- Email: [email protected]
Why does Quinine still work after 350 years of use?
Comment
W h y Does Quinine Still Work After 350 Years of Use? S.R. Meshnick P!asmodium strains have become resistant to almost every antimalarial dru~ introduced in the past 60 years. In contrast, even though quinine has been used to treat malaria for rnore than 3S0 years, moderate resistance has only recently developed in limited geographic areas i. What is so special about quinil;e~ There are three possible expbnations. FirsL quinine's intraparasitic target might be so specific that mutations that confer resistance carl occur only at an exceptionally slow rate. Second, modem strains of malaria parasite might in fact already be resistant to quinine when compared to the strains of malaria that existed 300 years ago. In other words, quinine resistance m~y have developed slowly over centuries of use, and we simply never noticed that the required doses of quinine had increased. Third, perhaps quinine had not been used Frequently enough to exert evolutionary selective pressure until recently, Understanding quinine's longevity could have important implications. For example, if quinine's target is so speciz! then it would be very useful to identify it. On the other hand, if modem malaria parasites already are quinine resistant, then we should try to understand how this resistance has affected parasite metabolism. Quinine has its origins in Peru in the early 17th century 24, where Indians were observed chewing on the bark of cinchona trees, and the bark was found to be an effective 'febrifuge' for 'ague' or 'intermittent Fever'. It is unlikely that the antimalarial action of the bark had been known to indigenous medicine, because malaria had only recently been imported inLo %uth America. Since too!aria was rampant in Europe, and would remain so until the early 20th century, there was soon a great demand for 'bark' or 'Jesuit's powder'. In 1820, a method of purifying quinine from cinchona bark. was developed 0y Pelletier and Caventou in Paris. Quinine became especially important to English settlers in India, who invented the practice of sweetening their prophylactic 'tonic' with gin 8. (Tonic water, toda) :, contains only 15 mg of quinine per bottle 9, not nearly enough to have antimala~4al activity. It was probably just a good excuse For ~in.) Parasitology lc~;ay,vol. 13, no. 3, 1997
Quinine is structura!ly similar to the other ouinoline an~i:nalarlals, especially to mefloquine, wh' : . is also a quinoline methanol. The mechanism by which quinine kills malar:a parasites is unknuwn, so the drug target could possibly be one t,,at is especially important and immutable. As mab ri~ parasites have become resistant to all of the ,xner quinolines, however, this seems unlikely. Pldsmodium fdlciporum develops resistance to chloroquine and mefioquine by becoming able to pump the drag out from the cell (reviewed in Ref. I 0). Both mefloquine dnd chloroquine have side chains that contain linear strings of carbon and nitrogen atoms. Quinine, on the other hand, has a ,much more complex side chain containing carbons and one nitrogen in a rigid two-dnl structure. Perilaps the malada proteins which pump out mefloquine and chloroqui,-,~ are unable to handle qutn~ne because of its side chain structure. However, microorganisms have become resi.~l"an~ to every antimicrobial agent ever put into widespread use. Malaria is no exception. It is highly unlikely that biochemistry alone can explain the relative longevity of quinine as an e,~ective antimala,'ial. Have quinine doses increased over the centuries? There are several reasons why we will never know the doses of quinine that were effective before the 20th century. First, prior to 1820, ague was treated with suspensions of finely powdered cinchona bar~. in wine. The amounts of absorbed (bioavailable) quinine mu~ have varied depending on how this suspension was prepared. Also, the amount of quinine contained it, c;nchona bark from different sources varied greatly (0.3-8%), and onchona bark contains vanable amounts of other antimalarial alkaloids such as cinchonine 3. Thus. it is imposs~b!e to de :rmine accu~a~c!y now much antimalar al material was actually given to pat~ent.i Second, standard practice: ~ideli~'s for drug use are invent!ons of the late 23~.h centuw. Prior to that, physicians ,vou!d use medicines at whatever doses they felt comfortable with (malpractice law was also ~ill in its infancy). And if those didn't weft<, they might repeat the treatment, double the dose,
