Identification of Cases
By June 7, 11 patients with confirmed or suspected monkeypox had been identified in central and southeastern Wisconsin. At that time, a suspected case of monkeypox in humans was defined as vesiculopustular skin lesions and fever in persons exposed to animals obtained from Distributor 1 or 2. All 11 patients were linked by direct contact with prairie dogs sold by Distributor 2. The index family (which included Patients 1, 2, and 3) had purchased two prairie dogs from Distributor 2. The eight patients identified in southeastern Wisconsin (Patients 4 through 11) included Distributor 2 and his wife, two employees of different pet stores (1 and 2) that had received prairie dogs from Distributor 2, two veterinarians from different clinics who had treated ill prairie dogs sold by Distributor 2, a person who had purchased two prairie dogs from Pet Store 2, and her houseguest (Figure 1). The earliest date of onset of illness was May 15.
Clinical Features
The 11 initial cases in Wisconsin occurred in 5 males and 6 females ranging in age from 3 to 43 years. Six of the 11 patients had previously received a single dose of smallpox vaccine during childhood. In all cases transmission of monkeypox virus appeared to be by direct contact with an infected prairie dog. However, Patients 2 and 3 provided direct care to their infected child, and the possibility of person-to-person transmission cannot be excluded. Patients 1 and 4 were scratched or bitten by an ill prairie dog. In three patients, the infection appeared to be transmitted directly to open wounds: a cat scratch on the hand of Patient 2, a cut on the hand of Patient 6, and brush scratches on the legs of Patient 7. The incubation period for the infection was difficult to determine owing to the lengthy intervals of exposure to infected prairie dogs and could have ranged from 4 to 24 days (median, 15; mean, 14.5).
Initial signs or symptoms were typically skin lesions or fever (temperature above 38°C) with drenching sweats and severe chills. Frequent signs and symptoms were skin lesions (100 percent), headache (100 percent), fever (82 percent), sweats (82 percent), chills (82 percent), persistent cough (73 percent), lymphadenopathy (55 percent), and sore throat (55 percent). Less frequent signs and symptoms were pharyngitis (27 percent), tonsillar hypertrophy (18 percent), tonsillar erosions (18 percent), malaise (18 percent), mild chest tightness (18 percent), diarrhea (18 percent), myalgias (9 percent), back pain (9 percent), nasal congestion (9 percent), blepharitis (9 percent), and nausea (9 percent). Pulmonary, cardiac, abdominal, neurologic, and musculoskeletal examinations were normal in all patients examined.
Figure 2. Figure 2. Primary Inoculation Reactions (Panels A, B, and C), Examples of the Smallpox-like (Panel D) and Umbilicated Varicella-like (Panel E) Disseminated Monkeypox Lesions, and the Morphologic Appearance of Disseminated Lesions over Time (Panels F, G, H, and I). Panel A shows a primary inoculation reaction at the site of a prairie-dog bite on Patient 1, Panel B a prairie-dog scratch on Patient 4, and Panel C a preexisting cat scratch on Patient 2. Panel F shows a disseminated lesion less than 24 hours after its appearance, Panel G lesions after six days, Panel H a lesion after 96 hours, and Panel I a lesion after more than nine days.
In five patients, primary skin lesions appeared as nodular swellings around the margins of bites or scratches (Figure 2A, Figure 2B, and Figure 2C). In most patients, including the six patients without identifiable primary lesions, 1 to 50 satellite and disseminated skin lesions developed over a period of several days. Lesions evolved from papules to vesiculopustules, some with prominent erythematous flares, and resolved leaving serous-to-hemorrhagic crusts that eventually detached. Different stages in the evolution of the lesions were seen simultaneously in all patients with multiple lesions. Lesion sites included the face, scalp, hands, arms, legs, trunk, perineum, conjunctivae, and buccal mucosa (Figure 2D, Figure 2E, Figure 2F, Figure 2G, and Figure 2H). Larger lesions left central scars (Figure 2I). Four of the 11 patients were hospitalized for their illnesses. The clinical course was self-limited in all cases. The median time to crusting of all skin lesions was 12 days (range, 3 to 25). Nine patients received antibiotics (six received ciprofloxacin, and eight were given doxycycline). Patient 1 received intravenous acyclovir, and Patients 2 and 4 received valacyclovir. No patients received vaccinia immune globulin.
Laboratory Investigations
A battery of laboratory tests was used to identify the etiologic agent of infection. The specific tests used to confirm the diagnosis of monkeypox for individual patients varied depending on the availability of specimens and the stage of illness. Results are summarized in Figure 1.
Figure 3. Figure 3. Histologic, Immunohistochemical, and Ultrastructural Evaluation of the Skin-Biopsy Specimen from Patient 2. Panel A shows scattered degenerating and necrotic keratinocytes within the epidermis and a moderate inflammatory-cell infiltrate within the epidermis and superficial dermis (hematoxylin and eosin, ×50). Panel B shows the boxed area in Panel A at a higher magnification (×200); a multinucleated cell (long arrow) and eosinophilic viral inclusion bodies (short arrows) are evident. Panel C shows immunohistochemical staining of orthopoxvirus antigen within the epidermis (horseradish peroxidase with hematoxylin counterstain, ×40). The inset shows immunoreactivity within individual keratinocytes (×250). Panel D shows virions within the cytoplasm of a keratinocyte and includes immature forms that are being assembled (long arrow) and clusters of mature virions (short arrow). N denotes nucleus. Panel E shows virions with dumbbell-shaped cores characteristic of poxviruses. Panel F shows a negatively stained virion from cell culture (phosphotungstic acid). The brick-shaped particle has regularly spaced, threadlike ridges on the exposed surface.
