In vivo response of Plasmodium falciparum to chloroquine in pregnant and non-pregnant women in Siaya District, Kenya*

Bulletin of the World Health Organization, 65 (6): 885-890 (1987) In vivo response of Plasmodium falciparum to chloroquine in pregnant and non-pregnant women in Siaya District, Kenya* R. W. STEKETEE,' A. D. BRANDLING-BENNETT,2' 3 D. C. 0. KASEJE, I. K. SCHWARTZ,2' 3 & F. C. CHURCHILL 5 Chemoprophylaxis using chloroquine (CQ) in suppressive doses has been recom- mended to protect pregnant women in malarious areasfrom the adverse effects ofmalaria during pregnancy. In a malaria-endemic area of western Kenya with CQ-resistant Plasmodium falciparum, we determined the prevalence and density offalciparum infection in gravid and nulligravid women and compared the in-vivo parasite response to CQ using two regimens: 25 mg/kg body weight (CQ25) divided over a period ofthree days (for high- density parasitaemias) and 5 mg/kg body weight (CQS) weekly for 4 weeks (for low- density parasitaemias). P. falciparum infections were present in 102 (42%) of 244 pregnant women. A greater proportion ofprimigravidae were parasitaemic (68%) than nulligravidae (50%, P =0.02) or multigravidae (33%, P1<-6). Primigravidae showed a higher geometric mean parasite density. In the CQ25 treatment group, failure to clear parasites by day 7 was more common in primigravidae than nulligravidae (P =0. 008) or multigravidae (P=0.15). In the CQ5 treatment group, primigravidae were more likely to show increasing parasite density than nulligravidae or multigravidae. In this area of Kenya, virtually all women in their first pregnancy are exposed to malaria and are at greatest risk for malaria infection; compared with other women, they show higher parasite densities and are least likely to respond to chloroquine treatment in areas ofchloroquine resistance. Malaria control strategies might be targeted to this group of women, but we are pessimistic about the efficacy of weekly CQ5 where there is chloroquine resistance. In areas with Plasmodium falciparum malaria, the vulnerability of pregnant women and their fetuses to malaria infection could result in stillbirth, abortion, or low-birth-weight infants (1). The World Health Organization has therefore recommended the use of an antimalarial drug in a regular suppressive dose throughout pregnancy for women living in such areas (2). Chloroquine (CQ) is the drug of choice at 5 mg/kg body weight (CQ5) weekly in areas where P. falciparum is sensitive to CQ. Weekly suppressive doses ofchloroquine in school- children have been effective in suppressing peripheral * From the Center for Infectious Diseases (CID), Centers for Disease Control (CDC), Atlanta, GA 30333, USA. 1 Malaria Branch, CID, CDC. Requests for reprints should be sent to this author. 2 Division of Parasitic Diseases, CID, CDC. 3 Kenya Medical Research Institute, Nairobi, Kenya. 4 Department of Community Medicine, University of Nairobi Medical School, Nairobi, Kenya. s Control Technology Branch, CID, CDC. parasitaemia and clinical malaria in areas with CQ- sensitive parasites (3). However, high levels of compliance in chemoprophylaxis programmes for children are difficult to maintain (4), and a com- munity trial of malaria chemoprophylaxis with the WHO-recommended regimen showed a low rate of coverage among pregnant women (5). CQ-resistant P. falciparum is now established across most of East and Central Africa where in-vivo parasite resistance has been observed in 10-60% of persons treated with CQ at a dose of 10 mg or 25 mg/kg body weight (CQ25) (6, 7). The failure rate of these higher dosage regimens suggests that weekly doses of CQ5 would not effectively maintain the peripheral blood free of parasites and reduce the parasite blood load, particularly in the placenta. Given a less than optimally effective drug and difficulty in delivering chemoprophylaxis, the strategy of antimalarial intervention in pregnant women in Africa must be assessed carefully. An understanding of drug efficacy in the target popu- 4844 -885 886 R. W. STEKETEE ET AL. lation is an important prerequisite to developing an effective antimalarial policy for pregnant women. To evaluate the efficacy of chloroquine in pregnant women living in a malaria endemic area of Kenya, we determined the prevalence of P. falciparum infection in gravid and nulligravid women in this district, and assessed the in-vivo