SARS-CoV-2 infection tracks closely with aplastic anemia onset in a single-center series
A retrospective analysis from University Hospital Tübingen links SARS-CoV-2 infection with the timing of aplastic anemia onset in a way that is biologically plausible, clinically concerning, and still short of proving causality. Across 65 patients diagnosed between 2009 and 2025, the annual case count rose modestly after the pandemic began, although the increase did not reach statistical significance. The sharper signal came from age: patients diagnosed from 2020 onward were older, with a median age of 64 years compared with 43 years before the pandemic.
The age shift matters at the bedside. Older patients bring more baseline marrow vulnerability, more clonal hematopoiesis, and more opportunities for viral immune activation to expose a limited hematopoietic reserve. In a disease already driven by immune-mediated stem cell destruction, any inflammatory trigger can tip the marrow from compensated cytopenia into frank failure.
Unrecognized infection was common near diagnosis
The in-depth cohort included 18 patients diagnosed during the early pandemic years, from 2020 through 2023. Only two reported a prior SARS-CoV-2 infection, yet 67% had anti-nucleocapsid antibodies, a pattern that points to a much higher rate of unrecognized infection than history alone suggested. Antibody kinetics indicated that most infections likely occurred close to the time aplastic anemia developed.
That temporal clustering is the central observation. It does not establish that the virus caused the marrow failure, but it does place viral exposure in the same window when immune dysregulation was unfolding. For a syndrome known to be triggered by infections such as hepatitis and herpes viruses, the association is difficult to ignore.
Inflammation was present, but not the same in every patient
Serological autoimmune testing was sparse and did not separate infection-associated cases from the rest of the cohort. Antinuclear antibodies, antineutrophil antibodies, antiphospholipid antibodies, anti-cardiolipin, anti-β2-glycoprotein I, anti-SSA, platelet antibodies, HLA class I antibodies, and anti-PF4/heparin antibodies were generally infrequent. Vaccination status also did not map onto a clear autoimmune pattern.
The cytokine data were more revealing. Irrespective of infection or vaccination history, patients showed a consistent inflammatory signature with markedly elevated TNFβ and IL-13. In three patients with serological evidence of infection, the profile broadened further, with pronounced activation of IL-1α, IL-6, and IL-10. Those cytokines fit the immunobiology of an active post-viral marrow assault: IL-1 and IL-6 can amplify innate and adaptive immune activation, while TNF-family signaling can suppress hematopoiesis and worsen progenitor loss.
One patient relapsed after a subsequent SARS-CoV-2 reinfection. The clinical course is too small to generalize, but it is hard to dismiss a second inflammatory hit in a marrow already primed for immune-mediated collapse.
Clonal hematopoiesis may mark the most vulnerable marrow
The three infection-associated patients with the strongest cytokine activation also had marked lymphocytic marrow infiltration and mutations linked to clonal hematopoiesis, including ASXL1, DNMT3A, and TET2. Those findings are important because clonal hematopoiesis can alter the inflammatory set point of the marrow microenvironment. A pre-existing clone may not cause aplastic anemia on its own, but it can shape how stem cells and immune cells respond when a viral trigger arrives.
In practical terms, this means SARS-CoV-2 may be less of a sole cause than a precipitating event in a biologically primed host. The marrow that fails after infection may already carry a mutational footprint that lowers the threshold for immune-mediated damage.
Vaccination did not emerge as the driver
Patients in the early-pandemic cohort had varied vaccination histories, including mRNA and viral vector vaccines, yet the analysis did not show a link between vaccination and autoimmune phenomena in this series. That point matters because the discussion around post-pandemic marrow failure can easily blur infection and vaccination into one indistinct exposure history. Here, the serology and timing point more convincingly toward infection, especially when anti-nucleocapsid antibodies are present, since those antibodies are not expected after vaccination alone.
The absence of a vaccination signal does not prove vaccines are irrelevant in every patient. It does mean this cohort did not support vaccination as the explanation for the inflammatory marrow phenotype observed during the pandemic years.
What the case series can and cannot prove
The study has the limits of a single-center retrospective design. The sample is small, the analyses are univariate, no multivariable model was fitted, and p values were not adjusted for multiple testing. The infection timing was inferred partly from antibody patterns rather than direct virologic confirmation, and several patients had incomplete molecular data. Those constraints leave causality unresolved.
Even so, the clinical pattern is coherent. During the pandemic, aplastic anemia cases were older on average, many carried serologic evidence of prior SARS-CoV-2 infection that had not been recognized clinically, and several of the most inflammatory cases also carried clonal hematopoiesis mutations. That combination suggests a three-part model: viral exposure, immune activation, and a vulnerable hematopoietic substrate.
For hematologists, the message is not that every new cytopenia after COVID-19 is aplastic anemia. It is that SARS-CoV-2 belongs on the list of possible triggers when marrow failure appears after a recent, even silent, infection. Future multicenter cohorts will need to test whether clonal hematopoiesis truly increases susceptibility, and whether infection-associated cases differ in trajectory, relapse risk, or response to immunosuppression.