Sections of formalin-fixed, paraffin-embedded skin-biopsy specimens obtained at the pustular stage from two patients and stained with hematoxylin and eosin showed marked ballooning degeneration of keratinocytes with epidermal necrosis and spongiotic edema (Figure 3A). In portions of the epidermis, keratinocytes exhibited multinucleation, nuclear clearing with margination of chromatin, and occasional eosinophilic cytoplasmic inclusions (Figure 3B). A moderate neutrophilic and lymphocytic inflammatory infiltrate was present within the epidermis and superficial dermis.
Orthopoxviral antigen was detected in skin-biopsy specimens by means of immunohistochemical staining with rabbit antivaccinia polyclonal antibody (Virostat).11 Antigen was most prominent in degenerating keratinocytes and follicular epithelium but was absent in normal-appearing skin at the edges of the specimen (Figure 3C). Immunohistochemical stains for herpes simplex viruses 1 and 2 and cytomegalovirus were negative.
Tissue samples from seven patients and a submandibular lymph node from Prairie Dog 1 were inoculated onto primary and continuous cell lines, including rhesus-monkey kidney, rabbit kidney, MRC-5, RD, HEp-2, B-SC-40, and VERO cells. Except for the HEp-2 cells, cytopathic changes occurred in all cell lines within one to four days and were characterized by plaques of elongated and rounded cells with prominent cytoplasmic bridging and formation of syncytium.
Transmission electron microscopy was performed on glutaraldehyde-fixed skin-biopsy specimens from two patients.11,12 Virions in various stages of assembly were observed within the cytoplasm of keratinocytes (Figure 3D). Cross sections of mature virions contained dumbbell-shaped cores characteristic of poxviruses (Figure 3E). Negative-stain electron microscopy with phosphotungstic acid13 was performed on cultures from Patient 2 and Prairie Dog 1 and revealed numerous brick-shaped virions with regularly spaced, threadlike ridges on the exposed surfaces (Figure 3F). Taken together, the cell-culture and ultrastructural features suggested that the virus was a member of the genus orthopoxvirus.
Figure 4. Figure 4. Polymerase-Chain-Reaction Amplification of the Orthopoxvirus Hemagglutinin Gene Followed by Restriction-Fragment–Length Polymorphism Analysis of Samples from Patient 4, Prairie Dog 1, and Reference Isolates of Other Orthopoxviruses (Panel A) and Phylogenetic Tree of Orthopox Isolates (Panel B). TaqI was used for the restriction-fragment–length polymorphism analysis. Panel B shows that the gene sequences of samples from Patient 4 and Prairie Dog 1 shown in Panel A are closely related according to neighbor-joining methods. Significant bootstrap values for major nodes of the resulting hemagglutinin gene dendrogram are also shown. MPV denotes monkeypox virus, CPV cowpox virus, RPV rabbitpox virus, and VAC vaccinia virus. Reference monkeypox strains in the upper node (e.g., MPV-UTC) originated primarily in West Africa, whereas the strains in the middle node (e.g., MPV-CONGO8) are typically of Congolese (central African) origin. Percentages in Panel B indicate the relative similarity of the strains.
For molecular diagnosis, DNA was extracted from tissues, specimens obtained with swabs, “touch” preparations (obtained by touching the patients' skin), and cell-culture supernatants.14 Clinical samples were positive for orthopoxvirus on a PCR assay that amplifies a conserved segment of the DNA polymerase gene (E9L) present in all Old World orthopoxviruses except variola. To characterize the virus further, PCR of the orthopoxvirus hemagglutinin and A-type inclusion genes was performed, followed by restriction-fragment–length polymorphism analysis.15-18 Restriction profiles obtained from samples from the patients and Prairie Dog 1 were identical to reference strains of monkeypox virus and differed from other known orthopoxviruses (Figure 4A). In addition, a PCR assay for the gene for the monkeypox extracellular-envelope virus protein was positive. In contrast, a real-time PCR assay designed specifically to identify the vaccinia virus cytokine-response modifier B gene sequences was negative.
The combined PCR test results involving independent gene sequences provided strong evidence that the DNA signatures in the clinical samples were caused by a “monkeypox-like” Old World orthopoxvirus. Excluded orthopoxviruses included ectromelia and gerbilpox viruses (both of which have been isolated from rodents but are not considered human pathogens), vaccinia virus (which has the potential to infect some rodents), cowpox virus (which has rodent reservoirs and can cause limited infections in humans), camelpox virus (which is not a cause of human infections), variola (smallpox) virus, and the recognized New World orthopoxviruses.
The remote possibility that an undiscovered Old World orthopoxvirus might have the same DNA fingerprints as monkeypox virus led us to sequence the entire orthopoxvirus hemagglutinin gene from the samples. The amplified template of both strands was sequenced to eightfold repetitive coverage. The generic primer sequence, first used to amplify the hemagglutinin gene product, was corrected for equivalent monkeypox-specific sequences. The observed hemagglutinin gene sequence obtained from the clinical tissues was compared with GenBank sequences. Closely related gene sequences were compared by means of neighbor-joining methods, and significant bootstrap values were obtained for major nodes of the resulting hemagglutinin gene dendrogram (Figure 4B). The entire hemagglutinin gene sequence of the novel North American isolates was identical to examples of hemagglutinin genes obtained from monkeypox virus isolated from humans in West Africa and from nonhuman primates in primate colonies.