These two independent studies document anti-malarial efficacy in 2004 from locations in the east and west of DRC, a country from which few such data have been reported previously. Efficacy of SP was poor for both sites. Efficacy of the AS-SP and AS-AQ combinations was poor in Boende but high in Kabalo. The comparison between the results of the two studies is limited by the difference in study methodology. The age and parasitaemia enrolment criteria were different and different genotypic methods were used to discriminate recrudescence from reinfection. Indeed, it is possible that the chance of being classified as a recrudescence was higher in Boende where a two-loci genotypic method was used compared to Kabalo, where a three-loci method was used. The high failure rate of SP in Boende is consistent with the molecular results for antifolate resistance, showing 61% had the triple dhfr mutation, which is known to be associated with an increased risk of treatment failure [3, 4].
The threshold of resistance to SP at which it is no longer useful to add artesunate is not well defined. In Kabalo, the day 28-failure rate of SP was close to 20%, dropping to zero when given with three days of AS. Comparable results have been found in studies from Bie province in Angola where the PCR-adjusted 28-day failure rate was reduced from 25.3% for SP alone to 1.2% after addition of AS and in Southern Benin where the PCR-corrected failure rate of AS+SP at 28 days was reduced from 44.1% to 5.3% when combining AS to SP [14, 15]. In contrast, in Boende the failure rate of SP alone was 36% and only dropped to 25% when AS was added.
It is not that surprising that in a large country, such as DRC, considerable differences in anti-malarial efficacy between provinces were observed. Site-specific patterns of treatment-seeking behaviour or malaria transmission intensity may explain geographical differences in the efficacy of anti-malarial treatments [16–18]. The higher reinfection rate and higher proportion of anaemic children in Boende suggest higher transmission intensity there.
A limitation of these studies is the disappointing external quality control results, which suggest a failure of the internal quality control system in place. In the case of the Boende study, it is important to note that for the majority of discordant results the parasitaemia detected by one of the laboratories was less than 100/μL. Also, these results affected the classification of outcome in only three cases in Boende and none in Kabalo. The delay of some months before the EQC will have resulted in a deterioration in slide quality, which may partially explain these poor results. Having an adequate standard of microscopy even in these circumstances is critical. The experience at these two sites suggests that internal quality control measures in place were not adequate to guarantee this standard. External quality control at regular intervals may be the answer but is expensive, labour-intensive and the second reading needs to be timely to avoid excessive deterioration of slides, or slides need to be mounted. Sending a random sample of slides seems an obvious approach but leads to problems when agreement is not high. It is then recommended that all the slides be sent, by which time several months may have elapsed since the study has finished.
Conducting in vivo efficacy studies in remote regions, in which there is political instability, is a major challenge. Proxy markers for resistance, such as molecular markers are invaluable tools to assess anti-malarial resistance in such a context, being less resource demanding than an in vivo study. Unfortunately, there are no markers, which reliably predict clinical resistance to drugs other than SP.
Data from neighbouring Congo-Brazzaville to the west indicate high level SP and AQ resistance [19, 20]. Efficacy of AS+AQ in studies in Burundi and Tanzania to the East in 2002 and 2004 respectively was high, although in the Tanzanian study efficacy of amodiaquine alone was evaluated but not reported [21, 22]. The observation of varying efficacy of the same anti-malarial in two different regions in the same country shows the problem of implementing a uniform national policy, especially in large country like DRC. While these results cannot be generalized to the whole provinces (Katanga or Equatorial) and even less to the whole country, they should alert the national programme to foci of resistance to AS+AQ, which became the national treatment policy in 2005. Efficacy of ACTs including partner drugs to which there is pre-existing resistance such as SP or AQ, need to be monitored at very regular intervals. According to the 2006 WHO Africa Malaria Report, eight sentinel sites are now operational in DRC .
Since these studies were performed WHO pre-qualified anti-malarials have become available, e.g. artemether-lumefantrine and AS+AQ fixed-dose combination (FDC). Dihydroartemisinin-piperaquine and the FDC of artesunate-mefloquine are in the process of being registered. Both artemether-lumefantrine and dihydroartemisinin-piperaquine have been evaluated in a number of African countries and appear highly efficacious [24, 25]. The NMCP of DRC may need to consider regional adjustments to its policy guidelines, although such an approach is challenging. DRC is participating in the World Bank Booster programme worth $US30 million over four years, which finances provision of long-lasting insecticidal nets (LLINs), intermittent preventive treatment (IPT) for pregnant women and ACT as first-line treatment of malaria http://web.worldbank.org. It is important that concomitant monitoring of anti-malarial efficacy is supported in order to inform policy makers .