Am. J. Trop. Med. Hyg., 80(2), 2009, pp.

188193 Copyright 2009 by The American Society of Tropical Medicine and Hygiene

Methemoglobinemia and Adverse Events in Plasmodium vivax Malaria Patients Associated with High Doses of Primaquine Treatment
Jaime Carmona-Fonseca, Gonzalo lvarez, and Amanda Maestre*
Grupo Salud y Comunidad and Grupo Malaria, Universidad de Antioquia, Medelln, Colombia

Abstract. Primaquine (PQ) is recommended to prevent relapses in patients with Plasmodium vivax malaria infection. However, treatment with PQ causes methemoglobinemia. In this study, we measured the methemoglobin (MetHB) levels in three groups of subjects who received PQ treatment at 0.58, 0.83, or 1.17 mg/kg/d. A total of 112 subjects were studied. MetHB levels were detected at > 4% in 4650% 1 day after PQ treatment in all three groups and 49% of subjects had MetHB levels > 4% 15 days after treatment. Only subjects receiving the highest doses of PQ had mild and brief adverse events, and 17% of them were associated with treatment. We conclude that when PQ is administered under certain conditions (i.e., normal glucose-6-phosphate dehydrogenase activity, in non-pregnant subjects and with a light meal), daily doses as high as 1.17 mg/kg do not represent a serious risk of high MetHB levels to patients. INTRODUCTION Primaquine (PQ; Neo-Quipenyl and Palum) is used for controlling gametocytes in Plasmodium falciparum infection to block transmission and against persistent liver forms (hypnozoites) in P. vivax and P. ovale to prevent relapses after a primary attack. The antimalarial is a powerful inductor of methemoglobin (MetHB), even with normal, glucose-6-phosphate dehydrogenase (G6PD) activity, transforming hemoglobin (HB; Fe2+) into reduced HB (Fe3+)1 within erythrocytes. When MetHB in blood increases to > 2%, the condition is known as methemoglobinemia.2 Under physiological conditions, the erythrocyte has evolved several mechanisms of protection by reversion of MetHB to HB.3 Methemoglobinemia can also be caused by dapsone, pamaquine, primaquine, and sulfonamides.3 Congenital methemoglobinemia caused by enzymatic deficit or by defect on the hemoglobin molecule itself has been reported.1,2,4,5 Within the normal red blood cell, a small amount of MetHB can be present, but this is continuously reduced to HB by action of the NADH-diaphorase (cytochrome 5 reductase) enzyme, which normally keeps the MetHB level < 1%. Clinical findings in patients with excessive MetHB correlate to blood levels.6 MetHB levels < 20% provoke no signs or symptoms in most people. When levels exceed 30% most people will suffer dyspnea, nausea, and tachycardia; 55% MetHB causes lethargy and stupor; > 55% MetHB results in cardiac arrhythmia and neural depression; and > 70% MetHB is usually fatal. In general, PQ does not cause serious side effects when administrated to healthy whites at standard doses (i.e., 0.25 0.30 mg/kg/d, 14 days for vivax malaria; 0.75 mg/kg single dose for falciparum malaria), all of which are recommended by the Colombian government. However, among African populations, methemoglobinemia7 and hemolytic anemia are relatively common effects because of hemolysis in individuals with an inborn G6PD deficit.79 Hemoglobin is very sensitive to oxidative stress, and a G6PD deficit increases this sensitivity,7,10,11 resulting in more pronounced methemoglobinemia when such patients are given PQ. Nevertheless, methemoglobinemia is routinely observed even in individuals with normal G6PD activity.1,6 G6PD deficiency rates among individuals with a clinical diagnosis of MetHB were 51.3% in comparison to 8.7% in individuals lacking MetHB.12 In Colombia, the standard treatment scheme of chloroquine (CQ) plus PQ (CQ at 10 mg/kg on Day 1, 7.5 mg/kg on Days 2 and 3, and PQ at 0.250.30 mg/kg/d for 14 days) had 100% efficacy against P. vivax primary attacks13 and resulted in recurrences only in 18% of patients during 6 months of follow-up.14 These two antimalarials can induce production of MetHB at therapeutic doses in normal individuals, but the risk increases when it is administered at higher than standard doses or given to those who are G6PD deficient.15 The only licensed drug against hypnozoites is PQ, and some countries have reported failure to prevent relapses in a significant proportion of patients.1619 Therefore, higher-dose regimens of PQ are being applied in several countries.20,21 Moreover, the 14-day regimen often proves ineffective because of poor compliance. Hence, the question of PQ toxicity at higher daily or total doses is relevant to further exploration of alternative PQ regimens. As part of research on the efficacy of different regimens of PQ and the frequency of recurrences after a primary P. vivax malaria attack, we administered 2-, 3-, and 5-fold daily standard doses in G6PD-normal subjects and compared MetHB levels before and after PQ treatment. In addition, we identified the type and frequency of adverse effects after administration of each of three PQ treatment regimens. MATERIALS AND METHODS We conducted a randomized, non-masked, controlled clinical study on P. vivaxinfected patients. Volunteers were part of a larger study that aimed at studying the efficacy of different CQ plus PQ regimens against recurrences in P. vivax infected patients. The patients were recruited and randomized into one of three experimental groups with different treatment regimens. All subjects received the same regimen of CQ treatment first (10 mg/kg on Day 1 and 7.5 mg/kg on Days 2 and 3) and then received a 3-day consecutive course of PQ (Sanofi~Synthelabo, Paris, France): Group I received PQ 0.58 mg (base)/kg/d (i.e., 50% total and twice daily standard dose); Group II received 0.83 mg (base)/kg/d (i.e., 71% total and 3-fold daily standard dose); and Group III received 1.17 mg (base)/kg/d (i.e., equal total and 5-fold daily standard dose). All treatments were supervised. Antimalarials were supplied

