It has been more than a year since the outbreak of the novel coronavirus pneumonia (COVID-19). Although the spread of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) that caused COVID-19 in China was effectively controlled, the global epidemic has not stopped. According to data from the World Health Organization, as of 16:05 on February 15, 2021, Central European Time, the cumulative number of confirmed COVID-19 cases worldwide reached 108, 579, 352, and the cumulative deaths reached 2, 396, 408^[1]. The COVID-19 epidemic as a major global public health event has become the primary health threat for all mankind, and impacted the world's political, economic and cultural greatly^[2-3]. SARS-CoV-2 is a β-coronavirus with RNA as genetic material, which enters cell through a spike protein combined with angiotensin converting enzyme 2^[4-5]. COVID-19 generally manifests as fever and dry cough, and injuries multiple organ, especially the lungs^{[2, 5-6]}. Wearing mask and maintaining social distancing have been confirmed as the most effective measures to stop the spread of the virus form China's experience of fighting the epidemic^{[3, 7-9]}, and isolation and symptomatic supportive treatment still dominate for COVID-19 patients^[5]. However, the efficacy of antiviral drugs and traditional Chinese medicines needs more evidence^[10-11]. Due to the low penetration rate of masks and the limitations of treatment options abroad^[12-13], more and more hopes are pinned on the development of a COVID-19 vaccine. According to different targets and technologies, vaccines can be divided into the following categories: inactivated vaccines, recombinant spike protein vaccines, viral vector vaccines, RNA vaccines, live attenuated vaccines and virus-like particle vaccines, etc^[14-16]. Currently, hundreds of COVID-19 candidate vaccine projects have been registered in the US clinical trial database (clinicaltrials.gov)^{[15, 17]}. Results of phase 3 clinical trials of several vaccines are published^[18-22]. As of January 1, 2021, China, Russia, the United States, Britain and other countries have approved their own mass vaccination plans for the population. This study evaluated the safety and effectiveness of the COVID-19 vaccine through systematic literature review and qualitative analysis for the published COVID-19 vaccine clinical trial results.

1 Information and methods

This systematic review was completed in accordance with the guidelines in the "Preferred Reporting Project for Systematic Evaluation and Meta-Analysis (PRISMA)"^[23-24].

1.1 Literature inclusion criteria

The literature inclusion criteria: (1) The healthy men or non-pregnant women aged 18 and above; (2) COVID-19 vaccination as the intervention measure; (3) The randomized, controlled, and blinded trials; (4) The clinical trial results indicators include at least one or more as following: local adverse reactions (pain, itching, redness, swelling and induration, etc.), systemic adverse reactions (fever, diarrhea, fatigue, nausea/vomiting, etc.), the last vaccine neutralizing antibody geometric mean titer (GMT), seroconversion rate and other laboratory test indicators measured by live virus neutralization test 14 days or 28 days after inoculation.

1.2 Literature exclusion criteria

Documents that meet one of the following conditions were excluded: (1) Medical news, popular science articles, non-medical papers, reviews, letters, comments, basic research, case reports, conference abstracts; (2) No full text or literature published in a third language other than Chinese and English; (3) One of overlapping two studies were excluded; (4) If the data of the literature included in the later published literature, The former was excluded.

1.3 Literature search

The English databases PubMed, Embase, Cochrane Library and clinicaltrials.gov databases were searched. The Chinese databases searched included CNKI, Wanfang Database, China Biomedical Literature Service System and China Clinical Trial Registration Center. In order to ensure the comprehensiveness of the search results, this system evaluation used Boolean logic to search by "subject words + free words". The main search terms include: COVID-19, 2019-nCoV, SARS-CoV-2, 2019 novel coronavirus, vaccines, vaccination, COVID-19 vaccines, mRNA-1273 vaccine, Ad5-nCoV vaccine, ChAdOx1 COVID-19 vaccine, BNT162 vaccine, controlled clinical trial, randomized controlled trials, controlled clinical trial, random, blind, placebo, trial, Meta, and etc. Chinese search terms include: 新型冠状病毒、新冠肺炎、新型冠状病毒肺炎、疫苗、试验、随机对照试验、随机对照研究、随机对照、随机、元分析、Meta、荟萃, etc.

