Skip Navigation

Progressive Vaccinia

  1. Mike Bray and
  2. Mary E. Wright
  1. Biodefense Clinical Research Branch, Office of Clinical Research, Office of the Director, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, Maryland
  1. Reprints or correspondence: Dr. Mike Bray, Biodefense Clinical Research Branch, OCR/OD/NIAID/NIH, 6700A Rockledge Dr., Rm. 5132, Bethesda, MD 20892 (mbray{at}niaid.nih.gov).

Abstract

The resumption of smallpox vaccination for health care workers and other first responders has raised concern about the occurrence of complications in people with immunodeficiency disorders, including those infected with human immunodeficiency virus. During the era of universal vaccination, roughly 1 person per million vaccinees in the general population developed progressive vaccinia, which is characterized by the relentless outward spread of infection from the vaccination site and eventual dissemination to other areas on the body. Review of 56 cases reported in the English-language medical literature from 1893 through 1997 indicates that the condition occurred only in persons with severe cell-mediated immunodeficiency. Progressive vaccinia was found to be lethal in infants who completely lacked cellular immune function, but infection resolved in many adults with acquired immunodeficiency. Almost all cases were treated with vaccinia immune globulin, but its efficacy has never been tested in a placebo-controlled trial. Further research is needed to develop effective forms of therapy.

The cessation of routine vaccination against smallpox that followed the successful global eradication campaign has rendered the majority of the human population susceptible to severe or fatal infection should the disease reappear. Concern that the agent of smallpox (variola virus) might be used as a bioterrorist weapon has prompted the US government to embark on a crash program of vaccine production and to prepare plans for large-scale vaccination in the event of an outbreak of smallpox [13].

Before mandatory vaccination was discontinued in the United States in 1972, one of its rarest but most feared complications, progressive vaccinia, occurred at a rate of roughly 1 case per million vaccinees [49]. The condition occurred in infants and older adults with defective cell-mediated immunity, and it was characterized by the relentless enlargement of the primary vaccination lesion (figure 1A and 1B) [1012]. Since that time, the spread of HIV infection and the increased number of people receiving immunosuppressive medications raises the concern that a vaccination campaign will result in an unprecedented number of cases of progressive vaccinia.

Figure 1

A, Progressive vaccinia in an infant with severe combined immunodeficiency who was vaccinated at age 3 months (case 4 from [10]). The infant's condition improved transiently after receipt of treatment with vaccinia immune globulin, methisazone, and exchange transfusion, but he died 2 months after vaccination. Note the complete absence of inflammatory response around the lesions. B, Progressive vaccinia in 71-year-old man who had lymphosarcoma treated aggressively with antimetabolites and who was vaccinated in preparation for overseas travel. The vaccination site is replaced by a necrotic ulcer, and the surrounding and distal skin contains multiple vesicles (V. Fulginiti, personal communication). (Both images copyright 2002 © Logical Images, Inc. Both photographs are printed with the permission of Logical Images, Inc., and may not be reproduced without the written permission of Logical Images, Inc. All rights reserved.)

Three decades after the end of routine smallpox vaccination, physicians must once again learn to assess the normal response to the inoculation of vaccinia virus and learn to recognize and to treat its associated complications. This article reviews the pathophysiology of vaccination; summarizes the underlying illnesses in, the treatments of, and the outcomes for 56 patients with progressive vaccinia described in the English-language medical literature during 1893–1997; and suggests potential investigative therapy.

Vaccination and its Complications

Inoculation against variola is performed using vaccinia, a related orthopoxvirus. The vaccine currently available in the United States, Dryvax (Wyeth), is derived from the New York City Board of Health (NYCBOH) strain used before 1972. It is expected to produce the same reaction and complications observed when it was used previously [13]. Nearly all side effects can be predicted from the unusual nature of smallpox vaccination, which essentially employs a small circle of skin as a “culture plate” in which to grow vaccinia virus.