* Address correspondence to Amanda Maestre, Grupo Grupo Salud y Comunidad, Universidad de Antioquia, Calle 62 # 5259, Medelln, Colombia. E-mail: aemaestre@quimbaya.udea.edu.co

188

METHEMOGLOBINEMIA AND PRIMAQUINE

189

by the regional health authorities and were administered once daily with 150200 mL of juice and a pastry simultaneously from Day 1. Samples were selected using the Lwanga and Lameshow methods22 to test the differences in proportion between two independent populations. Initial sample size for each treatment was 49 and was increased to 70 to compensate against dropouts during the long period of follow-up. Study area. The study was carried out in two malariaendemic regions in Colombia: Turbo (8542 N, 7644123 W) and El Bagre (73525 N, 744827 W). The majority of inhabitants of these regions are of African origin, although mixing with indigenous and Spanish descendants is common. Both regions have high perennial and unstable malaria transmission with mean annual parasites indices (malaria cases/1,000 exposed individuals) during 20012003 of 19.65 in Turbo and 37.35 in El Bagre.23,24 Malaria patient enrollment and follow-up. This study included subjects from local malaria clinics with acute symptomatic P. vivax malaria and parasitemia diagnosed by thick smear microscopy between September 2003 and September 2006. The study protocol was reviewed and approved by the Ethics Committee of the Faculty of Medicine, Universidad de Antioquia (Medelln, Colombia). Each participant gave informed consent. The inclusion criteria for the study were > 2 years of age, P. vivax parasitemia of > 1,000 asexual forms/L, and willingness to participate. A normal quantitative G6PD screening test25 was required for those administered > 20.25 mg/kg/d PQ, and only individuals with normal G6PD (> 2.29 UI/g HB) were included in the study. Subjects were excluded if they were pregnant, diagnosed with other infectious diseases, had a history of antimalarial medications during the previous 2 weeks, had the presence of diarrhea or vomiting (more than five episodes in 24 hours), had symptoms or signs of severe malaria (according to WHO), had hypersensitivity to antimalarials, or had severe undernourishment.26 Exclusion from the study resulted after intake of any antimalarial different from those provided by the researchers, failure to attend controls, or consent withdrawal. Clinical assessment. Subjects were examined by an experienced general physician before and after treatment. A list of adverse events was examined initially before treatment and used as a baseline for further assessments of the presence or absence of secondary adverse events. These were defined as the expression of any new clinical symptom or sign or the aggravation of an already existing one during the period of treatment. The protocol included qualitative analysis, as recorded by the physician, of 1) the relationship between the event and the treatment, 2) follow-up on the development, 3) the severity, and 4) the outcome. G6PD and MetHB evaluations. The normalized Beutler method25 was used to quantify G6PD activity in the field study sites. The tests were performed by the researcher(s), and the

