Immunogenicity of COVID-19 vaccines in solid organ transplant recipients: a systematic review and meta-analysis

Abstract

Background

Solid organ transplant (SOT) recipients are at increased risks of morbidity and mortality associated with COVID-19.

Objectives

This study aimed to evaluate the immunogenicity of COVID-19 vaccines in SOT recipients.

Data sources

Electronic databases were searched for eligible reports published from 1 December 2019 to 31 May 2022.

Study eligibility criteria

We included reports evaluating the humoral immune response (HIR) or cellular immune response rate in SOT recipients after the administration of COVID-19 vaccines.

Participants

SOT recipients who received COVID-19 vaccines.

Assessment of risk of bias

We used the Newcastle-Ottawa scale to assess bias in case-control and cohort studies. For randomised-controlled trials, the Jadad Scale was used.

Methods

We used a random-effects model to calculate the pooled rates of immune response with 95% CI. We used a risk ratio (RR) with 95% CI for a comparison of immune responses between SOT and healthy controls.

Results

A total of 91 reports involving 11 886 transplant recipients (lung: 655; heart: 539; liver: 1946; and kidney: 8746) and 2125 healthy controls revealed pooled HIR rates after the 1st, 2nd, and 3rd COVID-19 vaccine doses in SOT recipients were 9.5% (95% CI, 7–11.9%), 43.6% (95% CI, 39.3–47.8%) and 55.1% (95% CI, 44.7–65.6%), respectively. For specific organs, the HIR rates were still low after 1st vaccine dose (lung: 4.4%; kidney: 9.4%; heart: 13.2%; liver: 29.5%) and 2nd vaccine dose (lung: 28.4%; kidney: 37.6%; heart: 50.3%; liver: 64.5%).

Conclusions

A booster vaccination enhances the immunogenicity of COVID-19 vaccines in SOT; however, a significant share of the recipients still has not built a detectable HIR after receiving the 3rd dose. This finding calls for alternative approaches, including the use of monoclonal antibodies. In addition, lung transplant recipients need urgent booster vaccination to improve the immune response.

Graphical abstract

Introduction

COVID-19 significantly increases the risk of severe diseases and death in solid organ transplant (SOT) recipients because of underlying immunosuppression and concomitant comorbidities [1]. The hospitalisation and mortality rates of COVID-19 vary from 26% to 63% [2,3] and 13% to 30% [4] in SOT recipients, respectively.

It has been determined that COVID-19 vaccines are the most effective way to control the pandemic [5,6]. By July 2022, over 10 billion vaccination doses have been administered worldwide. The immunogenicity of COVID-19 vaccine has been well demonstrated in the general population in large-scale phase III trials [[7], [8], [9]]. Unfortunately, initial trials for these vaccines did not include SOT recipients. Recent studies have demonstrated reduced immune response in transplant recipients, however, results among studies vary widely [[10], [11], [12], [13]]. Therefore, it is necessary to integrate these findings to better understand the immunogenicity of the COVID-19 vaccine in these populations. The purpose of this meta-analysis was to assess the immunogenicity of the COVID-19 vaccine in recipients of SOT.

Methods

Overview

This systematic review and meta-analysis was conducted using the Meta-analyses of Observational Studies in Epidemiology [14] guidelines. This report was preregistered and submitted to the PROSPERO (CRD42022311886).

Data sources and searches

We performed a search in electronic databases (PubMed, Web of Science, Cochrane Library, and Embase) from 1 December 2019 to 31 May 2022. The search strategy details are shown in Table S1. The following terms were identified: ‘COVID-19’, ‘SARS-CoV-2’, ‘vaccine’, ‘vaccination’, ‘solid organ transplant’, ‘transplantation’, ‘liver transplant (LIT)’, ‘kidney transplant (KT)’, ‘heart transplant (HT)’, ‘lung transplant (LUT)’, ‘pancreas transplant’, ‘immunogenicity’, ‘seroconversion’, ‘humoral immune response (HIR)’, and ‘cellular immune response (CIR)’. Two reviewers independently performed the search, and a third reviewer resolved disagreements.

Study selection

Articles were included irrespective of the publication format (original papers, comments, abstracts, and letters). There were no language restrictions. We included reports evaluating at least one of the key outcomes. Reports with less than 10 patients were excluded.

Outcomes of interest

The primary outcomes of this study included HIR rates among SOT recipients and the comparison of HIR between SOT recipients and healthy controls. The secondary outcomes included CIR rates and comparisons of CIR between individuals in SOT and healthy controls.

Data extraction and quality assessment

We screened all eligible studies to extract the details of the first author, publication year, study design, gender, age, sample size, vaccine types, organ types, number of doses administered, the time post-transplant, the measure of immunogenicity, the time interval between the 2nd and 3rd dose, and the outcomes of interest. We contacted the authors for the missing data.

The Newcastle-Ottawa scale [15] was used in case-control studies, which was also suitable for the cohort studies. For the randomised-controlled trials (RCTs), the Jadad Scale [16] was used. The Newcastle-Ottawa scale ranges from 0 to 9, and the Jadad scale ranges from 0 to 5, with higher scores indicating a reduced risk of bias. Grades of Recommendation, Assessment, Development and Evaluation (GRADE) [17] was conducted to assess the reliability of evidence for the outcomes of interest.

Data synthesis and analysis

We used Stata version 12.0 to perform the meta-analysis. The random-effects model was used to estimate the pooled rates with 95% CI of immune response after the COVID-19 vaccine in SOT recipients. We used a risk ratio (RR) with 95% CI for a comparison of immune responses between SOT and healthy controls. We calculated pooled rates and RR using the ‘metaprop’ and ‘metan’ commands in Stata software, respectively. Heterogeneity was defined as mild (I² = 0–25%), moderate (I² = 26–75%), or considerable (I² >75%). Sensitivity analysis was performed by removing one study at a time to confirm the robustness of outcomes [18]. Subgroup analyses were conducted according to the organ type, type of vaccine, time post-transplant, the time interval between the 2nd and 3rd dose, and vaccine dosage (one-, two-, or three-dose). We did not perform subgroup analysis if the number of included reports was less than five.

A mixed-effects meta-regression analysis was conducted to assess the source and magnitude of heterogeneity. The dependent variable may be moderated by the following variables: (a) type of vaccine, (b) time post-transplant, (c) quality of studies, and (d) study sample size. If the number of included studies was more than ten, an assessment of publication bias was performed using Egger's test and funnel plots.

Results

Characteristics of included studies

Our electronic database search identified 2073 articles. One hundred fifty-one potentially relevant articles were identified for full-text evaluation after removing duplicate articles and critically evaluating abstracts and titles (Fig. 1 ). After applying the eligibility criteria, 91 [5,6,[10], [13],[19], [20], [21], [22], [23], [24], [25],[26], [27], [28], [29], [30], [31], [32], [33], [34], [35],[36], [37], [38], [39], [40], [41], [42], [43], [44], [45],[46], [47], [48], [49], [50], [51], [52], [53], [54], [55],[56], [57], [58], [59], [60], [61], [62], [63], [64], [65],[66], [67], [68], [69], [70], [71], [72], [73], [74], [75],[76], [77], [78], [79], [80], [81], [82], [83], [84], [85],[86], [87], [88], [89], [90], [91], [92], [93], [94], [95],[96], [97], [98], [99], [100], [101], [102], [103]] articles were included in the quantitative synthesis, of which two were RCTs. This meta-analysis involved a total of 11 886 transplant recipients (LUT: 655, HT: 539, LIT: 1946, and KT: 8746) and 2125 healthy controls. The details of the included studies are shown in Table 1 .

Fig. 1.

Flow diagram of the study selection process.

Table 1.