1.4 Literature screening and data extraction

The literature screening and data extraction were done independently by two researchers. Differences in the summary of the results will be discussed and dealt with by the two researchers or the third researcher. All results obtained in the database were imported into Note Express (Wuhan University Library Edition) software, and duplicate documents were removed mechanically using the software's duplicate check function. The initial screening by reading the title and abstract, and the secondary screening by reading the full text were completed. The extracted data included: the first author, vaccine type, inoculation dose, interval between inoculations, number of subjects and baseline characteristics (race, sex ratio, age range or average age), research design, local and systemic adverse reactions, laboratory indicators, as well as funds, sponsors and registration number.

1.5 Methodological quality evaluation

Assess the risk of bias according to the Cochrane Systematic Review Manual^[25-26].

1.6 Statistical analysis

The main results of this systematic review included the safety and effectiveness of the vaccine. Indicators for evaluating safety included local adverse reactions (pain, itching, redness, induration, etc.) and systemic adverse reactions (cough, diarrhea, fatigue, fever, headache, nausea/vomiting, itching, muscle pain, joint pain/discomfort, anorexia, etc.). The immunogenicity indicators included GMT, seroconversion rate, and the response of IgG or other specific antibodies to the receptor binding domain.

2 Results 2.1 Literature search results

There were 753 relevant articles published before December 31, 2020. After screening, 13 papers were included in the system evaluation^{[19-22, 27-35]}. The process of document screening was shown in Figure 1.

2.2 Methodological quality evaluation

The 13 included studies all adopted a randomized control method^{[19-22, 27-35]}, with a double-blind method in 10 studies^{[21-22, 27-32, 34-35]}, and a single-blind method in 2 studies^{[20, 33]}, and bothsingle-blind method and double-blind methodin one study^[19]. All trials hid the allocation plan. Nine trials had incomplete data or selective reports^{[19, 22, 27, 29-31, 33-35]}, of which 2 had more missing data in the preprint^{[22, 29]}, and the remaining 7 missed individual data^{[19, 27, 30-31, 33-35]}; 9 trials had other types of bias^{[19-20, 22, 29-32, 34-35]}, for example, Keech et al.^[30] did not perform virus neutralization test in the experimental design. In general, the included literature had a low risk of bias (Figure 2 & Table 1).

2.3 The characteristics of the included studies

The 13 included studies were randomized, blinded, and controlled trials, involving 5 inactivated vaccines^{[21-22, 27-29, 34]}, 2 recombinant spike protein vaccines^{[30, 32]}, 2 RNA vaccines^{[20, 31, 33]} and 2 adenovirus vector vaccines^{[19, 35]}. Table 2 for details of vaccine characteristics and developer information). There were 6 studies comparing the effects of single-dose and double-dose vaccination^{[19, 27, 30-31, 33, 35]}. Most of the 13 studies compared the difference of two doses of vaccine at intervals of 2, 3 or 4 weeks. Most studies also compared the difference between low, medium and high injection doses. Participants in all trials were adults, and 5 articles reported the results of vaccines in the elderly population^{[19-20, 32-33, 35]}. The baseline characteristics of the participants were shown in Table 3.

2.4 Qualitative analysis 2.4.1 The effectiveness and safety of vaccines

In 10 studies, the 28-day seroconversion rate of testee exceeded 80%^{[21-22, 27-34]}. The RNA vaccine (BNT162b2) reported by Polack achieved 95% efficiency^[20], the recombinant adenovirus vector vaccine (ChAdOx1 nCoV-19) reported by Voysey achieved an effective rate of 70.4%^[19], but Zhu reported that the 28-day seroconversion rate of the adenovirus recombinant vector vaccine in testee was less than 60%^[35].

