Specificity of the indirect haemagglutination test with Plasmodium falciparum test cells

Bull. Org. mond. Sant} 1974, 50, 513-519 Bull. Wid Hl1th Org.f Specificity of the indirect haemagglutination test with Plasmodium falciparum test cells Comparison of indirect haemagglutination and fluorescent antibody tests on African sera J. H. E. T. MEUWISSEN,1 ANNA D. E. M. LEEUWENBERG,l A. VOLLER,2 & Y. MATOLA3 The specificity of the indirect haemagglutination (IHA) test with Plasmodium falci- parum placental antigen-sensitized test cells was examined with sera from healthy blood donors and from patients with diseases other than malaria. Only I nonspecific antibody reaction was seen in more than 700 tests. A comparison of IHA titres and indirect fluor- escent antibody (IFA) titres with IgG and IgM conjugates on 503 sera from inhabitants of Mto Wa Mbu in Tanzania showed in successive age groups an increasing number of seropositive reactors with both tests. The increase in the proportion of positive reactions and in the mean titre levels started earlier but was more gradual in the IFA test than in the IHA test. Parasite carriers had higher antibody levels than people without an apparent parasitaemia. Parasite carriers in the younger age groups especially were more frequently seronegative in the IHA test than in the IFA test with anti-IgG conjugates. The reactivity of the IHA test with P. falciparum (Palo Alto/Aotus) antigen was higher than that with P. falciparum placental antigen and was thus closer to that of the IFA test with IgG conjugates. There is a need for a reliable and inexpensive method for the serologic assessment of malaria, for which we have investigated the possibility of using the indirect haemagglutination test (IHA). Pre- viously the weak point of this test concerned its reproducibility, but the introduction of glutaralde- hyde-fixed cells greatly improved the reliability of the test. A field test should not be hampered with elaborate laboratory equipment; as a way of overcom- ing this difficulty we introduced lyophilized test cells and control cells for use in the field (4) and we described a simple field method for the absorption of heterophile antibodies in minute quantities of plasma (Meuwissen & Leeuwenberg, unpublished document to WHO, 1973). In the present paper, we report on the specificity of the IHA test with fixed sheep cells sensitized with Plasmodium falciparum (Palo Alto/Aotus) anti- ' Laboratory of Medical Parasitology, University of Nijmegen, The Netherlands. 2 Nuffield Institute of Comparative Medicine, The Zo- ological Society of London, Great Britain. 3 East African Institute for Malaria and Vector-borne Diseases, Amani, Tanzania. gen and on the application of the IHA test to some 500 plasma samples collected from Africans of dif- ferent ages. The IHA titres are compared with those obtained in indirect fluorescent antibody (lFA) tests with anti-IgG and anti-IgM conjugates. MATERIALS ANID METHODS Sera. For the assessment of the specificity of the IHA test with P. falciparum (Palo Alto/Aotus) test cells, 545 sera from healthy Dutch blood donors and 189 sera from patients with diseases other than malaria were used. The IHA titres of the latter sera had been tested earlier (5). From the collection of serum samples from malarious areas, those of Mto Wa Mbu were chosen. In this isolated settlement in the Arusha region of Tanzania, medicated salt had been distributed since 1961 without achieving much permanent success (Lelijveld, 1971 a; 2). The 516 African blood samples were first collected in capil- lary tubes containing heparin; after centrifugation, a Lelijveld, J. (1971) Sero-epidemiological studies of ma- laria in Tanzania (thesis, University of Nijmegen). 