TRANSACTIONSOFTHEROYALSOCIETYOFTROPICALMEDICINEANDHYGIENE(1999)93,641-643

Randomized controlled trials of 5- and 14-days primaquine therapy against relapses of vivax malaria in an Afghan refugee settlement in Pakistan

Mark Rowland’J and Naeem Durrani’ ‘HealthNet International, University Town, Peshawar, Pakistan; ‘London School of Hygiene and Tropical Medicine, London WClE 7HT, UK

Abstract Primaquine is the only drug available that can eliminate hypnozoites from the liver and prevent relapses of vivax malaria. The World Health Organization recommends a course of 14-21 days depending on region and strain. The National Malaria Control and Eradication Programmes of Pakistan and India have adhered to a 5-day course as their standard as it is deemed more practical to implement and because facility for detecting glucose 6-phosphate dehydrogenase (GGPD) deficiency is seldom available at the periphery. Evidence for the effcacv of the 5-dav regimen is controversial or lackinn. Two. vear-long. randomized controlled trials were undertaken in ah Afghan refugee camp in north-western Pakistan usirg a process of passive case detection and treatment at the camp’s clinic: the first trial compared treatment with chloroquine alone versus chloroquine plus 5-days primaquine, the second trial compared chloroquine alone versus chloroquine plus 14-days primaquine. Chloroquine is not hypnozoitocidal and was administered to eliminate the erythrocytic stages and to alleviate clinical symptoms. The daily primaquine dose was 0.25 mg/kg bodyweight and the total chloroquine dose was 25 mg/kg over 3 days. During the first trial 52% (1291250) of the non-primaquine group recorded a 2nd clinical-parasitaemic episode and 23% recorded a 3rd, whereas 5 1% (128/250) of the 5-days primaquine group reported a 2nd episode and 21% recorded a 3rd. During the second trial 49% (49/100) of the non-primaquine group recorded a 2nd episode and 25% recorded a 3rd, whereas only 32% (32/100) of the 14-days primaquine group recorded a 2nd and only 2% recorded a 3rd. The 5-days primaquine regimen has no value as an anti-relapse therapy and should be abandoned. In extended tests in vivo in which vivax cases (n = 3 1) were treated with chloroquine 25 mg/kg and 14-days primaquine, there was no parasite recrudescence within 28 days and hence no evidence of resistance to chloroquine.

Keywords: malaria, Plasmodium uiuax, chemotherapy, primaquine, relapse, refugees, Pakistan

Introduction Primaquine is the only drug available that can elimi-

nate hvpnozoites of vivax malaria from the liver and prevent-relapses of the disease (GILLES & WARRELL, 1993). The World Health Ornanization recommends a 14-day course (0.25 mg/kg diily) as the standard ther- apy and longer courses against more tolerant strains from South-East Asia and the Pacific (WHO, 1990). The National Malaria Control Programme of Pakistan has adopted a truncated 5-day course as their standard because of the risk of haemolysis in glucose 6-phosphate dehydrogenase (G6PD)-deficient cases and because facility to detect this genetic condition is seldom available in clinics at the periphery. India and other Asian countries have the same policy as Pakistan and consider the 5-day course to be more practical and economical (BASAVARAJ, 1960; SHARMA et al., 1973; ROY et al., 1977). Evidence for the clinical efficacy of the 5-day regimen is controversial or lacking (ROY et al., 1977; SHARMA et al, 1990; SINGH et al., 1990; BAIRD, 1998; GOGTAY~LKSHIRSAGAR, ~~~~;GoGTAY etal., 1998) and a call for randomized placebo-controlled trials has recently been made (BAIRD, 1998).

