- Email: [email protected]
Filariasis due to Loa loa and Manspnella perstans: distribution in the region of Okondja, Haut-Ogooué Province, Gabon, with parasitological and serological follow-up over one year
441
TRANSACTIONS OF THE ROYALSOCXETY OF TROPICALMEDICINEAND HYGIENE(1987)81, 441446
Filariasis due to Loa loa and Mansonella perstans: distribution in the region of Okondja, Haut-Ogoou6 Province, Gabon, with parasitological and serological follow-up over one year
‘International
M. VAN HOEGAERDEN', B. CHABAUD’, J. P. AKUE’ AND B. IVANOFF’ Medical Research Centre, BP. 769, Franceville, Gabon; ‘Hopita G&&al d’Okomija,
Gabon
Abstract The prevalence of Loa loa and Mansonella perstans filariasis has been determined in 6 rural villages in eastern Gabon. Between 18.9 and 27.2% of people carry L. loa microfilariae with an overall microfilarial rate of 25.1%. The microfilarial rate for M. perstans was more variable, between 33.3 and 62.2% (average 49.1%). No significant difference was seen in the microfilarial rate with age over 15 years for either parasite, but men were infected more frequently than women. Anti-L. loa antibody titres were measured, using a homologous microfilarial antigen in ELISA. Taking the parasitological and immunological evaluations together, only 10%of the sample population appear to be free of these filarial infections. L. loa and M. perstans microfilaraemia and corresponding serology were also investigated twice in 150 people at a one-year interval. 99.1% of the caseswho had no circulating L. loa microfilaria in March 1984 still did not show any 12 months later. Similarly, 97.1% of the untreated, microfilaraemic casesstill harboured this parasite a year later. The samewas not observed for M. perstans, since microiilariae appeared or disappeared in 26.7% of the cases. This suggests different dynamics for the two filarial infections. Variation in individual anti-L.loa antibody titres was low. The possibility of a genetic influence on the expression of loiasis is discussed.
Introduction Filariasis due to Loa loa was first reported in Africa by BRUMMPT(1904). The endemic zone of the parasite tid its insect vector have been delimited by-several authors. It can be calculated that 20 to 30 million people live in endemic areas in West and Central Africa and a conservative estimate is that 3 million people ha&our the parasite (FAIN, 1981). Clinical manifestations are usuallv classified as benign (such as the characteristic &bar oedema of the face, the wrist or the ankle, or the passageof the adult worm under the conjunctiva), although serious pathologies have occasionally been reported. Diethylcarbamazine citrate (DEC), the drug commonly used for treatment of loiasis, may provoke severe shock reactions, sometimesending in coma or sudden death due to encephalitis (VAN BOGAERTet al., 1955; CAUCHIEet al., 1965). Moreover, DEC is inefficient in eliminating Mansonella perstans (HAWKING, 1981), another filarial parasite common in Western Equatorial Africa. Although various authors suggested the non-pathogenic character of this latter organism, others (ADOLPH et al., 1962; GELFAND& WESSELS, 1964; DUKES et al., 1968) and ourselves (unpublished) have associated the presence of micr&l&ae (mf) of this species with certain clinical manifestations. RICHARD-LENOBLEet al. (1980) have determined the prevalence of circulatina mf in some regions of Gab& but villages in the re$on of Okondja Gere not investigated. This paper analyses the distribution of both filarial diseases, associating parasitological and immunological evaluations. Most studies of L. loa prevalence show that a maximum of about 25% of people ha&our circulating mf although it was estimated, in Cameroon, that each
individual was exposed to an infecting bite every 5 davs ~G~RD~N et al.. 19481. On the other hand. pr&&nce of M. Persians is highly variable with & rates varying from 6.3 to 81.2% in different regions of Gabon (RICHARD-LENOBLEet al., 1980). GORDON (1955) hypothesized that the development of the filaria Wuchereria bancrofti could be related to the genetic background of the individual. It has been extensively demonstrated in experimental models that the susceptibility to hehninth infections is related to the major histocompatibility complex. A possible association of histocompatibility antigens with filarial elephantiasis has been reported recently (CHAN et al., 1984). It could be postulated in the caseof loiasis (but not mansonellosis) that individuals have a genetic susceptibility (or resistance) to the appearanceof mf in peripheral blood circulation. In order to determine whether untreated individuals lose or acquire L. loa and/or M. perstans mf, we have also investigated the parasitaemia and serology of 150 adults living in an endemic area, over a period of one year. Subjects, Materials and Methods Study area 6 villages, each of about 200 to 500 inhabitants, were selected easf of Okondja, in a region where loiasis was highly endemic. These villages are situated along a dirt road, the farthest being 80 km eastof Okondja,towardsthe borderof the Democratic Republic of Congo, at an altitude of approximately 400 m above sea level (Figure). Forest characteristics vary among villages: Aboumi, Oboli and
Ontogo,closerto the Congoleseborder,aresituatedin dense forest with high canopy, habitat favourable for Chtysops
silacea and C. dimidiata (common vectors of L. loa). In
contrast, Obili, Bouala and Ayanabo, closer fo Okondja, are situated in less dense forest with a lower canopy and thus a less favourable habitat for the insect vector. Detailed
442
FILARIASIS
IN
EASTERN
GABON
conical tubes and 200 pl saponin (2% w/v) were added for haemolysis. Tubes were then centrifuged (10 min, 500g) and the supernatant discarded. The entire pellet was scanned under a microscope and the number of mf counted. Parasite species was determined by size and movement characteristics. serology A crude soluble extract of L. loa was prepared from purified mf (VAN HOEGAERDEN & IVANOFF, 1986), frozen and thawed 3 times and ultrasonicated (Soniprep 150, MSE, 23 kHz, amplitude varying from 8 to 14 pm, 5 x 1 min at O’C). After ultracentrifugation (lO’gx45 min at 4”C), the supematant was stored at -85°C. Protein content was determined by the Biorad (Coomassie blue) method using BSA as a standard. Antibody levels were determined by enzyme-linked immuoosorbent assay (ELISA) microtirration using 0.1 ug of protein to coat wells. Serum samples were diluted l/20 and l/2000 respectively for IgE and IgG + IgM determinations. Diluting buffer was PBS containing 2% gelatin and 0.5% Tween 20. Secondantibody was purified IgG fraction of goat anti-human IgG + IgM, or anti-human IgE coupled to alkaline phosphatase (Sigma). Results were expressed as O.D. obtained for patient’s serum divided by the average O.D. of three serum samples obtained from Caucasianswho had neverlived ia regionsendemicfor L. loo or M. perstam. The test was performed on serum samples not only from people presenting M. perstunsmicro&uremia alone but also
from individuals harbouring intestinal hehniaths or other parasites that may be found in Gabon.
