Neurologic Adverse Events Associated With Smallpox Vaccination in the United States, 2002-2004 FREE

The United States reinstated vaccination against smallpox among military personnel in 2002 and among selected civilian groups in January 2003. Routine smallpox vaccination had been suspended among US civilians in 1973 and among military personnel in the late 1980s. Rare adverse events causally associated with smallpox vaccination include cutaneous,^1- 4 ocular,^1- 5 and neurologic^1- 4,6- 7 syndromes. Acute neurologic disease occurs after many vaccinations, including smallpox vaccination.^6- 7 Central nervous system (CNS) and peripheral nervous system complications of smallpox vaccination are rare but severe vaccine-associated adverse events. The most common CNS complication after smallpox vaccination is encephalitis (postvaccinial encephalomyelitis [PVE]).⁶

Other neurologic events have been associated temporally with smallpox vaccination. A recent assessment substantiates the frequency of postvaccinial headache.⁸ Other neurologic syndromes infrequently historically reported in temporal association with smallpox vaccination include Guillain-Barré syndrome (GBS),⁹ acute cranial neuropathies,¹⁰ poliomyelitis-like syndrome,¹¹ Bell palsy,¹² and transverse myelitis.^10,13

Between December 16, 2002, and March 11, 2004, we assessed reports of PVE and other neurologic adverse events among US Department of Health and Human Services (DHHS) and Department of Defense (DoD) smallpox vaccinees to characterize them clinically and assess their frequency. All vaccinations were done with the New York City Board of Health strain of vaccinia (Dryvax, Wyeth Pharmaceuticals, Pearl River, NY).

The Smallpox Vaccine Adverse Events Monitoring and Response Activities (SVAEMRA) identified adverse events after smallpox vaccination through clinical consultation and reporting to the Vaccine Adverse Event Reporting System (VAERS).

A team of Centers for Disease Control and Prevention (CDC) SVAEMRA clinicians with expertise in smallpox vaccine–related safety responded to calls from health professionals identifying a suspected adverse event in DHHS vaccinees. The SVAEMRA clinicians completed case report forms, collecting information on vaccination, demographics, clinical events, and diagnostic studies, and followed up with the reporting clinician until a diagnosis was reached. A similar monitoring and consultation system was instituted by the DoD; this information was reported to VAERS and shared between the DoD and CDC.^14- 16 Events reported between December 16, 2002, and March 11, 2004, that were primarily neurologic, including those that met the case definition for encephalomyelitis adverse events (Box), were reviewed in detail by the lead author (J.J.S.), a board-certified neurologist and epidemiologist. Case report forms, medical records, and interviews with reporting physicians were used to ascertain vaccinee status, characterize the adverse event, and assess its association with smallpox vaccination. We reviewed all cases and conclusions with the Smallpox Vaccine Safety Working Group (SVS WG), a joint expert working group of the CDC’s Advisory Committee on Immunization Practices and the DoD’s Armed Forces Epidemiological Board.

Case Definition for Acute Encephalitis

A. A confirmed case of encephalitis is defined by demonstration of acute cerebral inflammation (± meninges) or demyelination by histopathology
B. A probable case of encephalitis is defined by the acute onset of:

   1. Encephalopathy (eg, depressed or altered level of consciousness, lethargy, or personality change lasting ≥24 hours)
AND
   2. Additional clinical evidence suggestive of cerebral inflammation, including 2 or more of the following:

      a. Fever (temperature ≥38°C) or hypothermia (temperature ≤35°C)

      b. Meningismus (ie, nuchal rigidity, photophobia/phonophobia)

      c. CSF pleocytosis (>5 white blood cells/mm³)

      d. Presence of focal neurologic deficit

      e. Electroencephalography findings consistent with encephalitis

      f. Neuroimaging findings on magnetic resonance imaging (MRI) consistent with acute inflammation (± meninges) or demyelination of the nervous system

      g. Seizures, either new onset or exacerbation of previously controlled seizures
AND
   3. No alternative (investigated) etiologies are found for presenting sign and symptoms

C. A suspected case of encephalitis is defined as the presence of the acute onset of

   1. Encephalopathy, as outlined for a probable case
AND
   2. One of the criteria listed for probable encephalitis as clinical evidence suggestive of cerebral inflammation.
AND
   3. No alternative (investigated) etiologies are found for presenting signs and symptoms

Case Definition for Acute Myelitis

A. A confirmed case of myelitis is defined by demonstration of acute spinal cord inflammation (± meninges) or demyelination by histopathology.
B. A probable case of myelitis is defined by the acute onset of:

   1. Myelopathy (development of sensory, motor, or autonomic dysfunction attributable to the spinal cord, including upper- and lower-motor neuron weakness, sensory level, bowel or bladder dysfunction)
AND
   2. Additional evidence suggestive of spinal cord inflammation, including 2 or more of the following:

      a. Fever (temperature ≥38ºC) or hypothermia (temperature ≤35°C)

      b. CSF pleocytosis (>5 white blood cells/mm³)

      c. Presence of focal neurologic deficit

      d. Electromyographic studies suggestive of central (spinal cord) dysfunction

      e. Neuroimaging findings on MRI demonstrating acute inflammation (± meninges) or demyelination of the spinal cord
AND
   3. No alternative (investigated) etiologies are found for presenting sign and symptoms

C. A suspected case of myelitis is defined as presence of the acute onset of:

   1. Myelopathy as outlined for a probable case
AND
   2. One of the criteria listed for probable myelitis, as evidence suggestive of spinal cord inflammation
AND
   3. No alternative (investigated) etiologies are found for presenting sign and symptoms

Cases fulfilling the criteria for both encephalitis and myelitis in any category would be classified as encephalomyelitis.

VAERS is a passive vaccine safety surveillance system operated jointly by the CDC and the US Food and Drug Administration¹⁷ that receives reports of adverse events after administration of US-licensed vaccines. VAERS reports collect information on the nature of suspected vaccine-associated adverse events, on demographics and clinical signs and symptoms, and on vaccinee status.

All VAERS reports of neurologic events temporally associated with smallpox vaccination received between December 16, 2002, and March 11, 2004, were reviewed. A systematic database search identified entries containing 1 or more of 199 Coding Symbol for Thesaurus of Adverse Reaction Terms (COSTART) codes that suggested CNS or PNS dysfunction. The lead author (J.J.S.) reviewed VAERS reports of all identified entries to assess the clinical syndromes and the association with vaccination. Cases with minimal or no clinical detail were excluded. Demographic data on DHHS vaccinees were obtained from the Pre-Event Vaccination System and on DoD vaccinees from the Defense Eligibility Enrollment Reporting System.

We classified cases into categories according to the predominant complaint, sign, or symptom. When possible, we compared observed rates (number of reports per number of vaccinations) of neurologic syndromes among vaccinees to expected incidence rates in the general population.

Observations in this report are based on VAERS-associated surveillance and response activities, which are exempt from institutional review board review.

The SVAEMRA clinical team or DoD was consulted on 30 serious events (eg, resulting in hospitalization, disability, or death) temporally associated with smallpox vaccination events and features suggesting neurologic involvement. These 30 cases were actively investigated by the lead author. Twelve were neurologic, were also captured through VAERS report review, and are summarized in Table 1.

Between December 16, 2002, and March 11, 2004, approximately 665 000 civilians and military personnel received smallpox vaccine, including approximately  435 000 primary (ie, first-time) vaccinees (66%). During this period, 2060 VAERS reports involving smallpox vaccine were received; 670 contained COSTART codes related to neurologic illness, 320 involved DHHS vaccinees (primarily civilian health care workers), and 350 involved DoD vaccinees (primarily military personnel deployed to overseas field situations). A preliminary screen excluded 456 reports unrelated to neurologic illness; the remaining 214 reports underwent detailed review; 2 misclassified cases were subsequently excluded.

The 214 neurologic VAERS reports included 111 DHHS and 103 DoD vaccinees: 105 (49%) were women, 106 (50%) were men, and sex was not identified for 3; overall, 17% of vaccinees were women.¹⁶ One hundred seventy-two (80%) reports were for persons younger than 50 years; the 18- to 29-year group had the largest proportion of adverse events (66; 31%). One hundred thirteen (53%) were reported primary vaccinees. Of reported neurologic adverse events, 54% occurred within 1 week of vaccination and 86% occurred within 30 days. A number of clinical syndromes, including dizziness or vertigo, limb weakness, syncope, and paresthesias, were poorly defined and unassociated with short-term morbidity and will not be discussed herein.

Headache, the predominant symptom among 44% of reports (Table 2), was most common and often accompanied by fever, chills, and fatigue. Most headache cases had no other evidence of neurologic dysfunction and had favorable outcome. Headaches requiring hospitalization and categorized as serious have been described.¹⁸

A total of 39 serious neurologic adverse events associated with smallpox vaccination were reported to VAERS; 27 (69%) were among primary vaccinees and all but 2 occurred within 12 days of vaccination (Table 3).

