Adiponectin serum level is an independent and incremental predictor of all‐cause mortality after transcatheter aortic valve replacement

Abstract Background Quantifiable biomarkers may be useful for a better risk and frailty assessment of patients referred for transcatheter aortic valve implantation (TAVI). Hypothesis To determine if adiponectin serum concentration predicts all‐cause mortality in patients undergoing TAVI. Methods 77 consecutive patients, undergoing TAVI, were analyzed. The CT axial slices at the level of the fourth lumbar vertebra were used to measure the psoas muscle area, and its low‐density muscle fraction (LDM (%)). To assess the operative risk, the STS (Society of Thoracic Surgeons Predicted Risk of Mortality) score, Log. Euroscore, and Euroscore II were determined. A clinical frailty assessment was performed. ELISA kits were used to measure adiponectin serum levels. We searched for a correlation between serum adiponectin concentration and all‐cause mortality after TAVI. Results The mean age was 80.8 ± 7.4 years. All‐cause mortality occurred in 22 patients. The mean follow‐up was 1779 days (range: 1572–1825 days). Compared with patients with the lowest adiponectin level, patients in the third tertile had a hazards ratio of all‐cause mortality after TAVI of 4.155 (95% CI: 1.364–12.655) (p = .004). In the multivariable model, including STS score, vascular access of TAVI procedure, LDM (%), and adiponectin serum concentration, serum adiponectin level, and LDM(%) were independent predictors of all‐cause mortality after TAVI (p = .178, .303, .042, and .017, respectively). Adiponectin level was a predictor of all‐cause mortality in females and males (p = .012 and 0.024, respectively). Conclusion Adiponectin serum level is an independent and incremental predictor of all‐cause mortality in patients undergoing TAVI.


| INTRODUCTION
Over the last two decades, transcatheter aortic valve implantation (TAVI) has evolved from an experimental procedure to an appropriate treatment of symptomatic aortic stenosis in patients at high operative risk. 1,2 Nonetheless, the appropriate selection of patients, who will benefit from TAVI, remains challenging.
It has been demonstrated that conventional clinical cardiac surgery risk assessment, by EuroSCORE II 3 , logistic EuroSCORE, 3 and STS Score, 4 predicts mortality after TAVI. Clinical frailty measurements have been demonstrated to be predictive of mortality after TAVI. 5,6 More recently, imaging studies explored if computed tomography (CT) derived measurements of visceral adipocyte tissue (VAT), subcutaneous adipocyte tissue (SAT), and psoas muscular circumferential surface area (CSA psoas) can be used as surrogate markers of frailty or sarcopenia to predict all-cause mortality in patients, scheduled for surgery or TAVI. 7,8 Plasma adiponectin has been proposed as a potential biomarker of frailty. [9][10][11] Adiponectin is recognized as a plasma cytokine that is released from adipose and skeletal muscle; plasma levels are increased in patients with decreased skeletal muscle mass, weight loss, and reduced physical activity. 11 Elevated adiponectin levels are also noted in elderly cognitively impaired patients, and adiponectin has been proposed as a biomarker to detect earlier cognitive decline in frail elderly patients. 12 There may be sex differences between adiponectin levels, frailty and mental status. 13 In females, there is a negative relationship noted between elevated adiponectin levels and both frailty and anxiety: such that high levels of adiponectin are associated with low body mass index and low levels of anxiety.
Whereas in males, the relationship is more complex; increasing adiponectin levels were associated with increased fat mass but decreased lean mass. Overall, though, there does appear to be a negative correlation between lean mass and adiponectin levels in both sexes. 13 Thus, in elderly patients undergoing TAVR, adiponectin could provide useful prognostic data in terms of potential cognitive decline postprocedure or underlying frailty periprocedure. Adiponectin assessment could also help to determine those patients more likely to benefit from post-TAVR rehabilitation to improve outcomes.
Therefore, this study sought to determine whether plasma adiponectin levels could predict all-cause mortality after TAVI. We tested the ability and relative strength of adiponectin levels to predict all-cause mortality after TAVI in a multivariable model, that included clinical operative risk assessment, frailty assessment, vascular access of the TAVI procedure, CT imaging variables (LMD (%)) and adiponectin serum concentration.

