Attenuated humoral response against SARS‐CoV‐2 mRNA vaccination in allogeneic stem cell transplantation recipients

Abstract Antibody persistence several months after severe acute respiratory syndrome coronavirus 2 (SARS‐CoV‐2) mRNA vaccination in allogeneic stem cell transplantation recipients remains largely unknown. We sequentially evaluated the humoral response to two doses of mRNA vaccines in 128 adult recipients and identified the risk factors involved in a poor response. The median interval between stem cell transplantation and vaccination was 2.7 years. The SARS‐CoV‐2 S1 Ab became positive after the second vaccination dose in 87.6% of the recipients, and the median titer was 1235.4 arbitrary units (AU)/ml. In patients on corticosteroid treatment, the corticosteroid dose inversely correlated with Ab titer. Multivariate analysis identified risk factors for poor peak response such as an interval from stem cell transplantation ≤1 year, history of clinically significant CMV infection, and use of >5 mg/day prednisolone at vaccination. Six months after vaccination, the median titer decreased to 185.15 AU/ml, and use of >5 mg/day prednisolone at vaccination was significantly associated with a poor response. These results indicate that early vaccination after stem cell transplantation (<12 months) and CMV infection are risk factors for poor peak response, while steroid use is important for a peak as well as a persistent response. In conclusion, although humoral response is observed in many stem cell transplantation recipients after two doses of vaccination, Ab titers diminish with time, and factors associated with persistence and a peak immunity should be considered separately.


| INTRODUC TI ON
Allogeneic hematopoietic SCT recipients are immunocompromised and are at a high risk of developing critical COVID-19, which is caused by SARS-CoV-2. 1-3 Excellent efficacy and safety of two doses of mRNA vaccines in healthy individuals have been reported, 4,5 although chronological deterioration of immune responses in vaccinated people remains a matter of investigation. 6,7 Regarding the risk of aggravation of COVID-19 1,2,8 and the insufficient response after the second vaccination due to long-lasting immunodeficiency in and administration of immunosuppressive agents to SCT recipients, [9][10][11] pertinent evaluation of the peak response as well as persistence of vaccine response is essential for optimal patient care. However, previous studies have mainly focused on the peak titer or humoral response at a single time point, and the analysis of immune longevity, including determinants, is limited. 12 In addition, some studies have reported an association between immunosuppressive agent use and a poor vaccine response, 10,13 although the details remain unclear. The main objective of this study was to evaluate the humoral response to SARS-CoV-2 mRNA vaccines in 128 SCT recipients and the secondary objective was to identify the risk factors associated with poor peak and/or persistence of vaccine response in this population.

| Patients
In this single-center prospective observational study, patients older than 18 years of age who underwent allogeneic SCT at our hospital, received two doses of the SARS-CoV-2 mRNA vaccines (BNT162b2 or mRNA-1273), and agreed to participate in the study, were enrolled. The first dose of the vaccination was administered between June 2021 and November 2021, and only the BNT162b2 and mRNA-1273 vaccine types were available in Japan during the study period.
All SCT recipients who visited our hospital as outpatients were eligible for inclusion. However, the following were excluded: patients who were intravenous immunoglobulin preparation-dependent during vaccination, did not receive a second vaccination dose within 42 days after the first dose, and did not visit the hospital for sample collection. Patients undergoing prophylactic administration of chemotherapy were eligible for inclusion, but patients with hematological relapse were excluded. In general, the interval between the first and second doses of the vaccine was 3 weeks for BNT162b2 and 4 weeks for mRNA-1273. Serum samples were serially collected before vaccination, 2 weeks ± 7 days after the first vaccination dose, 5 or 6 weeks after the first vaccination dose (i.e., 2 weeks after the second vaccination dose), 3 months ± 1 month after the first vaccination dose, and 6 months ± 1 month after the first vaccination dose.
The detailed transplantation procedures have been previously described. 14