results were measured using two separate spectrophotometers (Analyzer RA 50, mod 7424, Bayer, Berlin, Germany; and Spectrum Genesys 20, model 4001/4, Thermo Electron Corp., Rochester, NY), which were previously calibrated with standards to assure the accuracy. Levels < 2.29 UI/g hemoglobin were considered as evidence consistent with G6PD deficiency.27 MetHB levels were measured in all subjects on Days 1 (MetHB-1) and 15 (MetHB-15) and followed up on Days 4 and 18 after the completion of 3-day PQ treatment. MetHB levels were determined by spectrophotometry (Spectronic Genesys 20, model 4001/4). The mean of the normal values for the regions under study is reported as 0.5783 0.5781% (median, 0.40; 95th percentile, 1.1562%).28 Statistical analysis. Data were analyzed using EpiInfo 6.04 and SPSS 14.0. The Kolmogorov-Smirnov test was used to test for normalizations of G6PD and MetHB levels. The KruskalWallis test was applied to median G6PD activity and MetHB levels according to sex. A level of significance of 5% was always assumed. In addition, median values of MetHB were compared by the Kruskal-Wallis test for independent k samples, and the Bartlett c2 test was used to compare variance of homogeneity. Post hoc multiple comparisons were performed using the Student-Newman-Keuls test, whereas median readings on Days 1 and 15 after treatment completion were compared by Wilcoxon test for dependent groups. Lineal correlation and regression coefficients were determined for MetHB, G6PD reading, and age. Finally, Pearson and Spearman lineal correlation/regression analyses were also applied, and significance of the model was tested using ANOVA. RESULTS Among 120 subjects under study, a total of 27 were enrolled in Group I, 28 in Group II, and 65 in Group III. For ethical reasons, inclusion in Groups I and II was halted because of the high frequency of recurrences detected during the period of follow-up (Table 1). but the required sample size was successfully reached in Group III. Subjects included were young adults of ~30 years of age. Men represented 66% of the population. Sex distribution within the groups was statistically comparable (P = 0.260). The body weight among the volunteers was ~60 kg, with similar mean values among the groups (P = 0.159110). The mean G6PD activity in all 120 subjects was 3.87 1.12 UI/g HB, with a minimum of 2.30 and a maximum of 7.00 UI/g HB. MetHB levels. MetHB formation was measured in 112 subjects after 1 day and in 105 after 15 days of treatment completion. Global mean MetHB-1 levels were 5.20 5.33% with 65% of subjects at the level > 2%, whereas MetHB-15 levels

Table 1 Study design: primaquine treatment administered based on subjects sex, age, and body weight
PQ group Daily (total) dose (mg/kg) Number [%] Gender (M/F)* Age (years) mean SD Weight (kg) mean SD Subjects (%) with weight > 60 kg

I II III Total

0.58 (1.75) 0.83 (2.49) 1.17 (3.51)

27 [22.5] 28 [23.3] 65 [54.2] 120 [100.0]

21/6 = 3.5 16/12 = 1.3 42/23 = 1.8 79/41 = 1.9

30.9 3.1 37.0 2.5 28.5 1.9 P = 0.000000

61.4 9.3 61.6 9.1 57.1 14.6 P = 0.159110

59 52 40 P = 0.185

* Male/female, for sex: c2 = 2.696; gl = 2; P = 0.260.

190

CARMONA-FONSECA AND OTHERS

Table 2 Global G6PD activity and MetHB levels in P. vivax malaria subjects
Statistics G6PD (UI/g HB) (N = 119) MetHB-1 (%)* (N = 112) MetHB-15 (%) (N = 105)

Mean Median SD Variance Asymmetry SE for asymmetry Kurtosis SE for kurtosis Range Minimum Maximum 5 10 20 25 30 40 50 60 75 80 90 95

3.87 3.60 1.12 1.27 0.84 0.22 0.20 0.44 4.70 2.30 7.00 2.50 2.68 2.91 3.00 3.17 3.32 3.60 3.85 4.58 4.90 5.80 6.20

5.20 3.58 5.33 28.38 2.22 0.23 7.05 0.45 30.71 0.29 31.00 0.44 0.67 0.92 1.26 1.65 2.39 3.58 4.98 7.87 8.69 10.99 16.25

1.27 0.72 1.54 2.39 3.10 0.24 11.42 0.47 9.57 0.10 9.68 0.29 0.30 0.32 0.33 0.52 0.62 0.72 0.97 1.44 1.71 2.98 4.89

* MetHB level on Day 1 after PQ treatment completion (i.e., Day 4 of the follow-up). MetHB level on Day 15 after PQ treatment completion (i.e., Day 18 of the follow-up).

