Abstract
Background In 2022-2023 a global outbreak of mpox affected mostly gay, bisexual and other men having sex with men (GBMSM). Outbreak control in the Netherlands included isolation, quarantine, post-exposure prophylaxis vaccination, and primary preventive vaccination (PPV). We describe the course of the outbreak, the vaccination programme, vaccine effectiveness (VE) of full vaccination against symptomatic disease, and trends in behaviour to generate hypotheses about factors that influenced the outbreak’s decline.
Methods Observational study. Data on notified cases, number of PPV invitations, and PPV doses administered were collected from public health services. PPV uptake and coverage were calculated. Monthly trends in behavioural data of GBMSM visiting Sexual Health Centres (SHC) were analysed for all consultations in 2022. VE was estimated using the screening method.
Findings A total of 1,266 mpox cases were reported until August 1, 2023. The outbreak peaked early July 2022 and sharply declined afterwards. PPV started July 25, 2023; in total 29,851 doses were administered. 45.8% received at least one dose, 35.4% was fully vaccinated. Estimated VE was 68.2% (95% CI 4.3%-89.5%). In our analyses of behavioural data, we did not observe an evident decrease in high-risk behaviour.
Interpretation It is unlikely that PPV was a driver of the outbreak’s decline, as incidence started to decline well before the start of the PPV programme. The possible impact of behavioural change could not be demonstrated with the available indicators. We hypothesise that infection-induced immunity in high-risk groups contributed most to the decline.
Funding Dutch Ministry of Health, Welfare and Sport.
Background
Since the first reported human case of mpox (caused by monkeypox virus (MPXV)) in 1970 in the Democratic Republic of the Congo, endemic transmission of mpox was for decades only reported from African countries. Sporadic travel-associated cases were described outside of the African continent,1,2 and in 2003 an outbreak in the US was caused by imported mammals.3–5 In 2022 a global outbreak of mpox occurred, affecting mostly gay, bisexual and other men who have sex with men (GBMSM).1,6,7 Up to August 1, 2023 there were over 88,000 confirmed cases worldwide in 113 countries.8 The first patient in the Netherlands was diagnosed May 20, 2022 and the outbreak peaked in the beginning of July. The first 1,000 cases in the outbreak of mpox in the Netherlands until August 8, 2022 have been described by van Ewijk et al.9 Since then, the incidence continued to decrease. Up to August 1, 2023 an additional 266 persons with laboratory confirmed mpox have been reported to the Dutch National Institute for Public Health and the Environment (RIVM), resulting in a total of 1,266 confirmed mpox cases. In 2023, only 6 infections were diagnosed, of which the most recently diagnosed person had a date of symptom onset of May 26, 2023.
Outbreak control in the Netherlands included isolation of persons with mpox, and quarantine (later replaced by lifestyle recommendations) and post-exposure prophylaxis (PEP) vaccination with Modified Vaccinia Ankara-Bavarian Nordic (MVA-BN; also known as Imvanex or Jynneos), a third-generation smallpox vaccine,10 for high- and medium-risk contacts of all individuals diagnosed with mpox. Also, a primary preventive vaccination (PPV) programme with MVA-BN started on July 25, 2022, targeting those at highest risk of infection.11 Some variability in starting dates existed between different regions, but all regions started before August 11, 2022. Vaccination against smallpox was ceased in 1976 in the Netherlands.12 It is estimated that at the start of the 2022 mpox outbreak, 57%9 to 68%13 of the Dutch population did not have protective antibodies against Orthopox viruses. GBMSM and transgender persons were alerted on behaviours that would put them at higher risk of mpox and preventive measures to mitigate the risk through a targeted national communication campaign.14 Here, we describe the mpox outbreak in the Netherlands, the design and implementation of the vaccination programme, and trends in behavioural data to generate hypotheses about the factors that may have contributed to the decline of the outbreak in the Netherlands. Furthermore, we aim to estimate the vaccine effectiveness of MVA-BN against symptomatic mpox disease.
Methods
Data collection
All individuals diagnosed with mpox were reported to the RIVM in an online reporting system as part of the mandatory notification (see Supplementary Information).15 For all individuals diagnosed with mpox, information was collected on, amongst others, demographics, sexual orientation, potential source(s) of infection and results of contact tracing. A detailed description of the variables collected can be found in van Ewijk et al.9
The following groups were eligible for PPV: 1a. GBMSM and transgender persons using HIV-pre-exposure prophylaxis (PrEP) via Sexual Health Centres (SHC); 1b. GBMSM and transgender persons on the waiting list for the HIV-PrEP-pilot via SHC; 1c. GBMSM and transgender persons using HIV-PrEP via their general practitioner (GP); 2. GBMSM and transgender persons living with HIV at high-risk for mpox infection; 3. Other GBMSM and transgender persons at high-risk for mpox infection.11 All Public Health Services (PHS) identified persons eligible for PPV, and personally invited them. All vaccinations were given subcutaneously. The estimated numbers of PPV invitations sent out per indication group and all administered PPV doses were registered by all PHS in a dedicated online reporting system. For vaccinated persons that did not give consent to share their vaccination information with the RIVM, only the following anonymised data were collected: year of birth in three categories (<1985, 1985-1994, >1994), childhood smallpox vaccination yes/no, vaccination number (1st or 2nd), week of vaccination, and PHS location. For persons that gave consent to share pseudonymised data, this was supplemented with: unique identifier, year of birth, gender, indication group, immunocompromised (yes/no), received PEP vaccination (yes/no), date of PEP vaccination, and date of PPV. Administered PEP vaccinations were not registered on a national level. Details on PEP vaccination, the PPV programme and selection of persons eligible for PPV can be found in the Supplementary Information.
In the Netherlands, data regarding all visits to SHC are available in a national surveillance database (SOAP) of the RIVM. This database contains pseudonymised data from all consultations, including behavioural data. Consultations can be divided into regular SHC consultations and HIV-PrEP-pilot consultations. Persons participating in the HIV-PrEP-pilot are regularly tested for STIs at SHC (three-monthly), independently of their sexual behaviour. Monitoring behaviour in this group will therefore give a more accurate representation of possible changes in behaviour, compared to persons who consult a SHC because of symptoms or possible risk incurred. Therefore, only HIV-PrEP consultations of GBMSM and transgender persons (including non-binary persons) were included. As persons could have multiple consultations in 2022, analyses were done on consultation level. Variables extracted from the SOAP database used as proxies for behaviour with high risk for mpox infection were number of sex partners, group sex (yes/no; defined as multiple sex partners at once or over a short period), chemsex (yes/no; defined as using drugs [cocaine, XTC/MDMA/Speed, Heroin, Crystal Meth, Mephedrone, 3-MMC, 4-MEC, 4-FA, GHB/GBL and ketamine] before or during sex), condomless anal sex (always condomless vs. sometimes/never; receptive and/or insertive) and sex work (yes/no). All behavioural variables are self-reported behaviour in the past six months. Furthermore, number of tests and percentage positivity for chlamydia, gonorrhoea and syphilis were also analysed.
Data on notifications were included until August 1, 2023, data on vaccinated persons were included until April 30, 2023; and behavioural data were collected from all SHC consultations in 2022.
Data analyses
For individuals diagnosed with mpox, we assumed people born ≥1976 in the Netherlands or ≥1980 in another country did not receive a childhood smallpox vaccination.
Mpox vaccine uptake was defined as the proportion of invited people receiving at least one dose. Mpox vaccination coverage was defined as the proportion of invited people being fully vaccinated. A person was considered fully vaccinated if they had received two MVA-BN vaccinations; or if they received one MVA-BN vaccination and a childhood smallpox vaccination and were not reported to be immunocompromised. For the mpox vaccinees we had no information on country of birth. We assumed that people born <1976 with an unknown childhood smallpox vaccination status did receive this vaccination as a child, while childhood smallpox vaccination status was set to ‘No’ for everybody born ≥1980. For those born between 1976 and 1980, smallpox vaccination status was used as reported. If it was unknown if a person was immunocompromised, we assumed they were not. For both the analyses of the uptake and vaccination coverage, vaccinations administered to people who were not in the predefined indication groups were excluded. Uptake and vaccination coverage were calculated for all invitees, and by indication group.
Vaccine effectiveness (VE) against laboratory confirmed symptomatic mpox disease was estimated for those who were fully vaccinated using the screening method as described by Farrington.16 This method compares the vaccination coverage in the cases with the coverage in the population they arose from, with the VE calculated as 1 – odds of vaccination in cases/odds of vaccination in the population. Cases were included in the analyses if they were male at birth, indicated they had sex with men, and if their vaccination status was known. For these analyses, persons were considered fully vaccinated using the definition described before, plus their last vaccination should be at least 14 days before symptom onset to allow time for vaccine-induced protection.17 Also, at least one of the vaccinations received should be a PPV, to make the cases comparable to the population where the vaccination coverage was calculated from. Cases that had received their last vaccination less than 14 days before onset of symptoms were regarded incompletely vaccinated and were included in the total number of cases, but not in the completely vaccinated cases. The proportion of fully vaccinated among the cases was calculated per week. The vaccination coverage in the population (i.e., the proportion of fully vaccinated persons of those invited for PPV) was calculated per week as well. Here, also a 14-day delay was considered to allow time for vaccine-induced protection. Increasing vaccination coverage was accounted for by fitting a logistic regression model with the logit of the proportion of the population vaccinated as offset.16 To reduce bias introduced by the small sample size, Firth correction was applied to the logistic regression model. As the number of mpox infections dropped dramatically by the end of 2022, with 0 persons to include in December, analyses were done for August up to and including November.
To describe trends in behaviour among GBMSM and transgender persons, the numbers and proportions of consultations in which each sexual behaviour variable was reported, and the number of tests and positivity rates for chlamydia, gonorrhoea and infectious syphilis were plotted per month. To assess a possible significant decrease in risk behaviour (p<0.05), we compared the proportions in June, July and August (outbreak months) to May (pre-outbreak month) using Chi square tests. For non-normally distributed continuous variables the Mann-Whitney U test was used. All analyses were done in R version 4.2.0 and Excel version 2022.
Role of the funding source
The funder had no role in in study design; in the collection, analysis, and interpretation of data; in the writing of the report; and in the decision to submit the paper for publication.
Results
Epidemiology
Of all individuals diagnosed with mpox in the Netherlands, nearly all (99%; 1,247/1,265, for 1 person sex was unknown) were male at birth, of whom 94% (1,171/1,247) reported to have sex with men. Of those for whom information about gender identity was known, 98% identified as a man (1,076/1,098). The median age was 37 years (IQR 31-45). Of those eligible for childhood smallpox vaccination (born <1976 in the Netherlands or <1980 in another country) and with complete smallpox vaccination information, only 62% (168/273) reported to have received this vaccination in the past. Of those with information about PEP vaccination status for mpox, 4% (52/1,171) reported to have received PEP, but only in two instances it was given at least 14 days before the date of symptom onset. Of those with a known PPV status, 7% (56/768) reported to have received PPV at any time during the outbreak. In 12 instances the last vaccination was given at least 14 days before the date of symptom onset. 51% (640/1,266) of the individuals diagnosed with mpox originated from the Amsterdam region. Demographical, epidemiological and clinical characteristics of the 266 more recently diagnosed persons are comparable to the first 1,000 cases. For further details on the characteristics of these cases, we refer to van Ewijk et al.9
Vaccination
In total, 39,657 invitations for vaccination were sent out by the 25 PHS in the Netherlands. Overall, 29,851 PPV doses were administered between July 25, 2022 and April 5, 2023. Consent to share vaccination information with the RIVM was given for 26,993 vaccinations (90.4%). In total, 17,792 first vaccinations were given, and 11,846 second. 213 vaccinations had an unknown ranking number. These first and second PPV doses were administered to 18,684 unique persons, predominantly identifying as man. Background characteristics of the vaccinated persons can be found in Table 1.
The overall uptake of vaccination (at least one dose received) was 45.8% (18,157/39,657) with 35.4% (14,048/39,657) being fully vaccinated by April 30, 2023, with considerable variation between the different indication groups (Table 2). The highest uptake and vaccination coverage were found in GBMSM and transgender persons using HIV-PrEP, either via a SHC (indication group 1a) or GP (indication group 1c). The vaccination coverage in relation to the epidemic curve can be found in Figure 1.
If the date of symptom onset was unknown, the earliest of either sampling date or notification date to the PHS was used. A bin width of 7 days was used and data were grouped by PHS Amsterdam (dark grey) versus the rest of the Netherlands (orange). Black line: vaccination coverage (fully vaccinated persons).
Vaccine effectiveness
Between week 32 and 48 of 2022 (August 8 – December 4), a total of 181 individuals diagnosed with mpox were notified to the RIVM. Of those, 19 were excluded as they did not meet inclusion criteria to calculate the VE (i.e., not male, did not report to have sex with men, or vaccination status unknown). Table 3 shows the 162 included cases and VE per week. Overall VE of full vaccination against laboratory confirmed symptomatic mpox disease was estimated to be 68.2% (95% confidence interval 4.3% - 89.5%). The three fully vaccinated cases all received two PPV doses during the current vaccination programme.
Behavioural change
In 2022, 73,885 consultations were registered among GBMSM and transgender persons at SHC, of which 28,570 (38.7%) were HIV-PrEP-pilot consultations. The median age of included persons consulting SHC for HIV-PrEP was 37 years (IQR 30-49). In this group, Chi square analyses revealed a small but statistically significant decrease in the proportion of consultations where sex work in the past six months was reported in June versus May (2.4% vs. 3.3%, p = 0.049). Also, a significant increase in the proportion of consultations where group sex in the past six months was reported in August versus May (43% vs. 38%, p < 0.001) (see Supplementary Files). No trend was seen in the other indicators.
Discussion
After the first detection of mpox outside of endemic regions in May 2022, a large global outbreak occurred among GBMSM with a peak in incidence in the summer of 2022. After the peak, the incidence dropped rapidly, both in the Netherlands and in most of the other non-endemic affected areas.8,18 In addition to a targeted risk communication campaign, the Netherlands rapidly introduced control measures including PEP for contacts and a targeted PPV programme to protect those at highest risk of mpox. This resulted in an overall vaccination coverage of 35.4% in the target group by April 30, 2023, with differences between the different indication groups. In addition, a VE of full vaccination against laboratory confirmed symptomatic mpox disease of 68.2% was found, which is comparable to estimates from other studies. For one MVA-BN dose, VE estimates range from 36%-86%,19–23 and for two doses from 66% to 89%.21–23 Persons who received a childhood smallpox vaccination were also included in the referenced studies. Those persons would have been considered fully vaccinated after one MVA-BN vaccination in our definition. Therefore, it was to be expected that our estimated VE is a bit lower than the estimate for two doses MVA-BN.
The impact of the vaccination programme on the course of the epidemic is difficult to establish. Generally, a good immune response is seen two weeks after the first dose of MVA-BN, with seroconversion seen in 91% of subjects in a study by Pittman et al.24 However, antibody titres increased significantly after a second vaccination with MVA-BN, given four weeks after the first vaccination. When also taking an incubation period of 8 to 9 days into account,25,26 it may take several days to weeks before a significant effect of the vaccination programme on infections could be noticed. In addition, the timing and speed with which the vaccination programme is implemented also influences its impact. In the Netherlands, the PPV programme started after the peak of the outbreak and by the time PPV would be effective, the incidence was already down to about 5 cases per day, making it unlikely that PPV contributed significantly to the decline of the outbreak.
Unfortunately, the effect of PEP given to high-risk contacts of mpox cases cannot be assessed, as national data are lacking. PEP is unlikely to have contributed substantially given that the incubation period of mpox is about equal to, or possibly even shorter than the time to generate vaccine-induced immunity.24,25 Furthermore, as it takes time to identify high-risk contacts of a mpox case, invite them for vaccination and administer the vaccine, the impact of PEP to prevent mpox after exposure is likely to be negligible. However, PEP vaccinations may have contributed to the overall effect of mpox vaccination, as they were administered to the persons at highest risk for mpox infection.
Correct calculation of the uptake, vaccination coverage and assessing VE was challenging due to limitations regarding the quality of data and lack of data on denominators. Identifying and personally inviting the population at risk through multiple health care providers was challenging. In addition, some people used an alias, and privacy guidelines prohibited direct comparison of persons possibly invited multiple times. In an online survey conducted from July 1st to mid-August 2023, 16.3% of the GBMSM invited for vaccination reported to have received more than one invitation for vaccination (Adam P, unpublished). Also, exact numbers of invitations sent out by GPs and HIV health care providers were not known. Finally, it is known that several HIV health care providers invited all their patients instead of only those at highest risk of mpox, which could explain the lower PPV uptake in this group. Overall, the denominator is likely overestimated, but it is unknown to what extent. On the other hand, the number of administered vaccinations might be underestimated, since not all data were available due to privacy constraints. Altogether, we believe we may have underestimated the uptake, vaccination coverage and VE. Thus, our results emphasise the importance of a vaccination register with national coverage and adequate privacy assurance, to facilitate adequate monitoring and evaluation of future vaccination programmes in outbreak settings.
Our observations regarding sexual activity differ from reports from external sources. A behavioural survey conducted in August 2022 among 2,460 GBMSM in the Netherlands (Adam P, unpublished), indicated that half (50.4%) of the respondents had reduced their number of partners since the start of the outbreak, and two-thirds (65.5%) had avoided sex on premises venues. Similar behavioural adaptations to mpox were observed among GBMSM in the US as well.27 However, when examining our data on SHC visits, we could not identify comparable changes in behaviour, nor in positivity rates of STI. Most behavioural indicator variables remained relatively stable in 2022. The changes observed simply entailed a minor reduction in reports of sex work in the past six months in June compared to May, and a slight increase in reports of group sex in the past six months in August versus May. Although these findings align with the hypothesis of decreased sexual activity during the outbreak followed by a reengagement in sexual activity as the outbreak waned, the extent of these changes observed in our data was extremely limited.
The fact that, at each point in time, our behavioural indicator variables consisted of reports for the past six months (rather than the last month, for instance) could have hindered the detection of more significant changes in behaviour, particularly if GBMSM’s behavioural adaptations primarily occurred during the brief period when high numbers of mpox cases were reported. Also, HIV-PrEP users who regularly visit SHC on which the analyses were based may not be representative of the entire population of GBMSM at risk for mpox.
Based on the above findings, a likely explanation for the declining course of the outbreak may be the impact of acquired immunity, either by natural infection or PEP, in a substantial proportion of those most at risk. This is also in line with results from several modelling studies that conclude that immunity due to infection in the most sexually active GBMSM may be the main driver of the decline.28–31 Modelling results using data on Dutch mpox cases before the start of the PPV programme indicate that increased immunity due to infections may have limited further growth of the mpox epidemic, but could not exclusively explain the fast decline after the peak. This suggests that behavioural adaptations among GBMSM may have accelerated the decrease observed before the start of PPV.30 It may also explain the observed vaccine uptake, which was notably lower than expected based on previous studies investigating vaccine acceptance among GBMSM in the Netherlands. They found that 70-85% of GBMSM in July-September 2022 were willing to accept mpox vaccination.32,33 The perceived sense of urgency may have been low at the time people were invited for PPV, especially for the second dose, as the outbreak was already declining. This may have led to a lower vaccine uptake and vaccination coverage than expected.
Although we hypothesise that it is unlikely that PPV was the main driver of decline of the mpox outbreak, vaccination might play an important role to prevent future outbreaks, especially since the VE is expected to be relatively high. Fast implementation of future vaccination programmes is essential to be effective in preventing new cases.18 Immunity due to infection and vaccination among the most sexually active GBMSM reduces the risk of resurgence of mpox28–30 and will protect against (severe) disease. The vaccination programme in the Netherlands continued in 2023 and took newly identified risk factors, such as visiting sex venues, into account.34,35 Continued vaccination programmes are needed to offer vaccination to new people meeting the criteria for vaccination or recently arrived migrant GBMSM. Surveillance of mpox and evaluation of the vaccination programmes, including assessment of VE, are of critical importance to inform future mpox prevention and control.
Data Availability
Deidentified individual participant data included in this study will not be made available to others because of the sensitive nature of participant data extracted from participating public health services.
Contributions
SJMH and MRH conceptualized the study. SJMH, MRH, ELMOdC, IJMW, CEvE designed the study. MRH, IJMW, CEvE, SJL, PJJ, CvB, GGCvR, EH collected the data. MRH, IJMW, CEvE did the statistical analyses. MRH and IJMW wrote the first draft of the manuscript. All authors contributed to writing the manuscript, and all approved the final version.
Declaration of interest
We declare no competing interests.
Data sharing
Deidentified individual participant data included in this study will not be made available to others because of the sensitive nature of participant data extracted from participating public health services.
Disclaimer
CvE is a fellow of the ECDC Fellowship Programme, supported financially by the European Centre for Disease Prevention and Control The views and opinions expressed herein do not state or reflect those of ECDC. ECDC is not responsible for the data and information collation and analysis and cannot be held liable for conclusions or opinions drawn.
Acknowledgments
We thank Maartje Visser from the RIVM for selecting the persons belonging to the indication groups from the national database. We also thank Birgit van Benthem, Susan van den Hof and Maria Xiridou from the RIVM for their critical view and suggestions for improvement of the manuscript. Furthermore, we thank all the PHS and SHC in the Netherlands for their commitment during the mpox outbreak and PPV programme.
This work was supported by the Dutch Ministry of Health, Welfare and Sport.
References