or combine medications according to their ir.÷,j~iVe f~lina~S Finally, it wasn't until the end of the 19th centu~7, that malaria parasites were identified under the microscope. Thus, many illnesse~ other than m a M a wen ~. mistakenly treated ';~'ith quinine; marly patients who didn'~ respond to quinine may really have had another infectious disease and not malaria. Conversely, patients who seemed to resp,:;~,! to low doses of quinine may have had illnesses (such as viral infections) that spontaneously resolved. So we will never know how sensitive malaria parasites were to quinine prior to the 20th century. What were the doses of quinine used once malarie could be deftnitively diagnosed? According to a textbook or, m a m a written in 1900, quinine was to be given intramuscuMy or subcutaneously with a 2 g loading dose, folIo'wed by I g every 6-8 h (Re[ I I ). This is very similar to current recommendations, suggesting that there has been no cha:~e i. ~usc~ptibil~)'. Wns quinine used in enough quantity to engender resistance? Plasmadiun ~faloparum is the most drug-resis~nt species of malaria and is also the species against which quinine is most. ofi.en used today. Historically, however, the majority of malaria in Europe and North America was due to P. vivax or P. malariae. Even in colonial India, a large proportion of the malaria was due to P. vivax. Therefore. prior to the 20th century, most quinine waS probably used to treat viva× not fa!ciparam malaria. Furthermore, quinine waS probably used by a relative!y small group of people when compared to the broad distribution and use of antimalarials today 3. There are several pharmacological factors which may also have contributed to the maintenance of quinine's efficacy. Fir~ quinine quite often causes annoying side effects such as nausea and vomiting, ringing in the ears, headn2 loss. dizziness, headache, and/or disturbed vision. So people are unlikely to take the drug unless absol,~¢e!ynecessary,.Second, quinine has a relatively short, half-life (I0 h) (Re[ 12), so it will not linger at subtherapeutic concentrations in the blood..ream for very long. Thus, P. faldl~arum parasites, all in all, have probably Lad ~ia~ively
Coprm~l©1997 EJs~e,'~c,erceLTa Al'7~"tsrese".~; 0 o~-47bfl'e?.'$17 'qO ~Ou~9-'iTSS(97)OlO}3-X
~9
Comment little exposure to quinine and thus little evolutionary selective pressure from it. Quinine has remained an effective antimalarial for 350 yeal%. Prior to recent times, it was probably never used at a high enough frequency against P. falciparum to induce resistance. This could serve as a lesson to us as we look for ways to preserve the efficacy of our current antimalarials and other drugs.
References 9 Carroll, bl. 0995) Quinipe: Too Much Tonic i Pukdttay-ak~mee,S. et aL (1994) "Irons. R. Sac. Can Be ToxJc,Ohio StateUniversity Trop.Me'J ~yg. 88, 324-327 I0 Foote, £J. and Cowman, A.F, (1994) Acta 2 larcho, 5. (19~'J} Qu;r,~n~'~ Predecess0~ Trap!ca56,157-171 Francesca rorti and the Edrly Histoo, of CinII hlarchiafava. E. ec ol. (1900) Malone and chona. Jc.hr,s NapkinsUniversityPress Microo~anisms, William Wood & Co. 3 Gramiccia.G. (1987)Acta Leidensla55.5-13 12 vvnrte. N.J. (1985) Chn. Pharmacokinet;cs I O. 4 Gramiccia,G. (1987)Aciu leidensid 55. 15-20 187-215 S Smith,D.C (1976)/. Hist. Med. 31,343-367 6 Ver4aave,JP. (1995) Trap. Geog. Med. 47 Steven R. Meshnlck is a[ the Department 252-258 oi" Epidemiology, University of Michi~(In ~chool ; Knell. ~,.j. 1991) Malaria. Oxford University of Public Health, Ann Arbor, MI 48109, USA. Press Te# + I 3"~3 647 2406, Fax: + I 313 764 B Burke.J.(1978)Connections.L~le Brown& Co al 192, e-mall: [email protected]