were 1.27 1.54% and 15% individuals had > 2% MetHB levels (Table 2). MetHB-1 levels were 6.01 6.88% in Group I, 5.29 6.17% in Group II, and 4.84 4.13% in Group III (P[K-W] = 0.934; Table 3). MetHB-1 > 4% was observed in 4650% of subjects in all groups; meanwhile, MetHB-15 > 4% was observed only in 49% of subjects from all groups. There were no differences in terms of MetHB levels and their body weights between male and female subjects [MetHB-1: P(t) = 0.061; MetHB15: P(t) = 0.252]. Overall comparison in all groups confirmed mean MetHB-1 in individuals < 60 kg of 5.59 5.08% (95% CI, 4.266.93), which was similar (4.80 5.71%, 95% CI, 3.21 6.39) to those > 60 kg [P(F) = 0.443]. Likewise, when values were compared within the groups, the similarity was confirmed (P > 0.112). The proportion of overweight (> 60 kg) individuals was 4059% in the different groups (P = 0.184776). When analysis was made based on nine strata using the PQ dose

and body weight, no significant association was observed between these variables and MetHB-1 or MetHB-15 levels (Table 4). Comparison between MetHB-1 and MetHB-15 confirmed a significant reduction within each group and in all groups according to Wilcoxon test (P = 0.0000.001). Because no significant differences in MetHB levels were detected between the three treatments, we further studied MetHB and G6PD correlations in all treated subjects. The results are as follows: 1) MetHB-1 as function of G6PD: rho (Spearman) = 0.087 and P = 0.363; 2) MetHB-15 as function of G6PD: rho (Spearman) = 0.035 and P = 0.723. These data indicated that the correlation between MetHB and G6PD at the different time points appeared inverse but not statistically significant. However, correlation between the two MetHB readings was positive, moderate, and significant [rho (Spearman) = 0.466; P = 0.000]. The lineal model Y = a + bX for regression is MetHB-15 = 0.548 + 0.127 (MetHB-1). Both coefficients (a and b) were significant in the model (P = 0.004 and P = 0.000; Figure 1). Adverse events. Before treatment, frequencies of individuals from Group III reporting adverse events were as follows: muscle/bone pain, 83%; anorexia, 65%; nausea, 59%; abdominal pain, 57%; vomiting, 32%; cough, 32%; dyspnea, 17%; diarrhea, 11%; odynophagia, 11%; pruritus, 6%. In addition, in this group, the physician observed conjunctivae or palmar pallor in 25%, abdominal tenderness in 29%, hepatomegaly in 28%, jaundice in 3%, choluria in 3%, and splenomegaly in 3%. After treatment, the more frequent signs-symptoms in Group III were abdominal pain (including epigastralgia) followed by pruritus and diarrhea (Table 5). Meanwhile, in Group I, these were pruritus, dizziness, and abdominal pain (including epigastralgia). Comparison of the type and frequency of adverse effects between Group III and a reference group reported elsewhere14 (administered the daily standard and the total standard dose of PQ at 0.25 mg/kg/d for 14 days) confirmed the presence of significant differences (Table 5). Furthermore, comparison of the frequency of the main reported events confirmed major differences between Group III and the reference group but not between Group III and Group I. These findings might suggest that adverse effects associated with high daily doses are different in type and frequency from those observed at standard dose. The

Percentiles

Table 3 G6PD activities and MetHB levels in P. vivax malaria subjects treated with different doses of PQ
95% CI (mean)* Variable PQ group (mg/kg/d) Mean (SD/SE) Lower Upper Range

G6PD (UI/g HB) MetHB-1 (%) MetHB-15 (%)

I (0.58) N = 26 II (0.83) N = 28 III (1.17) N = 65 I (0.58) N = 24 II (0.83) N = 28 III (1.17) N = 60 I (0.58) N = 23 II (0.83) N = 27 III (1.17) N = 55

4.24 (1.23/0.24) 4.22 (1.11/0.21) 3.57 (1.01/0.13) 6.01 (6.88/1.40) 5.29 (6.17/1.17) 4.84 (4.13/0.53) 1.65 (2.23/0.46) 1.17 (1.31/0.25) 1.16 (1.29/0.17)

3.74 3.79 3.32 3.10 2.90 3.77 0.68 0.66 0.81

4.74 4.65 3.83 8.92 7.69 5.90 2.61 1.69 1.51

2.506.40 2.806.50 2.307.00 0.2928.60 0.3231.00 0.3916.67 0.289.68 0.105.24 0.307.79

Homogeneity variance tests for the different PQ treatments: 1) G6PD P = 0.124; 2) MetHB-1 P = 0.188; MetHB-15 P = 0.057. Variances were statistically similar. 2) G6PD P(K-W) = 0.004; P(K-W) = 0.934; MetHB-15 P(K-W) = 0.539. According to sex, medians of the variables were similar (K-W test): G6PD (P = 0.442), MetHB-1 (P = 0.090), and MetHB-15 (P = 0.560). * 95% confidence interval for mean. MetHB level on Day 1 after PQ treatment completion. MetHB level on Day 15 after PQ treatment completion.

