| |||||||||||||||||||||||||||||||||||||||||||||||||
The issue of safety has received substantial attention, given that a predictable number of adverse events will occur among vaccine recipients. Furthermore, an extensive literature has established credible estimates of the complication rates.1,2,3,4 The risk of secondary transmission, however, is discussed much less, perhaps because relatively little is known. A report on vaccine-related deaths in the United States during the 1960s found that 12 of the 68 deaths occurred in unvaccinated persons exposed to recently vaccinated family members or friends, a finding that demonstrates the potential gravity of the problem.5 A recent reconsideration of the transmission rates during the 1960s concluded that spread is remarkably infrequent; this finding is quite reassuring for immunocompetent persons in the general population.4
However, the finding of infrequent transmission may not apply to hospitals, where large numbers of workers will be vaccinated, many for the first time. There is a large concentration of immunocompromised patients in hospitals, a situation distinctly unlike that in 1947, the last time a mass vaccination campaign was mounted in the United States. The prospect that a series of decisions might unwittingly introduce a live, transmissible, and potentially lethal virus into hospitals has dampened the enthusiasm of many for widespread vaccination.
In this article, I review the literature on the secondary transmission of vaccinia virus, including transmission in hospitals, among families, and in other circumstances. Many of the older articles would not pass modern peer review. However, the information they contain cannot be obtained elsewhere — a fact that makes them, however limited, of real value.
Nosocomial Spread
Nosocomial spread of vaccinia virus has been reported at least 12 times, from 1907 through 1975, and has resulted in 85 secondary cases6,7,8,9,10,11,12,13,14,15,16,17 (Table 1). Several additional outbreaks of Kaposi's varicelliform eruption unrelated to vaccinia virus have also been described. The cause of this diffuse skin eruption, whose name is often incorrectly used interchangeably with eczema vaccinatum (a known complication of vaccinia virus vaccination), was debated till the middle of the 20th century. Experts argued whether herpes simplex or vaccinia was the more likely cause; current thinking accepts both these and other viruses as etiologic agents. Studies that clearly demonstrated herpes simplex to be the cause of a patient's Kaposi's varicelliform eruption are therefore not discussed in this article.18,19,20
|
The incidence of secondary transmission of vaccinia virus is not easily calculated. In Glasgow, Scotland, after a three-year-old girl with eczema vaccinatum was hospitalized, all 11 children on her ward and 4 on an adjoining ward had generalized disease.7 Smaller series from Germany,6 Sweden,9 Philadelphia,10 and São Paulo, Brazil,15 demonstrated transmission to 16 of 27 susceptible patients (59 percent). In a single outbreak involving adults at a hospital in Brazil where vaccination was given to several patients with pemphigus foliaceus, 16 unvaccinated persons developed secondary disease.16 However, because many patients on the ward were vaccinated simultaneously, the opportunity for exposure increased. Furthermore, the denominator was not clearly defined but may have included 187 patients whose vaccine history was unknown, yielding an incidence of secondary vaccinia of about 9 percent.
A single French report examined the contribution of the duration of exposure to the risk of vaccinia virus transmission. An infant presented in the daytime with eczema vaccinatum, was hospitalized on an eczema ward, and by evening was transferred to isolation.13 Despite this, four secondary cases occurred in children on the eczema ward, though none had close contact with the index case.
The exact route of transmission is also uncertain. In the above study,13 all of the children were confined to cribs and were too ill to interact. In another, after hospitalization of the index patient, several cases of disease occurred in an adjoining ward.7 Although the affected children did not mix, they were cared for by the same professional staff. A sore throat developed in three treating nurses, one of whom had several "pustular bullae" on the forearms, but none were formally evaluated.7 In a carefully studied case of transmission from an adult with disseminated vaccinia to a woman with active mycosis fungoides in California, investigators remained uncertain how the virus moved from the isolation room to the woman, whose room was some 25 m (75 ft) away.17 They suggested that perhaps health care workers carried the virus or that the two patients occupied the same hall area for several hours, resulting in fomite-based spread. The studies also raise the possibility of aerosol transmission of vaccinia virus.7,8,13 Any of these potential methods of spread has substantial implications for infection-control teams that may be called on to isolate and care for a patient with eczema vaccinatum.
Yet another route of transmission was demonstrated by a unique outbreak in Italy, where vaccinia was spread by a contaminated urinary catheter.14 After her older brother received vaccine, a 13-month-old girl had initially undiagnosed genital lesions and dysuria resulting from vaccinia infection. At the hospital, she was catheterized, and the catheter was then placed in a pan of Citrosil solution for sterilization. Several other urinary catheters were soaking in the same pan. Within a five-week period, there were 23 secondary cases with vulvar–urethral vaccinia; each of the patients had been catheterized with one of the contaminated catheters. About half had high fevers, and some had gross hematuria. Virus was cultured from the urine of several children.