In 6 studies, the incidence of adverse reactions in volunteers within 28 days for vaccination was less than 30%^{[20-22, 27-28, 34]}. The adverse reaction rates of the recombinant spike protein vaccine (SCB-2019) reported by Richmond^[32] and the RNA vaccine reported by Walsh^[33] were 34.7% and 39.1%, respectively, and the adverse reaction rates of the RNA vaccine (BNT162b1) reported by Mulligan^[31] and the adenovirus recombinant vector vaccine reported by Zhu^[35] were 52.8% and 73.0%, respectively. Three studies could not obtain the adverse reaction rate^{[19, 29-30]}. The adverse reactions of all vaccinated testee were mostly mild to moderate, and could be relieved within 24 hours after vaccination. The most common local adverse reaction included pain or tenderness at the injection site^{[19-22, 27-35]}. Fatigue was reported as the most common systemic adverse reaction in 9 studies^{[19-20, 22, 28-29, 31, 33-35]}. In addition, fever was reported as the most common systemic adverse reaction in 2 studies^{[21, 27]}, and 2 studies reported somatic pain as the most common systemic adverse reaction^{[30, 32]} (Table 4).

2.4.2 Dose difference

The difference in injection dose is an important factor affecting the immunogenicity and safety of the vaccine. A total of 9 studies^{[21-22, 27-29, 32-35]} found significant differences in GMT and seroconversion rates obtained from testee with different doses of vaccination, 8 of which^{[20-22, 28-29, 31, 34-35]} found that GMT increased, and 4^{[22, 28-29, 32]} found that the seroconversion rate of testee increased with the increase of vaccine dose, but the incidence of adverse reactions also increases relatively ^{[22, 28-29, 32]}. Therefore, when the clinical trial entered Phase III, the researchers set the medium dose as the standard dose of the vaccine ^[19-20].

2.4.3 Difference of age

Four studies specifically recruited the elderly 60 years and older, and conducted a special subgroup analysis in the results. Richmond ^[32] reported that the GMT range measured by the micro-neutralization test in the elderly group was 1567-3625, which was lower than 2510-4452 in the 18-59-year-old group. The incidence of systemic adverse reactions in the elderly after the first injection was 17%, which was lower than 38% in the 18-59 years-old group. Xia ^[27] also reported that the GMT of the elderly group was lower than that of the 18-59 years-old group, and the seroconversion time was later than that of the 18-59 years-old group. The incidence of systemic adverse reactions in the elderly within 7 days after vaccination was 28.6%, which was lower than 41.7% of the 18-59 years-old group. Polack^[20] and Walsh^[33] also reported similar results. In short, compared with healthy people aged 18 to 59, the GMT detected in the serum was significantly lower in elderly population vaccinated with the same vaccine according to the same procedure, but the incidence of adverse reactions in the elderly population was also significantly lower ^{[20, 27, 32-33]}.

2.4.4 Differences in vaccination procedures

Although a number of studies designed a comparison of different vaccination procedures, the results of the experiment were complicated. Zhang 's research showed that testee who vaccinated at 2-week intervals got a faster immune response, but a stronger immune response at 4-week intervals^[34]. Che detected a stronger immune response in testee who were vaccinated at 2-week intervals^[28]. Xia also found that the incidence of adverse reactions in testee vaccinated at 2-week intervals was lower than that at 4-week intervals^[21]. In 6 studies that compared single-dose and double-dose vaccination, 4 studies showed that double-dose vaccination produced a stronger immune response than single-dose vaccination^{[19, 31, 33, 35]}.

2.4.5 Differences of vaccine type

The RNA vaccine (BNT162b2) reported by Polack^[20] and the recombinant adenovirus vector vaccine (ChAdOx1 nCoV-19) reported by Voysey^[19] involved more than 10, 000 people, and two both used relative risk to calculate the effective rate, showing that effective rate of the former was 95%^[20], and the latter was 70.4%^[19]. Owing to differences in the design, the small sample size, and different outcome indicators of other clinical trials, their effective rates were not yet comparable.

3 Discussion

The system evaluation draws the following conclusions: (1) All candidate vaccines have a good immunogenicity and safety except the vaccine reported by Zhu^[35]. Within 28 days after vaccination, the testee' serum GMT increased significantly, and the seroconversion rate was mostly greater than 80%. The adverse reaction rate of most vaccines was less than 30%, degree was mild to moderate, and symptoms were alleviated within 24 hours. (2) The potency and adverse reaction rate after vaccination were positively related to the dose. Most clinical trials chose the middle dose when the phase III. This might be the result of comprehensive consideration of effectiveness and safety. (3) Under the same conditions, the vaccine had poor immunogenicity to elderly people over 60, but the adverse reaction rate was also low. One of the possible reasons was low immunity of the older. A lot of studies on the tolerance of the elderly population to the vaccine still are needed. In addition, there are currently no published results of clinical trials targeting juveniles. (4) Most studies recommend double-dose vaccination, but the interval needs further study.