3218 - 513 514 J. H. E. T. MEUWISSEN ET AL. the part with the supernatant plasma was sealed and stored in a deep-freeze until used. For dilution of the plasma, a volume of 20 lI was taken from the capillary with the aid of a micro- pipette that had a disposable tip (Eppendorf pipette). This plasma was added to 180 ,ul of phosphate buf- fered saline, pH 7.2, to give a 1 : 10 dilution. IHA test. For the indirect haemagglutination test the procedures described earlier were used (5). All the African sera were examined with the same batch of glutaraldehyde-fixed sheep cells (sheep No. 22). The samples were absorbed with control cells following the method described (Meuwissen & Leeuwenberg, unpublished document to WHO, 1973). As control cells, fixed sheep cells sensitized with Aotus control antigen (Ao5) were used. Sensitiz- ation of test cells was carried out with P. falciparum placental antigen. The sera were examined in per- manent poly(methyl methacrylate) "U " type plates.a From various brands and types of disposable plates, only Biocult Linbro " U " type plates suited our purposes.b In addition, some of the sera were also examined with cells sensitized with P. fieldi antigen and with another P. falciparum Aotus antigen (Palo Alto strain, Ao8). The results of this survey were expressed in the various age groups as proportions of positive sera (i.e., titres >1: 40) and as the mean titre index. The individual titre index was found by calculating the exponent to the base of 2, after having divided the titre by 10. A titre of 1: 20 (=1: lOx 21) thus cor- responded to a titre index of 1, while a titre of 1: 40 (1: 1Ox22) corresponded to a titre index of 2. The mean titre index was calculated by dividing the sum of individual titre indices by the number of observations. IFA test. All the African sera were examined in the IFA test with P. falciparum antigen (Palo Alto strain), obtained from an owl monkey (Ao2), and processed according to the method described by Voller (7). The samples were tested with commer- cially available monospecific anti-IgG and anti-IgM conjugates.c Although the IFA titres were positive at a plasma dilution of 1: 20, the same criteria had to be used for comparing the IHA and IFA results. In this study, therefore, IFA titres >1 : 40 were considered to be positive and for the calculation of the mean titre index all titres <1: 40 were given the individual index of 1. a Cooke Engineering Co., Alexandria, Virginia, USA. b Biocult Laboratories Ltd., Glasgow, Scotland. c Nordic Company, Tilburg, The Netherlands. Parasitological examination. Giemsa-stained thick blood films from the inhabitants of Mto Wa Mbu were examined at the laboratory in Amani. The para- site rate was determined for every age group. The parasite density was estimated for each individual and was expressed for each age group as the mean number of parasites per 200 leucocytes. RESULTS Specificity of the IHA test In Table 1 are given the results of the IHA test with P. falciparum (Palo Alto/Aotus)-sensitized test cells on 734 absorbed sera taken from patients and healthy blood donors. Only 1 serum showed a posi- tive titre; this was from a boy who showed only IgM immunoglobulins in his serum. All the other sera were negative including those from patients with diseases such as rheumatoid arthritis, syphilis, infec- tious mononucleosis, disseminated lupus erythema- tosus, leukaemia, and myelomatosis. Table 1. Specificity of the IHA test for malaria. Exam- ination of nonmalarious sera with test cells sensitized with P. falciparum (Palo Alto/Aotus) antigen after absorption with control cells sensitized with Aotus control antigen (Ao5) Sera and age No. IHA titres Non-specifically groups (years) examined < 1 :40 positive IHAtitres (> 1 :40) from patients a 189 188 1 from healthy blood donors: 15-20 25 25 0 21-25 123 123 0 26-30 97 97 0 31-40 133 133 0 41-50 102 102 0 51-60 50 50 0 61-70 15 15 0 total 734 733 1 a For diagnoses, see Meuwissen et al. (5). Examination of African sera Parasitological and serologic examinations were carried out on 516 blood samples collected from people of various ages. The IHA test on plasma samples after absorption with control cells showed