METHEMOGLOBINEMIA AND PRIMAQUINE

191

Table 4 MetHB Levels in P. vivax subjects after treatment with PQ

Dose (mg) Weight range (kg) Stratum Number Mean MetHB predicted* Mean MetHB-1 observed (SD) Mean MetHB-15 observed (SD)

1.17 1.17 1.17 0.83 0.83 0.83 0.58 0.58 0.58

< 55 5565 > 65 < 55 5565 > 65 < 55 5565 > 65

High-low High-medium High-high Medium low Medium-medium Medium-high Low-low Low-medium Low-high

21 6 5 20 12 10 18 8 7

9 8 7 6 5 4 3 2 1

6.15 (4.12) 5.67 (2.86) 5.78 (5.96) 4.15 (3.83) 5.86 (8.60) 3.32 (2.74) 4.26 (4.39) 3.46 (4.18) 9.77 (10.84) P(K-W) = 0.369896

1.09 (0.88) 1.62 (1.46) 1.38 (1.32) 1.56 (2.07) 1.34 (1.61) 1.03 (0.97) 0.90 (0.65) 0.69 (0.43) 2.74 (3.73) P(K-W) = 0.918226

* 1 (minimum) 9 (maximum).

recorded events were frequent in individuals from Group I and very frequent in those from Group III (Table 5), predominantly presented as jaundice, splenomegaly, palm pallor, insomnia, and orthostatism. Moreover, in this group, 17% of the adverse effects were diagnosed by the physician as definitively associated with PQ treatment, 15% were probably associated, 28% were possibly associated, 9% were unlikely associated, and 31% were not associated. In all groups, a severe adverse event was never reported, and all events were relieved in 99% of individuals in 28 days after treatment. Treatment of pruritus was required in 2.4% cases. Adverse events were observed in at least 89% subjects before 48 hours after the completion of treatments. Most signs and symptoms observed before and after treatment were analogous. Random blood tests performed on 20 adult subjects from any group after 24 hours of PQ treatment were within normal ranges: hemoglobin, 11.7 1.3 g/dL; hematocrit, 35.1 3.9%; total white cell count, 7,700 1,200/L; alanine aminotransferase, 33.0 28.0 U/L; total bilirubin, 0.60 0.30 mg/dL; direct bilirubin, 0.20 0.13 mg/dL; creatinine, 0.74 0.12 mg/dL. Urine analysis was normal. MetHB-1 values of 28.6% and 31% were detected in two subjects; their MetHB-15 values were 7.8% and 9.7%, respectively. In these subjects, no clinical manifestation was observed.

DISCUSSION This study aimed at clarifying the effect of PQ on MetHB formation when administered at higher than standard daily doses in P. vivax malaria subjects in Colombia. We examined the effect on subjects at 1 and 15 days after malaria treatment with a standard CQ dose (10 mg/kg on Day 1 and 7.5 mg/kg on Days 2 and 3) plus a 3-day course of PQ at 0.58, 0.83, or 1.17 mg/kg/d. We confirmed that, in subjects with normal G6PD activity, levels of MetHB 24 hours after administration of PQ at 5-fold the daily standard dose (1.17 mg/kg/d for 3 days) were < 20% in 98% of the subjects. Among the few individuals who exhibited > 20% MetHB, clinical signs associated with methemoglobinemia were absent. Several authors have reported that clinical cyanosis was observed when MetHB reaches ~15%2 or 1520 g/L,29 but this could not be confirmed in this study, even though exhaustive exploration to detect these signs were carried out by the clinician. Therefore, our findings might be more close to the reports made by other authors pointing out the fact that the threshold for clinical signs and symptoms of methemoglobinemia is > 20%.6 A rise of MetHB levels of up to 4% when PQ is used at standard doses have been reported in healthy individuals.30 However, the expression standard dose refers not only to the total PQ dose but also to the time when the treatment is administered. In our study, one of the groups (Group III) was administered the total standard dose in 3 days instead of the 14 days currently recommended in Colombia. MetHB-1 at > 4% was detected in ~50% of the individuals in all three groups regardless of the difference of the total doses of PQ treatment. It was expected that administration of PQ at high doses during a short period would result in higher toxicity. All subjects in this study received higher than standard daily doses. We failed to confirm a direct relationship between PQ administration and MetHB levels [mean MetHB-1 = 5.20 5.33% P(K-W) = 0.934 and MetHB-15 = 1.27 1.54 P(K-W) = 0.539]. PQ has been reported to be as well tolerated as placebo, with low toxicity,8 even with prolonged administration of a total dose of 30 mg/d over a period of 1652 weeks.3135 Similarly, 60 mg/d also exhibited good tolerance,36,37 and high sensitivity to the PQ treatments has been mainly reported in MetHB reductasedeficient individuals.6,38 The affect of PQ on MetHB induction at higher than standard doses (> 210 mg; e.g., 315 or 420 mg), when administered in < 14 days (e.g., 5 or 10 days), has not been studied, but tolerance has been reported as adequate when administered with food.8

Figure 1. Regression analysis for MetHB-15 in MetHB-1 levels in peripheral blood. MetHB-15 = 0.548 + 0.127 (MetHB-1). r = 0.45458; adjusted r (five extreme values are excluded) = 0.59763.

