515
REVIEW ARTICLES
Meningococcal Meningitis in Sub-Saharan Africa: A Model for the Epidemic Process Patrick S. Moore
Center for AIDS Prevention, University a/California, San Francisco, San Francisco, California
The factors that cause the transition of a communicable disease from an endemic steady-state to a rapidly evolving epidemic are not well understood. Epidemics are difficult to study because they are unpredictable and occur infrequently. In sub-Saharan Africa, however, epidemics of group A meningococcal meningitis have a regular cyclic pattern that provides an unusual opportunity to study the factors leading to the emergence of a communicable disease epidemic.
Epidemiology of Epidemic Meningococcal Disease Neisseria meningitidis, the bacterium that causes meningococcal meningitis, can be classified into at least 13 distinct serogroups based on the antigenicity ofthe outer polysaccharide capsule [1, 2]. Although various meningococcal serogroups are responsible for sporadic cases of meningitis, major epidemics are caused primarily by group A [2] and, to a lesser extent, by group C meningococci [3-5], Meningococcal meningitis has a case-fatality ratio of5%-25% [6,7] and neurological sequelae are common among survivors [8]. N. meningitidis is also a common commensal of humans: 2%-10% of healthy persons are nasopharyngeal carriers of potentially pathogenic meningococci [9-11]. Meningococci are spread by person-to-person respiratory transmission, and
Received 5 February 1991; revised 14 May 1991. The views expressed in this paper are those ofthe author and not necessarily those of the Centers for Disease Control (Atlanta), Grant support: National Institute of Mental Health traineeship in AIDS prevention studies (T32MH 1905). Reprints or correspondence (present address): Dr. Patrick S. Moore. Arboviral Diseases Branch. Division of Vector-Borne Infectious Diseases, Centers for Disease Control, P.O. Box 2087. Fort Collins, Colorado 80522. Clinical Infectious Diseases 1992;14:515-25 © 1992 by The Universityof Chicago. All rights reserved. 1058-4838/92/1402-0022$02.00
meningococcal carriage rates do not necessarily correlate with the risk of invasive disease [12]. Meningococcal transmission often occurs in the absence of a corresponding increase in the rate of meningococcal disease [13], a circumstance indicating that transmission alone is not sufficient to trigger an epidemic [14]. Since World War II, group A meningococcal epidemics have been rare in industrialized nations, and attack rates for all serogroups combined generally do not exceed 5 cases/ 100,000 per year [2,15,16]. Limited outbreaks of group A meningococcal disease that primarily affect lower socioeconomic groups have taken place in Finland [17], New Zealand [18, 19], and the U.S. Pacific Northwest [20] over the last 15 years. These outbreaks, as well as outbreaks that occurred in the United States before the 1940s [2 I], had relatively low overall incidence rates « I5 cases/ I00,000 per year). In contrast, intense meningococcal epidemics have been reported in a number of developing countries, including Brazil [2], Nepal [22], China [23], and various subSaharan African nations [24-26]. Attack rates during these epidemics can approach 1% of the population; as many as 3 million cases occurred during the group A meningococcal epidemic in China in 1967 [23]. Health facilities in developing countries can be paralyzed during epidemics by the rapid and prolonged influx of patients with meningitis. A broad savannah region in Africa extending from The Gambia to Ethiopia is known as the "meningitis belt" (figure 1). This region is uniquely susceptible to intense group A meningococcal epidemics that occur in 8- to 14-year cycles [7, 27-29]; each epidemic wave follows a multiyear, crescendo-decrescendo pattern (figure 2). Rates of disease are also seasonally dependent, peaking during the dry season (December-May) and declining rapidly with the onset of rains (figure 3). Even during major epidemic waves, menin-
Downloaded from http://cid.oxfordjournals.org/ at Florida Atlantic University on July 23, 2015
Epidemic group A meningococcal meningitis follows a unique and distinctive pattern in subSaharan Africa. Advancesin molecular and fieldepidemiology have begun to elucidate the mechanisms of meningococcal meningitis epidemics. Epidemics result from a complex combination of host, organism, and environmental risk factors. Recent studies suggest that "antigenic shifts" in group A meningococcalclones may trigger an outbreak of disease by suddenly decreasing herd immunity within a population. Although the introduction of new group A meningococcal strains into a susceptible population contributes to the likelihood of an epidemic, the presence of additional environmental factors, such as low humidity and coincident respiratory tract infections, are also necessaryfor an epidemic to occur. Despite the unique behavior of group A meningococcal disease in sub-Saharan Africa, the application of similar methods of epidemiological analysis may be useful for determining epidemic processes for other diseases.
