Author + information
- Received August 10, 2012
- Revision received October 12, 2012
- Accepted November 7, 2012
- Published online February 19, 2013.
- Haïfa Mahjoub, MD,
- Patrick Mathieu, MD,
- Mario Sénéchal, MD,
- Eric Larose, MD,
- Jean Dumesnil, MD,
- Jean-Pierre Després, PhD and
- Philippe Pibarot, DVM, PhD⁎ ()
- ↵⁎Reprint requests and correspondence Dr.
Philippe Pibarot, Institut Universitaire de Cardiologie et de Pneumologie de Québec, 2725 Chemin Sainte-Foy, Québec, Quebec G1V-4G5, Canada
Objectives This study sought to identify the clinical and metabolic determinants of bioprosthetic valve degeneration.
Background Structural valve degeneration (SVD) is the major cause of bioprosthetic valve failure. Recent retrospective studies have reported an association between atherosclerotic risk factors and development of SVD.
Methods A total of 203 consecutive patients with aortic bioprosthetic valves were recruited. Doppler echocardiography and multidetector computed tomography (CT) examinations were performed for assessment of bioprosthesis calcification and abdominal adiposity. A cardiometabolic risk profile was also obtained. SVD was defined as an increase in mean transprosthetic gradient of ≥10 mm Hg and/or a worsening of transprosthetic regurgitation ≥1/3 class between 1-year post-operative echo and last follow-up echo (mean follow-up time was 8 ± 3 years).
Results Forty-two patients (20%) were identified as developing SVD. Patients with SVD had significantly higher plasma total-cholesterol (4.6 ± 1.1 mmol/l vs. 4.1 ± 0.9 mmol/l, p = 0.05), low-density lipoprotein-cholesterol (2.5 ± 1.0 mmol/l vs. 2.2 ± 0.7 mmol/l, p = 0.02), and apolipoprotein B (ApoB) levels (0.71 ± 0.22 g/l vs. 0.64 ± 0.17 g/l, p = 0.02) and higher ApoB/ApoA-I ratios (0.48 ± 0.17 vs. 0.41 ± 0.11, p = 0.004) than those with no SVD. Multivariate analysis revealed that increased ApoB/ApoA-I ratio (odds ratio [OR]: 1.41, 95% confidence interval [CI]: 1.10 to 1.82 per 0.1 increment; p = 0.007) and the use of bisphosphonates (OR: 3.57, 95% CI: 1.14 to 10.80 p = 0.02) were the strongest independent predictors of SVD.
Conclusions This is the first study to report a strong association between increased ApoB/ApoA-I ratio and the risk of developing SVD among patients with aortic bioprosthetic valves. Further studies are needed to determine if an elevated ApoB/ApoA-I ratio, which reflects the balance of proatherogenic and antiatherogenic lipoproteins, is a risk marker or a risk factor for SVD.
Approximately 275,000 prosthetic valves are implanted each year worldwide, one-half of which are mechanical and half are bioprosthetic valves, with a shift toward greater use of bioprostheses (BPVs) over the last decade (1). BPVs generally have a good hemodynamic profile and, unlike mechanical valves, have a low thrombogenic potential and therefore generally obviate the need for anticoagulation. However, BPVs are still plagued by their limited durability (1,2). Structural valve degeneration (SVD) is the major cause of bioprosthetic valve failure (3,4). SVD is expressed clinically by the development of a progressive stenosis due to leaflet calcification or by regurgitation due to leaflet tear. Reoperation rates for SVD of BPVs have been shown to be as high as 30% at 15 years (4). For years, SVD was seen as a passive degenerative process of calcification related to the chemical fixative treatment of porcine or bovine tissue before implantation, which is accelerated by mechanical stress (1). Recent studies have reported similarities between histological findings in native aortic stenosis valves and BPVs explanted for SVD such as the presence of lipids, foam cells, and inflammation (5,6). There is compelling evidence that aortic stenosis is an active disorder involving atherosclerotic-like processes (7–10). Similarly, it has been suggested that atherosclerotic risk factors may influence the development of SVD. For instance, recent retrospective studies have shown an association between some atherosclerotic risk factors such as diabetes, smoking, hypercholesterolemia, and metabolic syndrome and the development of SVD (11–13). These results suggest that atherosclerotic-like processes could also contribute to the pathogenesis of SVD. Identification of modifiable risk factors of SVD is an essential first step in the development of medical therapies that could inhibit or at least slow the progression of SVD, especially if initiated at the time of surgery, before the onset of BPV tissue degeneration. The objective of the present study was to identify the clinical and metabolic correlates of bioprosthetic SVD assessed by Doppler echocardiography.