192

CARMONA-FONSECA AND OTHERS

Table 5 Adverse events after primaquine treatments

Standard treatment Event 0.25 mg/kg/d for 14 days* Non-standard treatments 1.17 mg/kg/d for 3 days (Group III) 0.58 mg/kg/d for 3 days (Group I) P (c2)

Spontaneous abdominal pain Pruritus Diarrhea Jaundice* Vomit* Anorexia Splenomegaly* Insomnia* Dizziness Nausea Sweating Herpes labialis* Orthostatism* Conjunctival pallor Palm pallor*
Values are percentages. * Absent in Group I (0.58 mg/kg/d for 3 days). Simultaneous comparison of the three groups. Absent in the standard group (0.25 mg/kg/d for 14 days). Calculation was not done because of presence of values < 1.

5.8 7.6 8.1 0 6.4 3.5 0 0 7.0 7.6 11.0 0 0 8.7 0

17.8 16.8 9.9 6.9 6.9 5.9 5.9 5.9 4.0 4.0 4.0 3.0 3.0 3.0 3.0

10.2 15.3 6.8 0 0 6.8 0 0 11.9 5.1 5.1 0 0 0 5.1

0.006799 0.046654 0.774967 0.000201 0.125640 0.487884 0.000923 0.000923 0.171431 0.456061 0.075146 0.016433 0.023435

As higher PQ doses are administered, higher MetHB in individuals with lower body weight may be expected. In this study, subjects in all three groups had a mean body weight of 5762 kg; however, individuals > 60 kg were more frequent in Group I (lowest daily and total PQ) and less frequent in Group III (highest daily and total PQ dose) than in Groups I and II. The highest MetHB-1 was detected in Group I and the lowest in Group III, whereas the proportion of individuals over the 4% threshold for MetHB-1 was similar in all groups. In any case, the highest MetHB-1 levels were observed in individuals weighing < 60 kg, but no statistical difference could be confirmed between the effect of any of the three treatments and body weight. Our results confirmed that effects associated with high daily PQ dose were different in type and frequency compared with administration of the standard dose. Such events were frequent in Group I but predominant in Group III who were given the highest daily and total PQ dose. Clinical findings such as jaundice, choluria, splenomegaly, and palm pallor were more frequent in Group III, and they might be associated with hemolysis after PQ treatment79; however, in 20 randomly selected individuals after 24 hours of PQ treatment, we found no laboratory evidence of hemolysis and thus consider those clinical findings more likely linked to malaria rather than to PQ. The results of this study showed that the adverse events with 2- to 5-fold higher daily PQ were mild, short-lived, and without long-term side effects. There was no treatment termination required during the study. Furthermore, < 5% of subjects required anti-histamine preparations for pruritus, and no other reactions required specific treatment. Taken together, this evidence confirms than PQ was well tolerated at high daily doses. Efficacy against recurrences of PQ at a total standard dose during shorter periods (< 14 days) remains to be studied in detail. Our findings and the cumulative evidence confirm that it is feasible to administer PQ in < 14 days whenever required for research or clinical purposes but only after previous fulfillment of three simple criteria: normal G6PD activity, in

non-pregnant subjects, and with a light meal. Under these conditions, the major adverse effect after PQ administration, serious methemoglobinemia, is rare, but adverse events at even the highest daily PQ dose (mainly at 5-fold standard dose) were short-lived, mild, and without long-term side effects.
Received December 5, 2007. Accepted for publication August 20, 2008. Acknowledgments: The authors thank the directors of the local hospitals at Turbo and El Bagre for support, the volunteers for collaboration, the Grupo Malaria personnel in the field for technical support, and Dr. Roubing Wang for reviewing the manuscript. Financial support: This work was supported by Universidad de Antioquia, Colciencias (Code 1115-04-16497; contract RC-253-20043) and Direccin Seccional de Salud de Antioquia (DSSA). Authors addresses: Jaime Carmona-Fonseca and Amanda Maestre, Grupo Grupo Salud y Comunidad, Universidad de Antioquia, Calle 62 # 5259, Medelln, Colombia, Tel: 57-4-2196000, Fax: 57-4-2196487. Gonzalo lvarez, Grupo Malaria Universidad de Antioquia, Calle 62 # 5259, Medelln, Colombia, Tel: 57-4-2196486, Fax: 57-4-2196487. Reprint requests: Amanda Maestre, Universidad de Antioquia, Calle 62 # 5259, Medelln, Colombia. E-mail: aemaestre@quimbaya.udea .edu.co.

