Abstract
BackgroundIn April 2003, an outbreak of monkeypox occurred in the United States following the importation of monkeypox virus (MPXV)–infected animals in a consignment of exotic pets from West Africa. Transmission of the virus to non-African captive species, including prairie dogs, preceded human disease
MethodsWe evaluated the influence of the route of infection on clinical illness for persons with confirmed and probable cases of human monkeypox. Exposures were categorized as being “noninvasive” (e.g., the person touched an infected animal, cleaned an infected animal’s cage, and/or stood within 6 feet of an infected animal) or “complex” (e.g., invasive bite or scratch from an ill prairie dog plus potential noninvasive exposure), and associations between exposure, illness manifestation, and illness progression (i.e., elapsed time from first exposure to an ill prairie dog through various benchmarks of illness) were assessed
ResultsPatients with complex exposures were more likely than patients with noninvasive exposures to have experienced pronounced signs of systemic illness (49.1% vs. 16.7%; P=.041) and to have been hospitalized during illness (68.8% vs. 10.3%; P<.001). Complex exposures were also associated with shorter incubation periods (9 days for complex exposures vs. 13 days for noninvasive exposures) and the absence of a distinct febrile prodrome
ConclusionsThe findings of this study indicate that route of infection can influence monkeypox illness manifestations
Of the species of virus in the genus Orthopoxviridae that are capable of causing disease in humans, monkeypox virus (MPXV) and variola virus are unique in their capacity to cause severe systemic disease accompanied by a generalized vesiculopustular rash [1]. Naturally occurring smallpox was successfully eradicated as of 1977, in part because variola virus was an obligate human pathogen and was, therefore, not maintained in any nonhuman animal reservoir [1, 2]. MPXV, on the other hand, is a zoonotic virus (an animal reservoir for MPXV has not been definitively identified) that is considered to be endemic to regions of West Africa and the Congo basin
MPXV shares several distinctive biological features both with variola virus and with the zoonotic orthopox viruses, vaccinia virus, and cowpox virus. Like variola virus, MPXV (Congo basin variant) can be spread from person to person, although it is evident that interhuman spread of MPXV is less efficient than was historically observed for variola virus. Person-to-person spread of both viruses is thought to occur principally via infectious oropharyngeal exudates generated during the rash phase of illness or possibly during the preceding 2–3–day period of febrile prodrome [1, 3, 4]. Alternatively, like vaccinia virus and cowpox virus, MPXV can infect a broad range of mammalian animal species [5–14] and can be transmitted to humans by means of direct contact with infected animals, often by means of traumatic injury to the skin (or, for vaccinia virus, by means of inoculation). Rare instances of human-to-human spread of vaccinia virus and cowpox virus are attributed to percutaneous or mucocutaneous infections [15–18]
Therefore, it is evident that human monkeypox infection can be acquired naturally by means of varied routes (e.g., the respiratory tract, skin, and mucous membrane) and from varied sources (e.g., rodents, humans, and nonhuman primates). Neither the relative significance of the different potential sources of MPXV to naturally occurring human disease nor the influence of different routes of infection on clinical manifestations or illness progression have been systematically evaluated
In 2003, MPXV was brought into the United States from West Africa in a consignment of exotic animals intended for sale as pets [5]. The virus is thought to have been transmitted from African animals in the consignment to a number of susceptible non-African species with which they were cohoused, prairie dogs among them [5]. Forty-seven cases of human MPXV infection (confirmed and probable cases; J. Cono, personal communication) were documented, each stemming from direct or indirect contact with MPXV-infected prairie dogs. There were no instances of human-to-human transmission documented [5, 19]. Persons who became infected with MPXV were, however, exposed to ill prairie dogs in different settings and by means of varied types of contact [20]. Some individuals were bitten, many reported handling the animals without being bitten, and some reported only having been in the same room with a sick prairie dog, suggesting that MPXV infections may have been initiated in different persons by means of a variety of mechanisms, including percutaneous, inhalational, and mucocutaneous exposures
A previous study demonstrated that patient age and smallpox vaccination status had little influence on illness manifestation or severity among persons infected with MPXV during the outbreak in the United States in 2003 [21]. In this study, we reviewed confirmed and probable human monkeypox case histories from that outbreak and evaluated the influence of probable route of infection on various indicators of illness, including severity, presentation (symptoms), and illness progression
Subjects, Materials, and Methods