However, this systematic review has some limitations: (1) No evidence of the long-term effectiveness and safety of the vaccine. Due to the urgency of vaccine development, most trials only followed up to 28 days after vaccination. Whether neutralizing antibodies can be maintained for a long time and whether there are delayed adverse reactions after vaccination still require a longer period. (2) In order to get more up-to-date evidence, this systematic review also includes preprinted documents, which have not been peer reviewed and some of the data are not available. (3) Only randomized, double-blind, and controlled trials were included, while observational studies, retrospective case analysis, and early animal experiments were all excluded. For example, an open label trial conducted by Anderson^[36] found that mRNA-1273 vaccine had a good safety in the elderly population. Logonov^[37] reported two adenovirus recombinant vector vaccine preparations (rAd26) in a non-random clinical trial (rAd26-S and rAd5-S) had a good safety and immunogenicity in healthy people aged 18 to 60. (4) There were differences in the design of various clinical trials, which made it impossible to compare the advantages and disadvantages of different types of vaccines. For example, Voysey^[19] and Polack^[20] used relative risk to calculate the effective rate. Although the remaining 10 studies have completed the virus neutralization test, the experimental design schemes were quite different ^{[21-22, 27-29, 31-35]}. (5) Only Chinese and English documents were searched in this systematic review, and documents published in other languages such as Japanese and French were excluded.

In conclusion, this systematic review summarized the results of clinical trials related to the COVID-19 vaccine, showing that most vaccines had a good safety and effectiveness. It is believed that with the widespread vaccination of COVID-19, it is possible to control and end the global pandemic of COVID-19.

Conflict of interest: The authors have no conflicts of interest to disclose.

新型冠状病毒肺炎(COVID-19)疫情暴发至今已1年余。虽然COVID-19疫情在我国已经得到了有效控制, 但全球整体疫情形势依然严峻。根据世界卫生组织的数据, 截至欧洲中部时间2021年2月15日16 : 05, 全球累计COVID-19确诊病例达到108 579 352例, 累计死亡人数达到2 396 408人^[1]。作为全球的重大公共卫生事件, COVID-19疫情成为全人类首要的健康威胁, 世界政治经济文化也受到巨大冲击^[2-3]。导致COVID-19的严重急性呼吸综合征冠状病毒2(SARS-CoV-2)是以RNA为遗传物质的β属冠状病毒, 通过刺突蛋白结合血管紧张素转化酶2进入细胞^[4-5]。COVID-19患者的首发症状以发热和干咳多见, 在多脏器损伤中, 肺脏受损最为严重^{[2, 5-6]}。在疫情控制上, 佩戴口罩和保持社交距离已经在中国抗击疫情的实践中被确认为阻断病毒传播最为有效的措施^{[3, 7-9]}。在对COVID-19患者的治疗上, 隔离和对症支持治疗仍占主要地位^[5], 而关于抗病毒药物和中药等的疗效还需更多证据的支持^[10-11]。由于口罩在国外普及率的低下和治疗方案的局限性^[12-13], 越来越多的希望被寄托在COVID-19疫苗的开发上。根据靶点和技术的不同, 疫苗可以被分为以下几类: 灭活疫苗、重组刺突蛋白疫苗、病毒载体疫苗、RNA疫苗、减毒活疫苗和病毒样颗粒疫苗等^[14-16]。目前, 已有数百项COVID-19候选疫苗的项目在美国临床试验数据库(clinicaltrials.gov)注册^{[15, 17]}, 数种疫苗的3期临床试验结果予以发表^[18-22]。截至2021年1月1日, 中、俄、美、英等国家先后批准了本国疫苗在人群中的大规模接种计划。本研究通过系统文献复习及定性分析已发表的COVID-19疫苗临床试验结果, 评估COVID-19疫苗的安全性与有效性。

