Evidence for Infection of Camels

MERS-CoV infection in camels has been demonstrated through serologic investigations, molecular evidence using real-time reverse transcription PCR (rRT-PCR), and by virus isolation, as described in recent reviews (16,17). Geographically wide-ranging seroprevalence studies have identified MERS-CoV–specific antibodies in camels in countries across the Middle East and North, West, and East Africa, often with >90% seroprevalence in adult camels (18). Studies in many of these countries have shown molecular evidence of MERS-CoV RNA and isolation of infectious MERS-CoV in camels (16,17,19–21).

Viral Shedding in Camels

In naturally or experimentally infected camels, MERS-CoV appears to cause an upper respiratory tract infection with or without symptoms, including nasal and lachrymal discharge, coughing, sneezing, elevated body temperature, and loss of appetite (20,22,23). In naturally infected camels, MERS-CoV RNA has been recovered most commonly from nasal swabs but also from fecal swabs, rectal swabs, and lung tissue (20,24). No evidence of viral RNA has been demonstrated in camel serum, blood, or urine using rRT-PCR (25,26). In experimentally infected camels, infectious virus and RNA was detected in nasal swab and oral samples but not in blood, serum, feces, or urine (23). MERS-CoV RNA has been detected in raw camel milk collected using traditional milking methods, including using a suckling calf as stimulus for milk letdown; presence of live virus was not evaluated (27). Virus RNA may therefore have been introduced via calf saliva or nasal secretions or fecal contamination. Experimentally introduced virus can survive in milk but did not survive when heat treated (28). It is also not known if the virus would remain viable in milk from seropositive dams when antibodies could be found in the milk.

These shedding data indicate that contact with camel nasal secretions, saliva, and respiratory droplets carry potential risk for camel-to-human or camel-to-camel transmission. Contact with nasal secretions can occur when directly handling live camels, and virus from camel nasal secretions can contaminate fomites in the environment (29). Although rRT-PCR evidence of MERS-CoV and genome fragments have been detected in air samples from a camel barn (30), no live virus was detected, and no epidemiologic study has implicated airborne transmission. Transmission following exposure to camel feces may be biologically plausible, although no epidemiologic evidence indicates the likelihood of such transmission. Similarly, although transmission following consumption of raw camel milk may be biologically plausible, epidemiologic studies have not consistently identified milk consumption as a unique risk factor for MERS-CoV infection or illness, independent of other direct or indirect camel exposures (31,32). No epidemiologic evidence supports transmission associated with camel urine or meat.

Camel-to-Camel Transmission Dynamics

MERS-CoV RNA is detected most frequently in younger camels (22,25,33) but has been detected in camels >4 years of age (22). In a longitudinal study of a camel dairy herd, most calves became infected with MERS-CoV at 5–6 months of age, around the time maternal MERS-CoV antibodies wane. The calves then produced MERS-CoV antibodies by 11–12 months of age (34). In seroprevalence studies, camels <2 years of age demonstrated lower seroprevalence than camels >2 years of age (25,35). Across many countries, the seroprevalence of adult camels is >90% (16,17). Overall, these data suggest most camels are initially infected with MERS-CoV at <2 years of age. However, camels can shed virus despite preexisting MERS-CoV antibodies, suggesting that repeat infections are possible (36,37). Varying prevalence of MERS-CoV RNA in camels has been reported in different countries and settings, such as farms (33) and live animal markets (38). Risk for camel-to-camel or camel-to-human transmission may be influenced by crowding, mixing of camels from multiple sources, transportation, and characteristics of live animal markets (39). Phylogenetic modeling has provided supportive evidence that long-term MERS-CoV evolution has occurred exclusively in camels, with humans acting as a transient and usually terminal host (40).

Evidence and Risk Factors for Zoonotic Transmission

Persons in Saudi Arabia with occupational exposure to camels demonstrated higher seroprevalence of MERS-CoV–specific antibodies (camel shepherds, 2.3%; slaughterhouse workers, 3.6%) compared with the general population (0.2%) (41). A case–control study of primary human cases and matched controls also showed that camel exposure was more likely among case-patients than matched controls (31). Further evidence supporting camel-to-human transmission includes identical or nearly identical MERS-CoV sequences found in camels and humans (41–45).

