Impact of Omicron BA.1 infection on BA.4/5 immunity in transplant recipients

Abstract

Mutations in the spike protein of SARS-CoV-2 have allowed Omicron subvariants to escape neutralizing antibodies. The degree to which this occurs in transplant recipients is poorly understood. We measured BA.4/5 cross-neutralizing responses in 75 mostly vaccinated transplant recipients who recovered from BA.1 infection. Sera were collected at 1 and 6 months post-BA.1 infection, and a lentivirus pseudovirus neutralization assay was performed using spike constructs corresponding to BA.1 and BA.4/5. Uninfected immunized transplant recipients and health care worker controls were used for comparison. Following BA.1 infection, the proportion of transplant recipients with neutralizing antibody responses was 88.0% (66/75) against BA.1 and 69.3% (52/75) against BA.4/5 (P = .005). The neutralization level against BA.4/5 was approximately 17-fold lower than that against BA.1 (IQR 10.6- to 45.1-fold lower, P < .0001). BA.4/5 responses declined over time and by ≥0.5 log₁₀ (approximately 3-fold) in almost half of the patients by 6 months. BA.4/5-neutralizing antibody titers in transplant recipients with breakthrough BA.1 infection were similar to those in immunized health care workers but significantly lower than those in uninfected triple-vaccinated transplant recipients. These results provide evidence that transplant recipients are at ongoing risk for BA.4/5 infection despite vaccination and prior Omicron strain infection, and additional mitigation strategies may be required to prevent severe disease in this cohort.

1. Introduction

Immunocompromised solid organ transplant recipients (SOTRs) are predisposed to severe COVID-19.1, 2, 3 In addition, common preventive strategies, including vaccine boosters, are less immunogenic in this population.^{4 , 5} Exogenous immunosuppression may also promote ongoing viral replication and the emergence of variants,⁶ highlighting the importance of understanding immune responses in this vulnerable cohort.

The rapid emergence of Omicron subvariants provides significant challenges for immunocompromised patients such as SOTRs. Omicron subvariants have significant alterations in the sequence of the spike protein compared with earlier strains or ancestral SARS-CoV-2. These changes facilitate the evasion of immune responses induced by both prior infection and immunization.7, 8, 9, 10 The BA.4 and BA.5 subvariants of Omicron, originally identified in early 2022, have isomorphic spike sequences, despite having different virologic and clinical characteristics, primarily owing to differences outside the spike region.¹¹ Both are thought to be more transmissible than prior Omicron subvariants. At the time of writing this article, the BA.5 subvariant is the dominant circulating subtype in many regions around the world. This is likely due to the ability of the BA.4/5 spike protein to evade neutralizing antibody responses to a greater degree than prior SARS-CoV-2 strains, including antibodies induced by natural infection with Omicron subvariants such as BA.1 and BA.2.12, 13, 14, 15, 16

It is unknown how antigenic escape by BA.4/5 impacts neutralizing antibody responses in immunocompromised persons. In this study, we evaluated heterotypic neutralizing antibodies to BA.4/5 in a cohort of 75 largely immunized transplant recipients who recovered from infection with Omicron BA.1. These responses were compared with those of health care worker controls and immunized SOTRs with no prior history of infection.

2. Materials and methods

2.1. Study design and ethics

This study was approved by the University Health Network research ethics board. All participants or their delegates provided informed consent. We recruited 3 cohorts. The primary cohort consisted of 75 SOTRs with Omicron BA.1 infection, enrolled from December 25, 2021, to January 31, 2022 (n = 75). The majority of this cohort was vaccinated with mRNA vaccine, representing breakthrough Omicron BA.1 infection/hybrid immunity: 3 (4.0%) were unvaccinated, 2 (2.7%) received a single dose, 12 (16.0%) were double-vaccinated, 53 (70.7%) were triple-vaccinated, and 5 (6.7%) received 4 doses prior to infection. The second cohort consisted of uninfected transplant recipients who received 3 doses of mRNA-1273 vaccine (Moderna; n = 60). A lack of infection was confirmed by the absence of an antinucleocapsid antibody and the absence of any prior positive viral detection test.¹⁷ The third cohort consisted of a control group of uninfected health care workers vaccinated with 3 doses of BNT162b2 (Pfizer) (n = 20). Infections were documented using SARS-CoV-2 nasopharyngeal swab polymerase chain reaction (PCR) or rapid antigen tests. Variant determination of samples was performed using C19-SPAR-Seq¹⁸ in the clinical microbiology laboratory. Serum was isolated from whole blood at approximately 1 month and 6 months after symptom onset in infected transplant recipients and 4 to 6 weeks after the third dose of vaccine in the vaccinated-only cohorts and cryopreserved for batch testing.