Two reports have clearly defined the epidemic curve of vaccinia virus infection. In the 1935 outbreak in Glasgow, all secondary cases occurred between 8 and 18 days after exposure.7 Examination of patients treated at an infectious-disease hospital in Brooklyn, New York, after the 1947 mass vaccination in New York City found an average incubation period of 10.6 days (range, 5 to 19).11
Nine of the 85 reported patients (11 percent) died. Worsening severity of disease with each generation of transmission was seen in one outbreak in Germany.8 In another report, from Scotland, those in whom disease developed later tended to have milder symptoms.7 Death was typically due to encephalitis or the development of secondary bacterial pneumonia. Treatment included antibacterial agents and, for several, vaccinia immune globulin.
Spread within Families
Numerous reports have described the spread of vaccinia virus within families. The majority are instances of single transmission, usually from a recently vaccinated child to an unvaccinated younger sibling.22,23,24,25 However, two or more secondary cases have been reported in at least eight reports of family outbreaks published from 1931 to 198126,27,28,29,30,31,32,33 (Table 2). Many of the reports describe severe, sometimes fatal eczema vaccinatum in the first family member with secondary disease and substantially milder local inoculation disease in the rest of the family. These latter infections might have been overlooked had medical attention not been sought for the severe case.
|
In many of the family outbreaks,27,28,29,30 sharing close quarters was a significant factor, suggesting the need for sustained, intimate contact to transmit vaccinia between intact hosts. In one outbreak, a bed was shared by three persons in whom disease developed, further supporting this notion.27 An unusual aspect of the family outbreaks of vaccinia was the apparent tendency for lesions to be present in similar anatomical areas in all secondary cases, including the mouth32 and the face.33
Other Transmission
Scattered reports detail other cases of secondary transmissions of vaccinia, exclusively by inadvertent inoculation.34,35,36,37,38,39,40 Eyelids, lips, nose, and vulva were most commonly reported.32 Humphrey found 70 cases of vulvar vaccinia in the literature,37 including the 24 catheter-related cases described above,14 many due to auto-inoculation and several from sexual transmission.31,34 The mucosa may be involved because vaccinia can penetrate more easily into this tissue than into skin. Alternatively, vaccinia may have a tropism for mucosal surfaces. This phenomenon may be important, since many currently hospitalized patients, such as those receiving chemotherapy, have substantial mucosal abnormalities and therefore may be at higher risk for acquisition of secondary vaccinia virus infection.
Occupational spread to the hands of those working with vaccinia virus vaccine has been described, and many workers have repeated local infection despite previous vaccination.41 A sustained outbreak occurred among 22 farm workers and 450 cows on a dairy farm in El Salvador.42 One of the workers had received vaccine and resumed milking cows before his lesions had resolved, thereby spreading the virus to cows and thence to coworkers, including the woman who washed the towels used by the milkers. In all 22 affected workers, lesions were confined to the hands and genitals.
Implications for Vaccination Policy
The rate of adverse events after vaccinia virus vaccination is being carefully scrutinized as a national vaccination policy is developed.1,2,3,4 Relatively little is known, however, about the risk of secondary transmission of this live virus in the hospital setting.4 A review of the literature indicates that nosocomial transmission does occur and that the outcome may be fatal in up to 11 percent of cases. Nosocomial outbreaks seem to require relatively minor contact with a source case, whereas spread in the home appears to occur only with sustained, intimate exposure, perhaps owing to immunologic and dermatologic differences among the persons exposed.
Information regarding secondary transmission is particularly important for health care facilities, which will need to vaccinate workers while ensuring patients' safety. The composition of hospitalized patients in the 21st century is dramatically different from that in the mid-20th century. Patients treated before the 1950s were very unlikely to be immunosuppressed: cancer chemotherapy was just beginning; transplantation had not yet been performed; the human immunodeficiency virus (HIV) was unheard of; and corticosteroid therapy had only recently been introduced.
Now, approximately 506,154 persons in the United States are known to be living with HIV43; 1.2 million new non-skin cancers are diagnosed annually44; 2.1 million persons have rheumatoid arthritis and receive therapy with corticosteroids or other immunosuppressive agents45; and more than 14 million have asthma, many of whom require intermittent steroid use.46 Thousands of solid-organ and bone marrow transplantations are performed each year and tens of thousands of transplant recipients are alive and still receiving immunosuppressive therapy. Atopic dermatitis is also more common, with prevalence among children ranging from 6.8 percent to 17.2 percent.47 Finally, there are tens of thousands of patients in intensive care units and newborn nurseries. Current expert opinion recommends that vaccination of such persons should be avoided.48 Vaccination can be avoided, but contact with a recent vaccinee probably cannot.