This-concern is shared by the agencies responsible for the health care of Afahan refugees in Pakistan. Vivax malaria is a major problem in this population with over 100 000 episodes being treated each year at clinics run by the United Nations High Commissioner for Refugees (UNHCR), the Pakistan Ministry of Health, and non- governmental organizations (ROWLAND et al., 1994). Many patients given a 5-day course return to clinics with 2nd or 3rd episodes within months of treatment (ROW- LAND et al, 1997; SHAH et al., 1997). With the seeming failure of primaquine therapy to significantly reduce annual incidence rates, and with GGPD deficiency being a common condition among Afghan (15 %) and Pakistani (7%) Pathans (BOUMA et aZ., 1995), more evidence is

Author for correspondence: Mark Rowland, Disease Control and Vector Biology Unit, London School of Hygiene and Tropical Medicine, Keppel Street, London WClE 7HT, UK; phone t44 020 7927 2333, fax +44 020 7580 9075, e-mail m.rowland@lshtm.ac.uk

needed to justify continuing with the 5-day regimen. Between 1996 and 1998-2 randomized controlled trials of primaquine were undertaken at an Afghan refugee camp. The first compared the standard 5-day course of primaquine versus no anti-relapse therapy. When this failed to demonstrate a differential impact, a 2nd trial was undertaken to examine the efficacy of the WHO-recom- mended 14-day course since resistance to primaquine may, conceivably, have evolved. All cases were treated concurrently with chloroquine to eliminate erythrocytic stages and to alleviate clinical symptoms. Since chlor- oquine-resistant vivax malaria was becoming quite pre- valent in Asia (BAIRD et aZ., 1991; SCHUURUMP et al, 1992; MYAT-PHONE-KYAW et al., 1993; GARG et al., 1995), it seemed prudent to also carry out a 28-day in- vivo test on Plasmodium vivax infections (WERNSDOR- FER & PAYNE, 1988) in the same refugee camp.

Materials and Methods Patients with vivax malaria were recruited by a process

of passive case detection at the clinic and microscopy unit run bv HealthNet International in Adizai refugee camo (population 3950) near Charsadda in North W’est Fron- tier Province. All families within the camp were regis- tered at the clinic, censused each year, and received free diagnostic and treatment services. Clinical, parasitaemic cases with temperature >37.5”C or history of fever were recruited in July and August at the onset of the main transmission season (ROWLAND et al., 1997), tested for G6PD deficiency (Sigma Diagnostics, Poole, Dorset, UK), and allocated at random to either the placebo or the primaquine group. Cases not admitted to the study included patients with mixed infections or with evidence of recent antimalarial intake, cases of G6PD deficiency or severe anaemia, children aged <3 years, pregnant women, and the very elderly. Each case was followed-up each day by outreach workers who provided supervised treatment until the course had been comnleted.

All cases were treated with 25 mg chloioquine (Niva- auine-Pa. Rhone-Poulenc. France) ner ke bodvweieht in divided’doses over 3 days. Cases aliocatgd to the n&- primaquine groups were given no further treatment. Cases allocated to the primaquine groups were given a

642 MARK ROWLAND AND NAEEM DURRANI

daily dose of 0.25 mg primaquine (International Dis- pensary Association, Holland) per kg: for 5 days in the 1996-97 studv. and for 14 davs in the 1997-98 studv. Any relapse 0; new infection”was retreated using &e same regimen and procedure administered at admission.

A

In the extended in-vivo resistance test, clinical and parasitaemic cases were treated with 25 mg chloroquine per kg over 3 days and with daily 0.25 mgprimaquine per kg for 14 days (WERNSDORFER & PAYNE, 1988). Cases testing positive for chloroquine in Dill-Glazko urine tests were excluded. Blood smears were taken at I-3-day intervals for 28 days: 200 fields were examined before any blood film was classified as malaria-negative. The num- ber of parasites were counted against 200 leucocytes.

Results The admission variables were consistent within and

between the 2 controlled trials. The average age of cases in the non-primaquine and primaquine groups were, respectively, 10.0 years and 10.4 years in the 5-day trial, and 11 years and 12 years in the 14-day trial. Females comprised 5 1% of the 5-day and 54% of the 14-day trial. A total of 500 cases were enrolled into the 5-day trial but only 200 cases were enrolled into the 14-day trial owing to the greater demands placed on outreach workers in administering the extended treatment.