1 资料与方法

本系统评价遵循《系统评价和Meta分析的首选报告项目(PRISMA)》中的准则完成^[23-24]。

1.1 文献纳入标准

文献纳入标准包括: (1)试验对象为18岁及以上的健康男性或未孕女性; (2)干预措施为接种COVID-19疫苗; (3)试验类型为随机、对照、盲法试验; (4)临床试验结果指标至少包括以下一项或几项: 局部不良反应(疼痛、瘙痒、发红、肿胀和硬结等)、全身不良反应(发热、腹泻、疲劳、恶心/呕吐等)、末次疫苗接种14 d或28 d后以活病毒中和试验测得的中和抗体几何平均滴度(GMT)、血清转化率及其他实验室检测指标。

1.2 文献排除标准

具备以下条件之一的文献被排除: (1)文献类型为医学新闻、科普文章、非医学类论文、综述、信件、评论、基础研究、病例报告、会议摘要; (2) 无法获取全文或以中文、英文外的第三种语言发表的文献; (3)若两项研究的受试者存在重叠, 则其中之一被排除; (4)若文献的数据被之后发表的文献包含在内, 前者予以排除。

1.3 文献检索

对英文数据库PubMed、Embase、Cochrane图书馆和clinicaltrials.gov数据库进行了检索。检索的中文数据库包括中国知网、万方数据库、中国生物医学文献服务系统和中国临床试验注册中心。为了保证检索结果的全面性, 本系统评价运用布尔运算逻辑, 采取"主题词+自由词"方式进行了检索。主要检索词包括: COVID-19、2019-nCoV、SARS-CoV-2、2019 novel coronavirus、vaccines、vaccination、COVID-19 vaccines、mRNA-1273 vaccine、Ad5-nCoV vaccine、ChAdOx1 COVID-19 vaccine、BNT162 vaccine、controlled clinical trial、randomized controlled trials、controlled clinical trial、random、blind、placebo、trial、Meta等。中文检索词包括新型冠状病毒、新冠肺炎、新型冠状病毒肺炎、疫苗、试验、随机对照试验、随机对照研究、随机对照、随机、元分析、Meta、荟萃等。

1.4 文献筛选和资料提取

文献筛选和资料提取工作由两位研究者独立完成。若结果汇总时出现分歧, 由两位研究者讨论处理或交由第3位研究者决定。在数据库中获得的所有检索结果导入NoteExpress(武汉大学图书馆版)软件中, 使用软件的查重功能机械地去除重复文献。然后通过阅读标题和摘要完成初次筛选, 通过阅读全文完成二次筛选。在第二次筛选中, 每一篇文献被剔除的原因均被记录。所提取数据包括: 第一作者、疫苗类型、接种剂量、接种间隔时间、受试者人数及基线特征(种族、性别比例、年龄范围或平均年龄)、研究设计方案、局部和全身不良反应、实验室检查指标, 以及基金、赞助商和注册号等。

1.5 方法学质量评价

依据Cochrane系统评价手册评估偏倚风险^[25-26]。

1.6 统计学分析

本系统评价的主要结果包括疫苗的安全性和有效性。评估安全性的指标包括局部不良反应(疼痛、瘙痒、红肿、硬结等)及全身不良反应(咳嗽、腹泻、疲倦、发烧、头痛、恶心/呕吐、瘙痒、肌肉疼痛、关节痛/不适、厌食等)。评估免疫原性的指标包括GMT、血清转化率、IgG或其他特异性抗体对受体结合域的反应。

2 结果 2.1 文献检索结果

检索了截至2020年12月31日之前发表的所有相关文献, 共得到753篇。经过筛选后纳入13篇^{[19-22, 27-35]}进入本系统评价。文献筛选的具体流程见图 1。

2.2 纳入研究的方法学质量评价

纳入的13项研究^{[19-22, 27-35]}均采用了随机对照的方法, 其中10项^{[21-22, 27-32, 34-35]}实施了双盲法, 2项^{[20, 33]}实施了单盲法, 1项^[19]在不同试验地点分别使用了单盲法和双盲法; 所有试验均隐藏了分配方案; 9项^{[19, 22, 27, 29-31, 33-35]}数据不完整或选择性报告, 其中2项^{[22, 29]}预印本缺失数据较多, 其余7项^{[19, 27, 30-31, 33-35]}缺失个别数据; 9项^{[19-20, 22, 29-32, 34-35]}存在其他类型偏倚, 如Keech等^[30]在试验设计中未做病毒中和试验。总的来讲, 所纳入文献的偏倚风险较低。见图 2和表 1。

