Rethinking smallpox vaccination

February 2005

For most of us with responsibilities or interests in hospital epidemiology, the topic of smallpox vaccination — such a problematic topic only three years ago — has quietly come to an unofficial conclusion or at least quietly receded into the background. In truth, our problems have moved on, with the arrival and subsequent disappearance of severe acute respiratory syndrome (SARS), and today, the ever-increasing threat of pandemic avian influenza.

At least some members of the infectious disease community, however, have been troubled by the apparent disregard of most of us to this bioterror threat and the fact that most of us, myself included, seemed relieved when CDC’s smallpox immunization campaign for public health and health care personnel was quietly put on the back burner.

A report in the Dec. 1, 2004, issue of Clinical Infectious Diseases, entitled “Rethinking smallpox,” argues that we may have been overly hasty in dismissing the smallpox threat (39:1668-1673). It was written by Drs. Martin Weiss, Peter Weiss and Phyllis Guze, from several of the teaching hospitals of the University of California, Los Angeles, School of Medicine and the Veterans Administration. It is useful at this juncture to examine the argument made by these authors and reconsider the assumptions on which our response to the smallpox immunization program has been based.

Transmission routes

The first assumptions to be examined relate to how smallpox is spread. Although we are all aware that smallpox has occasionally been transmitted by the airborne route, we continue to believe, on the basis of evidence gathered from previous epidemiologic studies of natural smallpox, that most spread occurs by respiratory droplets, requiring relatively close and intimate contact, as might occur within a family unit. Thus, one index case has generally resulted in from one to three secondary cases. Thanks to its long incubation period (range, 7-21 days, with a mean of 12-14 days), spread takes place at a relatively slow pace, thus allowing ample time for intervention with vaccine, following the concept of “ring vaccination.” Control could be expected within two or three generations of cases.

We do know, however, that smallpox virus can be “weaponized” in an aerosol form, as was done in the Soviet Union previously. It is not clear how long this aerosol remains viable and, thus, how long the period of infectivity may be after an aerosol release. Undoubtedly, it is related in large part to the degree and intensity of ultraviolet (UV) light exposure. Previous studies of vaccinia virus, however, have demonstrated that aerosolized virus shielded from UV light may remain viable for 24 hours. A careful and competent release in a large enclosed area could possibly infect thousands of people and, if in a setting such as an international airport, could result in spread throughout the world in a very short time.

Another unknown in the debate about the contagiousness of smallpox is the possible effect of prior vaccination in limiting spread. The authors point out that few of the epidemiologic studies that described the spread of smallpox included information on prior vaccination of the exposed population. Could the prevailing notion that smallpox is not a highly contagious disease be in part a result of prior vaccination of some of the population in these studies?

Postexposure vaccination

One further assumption that most of us have made is that “postexposure” vaccination, if carried out within the first four to five days of exposure, would be quite effective and reduce the clinical case rate by as much as 50%. Furthermore, postexposure vaccination for people previously vaccinated would further modulate any disease and assure very low mortality. Those are large assumptions that overlook the opinions of some experts that postexposure vaccination is “at best of limited effectiveness.”

One further flaw in relying on postexposure vaccination is the obvious problem of how long after exposure it would take to establish the diagnosis. After a covert release, it could easily take three weeks or more for cases to appear, be recognized and be confirmed as smallpox, and for investigators to begin an epidemiologic study. By that time, a second generation of cases will already be incubating. In their mathematical model, Bozette and his colleagues estimated that over 50,000 deaths might well result from a “high-impact airport attack,” even with an aggressive postexposure vaccination program (N Eng J Med. 2003:348;416-425).

Another reason for foot-dragging on the smallpox immunization program was concern – justifiable concern – about the adverse reactions to smallpox vaccine. The recently published experience from the Department of Defense provides some reassurance on those issues. Among more than 600,000 military vaccinees, there were no cases of progressive vaccinia or eczema vaccinatum reported, and only one case of encephalitis, in a person who subsequently recovered. Viral culture and polymerase chain reaction (PCR) on that patient failed to reveal evidence of vaccinia virus as the etiology. There were 50 cases of contact transfer of vaccinia virus, primarily in spouses or intimate contacts. This lower-than-expected number of serious adverse reactions may have been due in part to more careful screening of potential vaccinees and covering of vaccination sites, which in earlier years had generally been left uncovered.

The major surprise in the vaccination program was the unexpected apparent increase in cases of myopericarditis; 83 cases occurred, and there was one death. Among the 64 cases that were followed up, all returned to normal cardiac function by electrocardiogram, echocardiogram and functional testing. However, in the much smaller civilian program – mostly health care and public health personnel – five vaccinees experienced myocardial infarctions, whereas, only two might have been expected. Nonetheless, the major lesson that emerged was reassuring as to the safety of smallpox vaccination, using the available calf lymph vaccine.

The authors go on in their report to discuss several nonvaccine approaches to preventing smallpox, including prophylactic drugs and the use of N-100 or N-95 respirators. Only two drugs are known that might prevent smallpox after exposure, cidofovir (Vistide, Gilead Sciences) and methisazone. The latter was used in India in the 1960s, but its efficacy remains somewhat controversial. Even more problematic is the fact that it is essentially an orphan drug, not available for use in the United States.

N-100 or N-95 respirators, correctly fitted, would likely be at least partially effective in postexposure prophylaxis. It is difficult for many people to breathe through a properly fitted N-100 respirator, and there would be enormous technical difficulties in doing proper fit testing. Thus, the use of these respirators in a postexposure situation would be limited to first responders, police and fire personnel and health care personnel.

None of the foregoing is meant to imply that the smallpox bioterror threat level has increased or that such an attack is inevitable. This report, however, certainly deserves review by hospital epidemiologists, hospital safety personnel and all others concerned with bioterrorism preparedness. We are making a number of assumptions in putting smallpox vaccination on the back burner, and it is well to review them from time to time to reassess our willingness to continue on that path.