Occupational groups with frequent exposure to camels have been assessed through seroepidemiologic studies. In Qatar, a study of 9 seropositive and 43 matched seronegative camel workers showed that regular involvement in training and herding of camels, cleaning farm equipment, not handwashing before and after camel handling, and milking camels were associated with seropositivity (46). In a Saudi Arabia study of 30 camel workers in which 50% were seropositive for MERS-CoV, no association was identified between seropositivity and factors including age, smoking, handwashing after camel contact, consuming camel meat or milk, or specific occupation (camel truck driver, handler, or herder) (47). Neither investigation controlled for possible confounding risk factors (e.g., age or duration of exposure to camels). In Abu Dhabi, an investigation of 235 market and slaughterhouse workers showed that 17% were seropositive for MERS-CoV and that daily contact with camels or their waste, working as a camel salesman, and self-reported diabetes were risk factors for seropositivity (32). Among market workers in the same study, handling live camels and either cleaning equipment (e.g., halters, water troughs, etc.) or administering medications to camels were risk factors on multivariable analysis (32). These studies generally support the hypothesis that direct physical contact with camels is a risk factor for transmission, although cleaning equipment could also result in indirect transmission.

Potential Seasonality of Human Cases

Previous findings suggest that MERS-CoV circulation among camels peaks in late winter through early summer (22,26,48). Initial MERS-CoV infection in camels is thought to occur at ≈6 months of age, after the typical winter calving season. Investigations have demonstrated different seasonal peaks for MERS-CoV infection in camels: November–January (22) and December–April (48) in Saudi Arabia, January–June in Egypt (26). A seasonal peak has been suggested to result in a corresponding peak in zoonotic transmission, and an April–July peak has been supported by phylogenetic modeling (40). However, camels have been found to be rRT-PCR positive for MERS-CoV throughout the year (22). Further investigations are needed to demonstrate seasonality of MERS-CoV infection in camels and link these patterns to seasonal peaks of zoonotic transmission.

Zoonotic MERS-CoV Transmission outside the Arabian Peninsula

Despite evidence of circulating MERS-CoV in camels in North, East, and West Africa (49), limited evidence of human infection exists from Africa. In Kenya, no MERS-CoV antibodies were detected among 760 persons with occupational exposure to camels (50). In a separate study in Kenya, 2 seropositive adults who kept other livestock but not camels were identified among 1,122 persons who were predominantly without occupational exposure to camels (Appendix reference 51). In Nigeria, no MERS-CoV antibodies were found in 261 slaughterhouse workers with exposure to camels (Appendix reference 52). MERS-CoV sequences from camels in Africa phylogenetically cluster separately relative to camel and human MERS-CoV from the Arabian Peninsula, suggesting single or few introductions into Saudi Arabia and limited contact (19). Differences in virus growth have been shown between MERS-CoV strains isolated from West Africa and those isolated from the Middle East (19); relative to human and camel MERS-CoV from Saudi Arabia, virus isolates from Burkina Faso and Nigeria had lower virus replication competence in ex vivo cultures of human bronchus and lung. These findings may suggest regional variation in the potential for MERS-CoV replication in humans. Other factors contributing to the limited evidence of zoonotic MERS transmission in Africa may include differences between virus surveillance, human populations, camel populations, camel husbandry, and the type of human–camel interactions in these regions.

Prevention of Zoonotic Transmission

WHO recommends that anyone presumed at higher risk for severe illness (e.g., persons who are older, have diabetes, or are immunocompromised) should avoid contact with camels and raw camel products (Appendix reference 53). Although no evidence definitively links raw camel products with MERS-CoV infection, WHO presents these precautions in the context of considerable knowledge gaps surrounding MERS-CoV transmission. In addition, WHO recommends basic hygiene precautions for persons with occupational exposure to camels (Appendix reference 53).