SPECIFICITY OF THE IHA TEST Table 2. Persistant heterophile antibodies after absorption with control cells in the IHA test with Ao5 control cells Age groups (years) Total < 2 2-4 5-9 10-19 20-39 40-59 >60 no. examined 53 85 132 107 80 45 14 516 no. positive at titre: 1:40 0 1 4 0 3 0 0 8 1:80 0 0 0 0 1 0 1 2 1:160 0 0 0 2 0 1 0 3 total no. positive(persistent heterophile 0 1 4 2 4 1 1 13 antibodies) % positive 0 1.2 3 1.9 5 2.2 7.1 2.5 no. of parasite carriers 0 0 3 1 1 1 0 6 that 13 samples (2.5 Y.) had persistent heterophile antibodies, which were not restricted to any par- ticular age group (Table 2). These 13 sera were ex- cluded when comparing the IFA and IHA tests. Fig. 1 shows the results of the tests with cells sen- sitized with P. fakciparum placental antigen on the remaining 503 sera. In the IFA test they were exam- ined with P. falciparum antigen as well as anti-IgG and anti-IgM conjugates. At the time of blood col- lection a high proportion of the population had a patent parasitaemia. The difference in parasite rates between children and adults was remarkably small; this also applied to the parasite density. In successive age groups the proportions of people with negative or low antibody titres showed a decrease, and with age there was a rise in the mean titre levels. At a serum dilution of 1: 40 the proportion of positive sera in the IHA test was higher than that obtained in the IFA test with lgM conjugates and lower than that observed in the IFA test with IgG conjugates. The mean titre indices indicate that, in the younger age groups, the IFA IgG antibody titres were higher than the IHA titres but in adults this was reversed. More- over, the data suggest that the reactivity in the IHA test started to increase in the age group 10-19 years; the titre level in the IFA test started to rise earlier but this increase occurred more gradually. Table 3 and Fig. 1 show a further analysis of the results on sera from people with positive or %with titrese 1.40 parasite rate - 50 -40 .0 - 30 1-. 20 - 10 10j7AjH 8-56IFA IHA + -. -+ parasitemia 6 2 6-11 12-23 2-4 5-9 10-19 20-39 40-59. 60+ months years Fig. 1. Results of the immunofluorescent antibody (IFA) tests with IgG and IgM conjugates and of the indirect haemagglutination (IHA) test in relation to age and parasitaemia. . . _ - - - .. ,_ 515 mFan t&t_;nd* J. H. E. T. MEUWISSEN ET AL. Table 3. Comparison of IFA IgM, IFA IgG, and IHA antibody titres in absorbed sera from people without (-) and with (+) parasitaemia (IFA IgM and IFA IgG tests with P. falciparum (Palo Alto/Aotus) antigen; IHA test with P. falciparum placental antigen) a No. Percentage with titres <1: 40 Mean titre index examined Parasite Age groups rate (%) IFA IgM IFA IgG IHA IFA IgM IFA IgG IHA - + + + + + - + + <3 months 3 1 (25) (100) (0) (67) (0) (67) (0) (0) (3) (2.7) (2.0) (2.0) (2.0) 3-5 3 2 (40) (100) (100) (33) (0) (33) (50) (1.0) (1.0) (2.0) (3.5) (2.7) (1.5) 6-11 19 8 30 84 50 58 13 63 50 1.4 1.6 1.9 3.0 1.7 2.5 12-23 9 8 47 67 25 44 13 67 13 1.9 3.0 2.6 3.5 2.0 3.1 2-4 years 56 28 33 61 67 29 11 50 25 1.7 1.8 3.2 3.8 2.3 3.3 5-9 70 58 45 60 38 24 5 44 29 1.7 2.3 3.4 4.4 2.4 3.4 10-19 60 45 43 47 31 15 2 33 9 1.9 2.5 3.8 4.4 2.9 4.3 20-39 56 20 26 34 15 11 5 14 5 2.4 3.2 4.1 5.0 4.9 5.9 40-59 34 10 23 33 20 0 0 12 10 2.7 3.2 5.1 5.0 5.8 6.6 >60 10 3 23 20 (67) 0 (0) 10 (0) 3.3 (1.7) 5.1 (4.0) 6.5 (5.3) total 320 183 36.3 50.4 37.9 20.6 5.5 35.4 19.8 2.0 2.4 3.7 4.3 3.4 4.0 a Figures in parentheses indicate data from > 5 people. negative blood smears for parasites. These groups differed in their serologic reactivity. In most age groups the mean titres observed with all 3 serologic tests were higher for parasite carriers than for those without a parasitaemia. Whereas there were among the parasite carriers about 4 times as many negative sera in the IHA test (19.8%) as in the IFA IgG test (5.5%), there were only about 11/2 times as many (35.4% compared with 20.6%) in persons without a patent