Iron and malaria

ParasitologyToday,vol, 5, no. 3, 1989

77

Iron and Malaria S.J.Oppenheimer Iron deficiency is common in the developing world; consequently, programmes of presumptive therapy and mass supplementation have been introduced in several countries. In this article Stephen Oppenheimer suggests caution, as recent evidence suggests that these practices may actually increase the likelihood of the subject developing patent malaria in endemic areas. This may be especially significant in infants, who are less likely to be immune to malaria, and in pregnant women, who are often routinely given iron supplements and in whom malaria may damage the foetus. Iron is indispensable as a micronutrient to almost all living organisms. Iron deficiency may have arisen relatively recently in human evolutionary history, when hunter-gatherer societies turned to agriculture, but it is now the commonest mi~:ronutrient deficiency in the world. The lowest non-haem iron stores in human populations have been recorded in India, coastal Papua New Guinea, and Burma I. Iron deficiency is not, however, confined to less developed countries; it is still the most important micronutrient deficiency in the USA, particularly in infancy2. The main effect of iron deficiency is a hypochromic microcytic anaemia. Many measurable non-haematological effects also occur, for instance in .epithelial tissues, brain and in cell-mediated immunity ~. These known deleterious and preventable effects have stimulated recommendations 3 for the introduction of programmes designed to prevent iron deficiency by oral iron supplementation, in both developing and developed countries. Iron supplementation to children and pregnant mothers is a world-wide practice, mainly bythe oral route, including milk fortification, but also parenterally. Certain developing countries, with a high prevalence of iron deficiency, have fortified other foodstuffs with iron. Common salt has been thus fortified in India4, and in Thailand, fortification offish sauce with iron and iodine has been proposed s. Iron S u p p l e m e n t a t i o n m~d M a l a r i a

Scattered clinical reports over the past two decades have suggested that iron therapy may increase susceptibility to infection with malaria in certain situations: Byles and D'Sa 6, in an uncontrolled study, reported II cases of clinical malaria among 917 pregnant women immediately following iron therapy, and Masawe et OI, 7 reported malaria in eight iron-deficient patients after starting therapy. Unfortunately, this anecdotal study gave no details of which patients were on oral or parenteral therapy. ~) 1989,ElsevierSciencePublishersLtd, (UK) 0165~S147/89/$0200

In an often quoted study, a family team, the Murrays 8, conducted a prospective single-blind, placebo-controlled randomized trial of 30 days' oral iron supplementation to 137 adult Somali nomads with iron deficiency anaemia. Iron treatment increased haemoglobin levels and transferrin saturation during the study. Although no malaria was noted at the start of the study in either group, 13 clinical cases of malaria had occurred in the group (n = 71) receiving iron and only one in the control group (n = 66) by the end of the trial. Unfortunately no further observations were made after the end of therapy. These adverse reports contrast with the apparently beneficial results of three longitudinal studies of dietary iron supplementation in infancy (reviewed in Ref. I). These studies, in the UK, New Zealand and the USA, claimed to show reductions in respiratory infections in supplemented groups I. Several features distinguish these results. First, in the reports of increased or exacerbated infections, observations were confined to the period shortly after starting therapy, when circulating iron levels were likely to have been high8. Second, malaria, which was the dominant infection in the adverse reports, is not endemic in the countries where the results of iron supplementation appeared to be beneficial. Non-physiological, acute effects of iron therapy need to be separated from

the long-term effects of improving iron status. Carefully controlled longitudinal studies are thus necessary in both malarious and malaria-free areas to determine the benefits and possible hazards of iron supplementation. A Prospective S t u d y

More recently, a prospective, randomized, double-blind, placebocontrolled trial of iron supplementation in infancy was carried out in Papua New Guinea9. Iron deficiency was highly prevalent among infants in the study population. Malarial endemicity in the area varied from mesa- to haloendemic. To avoid the known risks of iron therapy in the neonatal period ~°, a single dose of iron dextran (I 50 mg elemental iron) was administered at 2 months of age to the treatment group (n=236); control infants ( n = 2 5 0 ) received an injection of sterile pyrogenfree saline. Infants were fully reexamined one week after the injection, and at 6 and 12 months of age. At follow-up, blood was taken for a range of haematological indices including thick and thin film examination for malaria. In addition, all admissions to hospital were carefully documented. After breaking the code, several results became apparent: no significant difference in malaria rates was seen at the one-week follow-up visit, but at both the 6- and 12-month visits, malaria slide positivity and spleen rates were higher in the iron treatment group (Table I). The increase in malarial slide positivity rate associated with iron was 64% for each visit, while spleen rates were 30-40% higher9. A quarter of the iron treatment

