Evaluating association between epigenetic clocks and cancer survival among Black women diagnosed with high grade serous ovarian cancer Open Access

Rangaswamy Nandakumar, Dhanish Revanth (Spring 2023)

Permanent URL: https://etd.library.emory.edu/concern/etds/1c18dh21x?locale=en


Background:  Epigenetic clocks are tools used to estimate a person's epigenetic age based on DNA methylation levels, and numerous studies have demonstrated their effectiveness in predicting the risk of cancer mortality and progression. Despite their proven efficacy, the association between epigenetic clocks and mortality in Black women with high-grade serous ovarian cancer remains largely unexplored.


Methods: We conducted a study on 202 Black women with High Grade Serous Ovarian Cancer (HGSOC) from the AACES and NCOCS study cohorts. We calculated the mitotic age using Epitoc2 and MiAge and categorized these measures as dichotomous and trichotomous exposure variables. We also analyzed the correlation between the mitotic ages calculated from the clocks and chronological age and used ANOVA to investigate any differences in the mitotic ages across the risk factors. To evaluate the association between mitotic age and survival, we fit Cox proportional hazard models for different sets of covariates.


Results: The cohort had a median age of 57 years and median follow up of 4.4 years and 110 subjects had died. The median mitotic age using Epitoc and MiAge was 64.4 with annual cell turnover rate and 820.9 relative mitotic age, respectively. We observed that mitotic age obtained from Epitoc decreased with increasing chronological age and had a higher value for localized stage, whereas MiAge-derived mitotic age did not vary with age but had a higher value for localized stage. Cox proportional hazard models between MiAge-derived mitotic age and survival did not show a significant association. However, for Epitoc-derived mitotic age categorized into tertiles, we found a significant difference in survival between the Q2 and Q1 tertiles in all models (HR = 0.61, [95% CI: 0.35, 1.05; adjusting for age, stage, and debulking status], HR = 0.50 [95% CI: 0.29, 0.88; additionally adjusted for hormone use], HR = 0.51 [95% CI: 0.31, 0.84; adjusting for age and stage], and HR = 0.57 [95% CI: 0.33, 0.98; adjusting for age and debulking status].


Conclusion: The association between mitotic age and survival is not linear. Epitoc2 appears to be a more useful clock in risk stratification. A larger study is required to identify the threshold and evaluate the true relation between mitotic age and survival.

Table of Contents


1.1 Descriptive epidemiology

                       Table 1 Incidence and mortality rate across different racial groups in the US

           1.2 Pathological classification

                       Table 2: Characteristics of the most common histotypes of EOC

1.3 Factors associated with survival

1.4 Epigenetics and Cancer

1.5 Epigenetic age and clocks

1.6 Rationale for using mitotic clocks in HGSOC


           2.1 Data source

           2.2 Methylation data quality control

           2.3 Exposure, outcome and covariates

           2.4 Mitotic age calculators

2.5 Statistical Analysis


           3.1 Descriptive Statistics

           Table 3: Table 3: Description of the study population (significance at p < 0.05)

           3.2 Correlation between chronological age and mitotic age

           3.3 Mitotic age acceleration between different risk factors

           Table 4: Comparison of the mitotic age acceleration between age groups,

smoking groups and stage for the two mitotic clocks:

           3.4 Univariate Kaplan Meier Analysis

           3.5 Multivariable survival analysis (Cox-Proportional hazard)

                       3.5.1 Mitotic age as a dichotomous variable

                       3.5.2 Mitotic age as a trichotomous variable

                       Table 5: Cox proportional hazard ratio, 95% confidence interval and

their p values are tabulated below for the models along with mitotic age

calculated using EpiToc2 and MiAge: a) Exposure as a dichotomous variable

and b) exposure as a trichotomous variable.

           3.6 Sensitivity Analysis

Table 6: Comparing the HR estimates with only AACES samples and with both AACES and NCOCS combined




About this Master's Thesis

Rights statement
  • Permission granted by the author to include this thesis or dissertation in this repository. All rights reserved by the author. Please contact the author for information regarding the reproduction and use of this thesis or dissertation.
  • English
Research Field
Committee Chair / Thesis Advisor
Last modified

Primary PDF

Supplemental Files