From June 2008 to June 2010, 203 consecutive patients with a BPV in the aortic position were prospectively recruited for this study that included a Doppler echocardiography, a multislice computed tomography (CT) examination, and a blood sample analysis. All patients underwent an isolated aortic valve replacement procedure at the Quebec Heart and Lung Institute at least 3 years previously, with a complete Doppler echocardiographic examination available at 12 ± 6 months post-operatively (labeled “1-year post-op. Echo” hereafter). Exclusion criteria were: 1) presence of mild or increased paravalvular regurgitation; 2) significant concomitant mitral valve disease, defined by mild or increased mitral regurgitation or mitral valve effective orifice area (EOA) <1.5 cm2; 3) subvalvular flow acceleration precluding measurement of BPV valve EOA; 4) left ventricular (LV) systolic dysfunction, defined by an LV ejection fraction of <50%; and 5) congestive heart failure with New York Heart Association class III or IV. All patients recruited to the study had a clinical examination, a complete plasma glycemic and lipid profile, and a complete Doppler echocardiographic study (labeled “last follow-up echo” hereafter). Patients also had a multidetector CT scan for assessment of BPV calcification and for abdominal subcutaneous and visceral adiposity.
All Doppler echocardiographic studies were reviewed by the same cardiologist (H.M.). Operators were blinded to the results of clinical, laboratory, and CT data. Peak transprosthetic flow velocity was determined by continuous-wave Doppler. Mean transprosthetic gradient was calculated using the modified Bernoulli equation. BPV EOA was calculated using the standard continuity equation. The absolute and annualized changes in mean gradient and EOA were calculated as: absolute change = (value at last follow-up echo – value at 1-year post-op echo); annualized change = (value at last follow-up echo – value at 1-year post-op echo)/time between 1 year and last follow-up echocardiographic examinations, respectively.
Prosthetic regurgitation was detected by color Doppler echocardiography, and the origin of the jet was visualized in several views to differentiate periprosthetic from transprosthetic regurgitation. Transprosthetic regurgitation severity was assessed as recommended by the American Society of Echocardiography and classified as mild, moderate, or severe (14). Worsening of valve regurgitation was defined as an increase of at least one-third of the class in the severity of regurgitation during follow-up according to the following scheme: from none or mild to moderate or from moderate to severe.
SVD was defined as an increase in transprosthetic mean gradient ≥10 mm Hg and/or worsening of transprosthetic regurgitation ≥1/3 class between 1-year and last follow-up echocardiograms. Prosthesis-patient mismatch (PPM) was defined as not clinically significant (i.e., mild or no PPM) if the indexed EOA was >0.85 cm2/m2, moderate if it was >0.65 cm2/m2 and ≤0.85 cm2/m2, and severe if it was ≤0.65 cm2/m2.
Clinical and operative data
Medical history included history of smoking, documented diagnoses of hypertension (patients receiving antihypertensive medications or having known but untreated hypertension [blood pressure ≥140/90 mm Hg]), diabetes (fasting glucose ≥7 mmol/l), hypercholesterolemia (patients receiving cholesterol-lowering medication or, in the absence of such medication, having a total plasma cholesterol level >240 mg/l), coronary heart disease (history of myocardial infarction or coronary artery stenosis on coronary angiography), renal insufficiency (estimated glomerular filtration rate <60 ml/min/1.73 m2), and detailed information on current medication were collected. Body weight, height and waist circumference were measured following standardized procedures. Blood pressure, heart rate and NYHA class were also assessed. The clinical identification of patients with the features of the metabolic syndrome was based on the modified criteria proposed by the National Cholesterol Education Program Adult Treatment Panel III (NCEP-ATP III) (15). Operative data including model and size of BPVs were also recorded.