In the 5-day primaquine trial, 52% (1291250) of individuals in the non-primaquine group recorded a 2nd clinical-parasitaemic episode and 23% recorded a 3rd episode between August 1996 and June 1997, whereas 51% (128/250) of individuals from the prima- quine group recorded a 2nd episode and 21% recorded a 3rd over the same period (Fig. A). These rates in the non- primaquine and 5-day primaquine groups were not significantly different from each other. The mean inter- val between clinical episodes was similar in non-prima- quine and 5-day primaquine groups (Table).

In the 14-day primaquine trial, 49% (491100) of the non-primaquine group recorded 2nd episodes and 25% recorded 3rd episodes between August 1997 and June 1998, whereas only 32% (321100) of the primaquine group recorded 2nd episodes (x2 = 6, d.f. = 1, P = 0.014) and only 2% recorded 3rd episodes (x” = 17, d.f. = 1, P < 0.001) over the same period (Fig. B). The interval between clinical episodes was significantly long- er (t = 2.4, P = 0.0 17) in the 14-day primaquine group than in the non-primaquine group (Table). If one assumes that 2nd episodes in the 14-day primaquine group were due to new infections rather than to relapses, the relapse frequency in the placebo group is estimated to be 17%.

Individuals aged >15 years had fewer 2nd episodes (38% [36/95], P < 0.001) and fewer 3rd episodes (36% [13/36], P = 0.015) than individuals aged S15 years (55% [221/405] and 44% [98/221], respectively).

The 5-day and 14-day primaquine therapies appeared to be well tolerated in GGPD-deficient individuals. Although there was no systematic monitoring for hae- molysis, no instances were detected passively by clinic microscopists during or after therapy.

Geometric mean parasite density in the extended in-

2nd 3rd

Vivax episode

4th

0 non-primaquine E3 primaquine for 5 days

B

50

0 2nd 3rd 4th

Vivax episode

anon-primaquine 0 primaquine for 14 days

Figure. Frequency ofvivax malaria episodes over the course of 1 year in groups either non-treated or treated with primaquine (0.25 mg per kg bodyweight per day) (A) for 5 days, (B) for 14 days.

vivo test decreased from 65 15 to 0 per pL within 2 days of starting treatment. All cases remained clear of tropho- zoites and gametocytes until the end of the test on day 28.

Discussion The present study appears to be the first randomized

controlled trial of 5-days primaquine therapy against vivax malaria held under field conditions. The regimen was first adopted by the health authorities of India and Pakistan in the early 1960s (SINGH et aZ., 1954; BASA- VARAJ, 1960) because it was more practical to implement than the WHO-recommended 14-day course (COVELL et al., 1954) which had always had problems of com-

Table. Geometric mean interval between episodes of vivax malaria in Afghan refugees not treated with primaquine or treated with primaquine

From admission to 1 st episode From 1st to 2nd episode

Year Treatment group n Interval (95% CI) in days n Interval (95% CI) in days

1996-97 Non-primaquine 129 118 (102-135) 105 (86-128) Primaquine for 5 days 128 136 (117-156) 119 (98-145)

1997-98 Non-primaquine 49 68 (53-86)* - Primaquine for 14 days 32 103 (85-126)” -

*Significantly different, P = 0.017.