Of equal importance are the differences in the modern population of health care workers, some of whom are themselves immunocompromised. Previously, hospitals were staffed with workers who had received at least one vaccinia virus vaccine. Such persons were therefore unlikely to initiate or propagate an outbreak. In contrast, most current health care workers are susceptible to smallpox and vaccinia and so might play a dangerous supporting or even lead part in any nosocomial outbreak.
Both the rate and route of vaccinia transmission remain unknown. The incidence derived from the cited studies (9 to 59 percent) is certainly an overestimation of current risk, owing to erratic infection-control practices in past decades, differences in the virulence of the vaccinia virus used, and a substantial reporting bias. The current plan for an occlusive dressing at the vaccination site and other now-routine infection-control procedures, including hand hygiene and isolation for any patient with unexplained fever and rash, should effectively limit potential spread.49 Equally important is the need to ensure that the vaccine program develops slowly, with flexibility and ample time to make any necessary adjustments.
The actual route of transmission is not revealed by these outbreaks, but it may include several different paths.7,8,13,14,17 First, health care workers may carry virus on their clothes, on their hands, or even in the nasopharynx. Other evidence suggests transmission by fomites,17 and the widespread transmission from contaminated urinary catheters14 emphasizes the need for rigorous cleaning of any item that comes into contact with a recently vaccinated person. There is also the possibility of transmission of vaccinia virus by the aerosol route, since some secondary cases have occurred on the same hospital floor as a bedbound source patient.7,8
Other than those with underlying skin conditions, it is not known which patients are at high risk for secondary disease. Dozens of reports have described progressive vaccinia (also referred to as vaccinia necrosum and vaccinia gangrenosa) in immunocompromised patients, particularly those with hematologic neoplasms (especially chronic lymphocytic leukemia), hypogammaglobulinemia, or defects in cellular immunity.50,51,52,53,54,55 These infections, which are often fatal, may last for months and may respond poorly to frequent doses of vaccinia immune globulin.50 Progressive vaccinia in a newly vaccinated soldier with advanced, previously undiagnosed HIV infection has also been described.56 These studies demonstrate that vaccinia may be easily transmitted to hosts with severe dermatologic disorders, with substantial mortality in the absence of appropriate infection-control measures.
An additional important finding from these articles is the observation that secondary disease is manifested exclusively as eczema vaccinatum or contact inoculation. No secondary cases of the most severe complications — progressive vaccinia and postvaccination encephalitis (except in those with overwhelming eczema vaccinatum) — have been reported. Thus, the danger of nosocomial spread, though alarming, is mitigated by the limited range of clinical manifestations in secondary disease.
Because of the risk of secondary transmission of vaccinia, many hospitals remain uncomfortable with the recent recommendation against the provision of administrative leave for newly vaccinated health care workers.56 Also, the advisability of immunocompromised workers' remaining on the job while colleagues receive vaccine has not been determined. Until these controversies are settled, hospitals must be certain that the rush to vaccinate health care workers does not result in a self-inflicted epidemic — not of smallpox, but of infection with the live, potentially lethal virus, vaccinia.
Supported in part by a grant (K24 AI052239-01) from the National Institutes of Health.
I am indebted to Linda Han and Sara Tuttle for research assistance, and to Johan Herrlin, Roman Tuma, Gregoire Lauvau, Matthias Frank, and Svetolik Djurkovic for help in translating the articles cited.
Source Information
From the Infectious Disease Service, Memorial Sloan-Kettering Cancer Center, New York.
This article was published at www.nejm.org on December 19, 2002.
Address reprint requests to Dr. Sepkowitz at the Infectious Disease Service, Memorial Sloan-Kettering Cancer Center, 1275 York Ave., New York, NY 10021, or at sepkowik{at}mskcc.org.
References
| |||||||||||||||||||||||||||||||||||||||||||||||||
Related Letters:
Smallpox and Smallpox Vaccination
Letai A. G., Snyder K. M., Fett J. D., Worthington M. G., Ross J. J., Neff J. M., Lane J. M., Fulginiti V. A., Milton D. K., Bozzette S. A., Boer R., Mack T., Sepkowitz K. A.
Extract |
Full Text |
PDF
N Engl J Med 2003;
348:1920-1925, May 8, 2003.
Correspondence
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2008 Massachusetts Medical Society. All rights reserved. |
Previous