Figure. Map of GabonshowingHam-OgooueProvinceand the 6 villages where the surveyswere conducted.
entomological studies of Chryf0f~ density or infection rates were not performed. The same 6 villages were visited in March 1984 and March 1985. Sample choice Surveys were announced 2 weeks in advance in an attempt to ensure that most of the population was present in the villages. In each village, the chief and notables were infomed of the purposes of the study which were then communicated io local dialect, by the chief, to the community, Volunteers were registered and bled m the order they presented, without distinction of ageor sex, chiefs and elders being generally first. All people presenting were included in the surveys. 411 individuals were seen. Of the 237 people investigated in the 6 villages during March 1984, 150 spontaneously presented again a year later
and blood sampleswere collected.They were askedif they
had rece.ivedDEC in the preceding year. Those individuals who presented L. IOUmicrolilaraemia in 1984but not in 1985 were questioned again in April 1985, together with family members and, when present, the nurse of the village, to ascertain whether they had received DEC. People who had lost M. perstunsmf were not recalled as this parasite is not affected by DEC (HAWKING, 1981). Blood samples Venous blood (10 ml) was collected in Venoject vacuum tubes, without additive and on EDTA, between 9 a.m. and 4 p.m. Blood was allowed to clot at ambient temperature for 1 to 8 h, after which it was centrifuged at 4°C for 15min at 500 g. Serum samples were immediately frozen at -20°C. Micro6laraemia was determined by direct reading of 10 ul uncoagulated, fresh blood between microscope slide and coverslip. When more than 20 mf were found, the result was multiplied by 100 and parasitaemia expressed in mflrnl. When fewer than 20 mf were found,. Knott’s concentration technique was used: 1 ml blood was diluted with 9 ml PBS in
Results L. loa and M. perstansmf were counted in blood samples from 411 people (287 women and 124 men), whose ages ranged from 9 to 70 years. The prevalence
of microfilaraemia is given in Table 1. On the whole, 25% of the sample population was parasitized with L. loa mf and 49% with M. perstans mf (14% bad mixed infections). Little variation was observed in the L. lou mf rate between villages or surveys, whereas variation in M. perstans prevalence was considerably higher. Microfilaraemia varied between 1 and 86 2OO/mlfor L. loa and 1 to 18OO/mlfor M. perstans (82% of the sampleswere below 20 mf/ml for the latter parasite). The distribution according to sex and age is given in Table 2, showing a higher rate in males than females and a relatively constant mf rate among the different age groups (only 3 individuals were under 15 years of age in the 9 to 25-year age group). Only 40% showed no circulating mf during these surveys. Antibody levels detected by ELISA were between 1 and 20 times the background obtained with the reference negative samples for IgG + IgM, and between 1 and 40 times for IgE. Only those samples giving an 0 .D . at least 3 times superior to the controls were considered positive. The technique does not
appear to distinguish carriers of the two parasite species but cross-reaction was not observed with serum from people harbouring intestinal parasites or other parasitic diseasesthat may be found in the area. No Onchocerca volvulus or Mammella
strepwcerca mf
were found in skin snips of 300 people during the 1985 survey (M. Van Hoegaerden, C. Dupont & A. Dupont, unpublished). Results of immunological and parasitological investigation were obtained for 198 of the 237 persons seen in 1984 and are reported in Table 3. Taking these two
ht.
VAN
HOEGAERDEN
et
443
al.
Table l-Parasitological prevalence of L. loa and M. perstunsmf ia 6 villages ia eastern Gabon determined during two surveys made at aa interval of one year* No.
L. loa
examined
(including
it
17 9 (25Wo) (26.2%)
ggy Bouala Ayanabo
1;: 2:
22 (27.2%) 3; [p%,, . 0 14 (23.0%)
Total
411
Village
;:p
*All individuals
103 (251%)
seen in 1985 were different
Table 2-Distribution
M. perstans
Mixed hfCCti0n5
Negative
29 12 (33.3%) (44.6%)
; g;:::$q
27 19 (52.8%) (41.6%)
67 36 (44.4%) (51.1%) 23 (62.2%) 35 (57.4%)
19 10 (12.3;) (14.5%) 7 (18.9%) 10 (16.4%)
49 33 (40.7%) (37.4%) 14 (37.8%) 22 (36.1%)
202 (49.1%)
58 (14~1%)
164 (399%)
mixed infections)
from those seen in 1984.
of mf carriers according to sex and age
No.
L. loa
M. pentans
Sex
Age
examined
(including
Males
9-25 26-40
14 :i
4 (28.6%) 2 (16.7%)
9 (64.3%) 4 (33.3%)
41-55 >55
62
26 13 (419%) (36.1%)
22 35 (56.5%) (61.1%)
20 (32:3;)
Total
124
45 (36.2%)
70 (56.5%)
39 (31.5%)
8 11 17 22
14 (359%) ;; yj
19 34 37 35
Females
9-25 26-40 41-55 >55
ii29 i:
Total Total M + F
Table 3-Parasitological ssmpler
L. loll
M. pentam
-
-
+ + + + -
evidence
-
58 (20.2%)
132 (46.0%)
125 (43.6%)
202 (49.1%)
164 (399%)
19 ( 96%) 68 (34.3%)
-
+
3 ( 15%) 25 (12.6%)
+ +
+
4 ( 2W0) 21 (10.6%)
+ +
+
6 ( 3.0%) 52 (26.3%)
Table 4-Evohtioo of L. loa and M. perstunsmicrotilaraemia in 150 individuals over a period of one year 1985 Negative
L. loa 1984 M. @arms
1984
(48.7%) (41.5%) (43.5%) (43.2%)
103 (25.1%)
+
Negative
4 (28.5%) ;: [;g:“:o]
411
Number (percent)
SCCOlOgiCal
37 (45.7;)
Negative
287
and seroiogicai evaluati& of 198 se-
Parasitological evidence
(20.5%) (13.4%) (20*0%) (27.2%)
mixed infections)
Positive
Positive
112 10**
3* 25
Negative Positive
58 15
2s 52
l 213 caseshad taken partial DEC trcatmcnt less than a year before March 1984. **9/10 cases had a complete or partial DEC treatment course between the surveys.
criteria together, it appears that only 10% of this sample population had no evidence of filarial infection, whereas 35% had serological evidence of past or rTmteTearia! infection in the absenceof circuIating mammg 55% camed one or both mrcrofilarial species. Evolution VidtUlS
of parasitaemia and serology in 150 indi-
150 people were seen in both March 1984 and March 1985. The evolution of L. loa and M. petstans microlilaraemia, over this period, is shown in Table 4 and average mf counts (with standard deviation) are given in Table 5. L. loa. Of the 35 people parasitized with L. loa mf in 1984, 10 no longer had this organism in 1985. It was later confirmed that 9 of them (whose microfilaraemia varied from 1 to 60 OOO/mlbefore chemotherapy) had taken complete or partial (but at least microfilaricidal) DEC treatment in the interval. The last case(who had 3 mf/ml in 1984) had not received treatment but complained of pruritus and oedema, not pitting under pressure, which is typical of loiasis. This individual’s serology was positive for IgG in both 1984 and 1985. Among the 25 individuals who remained positive over this period, the microlilaraemia increased in 16 by a factor varying from l-3 to 3800 (although 4 had taken DEC) and decreasedin 9 by a factor of 1.7 to 1500 (2 had undergone complete treatment and one had received partial treatment). Of the 115 individuals not showing this parasite in 1984, 3 (2.6%) were carriers a year later with parasitaemias of 5,26 and 77 mf/ml. However, 2 of the 3 had taken
444
FILARIASIS
Table S-Mean
IN
EASTERN
GABON
(with standard deviation) of positive L. loa and M. perstans microtilaraemias
Infection
No.
Year
L. loa alone
in 1984 and 1985
examined
mean k s.d.
