Clinical manifestations

Presentation and course

A variety of neurologic complications of vaccination have been reported, including headache, dizziness or vertigo, paresthesias, postvaccination (“needle-stick”) syncope, seizures, encephalomyelitis, transverse myelitis, meningitis, Bell palsy, and polyneuritis or Guillain-Barré syndrome (128; 63; 64; 298; 224; 195). Almost all serious complications occur within 12 days of vaccination (298). Table 2 shows 1 method of categorizing neurologic adverse reactions to vaccination.

Table 2. Categorization of Neurologic Adverse Reactions to Vaccination

Postvaccination encephalomyelitis usually occurs 10 to 14 days after vaccination but may occur from 7 to 21 days after vaccination (298). Headache, vomiting, drowsiness, and fever are often the first symptoms observed (155; 128; 279; 55; 80; 195). In mild cases, these may be the only symptoms, and in such cases, recovery is often rapid and complete. In more severe cases, symptoms progress to include disorientation, confusion, stupor or coma, paralysis, incontinence, urinary retention, and seizures (155; 128; 279; 55; 80; 195). Cerebral, midbrain, medullary, and spinal cord lesions have been observed clinically.

Acute disseminated encephalomyelitis usually has an abrupt onset 7 to 14 days after vaccination, with fever, altered consciousness, focal neurologic findings, and possibly seizures. Neurologic sequelae occur in 25% to 30% and include personality changes, seizures, dysarthria, paresis, ataxia, and other cerebellar dysfunction.

Clinical vignette

Suspected postvaccinial encephalomyelitis. A 38-year-old man with a history of heavy cigarette usage was hospitalized with acute respiratory distress and hypoxia 10 days after primary smallpox vaccination (68). He was treated for acute epiglottitis with corticosteroids, bronchodilators, antibiotics, and intermittent lorazepam. He was discharged 1 week later on a steroid taper and bupropion (for smoking cessation) but was again readmitted after several days because of agitation, emotional lability, and confusion. He had difficulty concentrating but was afebrile and well-oriented and had no focal neurologic deficits. Head CT showed several punctate areas of hypointensity in the deep white matter. MRI with and without gadolinium contrast showed multiple nonspecific nonenhancing punctate areas of increased signal in the deep white matter on fluid attenuation inversion recovery (FLAIR) images. Laboratory studies were significant for an elevated creatine phosphokinase level in the serum (3000 u/L) and a urine toxicology screen positive for marijuana and benzodiazepines. CSF studies were normal, including cell counts, protein, glucose, and IgG indices, with no oligoclonal bands. Polymerase chain reaction studies of CSF for herpes simplex virus and vaccinia virus were also negative. EEG was normal with the exception of findings consistent with ingestion of benzodiazepines. He improved and was discharged several days later with a diagnosis of steroid-induced psychosis. Four days later (25 days after vaccination) he had a generalized tonic-clonic seizure. A repeat MRI was unchanged from the previous scan. He was treated with phenytoin, and his steroid dose was increased. He was discharged the following day with a diagnosis of postinfectious encephalomyelitis. A month after his vaccination he had mild residual confusion and emotional lability.

The Centers for Disease Control and Prevention (CDC) categorize this case as “suspected” (based apparently on the diagnosis of clinicians involved in the patient’s care) rather than “probable” or “confirmed.” Several other factors could have caused or contributed to the patient’s neurologic presentation, including the acute infection and hypoxia, steroids, lorazepam, and bupropion. Features that are not typical of postvaccinial encephalomyelitis include the absence of fever or focal neurologic deficits, normal CSF cell counts and protein, normal EEG, and absence of enhancing white matter lesions on MRI. The reported CT and MRI findings are not specific. As noted by the Centers for Disease Control and Prevention, “a temporal association with vaccination does not necessarily indicate causality” (68).

Macrophagic myofasciitis. A previously healthy 25-year-old man was admitted for generalized muscle pain with an insidious onset that began 3 years prior to admission (309). He had exercise intolerance and a significant decrease in muscle strength, requiring gait support. He was on no chronic medications and had no recent history of illicit drug use or exposure to recognized toxin. Hepatitis B and tetanus vaccines were administered, respectively, 10 and 2 years before symptom onset. No analytical, imaging, or electromyography changes were found. Muscle biopsy revealed a predominantly macrophagic inflammatory infiltrate with aluminium deposits suggestive of macrophagic myofasciitis, which was attributed to previously administered vaccines.

Biological basis

Etiology and pathogenesis

Many reported neurologic events following vaccination are not caused by the vaccine in question but, rather, are part of the background occurrence of such events. Well-conducted epidemiological studies are necessary to establish causality.

The etiology of true neurologic complications of vaccination may be mixed or multiple; these include vasovagal or anxiety reactions induced by the needle-stick of vaccination, febrile seizures induced by postvaccination fever, direct spread of the vaccine virus to the central nervous system (with live-attenuated vaccines), and induction of immune-mediated demyelination.

Postvaccination febrile seizures do not differ from febrile seizures occurring in other contexts and generally have no worse outcome or prognostic implications (66). The only recognized exception is vaccination-triggered occurrence of febrile seizures in Dravet syndrome; vaccine-induced fever may trigger an earlier manifest onset of this genetic disorder, but this does not alter the course of the disease (221; 378; 343).

In most cases, genetic or structural defects are the underlying cause of epilepsy with onset after vaccination, including cases with either preexistent encephalopathy or benign epilepsy with good outcome (344).

Postvaccination encephalomyelitis likely represents an autoimmune demyelinating process (330). In immunologically normal adults, most cases probably result from induction of autoimmune demyelination, ie, acute disseminated encephalomyelitis, as can occur with other viral exanthems or several vaccines (eg, smallpox, chickenpox, rabies). Limited pathologic studies have demonstrated perivenous demyelination, mononuclear cell perivascular inflammation, and diffuse cortical petechial hemorrhages (295; 310; 278; 279; 55; 80; 195). With the smallpox vaccine, postvaccination encephalomyelitis occurred much more frequently after primary vaccination than after revaccination (295; 55; 171; 195).

Experimental allergic encephalomyelitis has been an informative animal model for postvaccination encephalomyelitis (277; 259; 203), although its utility as a model for multiple sclerosis has been highly contested (312; 85; 154).

With some vaccines, there appears to be variation in individual susceptibility to development of adverse events after vaccination. For example, genetic polymorphisms in the methylenetetrahydrofolate reductase gene and interferon regulatory factor 1, an immunological transcription factor, have been associated with an increased susceptibility to adverse events after smallpox vaccination (273).

In rare cases, live-attenuated viruses may spread directly to the central nervous system. This is most likely in young infants and immunocompromised individuals. This may occur with reactivation of live-attenuated (Sabin) poliovirus and with the vaccinia virus used in smallpox vaccination. In a few cases, vaccinia virus has been recovered from cerebrospinal fluid and blood (136; 135; 137; 195; 197), but the significance of this is not entirely clear in the absence of similar studies on healthy vaccinees. Animal models suggest that vaccinia virus can migrate across the blood-brain barrier but that encephalitis only occurs when additional factors are operative, such as during co-infection by bacteria (116).

Neurologic complications have also resulted from manufacturing defects, as with the Cutter incident in 1955 in which some lots of the Salk inactivated poliovirus vaccine manufactured by Cutter Laboratories contained live polio virus (see further discussion below) (262; 248; 249).

Aluminium oxyhydroxide (alum), a nano-crystaline compound forming agglomerates, was first introduced in vaccines as an immunologic adjuvant effect in 1927. Alum is the most commonly used adjuvant in human and veterinary vaccines. Although generally well tolerated, alum may rarely cause disabling health problems in susceptible individuals. A small proportion of vaccinated people present with delayed onset of diffuse myalgia, chronic fatigue, and cognitive dysfunction, and exhibit very long-term persistence of alum-loaded macrophages at sites of previous intramuscular immunization, forming a granulomatous lesion called macrophagic myofasciitis (126; 127; 124; 125; 309). Macrophagic myofasciitis should be regarded as pathological only if detected at least 18 months after last aluminum-containing immunization (309). Macrophagic myofasciitis is considered part of the spectrum of the so-called autoimmune/inflammatory syndrome induced by adjuvants (“ASIA”) (304).

Epidemiology

The United States Vaccine Adverse Event Reporting System (VAERS). The United States Vaccine Adverse Event Reporting System (VAERS), a national passive surveillance system for reporting adverse events that occur after administration of United States-licensed vaccines, is important for detecting rare adverse events and assessing reporting trends but cannot be used to assess incidence rates or causality. Because reports to VAERS are submitted voluntarily by individuals with different professional backgrounds and personal interests (eg, manufacturers, health-care providers, patients, and family members), reports vary in quality and completeness, often lack detail, and frequently include inaccurate information. Actual rates for adverse events reported to VAERS cannot be calculated because (1) the cases reported are influenced by both underreporting and differential reporting, and (2) the corresponding number of doses administered is unknown. Differential reporting is often manifest by increased reporting for specific vaccines or specific adverse events, particularly for more serious and unexpected events, events occurring soon after vaccination, events surrounded by publicity, and reports related to litigation proceedings (372).

Standard case definitions for (some) adverse neurologic events. The Brighton Collaboration was inaugurated in 2000 as a voluntary international organization to facilitate the development, evaluation, and dissemination of high-quality information about the safety of human vaccines (53; 54). The Collaboration subsequently developed standardized case definitions of adverse events following immunizations (AEFI). Standardized case definitions of adverse events following immunizations are an essential element in the assessment of immunization safety because they establish a common “vocabulary” and understanding of adverse events following immunizations that facilitates the comparability of data from clinical trials and surveillance (54). In particular, Brighton Collaboration case definitions were developed for encephalitis, myelitis, and acute disseminated encephalomyelitis (ADEM) (297); aseptic meningitis (326); and facial nerve palsy including Bell palsy (272). In later work from the Brighton Collaboration, 45 maternal and 62 fetal/neonatal events were prioritized, and key terms and concept definitions were endorsed, including various adverse outcomes of pregnancy, microcephaly and macrocephaly, neonatal encephalopathy, hypoxic ischemic encephalopathy, neonatal seizure, hypotonia/hypertonia, hyporeflexia/hyperreflexia, meningitis, meningoencephalitis, intraventricular hemorrhage, and periventricular leukomalacia (237).

Assessment of a causal relationship ("causality"). Unfortunately, many associations between vaccination and putative neurologic complications have been asserted without sufficient objective data to establish a causal relationship (233; 65). As Gellin and Schaffner noted in 2001, events that occur after vaccination are likely to be attributed to vaccination, regardless of whether or not the vaccine actually caused the events (122). Many patients and some physicians interpret the temporal sequence of vaccination-followed-by-a-neurologic-event as sufficient evidence to support an interpretation of causality—ie, that the vaccine actually caused the neurologic event (post hoc ergo prompter hoc fallacy; “after this, therefore, because of this”). The consideration of a causal association is supported by apparent biological plausibility: for example, because hepatitis B vaccination stimulates the immune system and because multiple sclerosis has an immunologic basis, several authors have suggested that hepatitis B vaccination may trigger a paradoxical response that triggers the onset of (ie, causes) multiple sclerosis. Furthermore, published case reports and inadequately performed epidemiologic studies suggesting increased risk instill “the persuasive power of immediacy” to public policy development and medicolegal decision-making. However, these considerations are not sufficient to establish causality.

English epidemiologist and statistician Sir Austin Bradford Hill CBE FRS (1897-1991) pioneered the randomized clinical trial and, together with British physician epidemiologist Sir William Richard Doll CH OBE FRS (1912-2005), demonstrated the connection between cigarette smoking and lung cancer.

Hill is widely known for pioneering the "Bradford Hill" criteria for determining a causal association.

Braford Hill "considerations" to assess causality (presented in the original order):

The influential Bradford Hill "viewpoints" (often referred to by others as "criteria," although he did not see them as such) are still commonly used to assess causality within epidemiology (149). Bradford Hill's considerations or viewpoints can be misapplied and misinterpreted, and some of the viewpoints may even be misleading, for example, if studies reporting similar effects are similarly biased (which is not uncommon) (65).

Hill's approach also served as a stimulus for later causal frameworks, including some that tried to incorporate multicausal networks: (1) directed acyclic graphs (DAGs); (2) sufficient-component cause models ("causal pies"); and (3) the Grading of Recommendations, Assessment, Development and Evaluation (GRADE) methodology (284; 323; 299; 129; 87; 01; 03; 56; 341; 373; 31; 30; 98; 42; 51; 77; 293; 340; 377).

The GRADE approach, used to assess the quality of evidence for a specific outcome across studies, is most directly applicable to a meta-analysis undertaken in the context of a systematic review (but can also be applied to individual studies or nonquantitative syntheses). The different causal frameworks have substantial overlap, particularly for four of Hill's viewpoints: "strength of association (including analysis of plausible confounding); temporality; plausibility (encoded by DAGs or SCC models to articulate mediation and interaction, respectively); and experiments (including implications of study design on exchangeability)" (303).

The Clinical Immunization Safety Assessment (CISA) network in the United States developed a thoughtful and very detailed algorithm to assist in collecting and interpreting data, and to help assess causality after individual adverse events following immunizations (AEFI) (199; 140).

In 2019, the World Health Organization published an updated manual on causality assessment of an adverse event following immunization (AEFI) (371). WHO categorized adverse events following immunizations into 5 categories: (1) vaccine product-related reaction; (2) vaccine quality defect-related reaction; (3) immunization error-related reaction (ie, caused by inappropriate vaccine handling, prescribing or administration); (4) immunization anxiety-related reaction/Immunization stress related response (ISRR); and (5) coincidental event (371). For causality assessment, the WHO relied heavily on the Bradford Hill viewpoints. WHO developed a causality assessment checklist with 5 major sections: (1) Is there strong evidence for other causes? (2) Is there a known causal association with the vaccine or vaccination? (3) Was the event in section II within the time window of increased risk? (4) Is there strong evidence against a causal association? and (5) Other qualifying factors for classification (eg, history of prior similar response to a similar vaccine, or in the absence of vaccination) (371).

Butler and colleagues in 2021 provided another excellent resource from the United Kingdom that provided criteria for defining causality in vaccine-related adverse events adapted from WHO criteria and Bradford Hill viewpoints (149; 371; 60). Butler and colleagues also presented suggested criteria for labelling causality in neurologic adverse events following immunization as probable, possible, or unlikely (60).

Reference to these guides by clinicians, authors, and editors would greatly improve the reporting of adverse events following immunization (149; 140; World Heath Organization 2019; 60).

Criteria for labelling causality in neurologic adverse events following immunization (60):

In a review of 154 postmarketing individual case safety reports (ICSRs) reporting immunization errors and a fatal outcome that were submitted to EudraVigilance from 2001 to 2016, 20 cases were classified as consistent with causal association of which 12 had a large error contribution, 2 had a moderate error contribution, and the remainder had small or no error contributions or were unclassifiable; 2 general issues were responsible for the 12 cases with a large error contribution: (1) incorrect handling of the multidose vial after opening of the vial; and (2) administration of live-attenuated vaccines to immunocompromised patients (153).