Viral Shedding in Humans

MERS-CoV shedding in humans appears to differ from the pattern of viral shedding in camels. In humans, MERS-CoV RNA and live virus have been detected in both upper (nasopharyngeal and oropharyngeal swab) and lower (sputum, tracheal aspirate, and bronchoalveolar lavage fluid) respiratory tract samples, although RNA levels are often higher in the lower respiratory tract (Appendix reference 61). In humans, MERS-CoV is predominantly thought to infect the lower respiratory tract (Appendix reference 62), where the MERS-CoV dipeptidyl peptidase-4 (DPP4) receptor predominates, in contrast to camels, where DPP4 is found predominantly in the upper respiratory tract (Appendix reference 63). More severely ill patients typically have higher MERS-CoV RNA levels, as indicated by rRT-PCR cycle threshold (C_t) values and more prolonged viral shedding (Appendix reference 64). MERS-CoV RNA has been detected from the lower respiratory tract >1 month after illness onset (Appendix references 65,66), and live virus has been isolated up to 25 days after symptom onset (Appendix reference 67). RNA detection is prolonged in the respiratory tract of patients with diabetes mellitus, even when adjusting for illness severity (Appendix reference 66). Among mildly ill patients, who might typically be isolated at home, viral RNA levels in the upper respiratory tract have been detected for several weeks (Appendix references 68,69). Infectious virus has been isolated from the upper respiratory tract of a patient with mild symptoms (Appendix reference 68), suggesting a potential for transmission among less severely ill patients. However, there is no definitive evidence of transmission from asymptomatic cases, and epidemiologic evidence suggests that transmission from mildly symptomatic or asymptomatic cases does not readily occur (Appendix reference 70).

In humans, MERS-CoV RNA has been detected outside of the respiratory tract (Appendix references 66,71,72). Viral RNA has been detected in the whole blood or serum of mildly or severely ill patients (Appendix references 66,72) and in the urine of patients who subsequently died (Appendix reference 66), although virus isolation attempts on urine samples (Appendix reference 66) and serum (Appendix reference 71) have been unsuccessful. MERS-CoV RNA has also been detected from the stool of mildly and severely ill patients (Appendix reference 66). Subgenomic MERS-CoV RNA, an intermediate in the virus replication cycle, has been detected in stool, suggesting that MERS-CoV might replicate in the gastrointestinal tract (Appendix reference 73); however, it is not clear if this contributes to pathogenesis or transmission.

Reproduction Number and Attack Rates

The number of secondary cases resulting from a single initial case (reproduction number, R0) (Appendix reference 74) ranges widely for MERS-CoV, e.g., from 8.1 in the South Korea outbreak, compared with an overall R0 of 0.45 in Saudi Arabia (Appendix reference 74). Superspreading events, which generally describe a single MERS-CoV case epidemiologically linked to >5 subsequent cases, have been frequently described, particularly in healthcare-associated outbreaks (Appendix references 55,56). R0 estimates, however, can vary depending on numerous biologic, sociobehavioral, and environmental factors, and must be interpreted with caution (Appendix reference 75). Most studies estimating R0 across multiple areas, or at the end stage of an outbreak, result in estimates of R0<1, consistent with the knowledge that the virus does not continue to circulate in humans and that outbreaks are eventually contained. A wide range in published attack rates (the proportion of exposed persons who are infected) has also been reported (Appendix reference 74).

Transmission in Healthcare Facilities

Multiple studies have examined risk factors for MERS-CoV transmission in healthcare facilities. Higher viral loads (rRT-PCR C_t values) in respiratory tract samples have been linked to transmission risk (Appendix reference 76). Kim et al. described heterogeneity of transmission in South Korea in 2015, where 22 of 186 cases were associated with further transmission of MERS-CoV and 5 superspreading events accounted for 150 of 186 cases (Appendix reference 54). On multivariable analysis, transmission was associated with lower C_t values (indicating higher viral RNA load) and preisolation hospitalization or emergency department visits. Superspreading events were associated with a higher number of preisolation contacts, increased preisolation emergency room visits, and visiting multiple healthcare providers.

Alraddadi et al. studied risk factors for MERS-CoV infection in 20 healthcare workers in Saudi Arabia using serologic testing (Appendix reference 77) and found that N95 respirator use among healthcare workers decreased the risk for seropositivity. Conversely, wearing a medical mask (as opposed to not wearing a medical mask) increased the risk for seropositivity, but this finding was observed in a small number of persons and was strongly correlated with not wearing an N95 respirator. All 20 healthcare workers had been in the same room or automobile or within 2 m of a MERS patient. This study provided evidence to suggest that aerosol transmission of MERS-CoV may be possible at close range, as seen with other respiratory viruses (e.g., influenza) spread primarily by droplet or contact transmission, particularly during aerosol-generating procedures. Having participated in infection control training specific to MERS-CoV was associated with a decreased risk for seropositivity; in healthcare workers in South Korea, a lack of personal protective equipment (PPE) use was more likely in MERS-CoV–infected healthcare workers than among exposed uninfected healthcare workers (Appendix reference 78).