516
em 1992;14 (February)
Moore
gitic activity is suppressed by environmental factors during each rainy season [13, 30, 31]. Similar epidemic patterns occurred in the United States before World War II [21, 32, 33] and occur in present-day China [23]. In addition to the unique temporal pattern ofmeningococcal disease in Africa, there is a marked shift in the age of patients with meningitis during epidemics. In industrialized countries, highest rates of sporadic meningococcal meningi-
tis occur among young infants [15]. During group A meningococcal epidemics, however, older groups are often affected, with the age-specific attack rates among 5- to IO-year olds being higher than the rates among younger groups [2, 5, 24, 34]. This shift is not due to underreporting of cases in younger children [6].
100 I
500
IUJZ 1-0 «::2:
~
UJ« I-UJ
400
a:- 60 C/l8
«~ a: o
C/l0 -0
1 -' _0 CJO
Z'!:: -C/l ZUJ UJC/l ::2:«
80
i= 0,
300
-0
200
ZC/l UJUJ ::2:C/l «
CJo
~:!:::
40
S2. 20
S2. 100 0
1980 1940
1950
1960
1970
1980
YEAR Figure 2. Annual meningitis rates for Burkina Faso, 1939-1985 (from [29]).
1981
1982
1983
1984
1985
YEAR Figure 3. Monthly meningitis rates, Burkina Faso, 1981-1985. Meningococcal meningitis rates are seasonally dependent, even during major epidemic waves (from [29]).
Downloaded from http://cid.oxfordjournals.org/ at Florida Atlantic University on July 23, 2015
Figure 1. The Meningitis belt of subSaharan Africa (based on [7, 24, 26, 27]).
eID 1992;14 (February)
Epidemic Meningococcal Meningitis
No single epidemiological factor fully accounts for the patterns of African meningococcal epidemics [24]. However, recent developments in the quantitative and descriptive epidemiology of this disease suggest that epidemics result from a complex interaction of multiple factors. Interactions between the host, the organism, and the environment may explain the periodic and seasonal patterns of epidemics, as well as the unusual age distribution among persons who contract meningitis during an epidemic.
The Host
Natural immunity, however, may also be directed against non polysaccharide antigens shared between Neisseria species. N. lactamica is probably non encapsulated [40]. Therefore, cross-protective immunity induced by N. lactamica carriage is probably directed against non polysaccharide antigens. In addition, healthy persons are significantly more likely than patients with meningitis to have protective bactericidal antibody levels against a variety of meningococcal serogroups [33]. This is not due to an intrinsic inability of patients with meningitis to develop antipolysaccharide antibodies since recuperating patients develop high homologous bactericidal titers [33, 38]. This is consistent, however, with the possibility that bactericidal assays actually measure antibodies directed against nonpolysaccharide antigens shared among different meningococcal serogroups and different Neisseria species. Low premorbid bactericidal titers in patients with meningitis may be due to lack of previous carriage of Neisseria species. The development of cross-protective immunity against non polysaccharide epitopes would explain why many healthy American adults have high bactericidal titers against relatively rare meningococcal serogroups [33]. Shared antigens that might stimulate cross-protective antibodies include the lipooligosaccharide [40, 41] and various outer membrane proteins [42, 43]. The duration, efficacy, and specificity of immunity induced against nonpolysaccharide antigens has not been well characterized. Loss of herd immunity against group A meningococci may contribute to the regularity of epidemic cycles in sub-Saharan Africa [24]. Carriage rates of 20%-30% occur during epidemics in Africa [44], and the half-life of carriage is ~ 4 weeks [45, 46]. Carriage of group A meningococci can cause transient meningococcemia [47] and fever [14,47]. Under these conditions, most persons in a population can be expected to become meningococcal carriers and develop protective antibody titers [45]. Although immunization with purified group A polysaccharide vaccine does not prevent carriage [14, 48], natural immunity induced by whole group A meningococci appears to effectively reduce transmission [48]. Rising herd immunity due to widespread carriage would then limit meningococcal transmission, ending the epidemic wave [48]. A population's susceptibility to epidemic disease might return as antibody levels decline and herd immunity is diluted by new birth cohorts. In adults, protection afforded by the group A polysaccharide vaccine declines to 67% within 3 years of vaccination [39]. Vaccinees develop a rapid T cellindependent antibody response, but < 10% of children
REVIEW ARTICLES
Meningococcal Meningitis in Sub-Saharan Africa: A Model for the Epidemic Process Patrick S. Moore
Center for AIDS Prevention, University a/California, San Francisco, San Francisco, California
The factors that cause the transition of a communicable disease from an endemic steady-state to a rapidly evolving epidemic are not well understood. Epidemics are difficult to study because they are unpredictable and occur infrequently. In sub-Saharan Africa, however, epidemics of group A meningococcal meningitis have a regular cyclic pattern that provides an unusual opportunity to study the factors leading to the emergence of a communicable disease epidemic.