RANDOMIZED CONTROLLED TRIALS OF 5-AND 14-DAYS PRIMAQUINE THERAPY 643

pliance (WHO, 1966). Relapse rates in those early as well as inlater studies (SHARMA etaZ., 1973; ROY etaZ., 1977) never exceeded 7% in the 0.5-2 years of follow-up, and this observation has been used to justify the 5-day policy ever since (GOGTAY et al., 1998) even though a relapse rate of under 7% may have been entirely natural (BAIRD, 1998). Both SINGH et al. (1990) and SHARMA et al. (1990) compared the 5-day course of primaquine (plus chloroquine) against non-primaquine treatment (chlor- oquine given alone) in consecutive years. SINGH et al. (1990) obtained similar rates of 2nd infections (around 10%) in both primaquine and non-primaquine.groups whereas SHARMA et al. (19901 found fewer 2nd infec- tions in the 5-days primakuine’group. The problem with this method is that relapses cannot be distinguished from new infections, the rate of which may vary from year to year owing to unstable transmission. Where the 5-day regimen has been tested against a placebo simulta- neously, and this had only been done under artificial conditions in the USA on 10 non-immune volunteers, the rate of relapse was 100% in both groups (CONTACOS et al., 1974). The absolute failure of the 5-day regimen to prevent relapses is confirmed by the present field study in Pakistan. There is no justification to continue with it, and as BAIRD (1998) has emphasized, continued use of sub- optimal doses invites selection of resistance to the only drug available for anti-relapse therapy.

It is not proposed to switch to 14-days primaquine therapy since this is only justified where vivax transmis- sion is under control or where patient compliance with the 14-day regimen can be assured (WHO, 1990). Haemolytic reaction to primaquine can be particularly severe in Asian and Mediterranean variants of GGPD- deficiency (CLYDE, 198 1). Because daily medical super- vision for 14 days is unfeasible in areas of Pakistan where malaria is most common, the only sensible policy is to abandon use of primaquine until conditions for effective and safe use can be met.

Acknowledgements We are grateful to the staffofthe HealthNet clinic in Adizai for

their wo&, and to Dr Naveeda Bano of UNHCR for encoura- ging this study. HealthNet International’s Malaria Control Programme is supported by the European Commission (DGl) and UNHCR. This research was partly supported by the Department for International Development of the United King- dom and by HealthNet International.

References Baird, J. K. (1998). Primaquine as anti-relapse therapy for

Plasmodium vivax. Transactions of the Royal Society of Tropical Medicine and Hygiene, 92, 687.

Baird, J. K., Basri, H., Purnomo, Bangs, M. J., Subianto, B., Patchen, L. C. & Hoffman, S. L. (1991). Resistance to chloroquine by Plasmodium vivax: in Irian Jaya, Indonesia. American Journal of Tropical Medicine and Hygiene, 44, 547-552.

Basavaraj, H. R. (1960). Observations on the treatment of 678 malaria cases with primaquine in an area free from malaria transmission in Mysore State, India. Indian Journal of Malar- ioloD, 14, 269-28 1.

Bouma, M. J., Goris, M., Akhtar, T., Khan, N. & Kita, E. (1995). Prevalence and clinical presentation of glucose-6- phosphate dehydrogenase deficiency in Pakistani Pathan and Afghan refugee communities in Pakistan: implications for the use of primaquine in regional malaria control programmes. Transactions of the Royal Society of Tropical Medicine and Hygiene, 89, 62-64.

Clyde, D. F. (1981). Clinical problems associated with the use of primaquine as a tissue schizontocidal and gametocidal drug. Bulletin of the World Health Organization, 59, 39 l-395.

Contacos, P. G., Coatney, G. R., Collins, W. E., Briesch, P. E. & Jeter, M. H. (1974). Five day primaquine therapy-an evaluation of radical curative activity against vivax malaria infection. American Journal of Tropical Medicine and Hygiene, 22,693-695.

Covell, G., Coatney, G. R., Field, J. W. & Singh, J. (1954). Chemotherapy of Malaria. Geneva: World Health Organiza- tion, Monograph Series, no. 27.

Garg, M., Gopinathan, N., Bodhe, I’. & Kshirsagar, N. (1995). Vivax malaria resistant to chloroquine: case reports from Bombay. Transactions of the Royal Society of Tropical Medicine and Hygiene, 89,656-657.

Gilles, H. M. & Wanell, D. A. (1993). Bruce-Chwatt’s Essential Malariology, 3rd edition. London: Edward Arnold.