parasitaemia. For further exploration of this point the relationship between trophozoite den- sities at 3 levels of IHA titres in various age groups was studied and showed that the geometric mean trophozoite density was generally lowest in adult trophozoite carriers with IHA titres of > 1: 160 (Table 4). Some of the sera were also examined in the IHA test with cells sensitized with P. falciparum (Aotus) anti- gen and with P. fieldi antigen. The results confirmed the earlier observation that P. falciparum (Aotus) test cells were more reactive than the cells sensitized with our batch of P. falciparum placental antigen and that the P. fieldi test cells were less sensitive (Meuwissen & Leeuwenberg, unpublished report to WHO, 1973). The results of the IHA tests with placental falci- parum antigen and with P. falciparum Palo Alto strain (Aotus) antigen were compared separately for persons with and without a patent parasitaemia at 3 levels of reactivity in the IHA test (Table 5). It is evident that the number of negative or weak positive reactions among parasite carriers was considerably lower with P. falciparum (Aotus) antigen than with the placental antigen. DISCUSSION Information on the specificity of the tests used in serologic surveys is of great importance. For a satis- factory assessment of the effect of malaria control, the test should be sufficiently specific even when malaria infections become rare. When the serologic test is not fully specific and the incidence of malaria infections in a population decreases, there may be a rise in the proportion of positive titres as a result of nonspecific reactions. For testing the specificity of the test, cells sensitized with the most reactive antigen available, P. falciparum (Aotus) antigen, were used. After absorption of the sera with con- trol cells, only 1 positive reaction from more than 700 sera was obtained. It is our impression that the specificity of the IHA test is sufficiently well estab- lished. Persistent heterophile reactions after absorption 516 SPECIFICITY OF THE IHA TEST Table 4. Comparison in various age groups of the geometric mean number of tropho- zoites per 200 leucocytes in trophozoite carriers at different levels of IHA titres (IHA test with P. falciparum placental antigen) a Mean trophozoite density (geometric means) AgegroupNo. examined No. of in trophozoite carriers, according to IHA titresAge group No. examined trophozoite carriers <1: 40 1 :80 > 1:160 1 :160 < 3 months 4 0 - - - 3-5 5 1 (268) - - 6-11 27 8 (81) (271) (20) 12-23 17 8 (21) (27) (64) 2-4 years 84 24 73 105 40 5-9 128 51 36 28 56 10-19 105 42 75 30 34 20-39 76 19 (71) (37) 13 40-59 44 8 (18) - 5 > 60 13 3 - - (24) total 503 164 52.5 41.8 27.0 a Figures in parentheses indicate data from < 5 people. were shown by 2.5 % of the African sera. These sera were excluded when comparing the IFA and IHA tests because their titre values were unreliable. How- ever, since in this study about half these sera were obtained from parasite carriers, it should be deter- mined in future studies to what extent the exclusion of these sera caused a loss of valuable information. The results of the examination of the African sera indicated some differences between the IFA and IHA tests. The IHA test with P. falciparum placental antigen showed twice as many non-reactors at a serum dilution of 1: 40 as did the IFA IgG test (29.7% as compared with 15.1 %). It is also striking that there was a high proportion of negative sera among parasite carriers under 10 years old; in the age group 5-19 years seronegative reactions were about 5 times as frequent in the IHA test as in the IFA IgG test. In the literature, data on comparative studies of IHA and IFA tests are extremely scarce. Sadun et al. (6) compared the results of IHA tests with P. falciparum-sensitized test cells and those of the soluble antigen fluorescent antibody (SAFA) Table 5. Distribution of persons without and with apparent parasitaemia according to IHA titre levels in tests using P. falciparum placental antigen and P. falciparum Palo Alto/Aotus antigen Persons without apparent parasitaemia Persons with parasitaemia Type of Distribution (%) according Distribution (%) according P. falciparum Total to titre level Total to titre level antigen used examined 1 80 examined 1 80 1:3 1 :<40 1:6 1:320 + 1 :<40 1:6 320 + Placental 320 47.5 24.1 28.4 183 28.4 29.5 42.1 Palo Alto/Aotus 136 44.7 a 21.2 a 34.1 a 58 15.7 a 32.9 a 51.4 a a Adjusted figures obtained after standardizing for age against the placental antigen group. 