REFERENCES
1. van Heijst NP, Metahemoglobinemia. IPCSIntox databank. Available at: http://www.intox.org/databank/documents/treat/ treats/trt40_s.htm. Accessed November 1, 2007. 2. Hall AH, Kulig KW, Rumack BH, 1986. Drug- and chemicalinduced methaemoglobinaemia. Clinical features and management. Med Toxicol 1: 253260. 3. Toxicologia.net, ndice de txicos. 16. Metahemoglobinizantes. Available at: http://wzar.unizar.es/stc/toxicologianet/pages/x/ x16/01.htm. Accessed November 1, 2007. 4. Bradberry SM, 2003. Occupational methaemoglobinaemia. Mechanisms of production, features, diagnosis and management including the use of methylene blue. Toxicol Rev 22: 1327. 5. Medline-plus, Informacin en salud para todos. Enciclopedia mdica en espaol. U.S. National Library of Medicine; National Institutes of Health, Department of Health and Human Services. Available at: http://www.nlm.nih.gov/medlineplus/

METHEMOGLOBINEMIA AND PRIMAQUINE

193

6. 7.

8. 9. 10. 11. 12.

13.

14.

15. 16. 17. 18. 19.

20.

21.

22.

spanish/ency/article/000562.htm#Nombres%20alternativos. Accessed November 1, 2007. Coleman MD, Coleman NA, 1996. Drug-induced methaemoglobinaemia. Treatment issues. Drug Saf 14: 394405. Fonseca D, Mateus H, Silva C, Contreras N, Restrepo C, 2005. Deficiencia de glucosa 6-fosfato deshidrogenasa. Aspectos generales de la eritroenzimopata ms frecuente en el mundo. Acta Med Colomb 3064: 5964. Clayman CB, Arnold J, Hochwald RS, Yount EH, Edgcomb JH, Alving AS, 1952. Toxicity of primaquine in caucasians. JAMA 149: 15631568. Hockwald RS, Arnold J, Clayman CB, Alving MD, 1952. Toxicity of primaquine to negroes. J Am Med Assoc 149: 15681570. Frank JE, 2005. Diagnosis and management of G6PD deficiency. Am Fam Physician 72: 12771282. Luzzatto L, 2006. Glucose 6-phosphate dehydrogenase deficiency: from genotype to phenotype. Haematologica 91: 13031306. Santana MS, Rocha MA, Arcanjo AR, Sardinha JF, Alecrim WD, Alecrim MD, 2007. Associao de metemoglobinemia e deficincia de glicose-6-fosfato desidrogenase em pacientes com malria tratados com primaquina. Rev Soc Bras Med Trop 40: 533536. Carmona-Fonseca J, lvarez G, Blair S, 2006. Malaria por Plasmodium vivax: curacin del ataque agudo con tres dosis diferentes de primaquina y dosis fija de cloroquina. Antioquia, Colombia, 20032004. Biomedica (Bogota) 26: 353365. lvarez G, Pieros JG, Tobn A, Ros A, Maestre A, Blair S, Carmona-Fonseca J, 2006. Efficacy of three chloroquineprimaquine regimens for treatment of Plasmodium vivax malaria in Colombia. Am J Trop Med Hyg 75: 605609. Beutler E, Duparc S, the G6PD Deficiency Working Group, 2007. Glucose-6-phosphate dehydrogenase deficiency and antimalarial drug development. Am J Trop Med Hyg 77: 779789. Arnold J, Alving AS, Clayman CB, 1961. Induced primaquine resistance in vivax malaria. Trans R Soc Trop Med Hyg 55: 345350. Collins WE, Jeffery GM, 1996. Primaquine resistance in Plasmodium vivax. Am J Trop Med Hyg 55: 243249. Schwartz E, Regev-Yochay G, Kurnik D, 2000. Short report: a consideration of primaquine dose adjustment for radical cure of Plasmodium vivax malaria. Am J Trop Med Hyg 63: 393395. Signorini L, Matteelli A, Castelnuovo F, Castelli F, Oladeji O, Carosi G, 1996. Short report: primaquine-tolerant P. vivax in an Italian traveler from Guatemala. Am J Trop Med Hyg 55: 472473. Goller JL, Jolley D, Ringwald P, Biggs BA, 2007. Regional differences in the response of Plasmodium vivax malaria to primaquine as anti-relapse therapy. Am J Trop Med Hyg 76: 203207. Lalloo DG, Shingadia D, Pasvol G, Chiodini PL, Whitty CJ, Beeching NJ, Hill DR, Warrell DA, Bannister BA; HPA Advisory Committee on Malaria Prevention in UK Travellers, 2007. UK malaria treatment guidelines. J Infect 54: 111121. Lwanga SK, Lameshow S, 1991. Determinacin del Tamao Muestral en los Estudios Sanitarios. Ginebra: OMS.