Epidemiology of Epidemic Meningococcal Disease Neisseria meningitidis, the bacterium that causes meningococcal meningitis, can be classified into at least 13 distinct serogroups based on the antigenicity ofthe outer polysaccharide capsule [1, 2]. Although various meningococcal serogroups are responsible for sporadic cases of meningitis, major epidemics are caused primarily by group A [2] and, to a lesser extent, by group C meningococci [3-5], Meningococcal meningitis has a case-fatality ratio of5%-25% [6,7] and neurological sequelae are common among survivors [8]. N. meningitidis is also a common commensal of humans: 2%-10% of healthy persons are nasopharyngeal carriers of potentially pathogenic meningococci [9-11]. Meningococci are spread by person-to-person respiratory transmission, and
Received 5 February 1991; revised 14 May 1991. The views expressed in this paper are those ofthe author and not necessarily those of the Centers for Disease Control (Atlanta), Grant support: National Institute of Mental Health traineeship in AIDS prevention studies (T32MH 1905). Reprints or correspondence (present address): Dr. Patrick S. Moore. Arboviral Diseases Branch. Division of Vector-Borne Infectious Diseases, Centers for Disease Control, P.O. Box 2087. Fort Collins, Colorado 80522. Clinical Infectious Diseases 1992;14:515-25 © 1992 by The Universityof Chicago. All rights reserved. 1058-4838/92/1402-0022$02.00
meningococcal carriage rates do not necessarily correlate with the risk of invasive disease [12]. Meningococcal transmission often occurs in the absence of a corresponding increase in the rate of meningococcal disease [13], a circumstance indicating that transmission alone is not sufficient to trigger an epidemic [14]. Since World War II, group A meningococcal epidemics have been rare in industrialized nations, and attack rates for all serogroups combined generally do not exceed 5 cases/ 100,000 per year [2,15,16]. Limited outbreaks of group A meningococcal disease that primarily affect lower socioeconomic groups have taken place in Finland [17], New Zealand [18, 19], and the U.S. Pacific Northwest [20] over the last 15 years. These outbreaks, as well as outbreaks that occurred in the United States before the 1940s [2 I], had relatively low overall incidence rates « I5 cases/ I00,000 per year). In contrast, intense meningococcal epidemics have been reported in a number of developing countries, including Brazil [2], Nepal [22], China [23], and various subSaharan African nations [24-26]. Attack rates during these epidemics can approach 1% of the population; as many as 3 million cases occurred during the group A meningococcal epidemic in China in 1967 [23]. Health facilities in developing countries can be paralyzed during epidemics by the rapid and prolonged influx of patients with meningitis. A broad savannah region in Africa extending from The Gambia to Ethiopia is known as the "meningitis belt" (figure 1). This region is uniquely susceptible to intense group A meningococcal epidemics that occur in 8- to 14-year cycles [7, 27-29]; each epidemic wave follows a multiyear, crescendo-decrescendo pattern (figure 2). Rates of disease are also seasonally dependent, peaking during the dry season (December-May) and declining rapidly with the onset of rains (figure 3). Even during major epidemic waves, menin-
Downloaded from http://cid.oxfordjournals.org/ at Florida Atlantic University on July 23, 2015
Epidemic group A meningococcal meningitis follows a unique and distinctive pattern in subSaharan Africa. Advancesin molecular and fieldepidemiology have begun to elucidate the mechanisms of meningococcal meningitis epidemics. Epidemics result from a complex combination of host, organism, and environmental risk factors. Recent studies suggest that "antigenic shifts" in group A meningococcalclones may trigger an outbreak of disease by suddenly decreasing herd immunity within a population. Although the introduction of new group A meningococcal strains into a susceptible population contributes to the likelihood of an epidemic, the presence of additional environmental factors, such as low humidity and coincident respiratory tract infections, are also necessaryfor an epidemic to occur. Despite the unique behavior of group A meningococcal disease in sub-Saharan Africa, the application of similar methods of epidemiological analysis may be useful for determining epidemic processes for other diseases.