517 J. H. E. T. MEUWISSEN ET AL. test with P. falciparum antigen. They followed the serological response in 9 volunteers with P. falci- parum infections and 10 with P. vivax infections. The patients had intermittent parasitaemias for sev- eral months. It is reported that the patterns of anti- body development with time were similar in both tests and that the antibodies were detected at approxi- mately the same time after the onset of parasitaemia. Somewhat at variance with their comments, the 8 graphs accompanying the paper show that in 4 falciparum cases and in 3 of 4 vivax cases the IHA titres early during the parasitaemia were lower than the SAFA titres. in the falciparum cases, the IHA titres were higher than the SAFA titres after about 3 months. Wilson et al. (8) compared the titres in the IHA test using P. knowlesi test cells with those obtained in the IFA test using P. vivax antigen from sera taken from US soldiers with P. vivax infections. They concluded that the fFA test detected antibodies slightly more efficiently than the IHA test during the first 2 weeks after the onset of symptoms. How- ever, from the third week onwards and clearly after 7 months, the IHA test detected antibodies somewhat more efficiently than the WFA test. With the IHA test they also detected significantly high malarial antibody levels in patients who had previously had malaria; this they did not observe with the IFA test. The authors suggested that possible explanations might be the use of heterologous antigens, the occur- rence of nonprecipitating antibodies detected by the IFA test only, the inhibition of the IHA test by nonprecipitating antibodies, or the enhancement of IHA titres by the presence of autoantibodies in people with previous attacks. In the present study, the following possibilities may account for the dif- ferent reactivities observed in the IHA and IFA tests. (1) The reactivities may differ owing to the use of P. falciparum (Palo Alto/Aotus) antigen in the IFA test and the use of P. falciparum placental anti- gen in the IHA test. Table 5 shows that in the IHA tests with both falciparum antigens, a higher reactiv- ity was obtained with the P. falciparum (Palo Alto/Aotus) antigen. (2) Free circulating, soluble antigens in the sera of patients during parasitaemia might inhibit the IHA test more than the IFA test. Soluble antigens could be demonstrated in P. falciparum infections and the presence of S antigens was directly correlated with the degree of parasitaemia at the time of examination (3). The data in Table 4 do not rule out the possibility that soluble antigens could inhibit the IHA titre level. In most age groups the mean number of trophozoites was lowest in trophozoite carriers with the highest titre level. This could be the result of an absence of IHA inhibition by the soluble antigen; on the other hand, this could just as well be a feature of concomitant protective im- munity. Detailed longitudinal observations in experi- mental models are needed for the investigation of this point. (3) After a malaria infection, the IHA test may revert to negative sooner than the IFA test, but we have no evidence that this happens. (4) In a primary infection, it may very well take longer for the IHA test to become positive than for the IFA test. A low degree of affinity of malaria antibodies could remain inapparent in the IFA test but not in the IHA test. We