Pathophysiology of normal vaccination. Vaccinia is introduced into the skin by means of multiple punctures of a bifurcated needle. The virus replicates in the basal layer and disseminates from cell to cell, causing necrosis and the formation of fluid-filled vesicles. The initial spread of virus is slowed by innate antiviral mechanisms, and, by the second week, the cell-mediated immune response begins to eliminate infected cells. Neutrophils, macrophages, and lymphocytes infiltrate the inoculation site, forming a confluent pustule and releasing cytokines and chemokines that cause hyperemia and edema in surrounding tissues. Clinical manifestations may include complaints of malaise and other mild constitutional symptoms, fever, and tender, enlarged axillary lymph nodes. Some vaccinees develop additional local “satellite” pustules that resolve along with the primary lesion.

Although some vaccinia viruses commonly disseminate through the bloodstream, the NYCBOH strain reportedly causes only limited viremia in a small percentage of recipients during the period of pustule formation [13, 14]. The inflammatory process reaches its peak by days 10–12 after vaccination and begins to resolve by day 14, with shedding of the scab by day 21. This sequence of events, which simulates the development of a smallpox pock, is known as a “take” reaction. A successful take is required for the development of antivaccinia antibody and cell-mediated responses [1518].

Complications. Vaccination complications and their incidences reported in 2 large surveys are presented in table 1. With the exception of postvaccinial encephalitis, all complications are caused by the escape of virus from the inoculation site. The most common side effect results from the accidental transfer of virus to other areas on the body of the vaccinee or his close contacts [19]. Less often, internal viral dissemination produces a condition known as “generalized vaccinia,” in which additional pocks appear at scattered sites on the skin during the first 2 weeks after vaccination. The number of pocks tends to be small, and they usually resolve quickly without scarring, suggesting that the number and size of lesions are restricted by development of humoral and cell-mediated responses. The external and internal spread of virus may have serious consequences in persons with eczema and other forms of atopic dermatitis; in these persons, defects of innate immunity and a high level of Th2 cell activity render the skin unusually permissive to the initiation and rapid spread of vaccinia infection (known as“eczema vaccinatum”) [20, 21].

Table 1

Complications of smallpox vaccination.

Progressive Vaccinia

Progressive vaccinia, also known as “vaccinia necrosum,” “vaccinia gangrenosa,” and “disseminated vaccinia,” is rarer than the complications mentioned above [4, 5]. It is characterized by the inexorable enlargement of the primary vaccination lesion and, in most cases, the eventual appearance of similar foci of infection on other areas of the body (figure 1) [1012]. The slow but relentless spread of virus deep into the tissues causes necrosis and osteomyelitis, often with bacterial superinfection, leading to death weeks or months after vaccination. Although early studies attributed progressive vaccinia to an inability to produce antivaccinia antibodies, by the mid-1960s, it had become clear that progressive vaccinia resulted from defective cellular immune function [10].

The first recognizable case was reported in 1893 in a 3-month-old infant [22]. Although several patients with “vaccinia gangrenosa” were described during the next 50 years, a review of these cases indicates that the patients had either severe generalized vaccinia or eczema vaccinatum. The vaccination of >6 million residents of New York City in 1947 resulted in 1 clearly described but undiagnosed case of progressive vaccinia [23]. The condition was not clearly differentiated from other vaccination complications until the early 1950s, and most reported cases date from the subsequent 2 decades. Almost all cases occurred at the ends of the age spectrum: in infants with congenital immune defects (table 2) [10, 2237] and in adults aged >50 years with acquired immunodeficiency (table 3) [3866]. In contrast to other complications, the incidence of progressive vaccinia did not decrease among secondary vaccinees (table 1), because most adults who developed the condition had been vaccinated as children. With a single exception [52], all reported cases of progressive vaccinia resulted from vaccination rather than contact infection.

Table 2

Case reports of progressive vaccinia in children.

Table 3

Case reports of progressive vaccinia in adults.