Characteristics of the included studies in systematic review and meta-analysis of immune response to COVID-19 vaccines in SOT recipients

Study	Population	Sample size	Male/Female	Age (y),mean ± SD/median (IQR/Range#)	Time post-transplant (y), mean ± SD/median (IQR/Range#)	Type of vaccine	Immunogenicity detection method	Cutoff value for positivity	Detection time, mean ± SD/median (IQR/Range#)	Vaccine Dose	Time interval between 2-3 dose (d), mean ± SD/median§(IQR)	Immunotherapy drugs/regimen	NOS/Jadad
Alejo 2021 [20]	SOT KT LIT HT	18	9/9	54 (42–73)	4 (0.5–21)#	BNT162B2 mRNA-1273 JNJ-78436735	ECLIA ELISA	50 U/mL (ECLIA) 4 AU/mL (ELISA)	2–6 wk#	2/3/4	67 (54–81)	NA	4
Azzi 2021 [21]	KT	76	45/31	62 (52–70)	3.9 (1.6–8.7)	BNT162B2 mRNA-1273 JNJ-78436738	CLIA	Signal-to-cutoff ratios >1	NA	2	NA	CNI (98.7%), MPA/MMF (80.2%), Prednisone (100%)	5
Ben-Dov 2022 [22]	KT HC	252 71	168/84 25/46	53.5 ± 14.4 43.6 ± 14.3	4 (0.3–49)# NA	BNT162b2	CLIA	59 AU/mL	10–20 d/2–6 wk# 8–117 d#	1/2	NA	Tacrolimus (83%), Everolimus (9%), CyA (7%), Sirolimus (1%) NA	7
Ben-Dov 2022 [23]	KT HC	118 82	NA	NA	NA	BNT162b2	CLIA	59 AU/mL	NA	3	180§	NA	5
Benotmane 2021 [19]	KT	159	98/61	57.6 (49.6–66.1)	5.3 (1.9–11.1)	mRNA-1273	CLIA	50 AU/mL	28 (27–33) d	3	51 (48–59)	Tacrolimus + MMF/MPA + steroids (84%)	6
Benotmane 2021 [24]	KT	241	156/87	57.7 (49.3–67.6)	6.4 (2.9–13)	mRNA-1273	CLIA	59 AU/mL	4 wk	1	NA	CNI (89.2%), MPA/MMF (79.3%), AZA (2.9%), Belatacept (3.8%), mTORi(14.5%), and Steroids (58.9%)	6
Benotmane 2021 [25]	KT	204	130/74	57.7 (49.4–67.5)	6.2 (3–12.8)	mRNA-1273	CLIA	59 AU/mL	4 wk	2	NA	CNI (89.2%), MPA/MMF (78.9%), AZA (2.9%), Belatacept (2.5%), mTORi (13.2%), and Steroids (59.8%)	6
Bergman 2021 [26]	SOT HC	89 90	45/44 39/51	NA	NA	BNT162b2	ECLIA	0.80 U/mL	2 wk 35 d	2	NA	NA	8
Bertrand 2021 [27]	KT	45	23/22	63.5 ± 16.3	6.9 (0.22–30.2)#	BNT162b2	CLIA	59 AU/mL	3/4 wk	1/2	NA	Tacrolimus (53.3%), CyA (17.8%), MMF (82.2%), AZA (8.9%), Everolimus(6.7%), Belatacept (22.2%), and Steroids (46.7%)	5
Boyarsky 2021 [10]	SOT KT LIT HT LUT Pan	658	267/391	NA	NA	BNT162B2 mRNA-1273	ECLIA ELISA	0.80 U/mL 1.1 AU/mL	NA	2	NA	Antimetabolite (71.9%)	6
Boyarsky 2021 [28]	SOT KT LIT HT LUT	436	160/266	55.9 (41.3–67.4)	6.2 (2.7–12.7)	BNT162B2 mRNA-1273	ECLIA ELISA	0.80 U/mL 1.1 AU/mL	20 (17–24) d	1	NA	Antimetabolite (73.4%)	6
Bruminhent 2021 [29]	KT HC	35 38	21/14 7/31	50 (42–54) 39 (34–42)	4.5 (2–9.5) NA	CoronaVac	CLIA	50 AU/mL	4/2 wk	1/2	NA	Tacrolimus (63%), CyA (31%), MMF (60%), MPS (37%), Sirolimus (3%), Everolimus (3%), and Prednisolone (97% NA)	7
Buchwinkler 2021 [30]	KT	216	147/69	59.9 (50.7–68.5)	NA	BNT162B2 mRNA-1273	CLIA	13 AU/mL	4 wk	1/2	NA	Tacrolimus (70%), CyA (18%), AZA (15%), MPA (69%), Belatacept (6%), Glucocorticoids (82%), and mTORi (5%)	7
Cao 2021 [31]	SOT LUT HT	37	10/27	64 (50–69)	36 (1–365)#	BNT162B2 mRNA-1273	CLIA	50 AU/mL	21 (19–25) d	2	NA	Tacrolimus (81.1%), CyA (13.5%), Everolimus (2.7%), and Sirolimus (2.7%)	5
Chavarot 2021 [32]	KT	62	36/26	63.5 (51–72)	3.9 (2.1–6.6)	BNT162b2	CLIA	50 AU/mL	28 (28–33) d	3	NA	CNI (3%), Belatacept (100%), Steroids (100%), mTORi (Everolimus) (13%), MPA (71%), and AZA (5%)	7
Chavarot 2021 [33]	KT	101	68/33	64 (53–73)	4.9 (2.4–8.7)	BNT162B2 mRNA-1273	CLIA	50 AU/mL	28 d	1/2	NA	CNI (7.9%), Belatacept +MPA(78.2%), Steroids (96%), mTORi(11.9%), and AZA (2%)	5
Cholankeril 2021 [34]	LIT	69	48/21	63 (51–68)	3.3 (1.7–8.3)	BNT162b2	ELISA	Titer >20	30–70 d#	2	NA	Tacrolimus (90%), MMF(36%), and Prednisone(37%)	6
Correia 2022 [35]	KT	131	85/46	59.3 ± 9.6	9.2 ± 7.2	BNT162b2 mRNA-1273 AZD1222 JNJ-78436735	CLIA	50 AU/mL	20.3 ± 6.2 d	2	NA	Antimetabolite (64.1%)	6
Cotugno 2022 [36]	SOT HC	34 36	NA	14.6 ± 7.2 NA	14.7 ± 7.2 NA	BNT162B2	CLIA	NA	1 wk NA	2 NA	NA NA	NA	5
Crane 2021 [37]	KT	25	14/11	19 (18–20)	5 (4–9)	mRNA-1273 BNT162b2	CLIA	50 AU/mL	4 wk	2	NA	MMF (84%) and AZA (12%)	5
Crespo 2021 [38]	KT HC	90 32	55/35 5/27	59.7 ± 12.5 52.7 ± 10.7	NA	mRNA-1273 BNT162b2	CLIA	13 AU/mL	4 wk	2	NA	Prednisone (91.1%), MPA derivatives (54.4%), Tacrolimus (91.1%), CyA (3.3%),and mTORi (24.4%) NA	7
Cucchiari 2021 [39]	KT	117	79/38	57.62 ± 14.32	1.7 (0.8–4.9)	mRNA-1273	ELISA	Titers >1	2 wk	2	NA	Antithymocyte Globulins (11.5%), Rituximab (2%), Tacrolimus (84.5%), CyA (3.4%), MPA (62.8%), mTORi (32.4%), Prednisone (79.7%), AZA (2.7%), Belatacept (8.1%), and Eculizumab (1.4%)	7
Danthu 2021 [40]	KT HC	74 7	44/30 4/3	64.8 ± 11.5 51.6 ± 6.8	5.5 ± 7.1 NA	BNT162b2	CLIA	13 AU/mL	4 wk	1	NA	Antithymocyte serum (35.1%), Basiliximab (64.9%), CNI (91.8%), Belatacept (2.6%),Everolimus (10.8%), Antimetabolite (82.4%), and Steroids (45.9%) NA	7
Davidov 2021 [41]	LIT HC	76 174	43/33 86/88	59 ± 15 59 ± 13	7 (4– 16) NA	BNT162b2	ELISA	Titers >1.1	38 ± 24 d 36 ± 22 d	2	NA	Prednisone (16%), MMF (21.3%),and Everolimus (14.7%) NA	6
Debska-Slizien 2022 [42]	KT HC	142 36	83/59 21/15	54 (43–63) 48 (45–62)	8 (3.5–15) NA	mRNA-1273 BNT162b2	CLIA	12 AU/mL	2–3 wk#	2	NA	Steroids (8.5%) and MMF/MPS (21.1%) NA	7
Del Bello 2022 [43]	SOT	396	257/139	59 ± 15	NA	BNT162b2	ELISA	Titers >1.1	NA	3	59 (25–75)	NA	5
Devresse 2021 [44]	KT	90	47/43	60 (38–79)	8.5 (0.8–36.7)	BNT162b2	ECLIA	0.8 U/mL	4 wk	2	NA	Tacrolimus + MPA + Steroids (41%) and Antimetabolite (76%)	5
Eren Sadioglu 2021 [45]	KT	85	38/47	46.4 ± 12.5	6.8 ± 5.7	Sinovac	ELISA	>10 IU/mL	4 wk	2	NA	Tacrolimus (88.2%), CyA (5.9%), MMF-MPA (63.6%), AZA (32.9%), mTORi (1.1%), and Steroids (95.3%)	7
Erol 2021 [46]	SOT KT LIT HC	48 56	32/16 NA	36.5 (1–62)# 37.5 (22–52)	8 (1–21) NA	Sinovac BNT162b2 Sinovac	CLIA	50 AU/mL	4/4–6 wk#	1/2	NA	NA	7
Fernández-Ruiz 2021 [47]	SOT KT LIT HC without transplant	44 28	27/17 7/21	52.4 ± 11.5 43.1 ± 15.1	2.3 (1.3–4.8) NA	BNT162b2	ELISA	Titers >1.1	4/2 wk	1/2	NA	Tacrolimus (80%), MMF/MPS (80%), Prednisone (60%), and mTORi (20%) NA	7
Firket 2021 [48]	KT HC	10 10	5/5 7/7	49.7 ± 13.8 51.5 ± 10.5	10.1 ± 8.8 NA	BNT162b2	CLIA	1 U/mL	2 wk	2	NA	Antithymocyte globulin (21%) and Rituximab (50%) NA	7