516
em 1992;14 (February)
Moore
gitic activity is suppressed by environmental factors during each rainy season [13, 30, 31]. Similar epidemic patterns occurred in the United States before World War II [21, 32, 33] and occur in present-day China [23]. In addition to the unique temporal pattern ofmeningococcal disease in Africa, there is a marked shift in the age of patients with meningitis during epidemics. In industrialized countries, highest rates of sporadic meningococcal meningi-
tis occur among young infants [15]. During group A meningococcal epidemics, however, older groups are often affected, with the age-specific attack rates among 5- to IO-year olds being higher than the rates among younger groups [2, 5, 24, 34]. This shift is not due to underreporting of cases in younger children [6].
100 I
500
IUJZ 1-0 «::2:
~
UJ« I-UJ
400
a:- 60 C/l8
«~ a: o
C/l0 -0
1 -' _0 CJO
Z'!:: -C/l ZUJ UJC/l ::2:«
80
i= 0,
300
-0
200
ZC/l UJUJ ::2:C/l «
CJo
~:!:::
40
S2. 20
S2. 100 0
1980 1940
1950
1960
1970
1980
YEAR Figure 2. Annual meningitis rates for Burkina Faso, 1939-1985 (from [29]).
1981
1982
1983
1984
1985
YEAR Figure 3. Monthly meningitis rates, Burkina Faso, 1981-1985. Meningococcal meningitis rates are seasonally dependent, even during major epidemic waves (from [29]).
Downloaded from http://cid.oxfordjournals.org/ at Florida Atlantic University on July 23, 2015
Figure 1. The Meningitis belt of subSaharan Africa (based on [7, 24, 26, 27]).
eID 1992;14 (February)
Epidemic Meningococcal Meningitis
No single epidemiological factor fully accounts for the patterns of African meningococcal epidemics [24]. However, recent developments in the quantitative and descriptive epidemiology of this disease suggest that epidemics result from a complex interaction of multiple factors. Interactions between the host, the organism, and the environment may explain the periodic and seasonal patterns of epidemics, as well as the unusual age distribution among persons who contract meningitis during an epidemic.
The Host
Natural immunity, however, may also be directed against non polysaccharide antigens shared between Neisseria species. N. lactamica is probably non encapsulated [40]. Therefore, cross-protective immunity induced by N. lactamica carriage is probably directed against non polysaccharide antigens. In addition, healthy persons are significantly more likely than patients with meningitis to have protective bactericidal antibody levels against a variety of meningococcal serogroups [33]. This is not due to an intrinsic inability of patients with meningitis to develop antipolysaccharide antibodies since recuperating patients develop high homologous bactericidal titers [33, 38]. This is consistent, however, with the possibility that bactericidal assays actually measure antibodies directed against nonpolysaccharide antigens shared among different meningococcal serogroups and different Neisseria species. Low premorbid bactericidal titers in patients with meningitis may be due to lack of previous carriage of Neisseria species. The development of cross-protective immunity against non polysaccharide epitopes would explain why many healthy American adults have high bactericidal titers against relatively rare meningococcal serogroups [33]. Shared antigens that might stimulate cross-protective antibodies include the lipooligosaccharide [40, 41] and various outer membrane proteins [42, 43]. The duration, efficacy, and specificity of immunity induced against nonpolysaccharide antigens has not been well characterized. Loss of herd immunity against group A meningococci may contribute to the regularity of epidemic cycles in sub-Saharan Africa [24]. Carriage rates of 20%-30% occur during epidemics in Africa [44], and the half-life of carriage is ~ 4 weeks [45, 46]. Carriage of group A meningococci can cause transient meningococcemia [47] and fever [14,47]. Under these conditions, most persons in a population can be expected to become meningococcal carriers and develop protective antibody titers [45]. Although immunization with purified group A polysaccharide vaccine does not prevent carriage [14, 48], natural immunity induced by whole group A meningococci appears to effectively reduce transmission [48]. Rising herd immunity due to widespread carriage would then limit meningococcal transmission, ending the epidemic wave [48]. A population's susceptibility to epidemic disease might return as antibody levels decline and herd immunity is diluted by new birth cohorts. In adults, protection afforded by the group A polysaccharide vaccine declines to 67% within 3 years of vaccination [39]. Vaccinees develop a rapid T cellindependent antibody response, but < 10% of children