1984
16
5542 + 14983
1985
11
1858* f 3604
I 1984
19
M. L.perstans: loa:
3334 1336 152 +Z!I419
1 1985
17
L. loa: M. perstans:
1837 + 2314
1984
48
Mixed
perstans
100 + 142
19 -t 48
60 78 + 211 i 1985 *The drop in averageL. lou microfilaraemia observed between 1984and 1985is artefactual, due to DEC treatment done
M.
administered to the highly parasitized people. Table 6-Antibody titre changes in 133 samples at one year interval (X = 1985 value11984 value) IgG + IgM
x>3
SetOlOgy
Table 6 shows the evolution of individual antibody titres over a period of one year in 133 samples: a significant change (3 times or more) in IgG + IgM titres was noticed in 5 (3.8%) casesonly. IgE titres changed by the same factor in 10 (7.5%) cases. No correlation was found between the change in antibody titre and the appearanceor disappearanceof circulating mf of either species.
kE
4
3>X>l
2:
x=1 l>X>O*3
:i
35
x-zo.3
:I: 7
1
Discussion
partial (microfilaricidal) DEC treatment less than a year before the 1984 survey. In the third case, no history of DEC therapy could be identified. M. perstam. Of the 67 casesharbouring mfin 1984, 15 (22.4%) had none a year later. Of the 52 who remained parasitized, the microfilaraemia remained identical in 6, decreasedby a factor of 1.2 to 180 in 12’ casesand increased by a factor of 1.1 to 60 in 34 cases. M. pentans mf were found in 25 (30.1%) of the 83 people without them in 1984. Thus, 40 (26.7%) of 150 people inverted their status regarding M. perstans microfilaraemia. In 47 (31.3%) individuals, both parasites were absent from blood circulation in 1984 and 1985. Serology was negative in both 1984 and 1985 in 8 (17%) of these 47. Table ‘I--Reported L. loa mf
in Central Africa Percentage of adult mf carriers
Area
17.8 to 23.3 3.4 to 26.4 5.4 to 24.4 27
Gabon Gabon
Gabon Gabon Gabon
Ham Ogoout, Gabon Cameroun Cameroun Western Northern Northern
rates
Loiasis is widespread in tropical Africa (RODHAIN, 1980), but its prevalence, basedon the detection of mf in blood circulation, does not usually exceed 27% (Table 7). We too found values around 25%. The detection of circulating mf alone does not reveal the actual prevalence of the diseasesince a large proportion of the population, possibly the majority, harboured the adult worm in the absenceof circulating mf (FAIN, 1981). We found that 78% of individuals not harbouring circulating mf have immunological evidence of past or present exposure to the parasites (Table 3) and specific tiarial symptomatology was identified in most of these cases.Only 10%of the total sample population have neither parasitological nor immunological evidence of aaria infestation. This is in agreementwith GOUSSARD et al. (1984) who found
Zaire Zaire Zaire
12 to 27 25.1
25-9 9.7 to 16.8 25 (1971)* (1972)*
10 (7 to 23) 88 (71 to 97)
Reference 1914 RINGENBACH & GUYOMARCH, GALLIARD, 1932 GENTILINI et al., 1965 LANGUILLAT et al., 1978 RICHARD-LENOBLE et al., 1980
Present communication LANGUILLON, 1957 RIPERT et al., 1977 FAIN, 1978 PAMPIGLIONE PAMPIGLIONE
*This population belonged to a particular ethnic group, the Bambutis of the Pygmy race.
et al., 1979 et al., 1979
M. VAN HOEGAERDEN et al.
that, by the age of two, 95% of the population had some antibodies against L. loa. The observation that only 25% of the population have circulating mf cannot be ascribed to absenceof exposure since it was calculated that, at certain times of year, inhabitants of endemic areaswere exposed to an infecting bite every 5 days (GORDONet at., 1948). Since L. loa adult worms are estimated to hve up to 15 years (EVELAND er al., 1975), all individuals would consequently be expected to harbour parasites and, therefore, mf. We found mf rates to remain constant whatever the ecological conditions of the villages, over a period of one year and independently of the individual’s age. If lower mf rates can easily be explained by variations in the presence of the transmitting insect (savannah or urban areasfor example), figures higher than approximately one quarter of the population have not, to our knowledge, been published for this part of Africa. The question may then be raised whether this constant maximum of 25% of adults presenting L. Zoa microfilaraemia is linked to a long term mechanism of host susceptibility or defence towards the appearance of mf in peripheral circulation. Susceptibility could be induced by prenatal sensitization through transplacental transfer of specific antigens. This would influence the outcome of the disease (WEIL et al., 1983). but we found no evidence of such transfer in 92 mot&-cord blood samples (VAN HOEGAERDEN & AKUE, 1986). Although alternative mechanisms such as immune prophylaxis induced by zoonotic species (of simian origin for instance) or concomitant immunity induced by existing adult worms against infective larvae and mf could explain the observed stability, the hypothesis of genetic influence may be invoked. GORDON(1955) suggested that the development of Wuchti bancrofti infection was related to the genetic background, dividing a population into “good” hosts (presenceof the parasite and absenceof clinical manifestations) and “bad” hosts (pathological evidence and low or non-existent microfilaraemia). BRENGUES (1975) observed a family clustering of susceptibility to W. bum-of%, but added a third category defined as refractory (neither clinical nor parasitological evidence). An identical association was demonstrated in cattle naturally infected with Setaria labiatopapillosa (BRENGUES& GIDEL, 1972) and in numerous human helminth infections and experimental models of fdariasis. CHAN et al. (1984) found a significant difference in the frequency of HLA-B15 histocompatibility antigen between patients with filarial elephantiasis and normal controls. Similar evidence of family clustering, compatible with genetic transmission of susceptibility, to Bancroftian filariasis was found by OTTESEN et al. (1981), although it was not associated with the HLA-A or B locus. The existence of a genetic influence on the outcome of a filarial infection could explain the remarkable constancy of the maximal L. loa prevalence (a quarter of the population) and, possibly, the exceptionally high figure found by PAMPIGLIONE et al. (1979) in a genetically homogenous group of pygmies in Northern Zaire. The study of genetically controlled long term resistance or susceptibility to L. loa mf would necessitatea larger epidemiological survey, but analy-
445
sis of the appearance or disappearance of mf in the same individual over a certain period could help to verify the hypothesis. We have therefore investigated L. loa and M. perszansparasitaemia and serology in 150 people over one year. During this period, circulating L. Zoamf appeared or disappeared in only 1.3% of the sample, in the absence of specific DEC treatment. The stability in 99.1% of the L. loa amicrotilaraemic people suggests the existence of a mechanism for long term resistance to the appearance of mf, whereas the maintenance of the presence of circulating L. loa mf in 97.1% of the untreated cases over one year implies a long term mechanism of susceptibility. A studv of familial transmission of this mutative susceptibility/resistance towards the appear&ce of L. loa mf in peripheral blood or the development of symptomatology could help understand the dynamics of this parasitic disease. In agreement with several previous reports, we observed that mf rates are higher in men than women. This was not due to the fact that women were treated with DEC more often than men. In addition, of the 19 persons without any evidence of f&rial infection, 16 (84Oh) were women. If a genetic in&ence prevails on the development of the disease,it could be related to the sex chromosome. No significant difference in mf rate was found among the age classes we defined, another observation in favour of the hypothesis of long term susceptibility or resistance. This is in partial agreement with LANGUILLON (1957), who reported a regular increase in mf rate with age till 20 and a stable figure thereafter, but in contrast to the observations of RIPERT et al. (1977)? who found an increase in L. loa microfilaraemia with age in males only. In contrast to the stability of individual L. Zoa microfilaraemia, M. perstans mf appeared or disappeared in 26.7% of the sample in a period of one year. Considerable variation in the M . perstansmf rate was also found in different regions of Gabon (RICHARD-LENOBLE et al., 1980). It seemslikely that clearance of M. persms mf with time is essentially related to an active host mechanism of mf destruction, rather than to a genetic influence. Different prevalences could then be ascribed to different rates of infection. The mechanism of elimination of these unsheathed mf could also be different from that of the sheathed L. loa. This may explain the generally low microfilaraemias observed for this txuasite (Table 5) and the important inversion facto;. ’ ’ Serological values could not be correlated with the presence- or absence of one or both mf species. Similarlv. the sianificant (factor 3 or more) individual changes-i& antihdy titre‘observed in onli 5% of the casescould not be correlated with either the appearance or disappearanceof one or both parasite species, or to DEC therapy. It is possible that L. Ioa adult worms play a role-in the induction of the synthesis of antibodies reacting in our serological test, since amicrofilaraemia does not imply the absence of the corresponding adult worm. It has frequently been reported that people presenting clinical manifestations such as Calabar oedemasor ocular passageof the macrofiIaria rarely ha&our L. loa mf ‘h c&ulating blood (GALLIARD. 1932; FAIN & MAERTENS. 1973: EVELAND et al.,. 1975j but may present positive