parasite response to regimens of CQ25 divided over three days in women with high initial parasite densities and CQ5 weekly for 4 weeks in women with low initial parasite densities. MATERIALS AND METHODS The study was conducted in Siaya District, Nyanza Province, in Western Kenya from 13 June to 16 July 1986. The district is approximately 1000-1400 metres above sea level. P. falciparum is endemic and accounts for 70-90% of malaria infections, while P. malariae and P. ovale are transmitted with lower frequency. Although the area experiences two rainy seasons (March-May and October-November), malaria transmission occurs throughout the year. Parasitaemia screening. All pregnant women attending the prenatal clinic at the District Hospital on 13, 16, and 19 June 1986 were screened for P. falci- parum infection using a Giemsa-stained thick blood smear. All asexual parasites and leukocytes were counted in adjacent fields by oil-immersion micro- scopy until >300 leukocytes were counted and the asexual parasite density per mm3 of blood was esti- mated using a value of 8000 leukocytes/mm3 of blood. All nulligravid women 14-17 years of age at one local school were screened in the same way. In-vivo drug study. Women screened in the above fashion, who were infected with P. falciparum only and had no history of antimalarial drug use in the previous seven days and/or negative results on urine testing for chloroquine using a modified Haskins procedure (8) and who gave informed consent, were enrolled and followed on days 1, 2, 7, 14, 21, and 28. Gravid and nulligravid women with screening para- site densities < 1500/mm3 were treated with CQ5, weekly, for 4 weeks; women with parasite densities > 1500/mm3 of blood were treated with CQ25 divided over three days as 10, 10 and 5 mg/kg. All drug doses were administered by health staff. To avoid risk to women treated with weekly CQ5, these women were counselled to return to the clinic if they experienced any feverish illness or malaria symptoms and they were carefully questioned on each return visit regarding symptoms in the interval since the last visit. Women on the CQ5 regimen who exhibited an increasing parasite density or malaria symptoms were treated with a therapeutic dose of an alternative antimalarial drug (see below). Information collected at enrolment included name, age, reproductive history, and history of fever and antimalarial drug use in the previous seven days. On each follow-up visit, any fever or antimalarial drug use since the previous visit and pruritus or vomiting following CQ treatment were noted. At enrolment and on all follow-up visits except on day 1, the following specimens were obtained on all study subjects: Giemsa-stained thick and thin blood smears, and whole blood on filter-paper for chloro- quine determination (9). Non-pregnant status of the nulligravidae was confirmed on day 14 ofthe study by urine tests for human chorionic gonadotropin.a One 15-year-old who denied being pregnant was repeat- edly positive by urine testing and was included in the gravid group. In the CQ25 group, resistant infection (RH type or early RI type according to the WHO classification (10)) was defined as the presence of P. falciparum parasites in peripheral blood on day 7. Absence of parasites on day 7 but reappearance of parasites on day 14 or 21 was presumed to be due to either RI resistance or early new infection. The CQ5 regimen was evaluated on day 28 after 4 doses of drug; women on this regimen were categorized as follows: aparasitaemic, low-level parasitaemia (< 2-fold increase in parasite densities and/or density of <2000/mm3 of blood), or high-level parasitaemia (all others). Women on CQ25 with a resistant infection and women on CQ5 with a high-density para- sitaemia were treated with amodiaquine at a dose of 25 mg base per kg body weight (AQ25), divided over three days or with three tablets of pyrimethamine- sulfadoxine (fixed combination). Any woman who was parasitaemic on day 28 was treated with a therapeutic dose of amodiaquine or pyrimethamine- sulfadoxine. Statistical testing. Differences in proportions were tested using x2 and Fisher's exact test; differences in means were tested using Student's t-test. Significance was designated at P < 0.05. RESULTS Malaria parasite screening. A total of 244 pregnant women were screened and P. falciparum was identi- fied on peripheral blood smears in 102 (42%) (Table 1); 68% of 66 primigravidae were parasitaemic compared with 33% of 178 multigravidae (X2= 24.41, P< 10-6). Fifty percent of 98 nulligravidae a Sensi-Slide or Pregnosis from Roche Laboratories, Division of Hoffmann-La Roche Inc., Nutley, NJ, USA. Use of trade names is for identification only and does not imply endorsement by the U.S. Public Health Service or Department of Health and Human Services. RV VIVO RESPONSE OF PLASMODIUM FALCIPARUM TO CHLOROQUINE Table 1. Plasmodium falciparum prevalence and parasite density in women in Siaya District, Kenya Geometric mean Pregnancy No. of No. positive for parasite density number women P. falciparum (per mm3 of blood) 0 98 49 (50)" 462 1 66 45 (68) 1538 2-3 77 26 (34) 336 4-5 52 15 (29) 299 ,>6 49 16 (33) 330 Total pregnant 244 102 (42) 644 ' Figures in parentheses are percentages. were parasitaemic; this was lower than the prevalence in primigravid women (P=0.02) and higher than the prevalence in multigravid women (P=0.003). The geometric mean parasite density (GMPD) was sig- nificantly higher in primigravidae than in nulligravid or multigravid women (P<0.0001). Pregnant women who reported the use of an anti- malarial drug in the previous 7 days were less likely to be parasitaemic (20% of 40 women) than those who denied recent drug use (46% of 204 women; P=0.003). Among pregnant women with no recent history of antimalarial drug use, the prevalence of parasitaemia was similar in women with a history of fever in the previous seven days (46% of 117 women) and with no recent history of fever (45% of 88 women). In-vivo parasite resistance studies. Seventy-two gravid and 42 nulligravid women were enrolled and followed for at least 7 days. The distribution by CQ5 and CQ25 treatment group and parasite clearance is shown in Tables 2 and 3. In the CQ25 treatment group, failure to clear parasites by day 7 was more common in primi- gravidae (7 of 20) than nulligravidae (0 of 17, P=0.008) or multigravidae (1 of 10, P=0. 15). For both gravid and nulligravid women, the clearance of Table 2. Characteristics and follow-up of gravid and nulligravid women with Plasmodium falciparum malaria treated with chloroquine phosphate, 25 mg/kg body weight Mean Initial geometric No. of women parasitaemic on follow-up visit Pregnancy No. of pregnancy Mean age mean parasite status women number (years) density/mm3 Day 7 Day 14 Day 21 Day 28 Nulligravid 17 0 14.3 2537 0 (0)° 5 (29) 12 (71) 13 (76) Primigravid 20 1 17.9 4828 7 (35) 14 (70) 16/19 (84) 16/19 (84) Multigravid 10 3.4 24.6 3311 1 (10) 2/8 (25) 2/8 (25) 6/8 (75) Total gravid 30 1.8 19.9 4343 8 (27) 16/28 (57) 18/27 (67) 22/27 (81) ' Figures in parentheses are percentages. Table 3. Characteristics and follow-up of gravid and nulligravid women with Plasmodium falciparum malaria treated with chloroquine phosphate, 5 mg/kg body weight, weekly No. of women on day-28 follow-up with: Mean Initial geometric Pregnancy No. of pregnancy Mean age mean parasite Low High status women number (years) density/mm3 Aparasitaemia parasitaemia° parasitaemiaa Nulligravid 25 0 14.5 232 6 (24)b 14 (56) 5 (20) Primigravid 7 1 19.7 849 3 (43) 1 (14) 3 (43) Multigravid 26 4.2 24.9 158 13 (50) 10 (38) 3 (12) Total gravid 33 3.5 23.8 227 16 (48) 11 (33) 6 (18) ' Low-density parasitaemia was defined as 4 2-fold increase in day-O parasite density and were asymptomatic; these women were continued on the CQ 5 mg/kg regimen. High parasitaemia was defined as >2-fold increase in parasitaemia from the day-O level to a level of >2000/mm3; these women were treated with an alternative drug. b Figures in parentheses are percentages. 887 R. W. STEKETEE ET AL. 700 z 0 o c.UJ es A. LU z a 0 0 I 0 0 0Lu 0 600 500 400 300 200 1004 14 o = C25 treatment * =C5 treatment = NULLIGRAVID WOMEN *----o = GRAVID WOMEN 24 0 2 7 14 21 28 DRUG GIVEN: 4 DAY OF FOLLOW-UP C25 " C5 _. ,. . Fig. 1. Mean blood chloroquine levels at follow-up visits in gravid and nulligravid women treated with chloroquine 25 mg/kg body weight or 5 mg/kg body weight weekly. parasites was unrelated to initial parasite density. Among all women who were aparasitaemic on day 7, primigravidae were more likely to have reappearance of parasites on day 14 and 21 than nulligravidae or multigravidae. In the CQ5 treatment group (Table 3), gravid women were more likely to clear their parasitaemia by day 28 than nulligravid women. Primigravidae (3 of 7) were