Georgery 2021 [49]	KT	78	40/36	62 (18–84)	9.7 (0.3–50.7)	BNT162b2	ECLIA	1 U/mL	4 wk	1	NA	Tacrolimus/MPA/Steroids (50%), Tacrolimus/Steroids (24%), and Antimetabolite (31%)	5
Georgery 2021 [11]	KT	79	38/41	61 (18–88)	105 (5–609)	BNT162b2	ECLIA	0.8 U/mL	4 wk	2	NA	Tacrolimus/MPA/Steroids (59.5%) and Antimetabolite (29%)	4
Grupper 2021 [50]	KT HC	136 25	111/25 8/17	58.6 ± 12.7 52.7 ± 11.5	3.3 (1.5–5.2) NA	BNT162b2	CLIA	15 AU/mL	16.5 ± 6.2 d 16.8 ± 2.9 d	2	NA	High-dose steroids (23.5%), Antithymocyte globulin (7.4%), Rituximab (2.9%), CNI (90.4%), mTORi (7.4%), and MMF (76.5%) NA	6
Guarino 2022 [51]	LIT HC	492 307	371/121 NA	64.9 (57.2–70.1) NA	14.1 (5.7–20.1) NA	BNT162b2	CLIA	1	4 wk NA	2 NA	NA NA	CNI (80.3%), Antimetabolite (34.2%), mTORi (27.9%), and teroids (7.1%) NA	8
Haidar 2021 [52]	SOT HC	183 107	110/73 30/77	61.2 ± 13.4 43.7 ± 13.7	NA	BNT162B2 mRNA-1273 JNJ-78436735	CLIA	signal/cutoff ratios >1	2 wk	2	NA	NA	7
Hall 2021 [53]	KT	127	88/39	66.2 (63.4–70.6)	2.9 (1.6–6.3)	mRNA-1273	ECLIA	0.8 U/mL	4/4–6 w#	1/2	NA	Prednisone (69.3%), CNI (84.3%), MMF/MPS (64.6%), AZA (8.7%), and Sirolimus (7.9%)	6
Hall 2021 [54]	KT	60	37/23	66.9 (64.0 – 71.8)	3.6 (2.0 – 6.8)	mRNA-1273	ECLIA	100 U/mL	4 ± 1 wk	3	60§	Prednisone (83.3%), CNI (98.3%), MMF/MPS (73.3%), AZA (13.3%), Sirolimus (10%)	4∗
Hallett 2021 [55]	SOT HT LUT	237 134	110/127 67/67	62 (46–69) 60 (44–69)	5.1 (2.5–11.0) 5.5 (2.6–12.4)	mRNA-1273 BNT162b2	ELISA; ECLIA	Titers >1.1	21 (19–26) d NA	1/2	NA	Tacrolimus (86%), MPA (62%), Corticosteroids (57%), Sirolimus (14%), CyA (8%), AZA (8%), Everolimus (7%), and Belatacept (1%) NA	6
Haskin 2021 [56]	KT	38	25/13	18.6 ± 2.8	7.3 ± 5.6	BNT162b2	CLIA	50 AU/mL	37 (20.5–53) d	2	NA	Rituximab (23.7%) and Antithymocyte globulin (13.2%)	7
Havlin 2021 [57]	LUT HC	48 10	29/19 NA	52.1 ± 14.3 NA	4.3 (0.3–20.1) NA	BNT162b2	ELISA	NA	1/4–6 wk#	1/2	NA	CNI (100%), Tacrolimus (97.9%), CyA (2.1%), and MPA (91.7%) NA	5
Herrera 2021 [58]	SOT LIT HT	104 58	NA 40/18	NA 61.5 (18–88)#	5.4 (0.3–27) 4.6(0–26.8)	mRNA-1273	ELISA	Titers >1.1	4/4 wk	1/2	NA	CNI (60.5%), MPA (29.6%), Prednisone (31.5%), and mTORi (17.3%)	7
Hod 2021 [59]	KT HC	120 202	96/24 61/141	59.7 ± 13 57.0 ± 13.6	5.8 ± 6.3 NA	BNT162b2	ELISA	Titers >1.1	2–4 wk#	2	NA	Tacrolimus (85.8%), MPA (77.5%), Prednisone (79.2%), CyA (6.7%), AZA(2.5%),and mTORi (5.8%) NA	7
Hoffman 2022 [12]	LUT	91	NA	NA	5.4 (1.9–9.7)	BNT162b2 mRNA-1273	CLIA	59 AU/mL	4/6 wk	1/2	NA	NA	5
Holden 2021 [60]	SOT KT LIT HT LUT	80	44/36	58.9 (47.9–66.8)	9.6 (4.8–16.0)	BNT162b2	ELISA	Titers >1.1	5.6 (5.1–6.3) d	2	NA	Prednisolone (26.3%), CNI (93.8%), Proliferation inhibitor (93.8%), and mTORi (2.5%)	5
Huang 2022 [61]	SOT KT LIT HT LUT	1163	706/457	62.0 (52.0–68.0)	3.2 (1.1– 6.8)	mRNA-1273 BNT162b2	ELISA	Titers >0.02	2 wk–3 mo#	2	NA	Antithymocyte globulin (0.3%), Antimetabolites (63.5%), Belatacept (0.6%), CNI (78.5%), mTORi (15.3%), and Rituximab (0.2%)	7
Husain 2021 [62]	KT	28	17/11	66 (42–87)	8.0 (1.0–15.8)	mRNA-1273 BNT162b2	CLIA	59 AU/mL	29 (12–59) d	2	NA	Tacrolimus (75%), Belatacep t(21%), Prednisone (32%), MMF/MPA (61%), AZA (11%), Leflunomide (4%), and Sirolimus/everolimus (14%)	5
Itzhaki Ben Zadok 2022 [63]	HT	42	35/7	61 (44–69)	9.2 (2.7–13.8)	BNT162b2	CLIA	50 AU/mL	21–26 d#	1/2	NA	CNI (81%), mTORi (57%), Steroids (69%), and Antimetabolites (55%)	5
Kamar 2021 [64]	SOT KT LIT	101	70/41	58 ± 2	8.1 ± 0.7	BNT162b2	ELISA	Titers >1.1	4 wk	1/2/3	61 ± 1	CNI (72.3%), Antimetabolite (64.4%), mTORi (28.7%), Steroids (85.1%), and Belatacept (11.9%)	8
Kamar 2021 [65]	SOT LIT KT HT	37	20/17	60 ± 14	9.1 ± 7.0	BNT162b2	ELISA	Titers >1.1	4 w	4	NA	CNI (86.5%), MPA (86.5%), mTORi (24.3%), and Steroids (91.9%)	5
Kantauskaite 2022 [66]	KT HC	225 176	148/77 NA	62 (54–70) 60 (54–69)	6.8 (2.6–12.3) NA	mRNA-1273 BNT162B2	ELISA	>35.2 BAU/mL	14 ± 2 d 17 d	2	NA	CNI (96.4%), MPA (83.1%), mTORi (3.1%), Steroids (94.2%), and AZA (1.8%) NA	5
Karaba 2022 [67]	SOT HC	58 35 16	27/31 16/19 11/5	NA NA NA	NA NA NA	mRNA-1273 BNT162B2 BNT162b2 AZD1222 mRNA-1273 BNT162b2	ELISA	Titers >1.1	2 w	2 3 2	NA 83§ NA	Prednisone (66%), CNI (90%), mTORi (14%), and Antimetabolites (69%) Prednisone (49%), CNI (80%), mTORi (9%), and Antimetabolites (66%) NA	6
Korth 2021 [68]	KT HC	23 23	11/12 9/14	57.7 ± 13.5 44.4 ± 9.2	11.4 ± 9.2 NA	BNT162b2	CLIA	13 AU/mL	2 wk	2	NA	MPA (78%), Corticosteroids (60%), Tacrolimus (60%), CyA (17%), Sirolimus (22%), Everolimus (4%), belatacept (4%), and AZA (4%) NA	7
Kumar 2021 [69]	SOT KT LIT HT LUT SOT without vaccination	60 57	37/23 39/18	66.9 (64.0–71.8) 66.1 (63.0–70.6)	3.6 (2.0–6.8) 2.3 (1.5–5.8)	mRNA-1273	ELISA	>0.33	4–6 wk#	2/3 3	60§ NA	Prednisone (83%), CNI (98%), MMF/MPS (73%), AZA (13%), and Sirolimus (10%)	4∗
Marinaki 2021 [70]	SOT KT HT HC	34 116	27/7 NA	60 (49.1–68.4) NA	11.1 (7.3–15.8) NA	BNT162b2	CLIA	50 AU/mL	10 (9–10) d	NA	2	Antimetabolite (33.3%) NA	5
Marion 2021 [71]	SOT KT LIT	367	232/135	59 ± 1	9 ± 0.4	BNT162b2	ELISA	Titers >1.1	4 wk	2	NA	Anticalcineurins (85%), Tacrolimus (78.2%), CyA (7.1%), MPA (68%), mTORi (25%), Steroids (82%), and Belatacept (9%)	5
Marlet 2021 [72]	KT	97 160	58/39 103/57	NA	NA	mRNA-1273 BNT162b2	CLIA	7.1 BAU/mL	3 wk	2 3	NA 43 (33–63)	NA	6
Massa 2021 [73]	KT	61	44/17	58.0 (47.1–66.1)	4.5 (1.8–11.3)	BNT162b2	ELISA	50 AU/mL	4/4 wk	2/3	28§	Corticosteroids (88.5%), Antimetabolites (62.3%), CNI (93.4%), mTORi (9.8%), and Belatacept (1.6%)	7
Masset 2021 [74]	KT KT	456 136	NA	61.4 ± 12.1 63.7 ± 11.7	10.5 ± 8.5 9.4 ± 8.1	BNT162b2	ECLIA	50 U/mL	NA/4 wk 4 wk	1/2 3	NA 50§	CNI (84.6%), mTORi (15.6%), Antimetabolite (74.5%), and Steroid (34.4%)	5
Mazzola 2021 [13]	SOT KT LIT HT HC	133 25	92/41 7/18	NA 55 (38–62)	3.8 (1.8–8.8) NA	BNT162b2	CLIA	50 AU/mL	4/4 wk 4 wk	1/2 2	NA NA	Corticoids (60.1%), CNI (81.9%), MMF (71.4%), and mTORi (19.5%) NA	6
Medina-Pestana 2021 [75]	KT KT without vaccination	942 6510	592/350 3971/2539	40 (32–46) 45 (37–53)	7 (3–12) 6 (2.8–11.1)	CoronaVac	CLIA	50 AU/mL	4 wk	1/2 2	NA NA	Tacrolimus + prednisone + AZA (32%%), Tacrolimus + prednisone + MPA (43%), prednisone + AZA + CyA (10%), and Tacrolimus + prednisone + mTORi (11%)	7