Neurologic complications of vaccination: general considerations. Neurologic complications of vaccination are in fact rare (233). Most patients with vaccine-associated neurologic complications recover fully, although rare poor outcomes are recognized (233).

Febrile seizures. Multiple vaccines have been associated with an increased risk of febrile seizure: (1) measles; (2) measles-mumps-rubella (MMR); (3) measles-mumps-rubella-varicella (MMRV); (4) combined diphtheria, tetanus, acellular pertussis, polio, and Haemophilus influenzae type b vaccine; (5) whole-cell pertussis vaccine; (6) 7-valent pneumococcal conjugate vaccine; and (7) concomitant administration of the trivalent inactivated influenza vaccine with either the 7-valent pneumococcal conjugate vaccine or the diphtheria, tetanus, and acellular pertussis vaccine (365; 235). Children receiving these vaccinations should not receive prophylactic antipyretics because no significant reduction in the rate of febrile seizures has been documented and because prophylactic antipyretic use can potentially decrease the immune response to certain vaccines (235).

Vaccination-proximate febrile seizures represent only a small proportion of all febrile seizure hospital presentations, and vaccination-proximate febrile seizures have the same outcomes as other febrile seizures (91). Specifically, febrile seizures after vaccination do not increase the likelihood that epilepsy or other neurologic disorders will develop in these children (365). Children with a history of convulsions may be at increased risk for febrile convulsions after vaccination (particularly MMR or MMRV vaccination), but the risk appears to be minimal (365).

Postvaccination syncope. Postvaccination syncope is an uncommon but recognized complication of vaccination (and other needle-stick-related medical procedures) (72; 193; 311). From 0.28 to 0.54 episodes of syncope are reported per million doses of vaccine distributed, although for various methodological reasons these figures may not reflect the true rate of postvaccination syncope. Postvaccination syncope occurs primarily among girls aged 11 to 18 years. Syncope is most commonly reported after 3 vaccines given to adolescents: human papilloma virus, MCV4 (meningococcal disease), and Tdap (tetanus, diphtheria, and pertussis). Because the ingredients of these 3 vaccines are different, syncope is most likely due to the vaccination process and not to the vaccines themselves. Only about 6% of postvaccination syncopal cases are considered "serious" (including those with secondary injuries), and three fourths of these occurred in females. Rare serious outcomes with secondary injury include skull fracture, other head injury, subarachnoid and other intracranial hemorrhage, motor vehicle accident, and death. One half of "serious" cases of postvaccination syncope occur within 5 minutes of vaccination, and more than two thirds occur within 15 minutes.

Vaccination and CNS demyelinating disorders. In 1 study, vaccination (of any type) was associated with an increased risk of acquired central nervous system demyelinating syndromes within the first 30 days after vaccination only in younger (less than 50 years) individuals (OR, 2.32; 95% CI, 1.18-4.57), but there was no longer-term association of vaccines with multiple sclerosis or any other CNS demyelinating syndrome (190); the authors concluded that the apparent short-term increase in risk suggests that vaccines may accelerate the transition from subclinical to overt autoimmunity. Postvaccination acute disseminated encephalomyelitis has been associated with several vaccines, including those for rabies, diphtheria-tetanus-polio, smallpox, measles, mumps, rubella, Japanese encephalitis, pertussis, influenza, and hepatitis B (161). However, vaccination is not associated with an increased risk of multiple sclerosis or optic neuritis for the following vaccines: BCG, diphtheria, influenza, measles, MMR, polio, rubella, tetanus, and typhoid (92; 93; 94; 147; 104).

Although several studies suggested a slightly increased risk of multiple sclerosis onset with hepatitis B vaccination, most studies have not found such an association (334; 32; 92; 93; 263; 306; 152; 104; 206). This consideration has largely been dismissed (see further discussion in section on hepatitis B vaccination below).

In general, immunization of patients with multiple sclerosis has been effective and relatively safe, such that benefits of vaccination outweigh the risks. In particular, vaccination does not appear to increase the short-term risk of relapse in multiple sclerosis (83). A review in 2002 for the American Academy of Neurology found (1) strong evidence against an increased risk of multiple sclerosis exacerbation after influenza immunization, (2) no evidence that hepatitis B, varicella, tetanus, or BCG vaccines increase the risk of multiple sclerosis exacerbations, and (3) insufficient evidence for other vaccines, particularly live-virus vaccines (286). In addition, patients with multiple sclerosis are at a significantly heightened risk of severe adverse outcomes (including death) following infectious diseases for which vaccinations are now available. Therefore, such patients should not be denied the benefits of protective vaccinations because of potential undocumented risks of disease exacerbation. Decision-making regarding immunization may be complicated in multiple sclerosis patients who are receiving immunosuppressive or immunomodulatory therapy that might impair their ability to generate an appropriate protective response to immunization. Also, live-virus vaccines are contraindicated in patients on immunosuppressant therapy.

Selected vaccines

Anthrax vaccine (AVA). Fatigue and headache were the most commonly reported systemic adverse events following anthrax vaccine administration among clinical trial participants, but they were reported by fewer than 11% of participants (372). Serious disorders of the nervous system were infrequently reported after anthrax vaccine vaccination, and none have been fatal (372; 362). None of the studies found that the risk for adverse health effects or chronic diseases was higher after anthrax vaccination (372). There is no evidence that anthrax vaccine recipients had a higher risk than the general population for life-threatening or permanently disabling adverse events immediately after receiving anthrax vaccine (170; 372). In addition, in a matched case-control study, Vaccine Analytic Unit (VAU) investigators found no significant associations between optic neuritis and previous receipt of anthrax vaccine (257).

Hepatitis B vaccine. In general, there are minimal adverse reactions (eg, myalgias) after receipt of hepatitis B vaccine, and these are transient, mostly limited to within 24 hours. Serious side effects of hepatitis B virus vaccine are rare.

During early post-licensure surveillance, several types of neurologic adverse events following hepatitis B vaccination were reported, including Guillain-Barré syndrome, chronic fatigue syndrome, sudden infant death syndrome, optic neuritis, and multiple sclerosis. Guillain-Barré syndrome was reported rarely following immunization with the plasma-derived vaccine, but Guillain-Barré syndrome has not been associated with the newer recombinant hepatitis vaccines. In addition, in the late 1990s, there were reports of onset or relapse of multiple sclerosis following hepatitis B immunization, particularly following an extensive immunization program in France; consequently, in 1998, the French government suspended the school-based vaccination program because of fears that hepatitis B vaccination caused neurologic disorders, including multiple sclerosis. Some considered a causal explanation to be biologically plausible (a viewpoint in Bradford Hill's list of causality considerations) because immune stimulation from vaccination might precipitate the onset of multiple sclerosis in susceptible persons or could lead to relapses, possibly through molecular mimicry between vaccine antigens and human proteins (emphasizing the limitations of biological plausibility as a consideration of causality) (219). Critically, infection with wild-type hepatitis B virus (HBV) has not been identified as a risk factor for multiple sclerosis, and there is no evidence that hepatitis B virus protein may trigger an attack on nerve tissue (219).

However, since then, multiple epidemiological studies have demonstrated no causal association between receipt of hepatitis B vaccine and any of these conditions (139; 382; 290; 32; 318; 319; 94; 226; 227; 228; 29; 291). The well-designed and conducted study by Ascherio and colleagues identified no association between hepatitis B vaccination and the development of multiple sclerosis (32); the strengths of this study included careful selection of control groups, a high prevalence of hepatitis B virus immunization among study subjects, and validation of vaccination status. Furthermore, the lack of an association between hepatitis B virus vaccination and development of multiple sclerosis is supported by other studies, including a Canadian study of children vaccinated as part of a school-based program (290) and a retrospective cohort study in the United States (382).

In addition, no evidence of a causal association between Bell palsy (285), Guillain-Barré syndrome (44), optic neuritis (257; 45), sudden-onset sensorineural hearing loss (46), or other chronic illnesses and receipt of hepatitis B vaccine has been demonstrated through analysis of the CDC’s Vaccine Safety Datalink (VSD) data.

In fact, there are no well-conducted studies demonstrating a statistically significant association between hepatitis B virus vaccination and the subsequent development of multiple sclerosis. Therefore, prior concerns about the development of multiple sclerosis should not interfere with vaccination of individuals at increased risk of hepatitis B virus infection, nor does available evidence justify changes in national vaccination policy for hepatitis B virus (191).

Human papilloma virus (HPV) vaccines. Available evidence has not suggested any significant safety concern regarding the use of HPV vaccines, and in particular it has not identified any serious neurologic adverse effects of the vaccine (28; 368). The proportions of persons reporting a serious adverse event were similar in the vaccine and placebo groups, as were the types of serious adverse event reported. Among nonserious adverse events in postlicensure data, the most commonly reported generalized symptoms included syncope, dizziness, headache, and myalgias (81; 28). In a VSD collaborative study between the Centers for Disease Control and Prevention and 9 integrated health-care organizations, there were no statistically significant increased risks observed for any of the prespecified endpoints including Guillain-Barré syndrome, stroke, seizures, or syncope (121; 81). In a general safety assessment evaluating outcomes diagnosed in emergency room visits and hospitalizations among 189,000 females receiving at least 1 dose of HPV4, same-day syncope was associated with HPV4 (81). In France, a case-control study to evaluate autoimmune disorders following HPV4 (among 211 cases and 875 controls) found no increased risk for central demyelination disease, multiple sclerosis, or Guillain-Barré syndrome (81; 133). Another study in England similarly found no increased risk for Guillain-Barré syndrome (13). A review of post-licensure safety surveillance during more than 4 years of routine use of the bivalent vaccine found no patterns or trends for potential immune-mediated diseases after vaccination, and in particular the observed incidences of Bell palsy and (confirmed) Guillain-Barré syndrome were within the expected range in the general population (14). Although anecdotal concerns have been raised about complex regional pain syndrome and postural orthostatic tachycardia syndrome following HPV vaccination, reviews of pre- and postlicensure data provide no evidence that these syndromes are caused by the HPV vaccines (28; 368).

Herpes zoster (shingles) vaccines. Overall, with the live-attenuated vaccine Zostavax (Zoster Vaccine Live, ZVL), serious adverse events occurred at similar rates in vaccinated and placebo groups (04; 100). In rare instances, the ZVL vaccine strain may cause disseminated rash and herpes zoster in immunocompetent recipients (Food and Drug Administration, 2006; 04; 100), and life-threatening and fatal complications in immunocompromised recipients (100). In the Shingles Prevention Study, the number and types of serious adverse events during the 42 days after receipt of vaccine or placebo were similar (1.4%) (252). However, rates of serious adverse events in the safety substudy were higher in vaccine recipients than in placebo recipients (1.9% vs. 1.3%, respectively, with a relative risk of 1.5 and 95% CI of 1.0-2.3), but no temporal or clinical patterns of adverse events were observed in vaccine recipients to suggest a causal relation. Zoster-like rashes were less common in vaccine versus placebo recipients during this 42-day period (p< 0.05) and the Oka/Merck strain of VZV was not detected in any of 10 lesion specimens from vaccine recipients available for PCR testing (indicating that these cases had zoster from reactivation of previously acquired VZV). In contrast, in early studies conducted during vaccine development, samples from rashes in 2 vaccinated persons were confirmed to be Oka/Merck-strain VZV; both experienced noninjection-site varicella-like rashes (72). (Note: Zostavax is no longer available for use in the United States, as of November 18, 2020.)

With the new subunit vaccine Shingrix (Zoster Vaccine Recombinant, Adjuvanted; RZV), overall rates of serious adverse events over the study periods were similar in the RZV and placebo groups (100).

Zoster vaccines and Guillain-Barré syndrome. In a cohort study and self-controlled case series analyses, Goud and colleagues assessed the risk of Guillain-Barré syndrome after administration of the zoster vaccines (130). This study utilized Medicare claims data on beneficiaries aged 65 years or older. The cohort study compared approximately 849,000 individuals vaccinated with Shingrix and approximately 1,817,000 individuals vaccinated with Zostavax. Self-controlled analyses included events identified from approximately 2,114,000 eligible beneficiaries 65 years or older vaccinated with Shingrix. The authors compared the relative risk of Guillain-Barré syndrome after Shingrix versus Zostavax, followed by claims-based and medical record-based self-controlled case series analyses to assess the risk of Guillain-Barré syndrome during a postvaccination risk window (from days 1 to 42) compared with a control window (from days 43 to 183).

In the cohort analysis, the risk of Guillain-Barré syndrome was significantly increased among those receiving Shingrix compared with those receiving Zostavax, with a rate ratio point estimate of 2.3. In the self-controlled case series analyses, there was a significantly increased risk of Guillain-Barré syndrome in the risk window following Shingrix compared with the control window, with a corresponding rate ratio of 4.3, and an adjusted attributable risk of 5.1 excess cases of Guillain-Barré syndrome per million vaccinations. Secondary analysis included approximately 951,000 first doses of Shingrix and identified a significantly increased risk of Guillain-Barré syndrome with a rate ratio of 9.3, and an adjusted attributable risk of 9.5 excess cases for every million vaccinations administered. The medical record-based analysis confirmed this significantly increased risk, finding a rate ratio of 5.0, and an adjusted attributable risk of 5.2 excess cases for every million vaccinations administered. The extended study period self-controlled case series decreased the observed rate ratio to 2.8 and the adjusted attributable risk to 3.1 excess cases for every million vaccinations administered.

The authors concluded that there is a slightly increased risk of Guillain-Barré syndrome during the 42 days following Shingrix vaccination in the Medicare population, with approximately 3 excess Guillain-Barré syndrome cases per million vaccinations administered.

Vaccination against shingles in patients with multiple sclerosis. In 1997, Ross and colleagues conducted an uncontrolled longitudinal pilot trial of vaccination of 50 multiple sclerosis patients with a live-attenuated VZV vaccine (N.B., not ZVL/Zostavax) as a possible therapeutic agent (283). All of the patients were seropositive for VZV before the vaccination and all of them developed a rise in antibody titers following vaccination. The patients were followed for 1 year. There were no major adverse effects identified. Overall, at the end of the trial, of 47 patients for whom such information was available, 14 were reported to be improved, 4 were reported to be worse, and 29 were unchanged. Although not a controlled trial, these results suggest that live-attenuated VZV vaccines are not necessarily harmful to patients with multiple sclerosis and could possibly be helpful.

In general, despite the results of Ross and colleagues (283), live-virus vaccines are not recommended for people with multiple sclerosis because live-virus vaccines can potentially precipitate an increase in disease activity (194). However, most people have had chickenpox earlier in their lives, so the virus is already in their bodies, undercutting the need for avoidance of live-virus vaccines for prevention of shingles in patients with multiple sclerosis. The use of ZVL (Zostavax) was not addressed in the 2002 American Academy of Neurology guideline on vaccine use in multiple sclerosis (286), prescribing information from the manufacturer (Merck), the CDC recommendations (142), or the FDA licensing of Zostavax. Regardless of these considerations, ZVL (Zostavax) is contraindicated for patients with multiple sclerosis who are on immunosuppressant medications: “For patients on immunosuppressant medications, live attenuated vaccines should be avoided while patients are on therapy and for 3 months after treatment” (61).