SubjectsThis study included all persons (n=47) identified with confirmed (n=37) or probable (n=10) MPXV infections during the monkeypox outbreak in the United States during May–July 2003 [5] (Cono, personal communication). Standardized case definitions based on clinical symptoms, exposure information, and laboratory criteria [22, 23] were used to allow for systematic assignment of case status to individuals with suspected MPXV infections [24, 25]. Individuals who had illness onset within 21 days after exposure to MPXV who experienced fever (defined as a body temperature ⩾37.4°C) and vesicular pustular rash or rash (potentially uncharacterized) plus orthopox IgM antibodies were classified as having probable cases of infection. Monkeypox cases were confirmed on the basis of any of the following laboratory findings from clinically derived specimens: MPXV isolation, detection of MPXV-specific nucleic acid signatures, positive electron-microscopy findings, or positive immunohistochemical findings (in the absence of other orthopox virus exposures)
Data collection and sourcesInformation was collected during the public health response to the monkeypox outbreak using standardized questionnaires and interviews conducted by medical personnel or public health professionals. Information pertaining to demographic data (e.g., age, sex, and jurisdiction), exposure type (e.g., bite or scratch) and setting (e.g., veterinary clinic or home), clinical signs and symptoms, laboratory test results, illness progression, and laboratory findings was collected for analysis. In addition, detailed information pertaining to MPXV exposure was collected by means of a household exposure survey. The protocol for this survey, which involved unpaid volunteers, was approved by the Centers for Disease Control and Prevention and partner human subjects institutional review boards
Assessment and classification of human exposures to MPXVExposure type refers to human contact with a prairie dog known to be infected with MPXV [9, 20] (C. Hutson, personal communication) or an ill prairie dog with an epidemiologic link to an animal known to be infected [5]. All reported exposures were categorized as being “complex”—consisting of an invasive exposure, such as a bite or scratch sufficient to break the skin, plus potentially noninvasive exposures—or “noninvasive”—defined as contact with an infected animal not resulting in a break in the skin, such as physical handling or being in the proximity of an infected animal. The complex exposure category was used to account for situations in which there was opportunity for both invasive and noninvasive exposures to have occurred. Exposure setting was categorized as taking place inside or outside of the home
Assessment of clinical parametersFor purposes of comparative analysis across exposure groups, sign and symptom categories based on the type and number of individual symptoms were established, because the manifestations of MPXV infection can vary from general systemic involvement to include respiratory, gastrointestinal, and other systems [21, 26]. These categories are as follows: (1) gastrointestinal (includes diarrhea and nausea and/or vomiting), (2) upper respiratory (includes runny nose and sore throat), (3) lower respiratory (includes wheeze, cough, and respiratory distress), (4) general respiratory (includes combined symptoms from the upper and lower respiratory categories), and (5) systemic (includes lymphadenopathy, fever, sweats, chills, muscle ache, back pain, headache, and abdominal pain). Gastrointestinal and lower and upper respiratory symptom categories were classified as having either presence or absence of any of the designated signs or symptoms. Within the general respiratory and inflammatory symptom categories, the presence of ⩾6 systemic symptoms and ⩾4 respiratory symptoms, respectively (each represents the upper quartile per category within the population of cases), were defined as being pronounced
In keeping with previous studies, we evaluated the classic triad of fever intensity, rash burden, and “degree of prostration” as a measure of overall monkeypox illness severity [26–28]. Illness progression was measured according to elapsed time intervals from the date of first MPXV exposure to various benchmarks of illness. Median values for intervals from the first possible exposure to onset of the first symptom (n=42), fever onset (n=32), rash onset (n=45), and the date of maximum temperature (n=31) were calculated using the population of MPXV-infected persons to create general time courses of illness for different categories of exposure
In this study, a lesion is defined as a single circumscribed area, which may include a blister, papule, or eschar. A rash is defined as consisting of multiple lesions, including localized constellations or diffusely distributed lesions, involving macules, papules, and/or pustules
Statistical analysesThe association between exposure (setting and type) and clinical features of monkeypox (signs and symptoms, rash characteristics, and clinical laboratory parameters) were assessed using nonparametric statistical tests, because the data were not normally distributed [29]. Fisher’s exact test (2-tailed) was used for categorical variables, and the Wilcoxon&rank-sum test was used for continuous variables. A P value of <.050 was used to measure the significance of associations. For most variables of interest, >65% of the cases had available data
Results
Exposure settings and probable routes of exposureOf the 47 patients with confirmed or probable cases of human monkeypox, 27 (57%) reported having probable exposure to MPXV in a home environment (table 1) by means of contact with an ill pet. The remaining 20 patients (43%) were all exposed in settings associated with occupational animal care (e.g., pet stores, animal swap meets, and veterinary clinics). These locations were categorized as being “nonhousehold” exposure settings. The median time window for potential exposures was significantly longer among individuals exposed in the home, compared with that among individuals exposed in other settings (10 days and 1 day, respectively; P=.003)