Diagnosis. The essential diagnostic feature of progressive vaccinia is the continuing enlargement of the primary vaccination lesion (figure 1A and 1B). The occurrence of additional “metastatic” lesions is not required for diagnosis. The inappropriately weak or absent inflammatory response to the continuing spread of virus results in additional clinical features by which the condition can be recognized:

•The initial vesicles form against a background of normal skin, without surrounding erythema or edema.

•The vesicles fail to transform to pustules by the end of the first week after vaccination.

•The lesion is continuing to enlarge at the end of the second week, eventually forming a shallow ulcer with central necrosis and raised edges containing vesicles.

•Little or no axillary lymphadenopathy is observed.

•When new lesions appear, they develop in the same manner as the lesion at the primary vaccination site.

Equivocal cases can be confirmed by culturing vaccinia virus from the skin lesions. Although not required for diagnosis, a biopsy will show a paucity or absence of inflammatory cells. In infants who completely lack a cellular immune response, the enlarging ulcers are flat and painless (figure 1A), but, in adults, the tissues surrounding a necrotic vaccinia lesion are often swollen and tender, particularly if bacterial superinfection has occurred (figure 1B). “Metastatic” lesions tend to develop one by one on any area of the skin or mucous membranes over the course of weeks to months.

It is essential to differentiate progressive vaccinia from more common and benign causes of an abnormally large vaccination lesion, such as an unusually strong take reaction or bacterial superinfection, both of which are accompanied by an increased (rather than diminished) inflammatory response [13]. The slow pace of development of progressive vaccinia helps to distinguish it from generalized vaccinia, in which numerous additional pocks appear along with the primary lesion, and from eczema vaccinatum, in which multiple areas of spreading infection appear soon after vaccination or close contact with a recent vaccinee.

Occurrence in children. In the era when vaccination was routinely performed during the first few months of life, most cases of progressive vaccinia occurred in young infants. In a survey of 5 million vaccinations performed in Great Britain during 1951–60, all 8 cases of progressive vaccinia were observed in infants aged <1 year; 7 cases were fatal [67]. Some infants had a history of frequent infection, but, in many instances, vaccination provided the first unequivocal evidence of an immune-system defect. Of the 23 children described in table 2, only 4 survived infection; 3 of the survivors were aged >1 year and appeared to have a partial rather than complete immune-system defect. In 1966, the US government recommended that vaccination be deferred to the second year of life, in part to allow time for immunodeficient infants to be identified and deferred from vaccination [68, 69].

During the 1950s, progressive vaccinia in infants was usually attributed to a defect in γ-globulin production (table 2), and almost all cases were treated with repeated inoculations of vaccinia immune globulin (VIG). In most instances, skin testing also revealed a lack of delayed-type hypersensitivity to common antigens, providing evidence of a defect in cellular immunity. It was eventually realized that progressive vaccinia most often occurred in patients with severe, combined immunodeficiency [10]. The critical role of cell-mediated immunity in the healing of the vaccination lesion was supported by studies showing that children with agammaglobulinemia could be vaccinated safely [70].

Occurrence in adults. The other group at risk of developing progressive vaccinia consisted of adults with chronic lymphocytic leukemia, lymphoma, or connective tissue disorders, almost all of whom were receiving corticosteroids and/or antimetabolite therapy (table 3). The physician who administered the vaccine was usually aware of his patient's underlying illness but did not recognize that it was a contraindication for vaccination. In 10 of the reported cases, the vaccine was given as treatment for recurrent herpetic ulcers. Such therapy was first recommended in the 1920s as a result of the mistaken idea that herpes and vaccinia viruses were closely related [71]. The practice was not officially denounced until 1982 [61, 72, 73]. In retrospect, the herpes lesions can be seen to represent an opportunistic infection signaling the presence of immunodeficiency.

Overall, it is likely that the published cases represent only the most severe manifestations of progressive vaccinia. In 2 instances, the condition resolved without treatment [45, 50], and it is likely that many other chronically ill adults with less severe degrees of immunodeficiency experienced delayed healing of their vaccination-site lesions that went unreported.