Malaria Transmission and Vector Control &l'4. Greenwood Malaria has been contairled very effectively in some parts of the world through control of its vector by environmental management and household spraying and, in some ar~as, this approach has been so successful that eradication of the infection has been achieved. In areas of high endemicity, such as those found in much oftropica! Africa, vector control, initially by household spra?ing I-3 and, more recently, by use of insecticide-impregnated bednets and curtains 4-7, has been very successful in decreasing mortality and morbidity from malaria, even though transmission has not been inten~Jpted. However, in most of these trials, control was maintained for only a short period. Recently, it has been suggested that in areas of high malaria endemicity the bene£rts of effective vector-control programmes will be only transitory, and that deaths and disease will not be prevented bL.~ only postponed ~-'° because efficient vector contro! will prevent the natural development of immunity to the infection. If this is the case, spraying, inse~icide-impregnated bcdnets, genetically engineered mosquitoes or transmission-blocking vacones will have littie to offer malaria control in many parts of Africa where malaria is highly endemic and curtznt plans for lace-scale bednet pro~ m m e s in these communities would be ill advised. For this reason, it is important that the data on which this hypothesis is based are reviewed critically. Data on the relationship between malaria mor'~ality and the level of exposure to ;nfec[ion, expressed as the entemological inoculation rate (EIR), collected by Snow and Marsh9 come from 90
a variety of sources and are of very variable quality. The limited data available suggest that the relationship between mortality from malaria and the EIR is not a linear one and that some kind of plateau is reached but the data are too few to indicate at what EIR a threshold is achieved. Identification of this threshold, if it exists, by the collection of more and better-quality data is important because this will define the level of exposure above which vector control measures may have on!y a transitory protective effect, Below this level, any reduction in EIR as a result of vector conb~ol should have a sustained effecton mortality and morbidity due to the infection. The best data on the relationship between the incidence of severe disease and the EIR come from studies undertaken on the coast of Kenya II and in the city of Brazzaville in the Republic of Cqngo 12. On the Kenyan coast, little o.~rrelation was found between the incidence of severe malaria in different communities and their ElK However, a comparison of the incidence of severe malaria among children admitted to hospital in Ifakara, Tanzania, where the EIR is about 300 infective bites per person per year, or to hospital in Kilifi, Kenya, where the EIR is about 10, showed that the overall incidence of severe malaria was about twice as high at Ifakara as at Kilifi (12.0 and 6.6 per I000 children, respectively) ~s. However, these figures may have been brought about by differences in treatment-seeking behaviour between the two communities. Severe anaemia was seen approximately three times more frequently
Cc~r~ghtC~11997,ENecerSconceLtd A l l n g h ~ A ' d
016.~-4758/t)7,~17,~3 S0169-4758(9~}01002-8
at Ifakara than at Killfi, and cerebral malaria more frequently at Kilifi than at Ifakara. Why cerebral malaria should preferentially affect older children, and be mo~e common in areas of medium than of very high endemicity is not certain, but could be explained if only a few strains of malaria parasite can cause this complication ~';. At Brazzaville in the Republic of Cargo, the incidence of severe malari ~ in communities where the EIR w'.s less than two was about one half .nat seen in communities where it was in the range 12-52, but there was a suggestion t2 that the incidence of severe malaria fell at an EIR above 52. The influe:~ce of EIR oq the incidence of malaria infection was studied in detail in a group of Kenyan children by Beier and colleagues Is. They found a close correlation between Lhe incidence of infection and the EIR per day durin 8 the precedin 8 t w o weeks (correlation coefficient = 0.74), the attack rate increasing tenfold over a range of up to two infective bites per day. In this study, both symptomatic and asymptomatic infections were included, and only children aged six months to six years were studied. Trape and Rogier I° have recently studied morbidity from malaria in two Senegalese villages with very different EIRs (200 in Dielmo, and 20 in Ndiop) and shown that the average number of clinical attacks of malaria, defined as an axillary temperature of 38.0°C or a rectal temperature of 38.5°C accompanied by the appearance of parasites in a subject who was previously parasite negative or by an increase in parasite density in any subject who was already Parasitology Today, voL 13, no. 3, •997