Blood was obtained by venipuncture after patients fasted overnight. Plasma was isolated and immediately processed by the laboratory for the measurement of glucose, insulin, total cholesterol (total-C), high-density lipoprotein-cholesterol (HDL-C), and triglyceride (TG) levels. Low-density lipoprotein-cholesterol (LDL-C) was calculated with the Friedewald formula when TG levels were ≤5.0 mmol/l (16). No patient had TG levels ≥5.0 mmol/l. Plasma apolipoprotein A-I (ApoA-I) and apolipoprotein B (ApoB) concentrations were measured using a nephelometric method with polyclonal antibodies on a BN-Prospect machine (Dade-Behring) (17). Distribution of LDL particle sizes was determined by nondenaturing 2% to 16% polyacrylamide gradient gel electrophoresis (18). The relative proportion of small LDL particles was determined by computing the relative area of the densitometric scan corresponding to LDL particles <255 Å. The absolute concentration of cholesterol among small LDL particles was calculated by multiplying plasma LDL-C levels by the relative proportion of small LDL particles. The homeostatic assessment model (HOMA) index was calculated to assess insulin resistance. C-reactive protein, creatinine levels, calcemia, and phosphatemia were also measured.
Multidetector CT data
BPV leaflet calcification was quantified by multidetector CT with the use of the volumetric method that identifies calcium within the BPV leaflets as areas of at least two contiguous pixels with a density ≥130 HU as previously described (19). The volume of calcified tissue expressed in cubic millimeters was individually calculated by summing the lesion volumes for all sections containing calcium. Particular attention was paid to distinguishing calcifications located in the region of the BPV leaflets from those located in the region of the prosthesis sewing ring and aortic annulus. Cross-sectional areas of abdominal total, visceral, and subcutaneous adipose tissue were also assessed by CT using previously described procedures (20). Operators were blinded to the results of echocardiograms.
Continuous data were expressed as mean ± SD or median and interquartile ranges (IQR) and were compared using unpaired Student's t test or a Wilcoxon rank sum test. Categorical data were expressed as a percentage and compared with the chi-square test. A logarithmic transformation was used when variables did not follow a normal distribution. Correlations between variables were determined using linear regressions or Spearman's coefficients. A multiple logistic regression analysis was used to identify the factors independently associated with BPV SVD. A multiple linear regression was used to identify the independent predictors of the progression rate of transprosthetic mean gradient. Variables with p values of ≤0.1 on univariate analysis were entered into the multivariate models. Age at implantation was forced into the models. A p value of <0.05 was considered statistically significant.
Clinical and operative data
Among the 203 patients recruited to the study, 42 (20%) were classified as having SVD by Doppler echocardiography. Table 1 compares patients with SVD with those with no SVD, with respect to clinical and operative data. The follow-up time was significantly (p = 0.03) longer in the SVD group (9 ± 4 years) than in the No SVD group (7 ± 3 years). There were no significant differences between the two groups in terms of age at implantation, sex, and cardiovascular risk factors (Table 1). Of the 203 patients, 90 patients (44%) met the clinical criteria of the metabolic syndrome with 22 patients (52%) in the SVD group and 68 patients (42%) in the No SVD group (p = 0.2). Moreover, there was no significant difference between the two groups in terms of medical therapy such as statins (p = 0.1), angiotensin-converting enzyme inhibitors (p = 0.4), and angiotensin receptor blockers (p = 0.5). However, a significantly greater proportion of patients took bisphosphonates in the SVD group than in the No SVD group (18% vs. 11%; p = 0.04). As for operative data, there was no difference between groups regarding type of BPV (porcine vs. bovine or stentless vs. stented), prosthesis size, use of Teflon-felt pledgets for BPV suture, or prevalence of PPM. There was a borderline significant trend (p = 0.05) for higher proportion of concomitant coronary artery bypass graft at the time of valve replacement surgery in the SVD group.
Doppler echocardiographic and CT data
SVD was associated with an increase in gradient in 81% of cases (36 patients), and only 8 patients (19%) had diagnoses of SVD related to a worsening of valve regurgitation. Among these 8 patients, 4 had moderate valve regurgitation at last follow-up echocardiography, whereas 4 patients had a severe valve regurgitation caused by a leaflet tear with an indication for urgent aortic valve replacement. One of these 4 patients died immediately after surgery due to cardiogenic shock.
