Retrospective study of the immunogenicity and safety of the CoronaVac SARS-CoV-2 vaccine in people with underlying medical conditions

Study participants

Between 5 Jul and 30 Dec 2021, we recruited 1302 participants at the 14–28th day after the second dose of inactivate SARS-CoV-2 vaccination and collected safety surveys. Among them we enrolled 1266 who had completed all survey questions including demographic information, underlying medical conditions (disease type, duration, severity, and control status) as well as occurrence of adverse events after each dose of vaccination. 297 individuals were excluded from this analysis due to receiving two different products of inactivated vaccines. The remaining 969 participants comprised the analyzed samples: 740 people with underlying medical conditions and 229 people as the healthy control. All of them formed the safety population. For the immunogenicity analysis, 969 participants had completed the blood sampling in the 14–28th day after the second dose of vaccine. 178 participants failed to be on site and were excluded from the 90 days immunity assessment, and 40 participants with same reason were excluded from the 180 days analysis (Fig. 1). Each disease group was further divided into 40–59 years old and ≥60 years old subgroups. Baseline demographic characteristics of the comorbidities group and healthy control were well-balanced (Table 1) across age groups, except for the age in ≥60 years old subgroups.

Flow chart of recruitment and enrollment (including inclusion/exclusion criteria) for the clinical trial ChiCTR2200058281 to evaluate the immunogenicity and safety of the CoronaVac SARS-CoV-2 vaccine in people with underlying medical conditions.

Vaccine safety

150 (20.27%) of 740 comorbidities participants had at least one adverse event, compared with 32 (13.97%) of 229 healthy participants (Supplementary Table 2). Most adverse events were mild (grade 1) in severity and participants recovered within 48 h (Supplementary Table 3). The most frequently reported adverse events in healthy and comorbidities group were injection-site pain (20 [8.73%] of 229 versus 94 [12.70%] of 740), fatigue (9 [3.93%] of 229 versus 48 [6.49%] of 740) and fever (0 [0.00%] of 229 versus 4 [0.54%] of 740). There was no significant difference seen between healthy and comorbidities cohort at overall level (Supplementary Table 2). Although 4 cases of grade 3 adverse events have been reported in 4 individuals in the comorbidities group, including acute allergy, skin & mucosa abnormalities and fever, most events occurred 7 days after vaccination except for a fever that occurred at the 1st day post vaccination. Thus, none was considered to be related to the vaccination except for the fever, which recovered at the 2nd day post vaccination (Supplementary note 1).

When inspecting the adverse events after the first and the second dose of vaccination, respectively, the incidence of adverse events were 16 (6.99%) of 229 in the health group versus 97 (13.11%) of 740 in the comorbidities group after the first dose, 19 (8.30%) of 229 versus 99 (13.37%) of 740 after the second dose (Supplementary Table 2), with significant difference between health and comorbidities group in both doses (P = 0.0129, first dose; P = 0.0487, second dose). We then stratified participants by age group: adults (40–59 years old) and seniors (≥60 years old), to explore if seniors exhibit different response to inactivated SARS-CoV-2 vaccine. In the adults subgroup, 81 (27.27%) of 297 participants in the comorbidities group versus 20 (16.53%) of 121 participants in the healthy control, reported adverse events, with statistically significant difference between groups (P = 0.0231). Moreover, systemic adverse events of the comorbidities group was significantly higher than that of the healthy control (13.47% versus 4.96%, P = 0.0099), while local adverse events showed no significant difference (Table 2). The incidence of adverse events was 54 (18.18%) of 297 in comorbidities people and 9 (7.44%) of 121 in healthy people after the first dose, and 55 (18.52%) of 297 in comorbidities people and 12 (9.92%) of 121 in healthy people after the second dose of vaccine. Both doses showed significant differences between diseases and healthy control (P = 0.0062 and P = 0.0387). In the senior subgroup, comorbidities and healthy group did not show significant difference in the overall (15.58% versus 11.11%, P = 0.2895), the first dose (9.71% versus 6.48%, P = 0.3537) or the second dose (9.93% versus 6.48%, P = 0.3543) vaccination. Thus, the higher incidence of adverse events in the adult comorbidities population was the main driving factor for the difference between the comorbidities cohort and healthy control in the overall incidence of adverse events. More specifically, the major inter-group difference was contributed by the systemic adverse events, mainly fatigue (26 [8.75%] of 297 in comorbidities, 3 [2.48%] of 121 in health, P = 0.0199). Furthermore, the incidence of fatigue in adults group was 20 (6.73%) of 297 in the comorbidities group after the first dose, with significant difference from the health control (1 [0.83%] of 121, P = 0.0115); There was no significant difference between comorbidities and health groups after the second dose with 12 (4.04%) of 297 in the comorbidities group and 2 [1.65%] of 121 in the health control (P = 0.3678).