However, now that there is a subunit vaccine for prevention of shingles (Zoster Vaccine Recombinant, Adjuvanted; RZV; Shingrix), the evidence gap and the potential risks of the live-attenuated vaccine in this population group have been largely obviated: pending new information that would indicate an unsuspected risk in this population group with the subunit vaccine, the subunit vaccine is strongly preferred for patients with multiple sclerosis.

Influenza vaccines (inactivated). A matched case-control study found no significant associations between optic neuritis and previous receipt of influenza vaccines (257).

Influenza vaccines and Guillain-Barré syndrome. The 1976 swine influenza vaccine was associated with an increased risk of Guillain-Barré syndrome (160), with a rate of Guillain-Barré syndrome among vaccine recipients exceeding the background rate by approximately 10 cases per million vaccinated (20; 364). Multiple subsequent controlled studies did not identify a significant association between influenza vaccination and Guillain-Barré syndrome (160; 173; 280; 159; 172; 324; 315; 57; 58; 131; 364). However, during some influenza seasons since the 1976/77 influenza season, inactivated influenza vaccines have been associated with a slight increase in the risk of Guillain-Barré syndrome in some population groups (17; 172; 24; 266). For example, there was a small, but significantly increased, risk for hospitalization because of Guillain-Barré syndrome in Ontario, Canada, in the period from 1992 to 2004, although a separate time-series analysis from the same population found no statistically significant increase in hospital admissions because of Guillain-Barré syndrome after the introduction of the universal influenza immunization program (172). In the United States, there was a small excess risk of Guillain-Barré syndrome in vaccine recipients aged 18 to 64 years in the 1990/91 vaccine season (17), and a small excess risk of Guillain-Barré syndrome among the Medicare population after monovalent H1N1 influenza vaccination in 2009/10 vaccine season (266), but no such increase was found in some other years since 1976 (200). For comparison, influenza-related Guillain-Barré syndrome is 4 to 7 times more frequent than influenza-vaccine-associated Guillain-Barré syndrome (364). The weight of epidemiologic evidence over the last 40 years does not support a major causal relationship between influenza vaccinations and Guillain-Barré syndrome (364). Nevertheless, given the experience with the 1976 swine influenza vaccine, and some intermittent small increased risks with some subsequent influenza vaccines, a significant association cannot be confidently excluded for future vaccine strains (164). A history of Guillain-Barré syndrome within 6 weeks following a previous dose of any type of influenza vaccine is a relative contraindication to further influenza vaccination (134).

Influenza A (H1N1) vaccines and narcolepsy etc. A sudden increase in the incidence of childhood narcolepsy was observed in Finland in 2009 (255); among 54 cases of incident childhood narcolepsy, 93% had received the AS03-adjuvanted pandemic H1N1 influenza A vaccine (Pandemrix, GlaxoSmithKline Biologicals, Germany) made using the European inactivation/purification protocol and used in Europe during the 2009 H1N1 influenza A pandemic. The incidence of narcolepsy was 9.0 per 100,000 person-years in vaccinated individuals compared with only 0.7 per 100,000 person-years in the unvaccinated individuals (rate ratio 12.7 with 95% confidence interval 6.1-30.8) (246). The vaccine-attributable risk of developing narcolepsy was 1 per 16,000 vaccinated 4- to 19-year-olds (95% confidence interval 1:13,000-1:21,000) (246). Similar, but less marked, increases in narcolepsy occurred in association with this vaccine in other European countries (86; 230; 356; 250; 261), with some studies also demonstrating an increased risk of lower magnitude in adults (316). Antibodies from vaccine-associated narcolepsy sera cross-reacted with both influenza nucleoprotein and hypocretin receptor 2, a hypothalamic neuropeptide that is involved in the regulation of sleep and arousal states, and that was previously shown to be deficient in narcolepsy (05). In contrast, influenza vaccines containing the A(H1N1)pdm09 virus strain used in the United States, Saudi Arabia, and Korea were not associated with an increased risk of narcolepsy (79; 102; 269). Results were inconsistent in Canadian studies (143; 236). There are still many unanswered questions regarding this novel vaccine-related adverse effect, particularly given reports that narcolepsy developed in some nonvaccinated children infected with wild A(H1N1) pandemic influenza (suggesting a role for viral antigens in disease development, although this apparently did not occur in Finland) (225) and different rates observed in different countries, particularly using different inactivation or purification protocols (suggesting that the AS03 adjuvant alone was not sufficient for disease development).

Postvaccination narcolepsy with cataplexy has also since been reported with tick-borne encephalitis vaccine (148) and yellow fever vaccine (281).

Relative risks were significantly increased for Bell palsy, paraesthesia, and inflammatory bowel disease after vaccination, predominantly in the early phase of the A(H1N1) vaccination campaign, but there was no change in the risk of Guillain-Barré syndrome or multiple sclerosis (40).

Japanese encephalitis virus vaccines.

Mouse-brain-derived inactivated vaccines. Most mouse-brain-derived inactivated vaccines have been discontinued (367) because of serious, but rare, allergic and neurologic adverse events (372). Although these vaccines contained no myelin basic protein at the detection threshold of the assay, use of mouse brains as the substrate for virus growth raised concerns about potential neurologic side effects associated with the vaccine. Moderate-to-severe neurologic symptoms (eg, encephalitis, seizures, gait disturbances, and parkinsonism) were reported at a rate of 0.1 to 2 cases per 100,000 vaccine recipients (with variation noted by country, case definition, and surveillance method) (372). In addition, cases of severe or fatal acute disseminated encephalomyelitis were reported that were temporally associated with vaccination of children in Japan and Korea (109; 367). In 2005, in response to these cases, Japan suspended routine vaccination with mouse-brain-derived Japanese encephalitis vaccines (109). However, the WHO Global Advisory Committee on Vaccine Safety determined that no evidence existed of an increased risk for acute disseminated encephalomyelitis associated with the mouse-brain-derived Japanese encephalitis vaccine and that a causal link had not been demonstrated. Nevertheless, the committee recommended that the inactivated mouse brain-derived vaccine be gradually replaced by new-generation Japanese encephalitis vaccines (109).

Vero-cell-derived inactivated vaccine. Neurologic events (encephalitis, encephalopathy, convulsions, peripheral neuropathy, transverse myelitis, and aseptic meningitis) have been reported at a rate of 1 per 1,000,000 doses (367). Two cases with serious neurologic events after vaccination (neuritis and oropharyngeal spasm) were reported from 12-month postmarketing-surveillance data for 10 phase-III trials in adults; however, on review they were considered unrelated to vaccination, and both subsequently recovered (294).

Live-attenuated vaccine. In a randomized trial of the safety of Japanese encephalitis vaccine (SA14-14-2) in 26,239 children, no cases of encephalitis or meningitis or severe reaction consistent with anaphylaxis occurred in either group within 30 days of vaccination (367). No cases of vaccine-associated Japanese encephalitis were reported in a review of data covering a 20-year period as of 2005 (367).

Meningococcal vaccines.

Polysaccharide vaccines. Adverse reactions to polysaccharide meningococcal vaccines are usually mild (26). Neurologic adverse effects (eg, seizures, anesthesias, and paresthesias) have been observed infrequently (22).

Conjugate vaccines. The most frequently reported adverse events following vaccination with MenACWY-D include headache (16%), dizziness (13%), and syncope (10%) (222). Among the 6.6% of reports coded as serious, the most frequent adverse events reported included headache (38%) and Guillain-Barré Syndrome (15%), although the diagnosis was not validated by medical records for all reports (222).

In 2005, shortly after licensure of MenACWY-D, several cases of Guillain-Barré syndrome were reported to VAERS (69; 70). Symptom onset clustered approximately 14 days after vaccination. No deaths were reported, and most persons recovered fully. The Advisory Committee on Immunization Practices reviewed the data and concluded that vaccine-induced protection against meningococcal disease outweighed the potential small increased risk for Guillain-Barré syndrome (78). However, because the risk for recurrence of Guillain-Barré syndrome after meningococcal vaccination was unknown, the FDA considered a prior history of Guillain-Barré syndrome to be a contraindication for use of this vaccine (71). The attributable risk for Guillain-Barré syndrome ranged from 0 to 1.5 additional cases of Guillain-Barré syndrome per 1 million vaccines administered within the 6-week period following vaccination, based on a large retrospective cohort study conducted from 2005- to 2008 and involving approximately 1.4 million adolescents aged 11 through 21 years (342). No cases of Guillain-Barré syndrome were identified within 1 to 42 days following 889,684 vaccine doses of MenACWY-D administered from January 2005 to March 2010 (222). In June 2010, after reviewing the 2 safety studies, the Advisory Committee on Immunization Practices removed the precaution for persons with a history of Guillain-Barré syndrome because the benefits of meningococcal vaccination were felt to outweigh the risk for recurrent Guillain-Barré syndrome in these people (222). Although a history of Guillain-Barré syndrome continues to be listed as a precaution in the package inserts for MenACWY-D, no specific concerns have since been raised about the risk for Guillain-Barré syndrome in persons who have a history of this condition and have been vaccinated with meningococcal conjugate vaccine (43; 222).

Among approximately 5,000 infants studied through 6 months postvaccination in 5 clinical trials of MenACWY-CRM, local and systemic adverse events after administration and routine vaccinations were similar to those observed after routine vaccination alone (212); 11 serious adverse events were considered possibly related to MenACWY-CRM, including acute encephalomyelitis (n=1), complex partial seizure (n=1), epilepsy (n=1), and febrile seizure (n=3) (212). No deaths were considered related to MenACWY-CRM (212). During the 20-month period from February 2010 to September 2011, there were 284 reports of adverse events reported to VAERS following receipt of MenACWY-CRM in the United States: syncope was reported in 9%, but Guillain-Barré syndrome was not reported (222).

Serogroup B protein subunit vaccines. In clinical trials, both MenB vaccines (MenB-4C and MenB-FHbp) were generally well tolerated, with few serious adverse events that all resolved without sequelae (110; 213; 82; 256).

In 3 controlled clinical trials of MenB-4C, serious adverse events were assessed through 6 months postvaccination in 2,716 participants who received at least 1 dose of MenB-4C (110; 213). One patient developed a tremor, which was considered by the study investigator to be a serious adverse event possibly related to the vaccine (110; 213); the administration of the investigational vaccine and onset of tremor were considered reasonably related in time, and the serious adverse event could not be attributed to other causes.

In addition, information about serious adverse events after MenB-4C vaccination was collected during 3 prelicensure vaccination campaigns in response to 3 outbreaks of serogroup B meningococcal disease at universities in the United States and Canada (110; 213). A total of 59,091 participants received at least 1 dose of MenB-4C. Rates of serious adverse events were similar in the vaccine and the control groups. One patient developed rhabdomyolysis (that resolved with no sequelae), which was considered by the study investigator to be a serious adverse event possibly related to the vaccine (110; 213).

In 7 clinical trials, a total of 9,808 subjects received at least 1 dose of MenB-FHbp, the most common adverse reactions in the 7 days after vaccination included fatigue (at least 40%), headache (at least 35%), and myalgias (at least 30%) (110; 213; 256). Four subjects reported 7 serious adverse events that were considered by the study investigator to be possibly related to the vaccine, including 1 report each of headache and vertigo (110; 213). All vaccine-related serious adverse events resolved without sequelae. No increased risk for any specific serious adverse event considered to be clinically significant was identified in any of the studies. No deaths were considered to be related to MenB-FHbp (110; 213).

Measles vaccine. Adverse reactions following measles vaccination are generally mild and transient (369). Measles vaccines can cause febrile seizures (see discussion under MMR/MMRV below), and there has been (unrealized) concern that it would impose a significant risk of postvaccination encephalomyelitis, Guillain-Barré syndrome, or subacute sclerosing panencephalitis (SSPE).

Although natural measles infection can cause an immunologically mediated postinfectious encephalomyelitis with perivenular demyelination in approximately 1 per 1000 infected persons (365a), the Institute of Medicine did not find convincing evidence to establish a causal relationship between administration of live-attenuated measles vaccine and development of postvaccination encephalomyelitis (320). Furthermore, the British National Childhood Encephalopathy Study (NCES) did not identify an increased risk of permanent neurologic abnormality following measles vaccination after 10 years of follow-up (229). Analysis of claims for encephalitis following measles vaccination in the United States found clustering of reported adverse events at 8 to 9 days after immunization, which is consistent with, but does not prove, the possibility that the vaccine causes encephalitis (101; 355); however, the reported risk was less than 1 per million doses, or roughly 1,000 times less than the risk from measles infection (365a).

Another late adverse outcome of natural measles infection is the development of subacute sclerosing panencephalitis (SSPE), a progressive, disabling, and deadly brain disease caused by recrudescence of latent measles virus, typically years after the primary infection. Naturally, there was concern initially that a live-virus measles vaccine would pose a significant risk of later development of SSPE. However, the live-attenuated measles vaccine does not increase the risk for subacute sclerosing panencephalitis, even among individuals with a prior history of measles or measles vaccination (157; 101). Furthermore, vaccine-strain measles virus has never been identified in patients with SSPE.

Guillain-Barré syndrome has been reported following vaccination with MMR and its component vaccines; however, the United States Institute of Medicine found insufficient evidence to accept or reject a causal relationship (320). Subsequent studies also found no evidence to support a causal association between measles vaccination and Guillain-Barré syndrome (158; 307; 365a).

Measles inclusion body encephalitis. Measles inclusion body encephalitis has been reported in persons with immune deficiencies following measles vaccination and was documented by demonstration of intranuclear inclusions corresponding to measles virus (39; 267) or by isolation of measles virus from the brain among vaccinated persons (50). The interval from vaccination to development of measles inclusion body encephalitis ranged from 4 to 9 months, consistent with development of measles inclusion body encephalitis after infection with wild measles virus (238).

Mumps vaccine and aseptic meningitis. Adverse reactions to mumps vaccination are generally rare and mild, although sensorineural deafness (240; 314), acute myositis (282), and particularly aseptic meningitis have been reported after mumps vaccination (21; 365a).

Aseptic meningitis. Aseptic meningitis has been reported over a wide incidence range from 1 case/400 vaccinations to 1/1,500,000 vaccinations, with the extremely wide range apparently reflecting differences in vaccine strains and their preparation as well as variation in study design and diagnostic criteria (21). The onset of aseptic meningitis usually occurs 2 to 3 weeks after vaccine administration with a median of 23 days (range, 18 to 34 days) (220; 21; 365a). In November 2006, the Global Advisory Committee on Vaccine Safety (GACVS) reviewed adverse events following mumps vaccination with special reference to the risk of vaccine-associated aseptic meningitis; GACVS concluded that all available mumps vaccine preparations are acceptable for use in immunization programs but recommended implementing appropriate strategies for risk communication and case management if higher-risk strains of mumps vaccine are being used in mass vaccination campaigns (21).