As expected, patients in the SVD group had higher mean transprosthetic gradient (24 ± 7 mm Hg vs. 12 ± 5 mm Hg; p < 0.0001), peak gradient (44 ± 13 mm Hg vs. 25 ± 10 mm Hg; p < 0.0001), peak transprosthetic velocity (3.25 ± 0.5 m/s vs. 2.44 ± 0.5 m/s; p < 0.0001), and lower EOA (0.97 ± 0.36 cm2 vs. 1.36 ± 0.4 cm2, p < 0.0001) at last echocardiography follow-up than patients in the No-SVD group (Table 2). The average absolute progression rate of mean gradient was 12.4 ± 5 mm Hg vs. 0.71 ± 4 mm Hg (p < 0.0001), and the average annualized progression rate of mean gradient was 1.9 ± 1.6 (median: 1.38; IQR: 0.91 to 2.49) mm Hg/year versus 0.09 ± 0.6 (median: 0.11, IQR: −0.31 to 0.39) mm Hg/year (p < 0.0001), respectively, in the SVD versus No SVD groups, respectively. Furthermore, patients with SVD had significantly higher calcium deposit within the BPV leaflets, assessed by CT (181 ± 33 mm3 vs. 75 ± 37 mm3; p < 0.0001).
Metabolic profile of patients with bioprosthetic valve degeneration
Patients with SVD had significantly higher plasma levels of total-C (4.6 ± 1.1 mmol/l vs. 4.1 ± 0.9 mmol/l; p = 0.05), LDL-C (2.53 ± 1.0 mmol/l vs. 2.21 ± 0.7 mmol/l, p = 0.02), non–HDL-C (3.22 ± 1.09 mmol/l vs. 2.82 ± 0.85 mmol/l, p = 0.01), ApoB (0.71 ± 0.2 g/l vs. 0.64 ± 0.1 g/l, p = 0.02), and concentration of LDL-C <255 Å (0.66 ± 0.56 mmol/l vs. 0.55 ± 0.34 mmol/l, p = 0.03) than patients without SVD (Table 3). The ApoB/ApoA-I and the total-C/HDL-C ratios were also significantly higher in patients with SVD (0.48 ± 0.17 vs. 0.41 ± 0.11, p = 0.004 and 3.65 ± 1.3 vs. 3.2 ± 0.8, p = 0.04 respectively). Moreover, a higher HOMA index (3.9 ± 5.9 vs. 2.9 ± 2.6, p = 0.1) was also observed in patients within the SVD group.
Predictors of bioprosthetic valve degeneration
On univariable analysis, SVD was associated with longer follow-up duration (p = 0.03) and with higher plasma levels of total-C (p = 0.05), LDL-C (p = 0.02), non–HDL-C (p = 0.01), and ApoB (p = 0.02), higher total-C/HDL-C ratio (p = 0.04), higher ApoB/ApoA-I ratio (p = 0.004), and use of bisphosphonate therapy (p = 0.04) (Tables 1 and 3). On multivariable analysis, the strongest independent predictors of SVD were the ApoB/ApoA-I ratio (odds ratio [OR]: 1.41, 95% confidence interval [CI]: 1.10 to 1.82 per 0.1 increment; p = 0.007), and the use of bisphosphonates (OR: 3.57, 95% CI: 1.14 to 10.80 p = 0.02) (Fig. 1). When statin therapy was forced into the multivariate model, it did not reach statistical significance (OR = 0.91, 95% CI: 0.38 to 2.26, p = 0.8) and ApoB/ApoA-I ratio (OR: 1.39, 95% CI: 1.06 to 1.84 per 0.1 increment; p = 0.01), as well as the use of bisphosphonates (OR: 3.56, 95% CI: 1.14 to 10.79; p = 0.02) remained significant predictors of SVD. The use of bisphosphonates remained independently associated with SVD even after further adjustment for sex (OR: 3.62, 95% CI: 1.11 to 11.49, p = 0.03).
When we used a linear regression multivariate model, the independent predictors of higher progression rate of mean transprosthetic gradient (in log-transformed format) were a younger age at implantation (β = −0.03 ± 0.01; p = 0.008), a stented BPV (β = 0.97 ± 0.29; p = 0.001), and higher ApoB/ApoA-I ratio (β = 1.47 ± 0.65; p = 0.02).