2.3 纳入研究的基本特征

所纳入的13项研究均为随机、盲法、对照试验, 共涉及灭活疫苗5种^{[21-22, 27-29, 34]}、重组刺突蛋白疫苗2种^{[30, 32]}、RNA疫苗2种^{[20, 31, 33]}和腺病毒载体疫苗2种^{[19, 35]}, 疫苗特性、开发者等信息见表 2。有6项研究比较了疫苗单剂量与双剂量接种的效应^{[19, 27, 30-31, 33, 35]}。大部分研究比较了以2周、3周或4周为间隔注射两剂疫苗的差别。大部分研究也比较了低、中、高不同注射剂量的差别。所有试验的参与者均为成年人, 有5篇文献报道了疫苗在老年人群体中的结果^{[19-20, 32-33, 35]}。所纳入研究参与者的基线特征见表 3。

2.4 定性分析结果 2.4.1 疫苗的有效性和安全性

在10项研究中, 受试者的28 d血清转化率超过80%^{[21-22, 27-34]};在两项万人规模的临床试验中, Polack等^[20]报道的RNA疫苗(BNT162b2)取得了95%的有效率, Voysey等^[19]报道的腺病毒重组载体疫苗(ChAdOx1 nCoV-19)取得了70.4%的有效率; Zhu等^[35]报道的腺病毒重组载体疫苗在受试者中的28 d血清转化率低于60%。见表 4。

在6项研究中, 志愿者在接种疫苗后的28 d内不良反应发生率低于30%^{[20-22, 27-28, 34]};Richmond等^[32]报道的重组刺突蛋白疫苗(SCB-2019)和Walsh等^[33]报道的RNA疫苗的不良反应率分别为34.7%和39.1%;Mulligan等^[31]报道的RNA疫苗(BNT162b1)和Zhu等^[35]报道的腺病毒重组载体疫苗的不良反应率分别为52.8%和73.0%;3项研究无法获取不良反应率^{[19, 29-30]}。所有疫苗接种的受试者发生不良反应事件绝大部分都是轻度到中度, 且在接种后24 h内可缓解; 所有疫苗接种最常见的局部不良反应均为注射部位疼痛或压痛^{[19-22, 27-35]};疲劳在9项研究中被报道为最常见的系统性不良反应^{[19-20, 22, 28-29, 31, 33-35]}。此外, 发热在2项研究中被报道为最常见的系统性不良反应^{[21, 27]}, 也有2项研究报道躯体痛为最常见的系统性不良反应^{[30, 32]}。见表 4。

2.4.2 剂量差异的影响

注射剂量的不同是影响疫苗免疫原性和安全性的重要因素。共有9项研究^{[21-22, 27-29, 32-35]}发现接受不同剂量疫苗接种的受试者获得的GMT和血清转化率存在显著性差异, 其中8项^{[20-22, 28-29, 31, 34-35]}发现GMT随着疫苗剂量的增加而增加, 4项^{[22, 28-29, 32]}发现受试者血清转化率随疫苗剂量的增加而增加。但随着接种剂量的加大, 不良反应的发生率也相对增加^{[22, 28-29, 32]}。因此, 当临床试验进入Ⅲ期阶段, 研究者将中等剂量设定为疫苗的标准剂量^[19-20]。

2.4.3 年龄差异的影响

有4项研究专门招募了60岁及以上的老年人群, 并在结果中进行了专门的亚组分析。Richmond等^[32]报道使用微量中和试验在老年人组测得的GMT范围为1 567~3 625, 低于18~59岁组的2 510~4 452;而老年人在第1次注射后的全身不良反应发生率为17%, 低于18~59岁组的38%。Xia等^[27]也报道老年人组GMT低于18~59岁组, 且达到血清转化时间晚于18~59岁组; 而老年人在接种后7 d内的全身不良反应发生率为28.6%, 低于18~59岁组的41.7%。Polack等^[20]和Walsh等^[33]两项研究也报道了相似结果。总之, 相比于18~59岁的健康人群, 老年人群按照相同的程序接种同种疫苗后, 血清中所检测到的GMT显著偏低, 但相应地老年人群中不良反应发生率也显著偏低^{[20, 27, 32-33]}。