| Patient selection
Between January 2014 and June 2016, patients referred for TAVR were included in the analysis. The recruitment of patients to this study partially overlapped with another ongoing frailty study. 14 Inclusion criteria included having a serum sample of adiponectin, severe aortic stenosis, and undergoing TAVR CT. Exclusion criteria included the inability to obtain informed consent.
The protocol was approved by the local research ethics board.
The study has been carried out in accordance with The Code of Ethics of the World Medical Association (Declaration of Helsinki).

| Data collection
The UZAntwerp TAVI registry is a comprehensive electronic database of patients undergoing TAVI. Patient baseline characteristics, TAVI procedural information, and outcome data (all-cause mortality and day of death) were retrieved from the registry (Table 1). In addition, the registry also recorded preoperative assessment risk scores (the logistic EuroSCORE, 3 EuroSCORE II, 15 and STS scores 4 ) and clinical frailty scores (Fried score, 16

| TAVI scanning protocol
The CT scanning protocol has been previously described. 14 Combined prospective ECG-triggered cardiac CT, chest, abdominal, and intrapelvic CT imaging acquisition to select the appropriate bioprosthetic size and assess the vascular access is part of the routine workup before TAVR.
All CT scans were performed on a dedicated CT scanner (GE lightspeed; General Electric Company, Easton Turnpike). Images were acquired using 80-120 kVp and automated tube current modulation.
A triphasic intravenous radiology contrast (Iomeron 350) injection protocol was used: 80 ml contrast agent at 4 ml/s, followed by 60 ml contrast agent at 4 ml/s plus 60 ml 0.9%NaCl solution at 2 ml/s, followed by 28 ml 0.9% NaCl solution at 2 ml/s). 14

| Assessment of the CT measured psoas muscle variables
The acquisition protocol of the psoas muscle variables has been described previously. 14 In summary, the psoas muscle area (PMA) was measured by contouring the psoas muscle border on a CT axial slice at the fourth lumbar vertebra. A filter, selecting the voxels with attenuation values between 0 HU and 100 Hounsfield Units (HU), determined the area attributed to the psoas muscle.
Subsequently, the low-density muscle portion (LDM(%)) was measured by selecting the voxels ranging from 0 HU to 29 HU within the psoas muscle area. LDM% expresses LDM as a fraction of the PMA, therefore not requiring adjustment to body surface area (BSA) ( Figure S1).

| Adiponectin serum concentration measurement
The serum concentration of adiponectin was measured using an enzyme-linked immunosorbent assay (ELISA) from blood samples from the TAVI-biobank. The Human HMW Adiponectin/Acrp30 Quantikine ® ELISA-kit (R&D Systems) was used to determine the serum adiponectin concentration.
The adiponectin concentration was measured twice, and the results were averaged. The coefficient of variation (CV), which is the ratio of the standard deviation/mean of both measurements, was calculated. The results of the adiponectin serum concentration measurements were rejected if the CV exceeded 15%.

| Follow-up duration
Patients were followed for a mean of 1779.0 ± 81.2 days. The median follow-up was 1825. Twenty-two deaths occurred during the followup period.
3.3 | Distribution of patient characteristics among patients with and without mortality post-TAVI Statistically significant differences in sex, hospital stay, operative risk assessment score, and psoas muscle assessment were observed between patients with and without mortality after TAVI (Tables 1   and 2

| Serum adiponectin
Compared with survivors, the serum adiponectin concentration was statistically significantly higher in patients with mortality after TAVI (