Characteristics of persons infected with monkeypox virus in the United States in 2003, by exposure category
Individuals were queried regarding the nature of their exposure to infected animals, which varied independently of exposure setting. A specific invasive exposure (i.e., bite or scratch) occurred among persons 12 (44%) with a household exposure and among 5 persons (25%) with a nonhousehold exposure (P=.226). Overall, 17 individuals (36%) received a bite or scratch from an ill prairie dog in addition to other potential noninvasive exposures. Of these, 6 persons (35%) reported the development of a lesion (presumably an inoculation lesion) or eschar at the site of the bite or scratch, but this information was not routinely documented for every case in this study. The remaining individuals reported having only noninvasive exposures, which included handling (touching), cleaning the cage of, or being in the same room with an ill prairie dog. More women (11) than men (6) reported having received a bite or scratch, but this difference was not statistically significant (64.7% and 35.3%, respectively; P=.239)
Associations between exposure and clinical manifestationsof monkeypoxTo evaluate the influence of probable route of infection and exposure setting on monkeypox illness, different measures of illness manifestation and severity were assessed in relation to MPXV exposure. The relative frequencies of occurrence of individual symptoms by route of exposure are shown in figure 1
Frequency of individual symptoms reported among monkeypox virus–infected persons exposed to the virus by means of complex (combined invasive and noninvasive) or noninvasive routes. *P=.075
Among monkeypox cases identified in the US outbreak, no statistically significant associations were observed between exposure setting or probable route of infection and either fever intensity or rash burden (table 1). In addition, the distribution of outcomes for lesion count at height of illness and maximum recorded body temperature were similar for all categories of monkeypox cases, regardless of exposure setting or probable route of infection. The third element of illness severity, degree of prostration, was addressed by evaluating symptom tableau and, to a lesser extent, hospitalization. When symptom tableau and hospitalization for MPXV-infected persons exposed in the household and for persons exposed in other settings were compared, no significant differences were observed (table 1)
Differences in illness manifestations were, however, identified when the probable route of infection was examined. Individuals who reported a complex exposure were more likely than persons who had noninvasive exposures to have experienced pronounced systemic symptoms (8 persons [49.1%] and 5 persons [16.7%], respectively; P=.041) and to have been hospitalized during illness (11 persons [68.8%] and 3 persons [10.3%], respectively; P<.001). A difference in the frequency of gastrointestinal symptoms between these 2 groups was suggested (8 persons [47.1%] and 6 persons [20%], respectively; P=.095). Indeed, 78% of all MPXV-infected persons who experienced gastrointestinal symptoms (11 persons), regardless of exposure category, were hospitalized, and 43% of persons with pronounced systemic symptoms (13 persons) were hospitalized. Upper respiratory symptoms occurred at a high frequency (80%) among individuals with noninvasive exposures (24 persons), but this was not statistically significantly different from the frequency among individuals with complex exposures (P=.176)
Progression of illnessTo determine whether probable route of MPXV infection impacts progression of illness, we calculated median time intervals from the first possible exposure to various benchmarks of illness onset for individuals who had complex exposures and for persons who were exposed by means of a noninvasive route (figure 2). Aggregated data were used to create a generalized model of the time courses of illness for the different categories of exposure
Illness progression among monkeypox virus–infected persons exposed to the virus by means of complex and noninvasive routes. All values represent the median no. of days from the first reported date of exposure to the end date indicated. The timeline of illness shown is truncated at 21 days, but this does not imply illness end point. Our survey did not capture the no. of days to lesion desquamation. Rash is defined as consisting of multiple lesions, including localized constellations or diffusely distributed lesions, involving macules, papules, and/or pustules
Examination of illness progression (figure 2) suggested that individuals who were exposed to MPXV by means of noninvasive routes experienced illness progression similar to that described in large observational studies conducted in the Congo basin [26], with a ∼13-day incubation period followed by a 2–3–day febrile prodrome preceding onset of rash. In contrast, complex exposures were generally associated with a shorter incubation period, symptom onset in the absence of fever, and rash or lesion (defined above) development prior to the onset of pronounced fever (see below). The patterns did not differ substantively when the analysis was restricted to individuals who had well-defined windows of exposure (i.e., persons who were infected in nonhousehold settings [n=20]). The median durations of illness for smallpox-vaccinated and non–smallpox-vaccinated individuals were indistinguishable. None of the individuals in the former group had been vaccinated (or revaccinated) for smallpox since at least 1972
