1. Introduction
COVID-19 is a term familiar to almost everyone today. For the past few months, there has been a massive surge in cases, and along with it, fears regarding the Omicron variant of COVID-19. The B.1.1.529 COVID-19 variant was initially detected in South Africa, first reported to the World Health Organization (WHO) on 24 November, identified from a specimen collected on 9 November 2021. However, due to a quick rise in cases of this variant in South Africa, it was classified as a variant of concern (VOC) with the name Omicron by the World Health Organization Technical Advisory Group on SARS-CoV-2 Virus Evolution (WHO TAG-VE) on 26 November 2021. Currently, the most common sub-lineages of Omicron are BA.1, BA.1.1, and BA.2 [
1].
The rise of the Omicron variant remains a topic of great concern among scientists worldwide, as it harbours mutations that have potential implications on vaccine effectiveness, viral transmissibility, disease severity and immune evasion [
2,
3]. As of 29 December 2021, 37 mutations on the spike or S protein were identified, with some being found in other VOCs such as the alpha, beta, gamma and delta variant [
4]. The primary fear is the mutations associated with the spike protein’s receptor-binding domain (RBD) that facilitates binding with the angiotensin-converting enzyme 2 (ACE2) for viral entry into host cells [
5]. Research has shown that the mutations on the RBD of Omicron have caused stronger binding between RBD and ACE2 [
6]. These findings are worrying as most COVID-19 vaccines in the current market induce humoral immunity that identifies the spike protein as the target [
7]. Therefore, the mutations, particularly in the RBD region, are worrying because they can reduce the efficacy of current vaccination regimens, and thus cause a colossal setback in global efforts to make COVID-19 endemic.
Malaysia’s first Omicron case was reported on 3 December 2021, where the patient was a traveller from South Africa, the origin of the Omicron variant. Since then, the Omicron variant has been rampant across the country, surpassing the transmission of the previous majority Delta variant. From the genomic surveillance of COVID-19 in Malaysia, in December 2021, about 50% of the samples processed were of the Omicron variant, with all of them being imported cases detected upon international arrival. However, in January 2022, the percentage of Omicron cases increased to approximately 70%, with most of the cases being imported but some starting to appear in multiple states across the country. In February 2022, 98.7% of GISAID submissions by Malaysia were of the Omicron variant [
4,
8]. In order to handle the COVID-19 pandemic in Malaysia, Act 342—Prevention and Control of Infectious Diseases Act 1988 was enforced [
9]. Under this enforcement, as listed in the COVID-19 guidelines issued by the Ministry of Health Malaysia [
10], all suspected cases, probable cases, and close contacts must be tested using either an antigen rapid test kit (RTK-Ag) or reverse-transcription polymerase chain reaction (RT-PCR). It is mandatory by law for individuals with positive test results to report to the ministry. For close contacts, they are issued a home surveillance order to undergo compulsory home isolation and report antigen rapid test kit results during the quarantine period. Likewise, vaccination against COVID-19 is a nationwide government initiative, and all vaccination and verification is carried out through a single nationwide system known as MySejahtera, even if the vaccine is administered in private healthcare facilities or outstations. Official COVID-19 data is updated and released to the public on a daily basis through the COVIDNow website, which is sourced from a GitHub public repository maintained by the Ministry of Health Malaysia.
To overcome the waning of vaccine effectiveness caused by the Delta variant, the Malaysian government decided to introduce the National COVID-19 Immunization Programme—Booster (PICK-B) on 13 October 2021. At the start, the program prioritizes the administration of booster doses to high-risk individuals and frontliners [
11]. In addition, recipients of CoronaVac, as the primary vaccination regimen, were included in the priority list due to the observation that numerous studies showed that CoronaVac has lower efficacy than other brands, providing limited protection against COVID-19 VOCs [
12]. After that, booster doses were administered to all fully vaccinated individuals, involving the cooperation between both the government and the private health sector from 30 September 2021 onwards [
11]. The top three types of vaccines administered are BNT162b2 (Pfizer-BioNTech Comirnaty), CoronaVac (Sinovac) and AZD1222 or ChAdOx1 (Oxford-AstraZeneca Vaxzevria) [
13], administered to 41.4%, 30.7% and 6.2% of the population, respectively, as of 31 March 2022.