Middleton 2022 [76]	KT	70	NA	NA	NA	BNT162b2 AZD1222	Siemens immunoassay	NA	40 (12–79) d	1	NA	NA	5
Midtvedt 2021 [77]	KT	141	79/62	67.4 ± 17.2	11.7 ± 9.8	BNT162b2	Flow cytometric assay	1.0 BAU/mL	25–89 d#	2	NA	CNI + MPA + prednisolone (74%), CNI + prednisolone (13%), and MPA (82%)	5
Miele 2021 [78]	SOT	16	13/3	57 ± 15.7	9 ± 7.5	BNT162b2	CLIA	59 AU/mL	2 wk	2	NA	Tacrolimus (93.7%), Everolimus (6.3%), MMF 2.5%), and Corticosteroids 6.3%)	5
Narasimhan 2021 [79]	LUT	73	54/19	65 (53.5–69.5)	3.3 (1.6–5.3)	mRNA-1273 BNT162b2	CLIA	50 AU/mL	2 wk	2	NA	Antimetabolite (99%)	6
Noble 2021 [80]	KT	57	NA	NA	14.5 ± 28.8	mRNA-1273 BNT162b2	ELISA	Titers> 1.1	4/4 wk	2/3	NA	Belatacept (72%) and Tacrolimus (28%)	4
Ou 2021 [81]	KT	609	244/365	58 (45–68)	7 (3–15)	mRNA-1273 BNT162b2	ELISA	Titers 1.1	1/1 wk	1/2	NA	Prednisone (68.5%), MPA (71.9%), Tacrolimus (77.2%), AZA (9.7%), and Sirolimus (8.4%)	7
Pedersen 2021 [82]	KT HC	58 20	NA 4/16	NA	NA	BNT162b2	CLIA	34.8 AU/mL	4 wk	2	NA	Tacrolimus (77.6%), Everolimus (1.7%), CyA (15.5%), mTORi (87%), MMF/MPA (93.1%), AZA (6.9%), and Steroids (12.1%)	5
Peled 2021 [83]	HT HC	77 136	50/27 86/50	62 (49– 68) 63 ± 13	7.4 (3.3–15.1) NA	BNT162b2	ELISA	Titers >1.1	3 wk 13.3 ± 1.4 d	2	NA	MPA (75.3%), MPS (53.2%), MMF (22.1%), and Everolimus (26%) NA	6
Rabinowich 2021 [84]	LIT HC	80 25	56/24 8/17	60.1 ± 12.8 NA	6.4 ± 6.2 NA	mRNA-1273 BNT162b2	CLIA	15 AU/mL	14.8 ± 3.2 d 15.8 ± 2.9 d	2	NA	Prednisone (30%), CNI (93.7%), Everolimus (22.5%), AZA (5%), and MMF (50%) NA	6
Rashidi-Alavijeh 2021 [5]	LIT HC	43 20	26/17 9/11	47 (36– 54) NA	8 (4–12) NA	BNT162b2	CLIA	13 AU/mL	2 wk	2	NA	Tacrolimus use (93%), Tacrolimus + everolimus (55%), Tacrolimus + MMF (28%), Tacrolimus monotherapy (18%), CyA (5%), and Everolimus (2%) NA	6
Rincon-Arevalo 2021 [85]	KT HC	40 35	28/12 20/15	62.4 (51.3–69.5) 51 (34–80)	5 (2–10) NA	BNT162b2	ELISA	>0.3	1 wk 3–4 wk#	2	NA	Steroid + Tacrolimus + MMF(22%), Steroid + CyA + AZA (1%), Steroid + CyA + MMF (13%), mTORi + MMF + Steroid (3%), and mTORi + CyA + MMF (1%) NA	8
Rozen-Zvi 2021 [86]	KT	308	197/111	57.5 ± 13.8	7.1 ± 7.5	BNT162b2	CLIA	50 AU/mL	2–4 wk#	2	NA	Tacrolimus (92.5%), CyA(7.5%), mTORi (8.4%), CNI (58.8%), Rituximab (1.9%), and Antithymocyte globulin (4.5%)	7
Ruether 2021 [87]	LIT HC	138 52	79/59 19/33	55.0 ± 13.2 50.9 ± 11.6	7 (2–17) NA	mRNA-1273 BNT162B2 AZD1222	ECLIA	33.8 AU/mL	29 (25–39) d 36 (22–63) d	2	NA	Prednisone (31.2%) and CNI (92.8%) NA	7
Russo 2021 [88]	KT	82	47/35	58.5 (50.3–65.0)	5.8 (2.9–11.9)	BNT162B2	CLIA	12 AU/mL	43 (23–63) d	1/2	NA	Steroid (91.5%), Everolimus (12.2%), CNI (97.6%), CyA (18.3%), Tacrolimus (79.3%), and Antimetabolite (69.5%)	7
Sattler 2021 [89]	KT HC	39 39	28/11 20/19	57.4 ± 14.0 53.0 ± 17.6	8.2 ± 6.1 NA	BNT162B2	ELISA	Titer >1.1	1/3 wk 8 ± 1 d	2	NA	Corticosteroids + tacrolimus + MMF (56.4%), Corticosteroids + CyA + MMF (33.3%), mTORi + MMF(7.7%), and mTORi + MMF + CyA (2.6%) NA	7
Schmidt 2021 [90]	SOT KT LIT HT LUT LIT+KT	40	22/18	54.5 ± 12.7	6.5 ± 9.9	BNT162b2 mRNA-1273 AZD1222	ELISA	35.2 BAU/mL	15 ± 6/14 ± 1 d	1/2	NA	CNI + antimetabolite + glucocorticoids (70%), CNI + antimetabolite (12.5%), CNI + glucocorticoids (5%), and CNI only (5%), mTORi + Antimetabolite + glucocorticoids (5%), and mTORi + glucocorticoids (2.5%) NA	7
Schramm 2021 [91]	SOT HT LUT HT+LUT HC	50 50	32/18 17/33	55 ± 10 47 ± 10	NA 1.9 (1.4–2.4)	BNT162b2	ECLIA	0.8 U/mL	21/21 d	1/2	NA	Tacrolimus/MMF (82%), CyA/MMF (10%), and Tacrolimus/mTORi (8%) NA	6
Shostak 2021 [92]	LUT	168	112/56	60.5 (49.3–67.8)	NA	BNT162b2	CLIA	50 AU/mL	1/2 wk	1/2	NA	MTORi (17%) and Antimetabolite/Prednisone (92%)	6
Spinner 2022 [93]	HT	40	27/13	17.1 (15.7–18.4)	NA	BNT162B2 mRNA-1273 JNJ-78436735	CLIA	signal-to-cutoff ratio >1	118 (57–152) d	2	NA	MPA (42.5%), Prednisone (35%), Sirolimus (40%), Tacrolimus (85%), Anti−thymocyte globulin (7.5%), and Rituximab (7.5%) NA	5
Strauss 2021 [94]	LIT	161	69/72	64 (48–69)	6.9 (2.9–15)	BNT162b2 mRNA-1273	ELISA ECLIA	Titers >1.1 (ELISA); 0.8 U/mL (ECLIA)	4 wk	2	NA	Tacrolimus (81%), MPA (35%), Corticosteroids (22%), Sirolimus (11%), CyA (8%), AZA (6%), and Everolimus (3%) NA	5
Stumpf 2021 [95]	KT HC	368 144	241/127 34/110	57.3 ± 13.7 48.0 ± 11.9	9.9 ± 6.8 NA	mRNA-1273 BNT162b2	ELISA	Titers >1.1	4/3–4 wk#	1/2	NA	Corticosteroids (48.4%), CNI (87.5%), MMF/MPA (76.1%), mTORi (16%), and Belatacept (4.6%) NA	7
Stumpf 2021 [96]	KT	71	45/26	57.0 ± 14.4	NA	BNT162b2	ELISA	Titers >1.1	4/4 wk	2/3	68 ± 1	CNI (87%), MMF/MPA (73%), and mTORi (24%)	5
Thuluvath 2021 [97]	LIT	62	41/21	65.7 ± 8.7	NA	mRNA-1273 BNT162B2 JNJ-78436735	ECLIA	0.4 U/mL	4 wk	2	NA	AZA (12%), Prednisone (12%), and Tacrolimus (18%) NA	6
Timmermann 2021 [98]	LIT	118	75/43	66.1 (28–89)#	14.4 (0–37)	BNT162b2 mRNA-1273 JNJ-78436735	ELISA	Titers >1.1	3 wk	2	NA	Tacrolimus (69.1%), MMF (36.1%), Everolimus 4.4%), andCiclosporin (5.2%)	6
Tylicki 2021 [99]	KT	83	54/29	55 (42–63)	8 (3.5–15)	BNT162b2 mRNA-1273	CLIA	33.8 BAU/mL	2/3 wk	2/3	90§	Steroids (89.2%) and MMF/MPS (78.1%)	6
Vaiciuniene 2021 [100]	KT	136	84/52	NA	5.8 (0.9–21.8)#	BNT162b2	ELISA	35.2 BAU/mL	3–6 wk#	2	NA	Steroids (88.2%), MMF (83.1%), CNI (90.4%), CyA (34.6%), Tacrolimus (58.8%), and Sirolimus (9.6%)	6
Werbel 2021 [101]	SOT KT LIT HT LUT	30	13/17	57 (44–62)	4.5 (2.3–10.5)	mRNA-1273 BNT162B2 JNJ-78436735	ELISA ECLIA	Titers >1.1 (ELISA) 0.8 U/mL (ECLIA)	14 (14–17) d	3	67 (54–81)	Tacrolimus/CyA + MPA (83.3%), Corticosteroids (80%), Sirolimus (3.3%), and Belatacept (3.3%)	5
Westhoff 2021 [102]	KT	10	8/2	54 (41–74)	4.3 (0.8–15.8)	BNT162b2	ECLIA	0.8 U/mL	4/2 wk	2/3	70§	NA	5
Yanis 2022 [6]	KT	54	33/21	72.1 ± 3.6	7.0 (2.7–13.0)	BNT162b2	ELISA	NA	21–42 d#	2	NA	CNI (100%), Corticosteroids (50.1%), MPA (38.9%), AZA (13%), and mTORi (5.6%)	6
Yi 2021 [103]	KT	145	NA	NA	5 (3–10)	BNT162B2 mRNA-1273	NA	NA	NA	1	NA	NA	5