more likely to show an increased parasite density, however, and to require a treatment dose than nulligravidae (5 of 25, P=0.22) and multi- gravidae (3 of 26, P=0.09). Whole-blood CQ levels were similar within treat- ment groups on a given follow-up day for nulligravid, primigravid and multigravid women (Fig. 1) and did not differ significantly between women exhibiting and not exhibiting resistance on day 7. Although vomiting was reported in the interval between CQ administration and the follow-up visit by 11 (15%) of 72 pregnant women, vomiting within 4 hours of drug administration was reported by only 2 women, and their whole-blood CQ levels indicated adequate absorption of the drug. Pruritus associated with CQ administration was reported at least once during the study by 56 (77%) gravid women but by only 8 (19%) nulligravid women (P< 10-8). The prevalence of reported pruritus among pregnant women did not vary with drug dosage: itching was reported by 22 (73%) out of 30 women on CQ25 and 34 (79%) out of 43 women on CQ5. An antihistamine, chlorpheniramine, was used to counteract the pruritus, but its efficacy was not systematically evaluated. We followed 23 women who were treated with amodiaquine (AQ25) after failing to clear their parasitaemia on chloroquine regimens; 18 (78%) of these women were aparasitaemic seven days after starting treatment. All five failures were among the 16 primigravid women. DISCUSSION As chemotherapeutic interventions against malaria during pregnancy have increased in complexity owing to the spread of chloroquine-resistant P. falci- parum across sub-Saharan Africa, the possibility of focusing the strategy on a subgroup of pregnant women is of immediate concern. Although other investigators have shown that young women of low parity appear to be at increased risk for both malaria infection (1, 11-14) and delivery of low-birth-weight infants (1, 11-15), the relative importance of focusing on young women as opposed to women of low parity has not been clear. In our study area of Western Kenya, with a 68% point prevalence of malaria infection in primigravidae, we assume that virtually all women are exposed to malaria at some time during each pregnancy. The observed higher prevalence of parasitaemia and higher parasite den- sities in primigravidae compared with younger nulligravidae and older multigravidae suggests that in her first malaria-exposed pregnancy a woman is more susceptible to malaria infection, regardless of her age. Exposure to malaria parasites during a woman's first pregnancy may convey some degree of pro- tection against infection in subsequent pregnancies; this is supported by studies of Plasmodium berghei infections in pregnant mice (16). The immunological memory may be associated with cellular response(s) at the placental-uterine interface (11) leading to an anamnestic response in subsequent pregnancies. Although humoral immune mechanisms may differ between the pregnant and non-pregnant state, it is unlikely that a pregnancy-related alteration in the quantity or quality of malarial antibody pro- duction would explain the difference we observed in responses between the primigravidae and the multigravidae (17). In areas with lower rates of P. falciparum transmission, some women may not be exposed to malaria during their first pregnancy and would be susceptible in the second pregnancy. This 888 INV VIVO RESPONSE OF PL4SMODIUM FALCIPARUM TO CHLOROQUINE might explain the observed gradient between the first, second, and all subsequent pregnancies for parasite prevalence and mean parasite densities in women living in other African countries with lower rates of transmission (1, 18). The proportion of primigravidae who failed to clear their parasitaemia following treatment doses of chloroquine (CQ25) was significantly higher than that of nulligravid and multigravid women. This poor response is similar to that observed in 5-9-year-old children living in this area of Kenya (A. D. Brandling-Bennett, unpublished findings, 1985). Similarly, primigravid women with low initial parasite densities ( < 1500 parasites/mm3 of blood), after four weeks of CQ5, were more likely to show increasing parasite densities and require therapeutic doses of the drug. Although the parasite density was higher in primigravidae than nulligravidae and multi- gravidae, the difference in clearance rates could not be attributed to higher