AUTHORS
Atsushi Marumo, MD, Arda Durmaz, PhD2, Maxwell Hou, N/A3, Aashray Mandala4, Christopher Haddad, BS5, Felipe de Almeida Sartori6, Karol K Gostomczyk7, Akram Abushamma, MD8, Serhan Unlu, MD9, Zachary Brady, BS10, Matteo D'Addona, MD11, Kartik Lakhotiya, MS12, Luca Guarnera, MD13, Nandini Rajaram Siva, MD14, Mark Orland, MD15, Carlos Bravo-Perez, MD16, Suresh Kumar Balasubramanian, MD17, Carmelo Gurnari, MD,PhD18, Babal Kant Jha, PhD19, Lukasz P. Gondek, MD, PhD20, Lara Jehi, MD21, Stephan Hutter, PhD22, Torsten Haferlach, MD, PhD23, Thomas LaFramboise24, Valeria Visconte, PhD25, Jaroslaw P. Maciejewski, MD, PhD, FACP26
AFFILIATIONS
1 Graduate School of Medicine, Nippon Medical School, Tokyo, Japan, Taussig Cancer Institute, Cleveland Clinic, Cleveland, Ohio, United States
2 Taussig Cancer Institute, Cleveland Clinic, Cleveland, Ohio, United States
3 Taussig Cancer Institute, Cleveland Clinic, Cleveland, Ohio, United States
4 Taussig Cancer Institute, Cleveland Clinic, Cleveland, Ohio, United States
5 Taussig Cancer Institute, Cleveland Clinic, Cleveland, Ohio, United States
6 Taussig Cancer Institute, Cleveland Clinic, Cleveland, Ohio, United States
7 Taussig Cancer Institute, Cleveland Clinic, Cleveland, Ohio, United States
8 Taussig Cancer Institute, Cleveland Clinic, Cleveland, Ohio, United States
9 Taussig Cancer Institute, Cleveland Clinic, Cleveland, Ohio, United States
10 Taussig Cancer Institute, Cleveland Clinic, Cleveland, Ohio, United States
11 Taussig Cancer Institute, Cleveland Clinic, Cleveland, Ohio, United States, University Hospital “San Giovanni di Dio e Ruggi d'Aragona”, Salerno, Italy
12 Taussig Cancer Institute, Cleveland Clinic, Cleveland, Ohio, United States
13 Taussig Cancer Institute, Cleveland Clinic, Cleveland, Ohio, United States, University of Rome Tor Vergata, Rome, Italy
14 Taussig Cancer Institute, Cleveland Clinic, Cleveland, Ohio, United States
15 Taussig Cancer Institute, Cleveland Clinic, Cleveland, Ohio, United States
16 Hospital Universitario Morales Meseguer, University of Murcia, IMIB-Pascual Parrilla, CIBERER - Instituto de Salud Carlos III, Murcia, Spain, Taussig Cancer Institute, Cleveland Clinic, Cleveland, Ohio, United States
17 Karmanos Cancer Institute, Wayne State University, Detroit, United States
18 Taussig Cancer Institute, Cleveland Clinic, Cleveland, Ohio, United States, University of Rome Tor Vergata, Rome, Italy
19 Cleveland Clinic Research, Cleveland, Ohio, United States
20 Sidney Kimmel Comprehensive Cancer Center, Johns Hopkins University, Baltimore, Maryland, United States
21 Cleveland Clinic, Cleveland, Ohio, United States
22 MLL Munich Leukemia Laboratory, Munich, Germany
23 MLL Munich Leukemia Laboratory, Munich, Germany
24 Case Western Reserve University, Cleveland, Ohio, United States
25 Taussig Cancer Institute, Cleveland Clinic, Cleveland, Ohio, United States
26 Taussig Cancer Institute, Cleveland Clinic, Cleveland, Ohio, United States
ABSTRACT
In addition to somatic mutations in HLA and PIGA, mutations in myeloid-associated genes are frequently detected in patients with bone marrow failure (BMF). Their age-adjusted prevalence has been reported to substantially exceed the one expected in the general population (PMID: 26132940). In addition to somatic mutations, heteroplasmy has been used as a clonal marker and thus also increases with age (PMID: 39578475). In AA, clonal hematopoiesis may be a function of the dissociation of the hematopoietic aging curve, i.e., biologic rather than chronologic age should be the proper contextual reference. Such a scenario may be due to the effects of pathological processes underlying BMF, which may ultimately represent a maladaptive form of somatic gene rescue. In this context, epigenetic age acceleration differential (AAD), defined as the difference between biological age estimated using DNA methylation clocks and chronological age, may serve as a useful biomarker (PMID: 42426219, 32580750). We tested the hypothesis that BMF, whether immune-mediated or chemically induced, may result in hematopoietic aging and thus affect the rate of clonal hematopoiesis, as reflected by heteroplasmy, myeloid-associated mutations, or immune-escape mutations.