Rubella vaccines. Adverse reactions following vaccination with the RA27/3 rubella vaccine, whether monovalent or in fixed combinations, are generally mild, particularly in children (23). Common adverse reactions include irritability, myalgia, and paresthesias. Other reported neurologic side effects in fixed combination vaccines that include rubella vaccine appear to reflect the other components (eg, aseptic meningitis with mumps vaccine, and febrile convulsions with MMR/MMRV) (23).

Measles-Mumps-Rubella (MMR) and Measles-Mumps-Rubella-Varicella combination vaccines. The MMR vaccine is well-tolerated and rarely associated with serious adverse events (222). Headache is more common with the first vaccination than with revaccination (222). Expert committees at the Institute of Medicine concluded that the available evidence supports a causal relation between MMR vaccination and febrile seizures, and between MMR vaccination and measles inclusion body encephalitis in persons with demonstrated immunodeficiencies (222). Numerous studies in different countries have demonstrated no increased risk of permanent neurologic sequelae and no evidence to support an increased risk of Guillain-Barré syndrome or autism following administration of measles-containing combination vaccines (369).

MMR/MMRV vaccines and febrile convulsions. MMR vaccine is associated with an increased risk for febrile seizures during the first 2 weeks after vaccination when the peak in replication of the live-attenuated measles virus occurs (6 to 12 days) (41; 217; 209), but monovalent varicella vaccine at age 18 months was not associated with increased risk of febrile seizures (209). Among 12- to 23-month-olds who received their first dose of measles-containing combination vaccine, fever and seizure were elevated 7 to 10 days after vaccination (184). Across a range of studies, febrile seizures typically occurred 6 to 14 days after vaccination and were not associated with any recognized long-term sequelae (132; 105; 76; 41). The risk for febrile seizures is approximately 1 case for every 1,150 to 4,000 doses of MMR vaccine administered, with most studies in the range of 1 case for every 3,000 to 4,000 doses of MMR vaccine administered (105; 41; 365a). Children with a personal or family history of febrile seizures or family history of epilepsy might be at increased risk for febrile seizures after MMR vaccination (48; 346).

For unclear reasons, a first MMRV vaccine dose in children aged 10 to 24 months is associated with an elevated risk of seizure or febrile seizure beyond that of the component vaccines (184; 211; 333; 292; 208). Indeed, the risk of febrile seizures is more than 2 times higher in children aged 12 to 60 months who were given a first dose of MMRV than those given separate MMR and varicella vaccinations (167) and approximately 2 to 3.5 times higher in children aged 11 to 23 months who were given MMRV than those given separate MMR and varicella vaccinations (184; 292). Nevertheless, the absolute level of risk is small (184; 211). Among 12- to 23-month-olds who received their first dose of measles-containing combination vaccine, vaccination with MMRV resulted in 1 additional febrile seizure for every 2,300 to 2,600 doses given compared with vaccination with separate MMR and varicella vaccines (167; 184). Given the low number of MMRV-specific febrile seizure cases, their transient and benign nature, and the benefits of vaccination (including improved vaccine coverage), the MMRV vaccine continues to be recommended, although some now include the option to separately immunize with MMR and varicella vaccines (184; 217; 120; 211). MMRV as the second dose of measles-containing vaccine did not carry an increased risk of febrile seizures in children aged 11 to 23 months (210). MMRV administration to older children does not appear to carry the same excess risk: MMRV and the combination of separate MMR and varicella vaccines were not associated with increased risk of febrile seizures among 4- to 6-year-olds (185).

In 2010, the Advisory Committee on Immunization Practices adopted new recommendations regarding use of MMRV vaccine and identified a personal or family (sibling or parent) history of seizure as a precaution for use of the MMRV vaccine. For the first dose of measles, mumps, rubella, and varicella vaccines at age 12 to 47 months, either MMRV vaccine or the combination of MMR vaccine and varicella vaccine may be used. However, in the absence of a parental preference for MMRV vaccine, the CDC recommends that the MMR and varicella vaccines should be administered for the first dose in this age group. For the second dose of measles, mumps, rubella, and varicella vaccines at any age (from 15 months to 12 years) and for the first dose administered at age of 48 months or older, MMRV vaccine is preferred over separate injections of its component vaccines (217).

The MMR vaccine and autism controversy. In 1998 the Lancet published a paper by British surgeon Andrew Wakefield and colleagues that suggested a temporal connection between the measles, mumps, and rubella vaccine and subsequent development of nonspecific colitis and behavioral symptoms consistent with autism spectrum disorder and "possible postviral or vaccinal encephalitis" (349). Despite uncontrolled observations in a small number of cases (n=12), the authors suggested that the MMR vaccine can serve as a trigger for the development of autism, a claim that garnered extensive media attention and resulted in widespread public anxiety over vaccine safety. Wakefield reiterated this claim in 1999 (349).

Attempts to evaluate the Wakefield hypothesis produced uniformly negative results (243; 162; 163; 317; 321; 214; 263; 328; 357; 308; 156; 337; 222; 365a). Indeed, an early review of national data from the United Kingdom showed that the incidence of autism started rising over a decade before the introduction of the MMR vaccine into the country, and furthermore the trend in autism incidence did not change on introduction of the MMR vaccine (243). In 2001, the US Institute of Medicine, at the request of the CDC and the National Institutes of Health, convened an independent committee to critically review available information; the committee concluded that available evidence "favors rejection of the causal relationship at the population level between MMR vaccination and ASD [autism spectrum disorder]" (317). Subsequent systematic reviews of epidemiological evidence also found no support for a causal relationship between MMR vaccine and autism (321; 357). An independent review commissioned by the World Health Organization also concluded that “existing studies do not show evidence of an association between the risk of autism or autistic disorders and MMR vaccine”; based on the extensive review presented, the Global Advisory Committee on Vaccine concluded that “no evidence exists of a causal association between MMR vaccine and autism or autistic disorders” (360). The Institute of Medicine concluded that available evidence favors rejection of a causal association between MMR vaccine and autistic spectrum disorders/autism (162; Institute of Medicine 2003; 222).

In 2004, investigative journalist Brian Deer published compelling evidence that raised the possibility of research fraud, unethical treatment of children, and financial conflict of interest by Wakefield (195). Despite a rebuttal from Wakefield (350; 351), Deer’s report prompted an official inquiry by the UK General Medical Council, which ultimately (in 2010) found Wakefield "irresponsible" and guilty of unethical behavior involving study participant selection criteria, accepting funding from the Legal Aid Board, and misrepresenting key facts about the study. Wakefield was stricken from the United Kingdom medical register, and his 1998 paper was retracted by the Lancet because it was shown to be fraudulent. Nevertheless, for more than a decade after its publication (and since), Wakefield’s paper severely damaged public acceptance of important health-promoting disease-prevention efforts, undermined faith in medical research, and resulted in huge expenditures for litigation and reparative health promotion efforts (195). Deer later published further information concerning the extent of Wakefield’s fraudulent claims and his brazen efforts to use his self-generated fraud for financial gain (88; 89; 90).

Polio vaccines.

Inactivated polio vaccine. Beginning around 1951, Jonas Salk and his team of researchers in Pittsburgh utilized the Enders-Weller-Robbins method to grow poliovirus for development of an inactivated-virus polio vaccine.

They initially focused on optimizing the virus cultivation process, including selecting the optimal nonneural monkey tissue for growing large quantities of virus, which proved to be monkey kidney tissue. By 1952, Salk’s team had begun immunization experiments in monkeys using polioviruses inactivated with formalin to identify the viral strains that would induce the optimal antibody response and to determine the optimal use of formalin to inactivate the virus without compromising its ability to stimulate production of protective antibodies. In 1952, after a year of trials with animal immunizations, Salk began human studies of the antibody response to inactivated poliovirus immunization among children; the results from the first 161 subjects were published in 1953. Then, in anticipation of a large national trial, Salk conducted community-based pilot trials in 1953 and 1954. The Salk vaccine was field tested in 1954 in a huge clinical trial funded by the National Foundation for Infantile Paralysis, through the “March of Dimes” public fundraising campaign developed by President Franklin Delano Roosevelt and his law partner, D. Basil O’Connor (192; 196). The 1954 Field Trial of the Salk vaccine was the largest public health experiment ever conducted, involving 1.8 million children (labeled "polio pioneers") who were inoculated with either vaccine or placebo or were simply observed. On April 12, 1955, at a press conference in Ann Arbor, epidemiologist and virologist Thomas Francis Jr. declared that the Salk inactivated polio vaccine was both safe and effective. That same afternoon, an advisory committee to the Laboratory of Biologics Control, the federal agency that was responsible for licensing biological products, recommended that vaccine licenses be granted to 5 pharmaceutical companies.

The "Cutter incident". Shortly after the 1954 field trial of the Salk inactivated polio vaccine, manufacturing difficulties with clumping of material and inadequate formaldehyde inactivation of virus during large-scale processing resulted in a large outbreak of iatrogenic paralytic poliomyelitis (the so-called "Cutter incident") (262; 241; 150; 248; 249; 192; 196). Within 2 weeks of the 1955 report that the Salk vaccine was safe and effective, 5 cases of paralytic polio were reported in the inoculated arms of patients receiving vaccine manufactured by Cutter Laboratories of Berkeley, California. All unused Cutter vaccine was promptly recalled, but by then at least 220,000 people were infected with live poliovirus: approximately 70,000 developed muscle weakness, 164 developed severe paralysis, and 10 died. Subsequent investigation found that small-scale manufacture of inactivated polio vaccine produced predictable inactivation of poliovirus, but large-scale production produced inconsistent inactivation at all 5 companies producing the vaccine. Furthermore, the test procedures that had been used to ensure that vaccine had been inactivated were found to be insensitive to residual live virus. Clumping of cellular debris had prevented adequate exposure of the virus to formaldehyde. Federal requirements for the manufacture of poliovirus vaccine were revised to require a filtration step to ensure complete inactivation of virus.

Vaccine-associated poliomyelitis. In 1957, Albert Sabin developed a trivalent live-attenuated polio vaccine using the time-consuming process of infecting monkeys with the poliovirus (192; 197).

Because the inactivated-virus vaccine developed by Salk and colleagues in Pittsburgh had already been successfully introduced in the United States in 1954, Sabin’s oral vaccine was tested in Russia and was ultimately licensed in the United States in 1963. The live-attenuated vaccine soon thereafter became the polio vaccine of choice because it was less costly, required minimal training to administer (ie, orally rather than by injection), prevented the disease carrier state, and helped prevent the spread of wild poliovirus. However, by then rates of polio in the United States had already dropped from tens of thousands per year to just 50 to 100 cases per year because of the Salk vaccine; consequently, the Sabin oral polio vaccine had a relatively limited impact on overall polio incidence in the United States, although it did have an important role in other countries. By the early 1970s, remaining incident cases of paralytic poliomyelitis in the United States were almost exclusively either imported cases or caused by the vaccine itself. The Sabin oral polio vaccine was discontinued in the United States in 2000 because by then the continued risk of vaccine-associated paralytic poliomyelitis outweighed the potential benefits of a live-virus vaccine (268).

Following oral polio vaccine, the vaccine strains of poliovirus replicate in the gut of vaccine recipients and may be excreted for 4 to 6 weeks after vaccination. During this time, the few attenuating mutations in the vaccine strains may revert so that the virus reacquires its previous virulence. Reversion of attenuating mutations is the underlying cause of vaccine-associated paralytic poliomyelitis (VAPP) in oral polio vaccine recipients and their close contacts. Symptoms of VAPP usually develop 4 to 30 days after oral polio vaccine or within 4 to 75 days after contact with an recipient of the vaccine (366).

VAPP incidence is higher with the first dose of oral polio vaccine than with subsequent doses (366). The first-dose risk of VAPP among recipients of the vaccine was estimated at 1 case per 1.4 million children in the United States and 1 case per 2.8 million children in India (366). VAPP is more common in individuals who are immunocompromised (366). In industrialized countries, VAPP occurs mainly in early infancy associated with the first dose of oral polio vaccine, and incidence decreases sharply (more than 10-fold) with subsequent doses, whereas in low-income countries VAPP may occur with second or subsequent doses of oral polio vaccine, with cases concentrated among children aged 1 to 4 years (27). Transmission from 1 VAPP case resulting in secondary VAPP cases has not been demonstrated (366).

Vaccine-derived poliovirus (VDPV). The attenuated viruses in oral polio vaccine may, through replication (in an individual or a community), reacquire the neurovirulence and transmissibility characteristics of wild-type polioviruses (27). A vaccine-derived poliovirus is an oral polio vaccine strain that has reverted to a virulent form and has developed the capacity for sustained circulation. VDPVs differ from the parental Sabin strains found in oral polio vaccine by 1% to 15% of their nucleotides, a measure of genetic change. VDPVs are categorized as circulating (cVDPV), immunodeficiency-associated (iVDPV), or ambiguous (aVDPV) when isolated from people with no known immunodeficiency or sewage isolates of unknown origin (27). cVDPVs were first recognized in 2000 during an outbreak in Hispaniola (178). In under-immunized populations, VDPVs can circulate and can cause paralysis (27). cVDPVs can be imported and spread in any under-vaccinated community within an otherwise well-vaccinated developed country, as occurred in an Amish community in Minnesota in 2005 (after oral polio vaccine had been discontinued in the United States in 2000) (09). Most cVDPVs (more than 90%) are type 2 poliovirus (27). Persistent cVDPVs, or VDPVs that continue to circulate for more than 6 months following detection, represent programmatic failures to contain an outbreak (27). Oral polio vaccine protects against VDPVs and is used to contain outbreaks. A fully immunized population will be protected against both vaccine-derived and wild polioviruses.

Aseptic meningitis and encephalitis. Aseptic meningitis and encephalitis have been reported rarely after oral polio vaccine, particularly in immunodeficient infants (366).

Guillain-Barré syndrome. In the 1980s, studies in Finland suggested an increased incidence of Guillain-Barré syndrome following mass oral polio vaccine vaccination (181; 338; 19). Other factors were later found to have contributed to the increase in Guillain-Barré syndrome incidence in Finland at that time, including an influenza epidemic (182). Subsequent data do not indicate an increased risk of Guillain-Barré syndrome following receipt of oral polio vaccine (271; 18; 67; 182).

Provocation poliomyelitis. In persons incubating wild poliovirus infection, intramuscular injections (eg, other vaccines) may provoke paralysis in the injected limb (322).