Currently, most data on the vaccine effectiveness of COVID-19 vaccines have been gathered mainly through randomized controlled trials (RCTs). It is indeed understandable why information obtained from RCTs is generally preferred by many researchers, due to its strict inclusion criteria, continual observation and specific populations, which ensures high internal validity [
14]. However, the stringent controlled settings of RCTs do not represent the real-world effectiveness of the vaccines. Real-world data provide valuable information outside the scope of RCTs, such as any unpredicted outcomes in the real-world pertinence of antibody responses after vaccination. The study of real-world effectiveness also takes into account factors such as the rollout and distribution of vaccines, vaccine supply and demand dynamics and government policies [
15]. These elements are particularly relevant in low- and middle-income countries (LMICs) in the course of mitigating COVID-19 fatalities in respective countries via vaccination [
16].
Therefore, this study aims to investigate the effects of booster doses on the Omicron wave in Malaysia by using real-world data established from nationally representative statistics on COVID-19 cases, deaths and vaccinations, and thereby estimating vaccine effectiveness against COVID-19 infections and deaths, comparing a full vaccination and a boosted vaccination regimen. A comparison will also be made between the major three vaccine brands authorized for use in Malaysia.
4. Discussion
Marked by daily new infections, new deaths, ICU utilization, and Rt values of COVID-19, the Omicron wave commenced in January 2022, peaking in the first half of March 2022, and then showed a downwards trend. Such a trend signifies the probable success of booster vaccination in curbing the Omicron wave in Malaysia, as booster vaccination rates steadily increased, peaking in mid-January, with an average of over 200,000 new booster vaccines being administered daily [
13]. Although the wave had not officially ended at the time of writing, it was postulated that the downtrend would continue as the wave had already peaked, and ICU utilization and deaths were trending downwards as well. In addition, Rt values also indicated that the downtrend would continue, possibly towards the end of the Omicron wave. One of the potential factors that contributed to the downtrend was the achievement of herd immunity within the Malaysian population, through a high quantity of infections and the partial protection conferred by primary and boosted vaccinations. It is noteworthy that Malaysia’s Rt values were much lower compared to other countries [
21], due to Malaysia being a tropical country and not a four-season country. Hence, the warm weather all year round might have prevented the worsening of COVID-19.
This study evaluates the importance of the booster vaccine rollout to Malaysians in curbing the COVID-19 pandemic. Generally, the results from this study showed that boosted vaccination is more effective in preventing total and symptomatic COVID-19 infections during the Omicron wave in Malaysia than those with two-dose vaccination without a booster. The result of this study is consistent with other studies on the effectiveness of booster doses against VOCs [
22]. Unfortunately, vaccine effectiveness against COVID-19 infections could not be stratified by vaccine types as there is an absence of individual-level data on uninfected individuals, and the chosen screening method does not allow such stratification. Hence, vaccine effectiveness between different vaccines can only be postulated from the results of vaccine effectiveness against COVID-19 deaths that provides for such stratification. As for deaths among COVID-19 patients, the same trend could be observed where boosted vaccination is more effective than without a booster. Research has shown that boosters reactivate memory B cells against the spike protein, which were capable of recognizing the mutated Omicron RBD. Boosters also enhance T-cell reactivity towards Omicron [
23,
24].