| Evaluation of anti-SARS-CoV-2 Ab titer
The concentration of anti-SARS-CoV-2 IgG was measured in serum samples at each time point using an iFlash 3000 chemiluminescence immunoassay analyzer (Shenzhen YHLO Biotech) with an iFlash-SARS-CoV-2 IgG kit and iFlash-SARS-CoV-2 IgG-S1 kit, as previously described. 16 The iFlash-SARS-CoV-2 IgG kit primarily detects antinucleocapsid Abs; 16 therefore, patients with positive SARS-CoV-2 IgG results were considered to have a previous history of COVID -19 and were excluded from analyses. The iFlash-SARS-CoV-2 IgG-S1 kit detects IgG specific to the S1 subunit of the spike (S) protein.
According to the manufacturer's instructions, samples with values ≥10 AU/ml were considered positive; in addition, a value >500 AU/ ml was defined as a high titer. The following conversion formula applied: YHLO IgG-S: AU/ml × 1 = BAU/ml.

| Patient characteristics
One hundred and forty-two patients were included in this study.
Of these, 14 were excluded from the analysis for the following reasons: history of COVID-19 before or within 14 days after vaccination  Acute lymphoblastic leukemia 22

Reduced intensity 56
Ab test result for these two patients was negative, we included these patients in subsequent analyses for SARS-CoV-2 S1 Ab titers after vaccination. The GMT of SARS-CoV-2 S1 Ab 2 weeks after the first vaccination dose, that is, before the second vaccination dose, was 6.82 AU/ml (median, 4.28 AU/ml).

| Antibody titers after second vaccination
The SARS-CoV-2 S1 Ab became positive 2 weeks after the second vaccination dose in 87.6% of the SCT recipients, and the GMT was 456.48 AU/ml (median, 1235.4 AU/ml) ( Figure 1A). The association between the Ab titers after the second vaccination dose and the clinical parameters was analyzed. The association of Ab titers with patients' age, lymphocyte count, and IgG levels at the first vaccination showed a correlation tendency ( Figure 1B-D).  Figure 1H). We noted that the Ab titer in nine patients who were treated with TKI or azacitidine at the time of the first vaccination seemed comparable with that of other patients, although the number was limited ( Figure S1G). Furthermore, SARS-CoV-2 S1 Ab titer according to the combination of immunosuppressive agents was also evaluated, although the number of patients in some subgroups was low ( Figure S2). The titer 2 weeks after the second vaccination dose was significantly higher in patients treated with calcineurin inhibitor alone than in those treated either with a combination of calcineurin inhibitor and corticosteroid or a combi-

| Diminishing vs. persistent Ab titers 6 months after vaccination
The GMT of SARS-CoV-2 S1 Ab 3 months after the first vaccination dose was 253.74 AU/ml (median, 581.70 AU/ml). The SARS-CoV-2 S1 Ab was positive in 83.6% of SCT recipients 6 months after first vaccination. However, the Ab titer significantly and gradually decreased after the second vaccination dose (Figure 2A), with the GMT reaching   Figure 2C to clarify differences in short-and long-term effects. Similar to the interval from the SCT, for some clinical factors the association with Ab titers at 6 months was no longer significant ( Figure 2C).

| Higher Ab titers after second vaccination
Considering the natural waning of Abs, a higher Ab titer 2 weeks after the second vaccination dose could be beneficial for persistent humoral immunity. The optimal cut-off was unknown; thus, we set SARS-CoV-2 S1 IgG > 500 AU/ml as the cut-off value because all patients with levels >500 AU/ml after the second vaccination dose remained SARS-CoV-2 S1 IgG positive 6 months after first vaccination.
The SARS-CoV-2 S1 Ab titer 2 weeks after the second vaccination dose was >500 AU/ml in 86 patients (67.2%). Clinical factors associated with SARS-CoV-2 S1 IgG > 500 AU/ml 2 weeks after the second vaccination were quite different from those associated with levels  (Table S1). Interval from SCT p = 0.0075

| COVID-19 breakthrough infection after vaccination
Five patients were diagnosed with COVID-19 at least 2 weeks after the second vaccination, and the 300-day cumulative incidence of COVID-19 breakthrough infection was 7.4% ( Figure S4)