We classified 13 cases as suspected meningitis, defined by the presence of headache and lumbar puncture performance, indicative of warranting cerebrospinal fluid (CSF) examination. Of these 13, 1 had a documented pleocytosis, and 1 was reported as having CSF “consistent with meningitis,” without further data. The remaining 11 reported normal or negative CSF or other study results (eg, head computed tomography, brain magnetic resonance imaging [MRI]). No reports of meningitis occurred more than 14 days after vaccination. Eight (62%) were primary vaccinees.

We initially classified 4 cases as suspected encephalitis or myelitis, including the 2 cases on which SVAEMRA consulted and 2 identified through VAERS review alone. One report of transverse myelitis described normal CSF and onset 61 to 120 days following vaccination and was excluded. Of the remaining 3 cases, one man had probable encephalitis defined by altered mental status, pleocytosis, and multifocal demyelinating lesions on brain MRI 10 days after primary vaccination. Another man with suspected encephalitis developed fever, altered mental status, and pleocytosis 8 days after primary smallpox vaccination and several weeks after multiple other vaccinations; he recovered completely within 8 days. Cerebrospinal fluid polymerase chain reaction testing of both patients for vaccinia nucleic acid by the CDC’s Poxvirus Laboratory was negative. A third patient with suspected encephalitis 7 days after primary vaccination had mental “slowing” that resolved completely in 1 day; CSF white blood cell count was 6 cells/mm³ (normal <5 cells/mm³). The observed rate for encephalomyelitis with these 3 cases was 5 per million vaccinations (7/million primary vaccinations). No reports of encephalitis occurred more than 14 days after vaccination.

Bell palsy (facial weakness) was reported in 11 vaccinees. The median interval between vaccination and reported onset was 7 days (range, 3-8 days). Six (55%) were primary vaccinees. We estimate the rate of reported Bell palsy to be 1.7 per 100 000 vaccinations (0.9/100 000 primary vaccinations).

Eight patients had reported seizures 9 days after vaccination. Seven (88%) were primary vaccinees. Three (38%) had a history of seizures. Among the remaining 5, 1 patient’s seizure was attributed to hypoglycemia. Two had apparent new-onset seizures: neuroimaging, electroencephalography (EEG), and CSF analysis results were normal, and no recurrences were experienced 2 months after onset for one patient and 4 months for the other. The 2 remaining cases included one with apparent new-onset seizures with no EEG or MRI findings reported and another for whom seizures occurred in a clinical context of severe respiratory distress, cardiac abnormalities, and ultimately death. This case was reported in detail by a Sentinel Review process of the SVS WG and other DHHS entities¹⁹; the suspected underlying etiology was a new-onset autoimmune syndrome.

GBS was reported twice in primary vaccinees and once in a revaccinee. The referring physicians attributed 2 cases to other concomitantly administered vaccinations; the interval between these vaccinations and onset was not reported. One individual for whom neurologic examination, CSF, or EMG findings were not available had GBS onset 1 day after vaccination, an interval biologically unlikely to reflect vaccine causality. Presuming all 3 temporal associations are valid and these cases were not misclassified, the maximum observed rate of GBS was 0.5 cases per 100 000 vaccinations.

Vaccines have reduced neurologic morbidity and mortality from many infections.^20- 22 However, vaccines, including smallpox vaccine, carry a risk of adverse reactions, including rare severe neurologic syndromes.^23- 25

Historically, PVE was the most frequent severe neurologic adverse event associated with smallpox vaccination. Estimated PVE rates vary significantly among studies, presumably because of differences in case definitions, variability in completeness of ascertainment and clinical evaluation, and differences among vaccine strains. An assessment of PVE in the United Kingdom between 1951 and 1960 found a reporting rate of 14 cases per million vaccinations.²⁵ Assessments in Germany in the 1940s and 1950s determined rates ranging from 121 to 449 per million vaccinations.^26- 28 Studies done in the United States in the 1960s suggest a relative infrequency (9-51 cases/million vaccinations) but apparent severity of PVE, with 53% of 68 smallpox vaccine–associated deaths between 1959 and 1966 attributed to PVE.^1- 2,4 The case-fatality rate for PVE was approximately 25%.⁴ The rate and severity varied with the vaccinia strain, with certain reactogenic, neurotropic strains associated with higher rates.^29- 31