AZA, azathioprine; CLIA, chemiluminescence analysis; CNI, calcineurin inhibitor; CyA, cyclosporine A; ECLIA, electrochemiluminescence immunoassay analyser; ELISA, enzyme-linked immunosorbent assay; HC, healthy controls; HT, heart transplant; IQR, interquartile range; KT, kidney transplant; LIT, liver transplant; LUT, lung transplant; MMF, mycophenolate mofetil; mTORi, mammalian target of rapamycin inhibitor; MPA, mycophenolic acid; MPS, mycophenolate sodium; NA, not available; NOS, Newcastle-Ottawa scale; Pan, pancreas transplant; SOT, solid organ transplant; SD, standard deviation.

The NOS was performed for the rest of unmarked studies. Newcastle-Ottawa scale rates studies out of 9 and the Jadad scale rates studies from 1 to 5, with higher scores indicating lower risk of bias.

#indicates range; § indicates median.

∗The Jadad scale was used.

HIR in SOT recipients

Twenty-four studies [13,24,27,28,33,40,43,46,47,49,53,55,58,63,64,[74], [75], [76],81,[90], [91], [92],95,103] assessed the HIR after the administration of 1st vaccine dose in SOT recipients. The pooled response rate was 9.5% (95% CI, 7–11.9%) (Fig. 2a ) with considerable heterogeneity (I² = 89.1%; p < 0.001). Sensitivity analysis indicated that the result was not changed markedly (Fig. S1a).

Fig. 2.

Meta-analysis of the HIR after 1st, 2nd, and 3rd doses COVID-19 vaccine in SOT recipients and comparison of HIR. (A) HIR after 1st vaccine dose in SOT (9.5%); (B) HIR after 2nd vaccine dose in SOT (43.6%); (C) HIR after 3rd vaccine dose in SOT (55.1%); (D) Comparison of HIR after 1st vaccine dose (SOT vs. healthy controls, RR = 0.036); (E) Comparison of HIR after 2nd vaccine dose (SOT vs. healthy controls, RR = 0.382). HIR, humoral immune response; SOT, solid organ transplant; RR, risk ratio.

Seventy-five studies [5,6,[10], [11], [12], [13],20,21,23,[25], [26], [27],[29], [30], [31],[34], [35], [36], [37], [38], [39],[41], [42], [43], [44], [45], [46], [47], [48],[50], [51], [52], [53],55,56,[58], [59], [60], [61], [62], [63], [64],[66], [67], [68], [69], [70], [71], [72], [73], [74], [75],[77], [78], [79], [80], [81], [82], [83], [84],[86], [87], [88], [89], [90], [91], [92], [93], [94], [95],[97], [98], [99], [100]] assessed the HIR after the 2nd vaccine dose in SOT recipients. The pooled response rate was 43.6% (95% CI, 39.3–47.8%) (Fig. 2b) with considerable heterogeneity (I² = 95.4%; p < 0.001). Sensitivity analysis did not change the result (Fig. S1b).

Seventeen studies [19,20,23,32,43,54,64,67,69,[72], [73], [74],80,96,99,101,102] assessed the HIR after the administration of 3rd vaccine dose in SOT recipients, of whom 55.1% (95% CI, 44.7–65.6%) (Fig. 2c) exhibited a humoral response, with considerable heterogeneity (I² = 95.2%; p < 0.001). Sensitivity analysis revealed that the HIR rate was 60% (95% CI, 55.1–64.9%; I² = 69.9%; p < 0.001) (Fig. S1c) after removing the study by Chavarot et al. [32].

Comparison of HIR (SOT recipients vs. healthy controls)

Four studies [22,40,91,95] compared the HIR between SOT recipients and healthy controls after the 1st vaccine dose. All included studies used the mRNA vaccine. The result demonstrated that SOT recipients had a significantly lower rate of HIR than healthy controls (RR = 0.036; 95% CI, 0.014–0.091; I² = 55.9%, p = 0.078) (Fig. 2d). Sensitivity analysis did not change the result (Fig. S1d).