now have preliminary evidence from observations in owl monkeys that in primary infections the IHA test could be negative at a 1: 40 dilution of the sera for a period of several weeks, while the IFA test is already positive. It is well known that the affinity of the circulating anti- bodies increases, in general, during the process of immunization. Hypothetically, it could be presumed that early in malaria infection the affinity of malaria antibodies might be sufficient for demonstration in the IFA test but insufficient to form stable antibody bridges between adjacent sensitized cells. Later in the primary infection and also during reinfections, antibodies would be reactive in both the IHA test and in the IFA test. Longitudinal observations on the course of the IHA test during primary infections and reinfections are therefore needed to verify this hypothesis. (5) Antimalarial drug administrations may have a greater influence on the IHA antibody titre level than on the IFA titre level. This requires further study since Desowitz et al. (1) observed a significant drop in the IHA titres as a result of antimalarial treatment. A similar decline in IFA titre levels has not been reported. The irregular distribution of chloroquinized salt for more than 10 years in the area where our sera were collected may also have had an influence on our IHA test results. 518 SPECIFICITY OF THE IHA TEST 519 ACKNOWLEDGEMENTS The authors thank W. H. Doesburg and J. A. M. van Druten from the Department for Applied Statistics, University of Nijmegen, for their assistance in the preparation of this paper and Dr W. H. Wernsdorfer for his critical review of the manuscript. RJESUMt SPECIFICITE DE L'EPREUVE D'HEMAGGLUTINATION INDIRECTE PRATIQUEE AVEC DES CELLULES SENSIBILISEES PAR PLASMODIUM FALCIPARUM: ETUDE COMPARIE DE L'EPREUVE D'HEMAGGLUTINATION INDIRECTE ET DE L'EPREUVE D'IMMUNOFLUORESCENCE INDIRECTE SUR DES SERUMS RECUEILLIS EN AFRIQUE Sur des s6rums provenant de donneurs de sang bien portants et de malades atteints d'affections non palu- deennes, les auteurs ont etudie la specificit6 de 1'6preuve d'h6magglutination indirecte (IHA) realis6e avec des cellules sensibilis6es par l'antigene placentaire de Plas- modium falciparum. Sur plus de 700 6preuves, une seule s'est r6v6l6e non specifique. D'autre part, les titres IHA et les titres IFA (immuno- fluorescence indirecte) obtenus avec des conjugu6s anti- IgG et anti-IgM ont 6te compar6s sur 503 6chantillons de serum humain prelev6s dans une localite de Tanzanie. Aussi bien dans les epreuves IFA que dans les epreuves IHA, il est apparu que le nombre des r6actions positives augmentait avec l'age. L'age auquel commence l'augmen- tation de la reactivit6 et du titre moyen d'anticorps est moins eleve avec 1'epreuve IFA qu'avec l'6preuve IHA. L'augmentation en fonction de l'age est plus graduelle dans 1'6preuve IFA. Chez les porteurs de parasites, les titres d'anticorps 6taient plus eleves que chez les sujets ne presentant pas de parasit6mie apparente. Dans l'epreuve IHA, la r6activit6 semble etre moins grande que dans l'6preuve IFA pratiquee avec un conjugu6 anti-IgG, en particulier chez les porteurs de parasites. Lorsque 1'epreuve IHA est faite avec l'antigene de P. falciparum (Palo Alto/Aotus), la r6activit6 est plus forte qu'avec l'antig6ne placentaire de P. fakciparum et se rapproche de celle obtenue dans l'epreuve IFA pra- tiqu6e avec un conjug6 anti-IgG. REFERENCES 1. DEsowrrz, R. S. ET AL. Milit. Med., 131: 1157 (1966). 2. DRAPER, C. C. & VOLLER, A. Amer. J. trop. Med. Hyg., 21: 696 (1972). 3. MCGREGOR, I. A. Brit. med. Bull., 28: 22 (1972). 4. MEUWISSEN, J. H. E. TH. & LEEUWENBERG, A .D. E. M. Trans. roy. Soc. trop. Med. Hyg., 66: 666 (1972). 5. MEUWISSEN, J. H. E. TH. ET AL. Bull. Wld Hith Org., 46: 771 (1972). 6. SADUN, E. M. ET AL. Milit. Med., 134: 1294 (1969). 7. VOLLER, A. & O'NEILL, P. Bull. Wld Hith Org., 45: 524 (1971). 8. WILSON, M. ET AL. Amer. J. trop. Med. Hyg., 20: 6 (1971).