2.4.4 接种程序差异的影响

虽然多项研究设计了不同接种程序的对比, 但试验结果是复杂的。Zhang等^[34]的研究表明, 以2周为间隔接种疫苗的受试者获得了更快的免疫反应, 但以4周为间隔接种疫苗的受试者获得了更强的免疫反应。但Che等^[28]在以2周为间隔接种疫苗的受试者中检测到了更强的免疫反应, Xia等^[21]也发现以2周为间隔接种疫苗的受试者不良反应发生率低于以4周为间隔接种疫苗的受试者。在6项比较了疫苗的单剂量与双剂量接种的研究中, 4项研究显示疫苗双剂量接种比单剂量接种产生更强的免疫反应^{[19, 31, 33, 35]}。

2.4.5 疫苗类型差异的影响

Polack等^[20]报道的RNA疫苗(BNT162b2)和Voysey等^[19]报道的腺病毒重组载体疫苗(ChAdOx1 nCoV-19)受试者人数超过10 000人, 都采用相对危险度计算有效率, 显示前者有效率为95%^[20], 后者有效率为70.4%^[19]。其他临床试验的设计存在差异, 受试者规模较小, 结局指标也有所不同, 其有效率尚无法比较。

3 讨论

本系统评价得出以下结论: (1)除了Zhu等^[35]报道的疫苗外, 所有候选疫苗都具有良好的免疫原性和安全性。接种后28 d内, 受试者血清GMT显著增加, 血清转化率大多大于80%, 大部分疫苗的不良反应率低于30%, 且以轻到中度为主, 24 h内缓解。(2)接种后产生的效价和不良反应率与剂量呈正相关, 因此, 大部分临床试验进入Ⅲ期阶段后, 选择了中等剂量作为标准剂量, 这可能是对有效性和安全性综合考虑的结果。(3) 相同条件下, 疫苗对60岁以上的老年人的免疫原性较差, 但不良反应率也偏低, 一种可能的解释是这与人体的免疫衰老有关。老年人群对疫苗的耐受性需要继续研究。此外, 目前尚没有针对未成年人的临床试验结果发表。(4)大部分疫苗研究都推荐双剂量接种, 但接种间隔时间需进一步研究。

然而, 本系统评价有一定的局限性: (1)缺乏疫苗的长期有效性和安全性的证据。由于疫苗研发的急迫性, 大部分试验只随访到了接种后28 d, 中和性抗体能否长期维持, 接种疫苗后是否有迟发的不良反应, 仍需要更长时间的随访。(2)为了纳入更多最新证据, 本系统评价也将预印本文献包含在内, 这些文献没有经过同行评议, 且其中一些数据无法获取。(3)本系统评价只纳入了随机、双盲、对照试验, 而观察性研究、回顾性病例分析及早期的动物试验均被排除在外。如Anderson等^[36]实施的一项开放标签试验发现mRNA-1273疫苗在老年人群体具有较好的安全性, Logunov等^[37]在非随机临床试验中报道了两种腺病毒重组载体疫苗制剂(rAd26-S和rAd5-S)在18~60岁健康人群具有较好的安全性和免疫原性。(4)各项临床试验的设计存在差异, 导致无法对不同类型疫苗的优劣进行比较, 如Voysey等^[19]和Polack等^[20]采用相对危险度计算有效率, Keech等^[30]未做病毒中和试验, 其余10项研究虽然均完成了病毒中和试验, 但试验设计方案差异较大^{[21-22, 27-29, 31-35]}。(5)本系统评价只检索了中英文文献, 以日文、法文等其他语言发表的文献被排除在外。

综上所述, 本系统评价总结了COVID-19疫苗相关的临床试验结果, 表明大部分疫苗都具有较好的安全性和有效性。这让我们有理由相信, 随着COVID-19疫苗的广泛接种, 有望控制、终结COVID-19的全球大流行。

利益冲突声明：所有作者均声明不存在利益冲突。