Use of BCG vaccine in HIV-infected infants

Additional data from studies in Argentina and South Africa confirm the significantly high risk of disseminated BCG (dBCG) disease in HIV-positive infants, with rates approaching 1%. Other studies have shown that infection with HIV severely impairs the BCG-specific T-cell responses during the first year of life. Thus, BCG may therefore provide little, if any, protection against tuberculosis in HIV-infected infants. Considering the significant risk of dBCG disease, these data strongly support the WHO recommendation of not giving BCG to children who are known to be infected with HIV. Additional data suggest that highly active antiretroviral therapy (HAART) may reduce the rate of dBCG disease and that this benefit may be greater than the increased frequency of the immune reconstitution inflammatory syndrome that is observed in BCG-vaccinated HIV-infected children receiving HAART.

The new data do not provide arguments for modifying the current policy recommended by WHO,⁹ which has also been supported by recent statements from the International Union Against Tuberculosis and Lung Disease and the Childhood TB Subgroup of the WHO DOTS Expansion Working Group.¹⁰ The operational difficulties in implementing the WHO recommendations were noted, in particular the delayed vaccination approach,¹¹ which might be possible to implement only in situations where: good TB and HIV surveillance systems for pregnant women and infants existed; where strategies for the prevention of mother-to-child transmission of HIV were operating optimally and were closely linked to well functioning EPI programmes with good follow-up of all infants; and in situations where HAART coverage for mothers and children is high.

Disseminated BCG diseases have mainly been reported from Argentina and South Africa, countries that use the Pasteur and Danish strains respectively. Countries with good HIV and tuberculosis reporting systems, such as Brazil and Thailand, have not reported dBCG cases to the extent that the other 2 countries have. These last 2 countries use less reactogenic BCG vaccines (i.e. Japanese and Moreau vaccines) and whether strain is important in the genesis of dBCG in HIV infected children merits further evaluation.

⁹See No. 3, 2007, p. 22.

¹⁰ International Journal of Tuberculosis and Lung Disease, 2008;12:1376–1379.

¹¹ The current WHO recommendation is that for infants born to HIV-infected mothers where early HIV diagnostic testing can be performed, BCG can be deferred until diagnostic testing results are available.

Full report of GACVS meeting of 3-4 December 2009, published in the WHO Weekly Epidemiological Record on 29 January 2010

Safety of BCG vaccine in HIV-infected children

The Committee has reviewed the policy on the use of bacille Calmette–Guérin (BCG) vaccination for children infected with HIV. Data from retrospective studies from Argentina and South Africa indicate there is a substantiated higher risk of disseminated BCG disease developing in children infected with HIV who are vaccinated at birth and who later developed AIDS. The reported risk associated with vaccinating HIV-infected children may outweigh the benefits of preventing severe tuberculosis, especially since the protective effect of BCG against tuberculosis in HIV-infected children is not known.

WHO currently recommends administering a single dose of BCG vaccine to all infants living in areas where tuberculosis is highly endemic as well as to infants and children at particular risk of exposure to tuberculosis in countries with low endemicity. BCG vaccine is contraindicated in people with impaired immunity, and WHO does not recommend BCG vaccination for children with symptomatic HIV infection.

GACVS concluded that the findings indicated there is a high risk of disseminated BCG disease developing in HIV-infected infants and therefore BCG vaccine should not be used in children who are known to be HIV infected.

The Committee recognizes the difficulty in identifying infants infected with HIV at birth in settings where diagnostic and treatment services for mothers and infants are limited. In such situations, BCG vaccination should continue to be given at birth to all infants regardless of HIV exposure, especially considering the high endemicity of tuberculosis in populations with high HIV prevalence. Close follow up of infants known to be born to HIV-infected mothers and who received BCG at birth is recommended in order to provide early identification and treatment of any BCG-related complication. In settings with adequate HIV services that could allow for early identification and administration of antiretroviral therapy to HIV-infected children, consideration should be given to delaying BCG vaccination in infants born to mothers known to be infected with HIV until these infants are confirmed to be HIV negative.

Full report of GACVS meeting of 29-30 November 2006, published in the WHO Weekly Epidemiological Record of 19 January 2007

Safety of BCG vaccination in immunocompromised individuals

GACVS noted that there has been repeated reference to local or disseminated BCG infection several years after BCG immunization in HIV-positive persons. There needs to be closer monitoring of these adverse events in areas of high HIV prevalence, with specific efforts to distinguish BCG infection from tuberculosis. Currently, no change invaccination policy is recommended, although the risk–benefit relationship should be continually assessed, and surveillance of HIV-positive persons who receive BCG vaccines should be continued for at least 5–7 years. In the development of new live attenuated vaccines against tuberculosis, account needs to be taken of the special safety issues for immunocompromised recipients.

Full report of GACVS meeting of 11-12 June 2003, published in the WHO Weekly Epidemiological Record on 8 August 2003

Reference:

The recent isolation of strains of bacille Calmette–Guérin (BCG) resistant to isoniazid from 5 patients with lymphadenitis in the Netherlands has brought into question the clinical relevance of the finding and whether the use of isoniazid-resistant BCG vaccine should be discontinued. The Netherlands experience relates to the Danish 1331 strain of BCG, used in the Netherlands since 1998. It is well known that there is a certain degree of isoniazid resistance of isolated BCG strains.

