Booster Shots Lift Antibodies, but Blood Disease Patients Lag Behind

People with hematologic diseases such as leukemia, myelodysplastic syndrome and aplastic anemia face a well-documented vulnerability to COVID-19, driven by both their underlying conditions and the immunosuppressive treatments they receive. Yet real-world evidence on how these patients actually fare on SARS-CoV-2 testing, and how well they respond to China’s widely used inactivated vaccines, has remained scarce. A new retrospective study from Peking University People’s Hospital, published in Frontiers in Public Health, addresses this gap by analyzing three independently identified populations: more than 20,000 individuals tested by SARS-CoV-2 RT-PCR, nearly 2,500 tested for anti-SARS-CoV-2 IgG antibodies, and almost 2,000 adults tested after receiving two or three doses of an inactivated vaccine.

The first population comprised 20,813 individuals who underwent RT-PCR testing between October 2023 and September 2025. Based on diagnoses recorded in electronic medical records, 1,341 were classified as having a hematologic disease and 19,472 as having none. Perhaps counterintuitively, recorded RT-PCR positivity was substantially lower among tested individuals with hematologic diseases: 4.77 percent versus 10.74 percent among those without, an unadjusted risk ratio of 0.44 (95 percent confidence interval, 0.35 to 0.57; p < 0.001). The authors caution, however, that because testing arose from routine clinical care, differences in testing indications, exposure, healthcare contact and protective behavior may have shaped this comparison. It should not be read as a difference in actual infection incidence or susceptibility.

The more telling analysis focused on viral clearance. Among RT-PCR-positive individuals, the researchers compared the interval from the first positive result to the first documented subsequent negative result. This analysis included only individuals with a documented subsequent negative result: all 64 of the 64 positive individuals with hematologic diseases, but only 381 of the 2,091 positive individuals without them. Among these 445 people, the median interval to documented negativity was 5.05 days (interquartile range, 3.26 to 10.66) in the hematologic disease group versus 3.99 days (IQR, 2.95 to 5.03) in the comparison group, a difference in medians of 1.06 days (95 percent CI, 0.73 to 2.21; p < 0.001). The authors note that this statistically significant difference was modest, and that its clinical importance remains uncertain because no established threshold defines a clinically meaningful difference for this outcome. The differential availability of negative results also warrants caution, since individuals without a documented negative result were excluded rather than censored.

Within the hematologic disease group, two subgroups stood out. Prior immunosuppressive treatment, defined as documented systemic use of agents such as corticosteroids, cyclosporine, mycophenolate mofetil, methotrexate or antithymocyte globulin within 12 months before the first positive result, was reported in 46 of the 64 patients. These patients had a median interval to negativity of 5.99 days (IQR, 4.03 to 11.35), compared with 3.97 days (IQR, 2.93 to 5.32) among the 18 patients without prior immunosuppressive treatment, an absolute difference of 2.02 days (p = 0.010). Transplantation status also mattered descriptively: among the 64 positive patients, 15 had undergone HLA-matched allogeneic hematopoietic stem cell transplantation (allo-HSCT), 30 had undergone haploidentical allo-HSCT, and 19 had not undergone transplantation. The corresponding median intervals were 4.77, 6.97 and 4.00 days, with haploidentical recipients showing the longest interval. The authors suggest this may reflect delayed immune reconstitution, graft-versus-host disease or continued immunosuppression, but stress that the finding is descriptive and that small subgroup sizes and differences in repeat-testing timing prevent firm conclusions about transplant type.

Mortality findings in this group were sparse. Three deaths were recorded during follow-up among the 64 RT-PCR-positive patients with hematologic diseases, all among patients aged 41 to 60 years, corresponding to a mortality proportion of 13.64 percent (3 of 22) in that age band and no deaths in other age groups. Two of the three deaths occurred among haploidentical allo-HSCT recipients, and the third occurred in a patient who had not undergone transplantation before infection but subsequently received an HLA-matched transplant. Because only three deaths occurred, the authors treat these observations as purely descriptive and explicitly caution against interpreting them as evidence of elevated mortality risk among middle-aged patients, noting that treatment intensity and transplantation status may have differed across age groups.

The second population included 2,438 individuals who underwent serum anti-SARS-CoV-2 IgG testing between October 2022 and September 2025, of whom 100 had a hematologic disease. Overall, 2,194 (90.00 percent) tested positive. IgG positivity was numerically lower among individuals with hematologic diseases (86.00 percent) than among those without (90.16 percent), an absolute difference of 4.16 percentage points, but the difference did not reach statistical significance (p = 0.174). Because infection and vaccination histories were not uniformly available in this population, the source and durability of these antibodies could not be determined, although impaired B-cell function associated with hematologic diseases and their treatments offers a plausible explanation for any reduction.