Table I. Malarial parasitaemia rates and splenomegaly rates in infants 4 and I 0 months after starting administration of iron-dextran and placebo. Relative risk b Percentage positive (and 95 % (and number followed up) confidence limits) Investigation Iron-dextran Placebo At age 6 months parasitaemia

18.5 (200)

I 1.3 (212)

1.78 (I .02-3. I O)

Splenornegalya

26.9 (197)

19.6 (219)

1.51 (0.95-2.33)

At age 12 months parasitaemia

32.8 (177)

20.0 (190)

1.95 (I.21-3.13)

Splenomegaly~

47.6 (170)

36.9 (I 79)

1.56 (I .02-2.39)

aHackett'sgrade I or greater. blron dextran versus placebo.

78

Parasitology Today, vol. 5, no. 3, 1989

group was admitted to hospital with evidence of malaria in their first year of life, compared with only 17% of the placebo group (P < 0.05) ~. However, no effect of iron on parasite densities was detectable. An additional finding was that infants with higher birth haemoglobin levels (and thus a higher total body iron content) were significantly more likely to have malaria at follow-up and were also more likely to be admitted to hospital with malaria. The latter effect was synergistic with the iron treatment effect ~~. Thus infants who both received iron and had a higher birth haemoglobin had a 40% risk of admission with malaria in the first year of life (Fig. I ). However, death rates were similar in the two groups and none of the 12 deaths in the study could be directly attributed to malaria. These findings showed a deleterious effect of iron supplementation lasting up to ten months after administration. This differs from previous studies where only short-term adverse effects were reported and longer-term effects on infection rates were, if anything, beneficial ~. Several factors limit any general interpretation of the results of this study to other malarious areas. First, a parenteral iron preparation was used and second, the study group consisted of infants aged 2-10 months who were probably less immune than older inhabitants of malaria-endemic areas. In contrast, in a preliminary report of a more recent study in the same area, Harvey et a1.12 failed to show any adverse effect of oral iron supplementation to pre-pubescent schoolchildren, particularly in relation to malaria indices. These authors also made the suggestion that immunity may have affected the interaction between iron and malaria ~2.

Iron, Malaria and Pregnancy

Another group that appears to have a peculiarly lowered immunity to malaria in endemic areas is women in their first pregnancies ~3. In this context it was noted, in an observational study, that treatment of anaemia with iron infusion during pregnancy was associated in women in their first pregnancies with higher rates of maternal perinatal malaria; in contrast, this effect was not seen in women who had had several pregnancies. Oral iron is a standard supplement in pregnancy, and in many parts of the tropics parenteral iron is often administered for practical reasons as presumptive treatment of anaemia during pregnancy. Since anaemia during preg-

60+

~ 50-

"Q 40E O

::::::.,i

._~

E 30E~Ei::I

o~" 2 0 -

i!!!ii::l

I~iiiiiiil 0

liiiiiiii IEEE~I

<13.7 ]

13.7 >15.7 J --15.7 Haemoglobin level at birth (g/dl)

Fig. I. Iron dextran trial in Papua New Guinea: probability of any admission with evidence of malaria by one year of age according to treatment group(irondextran [] or placebo I-1) and by birth haemoglobin level.

nancy in the tropics is commonly due to malaria ~3, treatment with iron may carry a definite risk. This last example highlights a recurring problem for populations in malaria-endemic areas. Those groups of individuals with a high risk of anaemia secondary to malaria (infants, toddlers and mothers) are also selectively more likely to receive iron either as a supplement or as presumptive treatment for anaemia.