Next, we compared the incidence of adverse events between comorbidities and healthy control stratified by disease types (Supplementary Table 4). The overall incidence of adverse events was 46 (19.83%) of 232 in the hypertension group, 24 (20.34%) of 118 in the CAD group, 34 (19.21%) of 177 in the DM group, 22 (23.40%) of 94 in the CRD group, 19 (21.59%) of 88 in the cancer group, and 5 (6.13%) of 31 in the obesity group versus 32 (13.97%) of 229 in healthy control. The most frequently reported adverse events in six disease groups were same as those in the health control, mainly injection-site pain, and fatigue (Fig. 2). All adverse events showed no significant difference between six disease subgroups and health control. When focusing on seniors, the CRD group showed significantly higher incidence of injection-site pain (9 [14.29%] of 63 vs. 5 [4.63%] of 108 in the health control, P = 0.0404) than that of the healthy control. Regarding adults, fatigue in the hypertension group (11 [10.89%] of 101 versus 3 [2.48%] of 121 in health group, P = 0.0124), DM group (7 [9.86%] of 71, P = 0.0403) and cancer group (6 [12.77%] of 47, P = 0.0152), and headache in the CRD group (2 [6.45%] of 31 versus 0 [0.00%] of 121 in health group, P = 0.0405) showed significant differences compared to the healthy control.

n = 969 study participants. Adverse events post 14 days of first dose and the second dose of the vaccination were collected and graded according to the China National Medical Products Administration guidelines. The histogram shown the incidence of adverse events happened in the 40–60 and ≥60 years old groups. Green, grade 1; orange, grade 2; blue, grade 3; Red, grade 4.

Humoral immunogenicity

The immunogenicity of people with underlying diseases and healthy control was evaluated at 14–28 days, 3 months, and 6 months after the two-dose vaccination (Table 3). By day 14–28, the seroconversion rates of neutralizing antibodies were 99 (84%) of 118 in the CAD group, 206 (89%) of 232 in the hypertension group, 150 (85%) of 177 in the DM group, 73 (78%) of 94 in the CRD group, 28 (90%) of 31 in the obesity group, and 75 (85%) of 88 in the cancer group versus 204 (89%) of 229 in health group; The neutralizing antibody GMTs were 22.77 (95% CI 18.29–28.35), 33.89 (95% CI 29.08–39.49), 26.45 (95% CI 21.99–31.82), 22.44 (95% CI 17.33–29.06), 32.92 (95% CI 20.89–51.88), 31.96 (95% CI 24.00–42.56) versus 30.50 (95% CI 26.41–35.22), respectively. Most diseases groups showed no difference in seroconversion rates and GMTs from the healthy control, while CAD (P = 0.0287) and CRD group (P = 0.0416) exhibited statistically significant decrease of humoral immune response. By day 90, both the seroconversion rates and GMTs were significantly reduced in all people, with no difference between disease and health groups. By day 180, seroconversion rates and GMTs continued to drop, but declined at a much slower rate. Interestingly, cancer patients showed a significant elevation in the seroconversion rate (68% versus 46%, P = 0.0020) and GMT (11.50 versus 7.42, P = 0.0050) compared to the healthy control (Table 3).

In the adult subgroup (40–59 years old) analyses, GMTs of neutralizing antibodies showed elevations in the cancer (53.71 versus 28.71, P = 0.0016) and hypertension (49.02 versus 28.71, P = 0.0009) patients compared to healthy control (Supplementary Table 5). Interestingly, in the senior subgroup (≥60 years old), cancer patients (18.56 versus 32.44, P = 0.0200) showed the opposite trend with a significantly reduced GMT level. Similarly, senior CAD (20.05, P = 0.0052) and CRD (19.85, P = 0.0144) patients also showed the same trend of immunogenicity reduction (Supplementary Table 6). Notably, these inter-group variations were no longer significant at 3- and 6-months post vaccination.