The waning of vaccine effectiveness for two doses was expected due to the Omicron variant. Suah et al. reported that during the Delta wave in Malaysia, the efficacy of CoronaVac and BNT162b2 was reduced compared to the previous waves of COVID-19 [
20]. Nevertheless, COVID-19 booster vaccines of any kind were proven effective in lowering COVID-19 infections and deaths in Malaysia. Although two doses were still providing some degree of protection at the time of writing, it was still considered vital and highly encouraged for boosters to be taken to confer a higher degree of protection against COVID-19 including the Omicron variant, especially for CoronaVac recipients and senior citizens.
Looking at the VE against deaths, AZD1222 and BNT162b2 showed better protection against COVID-19, followed by CoronaVac. Despite this, the protection the two vaccines elicited without a booster against death was 75.1 (95% CI: 68.0, 80.6) and 72.3 (95% CI: 68.9, 75.3), respectively. VE could not be calculated when CoronaVac was taken as a booster for AZD1222 and BNT162b2 primary schedules. This is because under the PICK-B program, the recipients of both primary schedules were recommended to take either AZD1222 or BNT162b2 as boosters. Therefore, there were too few individuals who had taken CoronaVac as the booster to calculate the VE accurately. AZD1222 is an adenovirus vector-based vaccine encoding the full-length spike protein sequence, whereas BNT162b2 is an mRNA-based vaccine encoding the RBD of the S1 protein only [
25].
Various articles and preprints indicated that two doses of AZD1222 or BNT162b2 were insufficient to protect against the Omicron variant, and boosting was required to reinstate significant protection [
26,
27]. Our study has shown that a booster of any kind increased the VE against death to 95.1 (95% CI: 92.9, 96.6) and 93.3 (95% CI: 92.1, 94.3) for AZD1222 and BNT162b2 primary recipients, respectively. This result is in line with other studies. For instance, an AZD1222 booster increased neutralizing antibody titres by 3.25-fold and 5.33-fold for AZD1222 and BNT162b2 primary recipients, respectively, in a phase II trial conducted in the United Kingdom [
28]. Furthermore, a Danish study (preprint) revealed that the VE of BNT162b2 primary vaccination was only 55.2% (95% CI: 23.5, 73.7) against the Omicron variant, but the VE dropped significantly to 9.8% after five months. A homologous booster administered within five months post-second dose restored the VE to 54.6% (95% CI: 30.4, 70.4) [
29]. Moreover, in this study, it is interesting to note that taking BNT162b2 as a heterologous booster for AZD1222 primary recipients showed better VE compared to a homologous booster. It was proven that a heterologous prime-boost vaccination of AZD1222-BNT162b2 is more effective than BNT162b2-AZD1222. Nevertheless, for BNT162b2 primary recipients, a homologous prime-boost was more effective than heterologous prime-boost vaccinations [
30].
For CoronaVac, the primary two-dose regimen was marginally effective to confer protection in this study. A booster dose of any brand is needed to prevent deaths with 89% efficacy, presumably preventing COVID-19 infections. Research has shown that the RNA-based and viral vector-based vaccines generate a more robust immune response than their inactivated virus-based counterparts. The vaccine candidate for CoronaVac was the inactivated virus of the early outbreak CN2 strain from China [
31]. Clinical trials have also shown waning vaccine efficacy for CoronaVac towards VOCs, going down to a VE of 50.7% [
32]. Nevertheless, CoronaVac still has advantages. Antibodies elicited by BNT162b2 can only bind to the spike protein, while CoronaVac-generated antibodies can bind to both the spike protein and the nucleocapsid [
33]. Furthermore, CoronaVac is superior in stimulating CD4
+ and CD8
+ T-cell responses to the structural protein than BNT162b2, potentially compensating for the waning of neutralizing antibody protection [
34]. At the time of writing, we are currently determining whether CoronaVac boosters to primary vaccination with BNT162b2 or AZD1222 would stimulate better T cell responses against the spike protein than boosters of the other two vaccine types. Nevertheless, it should be noted that VE from other studies is not to be used for direct comparison of VE in our study, as the methods of analysis used and the sample sizes are vastly different. The VE from other studies is cited here to provide evidence regarding the waning of vaccine effectiveness over time and a crude comparison of VE between the three vaccine brands.