| DISCUSS ION
In this study, humoral response was achieved in more than 85% of SCT recipients after two doses of the SARS-CoV-2 mRNA vaccine.
This is the largest study to evaluate humoral response of mRNA vaccine in Japanese allogeneic SCT recipients, and the response rate was comparable with that reported in previous studies. [9][10][11]13,[18][19][20] These studies have identified that some clinical characteristics, such as lymphocytopenia, hypoglobulinemia, and a shorter interval from the SCT, were associated with a poor response. We further clarified that treatment with corticosteroid was important for poor humoral response and that the dose of corticosteroid also had a significant impact.
We also showed that many of the parameters were not signifi- Interestingly, a history of csCMVi was associated with a poor humoral response in this study. These results might reflect impaired immunity in the recipients. However, CMV reactivation is associated with modified immunity 25-28 and variable infections. [29][30][31] In addition, CMV reactivation has been reported to be associated with a poor response after vaccination against hepatitis B virus in SCT recipients. 32 These reports suggest an association between CMV infection and impaired acquired immunity against other viruses approximately 1 year later. The median interval from csCMVi to vaccination in our cohort was approximately 3 years; therefore, the possible immunological impact of CMV reactivation lasts for a few years, although further validation using other cohorts and elucidation of the underlying mechanisms are warranted.
The adverse effects of vaccination in SCT recipients were generally mild in this cohort. Adverse effects in SCT recipients have been reported to be generally tolerable. [33][34][35] Nonetheless, careful follow-up is warranted because severe adverse effects after SARS-CoV-2 vaccination could occur in SCT recipients. 36,37 Five breakthrough infections were noted in our study, and all occurred more than 6 months after the first vaccination dose.
TA B L E 2 Univariate and multivariate analyses for SARS-CoV-2 S1 Ab positivity after second vaccination  and that humoral immunity against variants of concern including the Omicron strain can be enhanced with a third dose. 40 Further analyses identifying factors associated with the response and persistence of cellular immunity in SCT recipients are warranted. Second, neutralizing Abs were not analyzed. However, previous reports showed that the SARS-CoV-2 S1 Ab titer correlated well with neutralizing Ab titers 46 ; therefore, we considered that the SARS-CoV-2 S1 Ab titer was a useful and practical marker in assessing SARS-CoV-2 humoral immunity. Third, our analyses of adverse effects were based on a questionnaire survey and thus F I G U R E 2 Clinical characteristics associated with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) S1 Ab titers and their persistence in allogeneic stem cell transplantation (SCT) recipients. (A) Box plot of transition of SARS-CoV-2 S1 Ab titer. (B) Plot of SARS-CoV-2 S1 Ab titers after second vaccination and 6 months after vaccination. (C) Univariate analyses of SARS-CoV-2 S1 Ab >10 AU/ml after second vaccination (left column) and 6 months after vaccination (right column) and clinical parameters. Characteristics with p value <0.1 are shown. The color of circles indicates odds ratio and the size indicates p value. (D) Univariate analyses of SARS-CoV-2 S1 Ab >500 AU/ml after second vaccination. ATG, antithymocyte globulin; AU, arbitrary unit; csCMV, clinically significant CMV; cGVHD, chronic graft-versus-host disease; Lym, lymphocyte count; MMF, mycophenolate mofetil.  In conclusion, a positive humoral response after two vaccination doses was observed in 87.6% of SCT recipients. A higher dose of corticosteroid given around the time of vaccination was associated with a poor response. A shorter interval after SCT, history of csCMVi, and steroid use were significantly associated with SARS-CoV-2 S1 IgG < 500 AU/ml after the second vaccination dose.
Steroid use was also associated with a lack of persistent response.
Finally, the SARS-CoV-2 mRNA vaccine appears to be safe for patients with SCT.

ACK N OWLED G M ENTS
We thank all the patients who participated in this study. We thank Fumihiko Yasui for his kind and valuable suggestions regarding the study design, and Junko Ohta and Atsuko Yoshida for attentive sample management. We also thank the medical and laboratory staff at

D I SCLOS U R E
The authors declare no conflict of interest.

E TH I C S S TATEM ENT
Approval of the research protocol by an institutional review board: