Pandemic potential of poxviruses: From an ancient killer causing smallpox to the surge of monkeypox

Abstract

Smallpox caused by the variola virus (VARV) was one of the greatest infectious killers of mankind. Historical records trace back smallpox for at least a millennium while phylogenetic analysis dated the ancestor of VARV circulating in the 20th century into the 19th century. The discrepancy was solved by the detection of distinct VARV sequences first in 17th‐century mummies and then in human skeletons dated to the 7th century. The historical records noted marked variability in VARV virulence which scientists tentatively associated with gene losses occurring when broad‐host poxviruses narrow their host range to a single host. VARV split from camel and gerbil poxviruses and had no animal reservoir, a prerequisite for its eradication led by WHO. The search for residual pockets of VARV led to the discovery of the monkeypox virus (MPXV); followed by the detection of endemic smallpox‐like monkeypox (mpox) disease in Africa. Mpox is caused by less virulent clade 2 MPXV in West Africa and more virulent clade 1 MPXV in Central Africa. Exported clade 2 mpox cases associated with the pet animal trade were observed in 2003 in the USA. In 2022 a world‐wide mpox epidemic infecting more than 80,000 people was noted, peaking in August 2022 although waning rapidly. The cases displayed particular epidemiological characteristics affecting nearly exclusively young men having sex with men (MSM). In contrast, mpox in Africa mostly affects children by non‐sexual transmission routes possibly from uncharacterized animal reservoirs. While African children show a classical smallpox picture, MSM mpox cases show few mostly anogenital lesions, low‐hospitalization rates and 140 fatal cases worldwide. MPXV strains from North America and Europe are closely related, derived from clade 2 African MPXV. Distinct transmission mechanisms are more likely causes for the epidemiological and clinical differences between endemic African cases and the 2022 epidemic cases than viral traits.

INTRODUCTION

At first glance, it is not clear why one should review the origin and development of an infectious disease for which the causative agent was eradicated nearly 50 years ago, as was the case for variola virus (VARV), the causative agent of smallpox, probably the greatest killer of human history (Hopkins, 1983). One might argue that such a disease is only of interest to historians of medicine. However, the more we understand about the origin and development of past pandemics, the better are we prepared to anticipate and respond to future epidemics (Wertheim, 2017). How close we came to a new pandemic with poxviruses, was vividly illustrated with the recent outbreak of monkeypox virus (MPXV), where knowledge about VARV was of critical importance.

The monkeypox (mpox) pandemic has increased the interest in viral infections occurring at the animal–human interface. All the great and deadly pandemics of the last 100 years have an at least strongly suspected zoonotic origin, for example, the Spanish flu, AIDS or Covid‐19. Into this list, one should also add smallpox. Smallpox is frequently forgotten in this compilation of zoonotic viral diseases because it was eradicated in a spectacular worldwide initiative guided by the World Health Organization (WHO). In one of the triumphs of modern medicine, the world was declared smallpox free in 1980. This success was achieved because smallpox was an exclusively human disease without an animal reservoir. However, the genome analysis of VARV leaves no doubt that also smallpox despite its antiquity as a human disease has a zoonotic origin. In fact, VARV's closest relatives are camelpox virus (CMLV), which causes highly virulent infections in camels, and taterapox virus (TATV). The latter infects gerbils (Tatera kempi) hence, the name of the virus, without however causing a disease in this species (Parker et al., 2017). With this distinct pathogenicity of these two poxviruses in their respective hosts, gerbils might be a reservoir species for a smallpox ancestor while camels might represent as humans a secondary host species. This argument rest on the assumption that viruses coevolve with their hosts and attenuate their pathogenicity after long‐evolutionary periods of coexistence. Despite the fact that VARV is now extinct (save for two stocks maintained in the US and Russia), there are good reasons to be interested in the origin and evolution of smallpox virus. First, smallpox was one of the most notorious killers in human history. In the 20th century, an estimated 400 million people were killed by smallpox despite availability of an effective vaccine. A historian of medicine called it “the most terrible of all the ministers of death” (Hopkins, 1983). These numbers dwarf even the death toll of the Spanish flu estimated to be 50 million victims and of Covid‐19 which might have killed up to 10–20 million people in one pandemic. In contrast to the Spanish flu pandemic, smallpox was endemic through several centuries on many continents causing continued misery throughout historical times. Second, poxviruses are widely distributed in wild and domesticated animals creating a large pool of viruses for potential zoonotic infections. Third, while smallpox is and remains extinct, the recent spread of MPXV infections in the human population underlines that Orthopoxviridae, the group to which VARV and MPXV belong, still need public health attention. Fourth, the changing pathogenicity of smallpox and the changing transmission pattern of mpox infection underlines the dynamic character of poxvirus infections.

THE HISTORIAN'S PERSPECTIVE ON SMALLPOX EPIDEMICS

How old is VARV and the smallpox disease? Curiously, different answers are given if VARV or the smallpox disease, genetical or historical arguments are considered and only recent archaeogenetics data start to reconcile both datasets. Let us begin with the evidence that historians compiled on smallpox (Fenner et al., 1988).

Start in ancient Egypt?

The earliest material evidence for smallpox as a disease is the mummy of pharaoh Ramses V who died 1157 before common era (BCE). The visible parts of the skin from this Egyptian king and a few other mummies from this time are covered with numerous homogeneous pustules suggestive of smallpox (Reardon, 2022). The archaeologists could not check whether pustules are also found on palms and soles which would differentiate smallpox from chickenpox (a herpesvirus infection). VARV does not cause a persistent infection, it ends in either recovery and elimination of the virus and long‐term immunity to smallpox or death. Maintenance of VARV in the human population, therefore, depends on the regular influx of new susceptible subjects into a sufficiently large population to entertain continuous infection chains. Influx can be either by birth or immigration. Only Egypt, Mesopotamia, the Indus valley or China as early centers of state building offer the population size for continuous circulation of VARV and are therefore logical places for the origin of smallpox. However, medical Egyptian papyri, the Bible or the Greek physician Hippocrates do not mention an infectious disease compatible with smallpox, which is a strong argument against smallpox—at least as we know it—in Egypt or Antiquity as a whole. However, there are further independent arguments for smallpox appearing in ancient Egypt. The close genomic relatedness of VARV with CMLV and TATV speak for human contact with these two animal hosts both living in an arid climate zone. Archaeologists have dated the use of camels as pack animals in the Iron Age, not occurring before the 12th century BCE, in southern Arabia or the Horn of Africa (Sapir‐Hen & Ben‐Yosef, 2014). Biblical reports (Queen of Sheba from Yemen visiting King Salomon with a camel caravan in the 10th century BCE) and the somewhat earlier use of camels for copper transport from Egyptian mines by pharaoh Shoshenq, founder of the 20th dynasty (to which Ramses V belongs), make it plausible that around this time period, the population of the Nile valley made the first contact with camels and their viruses. An origin of smallpox in humans at that time is thus not implausible.

Further worldwide spread and written reports

Long‐distance trade by camels from Egypt and Arabia to the Levant, Mesopotamia, Persia and India could have spread smallpox further into South Asia. Buddhist missionaries might then have carried VARV to China and Japan. The first unmistakable medical descriptions of smallpox and its clinical symptoms dates from the physician Ko Hung in China in 340 common era (CE, equivalent to AD) and from writers in India and the Mediterranean at around 600 CE and most clearly from the Persian physician Rhazes writing at about 910 CE. Rhazes distinguished smallpox from measles, describing a seasonal incidence of smallpox and that it affected primarily children (Behbehani, 1983; Fenner et al., 1988). Smallpox became subsequently endemic in Asia. An exceptional temple record from a district on Honshu Island/Japan maintained between 1771 and 1851 showed that smallpox occurred there as an endemic disease with regular‐spaced epidemics peaking every 5 years when smallpox caused the death of 2% of adults and up to 40% of children younger than 5 years. A similar cyclic epidemic pattern was described for British India between 1870 and 1940 with smallpox mortality peaks about every 5 years causing 0.1% of all deaths. The smallpox mortality in India decreased over time induced by an increased vaccination rate (Fenner et al., 1988).

The spread of smallpox into Europe has been variously associated with the invasion of the Huns into central Europe, or the Islamic expansion into Spain or the move of crusaders in the 11th and 12th century into the Levant. Whatever the introductory event, smallpox became endemic in Europe in the 15th century and was reported not to represent a major plague (Fenner et al., 1988). This changed when Spaniards introduced viral infections into the New World where the virus encountered people who were not exposed to it previously. Historical records noted that the populations of the densely populated Aztec and Inca states suffered between 1518 and 1520 epidemics of apocalyptic dimensions. Thirty years after the arrival an estimated 30 million Americans had died from smallpox or measles (difficult to distinguish in historical records of this period). In 1563, a large smallpox epidemic was noted in Portuguese Brazil. The loss of the indigenous population led to the slave trade from west Africa with the concomitant introduction of smallpox by affected slaves. Starting in 1630, smallpox was reported to occur in several epidemics in Canada around the Great Lakes which nearly annihilated several Indian tribes (Rodríguez‐Frías et al., 2021).

Virulence changes?

High lethality is generally reported when a “new” pathogen enters a virgin population which has previously not encountered this pathogen and thus lacks immunity and genetic adaptation to the pathogen. The encounter of people from the Old and New World led to an exchange of infectious diseases: smallpox (and measles) from Europe to America and syphilis from America to Europe. Notably, in the European perspective smallpox was described as less severe than syphilis leading in French to the distinction of “la grosse vérole” for syphilis and “la petite vérole” for smallpox (hence, the English name), which might also refer to the smaller size of the eruptions and the lower age of the affected subjects in smallpox compared to syphilis.

However, smallpox was notorious for its changing virulence. Starting with the Thirty Years War smallpox became the most dreaded infectious disease in Europe outcompeting plague, leprosy and syphilis. The 17th‐century physician Sydenham described higher fatality rates for patients with confluent in contrast to distinct eruptions of smallpox. For London, we have detailed death data starting with a wave in the 1630s followed by a long epidemic between 1650 and 1680, then a practically uninterrupted epidemic period throughout the 18th century where smallpox represented nearly 10% of all deaths. After 1800 one sees a gradual decline of the smallpox death toll to small numbers mainly due to the widespread use of Jenner's vaccine. A very mild form of smallpox initially known as kaffir pox was described shortly before 1900 in South Africa. At the same time, a disease described as variola minor was spreading from Brazil and the southern US (Fenner et al., 1988). Smallpox persisted until it was eradicated by a worldwide vaccination and tracing effort. The campaign consisted not only of mass vaccination but also of case identification and vaccination of contact persons of cases (“ring vaccination”) until the last case was reported in 1977 in Somalia. WHO declared smallpox officially as extinct in 1980. Between 1959 and 1974 WHO reported yearly still between 30,000 and 200,000 cases of smallpox with the majority coming from Asia, followed by Africa and the Americas and a few cases from Europe and Oceania (Benenson, 1982).

THE EPIDEMIOLOGIST'S VIEW ON SMALLPOX

Various clinical forms

Smallpox is a systemic, febrile rash disease. Clinicians and epidemiologists distinguished four main clinical types of variola major smallpox manifestations (Damon, 2005). The most common form, occurring in about 90% of cases, produces viremia, fever, prostration and rash and is called ordinary smallpox. Fever sets in after a 10–14 days incubation period associated with backache, headache and vomiting. Skin lesions show a sequence of rash forms from macules (flat discoloration), to papules (elevated solid lesion), to vesicles (fluid filled blister) which are often umbilicated (central depression) to pustules (pus‐filled vesicle) and end with crusting and scab formation by day 14, which then slough off leaving pock scars. Rash started on the face and forearms and all lesions were at the same stage of development. The mucosa of mouth and throat were also affected leading to difficulty in swallowing. Mortality was proportional to the extent of rash being 10% for ordinary discrete smallpox to 50% and more in ordinary confluent smallpox. Surveys in Madras/India during the 1960s (much epidemiological research was done in South Asia during the final phase of smallpox before eradication) reported a 32% case fatality rate (CFR) for ordinary smallpox in unvaccinated people. Modified smallpox is the infection occurring in vaccinated subjects: in Madras case fatality was reduced to 3% in people with primary vaccination and 0% in subjects with two vaccinations (Benenson, 1982). In flat smallpox, the skin eruptions do not project above the surrounding skin level, which occurred in Madras in 7% of cases and showed a stunning 94% CFR (67% in vaccinated subjects). Hemorrhagic smallpox is the most dreaded form: in Madras it occurred in 3% of cases with a mortality of 98% which was not reduced by prior vaccination. Patients showed bleeding from mouth, nose and vagina and into the urine. The death occurred sometimes even before a flat focal rash appeared. Some data indicate that the different forms of variola major smallpox reflect different patterns of host response and not different viral strains. For example, in India, haemorrhagic smallpox was observed with increased frequency in pregnant women (Rao, 1964). In addition, for ordinary smallpox, CFR was 28% in pregnant compared to 8% in non‐pregnant women from Madras (Rao et al., 1963). A different situation was seen in Africa where apparently distinct viral strains, variola minor (alastrim) and intermediate strains produced symptoms of smallpox characterized by discrete lesions and rapid recovery. CFRs for these smallpox forms were 6% in Tanganyika (Bedson et al., 1963) and 2% in Ethiopia (Benenson, 1982).

Epidemiological characteristics

Smallpox distribution was worldwide and no racial difference in susceptibility was seen. However, mortality was lower in populations where smallpox has been endemic for several generations. In Bangladesh and Pakistan, smallpox incidence rates were higher in males than in females, but that difference was explained by social restrictions on the movement of Moslem women. In South Asia, infections were more common among members of the lower socioeconomic classes which was explained by household overcrowding and migration to find an occupation by unskilled workers. The number of smallpox cases in Madras was highest among children younger than 4 years and case numbers dropped markedly with increasing age. In Bangladesh, smallpox showed a marked seasonality with a peak in the spring, while in Pakistan a winter peak was observed.

In Madras and Bangladesh about 40% of unvaccinated family contacts developed the disease. The attack rate among unvaccinated family members was 8% when the index case had previously been vaccinated. Virus is present in high concentrations in the deeper layers of the skin but transmission occurred predominantly during the first week of illness from saliva charged with virus produced in oral rashes. Transmission occurred by direct person‐to‐person contact via droplet spread. Bedding became contaminated with the virus but fomites represent a minor path of transmission in laundry workers and chambermaids. The virus survived in shed crusts from patients for over 6 months. Insect vectors (houseflies, mosquitos) are not an important way of disease transmission. There is no animal reservoir for smallpox. No persistent smallpox infections were described in humans, transmission was thus only from acutely infected subjects. Subclinical infections were considered to be rare, but studies from Madras showed subclinical infections in a third of family contacts of index cases. Subclinical contacts excreted small amounts of virus but disease transmission from them seemed not to occur (Benenson, 1982).

THE GENETICIST'S VIEW ON SMALLPOX VIRUS

20th century viral isolates

VARV contains a linear double‐stranded DNA genome of about 186 kilobase pairs with covalently closed ends adjoining small inverted terminal repeat regions. The central genome region of poxviruses mainly specify conserved proteins, essential for virus replication, while the terminal regions encode more divergent proteins, associated with host range and virulence determination. Esposito et al. (2006) sequenced 43 smallpox virus isolates mostly collected during the WHO Intensified Smallpox Eradication Program starting in 1967, representing isolates from different geographical areas, including smallpox viruses associated with minor (<1%) (including alastrim isolates) or major (>10%) CFRs. The unrooted tree split into three clades A, B and C. The tree split according to geographical origin: West Africa (clade A), South America (clade B) and Asia (clade C). Clade C also contained a subcluster of lower virulence isolates from Africa (but outside of West Africa). The central genome regions of the sequenced smallpox strains were related by direct descent while the terminal regions varied substantially and showed evidence for recombination.

Li et al. (2007) sequenced a collection of 47 VARV isolates collected in the 30 years before the eradication of smallpox in 1977. VARV displayed a slowly evolving DNA genome. Isolates from Bangladesh collected over 2 years showed nearly identical sequences. The authors used concatenated single nucleotide polymorphism (cSNP) matrices for VARV phylogenetic analysis. The phylogenetic tree split into two clades. Clade I comprised Asian VARV major (20%–30% CFR) and East, Central and South African VARV isolates which also contained low‐virulence isolates (1% CFR). Clade I showed a clear geographical division with branches composed by isolates from Bangladesh, Middle East and Far East. The Non‐West African minor isolates were diagnosed as a variant from Asian variola major isolates. Clade II comprises two branches: South American isolates known as alastrim minor types and West African isolates which show an intermediate virulence (10% CFR) between variola major and alastrim minor. To their surprise, the authors calculated the time to the most recent common ancestor of the two VARV clades to about 220 years before the present while the radiation in clade I occurred in the last 90 years. Obviously, these genetic data do not concur with historical data which suggested a relationship of South American isolates to slave trade from West Africa and which tentatively linked East African isolates with Muslim trade across the Indian Ocean. The authors tried to fit both datasets but the genetic conclusion is clear: the VARV strains circulating in the mid‐20th century are recently evolved viruses.

Back to early modern times mummies

It may seem unreasonable to conclude, against all historical evidence, that smallpox is a recent disease. However, some insights can be obtained again from mummies. In 2004, Russian and French researchers identified well‐conserved mummies buried in the permafrost of Yakutia in Siberia, which were dated to the late 17th to early 18th centuries. Histological investigation showed iron inclusion in the lung (Thèves et al., 2014) pointing together with positive PCR results for short fragments of VARV DNA to hemorrhagic smallpox as cause of death. Sequencing of the short DNA fragments indicated a virus ancestral to both clades I and II isolates described by Li et al. (2007). Based on the new tree the origin of the smallpox virus was dated to 930 CE (Biagini et al., 2012).

More insight into the evolution of smallpox viruses was provided by another mummy, this time a young child buried in the crypt of a Lithuanian church, dated to about 1650 CE. A DNA library enriched for an unrelated virus‐yielded poxvirus reads that allowed an 18‐fold coverage of a VARV genome (Duggan et al., 2016). This genome was upon phylogenetic analysis ancestral to all previously sequenced VARV genomes. Further analysis allowed to date the ancestor of clade I and clade II VARV from the 20th century to the second half of the 18th century, when Jennerian vaccination was introduced, and smallpox changed from a disease of adults to one of children and when the slave trade at its height caused a wide spread of African viruses. At this time, clades I and II VARV went through a major population bottleneck causing—as the authors suspected—the extinction of older VARV lineages. These researchers dated the ancestor of the until then sequenced VARVs at the earliest to 1530, the time of European expansion and colonization, much later than historical records of smallpox. They mentioned that one explanation might be that it is difficult to distinguish smallpox from measles in historical records. Others even stated that the finding challenges the presumed history of smallpox (Wertheim, 2017). When analysing the genetic distances of VARV isolates against the year of sampling clear evidence for a clock‐like evolution of VARV genomes was obtained instilling confidence into the recent origin of historical VARV isolates. This point was confirmed by sequencing a museum specimen of a British childhood smallpox victim dated to 1766 which formed a sister group with the Lithuanian isolate to all modern VARV isolates (Ferrari et al., 2020). A skin piece covered with pock lesions from a Czech museum dated to between 1810 to 1890 (when a devastating smallpox epidemic with 23,000 victims in the unvaccinated French as opposed to the vaccinated Prussian troops contributed to the Prussian victory in the 1870/1 war) yielded a VARV genome that was at the basis of the 20th century PII viral clade (Pajer et al., 2017).

Back to skeletons from early medieval times

In a thorough study from ancient DNA from skeletal and dental remains of 1867 humans living in Eurasia and the Americas between 31,000 and 150 years ago, the obtained sequences were screened for matches to VARV (Mühlemann et al., 2020). Twenty‐six samples matched VARV sequences, for 13 samples virus sequences could be enriched and 11 were from individuals who died between 603 CE and 1050 CE in northern Europe, western Russia, and the UK. The remaining two were from 19th‐century samples in Russia. Four samples yielded a higher coverage VARV genome that revealed the existence of a previously unknown, now‐extinct virus clade circulating in the Viking Age. The most recent common ancestor of these ancient VARVs and the modern VARVs was dated to 1700 years ago, that is into the 4th century CE. Interestingly, bishop Marius of Avenches/Switzerland, an early medieval historiographer living in the 6th century, used for the first time the Latin term variola for a disease which might have been smallpox (Fenner et al., 1988). Notably, no human samples from Eurasia and America predating the Viking era yielded VARV sequences. Based on the new sequences, the ancestor for the 20th century, VARVs circulated only 300 years ago. The consortium studied another important aspect of smallpox virus evolution. Orthopoxvirus species with a narrow host range such as VARV have fewer (n = 162) genes than those with broad host ranges such as cowpox virus (CPXV) species which have 209 functional genes. A complex gene deletion pattern was revealed: genome comparisons showed that five genes were already inactivated in both modern (mVARV) and ancient (aVARV) viral genomes; 14 genes were inactive in mVARV but not yet in aVARV; seven genes were inactive in aVARV but active in mVARV. VARV followed thus at least two evolutionary trajectories of gene deletion, one leading to aVARV and another leading to mVARV. The concept that gene deletion increases virulence is at first glance counterintuitive. CPXV infects a broad range of hosts, including humans, but causes limited pathology. In the Viking Age, only 2% of the skeletons showed evidence of smallpox infection while in the 18th century Britain 10%–20% of deaths were caused by smallpox. Various patterns of deletion might have changed the virulence of VARV as seen several times throughout the history of VARV—human interaction with both virulence increases (e.g., in the 17th and 18th centuries in Europe) or in the development of less‐virulent variola minor strains in the late 19th century. During the Viking Age smallpox might have been a mild disease (Alcamí, 2020) possibly explaining the lack of historical records on smallpox during that time period.

Calculating the split

The split of the ancestor of VARV from the animal poxviruses TATV/CMLV lineage was calculated in a recent report to 7700 years before present, but that does not imply that VARV entered human populations at that time. By including animal orthopoxvirus genomes in their calculations, these scientists (Forni et al., 2023) estimated the time for the ancestor of VARV to 2000 BCE; they calculated the ancestor of mVARV to 1690 and the split of clades I and II to 1910 and 1850, respectively. A researcher from the Russian smallpox virus depository calculated younger data and noted a divergence time of 3400 years ago for VARV from the TATV/CMLV lineage, and 7000 years ago for the divergence of the MPXV lineage from that leading to CMLV/TATV/ VARV (Shchelkunov, 2009). These two long chronologies would include the possibility that pharaoh Ramses V died of smallpox as suggested by the visible pox marks on his mummy. Therefore, smallpox might indeed have affected the human population in these early civilizations. The data from Mühlemann et al. (2020) give what one could call a short chronology of smallpox in human populations starting in 300 CE. Such a late entrance of smallpox into human populations would explain why no medical writer of Antiquity described the disease and why no older skeletons from their collection showed evidence of VARV DNA traces.

CONTINUING CONCERNS

By a heroic effort of WHO, smallpox was eradicated from the globe (Fenner et al., 1988). Vaccination against smallpox has since been abandoned since the probability of negative vaccine effects by vaccinia virus—including even deaths—is now much greater than the risk of dying from natural smallpox. However, there is a persisting fear that VARV could be weaponized for military or terrorist purposes. Since the impact of a spread of VARV would be dramatic in a now largely non‐immune population such concern is justified. In addition, there has also been a concern that viable smallpox virus would be liberated from human burials in permafrost soil with climate change. So, far no viable VARV virus could be isolated from such material (McCollum et al., 2014) and the archaeological work with the Siberian smallpox mummies showed that even VARV DNA in such samples was largely degraded since long‐distance (2 kb) PCR did not yield amplifiable VARV DNA (Biagini et al., 2012). While these data are reassuring with respect to an unlikely resurgence of smallpox as VARV (except for military and bioterrorist use from the available two stocks or resynthesis from the public available VARV genome sequence), the follow‐up of the smallpox eradication program has shown that poxviruses exist in animal reservoirs that can cause smallpox‐like disease in humans with CFRs of up 10% and that related poxviruses exist that have under particular epidemiological conditions even the potential to cause a pandemic, fortunately with a CFR of only 0.2%. These allusions are to MPXV. In view of the genetic, epidemiological and clinical dynamics shown by VARV, a close attention on MPXV is clearly warranted justifying a review of genetic, epidemiological and clinical data on monkeypox (mpox) and the monkeypox virus (MPXV).

MONKEYPOX VIRUS DATA FROM AFRICA

Veterinary poxvirus infections

For the success of the smallpox eradication plan, it was important to assure that VARV has no reservoirs in animals. Only three limited zootics with smallpox‐like skin lesions were noted during the 20th century, one occurred in orang‐outans from the Djakarta zoo, one in free‐living rhesus monkeys in Bengal and one in Cebus monkeys from the Brazilian forest. In addition, WHO started in 1967 a survey for smallpox‐like diseases in monkeys kept in 26 research organizations handling large numbers of monkeys. Several outbreaks with a laboratory proven poxvirus disease were documented, including one at the Statens Serum Institute in Copenhagen where MPXV was isolated in 1959. Another epidemic occurred in the Rotterdam zoo among anteaters, which spread to the monkey house, affecting 10 orang‐outans, 6 of whom died. Finally, a dozen of 2000 monkeys held in a US pharma company showed smallpox‐like disease (Arita & Henderson, 1968). So far as could be discerned at the time, none of these smallpox‐like diseases was caused by VARV.

Human poxvirus infections after smallpox eradication

The next task of WHO was to assure that VARV did not circulate any longer in the human population after the worldwide vaccination effort. Therefore, the smallpox eradication unit of WHO followed carefully human cases resembling smallpox in countries where smallpox had vanished. In 1970, WHO described a first case of a smallpox‐like disease in a 9‐month‐old boy from Zaire from whom MPXV was isolated. The family did not recall a contact with smallpox cases, but monkeys were occasionally eaten as delicacy by this family (Ladnyj et al., 1972). In the following 2 years, the US Centers for Disease Control and Prevention (CDC) supported by local scientists identified six sporadic human cases of mpox in unvaccinated subjects presenting with mild disease from Liberia, Nigeria and Sierra Leone. Susceptible household contacts were not infected. Most cases lived in the tropical rain forest areas where hunting bushmeat, including monkeys, was common (Foster et al., 1972). Subsequently, WHO described 47 mpox cases during the 1970s mostly from Zaire, but a few cases were from West Africa. All cases lived in villages from tropical forest areas where hunting of wild animals for food was common. Geographical cluster, but also family clusters were observed suggesting a secondary attack rate of 3%–10%, much smaller than 25%–40% seen for smallpox in Zaire. Most cases were in children <10 years (y), both sexes were equally affected. Clinical symptoms closely resembled smallpox, CFR was 17% in the acute phase.

Searching poxviruses

Starting in 1975 WHO conducted a facial pockmark survey in 6.5 million children from Central and West Africa. None showed characteristic marks. In addition, 1600 clinical specimen were collected from patients with fever and eruptive disease identifying two additional likely mpox cases. The WHO scientists reported absence of MPXV neutralizing antibodies in 2000 animals sampled in Western and Central Africa. However, neutralizing antibodies were found with up to 20% prevalence in rodents, larger mammals and birds in regions where also mpox cases were reported (Breman et al., 1980). Four “whitepox” virus strains have been identified in organs of animals captured in the wild. Whitepox virus are isolates that could not be distinguished from VARV phenotypically, but showed upon restriction analysis deletions and recombination in the terminal viral genome regions (Dumbell & Archard, 1980).

Endemic monkeypox in Central Africa

From a survey in Zaire (now Democratic Republic of Congo, DRC) conducted between 1980 and 1985, 282 patients with mpox were identified as verified by virology or serology diagnostics. Most of the cases were detected in children younger than 10 years with no difference between the sexes; practically none had a smallpox vaccination. Again, the disease resembled ordinary or modified smallpox. In the pre‐eruptive phase fever, headache and backache was noted. The eruptions resembled those of smallpox. A quarter of the children had lesions on the genitalia. Distinct from ordinary smallpox, an enlargement of the lymph nodes was observed in 84% of mpox cases. CFR was 15% in children <4 years, 7% in children <10 years; no children >10 years died (Jezek et al., 1987).

When 12 villages in DRC were investigated, 88 cases of mpox disease were detected between 1996 and 1997, suggesting an attack rate of 2% and a CFR of 4%. Peak incidence was in August. Household secondary attack rate was 13%. The cases showed virtually identical viral genome sequences, distinct from that of other African countries, but closely related to a virus isolated 10 years ago from a squirrel in this region. From captured wild animals, 42% of squirrels and 16% of Gambian rats showed orthopoxvirus‐neutralizing serum antibodies (Hutin et al., 2001). From 383 animals caught close to villages with a human mpox case, only one squirrel showed skin eruptions and yielded MPXV and displayed high‐serum antibody titers to MPXV (Khodakevich et al., 1986).

In DRC, smallpox vaccination campaigns were discontinued in 1980. Public health scientists resumed a systematic survey from 2005 to 2007 in areas that were already actively surveyed in the 1980s. They observed 760 cases of laboratory‐confirmed mpox, which represents a 20‐fold incidence increase over the situation in 1980s, which they attributed to the abandonment of smallpox vaccination. Unvaccinated subjects had a fivefold higher risk of mpox suggesting a 85% protection rate against mpox. Subjects in their 20s showed nearly a comparable incidence rate as children, while subjects older than 30 years were practically spared, 62% of cases were in males. The mpox incidence was 5‐fold higher in forested than in savannah regions (Rimoin et al., 2010). Animals were trapped in DRC around villages reporting mpox cases: from 105 caught animals, representing mainly rodents, two Cricetomys giant pouched rats were tested positive for orthopoxvirus‐specific IgG by Elisa (Doshi et al., 2019).

The 2017/2018 epidemic in Nigeria

Researchers detected 276 suspected and 118 laboratory‐confirmed cases of mpox in Nigeria with a peak incidence in August 2017. The median age of the cases was 29 years, with a clear male bias in all age groups older than 10 years. The majority of cases occurred in three states with rainforest or swamps or mangrove vegetation. Clustering of cases was noted in households and in a prison facility. Clinical symptoms resembled those seen in cases from central Africa. A CFR of 6% was observed, the majority of them had HIV coinfection. Viral genome sequences analysed for seven cases displayed 42 bp differences from an MPXV isolated 1971 in Nigeria. MPXVs from two cases were identical, another pair differed by 2 bp and three cases differed by 11 mutations. Only 10% of the cases reported suspicious animal contacts. The authors suspected for this epidemic separate viral introductions from an animal reservoir, followed by limited human‐to‐human transmission. Three cases were exported to UK and Israel by travellers returning from Nigeria (Yinka‐Ogunleye et al., 2019). An analysis of hospital records from Nigeria in 2017/2018 revealed 40 patients hospitalized with mpox rashes. A quarter of them was coinfected with HIV‐1 which prolonged mpox illness. Overall case fatality for mpox was 13% (Ogoina et al., 2020).

MONKEYPOX VIRUS OUTSIDE OF AFRICA

In 2003, a cluster of 11 human mpox cases was reported in Wisconsin/USA. The cases were all linked to prairie dogs sold as pet animals. A distributor 1 from Illinois received pet animals (including rodents) imported from Africa. A further distributor 2 purchased prairie dogs and transported them together with an ill giant Gambian rat (Cricetomys species). Distributor 2 and his wife contracted mpox disease as well as a family who bought a prairie dog from him, as well as employees from two pet shops selling these prairie dogs, two veterinarians and two further members of a household purchasing a prairie dog. All human cases were observed between mid‐May to early June 2003. Skin lesions occurred around bite sites by infected prairie dogs and showed some dissemination. Fever, headache, lymphadenopathy were prominent signs, four cases were hospitalized, none died. Half of the cases had a previous smallpox vaccination. Most of the identified prairie dogs showed clinical signs and many died. Poxvirus was identified in the human lesions and in a prairie dog and characterized as MPXV belonging to the West African clade (Reed et al., 2004). Invasive contact (bites, scratches) with the infected prairie dogs compared to noninvasive contact (touching, cleaning cages) was associated with a higher clinical symptom score and higher rate of hospitalization (Reynolds et al., 2006).

Subsequent whole viral genome sequencing revealed 99% nt sequence identity between MPXV isolates from a given African region, but only 95% nt sequence identity between viruses from Central Africa or Western Africa thus identifying two African clades of MPXV. The MPXV from the 2003 US outbreak belonged to the Western African MPXV clade. Only a single nt difference separated the MPXV from the prairie dog and the human cases of this outbreak, proving animal‐to‐human transmission. The 2003 US MPXV differed from the Central African MPXV in five genome regions, located at the right and left viral genome ends, representing mostly insertions/deletions (Likos et al., 2005).

Isolated cases of exported mpox infection were reported, as for a resident of Singapore attending a wedding in Nigeria where bushmeat was consumed. He developed the disease in Singapore, but secondary cases were not observed in contacts (Yong et al., 2020). Seven imported mpox infections were reported from the UK; four were imported from travel to Nigeria. The disease was transmitted to one health care worker and two household contacts (Adler et al., 2022).

THE 2022 WORLDWIDE MPOX OUTBREAK

Spain

In May 2022, several European countries reported many cases of mpox infections with atypical presentation. Three hospitals from Spain described 181 consecutive patients: 97% were male and the median age was 37 years; 40% were HIV positive but on antiretroviral therapy. Nobody reported a travel to endemic African areas but 92% had sex with men. Skin lesions were few and concentrated on the anogenital and oral regions. No clinical difference was seen whether the patients were HIV positive or not or whether they had received smallpox vaccination in the past or not. Anal swabs yielded more MPXV than oral swabs and the genomes of 23 sequenced viruses belonged to the West African clade of MPXV. Only 2% of the patients were hospitalized, but a third needed pain‐relief medication mostly for proctitis; none died (Tarín‐Vicente et al., 2022).

UK

In early May 2022, a mpox patient with travel history to West Africa was also noted in UK, followed in the next week by six further mpox cases without African travel history. By mid‐July 2022, 1735 mpox cases were counted in the UK, 96% of them were men having sex with men (MSM) and 79% lived in London. A London hospital treated between May and July 197 laboratory‐confirmed mpox patients. The patient characteristics were similar to those seen in Spain: all were in men (median age 37 years); all but one had sex with men; 36% had a treated HIV coinfection; a third had a concomitant sexually transmitted infection. The patients had few mucocutaneous lesions (median of five lesions); 86% showed systemic symptoms (fever, lymphopathy, myalgia); 12% were hospitalized for pain treatment (penis oedema, proctitis). Median hospital stay was 8 days, deaths were not observed (Patel et al., 2022). In May 2022, 68% of the patients from two sexual health clinics in London had a confirmed mpox infection. Inconsistent condom use was the rule and 96% were unaware of having been in contact with a case. Skin lesions were numerous and in various stages of evolution (Girometti et al., 2022).

Spread through Europe

Another report summarized the outcome in mpox cases from 16 countries (mainly Europe and Israel, four sites in the Americas and one in Australia). The clinical and epidemiological characteristics corresponded well to those reported for Spain and UK. For a subgroup of patients detailed data were available allowing to determine the time from exposure to development of symptoms (7 days); the first and last positive PCR for MPXV was 5 and 21 days, respectively, after symptom onset. Diagnostic PCR was positive for 97% of skin and anorectal sites, 90% of semen, 26% of nasopharyngeal, 3% of urine and 7% of blood samples. Transmission was likely by sexual contact: most had five sexual partners during the preceding 3 months, a third had visited sex‐on‐site venues, 20% reported “chemsex” (sex with drugs; Thornhill, Barkati, et al., 2022). Italian scientists reported the isolation of an infectious MPXV for cell culture from the semen of two mpox patients (Lapa et al., 2022). A UK consortium reported on 156 hospitalized mpox patients: 98% were male with a mean age of 35 years; 71% were white; 30% lived with HIV but had a median CD4 count above 500 cells per mm³. Proctitis, rectal pain and secondary bacterial infections were the main reasons for hospitalization. Viral load was higher in lesions and the rectum than in throat swabs (Fink et al., 2023). A study investigating 50 mpox cases from a Paris hospital measured the highest viral loads in skin lesions and anal swabs; after 2 weeks 15% of semen samples were still PCR‐positive (Palich et al., 2023). In a Spanish study with 77 non‐hospitalized mpox patients median time to viral clearance was 25 days for skin lesions, these also presented with 7 log₁₀ per mL the highest viral load. Time to skin viral clearance by PCR in 90% of cases was 41 days, however, no sample with a viral load lower than 6 log₁₀ yielded an infectious virus (Suñer et al., 2023).

Worldwide spread

CDC in collaboration with the International Society of Travel Medicine runs a global clinical‐care‐based surveillance system that monitors infectious diseases in international travellers and migrants called GeoSentinel, which comprises 71 clinical sites in 29 countries on six continents. In May/June 2022, this system recorded 226 proven cases of mpox (many from Spain and Canada). The patient description resembled closely that reported by other national and international mpox surveys in 2022 (Angelo et al., 2023).

Between May and November, 2022, more than 78,000 mpox cases were counted in 109 countries that had previously not reported this disease. Sustained spread outside MSM networks had not occurred. Nevertheless, CDC reported that among 25,000 mpox cases in the USA 3.8% occurred in cis‐women (born as women) and 0.8% in trans‐women. The Share‐Net international clinical network reported on mpox infection in 136 women (about half were trans‐women, half of them were current sex workers). The median number of lesions was 10. Anal sex was associated with anal lesions and vaginal sex with vaginal lesions. Notably, 7% of cis‐women reported no sexual contact during the month preceding the infection. Exposure in these women was occupational via health care (nurses taking samples from mpox patients) or via household contact and other close non‐sexual contacts. All cis‐women with mpox yielded PCR positive skin and vaginal samples and more than 70% of the tested nasopharyngeal and anal swabs were virus‐positive. Overall, 13% of the diseased women were hospitalized, all recovered and the pregnancy in two infected women continued (Thornhill, Palich, et al., 2022). An international consortium investigated mpox in 382 cases, mostly men from the Americas and Europe, who presented with HIV infection and a low CD4 cell count. Overall, 28% of the cases were hospitalized and 7% died; severe complications were more common in people with a very low compared with low CD4 cell count and included necrotizing skin lesions (54% vs. 7%), lung involvement (29% vs 0%) and secondary infections and sepsis (44% vs. 9%; Mitjà et al., 2023).

Viral genome characteristics

Scientists wondered what caused the new transmission dynamics and distinct clinical characteristics of MPXV infections outside of Africa. One possibility is a mutated MPXV genome. Researchers from Portugal sequenced viruses from the earliest mpox cases reported in April 2022 and compared them to viral genome sequences from various regions during the worldwide 2022 outbreak (Isidro et al., 2022). All 2022 viral genomes were closely related and belonged to clade 3, a derivative of the West African clade 2 viruses already observed during the Nigerian outbreak in 2017/2018 and seen in various exported cases from Nigeria to UK, Israel and Singapore in 2018/2019. However, the 2022 viral isolates differed from the 2018/2019 isolates by a mean of 50 single nucleotides polymorphism. This is surprising since Orthopoxviruses normally mutate by only 1–2 substitutions per genome per year. MPXV from the worldwide 2022 outbreak apparently experienced an accelerated evolution. Upon closer analysis, a mutational pattern was detected that suggested the activity of the host APOBEC3 enzymes that belong to the vertebrate innate immune system. APOBEC3 restricts the replication of exogenous viruses via its cytosine‐to‐uracil deaminase activity. Instead of inactivating the virus, this enzyme might here have caused a hypermutation of MPXV. Researchers investigating MPXV genomes from the US 2022 outbreak came to similar conclusions, they identified a predominant B.1 variants that differed by only 1–2 nucleotides from each other and also corresponding European isolates from 2022 but by 13 nucleotide differences from a travel‐associated virus import from Nigeria to the US in 2021. However, these researchers identified a further A.2 variant among the US 2022 MPXV isolates that also belonged to the West African clade but differed by about 80 substitutions from the B.1 variants. All the cases displaying A.2 viruses had travelled to the Near East or Africa and might therefore represent separate viral introductions to the US. In particular, the US variants showed a mutational pattern suggesting APOBEC3 activity (Gigante et al., 2022). It was concluded that the current MPXVs jumped into the human population in early 2016 just before the Nigerian outbreak and had circulated since then in humans. MPXV continues to evolve: an isolate from Germany had one gene duplicated and four genes deleted (Kupferschmidt, 2022a).

Epidemiological analysis

For understanding the epidemiology of mpox and for designing efficient public control measures, it is important to know whether MPXV can cause asymptomatic infections. Indeed, when Belgian clinicians retested left‐over samples from men screened for gonorrhoea and chlamydia infections, 4 out of 237 men tested positive for MPXV in PCR. While one man had a misdiagnosed mpox disease, three did not report any symptoms, but all had seroconverted to MPXV indicating that asymptomatic infections can occur while remaining undetected (De Baetselier et al., 2022).

Non‐sexual transmissions of mpox occurred, but seems to be rare: apart from occupational and household exposure (Thornhill, Palich, et al., 2022) a single case of mpox in a 10‐day old baby from UK was reported. Both parents had shortly before birth a rash and tested positive for MPXV. The child developed a respiratory failure and needed 4 weeks of intensive care before recovering (Ramnarayan et al., 2022). Furthermore, in a Spanish piercing and tattoo parlour 21 of 58 clients were infected with MPXV by contaminated items held in the parlour; 14 female clients without known risk factors developed a systemic rash and a secondary infection to a mother of a client was observed. None of the patients was hospitalized (Viedma‐Martinez et al., 2023).

Model building

What might have caused the surge of mpox outside of Africa? Some scientists suspect human behaviour as the driving force. Epidemiologists developed models according to which no genetic changes in the viral genome is needed to explain the dynamics of the current outbreak. The reproduction number R₀ for MPXV even in an unvaccinated population is below 1, excluding an outbreak as observed in 2022 outside the endemic areas. Using sociological data according to which 2% of the general population are MSM, heavy‐tailed sexual contact networks with a small number of subjects having more than 10 sex partners over 21 days would be enough to lead to the observed outbreak. In such a model R₀ in the MSM network would be substantially higher than 1 and lead to an outbreak. Based on this model, contact tracing and targeted ring vaccination are appropriate public health measures to curtail the epidemic complemented by messages to MSM subjects with multiple partners preferentially propagated by opinion leaders (Endo et al., 2022).

Dynamics and impact of the epidemic

While this model simulates rather well the rise phase of the 2022 MPXV outbreak, it did not predict the subsequently observed decline phase. The outbreak peaked in August 2022 with more than 1000 new cases per day, but subsequently fell over the next 3 months nearly with the same kinetics as it rose since May 2002 to its maximum in August 2022. In May 2023 less than 20 cases per day were registered. Until May 2023 more than 80,000 cases and 140 MPXV‐associated deaths were counted (Mathieu et al., 2022). Deaths occurred by brain swelling (encephalitis possibly as a consequence of an excessive immune response) and in severely immunocompromised patients (Kozlov, 2022). Such a case was described in a HIV‐infected mpox patient with a very low CD4 cell count from US. The patient developed disseminated lesions extending into the oesophagus inducing difficulty in swallowing and severe proctitis. Treatment with the antiviral tecovirimat failed because the infecting MPXV was resistant to this drug. The patient died on hospital day 27 (Alarcón et al., 2023). In Europe, case numbers dropped earlier than in the US, a substantial drop was already seen in September before extended vaccination with smallpox vaccines in at risk populations could have an effect. The ebbing of cases might be a result from behavioural changes since an August 2022 survey conducted by CDC in the MSM community showed that half of their members had reduced the number of sexual contacts which was also supported by data showing that syphilis cases dropped as well in this community (Kupferschmidt, 2022b).

Epidemiological models anticipate that MPXV infected subjects are protected against a reinfection. Overall, this seems to be true but a few cases of re‐infection have been reported but the second infection episode was characterized by low‐viral load and short symptom duration (Raccagni et al., 2023) making further transmission of the infection unlikely. Few data exist about anti‐MPXV immunity. In the US 2003 outbreak, cases showed orthopoxvirus‐specific IgM, IgG, CD4, CD8 and B‐cell responses. Preexisting immunity tended to be associated with milder disease, but smallpox vaccination failed to provide complete protection against human mpox. The analysis of immune markers provided evidence of asymptomatic infections (Karem et al., 2007). Sera from mpox cases from a survey in Africa showed elevated cytokine concentrations in all samples. Overproduction of interleukin [IL]‐2R, IL‐10 and granulocyte macrophage‐colony stimulating factor were observed in patients with serious disease (Johnston et al., 2015). Italian cases from the 2022 mpox epidemic displayed an early expansion of activated effector CD4+ and CD8+ T cells that persisted over time. Inflammatory cytokines (IL‐1β, IL‐6, IL‐8 and TNF) were elevated, irrespective of HIV coinfection status (Agrati et al., 2023).

OUTLOOK

What will the future bring? At best, predictions are that the mpox outbreak will continue to fizzle out as observed in current trends with new case numbers. At worst, the virus could become endemic outside Africa, possibly by reaching a new animal reservoir (Reardon, 2022). This is not a farfetched scenario since MPXV is known to infect more than 50 species of mammal, but so far only two cases of MPXV transmission from pet owner to dogs have been reported in France and Brazil (Mega, 2022).

Current case numbers seem to indicate that the worst of the mpox epidemic is over. One may argue that the 2022 mpox outbreak was not really a pandemic since it affected essentially a subgroup of the human population, namely MSM, albeit with a rapid worldwide spread. Hospitalization rate was fortunately relatively low and mortality was very rare. Maiming sequels in recovered patients were not reported. May the public now lean back with a sigh of relief over another could‐be pandemic that did not reach the proportion of the Spanish flu, AIDS or Covid‐19?

Complacency seems misplaced. While smallpox and VARV have not returned, the historical data clearly show that VARV causing smallpox was probably the greatest plague that has stricken mankind at least over several centuries if not longer. DNA viruses such as VARV with an intrinsically much more stable genome than RNA viruses such as influenzavirus, HIV‐1 virus and SARS‐CoV‐2 can display an astonishing genome dynamic where gene deletions can mediate major virulence changes. VARV cannot establish a persistent infection in a human carrier but VARV could nevertheless become endemic in the human population and could rapidly spread with catastrophic consequences when meeting a previously unexposed population. MPXV is not VARV, neither with respect to spread nor with respect to virulence. However, MPXV is also endemic in Africa where it continues to cause disease manifestations that are more severe than those seen in mpox cases from the 2022 outbreak. MPXV is therefore here to stay, and any niche where it succeeds to replicate, particularly when infecting 100′000 humans across the world, carry the risk that viral genomes evolve and display new and potentially more dangerous characteristics. The 2022 MPXV epidemic out of Africa also demonstrates that a poxvirus that has a reproduction number below 1 for the general population can still grow explosively if it succeeds to enter via specific transmission routes into highly connected human groups. Globalization with international travelling and animal trade allows viruses to spread rapidly over great geographical distances. On the other hand, public health and media reacted quickly spreading the message. WHO declared on 23 July 2022 a Public Health Emergency of International Concern (PHEIC) (Zarocostas, 2022). A potentially effective vaccine and even potentially active antivirals (tecovirimat or cidofovir) were available and derived from past experience with poxviruses and sentinel surveys rung early the alarm bells. Whether these tools combined with messaging in the MSM network helped to rapidly curb the epidemic or whether this reflects an intrinsic dynamic of the epidemic is currently not clear. However, the mpox epidemic highlights one clear message: viruses will remain a major threat to human health and increase in frequency. Pandemic preparedness is a difficult task when not knowing from what corner of the viral world the next attack will come. In any way, pandemic preparedness starts with sound scientific knowledge building about past pandemics and neglected viral diseases in resource‐poor countries.

AUTHOR CONTRIBUTION

HB designed and wrote the report.

FUNDING INFORMATION

No funding information provided.

CONFLICT OF INTEREST STATEMENT

There is no conflict of interest to report.

Brüssow, H. (2023) Pandemic potential of poxviruses: From an ancient killer causing smallpox to the surge of monkeypox. Microbial Biotechnology, 16, 1723–1735. Available from: 10.1111/1751-7915.14294