Data presented here have shown that it is recommended that CoronaVac recipients be boosted by BNT162b2 and AZD1222. Furthermore, many studies have supported heterologous prime-boost vaccination for CoronaVac recipients as it has been shown to improve the immunological responses generated. For instance, a study in Turkey showed that a heterologous booster of BNT162b2 given to CoronaVac recipients resulted in a 104.8-fold increase in antibody levels. In contrast, a homologous CoronaVac booster caused an 8.7-fold rise [
35]. Another study demonstrated that for CoronaVac recipients, boosters have resulted in a geometric fold-increase of IgG titre of 152 (95% CI: 134, 173) for BNT162b2, 90 (95% CI: 77, 104) for AZD1222, and 12 (95% CI: 11, 14) for CoronaVac. This information benefits older populations with weaker immune responses to elicit better protection against COVID-19 [
36].
The strength of this study is being one of the first to determine the efficacy of a booster vaccine against the Omicron variant using nationwide data instead of smaller-scale clinical trials. In addition, data on COVID-19 in Malaysia used in this study were obtained from a firmly established official and direct repository under the Ministry of Health Malaysia. Because of this, the results from this study will contribute greatly to real-world data on booster vaccine effectiveness. This study also demonstrates the effects of booster vaccinations in LMICs. However, this study also contains some limitations. First, using death as an indicator for the severe outcome of COVID-19 is insufficient in evaluating vaccine effectiveness. However, based on WHO guidelines [
37], using deaths as severe outcome is plausible in this case as our nationwide database in terms of COVID-19 infections, hospitalization and vaccination is accurate and up-to-date, as it is mandatory by law. Using this, we demonstrated a large vaccine impact on COVID-19. Still, the method of analysis in this study was dependent on the granularity of data. Aggregated data on COVID-19-related ICU admissions could not be used in the screening method as a severe outcome due to the lack of data stratified into the vaccination status, i.e., unvaccinated, fully vaccinated and boosted. Therefore, VE against infection can only be estimated, using a combination of mild and severe outcomes using screening method. Moreover, as the data on the type of vaccine received were only available for cases (infected) but not controls (uninfected), the VE against infection could only be estimated using the screening method, not multivariable logistic regression. Had it been possible, the same method would have been carried out for both analyses. Furthermore, at the time of writing, genomic surveillance of COVID-19 was still inadequate in Malaysia, as the submission of sequenced data to GISAID was voluntary. Hence, variant-specific vaccine effectiveness, including sub-lineages such as Omicron BA.1, BA.1.1 and BA.2, could not be determined. The spreading of the Omicron variant could only be postulated from the GISAID data trend and the media statement of the Ministry of Health Malaysia.
At the time of writing, the Ministry of Health Malaysia had ruled that all non-boosted CoronaVac and Sinopharm BBIBP primary vaccine recipients above 18 years old and all non-boosted senior citizens above 60 years old, regardless of primary vaccination, had lost their status as fully vaccinated. This move reinforced the urgency of getting a booster dose to enhance antibody levels, as CoronaVac’s efficacy had been shown to drop substantially over time, and senior citizens are high-risk individuals [
38]. After 1 April 2022, Malaysia officially entered the “Transition to Endemic” phase since all indicators such as cases, ICU admissions, and death were on a downward trend. As a result, national border restrictions were lifted, and more social activities were allowed with relaxed standard operating procedures compared to previous phases. Although the peak of COVID-19 cases was higher, COVID-19-related ICU admissions and deaths were, at the time of writing, 80–90% lower during the Omicron wave than during the Delta wave [
39]. This accomplishment can be attributed to the success of the PICK-B program in distributing and administrating boosters, thereby bringing down the cases to a controllable level, as proven by this study. However, at the time of writing, mask-wearing was mandatory and social distancing measures were highly encouraged to continue the promising downward trend of COVID-19 cases.