W h y Does Quinine Still Work After 350 Years of Use? S.R. Meshnick P!asmodium strains have become resistant to almost every antimalarial dru~ introduced in the past 60 years. In contrast, even though quinine has been used to treat malaria for rnore than 3S0 years, moderate resistance has only recently developed in limited geographic areas i. What is so special about quinil;e~ There are three possible expbnations. FirsL quinine's intraparasitic target might be so specific that mutations that confer resistance carl occur only at an exceptionally slow rate. Second, modem strains of malaria parasite might in fact already be resistant to quinine when compared to the strains of malaria that existed 300 years ago. In other words, quinine resistance m~y have developed slowly over centuries of use, and we simply never noticed that the required doses of quinine had increased. Third, perhaps quinine had not been used Frequently enough to exert evolutionary selective pressure until recently, Understanding quinine's longevity could have important implications. For example, if quinine's target is so speciz! then it would be very useful to identify it. On the other hand, if modem malaria parasites already are quinine resistant, then we should try to understand how this resistance has affected parasite metabolism. Quinine has its origins in Peru in the early 17th century 24, where Indians were observed chewing on the bark of cinchona trees, and the bark was found to be an effective 'febrifuge' for 'ague' or 'intermittent Fever'. It is unlikely that the antimalarial action of the bark had been known to indigenous medicine, because malaria had only recently been imported inLo %uth America. Since too!aria was rampant in Europe, and would remain so until the early 20th century, there was soon a great demand for 'bark' or 'Jesuit's powder'. In 1820, a method of purifying quinine from cinchona bark. was developed 0y Pelletier and Caventou in Paris. Quinine became especially important to English settlers in India, who invented the practice of sweetening their prophylactic 'tonic' with gin 8. (Tonic water, toda) :, contains only 15 mg of quinine per bottle 9, not nearly enough to have antimala~4al activity. It was probably just a good excuse For ~in.) Parasitology lc~;ay,vol. 13, no. 3, 1997
Quinine is structura!ly similar to the other ouinoline an~i:nalarlals, especially to mefloquine, wh' : . is also a quinoline methanol. The mechanism by which quinine kills malar:a parasites is unknuwn, so the drug target could possibly be one t,,at is especially important and immutable. As mab ri~ parasites have become resistant to all of the ,xner quinolines, however, this seems unlikely. Pldsmodium fdlciporum develops resistance to chloroquine and mefioquine by becoming able to pump the drag out from the cell (reviewed in Ref. I 0). Both mefloquine dnd chloroquine have side chains that contain linear strings of carbon and nitrogen atoms. Quinine, on the other hand, has a ,much more complex side chain containing carbons and one nitrogen in a rigid two-dnl structure. Perilaps the malada proteins which pump out mefloquine and chloroqui,-,~ are unable to handle qutn~ne because of its side chain structure. However, microorganisms have become resi.~l"an~ to every antimicrobial agent ever put into widespread use. Malaria is no exception. It is highly unlikely that biochemistry alone can explain the relative longevity of quinine as an e,~ective antimala,'ial. Have quinine doses increased over the centuries? There are several reasons why we will never know the doses of quinine that were effective before the 20th century. First, prior to 1820, ague was treated with suspensions of finely powdered cinchona bar~. in wine. The amounts of absorbed (bioavailable) quinine mu~ have varied depending on how this suspension was prepared. Also, the amount of quinine contained it, c;nchona bark from different sources varied greatly (0.3-8%), and onchona bark contains vanable amounts of other antimalarial alkaloids such as cinchonine 3. Thus. it is imposs~b!e to de :rmine accu~a~c!y now much antimalar al material was actually given to pat~ent.i Second, standard practice: ~ideli~'s for drug use are invent!ons of the late 23~.h centuw. Prior to that, physicians ,vou!d use medicines at whatever doses they felt comfortable with (malpractice law was also ~ill in its infancy). And if those didn't weft<, they might repeat the treatment, double the dose,