Siman papovavirus 40. From 1954 to 1962, both the inactivated and live-attenuated forms of polio vaccine were prepared in primary cultures of rhesus monkey kidney cells, some of which were derived from monkeys that were naturally infected with simian papovavirus 40 (SV40) (366). SV40 is an oncogenic DNA virus that can induce primary brain cancers in animals, and viruses from the same family are oncogenic in humans. However, long-term follow-up studies did not support a causal association between the receipt of oral polio vaccine and the subsequent development of malignancies in humans (59). Oral polio vaccine is now tested for SV40, and none has been found positive (366).

Rabies vaccines. There are 3 main types of rabies vaccine: (1) nerve-tissue vaccines (now outdated), (2) cell-culture vaccines, and (3) embryonated-egg vaccines. Cell-culture and embryonated-egg vaccines have replaced nerve tissue vaccines in industrialized countries and are the ones recommended by the World Health Organization; they are considered safe and well tolerated (363).

Earlier inactivated nerve-tissue rabies vaccines frequently induced severe adverse reactions (eg, encephalitis, encephalomyelitis, and myelitis) and were less immunogenic than modern, concentrated, purified cell culture and embryonated egg-based rabies vaccines (CCEEVs) (201; 363; 370). Nevertheless, they are still used in a limited and decreasing number of developing countries (363). Severe adverse events with nerve-tissue vaccines were mainly neurologic, typically occurred about 2 weeks after vaccination, and were caused by an autoimmune response (possibly triggered by myelin basic protein contained in the vaccine) (145; 146; 264). These included: (1) meningoencephalomyelitis, (2) meningomyelitis or transverse myelitis (which results in lower limb paralysis, decreased sensation and sphincter disturbance), (3) meningoradiculitis, (4) mononeuritis multiplex (which may affect cranial nerves, particularly II, III, VII, IX, or X), and (5) ascending paralysis of the Landry type (ie, sudden onset of flaccid paralysis of the legs, followed by paralysis of the arms, usually developing between 1 to 2 weeks after the first injection). Historically, with the Semple nerve-tissue vaccine (prepared from rabies virus-infected goat or sheep brain tissue), the incidence of neurologic complications varied from 0.14 to 7 per 1,000 cases per vaccine recipient with a case fatality rate of up to 10% (38; 363). The Fuenzalida-type vaccine (prepared from suckling mouse brain tissue with a decreased myelin content compared with the Semple vaccine) was reported to cause neurologic complications in 0.12 to 0.037 per 1000 courses (Noguiera 1998; 363) with a case fatality rate of 22% (363). In 1 study of 18 patients with neurologic complications after administration of the Fuenzalida vaccine, only 1 had Guillain-Barré syndrome, 1 had meningoradiculitis, 4 had myelitis, and 12 had diffuse involvement of the nervous system, especially of the spinal cord and meninges (meningomyelitis and meningoencephalomyelitis) (34).

Because of the problems with inactivated nerve-tissue rabies vaccines, the World Health Organization has since 1984 strongly recommended their replacement by inactivated CCEEVs (370). Mild systemic adverse events following rabies immunization with inactivated CCEEV vaccines, such as transient headache and dizziness, have been observed in 5% to 15% of vaccine recipients (361; 370). In general, rabies vaccines with smaller doses and more advanced processing techniques are relatively safer, especially for young children (260). From October 1997 through December 2005, the United States Vaccine Adverse Event Reporting System (VAERS) received 336 reports of adverse events after vaccination with the Purified Chick Embryo Cell vaccine against rabies: among these there were 13 neurologic events, representing 4% of reports, without a clear pattern to suggest a plausible relationship to vaccination (99). Although serious neuroparalytic reactions during and after the administration of CCEEV rabies vaccines are rare (49; 168; 99; 216; 361; 370), such reactions, when they occur, pose a serious dilemma for the patient and the attending physician because a decision to discontinue the vaccination series must be weighed against the risk for acquiring rabies (216).

Smallpox vaccine. Encephalitis as a complication of smallpox vaccination was initially recognized in 1905 (279), more than a century after Edward Jenner performed his first smallpox vaccination (on 8-year-old James Phipps) on 14 May 14, 1796.

Since then, the estimated incidence of encephalitis and other neurologic complications of smallpox vaccination has varied considerably over time and across countries (186). Some of the variation reflects use of different strains of virus in the vaccines: vaccination with the New York City Board of Health strain is associated with a low rate of postvaccination encephalitis (3 cases per 1 million vaccinations) whereas the Lister and Copenhagen strains are associated with intermediate rates (26 and 33 cases per 1 million vaccinations, respectively), and the Bern strain is associated with very high rates (45 cases per 1 million vaccinations). Some additional variation in reported rates over time is explained by changes in vaccine production methods, quality control, vaccination procedures, health care, and case ascertainment (eg, better ascertainment procedures may have reduced the number of adverse events misdiagnosed as postvaccination encephalitis) (186). Vaccine postmarketing surveillance has suggested no association between optic neuritis and receipt of smallpox vaccinations in the United States military (257), although this continues to be intermittently raised as a potential issue (258).

Yellow fever vaccines. A French neurotropic yellow fever vaccine was discontinued in 1982 due to the high frequency of neurologic adverse effects (313).

Modern yellow fever vaccines are usually fairly well tolerated. In clinical trials, headache and myalgia were among the mild clinical symptoms that were commonly reported post vaccination (25). With the 17D-204 strain vaccine, YF-VAX (manufactured outside of Brazil, including in the United States), there were 4.7 serious adverse effects reported per 100,000 doses in the period from 2000 to 2006 (313). Reporting rates of serious adverse effects were highest among persons aged 60 years and older (8.3 per 100,000 doses) (313; 25). YEL-AND (YELlow fever vaccine-Associated Neurologic Disease) is 1 of the most common, well-characterized categories of serious adverse effects of vaccination with the vaccine (313). All reported cases of YEL-AND have been in primary vaccine recipients (ie, first dose of vaccine), possibly due to viremia that occurs after the first dose of vaccine (which has not been documented in persons receiving a booster dose of vaccine) (25).

YEL-AND encompasses several distinct clinical syndromes: (1) meningoencephalitis (neurotropic disease), (2) Guillain-Barré syndrome, (3) acute disseminated encephalomyelitis, and (4) bulbar palsy (183; 223; 313; 25). Meningoencephalitis develops from direct viral invasion of the central nervous system by the vaccine virus, with infection of the meninges or the brain. The other neurologic syndromes encompassed by YEL-AND represent autoimmune reactions in which vaccine-induced antibodies or T-cells cross-react with the host’s neuronal epitopes, leading to central-nervous-system or peripheral-nerve damage.

The viral strains used to develop yellow fever vaccines can develop neurovirulence, either with too few or too many passages during the attenuation process (115; 242). To minimize this risk, a vaccine seed lot system was developed in 1945 to clearly define the number of passages allowed for particular strains, but still additional cases of vaccine-associated encephalitis continued to be reported (313). In addition, studies in the 1950s found that vaccine-associated encephalitis occurred at extremely high rates in infants younger than 6 months of age (ie, 4 cases per 1,000 children vaccinated in this age group), so in the 1960s use of yellow fever vaccine was restricted to individuals who were at least 6 months old (16; 313). Subsequently, the number of cases of vaccine-associated encephalitis diminished in infants. Now, age less than 6 months continues to be a contraindication for yellow fever vaccination; additionally, vaccination is discouraged for infants aged 6 to 8 months except during epidemics when the risk of yellow fever virus transmission may be very high (25).

The reported rate for YEL-AND with the 17D-204 strain vaccine are generally 0.25 to 0.8 per 100,000 vaccine doses (180; 204; 205; 25). From VAERS data, the reporting rate for YEL-AND is 0.4 to 0.8 cases per 100,000 doses distributed (180; 204; 205; 313). Higher rates were reported in the elderly: 1.6 cases per 100,000 doses distributed among persons aged 60 to 69 years and 1.1 to 2.3 cases per 100,000 doses distributed among persons aged 70 years and older. These values are likely underestimates, given underreporting to VAERS. Reported rates of YEL-AND with postmarketing surveillance reports in Brazil using the 17DD YF strain vaccine were somewhat higher, at 1.1 per 100,000 doses distributed (25) whereas rates during preventive mass vaccination campaigns in West African countries were extremely low, only 0.016 cases per 100,000 doses of vaccine administered (25), likely indicating incomplete ascertainment.

Among 29 YEL-AND cases that occurred from 1990 to 2006 were 9 of meningoencephalitis, 8 of Guillain-Barré syndrome, 3 of acute disseminated encephalomyelitis, 1 of bulbar palsy, and 8 cases with insufficient data to classify (177; 183; 223; 313). The cases had a varied period from vaccination to symptom onset (range: 3 to 28 days) and a varied age range (6 to 78 years). All of the cases occurred in first-time vaccine recipients.

Fifty cases of YEL-AND were reported from 3 tertiary referral centers in São Paulo, Brazil, during 2017 to 2018: 32 had meningoencephalitis (14 with reactive yellow fever IgM in cerebrospinal fluid), 2 died, and 1 may have transmitted infection to an infant through breast milk (276). Of 7 cases of autoimmune neurologic disease after yellow fever vaccination, 3 had Guillain-Barre syndrome , 2 had ADEM, and 2 had myelitis. Novel potential vaccine-associated syndromes identified include autoimmune encephalitis, opsoclonus-myoclonus-ataxia syndrome, optic neuritis, and ataxia.

Cases continue to reveal novel or diagnostically confusing presentations. For example, a case of longitudinal myelitis with onset 45 days after yellow fever vaccination was reported in a 56-year-old man who received no other vaccines (74). Yellow fever virus IgM antibodies were detected in his cerebrospinal fluid and serologic testing for other regional flaviviruses was negative. However, the case is atypical because the time to symptom onset was considerably longer than that reported for other YEL-AND cases. Another case was a 23-year-old man who was diagnosed originally with meningoencephalitis but was later determined to have acute disseminated encephalomyelitis based on magnetic resonance imaging of the brain (232).

YEL-AND is rarely fatal (15; 313). In 1 fatal case of encephalitis in a 3-year-old girl who was vaccinated against yellow fever in 1965, molecular studies of the virus isolated from the brain later demonstrated that the vaccine virus had mutated, and monkey studies demonstrated that the virus had become more neurovirulent (15; 169); this is the only documented case in which a mutation of the yellow fever vaccine virus has been linked to an adverse outcome in a recipient (although this has been reported numerous times with the live-attenuated polio vaccine). Another fatal case was reported in a 53-year-old man who had unrecognized asymptomatic human immunodeficiency virus (HIV) infection and a CD4 count of less than 200/mm3 (177); the man developed fever and malaise 3 days after vaccination, became encephalopathic by day 5, and died 9 days after vaccination.

During the 2001 yellow fever mass vaccination campaign in Juiz de Fora, Brazil, 12 cases of aseptic meningitis were temporally associated with yellow fever vaccination, but a causal relationship with the vaccine was never established because "clinical and laboratory data were not available to confirm nor deny causality" (107).

SARS-CoV-2 vaccines. The safety of the various SARS-CoV-2 vaccines in use around the world (eg, mRNA vaccines, subunit protein vaccine, inactivated virus vaccines, DNA vaccine, recombinant adenovirus type-5-vectored vaccine, recombinant spike protein nanoparticle vaccine) has generally been excellent (12; 111; 166; 175; 176; 189; 207; 265; 331; 353; 374; 380; 381; 37; 47; 52; 75; 270; 287; 327; 329; 348; 375; 379; 35; 96; 165; 202; 245). A systematic review and metaanalysis found that significantly more adverse events were reported in vaccine groups compared with placebo groups of various COVID-19 vaccine trials, but the rates of reported adverse events in the placebo arms were still substantial; headache and fatigue were the most common adverse effects in this study (138). In general, the most common adverse reactions after COVID-19 vaccination are injection site pain, fever, headache, fatigue or malaise, and myalgias, which were mild and self-limiting. A prospective study in Mexico identified very low frequency of potential adverse neurologic events after the Pfizer COVID-19 vaccination (118).

The dozens of anecdotal reports of supposed "complications" of COVID-19 vaccination, and narrative, systemic, or other reviews of such case reports that do not properly consider causation from an epidemiologic standpoint (10; 11; 95; 108; 73) do not withstand scrutiny; these include reviews labeled "systematic" that tabulate case reports and have no control group (11; 95; 234; 239; 251; 332; 73). Some case reports present caveats that the complication in question occurred after vaccination but cannot be proven to be caused by vaccination, or alternatively that " a coincidental relationship with this inactivated vaccine cannot be excluded." Others, despite such caveats, inappropriately proffer their respective cases as a "possible rare side effect," a "possible complication," a "possible neurological side effect," a "possible causal link," or an event "that may have been triggered by" or "possible [sic] caused by" vaccination. Rare events in an appropriate time window for vaccines can be used for hypothesis generation, but except in very unusual circumstances the causality of side effects cannot be established from individual events. Most published reviews of supposed neurologic complications have been of low methodological quality, have failed to consider standard criteria for vaccine complications or causality, and have accepted anecdotal reports as prima facie evidence of causality (10; 11; 95; 108; 119; 251; 332; 73). In addition, some anecdotal reports of supposed "complications" of vaccination represent functional or somatoform neurologic disorders (08; 123). Unfortunately, such dubious reports have nevertheless been widely reported in mainstream and social media, leading to public confusion and doubt, and adding to vaccine resistance or avoidance. Such anecdotal reports of vaccine complications are intentionally not cited in this chapter. Two key points are worth emphasizing: (1) the incidence of most of these side effects is not different between vaccinated and unvaccinated populations; and (2) the mortality and morbidity risks of COVID-19 infection far outweigh the risks associated with vaccination (103).

Currently, 4 vaccines are authorized and recommended in the United States to prevent COVID-19:

The Pfizer and Moderna vaccines are mRNA vaccines that give cells genetic instructions to make a harmless piece of the SARS-CoV-2 spike protein. After the protein piece is made, the cell breaks down the mRNA but displays the protein piece on its surface. The protein fragment is recognized as foreign by the immune system, which triggers an immune response and an antibody response like what happens in natural infection with COVID-19. For all 4 authorized vaccines, the most common side effects locally were pain, swelling, and redness, and systemically were chills, tiredness, and headache. Reactogenicity symptoms (side effects that happen within 7 days of getting vaccinated) are common but are mostly mild to moderate in severity. Systemic side effects (eg, fever, chills, tiredness, and headache) are more common after the second dose of the vaccine. Most side effects are mild to moderate.

The Johnson & Johnson vaccine uses a disabled adenovirus to deliver the instructions. However, according to the CDC, "In most situations, Pfizer-BioNTech, Moderna, or Novavax COVID-19 vaccines are recommended over the J&J/Janssen COVID-19 vaccine." This is because the Johnson & Johnson COVID-19 vaccine is less protective and has been associated with greater risks. Novavax COVID-19 vaccine is a subunit protein vaccine.

