The study was conducted between July and October 2001 in Purworejo District in the Menoreh Hills near the southern coast of Central Java, a region experiencing resurgent malaria at that time. The social, economic and demographic characteristics of malaria in the region are described elsewhere 478. Falciparum and vivax malaria (approximately 1:1) occurs among all ages with prevalence in high-risk areas ranging from 5% to 38%. Among 40 and 54 subjects treated with CQ or SP for uncomplicated falciparum malaria in this region, 47% and 22%, respectively, developed recurrent parasitaemia within 28 days 4. During the close study follow up, none of the treatment failures developed complicated or severe malaria.

Subject enrollment is depicted in Figure 1. Subjects with uncomplicated falciparum malaria were recruited, in part, by mass blood survey and by passive case detection (PCD) at local government-run clinics. This was necessary because few adults in this study area suffered from symptomatic malaria. After informed consent, all subjects were screened by medical history, physical examination, semi-quantitative glucose-6-phosphate dehydrogenase (G6PD) assay (G-6-PDH Screening Test 203-A, Sigma Diagnostics^®, St. Louis, MO USA), and if female of child-bearing potential, a urine human chorionic gonadotropin (hCG) test (TestPack^® +Plus™ hCG Urine, Abbott, USA). Inclusion criteria included age ≥ 15, asexual stage P. falciparum parasite density ≥ 400/μl on Giemsa thick smear examined by standard microscopy, and availability for 28-day follow-up. Exclusion criteria included pregnancy, breast feeding, body weight < 40 kg, G6PD deficiency, history of antimalarial or antibiotic consumption during the previous 7 days, severe or complicated malaria, history of allergy or adverse reaction to study medications, and P. vivax or mixed species infection. Any parasitemic potential study subject not meeting these criteria was referred back to the cooperating clinic for standard therapy. Those patients deemed to meet enrollment criteria and willing to participate were assigned a sequential study subject code by the screening physician. The study subject codes were pre-assigned to treatment arms by a random process and all treatment was pre-packaged accordingly.

Subjects were randomized to one of four treatment arms: (A) CQ 25 mg base/kg (body weight) over three days, (B) CQ 25 mg/kg over three days + SP 25/1.25 mg/kg single dose, (C) CQ 25 mg/kg over three days + SP 25/1.25 mg/kg single dose + primaquine 45 mg single dose on day 0, or (D) CQ 25 mg/kg over three days + SP 25/1.25 mg/kg single dose + primaquine 45 mg single dose on day 2. Study investigators directly observed and documented administration of each dose of medication: CQ (Resochin^®; chloroquine diphosphate tablets 150 mg base; PT Bayer, Indonesia) in three once daily doses of 10, 10 and 5 mg base/kg on days 0,1, and 2 respectively; SP (Fansidar^® tablets; 500 mg sulfadoxine/25 mg pyrimethamine; Hoffman La Roche, Indonesia) in a single dose on day 0; Primaquine phosphate (generic tablets, 15 mg base; PT Kimia Farma, Bandung, Indonesia) in a single dose of 3 tablets on day 0 or day 2.

Subjects were followed for 28 days after initiating therapy. Study personnel visited subjects on 10 separate days (1, 2, 3, 4, 7, 11, 14, 18, 21 and 28) to assess symptoms, clinical recovery, adverse events and to obtain finger-stick blood samples for preparation of Giemsa-stained blood smears. Subjects complaining of illness on any day of follow-up were immediately brought to clinic and evaluated. Blood blot specimens (100 – 200 μl) were also prepared on days 0, 2, 7, 14, 21, 28 or day of recurrent parasitaemia from fingerstick blood samples collected by heparinized, 100 μl micro-capillary tubes and expelled onto Whatman No. 1 filter paper (Whatman International, United Kingdom) for drying and storage for later analysis.

The microscopists in this study were certified as competent by a standardized, proctored examination. Microscopists examined all ocular fields of standard Giemsa (Sigma, USA) stained thick smears at high power (1000× oil immersion) for asexual and sexual forms of P. falciparum and other plasmodial species, and recorded sexual and asexual parasite densities as number of parasites per 200 white blood cells. Parasite densities are reported as parasites/μL assuming an average white blood cell count of 8,000/μl.

Msp-2 genotyping was performed on blood blot samples from day 0 and day of recurrent P. falciparum parasitaemia following Felger et al. 9. Recrudescence was defined by the presence of the same genotype on day 0 and day of recurrent parasitaemia while a new infection (i.e. reinfection) was defined by the presence of a different genotype on day 0 and day of recurrent parasitaemia.