Occurrence with HIV infection. Little is known about the risk of smallpox vaccination for HIV-infected people. Routine vaccination of the civilian population had been discontinued by the time HIV infection emerged in the early 1980s, but the US armed forces did not begin to phase out their program until 1984, ending it in 1989. This resulted in the only known vaccination complication in an HIV-infected individual, a soldier vaccinated in May 1984 [63]. Although apparently healthy at the time of vaccination, he became ill with cryptococcal meningitis 2 weeks later, and 4 weeks after vaccination, a large number of vaccinia lesions appeared on the lower half of his body. Studies revealed a T-helper cell count of <25 cells/µL, which is indicative of severe immunodeficiency. The patient was treated with 12 weekly doses of VIG, and his lesions resolved. He subsequently showed an increase in his T-helper cell count to 300 cells/µL, suggesting that his recovery may have resulted from improved cellular immune function.

Retrospective analysis indicates that many more HIV-infected enlisted personnel in the armed forces were vaccinated. When the armed forces began screening in late 1985, in an effort to prevent the enlistment of HIV-infected individuals, a prevalence of seropositivity of 1.5 cases per thousand was found among applicants for military service [7476]. Assuming a prevalence of 1–2 cases per thousand persons among the 909,541 people who joined the armed forces in 1983–1985 [77], it is clear that at least several hundred HIV-infected enlistees were vaccinated during those years without experiencing serious adverse effects, suggesting that the vaccine poses little risk to persons during the first few years of HIV infection. This contrasts with the outcome in severely immunocompromised individuals, such as the aforementioned soldier and 2 patients with CD4 cell counts of <50 cells/µL, who developed progressively expanding lesions after inoculation with a preparation derived from vaccinia-infected cells [64].

Therapy

After 1960, almost all reported cases of progressive vaccinia were treated with VIG or with VIG plus the antiviral drug methisazone, whereas a few persons received methisazone alone (tables 2 and 3). No placebo-controlled trials of either agent have been performed. A few patients were treated with human fibroblast interferon or with local injections of adenine arabinoside or idoxuridine, without apparent response. Attempts were also made to bolster cell-mediated immunity by transfusing immune cells or whole blood from previously vaccinated donors, but, in infants, this often resulted in a graft-versus-host reaction [10]. In many cases in adults, improvement may have resulted from the tapering or discontinuation of immunosuppressive therapies rather than from any therapeutic intervention.

VIG. VIG was developed at a time when progressive vaccinia was believed to arise from a defective antibody response to vaccination [78]. It was prepared as a ∼20-fold concentrate of γ-globulin from pooled plasma samples obtained from recently vaccinated military recruits. A single inoculation of 0.6–1 mL/kg of body weight was reported to produce prompt remission of generalized vaccinia and rapid improvement in many cases of eczema vaccinatum [11, 12, 27, 47, 78, 79]. By contrast, patients with progressive vaccinia often received multiple injections over the course of weeks to months. VIG was used to treat almost all published cases of progressive vaccinia after 1955; the patients' outcomes appear to be representative of the general experience. Thus, all 9 patients with progressive vaccinia in the United States in 1963 were treated with VIG and survived, whereas 10 of 11 patients received VIG treatment in 1968, 6 of whom recovered [4, 6]. Although positive outcomes were usually attributed to VIG therapy, no placebo-controlled trials were performed. In many cases, improvement in a patient's cell-mediated immune function may actually have played the critical role in determining survival.

Methisazone. During the 1950s, a number of thiosemicarbazone derivatives were found to inhibit the replication of vaccinia virus, and one, methisazone, became the first antiviral drug introduced into clinical use [80]. Methisazone was fairly toxic when administered systemically; nevertheless, it was quickly applied to the therapy of progressive vaccinia, almost invariably in combination with VIG [6, 10, 42, 46, 48, 54, 57]. All 4 patients treated with methisazone alone were aged >1 year; 2 of them survived [6, 10, 46]. The lack of controlled trials makes it difficult to judge whether treatment played any role in recovery. The drug is no longer in use.