| Survival analysis
Univariable predictors of survival were calculated to determine those to be used in a multivariable cox regression analysis (   (Table 3A) and multivariable (Table 3B) Cox hazard regressions for all-cause mortality after TAVI   [20][21][22][23][24] Little is known about the ability of adiponectin to predict adverse outcome after TAVI due to severe aortic stenosis (AS).
Our study, with a mean follow-up of 1779 days, demonstrates that adiponectin serum concentration is an independent and incremental predictor of all-cause mortality after TAVI in males and females.
On the other hand, adiponectin serum level was not a predictor for mortality 30 days after TAVI. This finding may suggest that adiponectin is no predictor of TAVI-procedure related complications, despite the fact that adiponectin has been associated with aortic valve calcification.
Adiponectin has been associated with protective cardiovascular effects due to its anti-atherosclerotic and anti-inflammatory properties and ability to reduce insulin intolerance. 25 However, there is little data about the role of adiponectin in patients with severe AS. 25 Prior investigators have considered the potential for adiponectin to prevent adverse events through proposed anti-inflammatory actions in aortic stenosis. [20][21][22][23] The role of adiponectin to predict frailty per se was not considered. These investigators selected patients who were younger than our cohort and often were surgical AVR candidates, so patient frailty was less of a concern. 20-22 Our study differs from the aforementioned studies by the chosen endpoint and the variables, included in the multivariable model. Our study sought to answer whether serum adiponectin levels before TAVI predict all-cause mortality after TAVI. Furthermore, our multivariable model focused on including clinical variables, such as operative risk assessment (STSS), vascular access of the TAVI procedure, and a CT-derived psoas muscle attenuation variable (as a surrogate marker for sarcopenia).
Nevertheless, it should be noted that our study results do not align well with the current insight that adiponectin has mainly cardioprotective effects, as patients within the highest serum adiponectin concentration had the highest all-cause mortality. 25,26 This may reflect competing effects of adiponectin in the elderly, where perhaps its proposed antiinflammatory effects are less protective.
On the other hand, our study results are consistent with the findings from frailty studies, demonstrating that adiponectin levels are elevated in elderly cognitively impaired patients. [9][10][11] In contrast to earlier frailty studies, showing sex differences between adiponectin levels, frailty, and mental status, our results confirmed a strong and independent association between adiponectin serum concentration before TAVI and all-cause mortality after demonstrate an association between the adiponectin serum concentration and the TAVI-procedure-related 30-days mortality.
Age, sex, and cachexia have been associated with mortality after TAVI. [30][31][32] Age and sex are incorporated in the operative risk assessment score systems and thus embedded in our multivariable model.

| Study limitations
Three study limitations need to be discussed. First, it could be argued that the small sample size of our study may have altered the study results. On the other hand, the ability of adiponectin levels to predict all-cause mortality after TAVI in this limited study population indirectly proves the prognostic strength of serum adiponectin levels.
Our sample size is small with 77 patients being included in the analysis. The finding of an increased risk for those in the highest tertile of adiponectin is hypothesis-generating, and further largescale prospective studies are required to validate our observations.
All-cause mortality was used as the outcome variable in our study. This reflected the fact that frailty may lead to death through multiple causes, for example: fracture hip, aspiration pneumonia, and head injury. In this small sample size, further analysis of the potential effects of adiponectin on other causes of mortality such as cardiovascular events or sepsis was not possible.
Adiponectin levels may have been influenced by multiple metabolic factors, such as lipid profile, diabetes, smoking status, and medication use. These potential confounders could only partially be extracted from the database. To mitigate this limitation, diabetes, body mass index (BMI), and LDL cholesterol were introduced in a supplemental multivariable model (Supplemental Table 1). The overall conclusion that adiponectin is an independent predictor of all-cause mortality after TAVI remained unchanged. It should be emphasized that strong predictors of mortality after TAVI, such as age and sex are part of the STS Score and thus included in the multivariable model.

| CONCLUSION
Our study, with a mean follow-up of 1779 days, demonstrates that adiponectin serum concentration is an independent and incremental predictor of all-cause mortality after TAVI. Further prospective studies are warranted to determine the role of serum adiponectin levels to predict frailty and outcomes in patients with severe AS, who are being considered for TAVI.

ACKNOWLEDGMENT
This study did not receive a grant from any funding agency in the public, commercial or not-for-profit sectors. CoreValve.

DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available from Dr.
J. Bosmans upon reasonable request.