2.2. Pseudovirus neutralization assay

Neutralization assays against BA.1 and BA.4/5 were performed using previously validated spike-pseudotyped lentivirus assays with results expressed as log₁₀ ID50 (inhibitory dilution with 50% virus neutralization). To perform the pseudovirus neutralization assay,¹⁹ viral packaging (psPAX2; Addgene); spike protein constructs (Omicron BA.1 and Omicron BA.4/5 SARS-CoV-2, generated from consensus sequences in https://outbreak.info); and the ZsGreen and luciferase reporter (pHAGE-CMV-Luc2-IRES-ZsGreen-W), kindly provided by Jesse Bloom, were cotransfected into HEK293T cells (obtained from American Type Culture Collection, #CRL-3216). Viral supernatants were recovered 48 hours after transfection, and a viral titer assay for each pseudovirus was performed by infecting HEK293T-ACE2/TMPRSS2 cells.²⁰ Patient serum samples were diluted 1:22.5 and then serially diluted 3-fold over 7 dilutions, followed by incubation with the diluted virus at a 1:1 ratio for 1 hour at 37°C before adding to cells. The infected cells were lysed after 48 hours using the BrightGlo Luciferase Assay System (Promega), and luciferase activity was measured using a PerkinElmer Envision instrument (PerkinElmer) and expressed in relative light units. All cells were tested for Mycoplasma contamination and were found to be free of it.

2.3. Spike-specific T cell assessment

Peripheral blood mononuclear cells (PBMCs) were collected for T cell assessment at 1 month after symptom onset in patients infected with Omicron BA.1 who had sufficient cells available (n = 64). Cryopreserved PBMCs were thawed, rested, and incubated overnight with overlapping peptides corresponding to the Omicron BA.1 spike protein, as previously described.¹⁷ SARS-CoV-2-specific CD4+ and CD8+ T cell responses were measured using intracellular cytokine staining for the effector cytokines interleukin-2 (IL-2) and interferon-gamma (IFN-γ) by flow cytometry. Both monofunctional and polyfunctional (IL-2+ IFN-γ+) cells were characterized.

2.4. Statistical analysis

Descriptive statistics were used to outline the baseline characteristics of each cohort and have been described and analyzed elsewhere.¹⁷ ID50 titers were calculated in Prism version 9.4.1 (GraphPad Software) using a nonlinear regression (log[inhibitor] vs normalized response–variable slope) algorithm and converted to a log₁₀ scale. A positive neutralization assay was defined as any dilution that resulted in 50% viral neutralization, as calculated based on the above-generated curve. For the comparison of proportions of patients with a positive neutralization assay, chi-square or Fisher exact test was used. Differences in the medians of paired observations were tested using the Wilcoxon matched-pairs signed-rank test, and unpaired data were compared using the Mann-Whitney U test. The Spearman rank correlation test was used to measure the degree of correlation between parameters. P values of <.05 were considered significant. Statistical analyses and figure rendering were performed using Prism version 9.2.0.

3. Results

In the primary cohort of 75 transplant recipients who recovered from BA.1 infection, transplant types included kidney (n = 30; 40%), liver (n = 11; 14.7%), lung (n = 16; 21.3%), heart (n = 6; 8%), and others (n = 12; 16%), with a median time from transplant of 5.9 years (IQR, 2.3–9.9 years) (Table 1 ). The vaccinated cohort (n = 60) was also mostly kidney and kidney-pancreas transplant recipients. The most common immunosuppression regimen in both cohorts consisted of tacrolimus, prednisone, and mycophenolate. No patient in either cohort had received rituximab or belatacept in the 12 months prior to infection or vaccination. The median age of health care workers (n = 20) was 47 years (IQR 38–52), and the majority were female (75%).

Table 1.

Demographics and transplant characteristics of BA.1-infected and vaccinated transplant recipients.