Correlates of ApoB/ApoA-I ratio
As ApoB/ApoA-I ratio was the most powerful independent predictor of SVD, we sought to determine which were the clinical and metabolic factors associated with increased ApoB/ApoA-I ratio in our study sample. ApoB/ApoA-I ratio was significantly and positively associated with clinical diagnosis of the metabolic syndrome (r = 0.36; p < 0.0001) and of hypercholesterolemia (r = 0.23; p = 0.0009), with the HOMA index (r = 0.18, p = 0.01) and with the CT cross-sectional areas of total abdominal (r = 0.21, p = 0.006) and visceral adipose (r = 0.14, p = 0.05) tissue (Table 4). There was also a significant negative correlation between ApoB/ApoA-I ratio and statin therapy (r = −0.38; p < 0.0001). On multivariable analysis, the metabolic syndrome (p < 0.0001) and the absence of statin therapy (p < 0.0001) were independently associated with higher ApoB/ApoA-I ratio (Table 4). The ApoB/ApoA-I ratio remained an independent predictor (OR = 1.35, 95% CI: 1.02 to 1.80 per 0.1 increment; p = 0.03) of SVD after adjustment for age, duration of follow-up, and presence of metabolic syndrome and diabetes.
This prospective study reports that higher plasma levels of LDL-C, non–HDL-C, and ApoB as well as higher ApoB/ApoA-I and total-C/HDL-C ratios are significantly associated with increased risk of SVD. Importantly, an increased ApoB/ApoA-I ratio was found to be the most powerful independent predictor of SVD. This novel finding suggests that the proportion of LDL and HDL particles and the ensuing dysregulation in cholesterol transport balance could play a key role in the tissue degeneration of BPVs. These findings provide further support to the concept that an active lipid-mediated process is involved in SVD and that this process is determined, in large part, by the quality rather than the quantity of the LDL and HDL particles.
Definition and prevalence of SVD
The incidence of SVD is most often derived from the incidence of reoperation for BPV failure. Several studies in large series of patients have reported an SVD incidence of 10% to 30% at 10 years and 20% to 50% at 15 years (4,21,22). However, this approach may substantially underestimate the real incidence of SVD (12,23). Several patients may indeed die before their failed BPV is replaced, and it cannot be excluded that the BPV dysfunction might have contributed directly or indirectly to a patient's death. Furthermore, an important proportion of patients with severe BPV dysfunction may be denied surgery owing to advanced age and/or severe comorbidities. The presence of SVD may also affect the quality of life of these patients and may contribute to the occurrence of adverse events. Hence, it is probably more accurate to define SVD on the basis of the development of BPV hemodynamic dysfunction documented by Doppler echocardiography. In the present study, the prevalence of SVD identified by Doppler echocardiography was 20% at a mean follow-up of 8 years. In previous studies where Doppler echocardiography was used to identify SVD, Flameng et al. reported an SVD incidence of 7% at a median follow-up of 6 years (23), whereas other investigators reported an incidence of 30% to 39% with no precise information about the time duration since implantation (12,24). Discrepancies among the SVD incidence rates reported in those studies may be due to differences in the Doppler echocardiographic parameters and criteria used to define SVD as well as to differences in the average time interval between the BPV implantation and the last echocardiographic follow-up. In the present study, we elected to define SVD on the basis of an absolute increase in mean transprosthetic gradient of ≥10 mm Hg and/or a worsening of at least one class of regurgitation between the 1-year and last follow-up echocardiographic examinations. We used this definition because it was clinically meaningful and it was well suited for this study where the time between the 1-year and last follow-up echocardiographic examinations was highly variable from one patient to the other (3 to 18.6 years). Use of the annualized change in mean gradient with a fixed cut-point value (e.g., >3 mm Hg/year such as that used in previous studies) to identify SVD would have been misleading in this context.
Risk factors of SVD
Similar to observations of native aortic valve stenosis, several retrospectives studies reported an association between cardiovascular risk factors and SVD, thus suggesting that atherosclerotic-like processes could be involved in the degeneration of BPV tissues (11–13,22,25). Several studies reported an association between increased cholesterol levels and SVD (13,25). Briand et al. (12) also reported that patients with metabolic syndrome have faster progression of BPV degeneration as documented by Doppler echocardiography. A large multicenter study (25) recently reported that type 2 diabetes is a powerful predictor of SVD and mortality in patients with aortic or mitral valve replacement with a BPV. However, those previous studies were retrospective and did not report the measures of the quality of lipoproteins including ApoB and ApoA-I levels and ApoB/ApoA-I ratios.