Comparing across age subgroups, we observed distinct patterns for disease and health groups. In healthy participants, the seroconversion and GMT of neutralizing antibodies were slightly higher in seniors (≥60 years old) than their younger counterpart (40–59 years old) on day 14–28 and 90 after the second dose. Conversely, senior people with chronic diseases had a lower neutralizing antibody level compared to their younger counterpart by day 14–28 post vaccination. These differences between age groups in seroconversion rate and GMT of neutralizing antibodies diminished by 3 and 6 months after the second dose (Fig. 3a). The reverse cumulative distribution plot showed that, at each time point, the overall distribution of neutralizing antibody titers was generally close across diseases in the senior and adult subgroups (Fig. 3b).

n = 969 study participants. The error bars represent the standard error of the experiment results. a Titers of neutralizing antibodies across age groups. red-, blue- and orange-color refer to 14–28 days, 3 months and 6months after two-dose vaccination, respectively. Orange line refer to 40–59 years old group, and blue line refer to senior group older than 60 years old. b The inverse cumulative distribution of neutralizing antibodies in hypertension (light blue triangle), CAD (pink plus), DM (blue cross), CRD (yellow rhombus), cancer (green triangle), obesity (gray square) subgroups, and healthy control (red circle), across age groups.

Cellular immunogenicity

The SARS-CoV-2-specific T-cell response was quantified utilizing an AIM assay at 3 and 6 months post the two-dose vaccination (Fig. 4). By day 90, the SARS-CoV-2-specific CD4⁺ T cell responses (OX40⁺ CD137⁺) were detected in 21 (100%) of 21 in the CAD group, 31 (86.1%) of 36 in the hypertension group, 21 (95.5%) of 22 in the DM group, 18 (100%) of 18 in the CRD group, 8 (88.9%) of 9 in the cancer group versus 50 (89.3%) of 56 in health control (Supplementary Table 7). The median fraction of SARS-CoV-2-specific CD4⁺ T cells among CD4⁺ T cells were 0.05% (IQR 0.019%–0.1%), 0.04% (IQR 0.012%–0.084%), 0.08% (IQR 0.02%–0.17%), 0.08% (IQR 0.03%–0.12%), 0.03% (IQR 0.01%–0.12%), and 0.05% (IQR 0.02%–0.11%) in the CAD, hypertension, DM, CRD, cancer, and health group, respectively (Supplementary Table 8). There was no significant difference between disease groups and healthy control on either SARS-CoV-2-specific CD4⁺ T cell positive rate or fraction. The SARS-CoV-2-specific CD8⁺ T cell responses (CD69⁺ CD137⁺) were detected in 8 (38.1%) of 21 in CAD, 14 (38.9%) of 36 in hypertension, 16 (72.7%) of 22 in DM, 11 (61.1%) of 18 in CRD, 6 (66.7%) of 9 in cancer versus 34 (60.7%) of 56 in the health control (Supplementary Table 7). The median fraction of SARS-CoV-2-specific CD8⁺ T cell among CD8⁺ T cell were 0% (IQR 0%–0.01%), 0% (IQR 0%–0.01%), 0.02% (IQR 0%–0.06%), 0.01% (IQR 0%–0.03%), 0.01% (IQR 0%–0.02%), and 0.01% (IQR 0%–0.03%) in the CAD, hypertension, DM, CRD, cancer, and health group, respectively. There were no significant difference between disease groups and healthy control on either SARS-CoV-2-specific CD8⁺ T cell positive rate or fraction (Supplementary Table 8); By day 180, the detected positive rate and cell fraction of SARS-CoV-2-specific CD4⁺ and CD8⁺ T cells were close to that of day 90, with no significant difference between disease groups and healthy control, except for the detected positive rate of the SARS-CoV-2-specific CD4⁺ T cell in the cancer group at day 180 (81.8% versus 97.6% in health control, P = 0.0440), with significant lower than the health control.

n = 229 study participants. Box indicates 25th, 50th, and 75th percentile absolute error, and whiskers indicate 5th and 95th percentile absolute error. Blue color refer to 3 months after two-dose vaccination, while orange color refer to 6 months. The detected positive rate and cell fraction of SARS-CoV-2-specific CD4⁺ T cell responses (OX40⁺ CD137⁺) (a) and the SARS-CoV-2-specific CD8⁺ T cell responses (CD69⁺ CD137⁺) (b) in different chronic diseases subgroups and healthy control, quantified by AIM assays.