Selected other COVID-19 vaccines in use in other countries. The AstraZeneca or Oxford-AstraZeneca recombinant vaccine (ChAdOx1-S/nCoV-19; AZD1222; trade name "Vaxzevria," previously "COVID-19 Vaccine AstraZeneca") is not an mRNA vaccine, but instead uses a replication-deficient adenoviral vector vaccine against SARS-CoV-2. The vaccine expresses the SARS-CoV-2 spike protein gene, which instructs the host cells to produce the protein of the S-antigen unique to SARS-CoV-2, which stimulates the body to generate an immune response and to retain that information in memory immune cells. Although this vaccine is not approved for use in the United States, it has been used by numerous countries in Europe (eg, Finland, France, Germany, Italy, Netherlands, Spain, Sweden, etc.), Asia (eg, Indonesia, South Korea, Thailand), as well as Canada and Australia; some have suspended its use due to concerns about clotting disorders, however.

CoronaVac, also known as the Sinovac COVID-19 vaccine, is an inactivated virus COVID-19 vaccine developed by the Chinese company Sinovac Biotech. Unlike the mRNA vaccines approved in the United States, CoronaVac relies on traditional technology.

Cerebral venous sinus thrombosis with thrombocytopenia. Cerebral venous sinus thrombosis (CVST) with “vaccine-induced immune thrombotic thrombocytopenia” (VITT) was described in Europe following receipt of the AstraZeneca COVID-19 vaccine, which uses a chimpanzee adenoviral vector. Some have since employed the term thrombosis and thrombocytopenia syndrome (TTS) instead of vaccine-induced immune thrombotic thrombocytopenia because thrombosis and thrombocytopenia syndrome does not assume that the temporal sequence is necessarily causal.

In the United States, after 6 cases of cerebral venous sinus thrombosis with vaccine-induced immune thrombotic thrombocytopenia were recognized among approximately 7 million vaccine recipients of the Johnson and Johnson COVID-19 vaccine, there was a temporary national pause in United States vaccination with this product on April 13, 2021.

In a summary of reports of cerebral venous sinus thrombosis with vaccine-induced immune thrombotic thrombocytopenia after vaccination with the Johnson and Johnson COVID-19 vaccine in the United States, 12 U.S. patients were reported to the Vaccine Adverse Event Reporting System from March 2, 2021 to April 21, 2021, with follow-up through April 21, 2021 (296). All the cases were adult women, with 11 patients younger than 50 years of age and the remaining cases younger than 60 years of age. Seven cases had at least 1 risk factor for cerebral venous sinus thrombosis, including obesity (n=6), hypothyroidism (n=1), and oral contraceptive use (n=1). None had documented prior heparin exposure. Time from vaccination to symptom onset ranged from 6 to 15 days. Eleven cases initially presented with headache, whereas the remaining case initially presented with back pain and later developed headache. Seven of the 12 cases also had intracerebral hemorrhage; 8 had other venous thromboses, including portal vein thrombosis, internal jugular vein thrombosis, lower extremity deep venous thrombosis, and pulmonary emboli. After diagnosis of cerebral venous sinus thrombosis, 6 patients initially received heparin treatment. The platelet nadir ranged from 9,000/μL to 127,000/μL, with normal platelet levels typically considered between 150,000 and 450,000 per microliter of blood. All 11 patients tested for the platelet factor 4 antibody (as occurs in heparin-induced thrombocytopenia or HIT) by ELIZA screening had positive results, although functional platelet HIT antibody test results were positive in only 1 out of 9 (11%) of the U.S. cases with results available. All patients were hospitalized, including 10 in an intensive care unit. As of April 21, 2021, 3 of the patients had died with intracerebral hemorrhage.

In a later systematic review and metaanalysis of cerebral venous sinus thrombosis after vector-based COVID-19 vaccines (ie, Johnson & Johnson vaccine available in the United States and the AstraZeneca vaccine available in Europe), Palaiodimou and colleagues identified 69 studies comprising 370 patients with cerebral venous sinus thrombosis among 4182 patients with any thrombotic event temporally associated with COVID-19 vector-based vaccine administration (254); 23 of these studies were included in the quantitative metaanalysis. Among thrombosis and thrombocytopenia syndrome cases, the pooled proportion of cerebral venous sinus thrombosis was 51%. Thrombosis and thrombocytopenia syndrome was independently associated with a higher likelihood of cerebral venous sinus thrombosis when compared to patients with thrombotic events but without thrombosis and thrombocytopenia syndrome after vaccination (odds ratio 13.8). The pooled mortality rate of thrombosis and thrombocytopenia syndrome was 28%, whereas the pooled mortality rate of thrombosis and thrombocytopenia syndrome-associated cerebral venous sinus thrombosis was 38%. Thrombotic complications developed within 2 weeks of exposure to vector-based SARS-CoV-2 vaccines with a mean interval of 10 days. Thrombotic complications affected predominantly women under age 45, even in the absence of prothrombotic risk factors (eg, birth control pills). Women comprised 71% of the cases with thrombosis and thrombocytopenia syndrome, 75% of the cases with cerebral venous sinus thrombosis, 75% of the cases with thrombosis and thrombocytopenia syndrome-associated cerebral venous sinus thrombosis, and 69% of all cases with thrombotic events. Women are apparently disproportionately susceptible to these complications following vaccination with adenovirus vector-based COVID-19 vaccines; however, the pre-pandemic incidence rates of cerebral venous sinus thrombosis show only a slight female preponderance: women comprise three quarters of the cases following vaccination with adenovirus vector-based COVID-19 vaccines.

In a population-based cohort study, Ashrani and colleagues compared incidence rates of cerebral venous sinus thrombosis after Johnson & Johnson COVID-19 vaccination with the pre-pandemic incidence rate (33). Incident cases of cerebral venous sinus thrombosis were identified in Olmsted County, Minnesota, from January 2001 through December 2015. Sex-and age-adjusted incidence rates were adjusted to the 2010 United States census population. Then, VAERS data from February 28 to May 7, 2021 were used to estimate the incidence of cerebral venous sinus thrombosis after Johnson & Johnson COVID-19 vaccination. Estimates were calculated separately for 3 plausible periods during which individuals might be at risk: 15, 30, and 92 days. From 2001 through 2015, 39 Olmsted County residents developed cerebral venous sinus thrombosis: 74% had a predisposing risk factor, such as infection, cancer, or oral contraceptives. The median age at diagnosis was 41 years (range 22 to 84 years). There was only a slight female predominance, with women comprising 56% of the cases. The overall age- and sex-adjusted incidence rate was 2.3. Age-adjusted incidence rates were similar for women and men (2.5 and 2.3, respectively). As of May 7, 2021, approximately 8.7 million Johnson & Johnson vaccine doses had been administered in the United States. After excluding potentially duplicate reports and cases that were not objectively diagnosed, there were 38 objectively diagnosed cases within 92 days after vaccination (71% women). The median age was 45 years (range 19 to 75 years). The median time from vaccination to cerebral venous sinus thrombosis was 9 days (range 1 to 51 days); 82% occurred within 15 days and 95% within 30 days. The overall postvaccination incidence rate was 8.7 per 100,000 person-years at 15 days, 5.0 per 100,000 person-years at 30 days, and 1.7 per 100,000 person-years at 92 days. The 15-day postvaccination incidence rate for women was 13.0 per 100,000 person-years, approximately 3 times higher than the rate of 4.4 per 100,000 person-years for men. The postvaccination incidence rate among women was 5.1-fold higher than the pre-COVID-19 pandemic rate. This risk was highest among women aged 30 to 49 years. The authors concluded that in this population, the incidence rate of cerebral venous sinus thrombosis 15 days after Johnson & Johnson COVID-19 vaccination was significantly higher than the pre-pandemic rate. The higher rate of this rare adverse effect must be considered against the effectiveness of the vaccine in preventing COVID-19, with an absolute reduction of severe or critical COVID-19 of 940 per 100,000 person-years. However, the mRNA vaccines are even more effective and do not appear to carry the same excess risk of cerebral venous sinus thrombosis.

A self-controlled case series study using national data on COVID-19 vaccination and hospital admissions was conducted in England from December 1, 2020 to April 24, 2021 (151). Among approximately 30 million people vaccinated during this period, about 2 out of 3 were vaccinated with the AstraZeneca vaccine and 1 out of 3 were vaccinated with the Pfizer vaccine. The risk of cerebral venous sinus thrombosis was increased 4-fold after the AstraZeneca vaccine (incidence rate ratio 4.0, 95% confidence interval 2.1 to 7.7), 3.6 fold after the Pfizer vaccine (incidence rate ratio 3.6, 95% confidence interval 1.49 to 9.3 at 15 to 21 days), and after a positive SARS-CoV-2 test (1 to 7 days: incidence rate ratio 12.9, 95% confidence interval 1.9 to 89.6; 8 to 14 days: incidence rate ratio 13.4, 95% confidence interval 2.0 to 90.6).

A systematic review subsequently identified 41 cases of cerebral venous sinus thrombosis and vaccine-induced immune thrombotic thrombocytopenia after the AstraZeneca vaccine and 13 cases after the Johnson and Johnson vaccine (301). Most of the patients were women (72% of 32 cases after the AstraZeneza vaccine for whom gender information was reported, and all 13 of the cases after the Johnson and Johnson vaccine). Symptom onset occurred within one week after the first dose of vaccination (range 4 to 19 days). Headache was the most common presenting symptom. Intracerebral hemorrhage and/or subarachnoid hemorrhage were reported in half (49%) of the patients. The platelet count of the patients ranged from 5 to 127 cells×109/L. Anti-PF4 IgG assay and d-dimer were positive in most of the reported cases. At least 19 patients (39%) died due to complications of cerebral venous sinus thrombosis and vaccine-induced immune thrombotic thrombocytopenia.

In patients with cerebral venous sinus thrombosis prior to the COVID-19 pandemic, baseline thrombocytopenia was uncommon, and heparin-induced thrombocytopenia (HIT) and anti-PF4/heparin antibodies were rare (288).

A large international consortium (from 81 hospitals in 19 countries) conducted a cohort study of consecutive cases with cerebral venous sinus thrombosis within 28 days of SARS-CoV-2 vaccination between March 29, 2021 and June 18, 2021 (289). For reference, cases with cerebral venous sinus thrombosis between 2015 and 2018 were identified from an existing international registry. Cases were classified as having thrombosis with thrombocytopenia syndrome (TTS) if they had new-onset thrombocytopenia without recent exposure to heparin. Of 116 cases with cerebral venous sinus thrombosis after SARS-CoV-2 vaccination, 78 (67%) had thrombocytopenia syndrome, of whom 76 (97%) had been vaccinated with the AstraZeneca vaccine (and 1 each with the Johnson and Johnson vaccine and the Pfizer vaccines), whereas 38 (33%) had no indication of thrombocytopenia syndrome (53% with the AstraZeneca vaccine, 40% with the Pfizer vaccine, 5% with the Chinese CoronaVac vaccine, and 3% with the Moderna vaccine). The cases of cerebral venous sinus thrombosis after SARS-CoV-2 vaccination were predominantly women (80%) and the mean age was 48 years. The control group included 207 patients with cerebral venous sinus thrombosis before the COVID-19 pandemic (70% women; mean age 42 years). In-hospital mortality rates were 47% (95% CI, 37% to 58%) for cerebral venous sinus thrombosis with thrombocytopenia syndrome after SARS-CoV-2 vaccination, 5% (95% CI, 1% to 18%) for cerebral venous sinus thrombosis without thrombocytopenia syndrome after SARS-CoV-2 vaccination, and 4% (95% CI, 2% to 7%) after other vaccinations in the pre-pandemic control group. The mortality rate was 61% among patients in the thrombocytopenia syndrome group diagnosed before the condition garnered attention and 42% among patients diagnosed later.

As of 9 June 2021, 390 thrombosis and thrombocytopenia syndrome cases (140 CVST, 250 other major thromboembolic events with concurrent thrombocytopenia) occurred within 28 days following vaccination with the AstraZeneca vaccine in the United Kingdom where there has been a high early uptake: 67 were fatal after an estimated 24.6 million first doses (2.7 deaths/million first-dose vaccinations) (65).

Cerebral venous sinus thrombosis with thrombosis and thrombocytopenia syndrome after SARS-CoV-2 vaccination has a distinct clinical profile and a high case-fatality rate (289). Moreover, the FDA has determined cerebral venous sinus thrombosis with vaccine-induced immune thrombotic thrombocytopenia to be a plausible complication of COVID-19 vaccination (ie, plausibly caused by COVID-19 vaccination). Although many cases occur in women under 40 years of age, very few have been identified as taking birth control pills. Cases have occurred predominantly with the AstraZeneca vaccine, but significant numbers have occurred following both the Johnson and Johnson and the Pfizer mRNA vaccines (174; 289). A mechanism like that of autoimmune heparin-induced thrombocytopenia has been proposed. If clinicians identify such cases, diagnostic laboratories should be notified that thrombosis with thrombocytopenia syndrome is being considered to optimize testing methods. Anti-PF4 antibodies and functional platelet HIT antibody testing should be obtained (296; 305). Early recognition of this complication should prompt acute treatment that addresses the autoimmune and prothrombotic processes (305). Intravenous immunoglobulin (1 g/kg for 2 days), consideration of plasma exchange, and nonheparin anticoagulation (argatroban, fondaparinux) are recommended (305).

Using an observational study with a case-control design, Garcia-Azorin and colleagues assessed whether a headache following COVID-19 vaccine-related cerebral venous sinus thrombosis is associated with a higher probability of death or intracranial hemorrhage (117). The study population was individuals vaccinated with nonreplicant adenovirus vector-based vaccines (ie, the Johnson & Johnson vaccine available in the United States, and the AstraZeneca vaccine available in Europe). The authors reviewed (1) all the published cases and case series during March and April 2021, and (2) reports from the CDC and the European Medicines Agency providing patient-level data. Controls were identified from cases of “headache” following COVID-19 vaccination reported to the VAERS up to April 30, 2021. The statistical analysis assessed whether the presence of headache in patients with cerebral venous sinus thrombosis was associated with a higher probability of intracranial hemorrhage of death. Of the 77 identified cases, 90% were women. Headache was described in 49 and intracranial hemorrhage in 43%, and 25% died. The median time between vaccination and onset of cerebral venous sinus thrombosis-related headache was 8 days (interquartile range 7 to 10 days). The presence of headache was associated with a significantly higher odds of intracranial hemorrhage (odds ratio of 7.4), but not of death. Therefore, patients with a new-onset headache 7 to 10 days after vaccination with an adenovirus vector-based vaccine should be carefully evaluated to identify or exclude cerebral venous sinus thrombosis.

