AoV calcification in non-calcified (A), mild (B) and moderate (C) human AoV RCL, NCL and LCL, shown by representative Alizarin red images (black arrowheads denote calcification, fibrosa layer shown on top). Quantification of average AoV calcification grade for the hinge, middle and tip regions in non-calcified (n=4), mild (n=10) and moderate (n=9) leaflets (D). Quantification of the average AoV calcification grade for RCL, NCL and LCL in mild and moderate leaflets combined (E). P values from Mann-Whitney test (D, E; *p<0.05). AoV, aortic valve; LCL, left coronary leaflet; NCL, non-coronary leaflet; RCL, right coronary leaflet.

AoV calcification proceeds from hinge to tip with disease progression

Increased calcification severity in moderate leaflets included an expansion of AoV calcification from the hinge to tip regions (figure 2). Calcification in the ‘hinge’ regions of mild (p=0.012) and moderate (p<0.001) leaflets was increased when compared with non-calcified leaflets, with a significant further increase in moderate versus mild (p<0.001). Additionally, RCL, NCL and LCL exhibited similar hinge calcification patterns (online supplementary figure 1B). ‘Middle’ leaflet segment calcification was increased in moderate, when compared to mild (p<0.001) or non-calcified leaflets (p<0.001). AoV calcification was not detected at the ‘tip’ region. RCL prevalence was associated with increased occurrence of calcification involving the middle region and entire leaflet when compared with LCL prevalence (online supplementary figure 1C,D), with no differences between NCL and LCL (figure 2E). Thus, mild calcification occurs predominantly at the hinge, whereas moderate calcification expands from the hinge to the more distal valve segments with RCL prevalence.

Intrinsic and nodular patterns of AoV calcification

The presence of nodular calcification is a key feature of late-stage CAVD. Notably, two distinct types of AoV calcification were observed in postmortem leaflets and surgical specimens (figure 3). In contrast to nodular, ‘intrinsic’ calcification occurred within the valve matrix and did not extend beyond the leaflet tissue boundaries. Only intrinsic calcification was observed in >80% of mild leaflets, with a similar pattern seen across all three leaflets (figure 3C,D). In contrast, >60% of moderate leaflets exhibited both intrinsic and nodular calcifications, with the RCL having the greatest number of samples with this pattern (figure 3C,E). Interestingly, >90% of surgical specimens had only nodular calcification (figure 3C). Thus, nascent disease is characterised by intrinsic calcification found primarily at the hinge. As disease progresses, increased nodular calcification is observed predominantly at the middle region of the AoV leaflet (online supplementary figure 3).

Figure 3

Identification of intrinsic and nodular AoV calcification. Alizarin red staining denotes intrinsic (black arrowheads) and nodular (yellow arrowheads) calcifications in postmortem mild and moderate leaflets, as well as surgical specimens with clinical AS (A, higher magnifications in B). Quantification of the per cent of samples that exhibit no calcification versus intrinsic, nodular, or intrinsic and nodular patterns of calcification is shown in C–E. Quantifications for no clinical AS (postmortem) and clinical AS are shown in C. The calcification patterns in individual AoV leaflets are shown for mild (D) and moderate (E) specimens. AoV, aortic valve; AS, AoV stenosis; LCL, left coronary leaflet; N, nodule; NCL, non-coronary leaflet; RCL, right coronary leaflet.

ECM abnormalities increase with AoV disease severity

AoV leaflet thickness was assessed in postmortem AoV leaflets (figure 4). Histological analysis revealed increased thickness in calcified (mild and moderate combined) when compared with non-calcified AoV leaflets (p=0.021; figure 4G). However, this increase in thickness was not seen in moderate when compared with mild leaflets (figure 4H). A healthy AoV leaflet is stratified into three ECM layers: collagen-rich fibrosa, proteoglycan-rich spongiosa and elastin-rich ventricularis (figure 5A,B). Although significant thickening was not observed with disease progression, intrinsic and nodular calcifications were associated with distinct ECM abnormalities. Intrinsic calcification in mild and moderate leaflets was associated with abnormal collagen deposition and proteoglycan expansion to the fibrosa layer mostly at the hinge region (figure 5C). In contrast, nodular calcification was associated with increased elastin fragmentation interspersed with collagen at the middle and tip leaflet regions, relative to non-calcified leaflets (figure 6A,B). Elastin fragmentation also occurred in the supraelastin layer (SE), composed mainly of collagens and proteoglycans, on the ventricular surface of the leaflet (figure 7A). Moreover, increased ectopic elastin (EE) characterised by elastin fragmentation was found on the aortic surface (figure 7A). The number of samples with both SE and EE increased with calcification severity, with over 80% of surgical specimens exhibiting both (figure 7B). Interestingly, elastin fragmentation was detected in ~20% of non-calcified leaflets (figures 6 and 7), suggesting an early pathological role for elastin abnormalities in the development of AoV calcification.