initial parasite density. Although clearance of P. falciparum in children in areas of CQ resistance appears to be age related (D. Heymann, personal communication from Malawi, 1986), age could not explain our observed differences in clearance rates in women, since both younger nulli- gravidae and older multigravidae showed higher rates of clearance than the primigravidae. In areas with CQ-resistant P. falciparum, the physiological mech- anisms which render the primigravida more suscept- ible to developing a patent malaria infection may also contribute to a decreased ability to clear the infection following treatment. The purpose of regular antimalarial chemopro- phylaxis during pregnancy is to clear or prevent placental parasite infections, since it is the placental infection that is associated with delivery of low-birth- weight infants (1). Peripheral parasitaemia does not necessarily mean that there is a placental parasite infection (I), and we cannot be sure that clearance of peripheral parasites will be accompanied by the clearance of placental infection. However, it is certain that persistence of parasites in the peripheral blood bodes ill for placental parasite clearance. For areas in sub-Saharan Africa, where there is a great need to introduce effective malaria interven- tions, we are pessimistic about the efficacy of chemoprophylaxis in pregnant women using weekly chloroquine 5 mg/kg. We believe that malaria control strategies could be focused on primigravid women or those in their first malaria-exposed preg- nancies, but this is the very group that is least likely to respond to a chemotherapeutic intervention. As chloroquine-resistant parasites spread, a better under- standing of the problem is needed leading to further chemotherapeutic options for pregnant women. In areas with moderate chloroquine resistance, it may be worthwhile to assess the effect of intermittent full- dose treatment with chloroquine rather than continual prophylaxis. Likewise, further understanding of the immunological mechanisms responsible for the in- creased susceptibility during the first malaria- exposed pregnancy is needed; with this knowledge, malaria prophylaxis could be provided alone or in conjunction with tetanus toxoid in a prenatal clinic. This would offer an attractive addition or alternative to chemoprophylaxis. ACKNOWLEDGEMENTS James Sande Odera, Pius Owuer, Daniel Karyuki, James Otieno, and Julius Ochieng from the Kenya Medical Research Institute and the Division of Vector-Borne Diseases, Ministry of Health, Nairobi, Kenya, performed technical field work and Shary Williams and Monica Overman from the Control Technology Branch, Division of Parasitic Diseases, Centers for Disease Control, Atlanta, GA, performed the laboratory work for this investigation. The authors would like to thank the Director, Kenya Medical Research Institute, for permission to publish this paper. This investigation was supported by USAID (PASA BST-0453-T-HC-2086-02). RESUME REPONSE 1N VIVO DE PLASMODIUM FALCIPARUM A LA CHLOROQUINE CHEZ DES FEMMES (ENCEINTES OU NON) DU DISTRICT DE SIAYA, KENYA On a recommande une chimioprophylaxie a la chloroquine du paludisme pendant la grossesse. Au Kenya occidental, (CQ) a des doses suppressives pour proteger les femmes en- dans une region d'endemie du paludisme oix Plasmodium ceintes des regions impalud6es contre les effets ind6sirables falciparum chloroquinoresistant est installe, nous avons 889 890 R. W. STEKETEE ET AL. determine la prevalence et la densitd des infestations a falciparum chez les femmes enceintes ou non, et nous avons compare la reponse in vivo des plasmodies a la chloroquine administree a deux posologies diff6rentes: 25 mg/kg de poids corporel (CQ25) sur une periode de 3 jours (pour les parasit6mies a forte densite parasitaire), et 5 mg/kg de poids corporel (CQ5) par semaine pendant 4 semaines (pour les parasitemies a faible densit6). P. falciparum a 6te retrouve dans les frottis sanguins de 102 (42%) femmes enceintes sur les 244 examinees. Une plus grande pro- portion de primigestes (68%) que de nulligestes (50%, P=0,02) ou de multigestes (33%, P< 10-6) etaient para- sit6es. Les primigestes presentaient egalement une moyenne geometrique de la densitd parasitaire sensiblement plus dlevee que les nulligestes ou les multigestes. Chez les 30 femmes enceintes et 17 femmes non en- ceintes du groupe traite par la CQ25, la