Characteristics	BA.1 infected (N = 75)	Vaccinated uninfected (N = 60)
Age, y (median; IQR)	54 (47–64)	67 (64–72)
Female sex, no. (%)	26 (35)	23 (38)
Transplanted organ, no. (%)
Kidney or kidney-pancreas	39 (52)	35 (58.3)
Lung	16 (21.3)	11 (18.3)
Liver	11 (14.7)	3 (5)
Heart	6 (8)	10 (16.7)
Kidney-liver	3 (4)	1 (1.7)
Years since transplant (median, IQR)	5.9 (2.3–9.9)	3.8 (2.0–6.7)
Immunosuppression, no. (%)
Prednisone	64 (85.3)	50 (83.3)
Tacrolimus	63 (84)	47 (78.3)
Cyclosporine	10 (13.3)	12 (20)
Mycophenolate	62 (82.7)	44 (73.3)
Azathioprine	2 (2.6)	7 (11.7)
Sirolimus	1 (1.3)	2 (3.3)
Antilymphocyte globulin last 3 mo	1 (1.3)	0 (0)
Number of vaccine doses, no. (%)
0	3 (4)	0 (0)
1	2 (2.7)	0 (0)
2	12 (16)	0 (0)
3	53 (70.7)	60 (100)
4	5 (6.7)	0 (0)
Hospitalization related to COVID-19, no. (%)	11 (15)	—
Supplementary oxygen requirement related to COVID-19, no. (%)	6 (8)	—

All mRNA vaccines except 4 patients in the BA.1-infected cohort received mixed vaccinations with adenoviral vector vaccine followed by mRNA vaccine.

no., number.

Fig. 1A shows serum neutralization against BA.1 and BA.4/5 in the primary cohort of convalescent SOTRs. The proportion of transplant recipients with detectable neutralizing antibody responses was reduced from 88.0% (66/75) against BA.1 to 69.3% (52/75) against BA.4/5 (P = .005, chi-square test). The neutralization level against BA.4/5 was approximately 17-fold lower compared with that against BA.1 (median, 16.9; IQR, 10.6- to 45.1-fold lower; Fig. 1B; P < .0001; Wilcoxon rank-sum test); however, titers were correlated (Spearman r = 0.69; P < .0001; Supplementary Fig.). We further performed a univariate analysis of factors associated with the development of detectable cross-neutralizing antibodies (Table 2 ). Patients more likely to have BA.4/5 neutralization were those who were infected longer after transplantation (P = .016). Lung transplant recipients were significantly less likely to have cross-neutralizing antibodies after BA.1 infection (P = .003). This may reflect a greater cumulative degree of immunosuppression received by lung transplant recipients in general. In this cohort, 30 patients (40%) were treated with sotrovimab, but this was not associated with the detection of BA.4/5-neutralizing responses (P = .36). Six patients received remdesivir in this cohort. Eleven (14.7%) were hospitalized for a median duration of 4 days (IQR, 3–6 days), and 6 (8.0%) received supplemental oxygen. No patients were admitted to the intensive care unit or died.

Fig. 1.

BA.4/5 neutralizing antibody responses in organ transplant recipients and comparators. (A) Dot plots of the 50% neutralization titers (log₁₀ ID50) of neutralizing antibodies against Omicron BA.1 and BA.4/5 in mostly vaccinated transplant recipients 1 month after infection with Omicron BA.1. P values were used to compare medians and were determined using a 2-sided Wilcoxon matched-pairs signed-rank test. (B) Dot plot of the fold change in neutralizing antibody level observed for BA.4/5 compared with BA.1 level. The fold change is calculated by dividing the BA.1 ID50 value by the BA.4/5 ID50 value. For cases where the latter was 0, an arbitrary value of 0.1 log₁₀ ID50 was used. (C) Paired longitudinal comparison of BA.4/5 neutralizing antibodies at 1 and 6 months after BA.1 infection in a subset of patients (n = 21) who were positive for BA.4/5 neutralization at 1 month after initial BA.1 infection. (D) Dot plots comparing the log₁₀ ID50 for neutralizing antibodies at 1 month against BA.4/5 in patients infected with BA.1 (n = 75), triple-vaccinated transplant recipients (n = 60), and triple-vaccinated health care workers (n = 20). P values were determined using the Mann-Whitney U test. In dot plots, each dot represents an individual patient neutralization level. Horizontal lines represent median values.

Table 2.

Factors associated with detection of cross-neutralizing antibody to BA.4/5 in patients infected with Omicron BA.1.