Twenty-nine studies [1,4,5,17,21,25,32,34,37,38,42,44,[47], [48], [49],54,56,65,66,68,73,[84], [85], [86],88,90,92,94,98] compared the HIR among SOT recipients and healthy controls after the 2nd vaccine dose. Only twenty-five studies [1,4,5,17,21,32,34,37,38,44,47,48,54,56,65,66,68,73,[84], [85], [86],88,92,94,98] used the mRNA vaccine, one study [25] used only inactivated vaccines, two [49,90] used mRNA or adenovirus vector vaccines, and one [42] used mRNA or inactivated vaccines. The result demonstrated that SOT recipients exhibited a significantly lower HIR rate than healthy controls (RR = 0.382; 95% CI, 0.313–0.468; I² = 96.1%, p < 0.001) (Fig. 2e). Sensitivity analysis did not change the result (Fig. S1e).

CIR in SOT recipients

Six studies [27,33,47,53,90,95] assessed the CIR after the 1st vaccine dose in SOT recipients. Five studies [27,33,47,53,95] used mRNA vaccines, and one study [90] used mRNA or adenovirus vector vaccines. The overall proportion of SOT recipients who exhibited CIR was 12.2% (95% CI, 6.5–17.8%, Fig. 3a ), with moderate heterogeneity (I² = 60.5%, p = 0.027). Sensitivity analysis did not change the result (Fig. S2a).

Fig. 3.

Meta-analysis of the CIR after 1st, 2nd, and 3rd doses COVID-19 vaccine in SOT recipients. (A) CIR after 1st dose in SOT (12.2%); (B) CIR after 2nd dose in SOT (48.3%); (C) CIR after 3rd dose in SOT (57.6%); (D) Comparison of CIR after 2nd dose (SOT vs. healthy controls, RR = 0.477). CIR, cellular immune response; SOT, solid organ transplant; RR, risk ratio.

Eighteen studies [6,27,29,33,38,39,44,47,53,57,58,78,79,87,89,90,95,102] assessed the CIR after the 2nd vaccine dose in SOT recipients. Fifteen studies [6,27,33,38,39,44,47,53,57,58,78,79,89,95,102] used mRNA vaccine, one [30] used inactivated vaccine, and two [87,90] used mRNA or adenovirus vector vaccine. The overall proportion of SOT recipients who exhibited CIR was 48.3% (95% CI, 34.2–62.4%, Fig. 3b) with considerable heterogeneity (I² = 96.5%, p < 0.01). Sensitivity analysis did not change the result (Fig. S2b).

Only two studies [96,102] reported the CIR after the 3rd vaccine dose in SOT recipients (Fig. 3c). The overall proportion of SOT recipients who exhibited CIR with considerable heterogeneity (I² = 96.5%, p < 0.001) was 57.6% (95% CI, −5.4% to 120.6%).

Comparison of CIR (SOT recipients vs. healthy controls)

Five studies [29,38,47,87,89] compared the SOT recipients and healthy controls after the 2nd vaccine dose (Fig. 3d). The meta-analysis result revealed that the SOT recipients had a significantly lower rate of CIR than the healthy controls (RR = 0.477; 95% CI, 0.257-0.885; I² = 96.9%, p < 0.001). Sensitivity analysis did not change the result (Fig. S2c).

Subgroup analysis based on the organ type for HIR

As shown in Fig. 4a and Table 2 , sixteen studies [13,24,27,28,33,40,46,49,[74], [75], [76],81,95,103] assessed the HIR after the 1st vaccine dose in KT recipients. The pooled rate was 9.4% (95% CI, 6–12.7%), with considerable heterogeneity (I² = 90.9%, p < 0.001). Sensitivity analysis did not change the result (Fig. S3a). Forty-two studies [5,10,11,13,21,23,25,27,29,30,35,37,39,42,44,48,50,53,56,59,62,66,68,[70], [71], [72], [73], [74], [75],77,[80], [81], [82],86,88,89,95,99,100] assessed the HIR after the 2nd dose in KT recipients. The overall proportion of KT recipients who exhibited HIR was 37.6% (95% CI, 33.5–41.6%; I² = 90.5%, p < 0.001). Sensitivity analysis did not change the result (Fig. S3b). Only 13 studies [19,23,32,43,64,[72], [73], [74],80,96,99,101,102] assessed the HIR after the 3rd vaccine dose in KT recipients, of whom 54.4% (95% CI, 40.8–68.1%) exhibited a humoral response, with considerable heterogeneity (I² = 96.3%, p < 0.001). Sensitivity analysis did not change the result (Fig. S3c).

Fig. 4.

Meta-analysis of the HIR and CIR after 1st, 2nd, and 3rd COVID-19 vaccine doses in different types of transplant recipients. (A) HIR in LUT (1st dose: 4.4%, 2nd dose:28.4%), HT (1st dose: 13.2%, 2nd dose: 50.3%), LIT (1st dose: 29.5%, 2nd dose: 64.5%) and KT (1st dose: 9.4%, 2nd dose: 37.6%, 3rd dose: 54.4%); (B) CIR in LIT (2nd dose: 66.3%) and KT (1st dose: 6.9%, 2nd dose: 42.6%, 3rd dose: 57.6%). HIR, humoral immune response; CIR, cellular immune response; KT, kidney transplant; LIT, liver transplant; HT, heart transplant; LUT, lung transplant.

Table 2.

Subgroup analysis based on the type of vaccine, type of organ, and the time post-transplant

Subgroup				No. of included article	Pooled rate (%)	95% CI (%)	I2	p
HIR	1st	Type of vaccine	mRNA	20	7.9	5.7–10.1	85.2	<0.001
		Type of organ	KT	16	9.4	6–12.7	90.9	<0.001
			KT vs. HCs	3	3.0∗	0.7–12.4	75	0.018
			LIT	4	29.5	12.2–46.7	86.5	<0.001
			HT	4	13.2	8.1–18.2	0	0.934
			LUT	2	4.4	0.9–7.9	17.7	<0.001
		Time post-transplant	≤5 year	6	5.2	2.5–7.8	53.2	0.058
			>5 year	15	11.2	7.9–14.4	89.1	<0.001
	2nd	Type of vaccine	mRNA	59	42.2	37.6–46.9	95.2	<0.001
			inactivated	3	24.8	3.5–46.1	96.3	<0.001
		Type of organ	KT	42	37.6	33.5–41.6	90.5	<0.001
			KT vs. HCs	12	32.2∗	24.2–42.9	90.7	<0.001
			LIT	17	64.5	57.4–71.6	89.5	<0.001
			LIT vs. HCs	7	67.6∗	58.9–77.6	83.7	<0.001
			HT	11	50.3	37.6–63	89.1	<0.001
			LUT	8	28.4	22.3–34.5	64.5	<0.001
		Time post-transplant	≤5 year	17	33.3	28.1–38.5	80.6	<0.001
			>5 year	46	45.0	39.4–50.5	95.9	<0.001
	3rd	Type of vaccine	mRNA	14	56.2	44.5–67.9	95.8	<0.001
		Type of organ	KT	13	54.4	40.8–68.1	96.3	<0.001
		Time post-transplant	≤5 year	8	51.2	30.6–71.8	95.7	<0.001
			>5 year	6	58.6	48.6–68.7	85.0	<0.001
CIR	1st	Type of vaccine	mRNA	5	12.2	5.7–18.8	67.4	<0.001
		Type of organ	KT	2	6.9	3.1–10.7	0	0.51
		Time post-transplant	≤5 year	3	11.5	2.0–21.1	68.1	0.043
			>5 year	3	13.7	4.4–23.0	68.3	0.027
	2nd	Type of vaccine	mRNA	15	51.6	39–64.2	93.8	<0.001
		Type of organ	KT	11	42.6	23.5–61.7	97	<0.001
			KT vs. HCs	3	41.6∗	41.6–122	98.1	<0.001
			LIT	3	66.3	30.1–102.5	96.1	<0.001
			LUT	2	56.9	14.5–99.2	88.9	<0.001
		Time post-transplant	≤5 year	8	44.5	21.1–68.0	96.6	<0.001
			>5 year	9	51.2	32.0–70.4	96.2	<0.001
	3rd	Type of organ	KT	2	57.6	-5.4–120.6	96.5	<0.001
SOT vs. HCs	2nd	Type of vaccine	mRNA	25	36.6∗	29.1–46.1	96.0	<0.001

CIR: cellular immune response; CI: confidence interval; HCs: healthy controls; HIR: humoral immune response; HT: heart transplant; KT: kidney transplant; LIT: liver transplant; LUT: lung transplant; SOT: solid organ transplant.

^∗risk ratio.

The overall proportion of LIT recipients who exhibited HIR after the 1st vaccine dose was 29.5% (95% CI, 12.2–46.7%), with considerable heterogeneity (I² = 86.5%, p < 0.001). Sensitivity analysis indicated that the pooled rate of response was 37% (95% CI, 29.1–44.8%) without heterogeneity (I² = 0, p = 0.951), after removing the study by Mazzola et al. [13] (Fig. S3d). Seventeen studies [5,10,13,34,41,46,47,51,53,58,60,61,71,84,87,94,97,98] assessed the HIR after the 2nd vaccine dose in LIT. The response rate was 64.5% (95% CI, 57.4–71.6%; I² = 89.5%, p < 0.01). Sensitivity analysis did not change the result (Fig. S3e).