In the present study, we found that several markers of lipid metabolism, including LDL-C, non–HDL-C, ApoB, and the ApoB/ApoA-I ratio were significantly associated with higher incidences of SVD. Of interest, the most powerful associations were found with ApoB levels and ApoB/ApoA-I ratio. Total ApoB level largely reflects the number of potentially atherogenic particles. ApoA-I is the main protein in HDL particles and is involved in reverse cholesterol transport. It has been suggested that the advantage of ApoB/ApoA-I ratio relative to standard lipid markers such LDL-C or HDL-C is that it better reflects the balance between concentrations of proatherogenic versus antiatherogenic lipoprotein particles. This may explain why the ApoB/ApoA-I ratio outperforms LDL-C as a predictor of ischemic events and as an index of residual cardiovascular risk, particularly in populations with LDL-C in the normal or low abnormal ranges (26–30). The INTERHEART study, a myocardial infarction case-control multicenter study of 30,000 patients, reported that among all measured cardiovascular risk factors including smoking, hypertension, abdominal obesity, and diabetes, an elevated ApoB/ApoA-I ratio was the strongest predictor of myocardial infarction (27). In the AMORIS (Apolipoprotein-Related Mortality Risk) study, a Swedish prospective study of 175,000 subjects, ApoB and ApoB/ApoA-I ratio were found to be superior to LDL-C levels in predicting cardiovascular risk, especially in the stratum of patients with normal LDL-C levels (31). These findings are consistent with the recent results of the Emerging Risk Factors Collaboration trial that showed that ApoB and Apo-AI improve cardiovascular risk stratification especially in individuals classified at intermediate risk (10% or <20% predicted 10-year risk) (32). In the present study including a series of patients where 87% of patients had an LDL-C level of <130 mg/dl (3.37 mmol/l), the ApoB/ApoA-I ratio was a strong predictor of SVD.
Bisphosphonates and SVD
Conflicting results have been reported concerning the association of bisphosphonates and calcific aortic valve disease (33,34). In the present study, bisphosphonate therapy was associated with SVD of BPVs, and this association persisted after adjustment for age and sex. Further studies are needed to examine the relationship between dysregulation of mineral metabolism and SVD of BPVs and to determine whether bisphosphonate therapy has a direct positive or negative effect on SVD and whether this effect is age-dependent as in native aortic valve disease.
Potential mechanisms underlying the association between lipoproteins and SVD
For years, degeneration of BPVs was seen as a purely passive mechanism, which was to a large extent related to the chemical fixative treatment of porcine or bovine tissues (1). This hypothetical model proposed that the fixative treatment with glutaraldehyde prior to BPV implantation combined to the repetitive mechanical stress after implantation induced nucleation and growth of hydroxyapatite crystals (1).
However, recent histological studies of failed BPVs explanted at reoperation also reported the presence of inflammatory cells, foam cells, ApoB, and oxidized lipids, as well as expression of metalloproteinases within BPV tissues (5,6). These findings suggest that similarly to native aortic valve stenosis, a lipid-mediated inflammatory process could contribute, at least in part, to the SVD of BPVs (22). The infiltration of LDL particles within the BPV tissues and their oxidation may trigger an inflammatory process and the formation of foam cells (6,22). In turn, the inflammatory cytokines and oxidized LDLs may induce active mineralization and disruption of extracellular matrix, thereby leading to BPV stenosis and/or regurgitation.