Comparison of prodrome characteristicsThe presence of a febrile prodrome in conjunction with various other inflammatory symptoms (e.g., lymphadenopathy, chills, back pain, and headache) is considered to be typical for human monkeypox. However, because a period of febrile prodrome was not prominent among MPXV-infected individuals who had complex exposures to an ill animal, we examined the first reported signs or symptoms of illness in these individuals, compared with the first signs or symptoms in individuals who had noninvasive exposures
The proportion of patients with monkeypox in each exposure category who reported an initial symptom of either lesion, rash, gastrointestinal complaint, upper respiratory complaint, or systemic illness complaint is shown in figure 3. Of persons for whom the information was available, 6 (50%) with complex exposures experienced either gastrointestinal symptoms or a lesion at illness onset, whereas none with noninvasive exposures reported early gastrointestinal symptoms, and only 1 (6.7%) reported a lesion at illness onset (P=.045 and P=.128 for gastrointestinal and lesion occurrence as first symptom, respectively). Instead, 40% of individuals with noninvasive exposures reported initial upper respiratory symptoms or rash (6 persons), both of which were infrequent early symptoms among persons who had complex exposures (P=.281 and P=.258 for upper respiratory symptoms and rash, respectively)
First symptom experienced by persons infected with monkeypox virus. Data for the first reported symptom were available for 12 and 15 individuals from the complex and noninvasive exposure groups, respectively. A lesion is defined as a single circumscribed area, which may include a blister, papule, or eschar; a rash is defined as consisting of multiple lesions, including localized constellations or diffusely distributed lesions, involving macules, papules, and/or pustules. *P=.045
Discussion
During the US monkeypox outbreak in 2003, individuals contracted MPXV infections from infected prairie dogs; no human-to-human transmission was documented, but there were many different potential scenarios of infection involving respiratory and/or mucocutaneous exposures, percutaneous and/or inoculation exposures, or both, providing a unique opportunity to evaluate whether differences in exposure might influence clinical aspects of monkeypox disease in humans. The results of this study demonstrate that the route of MPXV infection among patients with monkeypox during the US outbreak in 2003 had a measurable influence both on monkeypox illness manifestations and on disease severity
A previous study of a large population of MPXV-infected individuals in the Congo basin failed to identify any meaningful differences in illness presentation between persons who contracted infection from primary zoonotic sources versus persons who contracted infection from a human source [30]. However, in that study, the likely route of infection from the primary animal source was not specified and, in fact, may have encompassed both direct and indirect routes. Only a single account of an animal bite associated with a human MPXV infection has been reported in Africa; it involved an infant in the Democratic Republic of the Congo who received a bite on the left foot from a chimpanzee [11]. The child, who ultimately survived, experienced fever 6 days after the bite (suggesting a short incubation period) and developed a wound lesion along with pronounced inguinal lymph node swelling. The nodal swelling occurred earlier and was more significant on the left side. This single description is generally consistent with the clinical picture after invasive exposure to MPXV, as it is described in this report. The relative importance of potential transdermal exposures has not been explored in African settings, but such exposures should be considered, particularly because they may affect incubation periods and prodromal characteristics, both of which are important features used to differentiate monkeypox from other, more common rash-associated illnesses, such as chickenpox [31]
Illness severity is another feature of monkeypox that is likely influenced by route of infection. Historically, fever intensity, rash burden, and degree of prostration were often used to estimate severity of illness among patients with monkeypox [26–28]. In our study, we found that, overall, fever intensity and rash burden did not vary among patients by the manner in which they were exposed. However, regarding the third element of severity—“degree of prostration”—deriving an objective, empirical assessment for comparative purposes proved to be somewhat difficult. For example, individuals who had complex exposures were significantly more likely to have been hospitalized during illness (one potential indicator of prostration), but during the US outbreak, some patients with monkeypox were hospitalized for purposes of isolation, rather than for care or treatment. Nevertheless, in addition to being significantly more likely to have been hospitalized, individuals who had been bitten or scratched exhibited a greater number of systemic complaints (fever, chills, and backache) and were more likely to have experienced nausea or vomiting, especially early during illness. Together, these observations are suggestive of a greater degree of prostration. By contrast, individuals who were likely to have been exposed to MPXV by means of respiratory or mucosal routes (persons with noninvasive exposures) experienced some upper respiratory symptoms but had fewer overall systemic symptoms and a more classical monkeypox illness progression accompanied by a 2–3–day febrile prodrome without pronounced gastrointestinal involvement