Four studies [13,28,58,63] assessed the HIR after the 1st vaccine dose in HT. All the studies used the RNA vaccine. The pooled response rate was 13.2% (95% CI, 8.1–18.2%) without heterogeneity (I² = 0, p = 0.934). Eleven studies [10,13,31,53,55,58,61,63,70,83,93] assessed the HIR after the 2nd vaccine dose in HT recipients, of whom 50.3% exhibited a humoral response (95% CI, 37.6–63%) with considerable heterogeneity (I² = 89.1%, p < 0.001). Sensitivity analysis did not change the result (Fig. S3f).

Only two studies [28,92] detected the HIR after the 1st vaccine dose in LUT patients. The pooled rate of response was 4.4% (95% CI, 0.9–7.9%, I² = 17.7%). Eight studies [10,12,31,53,55,61,79,92] assessed the HIR after the 2nd vaccine dose in LUT recipients of whom 28.4% (95% CI, 22.3–34.5%), exhibited a response with considerable heterogeneity (I² = 64.5%, p < 0.001). Sensitivity analysis revealed that the overall proportion of HIR was 30.8% (95% CI, 26.2–35.3%; I² = 12.5; p = 0.335) (Fig. S3G), after removing the study by Shostak et al. [92].

Subgroup analysis based on the organ type for CIR

As shown in Fig. 4b and Table 2, only two included studies [33,95] reported the CIR after 1st vaccine dose in KT recipients. The overall proportion of KT recipients who exhibited CIR was 6.9% (95% CI, 3.1–10.7%) without heterogeneity (I² = 0, p = 0.51). Eleven studies [6,27,29,33,38,39,44,47,89,95,102] assessed the CIR after the 2nd vaccine dose in KT recipients. Ten studies [6,27,33,38,39,44,47,89,95,102] used mRNA vaccine and one [29] used inactivated vaccine. The overall proportion of KT recipients who exhibited CIR was 42.6% (95% CI, 23.5–61.7%) with considerable heterogeneity (I² = 97.0%, p < 0.001). Sensitivity analysis did not change the result (Fig. S3h).

Three studies [47,58,87] assessed the CIR after the 2nd vaccine dose in LIT recipients. Two studies [47,58] used mRNA vaccines, and one [87] used mRNA or adenovirus vector vaccines. The overall proportion of LIT recipients who revealed CIR rate was 66.3% (95% CI, 30.1–102.5%) with considerable heterogeneity (I² = 96.1%, p < 0.001). Sensitivity analysis showed that the pooled response rate was 85% (95% CI, 76.7–93.3%) without heterogeneity (I² = 0, p = 0.459) after removing the study by Ruether et al. [87].

Only two studies [57,79] assessed the CIR after the 2nd vaccine dose in LUT recipients. The overall proportion of LUT recipients with high heterogeneity (I² = 88.9%, p < 0.001) was 56.9% (95 CI% = 14.5–99.2%).

Subgroup analysis based on organ type: comparison of HIR (transplant recipients vs. healthy controls)

As shown in Table 2, only 3 studies [22,40,95] compared the HIR among KT recipients and healthy controls after the 1st vaccine dose. The result revealed that compared with healthy controls, KT recipients had a significantly lower rate of HIR (RR = 0.030; 95% CI, 0.007–0.124; I² = 75%, p = 0.018). Sensitivity analysis did not change the result markedly (Fig. S4a).

After the 2nd vaccine dose, 12 studies [29,38,42,48,50,59,66,68,82,85,89,95] showed that KT recipients had a significantly lower rate of HIR than the healthy controls, (RR = 0.322; 95% CI, 0.242–0.429; I² = 90.7%, p < 0.001). Sensitivity analysis did not change the result markedly (Fig. S4b). Seven studies [5,13,41,46,51,84,87] showed that LIT recipients exhibited significantly lower HIR rates than the healthy controls (RR = 0.676; 95% CI, 0.589–0.776; I² = 83.7%, p < 0.001). Sensitivity analysis did not change the result markedly (Fig. S4c).

Subgroup analysis based on organ type: comparison of CIR (transplant recipients vs. healthy controls)

As shown in Table 2, three studies [29,38,89] compared the CIR between KT recipients and healthy controls after the 2nd vaccine dose. The results showed that KT recipients exhibited a significantly lower CIR rate than the healthy controls (RR = 0.416; 95% CI, 0.416–1.220; I² = 98.1%, p < 0.001). Sensitivity analysis did not change the result markedly (Fig. S4d).

Subgroup analysis based on vaccines type in SOT recipients

As shown in Table 2, subgroup analysis revealed that the mRNA vaccine's pooled HIR rates were 7.9% (95% CI, 5.7–10.1%; I² = 85.2%), 42.2% (95% CI, 37.6–46.9%; I² = 95.2%) and 56.2% (95% CI, 44.5–67.9%; I² = 95.8%) after the administration of 1st, 2nd, and 3rd vaccine doses, respectively. However, the inactivated vaccine's pooled HIR rate was only 24.8% (95% CI, 3.5–46.1%; I² = 96.3%) after the administration of 2nd vaccine dose in SOT recipients.

Subgroup analysis revealed that the mRNA vaccine's pooled CIR rates were 12.2% (95% CI, 5.7–18.8%; I² = 67.4%) and 51.6% (95% CI, 39–64.2%; I² = 93.8%) after the administration of 1st and 2nd vaccine doses, respectively.

Subgroup analysis revealed that SOT recipients who received only mRNA vaccine had a significantly lower rate of HIR than the healthy controls (RR = 0.366; 95% CI, 0.291–0.461; I² = 96.0%). Sensitivity analysis showed that all the results in this subgroup analysis were not changed markedly.

Subgroup analysis of the time post-transplant in SOT recipients

As shown in Table 2 , subgroup analysis revealed that the pooled HIR rates of time post-transplant ≤5 years were 5.2% (95% CI, 2.5–7.8%; I² = 53.2%), 33.3% (95% CI, 28.1–38.5%; I² = 80.6%) and 51.2% (95% CI, 30.6–71.8%; I² = 95.7%) after the administration of 1st, 2nd, and 3rd vaccine doses, respectively. In contrast, the pooled HIR rates of time post-transplant >5 years were 11.2% (95% CI, 7.9–14.4%; I² = 89.1%), 45.0% (95% CI, 39.4–50.5%; I² = 95.9%) and 58.6% (95% CI, 48.6–68.7%; I² = 85%) after the administration of 1st, 2nd, and 3rd vaccine doses, respectively.

Subgroup analysis revealed that the pooled CIR rates of time post-transplant ≤5 years were 11.5% (95% CI, 2.0–21.1%; I² = 68.1%) and 44.5% (95% CI, 21.1–68.0%; I² = 93.8%) after receiving the 1st and 2nd vaccine doses, respectively. The rates of time post-transplant >5 years were 13.7% (95% CI, 4.4–23.0%; I² = 68.1%) and 51.2% (95% CI, 32.0–70.4%; I² = 93.8%) after receiving the 1st and 2nd vaccine doses, respectively. Sensitivity analysis did not change the result in the subgroup analysis of time post-transplant.

Subgroup analysis based on vaccine type in different types of transplants for HIR

As shown in Table 3 , the mRNA vaccine's pooled HIR rates were 6.6% (95% CI, 3.8–9.4%; I² = 85.3%), 37.6% (95% CI, 33.2–42%; I² = 89.8%), and 55.5% (95% CI, 41.2–69.8%; I² = 96.6%) after receiving the 1st, 2nd, and 3rd vaccine doses, respectively, in KT recipients. Besides, the inactivated vaccines' pooled HIR rate was 24.8% (95% CI, 3.5–46.1%; I² = 96.3) after the administration of 2nd vaccine dose in KT recipients. The mRNA vaccine's pooled HIR rates were 37% (95% CI, 29.1–44.8%; I² = 0%) and 60.7% (95% CI, 51.9–69.5%; I² = 90.9%) after the administration of 1st and 2nd vaccine dose in LIT recipients, respectively. The mRNA vaccine's pooled HIR rate was 48.1% (95% CI = 34.7%–61.4%, I² = 89.1%) after the administration of 2nd vaccine dose in HT recipients.

Table 3.