or combine medications according to their ir.÷,j~iVe f~lina~S Finally, it wasn't until the end of the 19th centu~7, that malaria parasites were identified under the microscope. Thus, many illnesse~ other than m a M a wen ~. mistakenly treated ';~'ith quinine; marly patients who didn'~ respond to quinine may really have had another infectious disease and not malaria. Conversely, patients who seemed to resp,:;~,! to low doses of quinine may have had illnesses (such as viral infections) that spontaneously resolved. So we will never know how sensitive malaria parasites were to quinine prior to the 20th century. What were the doses of quinine used once malarie could be deftnitively diagnosed? According to a textbook or, m a m a written in 1900, quinine was to be given intramuscuMy or subcutaneously with a 2 g loading dose, folIo'wed by I g every 6-8 h (Re[ I I ). This is very similar to current recommendations, suggesting that there has been no cha:~e i. ~usc~ptibil~)'. Wns quinine used in enough quantity to engender resistance? Plasmadiun ~faloparum is the most drug-resis~nt species of malaria and is also the species against which quinine is most. ofi.en used today. Historically, however, the majority of malaria in Europe and North America was due to P. vivax or P. malariae. Even in colonial India, a large proportion of the malaria was due to P. vivax. Therefore. prior to the 20th century, most quinine waS probably used to treat viva× not fa!ciparam malaria. Furthermore, quinine waS probably used by a relative!y small group of people when compared to the broad distribution and use of antimalarials today 3. There are several pharmacological factors which may also have contributed to the maintenance of quinine's efficacy. Fir~ quinine quite often causes annoying side effects such as nausea and vomiting, ringing in the ears, headn2 loss. dizziness, headache, and/or disturbed vision. So people are unlikely to take the drug unless absol,~¢e!ynecessary,.Second, quinine has a relatively short, half-life (I0 h) (Re[ 12), so it will not linger at subtherapeutic concentrations in the blood..ream for very long. Thus, P. faldl~arum parasites, all in all, have probably Lad ~ia~ively
Coprm~l©1997 EJs~e,'~c,erceLTa Al'7~"tsrese".~; 0 o~-47bfl'e?.'$17 'qO ~Ou~9-'iTSS(97)OlO}3-X
~9
Comment little exposure to quinine and thus little evolutionary selective pressure from it. Quinine has remained an effective antimalarial for 350 yeal%. Prior to recent times, it was probably never used at a high enough frequency against P. falciparum to induce resistance. This could serve as a lesson to us as we look for ways to preserve the efficacy of our current antimalarials and other drugs.
References 9 Carroll, bl. 0995) Quinipe: Too Much Tonic i Pukdttay-ak~mee,S. et aL (1994) "Irons. R. Sac. Can Be ToxJc,Ohio StateUniversity Trop.Me'J ~yg. 88, 324-327 I0 Foote, £J. and Cowman, A.F, (1994) Acta 2 larcho, 5. (19~'J} Qu;r,~n~'~ Predecess0~ Trap!ca56,157-171 Francesca rorti and the Edrly Histoo, of CinII hlarchiafava. E. ec ol. (1900) Malone and chona. Jc.hr,s NapkinsUniversityPress Microo~anisms, William Wood & Co. 3 Gramiccia.G. (1987)Acta Leidensla55.5-13 12 vvnrte. N.J. (1985) Chn. Pharmacokinet;cs I O. 4 Gramiccia,G. (1987)Aciu leidensid 55. 15-20 187-215 S Smith,D.C (1976)/. Hist. Med. 31,343-367 6 Ver4aave,JP. (1995) Trap. Geog. Med. 47 Steven R. Meshnlck is a[ the Department 252-258 oi" Epidemiology, University of Michi~(In ~chool ; Knell. ~,.j. 1991) Malaria. Oxford University of Public Health, Ann Arbor, MI 48109, USA. Press Te# + I 3"~3 647 2406, Fax: + I 313 764 B Burke.J.(1978)Connections.L~le Brown& Co al 192, e-mall: [email protected]