The time course of illness was also distinctly different for individuals in the 2 exposure categories. This observation is generally consistent with observed differences in the progression of illness between naturally acquired smallpox (respiratory) and illness resulting from inoculation smallpox (introduction of variola virus across the skin by means of vaccination [i.e., variolation]) [1, 32]. The progression of typical, naturally acquired smallpox is well defined, from initial respiratory exposure (involving localized virus replication and migration to regional lymph nodes), through incubation (virus proliferation in reticuloendothelial tissues), through a period of febrile prodrome (which follows onset of secondary viremia), and finally to rash (virus dissemination to dermis) [33]. In the case of inoculation smallpox, however, the incubation period (the estimated period from exposure to fever onset) was reduced relative to that for naturally acquired smallpox (9 vs. 13 days), and primary lesion development at the site of inoculation was observed to occur prior to the appearance of fever and generalized rash. The lack of initial infection and replication in the respiratory tract presumably accounts for the shortened time course of infection and the truncated incubation period associated with inoculation smallpox. Inoculation smallpox was typically, although not uniformly, less severe than naturally acquired smallpox. In our study, inoculation exposures to MPXV, in combination with potential noninvasive exposures, were associated with compressed incubation periods, compared with noninvasive exposures alone (8 and 13 days, respectively), and in parallel with observations described above, lesion development (presumably a consequence of the inoculation exposure) was frequently noted prior to the onset of fever. Of note, however, is that in contrast to historical observations of a reduced severity of smallpox illness following controlled transdermal inoculation of variola virus, no diminution of monkeypox illness severity was observed among persons with complex MPXV exposures; rather, these individuals appeared to have experienced somewhat more-severe courses of illness. There are many factors that could account for this disparity of outcome, including viral dose, quality of inoculum, inoculation mechanism, and, perhaps most significantly, the fact that inoculation smallpox occurred in the absence of potential additional sources of airborne virus, whereas in our study, it was difficult to envision invasive exposure in the complete absence of airborne virus
There are limitations to the inferences that can be drawn from this study. First, we assigned MPXV-infected individuals to discrete exposure categories on the basis of their contact with sick or infected prairie dogs reported ∼8 weeks after exposure, which may be affected by recall. Complex exposures occurred mainly in household settings where individuals tended to have longer opportunities for potential exposure, in theory affording greater opportunities for additional respiratory and/or mucosal exposures. Also, the small size of the available study population of infected individuals constrained our ability to use multivariate modeling approaches to identify independent variables associated with illness manifestations and outcomes. For several variables of interest (e.g., lesion count at height of illness and date of maximum recorded body temperature), the analyses were also limited by incomplete reporting or recall of information by patients. And, because of the retrospective nature of the study, we were unable to obtain highly detailed data pertaining to the duration and intensity of exposures
MPXV infections in humans have been observed to result in a spectrum of clinical presentations from mild to severe [26, 30]. Although some of this variation is likely due to differences in gene composition and sequence at key sites in the MPXV genome [28, 34], the results of this investigation suggest that, for a single genotype of virus, some observed variations in illness manifestations and severity may also be due to differences in the way people are exposed to the virus. Patient age, smallpox vaccination status, and exposure characteristics all varied among individuals who became infected with MPXV during the US outbreak in 2003; however, age and vaccination status ultimately had little influence on clinical disease manifestation [21]. The manner in which a person was exposed to MPXV, however, impacted both illness progression and symptom presentation, with complex exposures involving a combination of invasive (bite or scratch) and noninvasive contact found to be associated with a compressed incubation period, increased systemic illness, and greater acuity of health care. The findings of this study should have important implications for monkeypox case recognition and management
Acknowledgments
We gratefully acknowledge and thank the reviewers of this manuscript and all the individuals involved in obtaining the data used in this study, including but not limited to Aaron Curns, Joanne Cono, Tracee Treadwell, Craig Conover, Marybeth Graham, Mark Sotir, Pam Pontones, and Mark Wegner
References
Presented in part: 54th Annual Meeting of the American Society of Tropical Medicine and Hygiene, Washington, DC, December 2005 (abstract 43)
Potential conflicts of interest: none reported
Financial support: Centers for Disease Control and Prevention (federal allocation apportioned to the National Centers for Infectious Diseases)
The findings and conclusions in this presentation have not been formally disseminated by the Centers for Disease Control and Prevention and should not be construed to represent any agency determination or policy
Present affiliation: Rush University Medical Center, Chicago, Illinois