Subgroup analysis based on the vaccine type and the time post-transplant in different transplant types

Subgroup					Number of included articles	Pooled rate (%)	95% CI (%)	I2	p
KT	HIR	1st	Type of vaccine	mRNA	11	6.6	3.8–9.4	85.3	<0.001
			Time post-transplant	≤5 years	3	2.9	0.7–5.1	24.2	0.267
				>5 years	10	8.6	6.7–14.0	89.5	<0.001
		2nd	Type of vaccine	mRNA	35	37.6	33.2–42.0	89.8	<0.001
				inactivated	3	24.8	3.5–46.1	96.3	<0.001
			Time post-transplant	≤5 years	11	30.3	23.5–37.1	78.8	<0.001
				>5 years	29	37.8	33.0–42.6	91.2	<0.001
		3rd	Type of vaccine	mRNA	11	55.5	41.2–69.8	96.6	<0.001
			Time post-transplant	≤5 years	5	44.7	15.8–73.6	96.8	<0.001
				>5 years	6	58.5	48.4–68.6	84.6	<0.001
	CIR	2nd	Type of vaccine	mRNA	10	46.7	30.3–63.2	94.6	<0.001
			Time post-transplant	≤5 years	5	37.4	9.3–65.4	96.0	<0.001
				>5 years	5	45.5	15.0–76.1	97.5	<0.001
LIT	HIR	1st	Type of vaccine	mRNA	3	37	29.1–44.8	0	0.951
		2nd	Type of vaccine	mRNA	13	60.7	51.9–69.5	90.9	<0.001
			Time post-transplant	≤5 years	5	45.7	36.5–54.9	45.1	0.122
				>5 years	11	70.4	64.5–76.3	81.0	<0.001
	CIR	2nd	Type of vaccine	mRNA	2	85	76.7–93.3	0	0.459
HT	HIR	2nd	Type of vaccine	mRNA	9	48.1	34.7–61.4	89.1	<0.001
			Time post-transplant	≤5 years	4	30.6	16.3–44.9	54.5	0.086
				>5 years	6	52.4	35.0–69.8	92.9	<0.001
LUT	HIR	2nd	Time post-transplant	≤5 years	4	28.8	22.9–34.8	0	0.434
				>5 years	3	33.8	26.9–40.7	32.6	0.227

CIR, cellular immune response; HIR, humoral immune response; HT, heart transplant; KT, kidney transplant; LIT, liver transplant; LUT, lung transplant.

Subgroup analysis based on vaccine type in different types of transplants for CIR

The mRNA vaccine's pooled CIR rates were 46.7% (95% CI, 30.3–63.2%; I² = 94.6%) and 85% (95% CI, 76.7–93.3%, I² = 0%) after the administration of 2nd dose in KT and LIT recipients, respectively (Table 3).

Subgroup analysis based on time post-transplant in different types of transplants

As shown in Table 3, with the increased time post-transplant, the pooled rates of HIR and CIR were high in different organ transplant recipients after the administration of 1st, 2nd, and 3rd doses.

Subgroup analysis based on the time interval between the 2nd and 3rd dose vaccination in SOT and KT recipients

The subgroup analysis was performed based on the time interval between the administration of 2nd and 3rd doses. Moreover, the result showed that the pooled HIR rates did not change significantly both in SOT and KT recipients (Fig. S5).

Grading the quality of evidence

The GRADE approach indicated that the overall quality of evidence was low because most of the data were from observational studies (Table S2).

Meta-regression analysis

Meta-regression analysis indicated that the heterogeneity of HIR after the 1st and 2nd doses originated from the vaccine type and the time post-transplant, respectively (Table S3). Sample size and risk-of-bias scores at the study level did not show significant effect modifiers in the meta-regression analysis.

Publication bias

Publication bias was determined via funnel plot effect sizes. No publication bias was found in the main outcomes (Fig. S6).

Discussion

This systematic review and meta-analysis summarised the cumulative evidence of immunogenicity of the COVID-19 vaccines in SOT recipients. We demonstrated that the immunogenicity of the vaccine was poor in SOT especially for LUT recipients, varied among target organs, and was significantly lower than that of healthy controls. A booster vaccination could induce a stronger immune response. Besides, the longer the time post-transplant was, the higher the probability of a detectable HIR and CIR after the vaccination in SOT recipients. Moreover, the immunogenicity of mRNA-based vaccines was stronger than the inactivated vaccines in SOT recipients.

Our meta-analysis showed that the immunogenicity of COVID-19 vaccines in cellular or humoral immune responses was significantly impaired in SOT patients. However, with an increased vaccination dose, SOT patients' immune capacity can be enhanced, consistent with previous studies [64,67,69,74]. In our analysis, we found that, after receiving the 2nd dose, SOT patients' immune response rate significantly increased from 9.5% to 43.6% in HIR, and from 12.2% to 48.3% in CIR. In addition, the rates of HIR and CIR reached 56.2% and 57.6% after receiving the 3rd vaccine dose, respectively. The same trend was observed in different transplant recipients. The HIR rates after the 2nd vaccine dose increased from 4.4% to 28.4% in LUT recipients and from 13.2% to 50.3% in HT recipients. A higher proportion of LIT and KT recipients exhibited a humoral response after the 2nd vaccine dose.

Although vaccine immunogenicity in SOT patients improved with an increased dose, immune response varied among different transplant recipients. Therefore, we performed a subgroup analysis of different organ transplants. Our results revealed that vaccination appeared to induce a relatively poor HIR in LUT recipients (1st dose: 4.4%, 2nd dose: 28.4%) but a relatively strong one in LIT recipients (1st dose: 29.5%, 2nd dose: 64.5%). We found that the high proportion of older patients comprising the LUT group may be responsible for the low immune response. Narasimhan et al. [79] and Hallett et al. [55] included 63% and 58% of LUT patients aged >60 years, respectively. However, <50% of older recipients were included in LIT. In addition, the proportion of recipients receiving antimetabolite therapy was different [57,92]. Antimetabolites were used in a high proportion of LUT recipients, a high proportion of 99% reported by Narasimhan et al. [79], but less than 40% in LIT patients. Furthermore, patients who have a short time post-transplant have a strong immunosuppressive state [51,55,94]. Therefore, 34.9% of LUT recipients had a time post-transplant of fewer than 3 years [55], but only 28.4% in HT recipients [94]. Our subgroup analysis results also indicated that with the longer time post-transplant, the immunogenicity was strong in SOT recipients after vaccination. We speculate that this may be due to the longer time since immunosuppression induction.

The immunogenicity appears to vary with the vaccine type. Therefore, the subgroup analysis based on vaccine type was performed. We found that the mRNA vaccine may cause a stronger HIR than the inactivated vaccines (56.2% vs. 24.8%), consistent with previous studies [[104], [105], [106]]. Sauré et al. [106] compared the seroconversion rates of CoronaVac and BNT162b2 by recruiting 56 261 individuals. They found that after the 1st dose of CoronaVac in week 4, 28.1% of individuals had positive IgG antibodies for SARS-CoV-2. The peak was 77.4% after the 2nd dose in week 3. However, the seroconversion was 79.4% during week 4 after receiving the 1st dose of BNT162b2, which increased to 96.5% during week 3 after the 2nd dose. In addition, compared with Sinovac, vaccination with BNT162b2 resulted in a higher antibody titer in immunocompromised populations [104]. The underlying mechanism might be that the mRNA vaccine utilized host cells to synthesize antigens for SARS-CoV-2 and triggered a strong immune response [107,108].

There are some limitations in this study. First, most articles were observational studies and many healthy controls in the included studies were healthcare workers. Secondly, most included studies assessed the HIR to COVID-19 vaccines. However, there is a lack of data on the CIR. Third, many studies used mRNA vaccines, but the other types are still lacking. Lastly, different testing methods for HIR and CIR were performed in different individual studies. The cutoff value to define a positive immune response was different across studies, which may lead to potential bias.

Conclusion

A booster vaccination enhances the immunogenicity of COVID-19 vaccines in SOT; however, a significant share of the SOT recipients still has not built a detectable humoral immune response after the 3rd dose. This finding calls for alternative approaches, including the use of monoclonal antibodies. In addition, lung transplant patients need urgent booster vaccination to improve immune response. However, a lack of a serological cutoff value correlated with protective immunity and multiple-site mutated SARS-CoV-2 variants may trigger immune escape against existing humoral and cellular immune responses, leading to breakthrough COVID-19 infection.

Transparency declaration

All data were derived from sources available in the public domain, as referred to in the reference list. The authors declare that they have no conflicts of interest.

Funding

No external funding was received for this study.

Author contributions

X.P.C. and D.L.: designing study, writing original draft, and editing the manuscript; B.J.M., J.D., and X.D.L.: search and data extraction; H.X. and L.L.: the risk of bias assessment and the statistical analysis; S.S. and G.M.: oversight, critical evaluation, and verification of the manuscript. All authors contributed to the data interpretation and approved the final manuscript. X.P.C. and D.L. have contributed equally to this work.

Editor: A. Kalil

Appendix A. Supplementary data

The following are the Supplementary data to this article.