To that end, we first analyzed AAD using DNA methylation arrays in patients with AA/PNH and control individuals and next analyzed AA/PNH and control groups using MitoHPC and estimate the frequency of heteroplasmy according to previously reported criteria (PMID: 39578475). Finally, samples from patients with AA/PNH were subjected to FLAER-based cell sorting, and AAD was evaluated in PNH-positive and -negative cell populations using nanopore sequencing. We used Horvath, Hannum, GrimAge2, epiTOC2, and DunedinPACE as biological age (PMID: 42426219, 32580750). Fisher’s exact test was used for categorical variables, and the Mann–Whitney U test was used for continuous variables. Propensity score matching was performed using age and sex as covariates, with a caliper width of 0.2
With these methodologies, we have found that AAD is higher in AA/PNH and thus a higher rate of clonal myeloid mutation and heteroplasmy-related clonality is also increased. In 30 propensity score–matched pairs, patients with AA/PNH had significantly higher Horvath AAD and DunedinPACE values than controls (mean Horvath AAD: 9.5 vs -0.6, P = 0.001; mean DunedinPACE: 1.2 vs1.0, P < 0.01). We next compared the frequency of heteroplasmy between 1,059 controls and 126 patients with AA/PNH across three age groups: ≤40, >40 to ≤70, and >70 years. Among controls, heteroplasmy was more frequent in older individuals (≤40 vs >70 years: 27% vs 40%, P < 0.05; >40 to ≤70 vs >70 years, 28% vs 40%, P < 0.001), whereas no significant age-dependent difference was observed in the AA/PNH group. In 124 propensity score–matched pairs, heteroplasmy and clonal myeloid mutation was significantly more frequent in patients with AA/PNH than in controls (heteroplasmy: 55% vs 22%, P < 0.0001; clonal myeloid mutation: 20% vs 1%, P<0.0001). Among 30 patients with AA/PNH at diagnosis, those with PNH clones and those harboring PIGA mutations had significantly lower DunedinPACE values than their respective counterparts (PNH positive vs negative, mean: 1.1 vs 1.3, P = 0.008 and PIGA+ vs PIGA-, mean:1.0 vs 1.2, P = 0.029). DunedinPACE and AAD did not improve upon hematopoietic recovery as seen in as serial samples, with a median follow-up period of 54 months. Finally, instead of chronologic, biologic age estimate was applied to six paired PNH-positive and -negative fractions. Although no significant differences were observed, the epiTOC2 value normalized to chronological age (estimated annual number of cell divisions) was significantly lower in PNH-positive cells than in PNH-negative cells (2.8 vs 16.6, P = 0.02).
Collectively, these findings indicate that both epigenetic age acceleration and heteroplasmy are increased in patients with AA/PNH compared with controls, suggesting accelerated biological aging. In contrast, within the AA/PNH group, PNH clones were associated with a slower pace of biological aging. Furthermore, sorted PNH-positive cells exhibited a lower estimated annual number of cell divisions, raising the possibility that escape from immune surveillance attenuates biological aging in PNH clones.
DISCLOSURES
Atsushi Marumo, MD: No relevant conflicts of interest to declare.