Malaria Transmission and Vector Control &l'4. Greenwood Malaria has been contairled very effectively in some parts of the world through control of its vector by environmental management and household spraying and, in some ar~as, this approach has been so successful that eradication of the infection has been achieved. In areas of high endemicity, such as those found in much oftropica! Africa, vector control, initially by household spra?ing I-3 and, more recently, by use of insecticide-impregnated bednets and curtains 4-7, has been very successful in decreasing mortality and morbidity from malaria, even though transmission has not been inten~Jpted. However, in most of these trials, control was maintained for only a short period. Recently, it has been suggested that in areas of high malaria endemicity the bene£rts of effective vector-control programmes will be only transitory, and that deaths and disease will not be prevented bL.~ only postponed ~-'° because efficient vector contro! will prevent the natural development of immunity to the infection. If this is the case, spraying, inse~icide-impregnated bcdnets, genetically engineered mosquitoes or transmission-blocking vacones will have littie to offer malaria control in many parts of Africa where malaria is highly endemic and curtznt plans for lace-scale bednet pro~ m m e s in these communities would be ill advised. For this reason, it is important that the data on which this hypothesis is based are reviewed critically. Data on the relationship between malaria mor'~ality and the level of exposure to ;nfec[ion, expressed as the entemological inoculation rate (EIR), collected by Snow and Marsh9 come from 90
a variety of sources and are of very variable quality. The limited data available suggest that the relationship between mortality from malaria and the EIR is not a linear one and that some kind of plateau is reached but the data are too few to indicate at what EIR a threshold is achieved. Identification of this threshold, if it exists, by the collection of more and better-quality data is important because this will define the level of exposure above which vector control measures may have on!y a transitory protective effect, Below this level, any reduction in EIR as a result of vector conb~ol should have a sustained effecton mortality and morbidity due to the infection. The best data on the relationship between the incidence of severe disease and the EIR come from studies undertaken on the coast of Kenya II and in the city of Brazzaville in the Republic of Cqngo 12. On the Kenyan coast, little o.~rrelation was found between the incidence of severe malaria in different communities and their ElK However, a comparison of the incidence of severe malaria among children admitted to hospital in Ifakara, Tanzania, where the EIR is about 300 infective bites per person per year, or to hospital in Kilifi, Kenya, where the EIR is about 10, showed that the overall incidence of severe malaria was about twice as high at Ifakara as at Kilifi (12.0 and 6.6 per I000 children, respectively) ~s. However, these figures may have been brought about by differences in treatment-seeking behaviour between the two communities. Severe anaemia was seen approximately three times more frequently
Cc~r~ghtC~11997,ENecerSconceLtd A l l n g h ~ A ' d
016.~-4758/t)7,~17,~3 S0169-4758(9~}01002-8
at Ifakara than at Killfi, and cerebral malaria more frequently at Kilifi than at Ifakara. Why cerebral malaria should preferentially affect older children, and be mo~e common in areas of medium than of very high endemicity is not certain, but could be explained if only a few strains of malaria parasite can cause this complication ~';. At Brazzaville in the Republic of Cargo, the incidence of severe malari ~ in communities where the EIR w'.s less than two was about one half .nat seen in communities where it was in the range 12-52, but there was a suggestion t2 that the incidence of severe malaria fell at an EIR above 52. The influe:~ce of EIR oq the incidence of malaria infection was studied in detail in a group of Kenyan children by Beier and colleagues Is. They found a close correlation between Lhe incidence of infection and the EIR per day durin 8 the precedin 8 t w o weeks (correlation coefficient = 0.74), the attack rate increasing tenfold over a range of up to two infective bites per day. In this study, both symptomatic and asymptomatic infections were included, and only children aged six months to six years were studied. Trape and Rogier I° have recently studied morbidity from malaria in two Senegalese villages with very different EIRs (200 in Dielmo, and 20 in Ndiop) and shown that the average number of clinical attacks of malaria, defined as an axillary temperature of 38.0°C or a rectal temperature of 38.5°C accompanied by the appearance of parasites in a subject who was previously parasite negative or by an increase in parasite density in any subject who was already Parasitology Today, voL 13, no. 3, •997