The infiltration and oxidation of LDL particles within the vascular and valvular tissues appears to be more closely related to the number, size, and density of LDL particles rather than to the total plasma concentration of LDL-C per se (6,35,36). Indeed, the small dense LDL particles are the most atherogenic because they persist in circulation for a longer period of time, have a greater ability to penetrate within the arterial and valvular tissues, and have a high susceptibility to oxidation. In native aortic valve stenosis, the proportion of small, dense LDL particles was the only blood metabolic marker found to correlate with the amount of oxidized LDL deposited within the valve tissue (36). In a recent retrospective study (37), we also found a significant association between the prevalence of BPV hemodynamic dysfunction and increased proportion of small LDL particles. In the present study, the concentration of LDL-C <255 Å was significantly associated with SVD. Furthermore, the phenotype of HDL particles may alter their anti-atherogenic effects, regardless of the total plasma concentration of HDL-cholesterol. Hence, abnormal phenotypes of LDL and/or HDL particles, as reflected by increased ApoB/ApoA-I ratio, could enhance the production of oxidized LDLs within the BPV tissues and thus promote inflammatory and calcifying processes leading to SVD.
Viscerally obese individuals often develop insulin resistance and an atherogenic dyslipidemia characterized by an increased proportion of small, dense LDL and HDL particles. These individuals may thus be at higher risk for SVD, as suggested by previous retrospective studies (12,25). In the present study, we found that the presence of the metabolic syndrome was the most powerful determinant of increased ApoB/ApoA-I ratio. Nonetheless, the ApoB/ApoA-I ratio remained a significant independent predictor of SVD even after adjustment for metabolic syndrome and type II diabetes.
SVD is the main cause of valve-related morbidity and mortality in patients with BPVs. There was until now no medical treatment to prevent the development of SVD. An increased ApoB/ApoA-I ratio may be useful to identify the patients who may be at higher risk for SVD. Further studies are needed to determine if the imbalance between proatherogenic and antiatherogenic lipoproteins reflected by increased ApoB/ApoA-I contributes directly to SVD or whether it is simply a marker for other risk factors. Nonetheless, when analyzing collectively the data presented in this study as well as those reported in previous studies (12,25), the main factors that have been found to be independently associated with increased risk of SVD (i.e., metabolic syndrome, diabetes, increased ApoB/ApoA-I ratio) are, at least in part, related to visceral obesity and ensuing insulin resistance. Hence, aggressive changes in lifestyle, including increase in physical activity and implementation of dietary changes (38), should probably be considered as the first-line intervention for SVD prevention in those patients with a BPV presenting with one or more of these factors. On the other hand, specific pharmacological therapies directed toward visceral obesity, metabolic syndrome, diabetes, or the associated atherogenic dyslipidemia will need to be supported by further mechanistic studies and validated in future randomized clinical trials. One of the advantages inherent to this particular context is that, as opposed to patients with native aortic valve disease, the pharmacological treatment could be instituted at the time of AVR before the initiation of the pathologic processes leading to SVD. This study also revealed that the association between ApoB/ApoA-I persisted after adjustment for metabolic syndrome and diabetes, which suggest that other factors may modulate the ApoB/ApoA-I ratio and the associated risk of SVD.
Statin drugs have been shown to decrease ApoB by 15% to 40% and increase ApoA-I by 5% to 15%; therefore, decreasing the ApoB/ApoA-I ratio by 25% to 45% (26,39,40). Consistently, in the present study, statins were independently associated with decreased ApoB/ApoA-I ratio. However, 77% of patients were taking statins in the present study, and there was no evidence of significant protective effect of such therapy with regard to SVD. In patients with native aortic valve stenosis and normal cholesterol levels, three randomized trials (SALTIRE [Scottish Aortic Stenosis and Lipid Lowering Trial], SEAS [Simvastatin and Ezetimibe in Aortic Stenosis] trial, and ASTRONOMER [Aortic Stenosis Progression Observation: Measuring Effects of Rosuvastatin] trial) failed to demonstrate a significant effect of statins on stenosis progression and clinical outcomes (41–43).
As for BPV degeneration, one retrospective study in 167 patients reported that statin therapy reduced the development of SVD (24). Other retrospective studies in larger series of patients, however, have failed to demonstrate a benefit for statin therapy in the prevention of SVD (12,25,44).
The main limitation of these studies of patients with BPV, including the present investigation, is that they are non-randomized and their results may thus have been affected by: 1) the inherent differences between the patients treated with statins versus those without statins; and 2) the inter-individual variability in the timing of initiation and in the duration of statin therapy.