446
FILARIASIS
IN EASTERN GABON
serology (FAIN, 1978). In hospital, we have examined 18 patients with proven Calabar oedema and found no circulating L. Zoumf in 16 (Flocard & Van Hcegaerden, unpublished). We observed that 80% of the L. Zou amicrohlaraemic cases had significant antibody titres and symptoms of fihtrial infection, which suggests that diagnosis by detection of circulating mf only hugely underdiagnoses the prevalence of the disease. Achowledgments The authors are indebted to Dr E. Frost for his suggestions and help in reviewing the manuscript. References Adolph, P. E., Kagan, I. G. & McQuay, R. M. (1962). Diagnosis and treatment of Acanthoche&memaperstans hlariasis. American Journal of Tropical Medicine and Hygiene, 11, 76-88. Brumpt, E. (1904). Les filarioses humaines en Afrique. Comptes rendus de la Soci& de Biologie, 56, 758-760. Brengues, J. (1975). La 6lariose de Bancroft en Afrique de I’Ouest. M&wires ORSTOM, 79, 50. Brengues, J. & Gidel, R. (1972). Recherches sur Setaria labiatopapillosa (Perroncito, 1882) en Afrique occidentale. Il. Dynamique de cette hlariose dans les conditions naturelles. Annaks de Parasitologic Hwnaine et Comparee, 47, 597-611. Cauchie, R., Rutsaert, J.., Thys, O., Bonnyns, M. & Perier, 0. (1965). Enctphahte d Lou loa trait&r par I’association de cortisone et de carbamazine. RevueBelge dePathokgie et de M&iecine Exp&nentak, 31, 232-244. Chan, S. H., Dissanaiake, S., Mak, J. W., lsmail, M. M., Wee, G. B., Srinivasan, N:, Soo, B. H. & Zaman, V. (1984). HLA and tilariasis m Sri Lankans and Indians. SoutheastAsian Journal of Tropical Medicine and Public Health, 15, 281-286. Dukes, D. C., Gelfand, M., Gadd, K. G., Clarke, V. de V. & Goldsmid, J. M. (1968). Cerebral fllariasis caused by Acanthocheiknema perstans. Central African Journal of Medicine, 14, 21-27. Eveland, L. K., Yermakoy, V. & Kenney, M. (1975). Loa loa infection without ~crohlaraemia. Transactionsof the f5~~5Gcsety of Troprcal Medicine and Hygiene, 69, Fain, A. (1978). Les problemes actuels de la lease. Bulletin de POrganisation Mondiak de la Sante, 56, 155-167. Fain, A. (1981). Epid6miologie et pathologie de la lease. f;7~2$ de la SocseteBelge de Medecme Tropicale, 61, Fain, A. & Maertens, K. (1973). Notes sur la ponte des microfilaires chez Loa ka et sur le degre de maturit6 des vers en migration. Bulletin de la Soci& de Pathologie Exotique, 66, 737-742. Galliard, H. (1932). Recherches sur les filarioses au Gabon occidental. Bulletin de la Soci&%de Pathologie Exotique, 25. 167-174. Gelfand, M. & Wessels, P. (1964). Acanthocheilonema perstansin a European female: a discussion of its possible pathogenicity and a suggested new syndrome. Transactionsof the Royal Sock@of Tropical Medicine and Hygiene, St? 552-556. Genuhni & Vuylsteke (1965). Quoted by RICHARD-LENOBLE et al., 1980.
Gordon? R. M. (1955). The host-parasite relationship in filanasis. Transactionsof the Royal Society of Tropical Medicine and Hygiene, 49, 496-507. Gordon. R. M.. Chwatt. L. 1. 81 Tones.C. M. (1948). The results of a preliminary eiomological survey‘of loiasis at Kumba, British Cameroons, together with a description of the breeding places of the vector, and suggestionsfor future research and possible methods of control. Annals of Tropical Medicine and Parasiwlogy, 42, 364-376. Goussard, B., lvanoff, B., Frost, E., Garm, Y. 81Bourderiou, C. (1984). Age of appearance of IgG, IgM, and IgE antibodies specific for Loa lou in Gabonese children. Microbiology and Immunology, 28, 787-792. Hawking, F. (1981). Chemotherapy of filariasis. Antibiotics and Chemotherap+ 30, 135-162. Languillat, G., Garm, Y., Tursz, A., Beam&, B. & Larivitre, M. (1978). Enquete sur l’etiologie de l’hypofecot&C au Gabon oriental. Revue d’Epidemiologie et de Sante Publique, 26, 273-282. Languillon, J. (1957). Carte des 8laires du Cameroun. Bulktin de la So&J de PathologieExotique, 50,417-427. Ottesen, E. A., Mendell, N. R., MacQueen, J. M., Weller, P. F;, Amos, D. B. & Ward, F. E. (1981). Familial predisposition to lilarial infection - not linked to HLA-A or -B locus specificities. Acta Tropica, 38, 205-211. Pampiglione, S., Najera, E., Ricciardi, M. L. & Junginger, L. (1979). Parasitological survey on Pygmies in Central Africa. Ill. Bambuti group (Zaire). Rivista di Parassiwl+ gia, 40, 187-234. Richard-Lenoble, D., Kombila, M., Carme, B., Gilles, J. C. & Delattre, P. Y. (1980). Prevalence des filarioses humaines sanguicolesau Gabon. Bulktin de la So&% de Pathologie Exotique, 73, 192-199. Ringenbach, J. & Guyomarch (1914). La filariose dans les regions de la nouvelle front&e Congo-Cameroun. Observations sur la transmission de la Micro$luria diuma et de Microfilaria~ perstans. Bulletin de la So&J de Pathologie Exotique, 7, 619-626. Riwrt. C.. Ambroise-Thomas. P.. Riedel. D.. RousselleSauer, C., Zimflou, A. & ibrahima, H: (1977). Epid& miologie des filarioses a L. loa et D. perstansdans sept villages de la province du Centre-Sud du Cameroun. Bulktin de la Societede Patholoeie Exotioue. 70.504-515. Rodhain, F. (1980). Hypotheses c&cemani I’&ologie dynamique des infections % Lou. Bulletin de la So&Q de Pathologie Exotique, 73, 182-191. Van Bogaert, L., Dubois, A., Janssens,P., Radermecker, J., Tverdy, G. & Wanson, M. (1955). Encephalitis in Loa loa iilariasis. 3ournal of Neurology, Neurosurgeryand Psychiatry, 18, 103-119. Van Hoegaerden, M. & AkuC, J. P. (1986). Lack of evidence for transplacental transfer of microfilarial antigens in filariasis due to Loa loa and Mansonellaperstans.Tropical Medicine and Parasitology, 37, 121-123. Van Hoegaerden, M. & lvanoff, B. (1986). A rapid, simple method for isolation of viable microfilariae. American 3ournal of Tropical Medicine and Hygiene, 35, 148-151. Weil, G. J., Hussain, R., Kumaraswami, V., Tripathy, S. P., Philips, K. S. & Ottesen, E. A. (1983). Prenatal allergic sensitization to helminth antigens in offspring of parasite-infected mothers. 3ournal of Clinical Znvestigarion, 71, 1124-1129. Accepted
for publication 15 May 1986
TRANSACTIONS OF THE ROYALSOCXETY OF TROPICALMEDICINEAND HYGIENE(1987)81, 441446
Filariasis due to Loa loa and Mansonella perstans: distribution in the region of Okondja, Haut-Ogoou6 Province, Gabon, with parasitological and serological follow-up over one year
‘International
M. VAN HOEGAERDEN', B. CHABAUD’, J. P. AKUE’ AND B. IVANOFF’ Medical Research Centre, BP. 769, Franceville, Gabon; ‘Hopita G&&al d’Okomija,
Gabon
Abstract The prevalence of Loa loa and Mansonella perstans filariasis has been determined in 6 rural villages in eastern Gabon. Between 18.9 and 27.2% of people carry L. loa microfilariae with an overall microfilarial rate of 25.1%. The microfilarial rate for M. perstans was more variable, between 33.3 and 62.2% (average 49.1%). No significant difference was seen in the microfilarial rate with age over 15 years for either parasite, but men were infected more frequently than women. Anti-L. loa antibody titres were measured, using a homologous microfilarial antigen in ELISA. Taking the parasitological and immunological evaluations together, only 10%of the sample population appear to be free of these filarial infections. L. loa and M. perstans microfilaraemia and corresponding serology were also investigated twice in 150 people at a one-year interval. 99.1% of the caseswho had no circulating L. loa microfilaria in March 1984 still did not show any 12 months later. Similarly, 97.1% of the untreated, microfilaraemic casesstill harboured this parasite a year later. The samewas not observed for M. perstans, since microiilariae appeared or disappeared in 26.7% of the cases. This suggests different dynamics for the two filarial infections. Variation in individual anti-L.loa antibody titres was low. The possibility of a genetic influence on the expression of loiasis is discussed.