Characteristics	BA.4/5 nAb not detected N = 23	BA.4/5 nAb detected N = 52	P-value
Age, y (median; IQR)	54 (47–62)	56 (46–64)	0.78
Female sex, no. (%)	10 (43.5)	16 (30.8)	0.28
Transplanted organ, no. (%)
Kidney or kidney-pancreas	9 (39.1)	30 (57.7)	0.14
Lung	10 (43.5)	6 (11.5)	0.003
Liver	1 (4.3)	10 (19.2)	0.16
Heart	2 (8.7)	4 (7.7)	0.99
Kidney-liver	1 (4.3)	2 (3.8)	0.99
Years since transplant, median (IQR)	3.6 (2.1–5.5)	7.1 (3.3–12.0)	0.016
Immunosuppression, no. (%)
Prednisone	18 (78.2)	46 (88.5)	0.29
Tacrolimus	21 (91.3)	42 (80.8)	0.32
Cyclosporine	2 (8.7)	8 (15.4)	0.50
Mycophenolate	17 (73.9)	45 (86.5)	0.20
Azathioprine	0 (0.0)	2 (3.8)	0.57
Sirolimus	1 (4.3)	1 (1.9)	0.99
Antilymphocyte globulin last 3 months	0 (0.0)	1 (1.9)	0.99
3 or more vaccine doses, no. (%)	17(73.9)	41 (78.8)	0.64
Hospitalization related to COVID-19, no. (%)	3 (13.0)	8 (15.4)	0.99
Supplementary oxygen requirement related to COVID-19, no. (%)	3 (13.0)	3 (5.8)	0.36

Continuous variables were analyzed using the Mann-Whitney U test, and categorical data were analyzed using the chi-square test.

nAb, neutralizing antibodies; no., number.

A longitudinal analysis was performed in a subset of SOTRs (n = 21) with detectable BA.4/5 responses and who agreed to provide follow-up serum samples. Longitudinal responses are shown in Fig. 1C. Neutralizing antibody levels against BA.4/5 waned over time, although they were still detectable in the majority of patients. BA.4/5 responses declined by ≥0.5 log₁₀ (approximately 3-fold) in almost half of the SOTRs measured by 6 months (9/21, 42.9%) (Fig. 1C).

Omicron BA.1-specific CD4+ and CD8+ T cells were previously described for this cohort in greater detail.¹⁷ Here we analyzed the polyfunctional (IL-2+ IFN-γ+) T cell response in those with positive (n = 45) or negative (n = 19) BA.5 neutralization. The median frequency of polyfunctional CD4+ T cells was 794 cells/10⁶ (IQR, 155–2093) in those with no detectable BA.5-neutralizing antibodies and 764 cells/10⁶ (IQR, 346–1816) in those positive for BA.5 neutralization (P = .74) (Fig. 2 ). Similarly, the median polyfunctional CD8+ T cell frequencies were 28 cells/10⁶ (IQR, 1–144) versus 35 cells/10⁶ (IQR, 1–36) in negative and positive BA.5 antibody groups, respectively (P = .91). Although BA.4/5-specific T cell responses were not measured, we and others have shown that SARS-CoV-2 T cell responses and epitopes are generally more conserved across variants compared with antibody responses.21, 22, 23, 24

Fig. 2.

Polyfunctional T cell frequencies in transplant recipients with and without cross-neutralizing antibody responses to BA.5 variant. Each dot represents an individual patient. Horizontal lines represent median and interquartile ranges. IL2, interleukin-2; IFN-γ, interferon-gamma.

Finally, a comparative analysis was performed between the primary study cohort of 75 SOTRs with mostly vaccine breakthrough BA.1 infection and a cohort of uninfected SOTRs immunized with 3 doses of mRNA vaccine and a control group of immunocompetent health care workers. The results indicate that a lower BA.4/5 neutralization titer was detected in the uninfected, vaccinated cohort (11.7%, 7/60; P < .0001; chi-square test), whereas 80% (16/20) of uninfected vaccinated health care workers had detectable BA.4/5-neutralizing antibodies (Fig. 1D). Antibody titers in the primary cohort of 75 SOTRs with hybrid immunity were similar to the levels in uninfected but vaccinated health care worker controls (P > .99).