non-6limination des parasites au 7C jour s'est rencontree plus couramment chez les primigestes que chez les nulligestes (P=0,008) ou les multigestes (P=0, 15). Chez les 33 femmes enceintes et 25 femmes non enceintes du groupe traitd par la CQ5, les primigestes (43 %) etaient plus susceptibles de presenter une remontee de la densite parasitaire que les nulligestes (20%) ou les multigestes (12%). Vingt-trois femmes ont dtd traitees par I'amodiaquine, a raison de 25 mg/kg de poids corporel, apres 1'echec de l'elimination de la parasitemie; 18 d'entre elles (78%) presentaient une parasitemie nulle 7 jours apr6s le traitement. Les cinq echecs ont tous ete observ6s chez des primigestes (au nombre de 16). La forte prevalence de la parasitemie chez les primigestes laisse a penser que, dans cette region du Kenya, toutes les femmes sont exposees au paludisme au cours de leur premiere grossesse. Nos resultats portent a croire que les femmes presentent, au cours de leur premi6re grossesse en region d'endemie, un risque maximal d'infestation, avec des densites parasitaires elevees, et qu'elles sont moins aptes a repondre au traitement a la chloroquine dans les regions de chloroquinoresistance. Les strategies de lutte antipaludique pourraient etre axees sur ces femmes mais nous sommes pessimistes quant a l'efficacite du traitement par la chloroquine a raison de 5 mg/kg de poids corporel par semaine, dans les regions de chloroquinoresistance. REFERENCES 1. McGREGOR, I. A. Epidemiology, malaria and preg- nancy. American journal of tropical medicine and hygiene, 33: 517-525 (1984). 2. WHO Technical Report Series No. 735, 1986 (WHO Expert Committee on Malaria: eighteenth report). 3. BRADLEY-MOORE, A. M. ET AL. Malaria chemo- prophylaxis with chloroquine in young Nigerian children. Annals of tropical medicine and para- sitology, 79: 549-562 (1985). 4. MACCORMACK, C. P. & LWIHULA, G. Failure to participate in a malaria chemosuppression programme: North Mara, Tanzania. Journal of tropical medicine and hygiene, 86: 99-107 (1983). 5. SPENCER, H. C. ET AL. Malaria chemoprophylaxis to pregnant women provided by community health workers in Saradidi, Kenya. H. Effects on para- sitaemia and haemoglobin levels. Annals of tropical medicine andparasitology, 81 (suppl. 1): 83-89 (1987). 6. NGUYEN-DINH, P. Etudes sur la chimiordsistance de Plasmodiumfalciparum en Afrique: donndes actuelles. Annales de la Societe# Belge de Medecine Tropicale, 65(suppl. 2): 105-113 (1985). 7. SPENCER, H. C. Drug-resistant malaria-changing patterns mean difficult decisions. Transactions of the Royal Society of Tropical Medicine and Hygiene, 79: 748-758 (1985). 8. HASKINS, W. T. A simple quantitative test for chloro- quine in urine. American journal of tropical medicine and hygiene, 7: 279-287 (1985). 9. PATCHEN, L. C. ET AL. Analysis of filter-paper- absorbed, finger-stick blood samples for chloroquine and its major metabolites using high-performance liquid chromatography with fluorescence detection. Journal of chromatography, 278: 81-89 (1983). 10. BRUCE-CHWATT, L. J. ET AL. Chemotherapy of malaria, 2nd edition. Geneva, World Health Organiz- ation, 1981, pp. 105-109. 11. McGREGOR, I. A. ET AL. Malaria infection of the placenta in the Gambia, West Africa; its incidence and relationship to stillbirth, birthweight and placental weight. Transactions of the Royal Society of Tropical Medicine and Hygiene, 77: 232-244 (1983). 12. JELLIFFE, E. F. P. Low birth-weight and malarial infection of the placenta. Bulletin of the World Health Organization, 33: 69-78 (1968). 13. KORrMANN, H. F. C. M. Malaria and pregnancy (thesis). Utrecht, Drukkerij Elinkwijk, 1972, pp. 1-98. 14. BRAY, R. S. & ANDERSON, M. J. Falciparum malaria and pregnancy. Transactions of the Royal Society of Tropical Medicine and Hygiene, 73: 427-431 (1979). 15. BANTJE, H. A multiple regression analysis of variables influencing birth-weight. Tropical and geographic medicine, 38: 123-130 (1986). 16. VAN ZON, A. A. J. C. & ELING, W. M. C. Pregnancy-associated recrudescence in murine malaria (Plasmodium berghei). Zeitschrift fur Tropenmedizin und Parasitologie, 31: 402-408 (1980). 17. LEDERMAN, M. M. Cell-mediated immunity and pregnancy. Chest, 86(suppl. 3): 6S-9S (1984). 18. STEKETEE, R. W. ET AL. Malaria infection in pregnant women in Zaire: the effects and the potential for inter- vention. Annals of tropical medicine and parasitology (in press).