Introduction Filariasis due to Loa loa was first reported in Africa by BRUMMPT(1904). The endemic zone of the parasite tid its insect vector have been delimited by-several authors. It can be calculated that 20 to 30 million people live in endemic areas in West and Central Africa and a conservative estimate is that 3 million people ha&our the parasite (FAIN, 1981). Clinical manifestations are usuallv classified as benign (such as the characteristic &bar oedema of the face, the wrist or the ankle, or the passageof the adult worm under the conjunctiva), although serious pathologies have occasionally been reported. Diethylcarbamazine citrate (DEC), the drug commonly used for treatment of loiasis, may provoke severe shock reactions, sometimesending in coma or sudden death due to encephalitis (VAN BOGAERTet al., 1955; CAUCHIEet al., 1965). Moreover, DEC is inefficient in eliminating Mansonella perstans (HAWKING, 1981), another filarial parasite common in Western Equatorial Africa. Although various authors suggested the non-pathogenic character of this latter organism, others (ADOLPH et al., 1962; GELFAND& WESSELS, 1964; DUKES et al., 1968) and ourselves (unpublished) have associated the presence of micr&l&ae (mf) of this species with certain clinical manifestations. RICHARD-LENOBLEet al. (1980) have determined the prevalence of circulatina mf in some regions of Gab& but villages in the re$on of Okondja Gere not investigated. This paper analyses the distribution of both filarial diseases, associating parasitological and immunological evaluations. Most studies of L. loa prevalence show that a maximum of about 25% of people ha&our circulating mf although it was estimated, in Cameroon, that each
individual was exposed to an infecting bite every 5 davs ~G~RD~N et al.. 19481. On the other hand. pr&&nce of M. Persians is highly variable with & rates varying from 6.3 to 81.2% in different regions of Gabon (RICHARD-LENOBLEet al., 1980). GORDON (1955) hypothesized that the development of the filaria Wuchereria bancrofti could be related to the genetic background of the individual. It has been extensively demonstrated in experimental models that the susceptibility to hehninth infections is related to the major histocompatibility complex. A possible association of histocompatibility antigens with filarial elephantiasis has been reported recently (CHAN et al., 1984). It could be postulated in the caseof loiasis (but not mansonellosis) that individuals have a genetic susceptibility (or resistance) to the appearanceof mf in peripheral blood circulation. In order to determine whether untreated individuals lose or acquire L. loa and/or M. perstans mf, we have also investigated the parasitaemia and serology of 150 adults living in an endemic area, over a period of one year. Subjects, Materials and Methods Study area 6 villages, each of about 200 to 500 inhabitants, were selected easf of Okondja, in a region where loiasis was highly endemic. These villages are situated along a dirt road, the farthest being 80 km eastof Okondja,towardsthe borderof the Democratic Republic of Congo, at an altitude of approximately 400 m above sea level (Figure). Forest characteristics vary among villages: Aboumi, Oboli and
Ontogo,closerto the Congoleseborder,aresituatedin dense forest with high canopy, habitat favourable for Chtysops
silacea and C. dimidiata (common vectors of L. loa). In
contrast, Obili, Bouala and Ayanabo, closer fo Okondja, are situated in less dense forest with a lower canopy and thus a less favourable habitat for the insect vector. Detailed
442
FILARIASIS
IN
EASTERN
GABON
conical tubes and 200 pl saponin (2% w/v) were added for haemolysis. Tubes were then centrifuged (10 min, 500g) and the supernatant discarded. The entire pellet was scanned under a microscope and the number of mf counted. Parasite species was determined by size and movement characteristics. serology A crude soluble extract of L. loa was prepared from purified mf (VAN HOEGAERDEN & IVANOFF, 1986), frozen and thawed 3 times and ultrasonicated (Soniprep 150, MSE, 23 kHz, amplitude varying from 8 to 14 pm, 5 x 1 min at O’C). After ultracentrifugation (lO’gx45 min at 4”C), the supematant was stored at -85°C. Protein content was determined by the Biorad (Coomassie blue) method using BSA as a standard. Antibody levels were determined by enzyme-linked immuoosorbent assay (ELISA) microtirration using 0.1 ug of protein to coat wells. Serum samples were diluted l/20 and l/2000 respectively for IgE and IgG + IgM determinations. Diluting buffer was PBS containing 2% gelatin and 0.5% Tween 20. Secondantibody was purified IgG fraction of goat anti-human IgG + IgM, or anti-human IgE coupled to alkaline phosphatase (Sigma). Results were expressed as O.D. obtained for patient’s serum divided by the average O.D. of three serum samples obtained from Caucasianswho had neverlived ia regionsendemicfor L. loo or M. perstam. The test was performed on serum samples not only from people presenting M. perstunsmicro&uremia alone but also
from individuals harbouring intestinal hehniaths or other parasites that may be found in Gabon.