4. Discussion

To our knowledge, this is the first published data on crossreactive immune responses to Omicron BA.4/5 in transplant recipients. The major findings of our study can be summarized as follows: although most SOTRs who recovered from BA.1 infection developed homotypic BA.1-neutralizing responses, a subset of these patients (69.3%) also mounted crossreactive neutralizing antibodies against the BA.4/5 spike protein. However, levels of neutralizing antibodies, expressed in terms of log₁₀ ID50, were approximately 17-fold lower against BA.4/5 compared with that against BA.1 but were comparable to triple-vaccinated health care workers. On the other hand, triple-vaccinated, uninfected SOTRs had minimal neutralization responses against BA.4/5 (11.7%), which were significantly lower than the other cohorts studied.

These results may have important implications regarding the risk of SARS-CoV-2 BA.4/5 infection in immunocompromised SOTRs in the ongoing BA.5 wave. Neutralizing antibodies correlate with protection from infection in studies of primates and immunocompetent adults.^{25 , 26} Given the lower levels of neutralizing antibodies and the lower proportion of SOTRs responding against BA.4/5, these results suggest that SOTRs remain at risk for Omicron BA.4 or BA.5 infection regardless of previous vaccination and infection status. The ability of BA.4/5 to evade neutralizing responses in immunized and previously infected immunocompetent cohorts has also been reported. Hachmann et al¹³ showed that 2 weeks after a vaccine booster and compared with the response against the ancestral WA1/2020 isolate, neutralizing antibody titers were 21.0-fold lower against BA.4/5 and 18.7-fold lower among those who were previously infected. Compared with the BA.1 subvariant, the BA.4/5 titer was reduced by a median factor of 3.3. Similar observations have been made in other studies.^{12 , 14 , 15} We show a similar reduction in comparative neutralization of BA.4/5 for transplant recipients.

The risk of contracting BA.4/5 may be partly mitigated in vaccinated SOTRs with prior BA.1 infection, where most (69.3%) of the crossreactive neutralizing antibodies against BA.4/5 are generated. The additive effect of natural infection plus immunization results in higher rates of crossprotective immunity compared with vaccination alone. Infection, however, is to be avoided because of the potential for severe disease, long COVID, and public health implications. Therefore, a potential solution to this may be specific BA.4/5 vaccine boosters that are anticipated to become available soon or passive immunization with monoclonal antibodies that are active against BA.4/5. We also showed in a subset of patients that longitudinal immunity to BA.4/5 after BA.1 infection appears to decline over a 6-month period. This may have implications for the timing of booster vaccines after infection.

Our study had some limitations. We acknowledge that most patients enrolled in this study experienced mild-to-moderate COVID-19. This could be due to improved outcomes because of vaccination and early treatment or could potentially reflect the undersampling of severe cases. Neutralization assays were not performed using live SARS-CoV-2 virus. However, the use of spike-pseudotyped lentiviral neutralization assays for SARS-CoV-2 immunogenicity is extensively reported in the literature and showed strong correlation to live-virus assays.²⁰ Another limitation is that early posttransplant patients were underrepresented in our study cohorts. Because early posttransplant patients are typically on greater immunosuppression, the results shown here may not be generalizable to that group of patients. Finally, the assessment of immune response just prior to or at the onset of infection would provide a more robust comparison, but such samples were not available to us and are generally more difficult to collect.

In summary, we demonstrate that SOTRs who recovered from BA.1 infection develop BA.4/5 cross-neutralizing responses, but at a significantly lower frequency and lower titer, with levels waning over time in most patients. In addition, triple mRNA vaccination on its own (in the uninfected transplant recipient cohort) induced minimal BA.4/5 neutralization. It is likely that during the current BA.5 wave, a significant portion of SOTRs remain predisposed to serious infection, despite vaccination and prior infection. Protective strategies, such as passive antibody prophylaxis or BA.4/5-specific boosters, which are now available in some countries, may be important to mitigate severe disease in this cohort.

Disclosure

The authors of this manuscript have conflicts of interest to disclose as described by the American Journal of Transplantation. D. Kumar has received a research grant from Roche and GSK and advisory fees from Roche, GSK, Sanofi, Merck, and Exevir. A. Humar has received research grants from Roche and Merck and advisory fees from Merck and Takeda. A.C. Gingras has received research funds from a research contract with Providence Therapeutics Holdings, Inc, for other projects. No other authors have conflicts of interest to disclose.

Data availability

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Appendix A. Supplementary data

The following is the Supplementary data to this article:

Multimedia component 1

Correlation of neutralizing antibodies (log₁₀ID50) against BA.1 compared with BA.4/5 in serum collected 1 month after BA.1 infection in transplant recipients (n=75). Each dot represents an individual patient. A Spearman correlation test was performed.