The third and primary population for vaccine analysis comprised 1,946 adults aged 18 or older who underwent antibody testing between January 19, 2021, and March 3, 2022, after receiving two or three doses of inactivated SARS-CoV-2 vaccines manufactured by the Beijing and Wuhan Institutes of Biological Products. Participants with documented prior SARS-CoV-2 infection were excluded. Of these, 1,092 had received two doses and 854 had received three. Neutralizing antibody (NAb) positivity, measured by a competitive chemiluminescence assay based on inhibition of binding between the viral receptor-binding domain and ACE2, was 94.1 percent in the three-dose group versus 77.3 percent in the two-dose group, an absolute difference of 16.8 percentage points. Anti-SARS-CoV-2 IgG positivity was likewise higher after three doses (97.2 percent versus 83.4 percent). After adjustment for age, sex, the interval from the most recent vaccine dose to antibody testing, and disease group, three doses remained associated with higher NAb positivity (adjusted prevalence ratio, 1.281; 95 percent CI, 1.232 to 1.332) and higher IgG positivity (aPR, 1.229; 95 percent CI, 1.189 to 1.271). Logistic regression sensitivity analyses yielded consistent results.

A small paired subanalysis lent within-person support to these cross-sectional findings. Of 171 participants with serial antibody measurements, only 20 had measurements available after both the second and third doses; among these 20, 19 showed higher NAb levels after the third dose. The authors describe this as an exploratory finding consistent with a stronger humoral response after the booster, but emphasize the small sample. Cross-sectional monthly comparisons, in which different participants contributed to each time category, showed NAb positivity in the two-dose group declining from 82 percent at month 1 to 33 percent by month 7, while the three-dose group ranged from 86 to 100 percent across examined intervals, with later estimates based on very few participants. Because the monthly categories comprised different people, these patterns describe population-level associations with time since vaccination rather than within-participant antibody decline, and the study could not define the duration or rate of antibody waning.

Disease group differences among three-dose recipients were striking. NAb positivity was 25 of 32 (78.13 percent) among participants with hematologic diseases, 36 of 42 (85.71 percent) among those with solid tumors, and 743 of 780 (95.26 percent) among those without documented hematologic disease or solid tumor (overall p < 0.001). Compared with the reference group, NAb positivity was significantly lower among participants with hematologic diseases (p = 0.001), while the solid-tumor subgroup result was considered exploratory owing to its limited size. The authors stress that antibody positivity does not directly indicate clinical protection, that the overall vaccination findings primarily reflect participants without hematologic disease or solid tumors and should not be directly generalized to hematologic patients, and that the hematologic subgroup of only 32 three-dose recipients limits precision.

The study’s limitations are substantial and acknowledged. It was single-center and retrospective; detailed treatment-related variables and measures of immunosuppression were unavailable, leaving potential residual confounding; the interval-to-negativity comparison was subject to differential follow-up and selection bias; and cellular immunity was not systematically evaluated. Nevertheless, the work adds real-world evidence that a third dose of inactivated vaccine is associated with markedly higher neutralizing and IgG antibody positivity, that most paired participants showed increased NAb levels after the booster, and that among RT-PCR-positive individuals with documented subsequent negativity, clearance intervals were modestly longer in patients with hematologic diseases, particularly after immunosuppressive treatment and among haploidentical transplant recipients. The authors argue these findings support closer virological and serological monitoring of hematologic disease patients and underscore the need for multicenter prospective studies with standardized longitudinal sampling, cellular immune assessments, viral variant characterization and long-term clinical outcomes to inform vaccination strategies for this vulnerable population.

Subject of Research: COVID-19 outcomes and antibody responses to inactivated vaccine boosters in patients with hematologic diseases

Article Title: Study on the clinical outcomes of COVID-19 infection and immune responses following the third booster dose of inactivated COVID-19 vaccine in hematologic disease patients

Article References: Zhang, H., Wang, X., Zhang, R., Liu, F., Chen, H., Jia, Y., & Rao, H. (2026). Study on the clinical outcomes of COVID-19 infection and immune responses following the third booster dose of inactivated COVID-19 vaccine in hematologic disease patients. Frontiers in Public Health, 14, Article 1917493. https://doi.org/10.3389/fpubh.2026.1917493

Image Credits: AI Generated

DOI: 10.3389/fpubh.2026.1917493

Keywords: COVID-19, SARS-CoV-2, hematologic diseases, inactivated vaccine, booster dose, neutralizing antibodies, RT-PCR, allo-HSCT, immunosuppression, IgG antibodies, retrospective study, Peking University People's Hospital