Figure 4

Leaflet thickness in postmortem AoV. Representative images of Alizarin red staining in non-calcified (A), mild (C) and moderate (E) AoV leaflets (black arrowheads: intrinsic; yellow arrowheads: nodular). Representative pentachrome images used for ECM analysis in relation to AoV calcification development in B, D and F (higher magnification in B’, D’ and F’). Using pentachrome images, the average AoV thickness per sample (including all three leaflets) was quantified for non-calcified (n=4) and calcified (n=19) leaflets (G), as well as non-calcified, mild and moderate leaflets (H). P values from Mann-Whitney test (G,H; *p<0.05). AoV, aortic valve; AS, AoV stenosis; ECM, extracellular matrix; F, fibrosa; S, spongiosa; V, ventricularis.

Figure 5

ECM abnormalities associated with intrinsic and nodular calcification. Alizarin red (A) and pentachrome (B) images (higher magnifications in C and D) for mild, moderate and clinical AS AoV samples. Non-calcified leaflets (A, black arrowheads: minimal calcification) have a normal stratified ECM (B, first column). The aortic valve surface is denoted by yellow arrows (D), while elastin fragmentation is evident nearby nodular calcification (D, yellow arrowheads). AoV, aortic valve; AS, AoV stenosis; ECM, extracellular matrix; F, fibrosa; N, nodule; S, spongiosa; V, ventricularis.

Figure 6

Elastin fragmentation in postmortem and surgical AoV leaflets. Representative pentachrome images are shown in A for non-calcified, mild, moderate and clinical AS leaflets (A, higher magnification in B; yellow arrowheads: normal elastin; black arrows: elastin fragmentation; white arrows: collagen deposition). Quantification of per cent sample with normal versus fragmented elastin for non-calcified, mild, moderate and clinical AS leaflets (C). AoV, aortic valve; AS, AoV stenosis; F, fibrosa; S, spongiosa; V, ventricularis.

Figure 7

Abnormal elastin distribution in postmortem and surgical AoV leaflets. Representative pentachrome images for mild, moderate and clinical AS AoV samples (A, black arrows: supraelastin (SE) on the ventricular surface of the leaflet; yellow arrows: ectopic elastin (EE) on the aortic surface of the leaflet). Quantification of the per cent of samples that exhibit SE, EE, or SE and EE for non-calcified, mild, moderate and clinical AS (B, note EE was not observed by itself). AoV, aortic valve; AS, AoV stenosis; F, fibrosa; S, spongiosa; V, ventricularis.

Pre-TAVI CT calcification analysis

To compare patterns of AoV calcification observed by histology with in vivo images, we analysed CT images prior to TAVI (online supplementary tables 3 and 4). The average patient age was 80.2 years (46% male), and all but one patient had obvious calcification of all three leaflets. CT images revealed a similar amount of calcification across all three leaflets in 36% of patients, while 29% involved primarily the NCL and 12% involved primarily the RCL (figure 8, online supplementary video). Analysis of calcification distribution across the leaflet revealed predominant involvement of the middle and proximal tip regions of the leaflet in 82% of patients, whereas thickening was observed at the distal leaflet tip but with minimal calcification. The CT method detected the presence of nodular calcification, but failed to detect any intrinsic calcification (figure 8). Notably, minimal to no calcification was detected at the hinge region of the leaflets. Whether this is a technical issue with detection of nodular versus intrinsic calcification, or real pathological presentation of calcification in late-stage CAVD, remains to be determined.

Supplemental material

Figure 8

Common patterns of AoV calcification demonstrated on three-dimensional (3D) contrast-enhanced CT reconstructions of TAVI patients. 3D CT reconstructions are shown as two-dimensional images. (A) Symmetrical and (B) asymmetrical (predominately involving the non-coronary leaflet (N)) regional distribution of nodular calcification is demonstrated in short axis with the most common extent of leaflet involvement observed. In B, the right (R) and left (L) coronary leaflets are less affected, with sparing of the central tip or nodules of Arantius. B is then dissected in its long axis (C), again highlighting the extent of the non-coronary leaflet involvement, predominately involving the mid-leaflet (M) and extending to the proximal tip with thickening and sparing of the distal tip (T). There is no involvement of the proximal third or hinge (H) of the leaflet. Higher magnifications are shown in A'-C'. Ao, aorta; AoV, aortic valve; LA, left atrium; LV, left ventricle; NCL, non-coronary leaflet; RA, right atrium; TAVI, transcatheter aortic valve implantation.