23. Carmona-Fonseca J, 2003. La malaria en Colombia, Antioquia y las zonas de Urab y Bajo Cauca: panorama para interpretar la falla teraputica antimalrica. Parte 1. Iatreia 16: 299318. 24. Carmona-Fonseca J, 2004. La malaria en Colombia, Antioquia y las zonas de Urab y Bajo Cauca: panorama para interpretar la falla teraputica antimalrica. Parte 2. Iatreia 17: 3453. 25. Beutler E, 1967. Glucose-6-phosphate dehydrogenase deficiency. Diagnosis, clinical and genetic implications. Am J Clin Pathol 47: 303311. 26. World Health Organization W.H.O., 2000. Severe falciparum malaria. Trans R Soc Trop Med Hyg 94: 190. 27. Menndez-Capote R, Daz-Prez L, Luzardo-Surez C, 1997. Hemlisis y tratamiento con primaquina. Informe Preliminar. Rev Cubana Med Trop 49: 136138. 28. Ude A, 1993. Manual de Toxicologa. Medelln: Universidad de Antioquia. 29. Rodrguez-Fernndez A, Del Pozo-Hessing C, 2005. Intoxicacin por cianuro. Medisan 5: 103107. 30. Fletcher KA, Evans DA, Gilles HM, Greaves J, Bunnag D, Harinasuta T, 1981. Studies on the pharmacokinetics of primaquine. Bull World Health Organ 59: 407412. 31. Baird JK, Lacy MD, Basri H, Barcus MJ, Maguire JD, Bangs MJ, Gramzinski R, Sismadi P, Krisin, Ling J, Wiady I, Kusumaningsih M, Jones TR, Fryauff DJ, Hoffman SL; United States Naval Medical Research Unit 2 Clinical Trials Team, 2001. Randomized, parallel placebo-controlled trial of primaquine for malaria prophylaxis in Papua, Indonesia. Clin Infect Dis 33: 19901997. 32. Fryauff DJ, Baird JK, Basri H, Sumawinata I, Purnomo, Richie TL, Ohrt CK, Mouzin E, Church CJ, Richards AL, 1995. Randomised placebo-controlled trial of primaquine for prophylaxis of falciparum and vivax malaria. Lancet 346: 11901193. 33. Soto J, Toledo J, Rodriquez M, Sanchez J, Herrera R, Padilla J, Berman J, 1998. Primaquine prophylaxis against malaria in nonimmune Colombian soldiers: efficacy and toxicity. A randomized, double-blind, placebo-controlled trial. Ann Intern Med 129: 241244. 34. Soto J, Toledo J, Rodriquez M, Sanchez J, Herrera R, Padilla J, Berman J, 1999. Double-blind, randomized, placebo-controlled assessment of chloroquine/primaquine prophylaxis for malaria in nonimmune Colombian soldiers. Clin Infect Dis 29: 199201. 35. Weiss WR, Oloo AJ, Johnson A, Koech D, Hoffman SL, 1995. Daily primaquine is effective for prophylaxis against falciparum malaria in Kenya: comparison with mefloquine, doxycycline, and chloroquine plus proguanil. J Infect Dis 171: 15691575. 36. Clyde DF, McCarthy VC, 1977. Radical cure of Chesson strain vivax malaria in man by 7, not 14, days of treatment with primaquine. Am J Trop Med Hyg 26: 562563. 37. Wilairatana P, Silachamroon U, Krudsood S, Singhasivanon P, Treeprasertsuk S, Bussaratid V, Phumratanaprapin W, Srivilirit S, Looareesuwan S, 1999. Efficacy of primaquine regimens for primaquine-resistant Plasmodium vivax malaria in Thailand. Am J Trop Med Hyg 61: 973977. 38. Cohen RJ, Sachs JR, Wicker DJ, Conrad ME, 1968. Methemoglobinemia provoked by malarial chemoprophylaxis in Vietnam. N Engl J Med 279: 11271131.