The present study is a prospective cross-sectional study that reported the prevalence of SVD at the time of the study. Most patients did not have serial echocardiographic follow-up. Hence, the information about the exact timing of SVD onset is unknown. In addition, the blood metabolic data were obtained at the time of the study (i.e., at last echocardiographic follow-up, which limits the inferences with regards to the value of the ApoB/ApoA-I ratio at the time of surgery to predict future development of SVD). The apparent lack of significant association between SVD and some clinical factors, including metabolic syndrome, diabetes, and statin therapy, may be a type II error because of the relatively small sample size. Given the small number (n = 45) of patients with SVD, some of the multivariate logistic regression models were overfitted. Nonetheless, this limitation does not affect the validity of the main result of this study, which is the demonstration of a strong association between increased ApoB/ApoA-I ratio and SVD.
This is the first study to report a strong association between increased ApoB/ApoA-I ratio and the risk of SVD. These findings support the concept that SVD is not a passive degenerative process but rather an active multifaceted process, which includes lipid-mediated mechanisms leading to mineralization and/or disruption of BPV tissues. Further studies are needed to determine if increased ApoB/ApoA-I ratio, which reflects the balance of pro-atherogenic and anti-atherogenic lipoproteins, is simply a risk marker or rather a true risk factor for SVD. Nonetheless, the findings of the present study can be used to generate the hypothesis that the initiation at the time of aortic valve replacement of aggressive changes in lifestyle and/or specific pharmacological therapy aiming at reduction of ApoB/ApoA-I ratio might help to reduce the incidence of SVD.
The authors thank Jacinthe Aubé and Dominique Labrèche for data collection and technical assistance.
This study was funded by a research grant (MOP #86666) from Canadian Institutes of Health Research (CIHR), Ottawa, Ontario, Canada. Dr. Mahjoub is the recipient of a PhD student scholarship from the International Chair on Cardiometabolic Risk. Dr. Pibarot holds the Canada Research Chair in Valvular Heart Diseases, CIHR. Dr. Mathieu is a research scholar at the Fonds de Recherche en Santé du Québec, Montreal, Québec, Canada. Dr. Després is the Scientific Director of the International Chair on Cardiometabolic Risk, Université Laval. All authors have reported that they have no relationships relevant to the contents of this study to disclose.
- Abbreviations and Acronyms
- apolipoprotein A-I
- apolipoprotein B
- bioprosthetic valve
- computed tomography
- effective orifice area
- high-density lipoprotein-cholesterol
- low-density lipoprotein-cholesterol
- left ventricular
- mean transprosthetic gradient
- structural valve degeneration
- total cholesterol
- Received August 10, 2012.
- Revision received October 12, 2012.
- Accepted November 7, 2012.
- American College of Cardiology Foundation
- Ruel M.,
- Kulik A.,
- Rubens F.D.,
- et al.
- Otto C.M.,
- Kuusisto J.,
- Reichenbach D.D.,
- Gown A.M.,
- O'Brien K.D.
- Freeman R.V.,
- Otto C.M.
- Cote C.,
- Pibarot P.,
- Despres J.P.,
- et al.
- Zoghbi W.A.,
- Chambers J.B.,
- Dumesnil J.G.,
- et al.
- Warnick G.R.,
- Knopp R.H.,
- Fitzpatrick V.,
- Branson L.
- St-Pierre A.C.,
- Ruel I.L.,
- Cantin B.,
- et al.
- Bonow R.O.,
- Carabello B.A.,
- Chatterjee K.,
- et al.
- Pibarot P.,
- Dumesnil J.G.
- Flameng W.,
- Herregods M.C.,
- Vercalsteren M.,
- Herijgers P.,
- Bogaerts K.,
- Meuris B.
- Lorusso R.,
- Gelsomino S.,
- Luca F.,
- et al.
- Meisinger C.,
- Loewel H.,
- Mraz W.,
- Koenig W.
- Walldius G.,
- Jungner I.,
- Aastveit A.H.,
- Holme I.,
- Furberg C.D.,
- Sniderman A.D.
- Elmariah S.,
- Delaney J.A.,
- O'Brien K.D.,
- et al.
- Mohty D.,
- Pibarot P.,
- Despres J.P.,
- et al.
- Chan K.L.,
- Teo K.,
- Dumesnil J.G.,
- Ni A.,
- Tam J.
- Kulik A.,
- Masters R.G.,
- Bedard P.,
- et al.