Ribavirin. The broad-spectrum antiviral drug ribavirin was used to treat 1 case of progressive vaccinia, which occurred in a patient with leukemia who had inadvertently been inoculated with vaccinia-infected material (figure 1) [65, 66]. Some new lesions developed while the patient was receiving ribavirin alone, but none appeared after VIG was added to the treatment regimen 12 days later.

Cidofovir. A compound with greater antivaccinia activity than ribavirin—but more potential for systemic toxicity—is the phosphonate derivative of cytosine, cidofovir (Vistide; Gilead Sciences), which is licensed for treatment of cytomegalovirus infection [81]. Cidofovir is highly protective against rapidly lethal orthopoxvirus infection in laboratory animals [82], suggesting that it would be beneficial for treatment of comparatively indolent vaccinia infections in humans. The drug must be administered intravenously, accompanied by probenecid and hydration to avoid renal toxicity [83]. Cidofovir's long intracellular half-life permits infrequent dosing. Modified forms that can be taken by mouth are under development [84].

The combination of cidofovir and VIG eliminated vaccinia infection in immunodeficient mice [85], and it is likely that any investigational use of cidofovir in humans would employ both agents, given either sequentially or together. Cidofovir is also effective when applied topically to 2 poxviral infections of the skin, molluscum contagiosum and orf [86, 87]. It will be important to determine whether similar treatment of a vaccinia ulcer would eliminate the virus.

Immunomodulators. Another approach worthy of study is the use of cytokines or other immunomodulators to stimulate a compromised host's residual cell-mediated immune function. Studies involving mice have shown that the Th1 cytokines IL-2 and IFN-γ stimulate orthopoxvirus clearance, whereas IL-4 and other Th2 cytokines facilitate viral spread [21, 8890]. Thus, athymic nude mice that are normally incapable of controlling vaccinia infection are able to eliminate recombinant viruses that encode IL-2 or IFN-γ [89, 90]. The replication of the IL-2-encoding virus is markedly restricted in nonhuman primates [91]. These findings suggest that local or systemic treatment with immunomodulators that potentiate a Th1 response could help suppress vaccinia infection in immunodeficient patients.

Conclusions

Vaccination against smallpox requires the host to recognize and eliminate vaccinia virus infection through the development of a cell-mediated immune response. The failure of that process leads to progressive vaccinia. This review of published cases indicates that, in the past, the condition was not identical in all immunocompromised individuals. In particular, progressive vaccinia had a strikingly different prognosis in infants and adults. Infants congenitally lacking cell-mediated immune function almost invariably died, whereas most adults with acquired immunodeficiency disorders were either able to eliminate their vaccinia infections or showed partial resolution before dying of their underlying disease.

It is unlikely that individuals who are known to be immunocompromised will be at significant risk for progressive vaccinia in future, because they are excluded from US vaccination plans. However, concern remains that some apparently healthy people with undiagnosed immunosuppressive conditions, such as HIV infection, may inadvertently be inoculated. Data from the military vaccination program suggest that the risk of progressive vaccinia is low in otherwise healthy HIV-infected individuals, but this has not been confirmed by clinical testing.

Almost all patients who had progressive vaccinia were treated with VIG, but its efficacy has never been proven in a controlled, prospective study. This and other questions regarding the pathogenesis of and therapy for progressive vaccinia merit further investigation in animal models. In particular, the efficacy of VIG and of systemic or topical cidofovir should be studied in nonhuman primate models of retroviral infection that simulate HIV disease in humans. It is hoped that cases of progressive vaccinia will be rare in any future vaccination campaign, but it is essential to identify effective forms of therapy.

  • Received November 12, 2002.
  • Accepted December 16, 2002.

References

| Table of Contents