Figure. Map of GabonshowingHam-OgooueProvinceand the 6 villages where the surveyswere conducted.
entomological studies of Chryf0f~ density or infection rates were not performed. The same 6 villages were visited in March 1984 and March 1985. Sample choice Surveys were announced 2 weeks in advance in an attempt to ensure that most of the population was present in the villages. In each village, the chief and notables were infomed of the purposes of the study which were then communicated io local dialect, by the chief, to the community, Volunteers were registered and bled m the order they presented, without distinction of ageor sex, chiefs and elders being generally first. All people presenting were included in the surveys. 411 individuals were seen. Of the 237 people investigated in the 6 villages during March 1984, 150 spontaneously presented again a year later
and blood sampleswere collected.They were askedif they
had rece.ivedDEC in the preceding year. Those individuals who presented L. IOUmicrolilaraemia in 1984but not in 1985 were questioned again in April 1985, together with family members and, when present, the nurse of the village, to ascertain whether they had received DEC. People who had lost M. perstunsmf were not recalled as this parasite is not affected by DEC (HAWKING, 1981). Blood samples Venous blood (10 ml) was collected in Venoject vacuum tubes, without additive and on EDTA, between 9 a.m. and 4 p.m. Blood was allowed to clot at ambient temperature for 1 to 8 h, after which it was centrifuged at 4°C for 15min at 500 g. Serum samples were immediately frozen at -20°C. Micro6laraemia was determined by direct reading of 10 ul uncoagulated, fresh blood between microscope slide and coverslip. When more than 20 mf were found, the result was multiplied by 100 and parasitaemia expressed in mflrnl. When fewer than 20 mf were found,. Knott’s concentration technique was used: 1 ml blood was diluted with 9 ml PBS in
Results L. loa and M. perstansmf were counted in blood samples from 411 people (287 women and 124 men), whose ages ranged from 9 to 70 years. The prevalence
of microfilaraemia is given in Table 1. On the whole, 25% of the sample population was parasitized with L. loa mf and 49% with M. perstans mf (14% bad mixed infections). Little variation was observed in the L. lou mf rate between villages or surveys, whereas variation in M. perstans prevalence was considerably higher. Microfilaraemia varied between 1 and 86 2OO/mlfor L. loa and 1 to 18OO/mlfor M. perstans (82% of the sampleswere below 20 mf/ml for the latter parasite). The distribution according to sex and age is given in Table 2, showing a higher rate in males than females and a relatively constant mf rate among the different age groups (only 3 individuals were under 15 years of age in the 9 to 25-year age group). Only 40% showed no circulating mf during these surveys. Antibody levels detected by ELISA were between 1 and 20 times the background obtained with the reference negative samples for IgG + IgM, and between 1 and 40 times for IgE. Only those samples giving an 0 .D . at least 3 times superior to the controls were considered positive. The technique does not
appear to distinguish carriers of the two parasite species but cross-reaction was not observed with serum from people harbouring intestinal parasites or other parasitic diseasesthat may be found in the area. No Onchocerca volvulus or Mammella
strepwcerca mf
were found in skin snips of 300 people during the 1985 survey (M. Van Hoegaerden, C. Dupont & A. Dupont, unpublished). Results of immunological and parasitological investigation were obtained for 198 of the 237 persons seen in 1984 and are reported in Table 3. Taking these two
ht.
VAN
HOEGAERDEN
et
443
al.
Table l-Parasitological prevalence of L. loa and M. perstunsmf ia 6 villages ia eastern Gabon determined during two surveys made at aa interval of one year* No.
L. loa
examined
(including
it
17 9 (25Wo) (26.2%)
ggy Bouala Ayanabo
1;: 2:
22 (27.2%) 3; [p%,, . 0 14 (23.0%)
Total
411
Village
;:p
*All individuals
103 (251%)
seen in 1985 were different
Table 2-Distribution
M. perstans
Mixed hfCCti0n5
Negative
29 12 (33.3%) (44.6%)
; g;:::$q
27 19 (52.8%) (41.6%)
67 36 (44.4%) (51.1%) 23 (62.2%) 35 (57.4%)
19 10 (12.3;) (14.5%) 7 (18.9%) 10 (16.4%)
49 33 (40.7%) (37.4%) 14 (37.8%) 22 (36.1%)
202 (49.1%)
58 (14~1%)
164 (399%)
mixed infections)
from those seen in 1984.
of mf carriers according to sex and age
No.
L. loa
M. pentans
Sex
Age
examined
(including
Males
9-25 26-40
14 :i
4 (28.6%) 2 (16.7%)
9 (64.3%) 4 (33.3%)
41-55 >55
62
26 13 (419%) (36.1%)
22 35 (56.5%) (61.1%)
20 (32:3;)
Total
124
45 (36.2%)
70 (56.5%)
39 (31.5%)
8 11 17 22
14 (359%) ;; yj
19 34 37 35
Females
9-25 26-40 41-55 >55
ii29 i:
Total Total M + F
Table 3-Parasitological ssmpler
L. loll
M. pentam
-
-
+ + + + -
evidence
-
58 (20.2%)
132 (46.0%)
125 (43.6%)
202 (49.1%)
164 (399%)
19 ( 96%) 68 (34.3%)
-
+
3 ( 15%) 25 (12.6%)
+ +
+
4 ( 2W0) 21 (10.6%)
+ +
+
6 ( 3.0%) 52 (26.3%)
Table 4-Evohtioo of L. loa and M. perstunsmicrotilaraemia in 150 individuals over a period of one year 1985 Negative
L. loa 1984 M. @arms
1984
(48.7%) (41.5%) (43.5%) (43.2%)
103 (25.1%)
+
Negative
4 (28.5%) ;: [;g:“:o]
411
Number (percent)
SCCOlOgiCal
37 (45.7;)
Negative
287
and seroiogicai evaluati& of 198 se-
Parasitological evidence
(20.5%) (13.4%) (20*0%) (27.2%)
mixed infections)
Positive
Positive
112 10**
3* 25
Negative Positive
58 15
2s 52
l 213 caseshad taken partial DEC trcatmcnt less than a year before March 1984. **9/10 cases had a complete or partial DEC treatment course between the surveys.
criteria together, it appears that only 10% of this sample population had no evidence of filarial infection, whereas 35% had serological evidence of past or rTmteTearia! infection in the absenceof circuIating mammg 55% camed one or both mrcrofilarial species. Evolution VidtUlS
of parasitaemia and serology in 150 indi-
150 people were seen in both March 1984 and March 1985. The evolution of L. loa and M. petstans microlilaraemia, over this period, is shown in Table 4 and average mf counts (with standard deviation) are given in Table 5. L. loa. Of the 35 people parasitized with L. loa mf in 1984, 10 no longer had this organism in 1985. It was later confirmed that 9 of them (whose microfilaraemia varied from 1 to 60 OOO/mlbefore chemotherapy) had taken complete or partial (but at least microfilaricidal) DEC treatment in the interval. The last case(who had 3 mf/ml in 1984) had not received treatment but complained of pruritus and oedema, not pitting under pressure, which is typical of loiasis. This individual’s serology was positive for IgG in both 1984 and 1985. Among the 25 individuals who remained positive over this period, the microlilaraemia increased in 16 by a factor varying from l-3 to 3800 (although 4 had taken DEC) and decreasedin 9 by a factor of 1.7 to 1500 (2 had undergone complete treatment and one had received partial treatment). Of the 115 individuals not showing this parasite in 1984, 3 (2.6%) were carriers a year later with parasitaemias of 5,26 and 77 mf/ml. However, 2 of the 3 had taken
444
FILARIASIS
Table S-Mean
IN
EASTERN
GABON
(with standard deviation) of positive L. loa and M. perstans microtilaraemias
Infection
No.
Year
L. loa alone
in 1984 and 1985
examined
mean k s.d.