Gogtay, N. & Kshirsagar, N. A. (1998). Primaquine as anti- relapse therapy for Plasmodium vivax: a reply. Transactions of the Royal Society of TroPical Medicine and Hvpiene. 92. 687.

Gogtay,N., Garg, A.%., Kidam, V., Kamtekar; k. &-Kshirsagar, N. A. (1998). A 5 days primaquine regimen as anti-relapse therapy for Plasmodium viuax. Transactions of the Royal Society of Tropical Medicine and Hygiene, 92, 34 1.

Myat-Phone-Kyaw, Myint-00, Myint-Lwin, Thaw-Zin, Kyin- Hla-Ave & Nwe-Nwe-Yin (1993). Emergence of chloro- quine-resistant Plasmodium &ax in Myanmar (Burma). Transactions of the Royal Society of Tropical Medicine and Hygiene, 87, 687.

Rowland, M., Hewitt, S. & Durrani, N. (1994). Prevalence of malaria in Afghan refugee villages sprayed with lambdacyha- lothrin or malathion. Transactions of the Royal Society of Tropical Medicine and Hygiene, 88, 378-379.

Rowland, M., Hewitt, S., Durrani, N., Bano, N. & Wirtz, R. (1997). Transmission and control of vivax malaria in Afghan refugee settlements in Pakistan. Transactions of the Royal Society of Tropical Medicine and Hygiene, 91,252-255.

Roy, R. G., Chakrapani, K. P., Dhinagaran, D., Sitaraman, N. L. & Gosh. R. B. (1977‘1. Efflcacv of ii-dav treatment of PlasmodiumPvivax iniection in Tamii Nadu. IidianJoumal of Medical Research, 65, 652-656.

Schuurkamp, G. J., Spicer, P. E., Kereu, R. K., Bulungol, P. K. & Rieckmann, K. H. (1992). Chloroauine-resistant Plasmo- dium vivax in-Papua hew Guinea. Tiansactions of the Royal Society of Tropical Medicine and Hygiene, 86, 12 1 - 122.

Shah, I., Rowland, M., Mehmood, P., Mujahid, C., Raziq, F., Hewitt, S. & Durrani N. (1997). Chloroquine resistance in Pakistan and the uusuree of falcioarum malaria in Pakistani and Afghan refugee populations. Annals of Tropical Medicine and Parasitology, 91, 591-602.

Sharma, M. I. D., Sehgal, P. N., Vaid, B. K., Dubey, R. C., Nagendra, S., Paithne, P. K. & Toshi, M. L. (1973). Effec- tiveness of drug sched&le being fillowed unde; the i$ational Eradication Programme, India, for radical cure of vivax malaria cases. Journal of Communicable Diseases, 5, 167- 174.

Sharma, R. C., Gautam, A. S., Orlov, V. & Sharma, V. P. (1990). Relapse pattern of Plasmodium vivax in Kheda District, Gujerat. IndianJournal of Malariology, 27, 95-99.

Singh, J., Ray, A. P., Misra, B. G. & Nair, C. P. (1954). Anti- relapse treatment with primaquine and pyrimethamine. Indian 7ournal ofMala&lopv, 8. 127-136.

Singh, M, Mishia, A. K. &“&h&ma, V. P. (1990). Radical treatment of vivax malaria in Madhya Pradesh, India. Indian Journal of Malariology, 27, 55-56.

Wernsdorfer, W. H. & Payne, D. (1988). Drug sensitivity tests in malaria parasites. In: Malaria-Principles and Practice of Malariklogy, Wemsdorfer, W. H. & McGregor, I. (editors). Edinburgh: Churchill Livingstone, pp. 1765- 1800.

WHO (1966). World Health Organization Expert Committee on Malaria. Twelfth report. Geneva: World Health Organization, Technical Report Series, no. 324.

WHO (1990). Practical Chemotherapy of Malaria. Geneva: World Health Organization, Technical Report Series, no. 805.

Received 15Jun.e 1999; revised 12 August 1999; acceptedfor publication 23 August 1999