Krzywicka and colleagues estimated the absolute risk of cerebral venous sinus thrombosis with and without thrombocytopenia within 28 days of a first dose of 4 SARS-CoV-2 vaccinations using data from the European Medicines Agency’s EudraVigilance database and data on vaccine delivery from 31 European countries (188); the risk of cerebral venous sinus thrombosis with thrombocytopenia was considerably increased in patients receiving the AstraZeneca and slightly increased in patients receiving the Johnson & Johnson vaccine compared with the estimated background risk, but the risk was not increased in recipients of SARS-CoV-2 mRNA vaccines. The absolute risk of cerebral venous sinus thrombosis within 28 days of first-dose vaccination was 7.5 per million of first doses for the AstraZeneca vaccine (95% CI 6.9–8.3), 0.7 per million of first doses for the Johnson & Johnson vaccine (95% CI 0.2–2.4), 0.6 per million of first doses for the Pfizer vaccine (95% CI 0.5–0.7), and 0.6 for the Moderna vaccine (95% CI 0.3–1.1). The risk of cerebral venous sinus thrombosis with thrombocytopenia within 28 days of first-dose vaccination with ChAdOx1 nCov-19 was higher in younger age groups.

In Singapore, Tu and colleagues found cerebral venous sinus thrombosis was rare after mRNA-based SARS-CoV-2 vaccines, and the incidence rate of cerebral venous sinus thrombosis after SARS-CoV-2 infection was significantly higher than with mRNA-based SARS-CoV-2 vaccination (336).

Facial paralysis (Bell palsy). Data from the Pfizer-BioNTech and Moderna vaccine trials suggested an imbalance in the incidence of Bell palsy following vaccination compared with the placebo arm of each trial: among nearly 40,000 participants who received a vaccine, there were 7 Bell palsy cases, compared with only one among participants receiving placebo, a difference though that was not statistically significant (p=0·07) (253). However, because the 40,000 vaccine-arm participants were followed for a median of 2 months (roughly 6700 person-years of observation time), only 1 or 2 cases would be expected based on the estimated incidence rate of Bell palsy in the general population of from 15 to 30 cases per 100 000 person-years (253). Notably, the number observed in the placebo arms of the trials was consistent with what would be expected. Thus, the observed incidence of Bell palsy in the vaccine arms is between 3.5-times and 7-times higher than would be expected in the general population.

Peripheral facial nerve palsy has been widely suggested as a possible complication of the Pfizer vaccine. In the original efficacy trial published in December 2019, Bell palsy was reported in 4 cases among vaccinated participants but in none of the controls. Subsequent case reports and commentaries fostered speculation that the Pfizer vaccine is associated with an increased risk of Bell palsy. However, a case-control study in Israel found no association between recent vaccination with the Pfizer vaccine and risk of Bell palsy (302). Several factors made Israel an ideal place to conduct this study: (1) Israel exclusively uses the Pfizer vaccine; (2) all residents of Israel are obligatory members of a national digital health registry system; (3) in Israel it is standard practice to refer all new-onset cases of Bell palsy for evaluation in the emergency department. Patients with Bell palsy admitted in January and February of 2021 at the emergency department of a tertiary referral center in Israel were matched by age, sex, and date of admission with control patients admitted for other reasons (302). Thirty-seven patients were admitted to the emergency department for Bell palsy during the study period: 2 out of 3 were male, and the mean age was 51 years. Fifty seven percent of cases with Bell palsy had been recently vaccinated compared with 60% of matched controls with identical age, sex, and admittance date. The mean time from vaccination to occurrence of Bell palsy among cases was 9.3 days with a range of 3 to 14 days from the first dose and 14.0 days with a range of 1 to 23 days from the second dose. The adjusted odds ratio for exposure was 0.84 with a 95% confidence interval of 0.37 to 1.90; this was not statistically significant. For comparison, the number of admissions for Bell palsy during the months of January and February in preceding years from 2015 to 2020 ranged from 17 to 35 cases with a mean of 26.8 cases, a median of 27.5 cases, and with 35 cases in the immediately preceding year.

In a retrospective study, data were collected from 41 health care organizations worldwide to identify patients diagnosed with COVID-19 with or without a diagnosis code of Bell palsy within 8 weeks of the COVID-19 diagnosis (325). To account for vaccination, the queries were restricted to the period from January 1, 2021 to March 31, 2021. The authors matched 63,551 nonvaccinated patients with COVID-19 to those who were vaccinated against the disease and had no history of COVID-19 infection. Among these patients, the authors identified those with a history of Bell palsy. Approximately 348,000 patients with COVID-19 were identified. Of these patients, 284 (or 82 per 100,000 cases) were diagnosed with Bell palsy within 8 weeks of the initial COVID-19 diagnosis. One hundred fifty three (54%) of these had no prior history of Bell palsy, whereas 131 or 46% did have a prior history of Bell palsy. Among the 1525 patients with a prior history of Bell palsy before developing COVID-19, there was an 8.6% recurrence rate within 8 weeks of COVID-19 diagnosis. There was a statistically significant increased relative risk of 6.8 for a diagnosis of Bell palsy in those with COVID-19 compared with those who were vaccinated. The observed incidence of Bell palsy temporally associated with COVID-19 infection is 2.7 to 5.5 times higher than the baseline annual incidence of Bell palsy, which ranges from 15 to 30 per 100,000 people.

To assess whether COVID-19 vaccination is associated with development of postvaccination facial paresis, Renoud and colleagues used data from the World Health Organization pharmacovigilance database (274). The authors employed disproportionality analyses, which are hypothesis-generating methods that quantify the extent to which a drug-event combination occurs disproportionally compared with what would be expected in the absence of any association. They do not quantify risk because the population that is exposed to the drugs is unknown. The authors performed 4 analyses with 2 control groups and 2 facial paralysis definitions (broad and narrow); the first control group was comprised of individuals exposed to all other viral vaccines and the second was restricted to influenza vaccines. Analyses were adjusted for sex and age. When compared with other viral vaccines, mRNA COVID-19 vaccines did not show a signal of facial paralysis (274). The incidence of facial paralysis after mRNA COVID-19 vaccination was not higher than that observed with other viral vaccines (274). Therefore, any association between facial paralysis and mRNA COVID-19 vaccines is likely comparable to other viral vaccines.

In summary, data so far suggest that the rate of Bell palsy is 2.7 to 7 times more likely following mRNA vaccination than the baseline incidence of Bell palsy in the population, but these values are nevertheless comparable to rates occurring with other viral vaccines. The U.S. Food and Drug Administration has recommended monitoring vaccine recipients for facial paralysis.

In a case series and nested case-control study completed in Hong Kong, the age-standardized incidence of clinically confirmed Bell palsy was 66.9 cases per 100,000 person-years (95% CI 37.2 to 96.6) following CoronaVac (inactivated virus) vaccination and 42.8 per 100,000 person-years (95% CI 19.4 to 66.1) following the Pfizer vaccination (354). The age-standardized difference in incidence compared with the background population was 41·5 (95% CI 11.7 to 71.4) for CoronaVac and 17.0 (95% CI –6.6 to 40.6) for the Pfizer vaccine, equivalent to an additional 4.8 cases per 100,000 people vaccinated for CoronaVac and 2.0 cases per 100,000 people vaccinated with the Pfizer vaccine. In the nested case-control analysis, 298 cases were matched to 1181 controls, and the adjusted odds ratios were 2.4 (95% CI 1.4 to 4.0) for CoronaVac and 1.8 (95% CI 0.9 to 3.5) for the Pfizer vaccine. Only the Chinese CoronaVac vaccine was associated with a significantly increased risk of Bell palsy (354).

Using data from more than 17 million patients in England in a self-controlled case series design, Walker and colleagues found an increased rate of Bell palsy in the period from 4 to 28 days after receiving the first dose of AztraZeneca vaccine (7,783,441 vaccinees; incidence rate ratio 1.39; 95% CI 1.27-1.53), corresponding to 17.9 cases of Bell palsy per 1 million vaccinees if causal (352); as the authors noted, "The absolute risk, assuming a causal effect attributable to vaccination, was low." In contrast, among vaccinees who received the Pfizer and Moderna vaccines, there was no evidence of any association with Bell palsy (Pfizer: 5,729,152 vaccinees; incidence rate ratio 1.09; 95% CI 0.75-1.57; Moderna 255,446 vaccinees; rate ratio 0.88, 95% CI 0.32-2.42).

Sudden sensorineural hearing loss. Preliminary analyses of VAERS data in the early phase of societal COVID-19 vaccination using 2 messenger RNA vaccines (Phizer and Moderna) found no association between inoculation with a SARS-CoV-2 messenger RNA vaccine and incident sudden hearing loss (113).

In a subsequent cross-sectional study, an updated analysis of VAERS data and a case series of patients who experienced sudden sensorineural hearing loss after COVID-19 vaccination did not suggest an association between COVID-19 vaccination (Phizer, Moderna, or Johnson & Johnson) and an increased incidence of hearing loss compared with the expected incidence in the general population (114). An annualized incidence estimate of the post-vaccination incidence of sudden sensorineural hearing loss was 0.6 to 28.0 cases per 100,000 people per year, which is comparable to the annual incidence of idiopathic sudden sensorineural hearing loss (11 to 77 cases per 100 000 people per year, depending on age). The rate of incident reports of sudden sensorineural hearing loss was similar across all 3 vaccines (0.16 cases per 100,000 doses for both Pfizer-BioNTech and Moderna vaccines, and 0.22 cases per 100,000 doses for the Johnson & Johnson vaccine).

In a population-based, retrospective, cohort study using data from the largest health care organization in Israel, Yanir and colleagues found a small increased risk of sudden sensorineural hearing loss in individuals who received the Pfizer vaccine (376). The age- and sex-weighted standardized incidence ratios were 1.35 (95% CI, 1.09-1.65) after the first vaccine dose and 1.23 (95% CI, 0.98-1.53) after the second vaccine dose. The attributable risks were generally small, and the results were similar when 2019 was used as a reference to estimate the expected number of cases.

The analysis of United States VAERS data found that the incidence of sudden sensorineural hearing loss after vaccination with either mRNA vaccine (Phizer or Moderna) or the Johnson & Johnson vaccine did not exceed the background rate observed in the general United States population, but data from Israel revealed a slightly higher risk of sudden sensorineural hearing loss after vaccination with the Pfizer mRNA vaccine compared with expected rates (114; 339; 376). However, in the Israeli study, sudden sensorineural hearing loss occurred in fewer than 1 per 100 000 vaccinated individuals. Neither study reported the severity and duration of hearing loss or the clinical outcomes after treatment (339).

In a register-based country-wide retrospective cohort study of 5.5 million Finnish residents conducted from January 1, 2019 to April 20, 2022, the incidence of sudden sensorineural hearing loss following COVID-19 vaccination was compared with the incidence before the COVID-19 epidemic in Finland (244). Before the COVID-19 epidemic, 18.7 out of 100,000 people received a diagnosis of sudden sensorineural hearing loss annually. There was no increased risk for sudden sensorineural hearing loss following any COVID-19 vaccination. In particular, adjusted incidence rate ratios with 95% confidence intervals for the 3 doses of the Pfizer vaccine were 0.8 (95% CI, 0.6-1.0), 0.9 (95% CI, 0.6-1.2), and 1.0 (95% CI, 0.7-1.4), respectively.

Guillain-Barré syndrome. Guillain-Barré syndrome (GBS) has been recognized in anecdotal reports following vaccination for COVID-19, but such anecdotal reports cannot distinguish sporadic cases from those that are precipitated by vaccination. A significant increased risk of Guillain-Barré syndrome has been demonstrated following administration of the Johnson & Johnson COVID-19 vaccine, but not after administration of mRNA vaccines.

In the Johnson and Johnson COVID-19 vaccine trial, 2 subjects developed Guillain-Barré syndrome: 1 each in the placebo and active vaccine arms. Consequently, the incidence of Guillain-Barré syndrome was identical in both arms of the trial (218). From available evidence it is not possible to infer that this case was (or was not) caused by the coronavirus vaccine she received. The main reason for publishing this case was precisely because of its causally indeterminate nature, which is certainly an unusual reason for a case report. In the present circumstances, though, the authors were trying to raise awareness of the possibility of misattributing postvaccination cases of Guillain-Barré syndrome to vaccination. An increased risk of postvaccination complications can only be determined with comparative trials, meta-analyses of such trials, or large-scale epidemiologic studies. The authors pointed out that as countries collectively move closer to vaccinating 1 billion people, some 8000 to 19,000 cases of Guillain-Barré syndrome can be expected per year based on the background incidence of Guillain-Barré syndrome. Accounting for single- or dual-dose vaccine strategies, perhaps 1000 to 4000 cases might be anticipated in the weeks after vaccination, but not caused by the vaccination. Misattributing these cases to vaccination could undermine global efforts at combatting COVID-19.

A systematic review summarized anecdotal reports of Guillain-Barré syndrome after COVID-19 vaccination and concluded that "COVID-19-associated Guillain-Barré syndrome seems to share most features of classic post-infectious Guillain-Barré syndrome and possibly the same immune-mediated pathogenetic mechanisms," which could arise if the association is causal but which would certainly be the case if the observed association was not causal (02). Interestingly, another systematic review summarized anecdotal reports of Guillain-Barré syndrome after COVID-19 itself (ie, SARS-CoV-2 infection) (07). Neither anecdotal reports, nor case series, nor aggregated systematic reviews of such anecdotal reports can establish causality for Guillain-Barré syndrome caused by either the virus or the vaccine.

As part of postauthorization safety surveillance, the U.S. Food and Drug Administration identified a potential safety concern for Guillain-Barré syndrome following receipt of the Johnson & Johnson COVID-19 vaccine. To assess the risk of Guillain-Barré syndrome following the Johnson & Johnson COVID-19 vaccine, Woo and colleagues examined reports of presumptive Guillain-Barre syndrome from February 2021 to July 2021 in the U.S. VAERS passive reporting system (359). The observed-to-expected ratio of Guillain-Barré syndrome was calculated based on background rates and vaccine administration data. The comparator was the background rate of Guillain-Barré syndrome in the general (unvaccinated) population that had been estimated and published based on a standardized case definition. Because of limited availability of medical records, cases were not assessed according to the Brighton criteria for Guillain-Barré syndrome. As of July 24, 2021, 130 reports of presumptive Guillain-Barre syndrome were identified following vaccination with the Johnson & Johnson COVID-19 vaccine. With approximately 13,210,000 doses of vaccine administered to adults in the United States to this point, the estimated crude reporting rate of Guillain-Barré syndrome was 1 case per 100,000 doses administered. The overall estimated observed-to-expected rate ratio was 4.2 for the 42-day window. For both risk windows, the observed-to-expected rate ratio was elevated in all age groups, except for the age group from 18 to 29 years. The authors concluded that there is a small but statistically significant increased risk of Guillain-Barré syndrome following receipt of the Johnson & Johnson COVID-19 vaccine. Following availability of these results, the FDA warned in July 2021 that the Johnson & Johnson COVID-19 vaccine is associated with an increased risk of Guillain-Barré syndrome. This, and the lower preventive efficacy of the Johnson & Johnson vaccine against COVID-19, greatly curtailed use of this vaccine in the United States.