1984
16
5542 + 14983
1985
11
1858* f 3604
I 1984
19
M. L.perstans: loa:
3334 1336 152 +Z!I419
1 1985
17
L. loa: M. perstans:
1837 + 2314
1984
48
Mixed
perstans
100 + 142
19 -t 48
60 78 + 211 i 1985 *The drop in averageL. lou microfilaraemia observed between 1984and 1985is artefactual, due to DEC treatment done
M.
administered to the highly parasitized people. Table 6-Antibody titre changes in 133 samples at one year interval (X = 1985 value11984 value) IgG + IgM
x>3
SetOlOgy
Table 6 shows the evolution of individual antibody titres over a period of one year in 133 samples: a significant change (3 times or more) in IgG + IgM titres was noticed in 5 (3.8%) casesonly. IgE titres changed by the same factor in 10 (7.5%) cases. No correlation was found between the change in antibody titre and the appearanceor disappearanceof circulating mf of either species.
kE
4
3>X>l
2:
x=1 l>X>O*3
:i
35
x-zo.3
:I: 7
1
Discussion
partial (microfilaricidal) DEC treatment less than a year before the 1984 survey. In the third case, no history of DEC therapy could be identified. M. perstam. Of the 67 casesharbouring mfin 1984, 15 (22.4%) had none a year later. Of the 52 who remained parasitized, the microfilaraemia remained identical in 6, decreasedby a factor of 1.2 to 180 in 12’ casesand increased by a factor of 1.1 to 60 in 34 cases. M. pentans mf were found in 25 (30.1%) of the 83 people without them in 1984. Thus, 40 (26.7%) of 150 people inverted their status regarding M. perstans microfilaraemia. In 47 (31.3%) individuals, both parasites were absent from blood circulation in 1984 and 1985. Serology was negative in both 1984 and 1985 in 8 (17%) of these 47. Table ‘I--Reported L. loa mf
in Central Africa Percentage of adult mf carriers
Area
17.8 to 23.3 3.4 to 26.4 5.4 to 24.4 27
Gabon Gabon
Gabon Gabon Gabon
Ham Ogoout, Gabon Cameroun Cameroun Western Northern Northern
rates
Loiasis is widespread in tropical Africa (RODHAIN, 1980), but its prevalence, basedon the detection of mf in blood circulation, does not usually exceed 27% (Table 7). We too found values around 25%. The detection of circulating mf alone does not reveal the actual prevalence of the diseasesince a large proportion of the population, possibly the majority, harboured the adult worm in the absenceof circulating mf (FAIN, 1981). We found that 78% of individuals not harbouring circulating mf have immunological evidence of past or present exposure to the parasites (Table 3) and specific tiarial symptomatology was identified in most of these cases.Only 10%of the total sample population have neither parasitological nor immunological evidence of aaria infestation. This is in agreementwith GOUSSARD et al. (1984) who found
Zaire Zaire Zaire
12 to 27 25.1
25-9 9.7 to 16.8 25 (1971)* (1972)*
10 (7 to 23) 88 (71 to 97)
Reference 1914 RINGENBACH & GUYOMARCH, GALLIARD, 1932 GENTILINI et al., 1965 LANGUILLAT et al., 1978 RICHARD-LENOBLE et al., 1980
Present communication LANGUILLON, 1957 RIPERT et al., 1977 FAIN, 1978 PAMPIGLIONE PAMPIGLIONE
*This population belonged to a particular ethnic group, the Bambutis of the Pygmy race.
et al., 1979 et al., 1979
M. VAN HOEGAERDEN et al.
that, by the age of two, 95% of the population had some antibodies against L. loa. The observation that only 25% of the population have circulating mf cannot be ascribed to absenceof exposure since it was calculated that, at certain times of year, inhabitants of endemic areaswere exposed to an infecting bite every 5 days (GORDONet at., 1948). Since L. loa adult worms are estimated to hve up to 15 years (EVELAND er al., 1975), all individuals would consequently be expected to harbour parasites and, therefore, mf. We found mf rates to remain constant whatever the ecological conditions of the villages, over a period of one year and independently of the individual’s age. If lower mf rates can easily be explained by variations in the presence of the transmitting insect (savannah or urban areasfor example), figures higher than approximately one quarter of the population have not, to our knowledge, been published for this part of Africa. The question may then be raised whether this constant maximum of 25% of adults presenting L. Zoa microfilaraemia is linked to a long term mechanism of host susceptibility or defence towards the appearance of mf in peripheral circulation. Susceptibility could be induced by prenatal sensitization through transplacental transfer of specific antigens. This would influence the outcome of the disease (WEIL et al., 1983). but we found no evidence of such transfer in 92 mot&-cord blood samples (VAN HOEGAERDEN & AKUE, 1986). Although alternative mechanisms such as immune prophylaxis induced by zoonotic species (of simian origin for instance) or concomitant immunity induced by existing adult worms against infective larvae and mf could explain the observed stability, the hypothesis of genetic influence may be invoked. GORDON(1955) suggested that the development of Wuchti bancrofti infection was related to the genetic background, dividing a population into “good” hosts (presenceof the parasite and absenceof clinical manifestations) and “bad” hosts (pathological evidence and low or non-existent microfilaraemia). BRENGUES (1975) observed a family clustering of susceptibility to W. bum-of%, but added a third category defined as refractory (neither clinical nor parasitological evidence). An identical association was demonstrated in cattle naturally infected with Setaria labiatopapillosa (BRENGUES& GIDEL, 1972) and in numerous human helminth infections and experimental models of fdariasis. CHAN et al. (1984) found a significant difference in the frequency of HLA-B15 histocompatibility antigen between patients with filarial elephantiasis and normal controls. Similar evidence of family clustering, compatible with genetic transmission of susceptibility, to Bancroftian filariasis was found by OTTESEN et al. (1981), although it was not associated with the HLA-A or B locus. The existence of a genetic influence on the outcome of a filarial infection could explain the remarkable constancy of the maximal L. loa prevalence (a quarter of the population) and, possibly, the exceptionally high figure found by PAMPIGLIONE et al. (1979) in a genetically homogenous group of pygmies in Northern Zaire. The study of genetically controlled long term resistance or susceptibility to L. loa mf would necessitatea larger epidemiological survey, but analy-
445
sis of the appearance or disappearance of mf in the same individual over a certain period could help to verify the hypothesis. We have therefore investigated L. loa and M. perszansparasitaemia and serology in 150 people over one year. During this period, circulating L. Zoamf appeared or disappeared in only 1.3% of the sample, in the absence of specific DEC treatment. The stability in 99.1% of the L. loa amicrotilaraemic people suggests the existence of a mechanism for long term resistance to the appearance of mf, whereas the maintenance of the presence of circulating L. loa mf in 97.1% of the untreated cases over one year implies a long term mechanism of susceptibility. A studv of familial transmission of this mutative susceptibility/resistance towards the appear&ce of L. loa mf in peripheral blood or the development of symptomatology could help understand the dynamics of this parasitic disease. In agreement with several previous reports, we observed that mf rates are higher in men than women. This was not due to the fact that women were treated with DEC more often than men. In addition, of the 19 persons without any evidence of f&rial infection, 16 (84Oh) were women. If a genetic in&ence prevails on the development of the disease,it could be related to the sex chromosome. No significant difference in mf rate was found among the age classes we defined, another observation in favour of the hypothesis of long