The use of isoniazid as single drug treatment for BCG vaccine-induced lymphadenitis is recommended only in the Netherlands. Elsewhere, most cases of BCG lymphadenitis are managed either by doing nothing or by surgical excision if necessary. For disseminated infection in patients with immune deficiency, treatment with 3 or more anti-tuberculosis drugs is recommended. A distinction should be made between such cases and those with lymphadenitis arising as a frequent consequence of BCG vaccination. Even in countries with a high prevalence of human immunodeficiency virus (HIV) infection in newborn infants, lymphadenitis is regarded as having no clinical relevance provided the correct injection technique has been used and adverse reactions are adequately treated. The Committee concluded that the isolation and identification of a low level of isoniazid resistance of BCG strains from 5 patients presenting with lymphadenitis do not justify a change in standard policy.

Full report of GACVS meeting of 9-10 June 2005, published in the WHO Weekly Epidemiological Record on 15 July 2005

GACVS reviewed the safety profile of BCG vaccines in preparation for an updated WHO position paper on BCG vaccines, and safety information sheet.⁸ Although the current global BCG vaccine supply is limited to only 3 prequalified manufacturers, the vaccine has been shown to be consistently protective against infant tuberculous meningitis and miliary tuberculosis, and remains an important tool for the prevention of tuberculosis. BCG vaccination has been part of immunization programmes since the 1960s and part of EPI programmes since 1974. While being shown to be effective in infants, evidence for BCG protection against pulmonary tuberculosis in older children and adults is more variable, ranging from 0% to 80%, and tends to be higher in individuals with no detectable prior exposure to mycobacterial infection or environmental mycobacteria. For this reason, the policy of BCG vaccination varies, particularly in low burden countries, with some countries electing not to administer it routinely. The currently available WHO BCG vaccine position paper was published in 2004.⁹ Revised BCG vaccination guidelines for infants at risk for HIV infection were published separately in 2007.¹⁰ The BCG position paper will be updated following a SAGE review of evidence currently scheduled for October 2017.

BCG has a well-established safety profile. Usually, a correct administration – intradermally on the upper arm (deltoid) – results in a local reaction consisting of a small pustule followed by a small scar. When administered subcutaneously, lymphadenitis can occur, with rare instances of suppuration and fistulae formation. Other severe reactions include injection site abscesses or severe ulcerations; regional adverse reactions such as ipsilateral regional lymph node enlargement; and rare episodes of distant disease, particularly in the skin, gut or bones. Osteitis or osteomyelitis may be detected, sometimes after 12 months following vaccination. Disseminated BCG disease having high case fatality is rare and is associated almost entirely with HIV and primary immune deficiency syndromes. BCG immune reconstitution inflammatory syndrome (IRIS) and eye problems including uveitis, optic neuritis and lupus vulgaris are also among the recognized adverse reactions to the vaccine.

The reactogenicity of BCG vaccine is influenced by the BCG strain, age at administration, immune status and revaccination. Most of the available data are obtained through passive surveillance methods, which could underestimate the actual number of AEFIs. The switching of BCG vaccine strains and manufacturing processes have occasionally been associated with increased reports of AEFIs.

The Committee noted that the live vaccine requires culture from controlled seed strains and comparison with reference standards regarding the number of culturable units per dose and laboratory tests of immunogenicity, lack of virulence and skin reactogenicity. However, the manufacturing process of this live vaccine, and not just vaccine strain differences, has been reported to affect reactogenity and notifications of local and regional adenitis.¹¹

A new systematic review of the published literature on safety is currently underway. Preliminary findings note mostly case reports of local reactions such as lupus vulgaris, adenitis, granuloma and other skin lesions. Case reports are also available on distal or disseminated adverse events such as osteomyelitis, meningitis, miliary tuberculosis and disseminated BCG. These case studies have a high risk of bias. Retrospective cohort studies are available that will be used to provide observed rates for some of the most frequent, known BCG reactions.

The review also identified limitations in the characterization of the safety profile of BCG vaccines. Data on strain-specific reactogenicity or on variations in reactogenicity within the same manufacturing process are fairly limited. New information will also be sought on reactogenicity rates and adverse events in newborns compared to the late neonatal period and post neonatal period in HIV and non-HIV affected infants to help inform any changes in the advice on use of BCG vaccine and schedules in Member States.

⁸ See http://www.who.int/vaccine_safety/initiative/tools/BCG_Vaccine_rates_information_sheet.pdf?ua=1

⁹ See No. 4, 2004, pp. 27–38.

¹⁰ See No. 21, 2007, pp. 193–196.

¹¹ WHO Expert Committee on Biological Standards 2011. Recommendations to assure the quality, safety and efficacy of BCG vaccines, Annex. Available at: http://www.who.int/biologicals/BCG_DB_HK_23_April_2012.pdf, accessed June 2017.

Full report of GACVS meeting of 7-8 June 2017, published in the WHO Weekly Epidemiological Record of 14 July 2017