Historically, PVE has been nearly 10-fold more common after primary vaccinations than revaccinations.³² Clinicohistopathologic assessments in the 1920s and 1960s identified 2 clinicopathologic manifestations: microglial encephalitis and postvaccinial encephalopathy.^6,33- 35 Microglial encephalitis was more frequent in persons older than 2 years, occurred 10 to 20 days after vaccination, and was characterized by fever, vomiting, headache, and malaise, followed by decreased consciousness, seizures, and coma. Neuropathologically, these persons had widespread demyelination of subcortical white matter. Postvaccinial encephalopathy occurred mainly in children younger than 2 years at 6 to 12 days after vaccination and presented as fulminant seizures, hemiplegia, and elevated intracranial pressure. These patients had diffuse cerebral edema, lymphocytic meningeal infiltration, and perivascular hemorrhages. Vaccinial viremia and virus isolation from brain or CSF was sometimes detected.^36- 37 Although clinicopathologic correlates of PVE during the era of modern neuroimaging and neuroimmunology have not been possible, the demyelinating form of PVE probably corresponds to acute disseminated encephalomyelitis (ADEM) or postvaccination encephalomyelitis, whereas the cytotoxic form may represent an acute neuroinvasive vaccinial encephalomyelitis.

We identified 3 persons with probable or suspected encephalitis after smallpox vaccination, with cerebral inflammation evidenced by altered mental status and pleocytosis. One patient had a longer interval between vaccination and illness onset and MRI findings suggestive of ADEM, and the other 2 had features more consistent with acute viral encephalitis, including a shorter interval between vaccination and onset and rapid recovery. Our estimated reporting rate of probable or suspected PVE cases of 5 cases per million vaccinees is consistent with previous assessments conducted in the United States. Wide availability of neuroimaging and CSF examination allow distinction between mild neurologic symptoms and true cerebral inflammation, reducing misclassification of encephalitis. None of the persons in this study with suspected encephalitis died, perhaps because of improved supportive neurocritical care.

Aseptic meningitis was reported in 13 patients. However, only 2 demonstrated pleocytosis, suggesting that aseptic meningitis may have been overdiagnosed in earlier studies.

Bell palsy and GBS have been reported after infections and immunizations. Most population-based assessments have not found higher-than-expected rates of these phenomena after vaccination, but a statistically significant association has been made for Bell palsy after intranasal influenza vaccine and hepatitis B vaccination^38- 39 and for GBS after rabies vaccination and some formulations of influenza vaccine.^40- 41 However, specific antecedent events are not usually evident. The expected rate of Bell palsy among the general population ranges between 15 and 40 cases per100 000 per year.^42- 43 The 11 cases of Bell palsy yields a rate of 1.7 per 100 000 vaccinations, about or lower than the expected incidence. Published estimates of GBS among the general population range from 0.4 to 4.0 cases per 100 000 per year.⁴⁴ The 3 cases of suspected GBS (estimated reporting rate of 0.5/100 000 vaccinations) is much lower than that observed within the general population. Further, 1 case of GBS was reported 1 day after vaccination, which suggests it was misclassified because a biological link with vaccination would be implausible.

Our estimations of the occurrences of GBS, Bell palsy, and PVE per number of vaccinations are ratios of reported events compared with the number of vaccinations. They are not expressed relative to units of time and are not directly comparable to the published population-based incidence rates. Although not statistically valid, comparison of these ratios with known background rates may be a useful indicator of adverse-events trends associated with vaccinations.^45- 46 Overall, 69% of the serious neurologic adverse events, including all 3 cases of encephalitis, occurred among primary vaccinees, which is consistent with historical data describing more adverse events among primary vaccinees than among revaccinees.

Our assessment has limitations. Like all passive surveillance systems, VAERS reports are subject to underreporting, ascertainment bias and differential reporting, and variability in report quality and completeness.⁴⁶ We did not attempt to obtain additional information from medical record reviews beyond the 12 cases prospectively identified through clinical consultation. Some military personnel receive multiple vaccinations, making it difficult to associate any one with the presumed adverse event. Data on intercurrent infections in cases of suspected meningitis and encephalitis were not available.

Our findings identified many milder neurologic adverse events temporally but not necessarily causally associated with smallpox vaccination. They suggest that such events are generally self-limited, nonserious, and not associated with severe morbidity or mortality when screening defers persons with high-risk conditions.

Smallpox vaccine is given to healthy people, which creates a low tolerance for associated risk. Risks associated with vaccines are best identified through population-based assessments.^47- 48 New, possibly less reactogenic smallpox vaccines are currently under development. Continued monitoring for neurologic events is needed to assess the safety of smallpox vaccines and to better characterize the spectrum of neurologic illness associated with them.