term susceptibility or resistance. This is in partial agreement with LANGUILLON (1957), who reported a regular increase in mf rate with age till 20 and a stable figure thereafter, but in contrast to the observations of RIPERT et al. (1977)? who found an increase in L. loa microfilaraemia with age in males only. In contrast to the stability of individual L. Zoa microfilaraemia, M. perstans mf appeared or disappeared in 26.7% of the sample in a period of one year. Considerable variation in the M . perstansmf rate was also found in different regions of Gabon (RICHARD-LENOBLE et al., 1980). It seemslikely that clearance of M. persms mf with time is essentially related to an active host mechanism of mf destruction, rather than to a genetic influence. Different prevalences could then be ascribed to different rates of infection. The mechanism of elimination of these unsheathed mf could also be different from that of the sheathed L. loa. This may explain the generally low microfilaraemias observed for this txuasite (Table 5) and the important inversion facto;. ’ ’ Serological values could not be correlated with the presence- or absence of one or both mf species. Similarlv. the sianificant (factor 3 or more) individual changes-i& antihdy titre‘observed in onli 5% of the casescould not be correlated with either the appearance or disappearanceof one or both parasite species, or to DEC therapy. It is possible that L. Ioa adult worms play a role-in the induction of the synthesis of antibodies reacting in our serological test, since amicrofilaraemia does not imply the absence of the corresponding adult worm. It has frequently been reported that people presenting clinical manifestations such as Calabar oedemasor ocular passageof the macrofiIaria rarely ha&our L. loa mf ‘h c&ulating blood (GALLIARD. 1932; FAIN & MAERTENS. 1973: EVELAND et al.,. 1975j but may present positive
446
FILARIASIS
IN EASTERN GABON
serology (FAIN, 1978). In hospital, we have examined 18 patients with proven Calabar oedema and found no circulating L. Zoumf in 16 (Flocard & Van Hcegaerden, unpublished). We observed that 80% of the L. Zou amicrohlaraemic cases had significant antibody titres and symptoms of fihtrial infection, which suggests that diagnosis by detection of circulating mf only hugely underdiagnoses the prevalence of the disease. Achowledgments The authors are indebted to Dr E. Frost for his suggestions and help in reviewing the manuscript. References Adolph, P. E., Kagan, I. G. & McQuay, R. M. (1962). Diagnosis and treatment of Acanthoche&memaperstans hlariasis. American Journal of Tropical Medicine and Hygiene, 11, 76-88. Brumpt, E. (1904). Les filarioses humaines en Afrique. Comptes rendus de la Soci& de Biologie, 56, 758-760. Brengues, J. (1975). La 6lariose de Bancroft en Afrique de I’Ouest. M&wires ORSTOM, 79, 50. Brengues, J. & Gidel, R. (1972). Recherches sur Setaria labiatopapillosa (Perroncito, 1882) en Afrique occidentale. Il. Dynamique de cette hlariose dans les conditions naturelles. Annaks de Parasitologic Hwnaine et Comparee, 47, 597-611. Cauchie, R., Rutsaert, J.., Thys, O., Bonnyns, M. & Perier, 0. (1965). Enctphahte d Lou loa trait&r par I’association de cortisone et de carbamazine. RevueBelge dePathokgie et de M&iecine Exp&nentak, 31, 232-244. Chan, S. H., Dissanaiake, S., Mak, J. W., lsmail, M. M., Wee, G. B., Srinivasan, N:, Soo, B. H. & Zaman, V. (1984). HLA and tilariasis m Sri Lankans and Indians. SoutheastAsian Journal of Tropical Medicine and Public Health, 15, 281-286. Dukes, D. C., Gelfand, M., Gadd, K. G., Clarke, V. de V. & Goldsmid, J. M. (1968). Cerebral fllariasis caused by Acanthocheiknema perstans. Central African Journal of Medicine, 14, 21-27. Eveland, L. K., Yermakoy, V. & Kenney, M. (1975). Loa loa infection without ~crohlaraemia. Transactionsof the f5~~5Gcsety of Troprcal Medicine and Hygiene, 69, Fain, A. (1978). Les problemes actuels de la lease. Bulletin de POrganisation Mondiak de la Sante, 56, 155-167. Fain, A. (1981). Epid6miologie et pathologie de la lease. f;7~2$ de la SocseteBelge de Medecme Tropicale, 61, Fain, A. & Maertens, K. (1973). Notes sur la ponte des microfilaires chez Loa ka et sur le degre de maturit6 des vers en migration. Bulletin de la Soci& de Pathologie Exotique, 66, 737-742. Galliard, H. (1932). Recherches sur les filarioses au Gabon occidental. Bulletin de la Soci&%de Pathologie Exotique, 25. 167-174. Gelfand, M. & Wessels, P. (1964). Acanthocheilonema perstansin a European female: a discussion of its possible pathogenicity and a suggested new syndrome. Transactionsof the Royal Sock@of Tropical Medicine and Hygiene, St? 552-556. Genuhni & Vuylsteke (1965). Quoted by RICHARD-LENOBLE et al., 1980.
Gordon? R. M. (1955). The host-parasite relationship in filanasis. Transactionsof the Royal Society of Tropical Medicine and Hygiene, 49, 496-507. Gordon. R. M.. Chwatt. L. 1. 81 Tones.C. M. (1948). The results of a preliminary eiomological survey‘of loiasis at Kumba, British Cameroons, together with a description of the breeding places of the vector, and suggestionsfor future research and possible methods of control. Annals of Tropical Medicine and Parasiwlogy, 42, 364-376. Goussard, B., lvanoff, B., Frost, E., Garm, Y. 81Bourderiou, C. (1984). Age of appearance of IgG, IgM, and IgE antibodies specific for Loa lou in Gabonese children. Microbiology and Immunology, 28, 787-792. Hawking, F. (1981). Chemotherapy of filariasis. Antibiotics and Chemotherap+ 30, 135-162. Languillat, G., Garm, Y., Tursz, A., Beam&, B. & Larivitre, M. (1978). Enquete sur l’etiologie de l’hypofecot&C au Gabon oriental. Revue d’Epidemiologie et de Sante Publique, 26, 273-282. Languillon, J. (1957). Carte des 8laires du Cameroun. Bulktin de la So&J de PathologieExotique, 50,417-427. Ottesen, E. A., Mendell, N. R., MacQueen, J. M., Weller, P. F;, Amos, D. B. & Ward, F. E. (1981). Familial predisposition to lilarial infection - not linked to HLA-A or -B locus specificities. Acta Tropica, 38, 205-211. Pampiglione, S., Najera, E., Ricciardi, M. L. & Junginger, L. (1979). Parasitological survey on Pygmies in Central Africa. Ill. Bambuti group (Zaire). Rivista di Parassiwl+ gia, 40, 187-234. Richard-Lenoble, D., Kombila, M., Carme, B., Gilles, J. C. & Delattre, P. Y. (1980). Prevalence des filarioses humaines sanguicolesau Gabon. Bulktin de la So&% de Pathologie Exotique, 73, 192-199. Ringenbach, J. & Guyomarch (1914). La filariose dans les regions de la nouvelle front&e Congo-Cameroun. Observations sur la transmission de la Micro$luria diuma et de Microfilaria~ perstans. Bulletin de la So&J de Pathologie Exotique, 7, 619-626. Riwrt. C.. Ambroise-Thomas. P.. Riedel. D.. RousselleSauer, C., Zimflou, A. & ibrahima, H: (1977). Epid& miologie des filarioses a L. loa et D. perstansdans sept villages de la province du Centre-Sud du Cameroun. Bulktin de la Societede Patholoeie Exotioue. 70.504-515. Rodhain, F. (1980). Hypotheses c&cemani I’&ologie dynamique des infections % Lou. Bulletin de la So&Q de Pathologie Exotique, 73, 182-191. Van Bogaert, L., Dubois, A., Janssens,P., Radermecker, J., Tverdy, G. & Wanson, M. (1955). Encephalitis in Loa loa iilariasis. 3ournal of Neurology, Neurosurgeryand Psychiatry, 18, 103-119. Van Hoegaerden, M. & AkuC, J. P. (1986). Lack of evidence for transplacental transfer of microfilarial antigens in filariasis due to Loa loa and Mansonellaperstans.Tropical Medicine and Parasitology, 37, 121-123. Van Hoegaerden, M. & lvanoff, B. (1986). A rapid, simple method for isolation of viable microfilariae. American 3ournal of Tropical Medicine and Hygiene, 35, 148-151. Weil, G. J., Hussain, R., Kumaraswami, V., Tripathy, S. P., Philips, K. S. & Ottesen, E. A. (1983). Prenatal allergic sensitization to helminth antigens in offspring of parasite-infected mothers. 3ournal of Clinical Znvestigarion, 71, 1124-1129. Accepted
for publication 15 May 1986