Studies have shown infection among persons without close and prolonged exposure to MERS case-patients during healthcare-associated outbreaks. In Jeddah during 2014, a total of 53 healthcare-associated cases were reported among hospitalized patients, of whom only 5 had documented presence in the same room as a MERS case-patient (Appendix reference 79). Among the remaining healthcare-associated cases, many shared common treatment locations (e.g., dialysis unit) but denied being in the same room as a MERS case-patient. Similar observations were documented during a multihospital outbreak in Jordan in 2015, and anecdotal evidence suggested a potential role for fomite transmission associated with a common imaging table and portable echocardiogram machine (Appendix reference 80).

MERS-CoV has been cultured from environmental objects, such as bed sheets, bedrails, intravenous fluid hangers, and radiograph devices, suggesting the potential for environmental transmission (Appendix reference 67). Viral RNA has also been identified in air samples from the hospital rooms of MERS patients (Appendix reference 81). MERS-CoV has also been shown to be relatively stable in the environment under various conditions (Appendix reference 82), supporting the possibility of fomite transmission, although definitive epidemiologic evidence implicating fomite or aerosol transmission is lacking.

Prevention of Healthcare-Associated Transmission

Studies have described delays in case recognition and establishment of infection control precautions as factors in healthcare-associated transmission (Appendix references 54,56,79). Triage practices that result in rapid isolation of suspected MERS case-patients and application of transmission-based precautions can decrease opportunities for early transmission. However, implementing triage procedures to quickly identify potential MERS cases in areas with local MERS-CoV transmission (e.g., Arabian Peninsula) is challenging because signs and symptoms are often nonspecific (Appendix reference 83). In addition, complications or exacerbations of concurrent conditions, including chronic kidney or heart disease, can manifest with acute or worsening respiratory symptoms that delay suspicion for MERS. Patient crowding has been associated with transmission in healthcare facilities, particularly in emergency departments (Appendix references 54,79). In multiple outbreaks, inconsistent PPE use has been reported as contributing to transmission (Appendix references 56,58), and transmission to healthcare personnel despite reported use of appropriate PPE (Appendix references 56,78) suggests that improper PPE use may contribute to transmission. Transmission risk may be particularly high during aerosol-generating procedures, in which large quantities of virus might be aerosolized.

Household Transmission

Human-to-human transmission of MERS-CoV has been reported among household contacts. Drosten et al. described 12 probable cases among 280 household contacts of 26 index case-patients (13). Arwady et al. investigated MERS-CoV infections in an extended family of 79 relatives, of whom 19 (24%) tested positive for MERS-CoV by rRT-PCR or serology (Appendix reference 84). Risk factors for acquisition included sleeping in an index case-patient’s room and touching their secretions. A study of MERS-CoV infection in a group of expatriate women housed in a dormitory in Riyadh, Saudi Arabia, showed an overall infection attack rate of 2.7% (Appendix reference 85). Risk factors for infection include direct contact with a confirmed case-patient and sharing a room with a confirmed case-patient; a protective factor was having an air conditioner in the bedroom. However, transmission among household contacts is variable; Hosani et al. found that none of 105 household contacts of 34 MERS-CoV case-patients showed antibodies to MERS-CoV (Appendix reference 70). Among those 34 patients, 31 were asymptomatic or mildly symptomatic, suggesting a lower risk for transmission among this group.

Viral Factors Affecting Human-to-Human Transmission

No evidence has been reported that mutations or recombinations in MERS-CoV have led to changes in human-to-human transmission. Recombination has been documented among coronaviruses (Appendix reference 86) and has been linked to increasing pathogenicity in strains of other animal RNA viruses (Appendix reference 87). Circulation of recombinant MERS-CoV has beenoc described in Saudi Arabia in camels (48) and humans (Appendix references 88,89) but no substantial change in human epidemiology was seen with this circulating variant (Appendix reference 89). Deletion variants of MERS-CoV were identified in humans in Jordan (Appendix reference 57), also without notable changes in epidemiology (Appendix reference 80).