Possible Mechanisms of the Interaction

How iron deficiency and iron administration interact with malaria is still a matter of conjecture. One hypothesis relates to red cell age. Plasmodium vivax can grow only in reticuIocytes f4. Pasvol eta/. ~s showed, using red cells of different ages separated by centrifugation, that P. falciparum preferentially invaded younger cells; the multiplication of parasites, and hence the rate of destruction of red cells, could be to a certain extent dependent on a continuing supply of younger cells and reticulocytes, which would be stimulated by administration of iron. A second hypothesis relates to iron availability. Some observers have shown that increased saturation of transferrin may help bacteria and other microorganisms to grow ~6. Red cells normally lose the capacity to acquire transferrinbound iron at the reticulocyte stage ~7. The adult red cell, infected with malaria,

would therefore be likely to be dependent on an intracellular iron pool unless specific transferrin receptors were elaborated on the cell surface by the parasite. However, a recent report suggests that cells infected with P. falciparum do indeed have a transferrin receptor ~8. There is also evidence from experiments with animals that desferrioxamine modifies malaria infections ~9. On the other hand, Peto et ol. 2° suggested that P. falciparum may not necessarily be dependent on serum transferrin-bound iron, but that desferrioxamine may act on a labile iron pool available for immediate use by the parasites. A third hypothesis, that microcytosis secondary to iron deficiency adversely affects parasite multiplication, has yet to be studied in depth. Nurse 21 suggested that a shortage of intracellular nutrients essential to the parasites might occur in small cells with a poor haemoglobin content. There might also be increased susceptibility to oxidant damage in a smaller intracellular environment.

Recommendations

Given the present incomplete knowledge of the interactions of iron and malaria, what recommendations can be given to health planners and clinicians? Iron dextran prophylaxis to infants in malarious areas should be definitely contraindicated. Oral iron supplementation in the same age group may also be contraindicated. Oral iron supplementation to older children may not be associated with adverse effects, though further work is needed to confirm this. The practice of treating anaemia in malarious areas with parenteral iron should be reviewed urgently. It should be contraindicated in infancy; it may be contraindicated during pregnancy, especially for women in their first pregnancies. Further controlled clinical studies are clearly needed before making more definite recommendations. Iron treatment for anaemia in a malarious area should be covered or preceded by effective antimalarial therapy. Iron therapy should be oral and, where possible, the decision to use iron should be based on laboratory evaluation of the cause of anaemia. Policies for standard presumptive treatment for anaemia at the primary health care level in developing countries need to be reviewed with particular reference to age group and identifiable causes, preferably with the aid of results from prospective controlled clinical studies.

ParasitologyToday,vol. 5, no. 3, 1989 References I Oppenheimer, S.J. and Hendrickse, R.G. (1983) Nutr. Abstr. Rev. 53,585-598 2 Lane, M.J. and Johnson, C.L. (1981) in Iron Nutrition Revisited - Infancy, Childhood, Adolescence (Report of the Eighty-second Ross Conference on Pediatric Research) (Oski, F.A. and Pearson, H.A., eds), pp 31-46, Ross Laboratories 3 Committee on Nutrition (1978) Pediatncs 62, 246-250 4 Datta, R.N. et aL (I 982) Am. J. Clin. Nutr. 35, 1442-1451 5 Suwanik, R. et al. (I 980)]. Mle:l. Assoc. Thailand 63,611-615 6 Byles, AB. and D'Sa, A. (1970)Br. Mled]. 3, 625-627

79 7 Masawe, A.E.J., Muindi, J.M. and Swai, G.B.R. (1974) Lancet 2, 314-317 8 Murray, M.J. et al. (1978) Br. Mled. ]. 2, 1113~1115 9 Oppenheimer, S.J. et oL (1986) Trans. R. Sac. Trap. Mted. Hyg. 80,603-612 I 0 Barry, D.M.J. and Reeve, A.W. (1977) Pediatrics 60, 908-912 II Oppenheimer, S.J. et aL (1986) Trans. R. Sac. Trap. MeG( Hyg. 80, 596-602 12 Harvey, P. et aL (1987) Am. Sac. Exp. Biol. 46, 4 13 Brabin, B.J.(I 983)Bu11. WHO 61, 1005-1006 14 Garnham, P.C.C. (1966)/Malarial Parasites and Other Haemospondia Blackwell Scientific Publications 15 Pasvol, G., Weatherall, D.J. and Wilson, R.J.M. (1980) Br.]. Haematol. 45, 285-295

16 Weinberg, E.D. (1974) Science 184, 952-956 17 Enns, C.A. et al. (I 981 ) Proc. Natl Acad. Sci. USA 78, 4222-4225 18 Rodriguez, M.H. and Jungery, M. (1986) Nature 324, 388--39 I 19 Raventos-Suarez, C., Pollack, S. and Nagel, FLL. ( 1982)Am.J. Trap./Vled. Hyg. 3 I, 919-922 20 Peto, T.E.A. and Thompson, J.L. (1986)Br. ]. Haematol. 63,273-280 21 Nurse, G.T. (1979) Lancet 2, 939-941

Stephen Oppenheimer is with the Department of Paediatrics, UniversitySainsMalaysia, 15990 Kota Bharu, Kelantan,Malaysia.

