Antibody response after second BNT162b2 dose in allogeneic HSCT recipients

The prognosis of COVID-19 infection is poor in hematopoietic stem-cell transplant (HSCT) recipients.1, 2 In a large multicentric series of 318 HSCT recipients (184 allogeneic HSCT recipients and 134 autologous HSCT recipients), the probability of overall survival at 30 days after the diagnosis of COVID-19 infection was notably dismal, at 68% (95% CI 58–77) and 67% (55–78) for allogeneic HSCT recipients and autologous HSCT recipients, respectively.¹ Immunocompromised patients have been excluded from initial studies of SARS-CoV-2 mRNA vaccine efficacy, so the efficacy of vaccination in this population warrants evaluation.

To analyse the immunogenicity of the BNT162b2 mRNA vaccine (Pfizer–BioNTech), we used the IgG II Quant Assay (Abbot Laboratories, Wiesbaden, Germany) to quantify spike glycoprotein-specific IgG receptor-binding domain (IgG[S-RBD])levels at a median of 28 days (IQR 26–31) after the second vaccine dose in 88 recipients who had received two successive doses (at 4-week interval) at a median of 23 months (range 3–213 [IQR 9–30]) after allogeneic HSCT. IgG(S-RBD) titres could be quantified in 69 (78%) participants, whereas IgG(S-RBD) was detected but not quantifiable in three participants (anti-S titre <21 arbitrary unit [AU] per mL) and not detected in 16 participants (anti-S titre <6·8 AU/mL). In parallel, nucleoprotein-specific IgG was detected in seven of 88 participants, denoting previous SARS-CoV-2 exposure.

As previously reported for surrogate measure of vaccine protection, we stratified samples by IgG(S-RBD) titres above or below 4160 AU/mL as this threshold has previously been shown to correspond to a 0·95 probability of virus neutralisation in in-vitro plaque reduction neutralisation tests.³ In a comparison of characteristics of patients with IgG(S-RBD) titres above (n=52) and below (n=36) this threshold, a time interval greater than 12 months between HSCT and vaccination, as well as an absolute lymphocyte count in peripheral blood above 1G/L at the time of vaccination correlated with protective IgG(S-RBD) titres after vaccination (appendix). In comparison, participants who had received systemic immunosuppressive drugs within 3 months of vaccination had subprotective IgG(S-RBD) titres. Systemic immunosuppressive treatments within 3 months of vaccination, together with a lymphocyte count below 1 G/L in peripheral blood, remained independently correlated with low IgG(S-RBD) titres in multivariable analysis, whereas the correlation with the time interval between HSCT and vaccination was lost. With a median follow-up of 84 days (range 44–121 [IQR 65–110]) after the first vaccination dose, we did not observe any COVID-19 infection in this cohort.

In this first evaluation of immunogenicity in allogeneic HSCT recipients after two vaccine doses, we observed overall frequent and high levels of humoral responses, which contrasts with recent observations in solid organ transplant recipients who are receiving very long-term pharmacological immunosuppression.⁴ We identified lymphocyte count as well as recent pharmacological immunosuppression, rather than the sole timing of vaccination after HSCT, as determinants of humoral response. Our findings support the large scale vaccination of allogeneic HSCT recipients, although additional multicentre and long-term studies are needed to specify the level of immunological protection against infection, also taking into account the effect of a third vaccine dose in non-responding patients.

We declare no competing interests.