Discussion

Here, we show calcification in all examined human AoV samples from individuals >50 years old (n=23) without clinical AS. This is important, as progression from AoV sclerosis to AS is often uncertain, with patients commonly presenting late in the disease process.5 6 Notably, distinct types of calcification and ECM abnormalities were detected depending on the severity of the disease. Early disease development is characterised by ‘intrinsic’ calcification, primarily found within the matrix at the leaflet’s hinge, accompanied by abnormal collagen deposition and proteoglycan expansion. With increased severity, nodular calcification extends from the middle to the proximal tip regions on the aortic surface of the leaflet. Although EE has been previously observed on the fibrosa side of clinical AS specimens,1 we show increased elastin fragmentation on both the fibrosa and ventricularis sides of postmortem leaflets, suggesting a pathological role for ECM abnormalities early in disease development.

Current clinical methodologies for disease diagnosis and assessment of CAVD severity include measurement of leaflet thickening. However, in this study AoV thickening was not always present prior to calcification in postmortem leaflets, and AoV thickness did not increase with increased calcification severity. Instead, a major difference observed in leaflet pathology was related to distinct forms of calcification, in association with the development of ECM abnormalities. Interestingly, calcification progression was commonly asymmetric, involving primarily the RCL in postmortem leaflets, and the NCL or RCL in the larger clinical TAVI cohort. These findings are supported by anatomical studies demonstrating that RCL and NCL exhibit increased shear stress16–19 predominantly in the leaflet’s middle region in patients >50 years old.20 While intrinsic calcification was detected with no clinical AS, both histological analyses and in vivo images of surgical CAVD demonstrated only nodular calcification. However, the small number of specimens examined is a limitation, as the statistical power was low to detect moderate differences between groups. The significance of intrinsic calcification, its effects on valve function and disease progression, as well as its relationship to the development of nodular calcification, remain unknown. Advances in non-invasive clinical imaging can now help assess the presence of calcification in human AoV in vivo.21 22 However, our data suggest that nodular calcification is easily detected by contrast-enhanced CT, while intrinsic calcification may not be. This may be related to the particle size of this form of calcification, or possibly obscured by the adjacent contrast. Even if intrinsic calcification could be appreciated by non-contrast CT, however, this technique will lack the ability to demonstrate the necessary anatomical detail of the AoV leaflets.

Specific medical therapies, such as statins,23–26 have demonstrated either no effect or reduced progression of AS. Here, we report histologically distinct forms of calcification, intrinsic and nodular, which may have different aetiologies and responses to medical therapies. Progression from AoV sclerosis to AS has been observed to occur in ~2% of affected patients yearly.27 28 Data on the prediction of which patients will progress from AoV sclerosis to AS are limited. The development of a non-invasive method for early detection of intrinsic and/or nodular calcification could be critical to early diagnosis and treatment of CAVD. However, long-term follow-up studies will be necessary to better understand the progression of both types of calcification and how they ultimately contribute to AoV dysfunction.

Key messages

What is already known on this subject?

  • Late-stage calcific aortic valve disease (CAVD) is characterised by the presence of aortic valve (AoV) calcification leading to clinically evident AoV stenosis (AS); however, disease development prior to clinical presentation is poorly understood.

What might this study add?

  • All AoV samples from individuals >50 years old had calcification, with severity correlating with age, independent of CAVD risk factors.

  • Two distinct types of calcification were identified: ‘intrinsic’, primarily found at the hinge of postmortem leaflets with no clinical AS; and ‘nodular’, primarily found in the middle region of more severely affected postmortem leaflets and in clinical AS.

How might this impact on clinical practice?

  • Understanding the pathological role of intrinsic calcification during the development of AoV disease may improve imaging interpretation for early detection of AoV pathological changes.

Acknowledgments

We thank the Burns Blaxall laboratory at Cincinnati Children’s for providing us with the whole human hearts procured from Restore Life USA (Elizabethton, Tennessee).

References

Footnotes

  • Contributors Study design: MVG-S, OG-R, JTT, WM, DK, KY. Data generation and analysis: MVG-S, JTT, KH, OG-R, DA, NO, KY. Statistical analysis: MVG-S, NO. Procurement of human tissue, imaging and study design: OG-R, JTT, WM, JC, JMS, DK. Wrote the manuscript with feedback from all authors: MVG-S, JTT, KY.

  • Funding This study was funded by NIH R01 HL114682 (KY).

  • Competing interests DK is on the scientific advisory board and is a consultant for Boston Scientific.

  • Ethics approval All studies have been approved by the institutional review boards of Cincinnati Children’s Hospital Medical Center, TriHealth and The Christ Hospital. Cincinnati Children’s is the institution of record. The study IDs are 2016-6506 for Cincinnati Children’s and 17-43 for The Christ Hospital. Informed consent was obtained from all subjects prior to inclusion in the study.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Patient consent for publication Not required.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Linked Articles