The Application of Nuclear Magnetic Resonance Spectroscopy to Parasite Metabolism W.J. O'Sullivan, M. R. Edwards and R, S. Norton Nuclear magnetic resonance (NMR) has become a valuable tool for the study of metabolism in a wide variety of biological systems. Its inherent advantages are that it is non-destructive and non-invasive. Observations can be carried out not only on extracts and media but also on whoh~. cells and whole tissues under varying conditions and over varying times. The information gained gives considerable insight into cellular metabolism. There has been, to date, relatively little literature on the application of NMR to the biochemistry of parasites, presumably reflecting the paucity of interfaces between parasitologists and NMR practitioners as well as the inherent di~culties in obtaining sufficient parasite material ~)r NMR experiments. These di~culties are being overcome and William O'Sullivan, Michael Edwards and Raymond Norton believe that NMR has a great deal to offer those interested in parasite metabolism. In particular, it has the capacity to turn up the unexpected, an important factor as so many parasites appear to have developed their own var.iations on orthodox metabolic pathways. Many reviews have dealt with the application of NMR to biological systems in general. A recent, brief article by Shulman highlights applications to whole tissues J and a review by CampbellBurk and Shulman2 describes experiments on yeast which are analogous to what might be achieved with, parasites. A more general review has been given by Williams and Gadian 3. The NMR experiment is dependent on the fact that, in a magnetic field, nuclei with unpaired spins can exist in more than one energy state and transitions between the energy states are accompanied by the absorption of energy. The spectrum obtained from modern instruments is such that it provides a characteristic profile or 'finger print' of a compound, dependent upon its chemical nature. A brief description of the NMR experiment and a spectrum of ATP is given in Box I. In common with other biological systems, the three most important nuclei for the study of parasite metabolism are 1989,ElsevierSciencePublishersLtd,(UK)0 r65-6f47/89/$02.00

JH, ~3C and 3 t p. Examples of the use of these three nuclei are given below, and an assessment of their relative advantages and disadvantages is given in Box 2.

extracts, and third, observations on whole living parasites. As with other procedures, the timing of observations is important as metabolites may be transient. The emphasis in this article is on the use of NMR to follow metabolism as a complement to classical approaches. We are not concerned here with situations where NMR has been used as an organic chemistry tool to aid the determination of the structure of unusual compounds isolated from parasites (for example the determination of the structure of the glycosyl-phosphatidylinositol moiety that anchors Trypanosoma brucei surface glycoproteins to the membrane ~l). M o n i t o r i n g of M e d i a

NMRfor

Parasitologists

NMR offers the opportunity of exploring metabolism without the necessity of any assumptions, effectively 'opening up a window' on the cell's metabolic processes. Examples of new insights into parasite metabolism that have been gained from NMR experiments are, the unexpected finding by Chapman et al. to using ~3C NMR, that glycerol was a major metabolite of Trichomonas vaginalis, and our recent observations that alanine is a major metabolic product of Giardia lamblia (see below). To illustrate the type of information that may be obtained, we have selected three examples, first, the monitoring of changes in the medium in which parasites are grown, second, the use of parasite

The concept is simple. A spectrum is taken of the medium in which a parasite is to be grown. During various stages of parasite growth, samples are taken, the cells spun down and spectra run on the supernatant media. Significant changes can usually be attributed to specific metabolites, and generally, only peaks from major metabolites will be clearly distinguishable in very complex mixtures, depending on the degree of overlap of peaks in the area of interest. Considerable discrimination can be achieved by difference spectroscopy, subtracting the 'before' spectrum from the 'after'. Such an experiment provides direct information on major fuel metabolism; what is used and what is produced. Further, the approach is open to