27. Cellular aging for risk stratification in adult deformity surgery: utilization of seven epigenetic clocks and two telomere length measurements in the analysis of comorbidity burden, frailty, disability and complications in adult deformity surgery

Abstract

<h3>BACKGROUND CONTEXT</h3> Recent advances in risk stratification have led to improvements in our ability to predict complications, cost and outcomes. Chronological age is important in almost all risk calculators; however, genetic or biological age may be more informative. Several techniques exist to assess genetic age including DNA methylation patterns with epigenetic clocks, telomere length and analysis of both mitochondrial DNA and single nucleotide polymorphisms. Our lab recently demonstrated a novel association between telomere length and 90-day complications in adult deformity surgery. There are no current risk stratification tools that integrate markers of cellular aging or biophysiologic reserve. Determining the optimal aging biomarkers to include in risk assessment models will be critical to future work in preoperative risk assessment. <h3>PURPOSE</h3> To assess the utility of epigenetic clocks and telomere length in evaluating comorbidity burden, frailty, disability and postoperative complications. <h3>STUDY DESIGN/SETTING</h3> Prospective cohort of adult deformity patients performed at a single academic center. <h3>PATIENT SAMPLE</h3> Adult patients with spinal deformity undergoing elective surgery. <h3>OUTCOME MEASURES</h3> Any medical or surgical complication within 90 days from surgery. <h3>METHODS</h3> Adult spinal deformity patients were prospectively enrolled. Demographics, frailty index (ASD-FI), Charlson comorbidity index (CCI), Oswestry Disability Index (ODI), and whole blood were collected along with presence of any medical or surgical complications at 90 days. CpG methylation was quantified using the Illumina EPIC array platform. Raw methylation data were processed using a custom pipeline in R to extract 7 epigenetic metrics: Horvath Age, Hannum Age, instantaneous pace of aging (DunedinPACE), and 4 principal component (PC) corrected epigenetic age metrics (PC Horvath, PC Hannum, PC PhenoAge, and PC GrimAge). Telomere length was measured directly by qPCR (TSR) or indirectly by methylation-based estimation (PC DNAmTL), which estimates telomere length using specific methylation sequences and correlation to leukocyte telomere length. Linear regression with biweight midcorrelation (bicor) was used to compare continuous variables. Logistic regression with odds radio (OR) was used for complications. Gender and BMI were treated as fixed-effects and added to each model. <h3>RESULTS</h3> A total of 43 patients were included with mean age of 65 years and 21 women (49%). A mean of 11 levels were fused posteriorly with 21 combined anterior-posterior approaches (43%) and 18 three-column osteotomies (42%). There were no significant differences in age, gender, BMI, preoperative ASA classification, CCI score, ODI, or surgical variables between patients who had complications compared to those without. CCI correlated with epigenetic age using 6 epigenetic clocks (Hannum Age, Horvath Age, PC Hannum, PC Horvath, PC PhenoAge, PC GrimAge) and estimated telomere length (PC DNAmTL). ASD-FI correlated with epigenetic age using the DunedinPACE epigenetic clock (bicor=0.439, p=0.003). ODI showed a trend toward correlation with epigenetic age using DunedinPACE, but did not meet statistical significance (bicor=0.262, p=0.098). Both direct and indirect telomere measures were associated with 90-day complications (TSR, OR=1,466, p=0.014 and PC DNAmTL, OR=279, p=0.010), while epigenetic age using the PC GrimAge clock showed only a trend (OR=1.09, p=0.067). <h3>CONCLUSIONS</h3> Epigenetic clocks seem to correlate with comorbidity (CCI) and disease-specific frailty (ASD-FI), while telomere length is significantly associated with complications. Perhaps most importantly, both direct and indirect measures of telomere length showed significant associations with complications, validating both approaches for this application. Integration of biomarkers into current risk calculators has significant potential for improved prediction accuracy, preoperative optimization and tailored surgical planning. <h3>FDA DEVICE/DRUG STATUS</h3> This abstract does not discuss or include any applicable devices or drugs.