In a cohort study of COVID-19 vaccines in the Unites States, the incidence of Guillain-Barré syndrome was elevated after receiving the Johnson & Johnson vaccine (141). Hanson and colleagues used surveillance data on 10,158,003 participants aged at least 12 years from the Vaccine Safety Datalink at 8 integrated health care systems in the United States. The unadjusted incidence rate of Guillain-Barré syndrome was 32.4 per 100,000 person-years in the 1 to 21 days after vaccination with the Johnson & Johnson vaccine (95% CI, 14.8-61.5), significantly higher than the background rate of 1 to 2 per 100 000 person-years. The unadjusted incidence rate of Guillain-Barré syndrome was 1.3 per 100,000 person-years in the 1 to 21 days after mRNA vaccines (95% CI, 0.7-2.4). In a comparison of the Johnson & Johnson vaccine with mRNA vaccines, the adjusted rate ratio was 20.6 (95% CI, 7.0-64.7).

In a cohort study of 702 patients with a prior history of Guillain-Barré syndrome, only 1 needed short medical care for postvaccination relapse, which was considered a minimal risk (300).

Using data from more than 17 million patients in England in a self-controlled case series design, Walker and colleagues found an increased rate of Guillain-Barré syndrome in the period from 4 to 42 days after receiving the first (but not the second) dose of AztraZeneca vaccine (7,783,441 vaccinees; incidence rate ratio 2.85; 95% CI 2.33-3.47), corresponding to 11.0 additional cases of Guillain-Barré syndrome per 1 million vaccinees if causal (352). As the authors noted, "The absolute risk, assuming a causal effect attributable to vaccination, was low." In contrast, among vaccinees who received the Pfizer vaccine, there was no evidence of any association with Guillain-Barré syndrome (5,729,152 vaccinees; incidence rate ratio 1.09; 95% CI 0.75-1.57). Among the 255,446 recipients of Moderna vaccine, there were too few outcomes to investigate Guillain-Barré syndrome.

CNS demyelination. Various anecdotal reports of central nervous system demyelination after SARS-CoV-2 vaccination have appeared, most of which followed receipt of the AstraZeneca vaccine. So far, none of the studies have established an increased risk of CNS demyelination following any vaccine for SARS-CoV-2.

Using data from more than 17 million patients in England in a self-controlled case series design, Walker and colleagues found no evidence of a significant association with transverse myelitis for either the AztraZeneca vaccine (7,783,441 vaccinees; incidence rate ratio 1·51; 95% CI 0.96-2.37) or the Pfizer vaccine (5,729,152 vaccinees; incidence rate ratio 1.62; 95% CI 0.86-3·03). Among the 255,446 recipients of Moderna vaccine, there were too few outcomes to investigate transverse myelitis.

In a prospective, self-controlled, multicentric observational study, involving 25 Italian tertiary multiple sclerosis centers, Di Filippo and colleagues assessed whether mRNA COVID-19 vaccines increase the short-term risk of clinical relapses in multiple sclerosis (97). Individuals with multiple sclerosis (n=324; 84% with relapsing-remitting multiple sclerosis), diagnosed according to the 2017 McDonald criteria, who underwent the first dose of an mRNA COVID-19 vaccine in January 2021, were recruited from each participating center in Italy. Based on vaccine availability in Italy at that time, all of them received the Pfizer vaccine. The investigators recorded the presence, characteristics, and number of relapses in the 60 days after the first dose of the vaccine. In the 2 months before vaccination, 6 clinical relapses were reported in 6 of the 324 patients, giving a relapse rate of 1.9%. In the 2 months after vaccination, 7 clinical relapses occurred in 7 of the 324 patients, giving a relapse rate of 2.2%. The mean interval from the first dose of vaccination to the clinical relapse was 44 days. The relapse rates in the 2 months before and after vaccination were not significantly different. Age, gender, disease duration, and disability status had no effect on relapses. The authors concluded that the Pfizer COVID-19 vaccine does not increase the short-term risk of clinical reactivation in people with multiple sclerosis.

Herpes zoster. Multiple anecdotal reports of herpes zoster after the first and second doses of a COVID-19 vaccine have appeared in the medical literature in the United States and abroad.

In a cohort study using a self-controlled risk interval (SCRI) design applied to 1451 people who had herpes zoster during either a risk or control interval (among 2,039,854 vaccinated individuals), Akpandak and colleagues found no significant association between COVID-19 vaccination and herpes zoster (incidence rate ratio, 0.91; 95% CI, 0.82-1.01; P = .08) (06). In the supplementary cohort analysis, COVID-19 vaccination was not associated with a higher risk of herpes zoster compared with influenza vaccination in the prepandemic period (first dose of COVID-19 vaccine: hazard ratio, 0.78; 95% CI, 0.70-0.86; p < .001); second dose of COVID-19 vaccine: hazard ratio, 0.79; 95% CI, 0.71-0.88; p < .001]) or the early pandemic period (first dose of COVID-19 vaccine: hazard ratio, 0.89; 95% CI, 0.80-1.00; P = .05; second dose: hazard ratio, 0.91; 95% CI, 0.81-1.02; p = .09) (06).

Adverse pregnancy outcomes. Vaccination with mRNA vaccines administered during the second and third trimesters of pregnancy is not associated with an increased risk of adverse pregnancy outcomes (106; 179; 215). In a population-based retrospective cohort study of 157,521 pregnant women with singleton pregnancies in Sweden and Norway, COVID-19 vaccination against SARS-CoV-2 during pregnancy was not significantly associated with an increased risk of adverse pregnancy outcomes (215). Similarly, in a population-based retrospective cohort study of 97,590 of pregnant women in Ontario, Canada, COVID-19 vaccination during pregnancy was not significantly associated with an increased risk of adverse peripartum outcomes (106).

Anxiety-related adverse event clusters. On April 7, 2021, after 5 weeks of use of the Johnson and Johnson vaccine, the U.S. Centers for Disease Control and Prevention (CDC) received reports of clusters of anxiety-related events from 5 mass vaccination sites, all in different states (144). An anxiety-related event was defined as any of the following occurring in a person during the 15-minute postvaccination observation period: tachycardia, hyperventilation, dyspnea, chest pain, paresthesia, light-headedness, hypotension, headache, pallor, or syncope. Four of the 5 sites temporarily closed while an investigation took place. Overall, 64 anxiety-related events, including 17 reports of syncope, among 8624 vaccine recipients, were reported from these sites for vaccines administered during April 7 to 9, 2021. Approximately one quarter of the syncopal and other anxiety-related events occurred in persons who reported a history of similar events after prior vaccinations. Approximately half of the reports to the Vaccine Adverse Event Reporting System (VAERS) concerning syncope after the Johnson and Johnson vaccination were for persons in the youngest age group (18 to 29 years) recommended for vaccination, an age group prone to vasovagal reactions to needlesticks and vaccinations. Based on reports to VAERS, syncope occurred at a rate of 8.2 episodes per 100,000 doses following the Johnson and Johnson vaccine compared with 0.05 episodes per 100,000 doses after influenza vaccination.

Prevention

Among the vaccines available today, the cost-benefit analysis of vaccinations and complications strongly argues in favor of vaccination (233). Sadly, the perception of significant risk associated with vaccinations has limited the success of disease-eradication measures. Evaluating whether there is causal relationship between vaccinations and putative neurologic complications is essential if we are to confront the widespread public misperception of vaccination risk.

With mass vaccinations, significant numbers of adverse events can be anticipated (only some of which are directly attributable to the vaccine in question whereas many represent cases that would have occurred regardless of vaccination). All potential vaccine recipients need to be screened for contraindications to vaccination, including prior allergic reactions to the vaccine or its components. Any person with an evolving central nervous system disorder should generally not receive vaccination until the condition has resolved or stabilized. Individuals who are immunosuppressed or have an immunological disorder should generally not receive live-attenuated vaccines.

Because half of "serious" cases of postvaccination syncope occur within 5 minutes of vaccination, and more than two thirds occur within 15 minutes, the Advisory Committee on Immunization Practices and the Centers for Disease Control and Prevention recommend observing patients (keeping them sitting or lying down) for 15 minutes after they are vaccinated, and if syncope develops, patients should be observed until symptoms resolve (187; 72; 193).

In some cases, where neuroparalytic adverse reactions to vaccines are particularly common within a specific age group, or where 1 type of vaccine has a relatively higher incidence of such serious adverse reactions, CDC’s Advisory Committee on Immunization Practices and the WHO’s Global Advisory Committee on Vaccine Safety (GACVS) will provide guidance on vaccine selection or age-specific restrictions. For example, when it became clear that yellow fever vaccines produced an unacceptably high rate of vaccine-associated encephalitis in infants younger than 6 months of age, the vaccine was restricted to individuals at least 6 months old or older. Similarly, when alternatives to nerve-tissue rabies vaccines became available, the World Health Organization strongly recommended their replacement with safer inactivated CCEEV rabies vaccines.

Differential diagnosis

Confusing conditions

It remains unclear how many cases of reported encephalitis or meningoencephalitis following vaccination are simply coincidental in time but not due to the vaccination itself. A number of different infectious agents and noninfectious processes can be responsible for encephalitis, and it is often impossible to establish the etiology.

A variety of encephalitic syndromes can mimic those attributed to vaccination. The possible causes are numerous and include, but are not limited to, Epstein-Barr virus, herpes viruses, enteroviruses, measles, mumps, Mycoplasma pneumoniae, varicella-zoster virus, and arboviruses (including West Nile virus). A variety of toxic encephalopathies can produce identical symptoms.

In most cases, genetic or structural defects are the underlying cause of epilepsy with onset after vaccination, including cases with either preexistent encephalopathy or benign epilepsy with good outcome (344). Vaccination-associated seizures present in the setting of various epilepsy syndromes, including severe childhood epilepsies, in more than 10% of cases (347). Genetic conditions, such as Dravet syndrome (a severe epileptic encephalopathy starting in the first year of life resulting from mutations in SCN1A) and Angelman syndrome (a complex genetic disorder that primarily affects the nervous system, due to a deletion or mutation of the UBE3A gene on chromosome 15), can cause vaccine-associated encephalopathies (typically precipitated by vaccine-induced fever) that are misinterpreted as vaccine complications (62; 247); numerous alleged cases of vaccine encephalopathy were reevaluated, sometimes years later, and diagnosed as having Dravet syndrome (275), and in 1 case intellectual disability and epilepsy attributed to a vaccine encephalopathy following smallpox vaccination were later recognized as resulting from Angelman syndrome (247). Seizures following vaccinations were reported in about one quarter to one half of patients diagnosed with Dravet syndrome, and in more than half (58%), vaccination-related seizures represented the first clinical manifestation (66; 335; 358; 36); the majority of seizures occurred after DPT vaccinations and within 72 hours after vaccination, and two thirds occurred in the context of fever (335). Even though Dravet syndrome is a rare disorder, 2.5% of reported seizures following vaccinations in the first year of life in our cohort occur in children with this disorder (345). Importantly, although vaccination might trigger an earlier onset of Dravet syndrome in children who, because of a SCN1A mutation, are destined to ultimately develop the disease, vaccination-associated earlier seizure onset in Dravet syndrome does not alter the disease course (221; 378; 343). Consequently, vaccination should not be withheld from children with SCN1A mutations (221).

Diagnostic workup

For cases with vaccine-associated meningitis or encephalitis due to CNS of the vaccine virus with live-attenuated vaccines [eg, oral polio vaccine, smallpox (vaccinia) vaccine, yellow fever vaccine], detection of the vaccine virus in CSF either by culture or nucleic acid amplification is diagnostic. The detection of virus-specific IgM and specific neutralizing antibodies in the CSF also supports the diagnosis (313): because virus-specific IgM antibodies do not normally cross the blood-brain barrier, their presence in CSF is considered indicative of local CNS infection (313).

For autoimmune-mediated events (eg, acute disseminated encephalomyelitis, Guillain-Barré syndrome), diagnosis should be made using appropriate studies (eg, neuroimaging, electromyography, and nerve conduction studies). No confirmatory vaccine-specific laboratory tests exist for vaccine-induced autoimmune syndromes.

No clinical, radiological, or laboratory findings are specific for a diagnosis of postvaccination encephalomyelitis. Temporal association with vaccination (10 to 14 days from vaccination to onset of symptoms on average) and meningoencephalitic symptoms suggest a diagnosis of postvaccination encephalomyelitis. In such cases with suspected postvaccination encephalitis, enhanced MRI and spinal fluid studies (including myelin basic protein studies) should be obtained (55). CSF findings are nonspecific but include an increase in opening pressure, a modest mononuclear pleocytosis (less than 500 cells/cc), and elevated protein concentration (155; 279; 55; 84). CSF glucose concentration is normal. EEG may show high-voltage slow-waves (55). Bacterial infection should be ruled out by examination and culture of CSF as well as other appropriate cultures, including blood. Polymerase chain reaction testing of CSF should be performed for agents commonly associated with meningoencephalitis.

Because VAPP is clinically indistinguishable from poliomyelitis caused by wild-type polioviruses (27), diagnosis of VAPP requires demonstration of acute flaccid paralysis with residual weakness present 60 days after symptom onset and confirmation with a stool sample that is negative for wild-type poliovirus but positive for vaccine virus (366).

Management

Treatment for vaccine-associated neurologic disorders depends on the particular clinical syndrome.

For vaccine-associated autoimmune-mediated neurologic manifestations, treatment typically includes intravenous immunoglobulin or plasmapheresis for Guillain-Barré syndrome and corticosteroids, intravenous immunoglobulin, or plasmapheresis for acute disseminated encephalomyelitis (313).

There is no specific therapy for encephalitis following vaccination (84). Supportive care, anticonvulsants, and intensive care may be required in individual cases (313). No treatment is recognized to alter the clinical course, although some authors advocate intravenous methylprednisolone and consideration of intravenous immunoglobulin and plasmapheresis (231).

For postvaccination encephalitis following smallpox vaccination, vaccinia immune globulin is not effective, and the antiviral cidofovir is not indicated (128; 279; 55; 80; 84; 231).

Healthcare providers who need clinical consultation for potential vaccine adverse reactions should contact their state or local health department or the CDC Clinician Information Line at (877) 554-4625. Military healthcare providers (or civilian providers treating a military healthcare beneficiary) should call (866) 210-6469 for clinical consultation.

Healthcare providers should report vaccine adverse events to their state or local health department and to the Vaccine Adverse Event Reporting System at the VAERS Web site or (800) 822-7967. The Vaccine Adverse Event Reporting System (VAERS) and the National Vaccine Injury Compensation Program track vaccine-associated adverse events and allow compensation for documented vaccine-related injury.

Outcomes

Encephalomyelitis following vaccination is a severe disease with a relatively high mortality and morbidity (55). Reported case-fatality rates for postvaccination encephalomyelitis vary considerably over different studies from approximately 10% to 50%. Some of the variation may be due to different levels of supportive care and to different vaccine strains. In the United States in 1968, mortality from postvaccination encephalomyelitis was approximately 10% (80). Death can occur suddenly, usually within a week of onset of symptoms. Approximately 25% to 30% of all survivors have some residual neurologic defect.