Therapeutic efficacy (parasitological cure) was defined as the proportion of subjects reaching day-28 without recrudescence. Therapeutic failures occurred when asexual parasitaemia increased between day 0 and day 2, failed to clear by day 4, or recurred between days 4 and 28. These data were used to determine the crude therapeutic efficacy of each regimen. Infections initially classified as therapeutic failures but later proven to be reinfections by msp-2 genotyping were reclassified as therapeutic successes with abbreviated follow-up (until the point of censure). These results were used to determine the "adjusted" efficacies of CQ and CQ + SP. EpiInfo 2000 1.1.2 (CDC, Atlanta USA) and SPSS version 9.0 (SPSS, Inc. USA) were used to calculate confidence intervals for odds ratios (OR) and the 95% confidence interval around those ratios, and for estimating the p-value using the Yates' corrected or Fisher's exact test when appropriate. Cumulative incidence of therapeutic failure was estimated by actuarial analysis as described elsewhere 10.

Among 124 parasitaemic persons identified during mass blood screening and passive case detection from outpatient clinics, 117 were enrolled and randomized to one of the 4 treatment regimens. Seven subjects did not meet study eligibility criteria and were returned to the referring clinic for treatment for the following reasons: taking quinine (1), pregnancy (1), G6PD deficiency (1), low body weight (2), declined participation (1) and administrative error (1).

Among the 117 subjects enrolled and randomly assigned, no statistically significant demographic differences between the four study arms were identified; age ranged from 16 to 65 years, and the male to female ratio was 1.2:1 (Table 2). No statistically significant differences were noted between treatment arms with regards to fever rates (range 24–41%) or geometric mean asexual parasite densities (range 653-1465/μL) before enrollment.

There were no significant differences in parasite or fever clearance times among the 4 treatment groups. Virtually all (115 of 117) subjects cleared parasitaemia by day 3 regardless of the therapy assignment, and all 38 subjects with fever prior to therapy had a normal axillary temperature by day 2 (Figure 2).

Asexual parasitemia outcomes were similarly efficacious among the three groups receiving (CQ+SP, CQ+SP+PQ0, and CQ+SP+PQ2), so these groups were combined specifically for comparison with CQ monotherapy against asexual blood stages.

Among the 29 subjects treated with CQ alone, 1 (3%) withdrew from the study after day 18 and 3 (10%) developed intercurrent P. vivax during the second and third week of post-therapy observation. Fourteen (48%) completed the study without recurrent parasitaemia. The remaining 11 (38%) subjects developed recurrent P. falciparum during the 28-day follow-up. The crude cumulative incidence of failure of CQ therapy for uncomplicated P. falciparum was 42% (Table 3); therefore, the crude therapeutic efficacy was 58%.

Among the 88 subjects treated with CQ + SP, six (7%) withdrew before day 28 (three declined further participation between days two and 21 and three moved away from the study site between days one and 18). One subject (1%) developed intercurrent P. vivax (day 22). Seventy-six (86%) completed the study without recurrent parasitaemia. The remaining 5 (6%) subjects developed recurrent P. falciparum during follow-up, therefore providing a crude therapeutic efficacy of 94%. The crude 28-day cumulative incidence of failure of CQ + SP therapy was 6% (Table 3). The relative risk of treatment failure with CQ compared to CQ + SP among those who completed the 28-day observation period was 7.13 (95% CI 2.74 – 18.57, p = 0.00003).

Three of 11 apparent therapeutic failures among subjects receiving CQ alone could not be evaluated because P. falciparum DNA could not be amplified for msp-2 genotyping, leaving eight evaluable recurrent infections. One case of day 28 recurrent parasitaemia was reclassified as a cure with late reinfection based on differing day 0 and day of recurrence msp-2 genotypes. From the remaining seven recurrent infections, the parasite genotypes matched and their classification as recrudescence affirmed. The adjusted cumulative incidence of failure of CQ therapy for uncomplicated P. falciparum was therefore 30% (therapeutic efficacy 70%) (Table 4).

In the combined CQ + SP groups, one of five recurrent parasitaemias was excluded from analysis because parasite DNA could not be amplified, leaving four evaluable recurrent infections. Based on discordant day 0 and day of recurrence msp-2 genotypes, three apparent therapeutic failures were reclassified as successful cures with reinfection. The 28-day cumulative incidence of failure of CQ+SP therapy for uncomplicated P. falciparum was therefore only 1% (therapeutic efficacy = 99%) (Table 4). Subjects treated with CQ alone were 25 times more likely to suffer therapeutic failure compared to CQ combined with SP (95% CI 3.3 – 196.06; p = 0.0006).

Figure 3 illustrates the course of gametocyte rates among groups after treatment. Comparing the two groups receiving primaquine, gametocyte rates declined steadily to 0% by day 11 after therapy and did not reappear during follow-up. Gametocyte rates declined more slowly in the non-primaquine groups. The difference in gametocyte clearance rates on day 11 between groups receiving primaquine (0%) and those not (33%) was significant (p = 0.025). A single dose of primaquine greatly improved the clearance rate of gametocytes, and administering the dose on day 2 vice day 0 accelerated clearance time (0% vs. 7% on day 7); however this difference was not statistically significant.