Read original article, http://www.cdc.gov/ncidod/eid/vol6no6/king.htmRead letter by Coles et al., http://www.cdc.gov/ncidod/eid/vol7no6/coles_letter.htm

To the Editor: Drs. Coles, Liang, and Doenhoff have raised important issues regarding the emergence of praziquantel resistance in human populations. We agree that praziquantel resistance will undoubtedly emerge. In our recent modeling paper (1), we attempted to address the question of how soon such resistance will become clinically significant.

As Dr. Cole and colleagues discuss, in the future the best means for detecting resistance will be through laboratory testing of field isolates for resistance genes. In the meantime, in the absence of validated laboratory testing, analysis of ongoing clinical experience provides Schistosoma haematobium control programs some useful insight into the potential emergence of drug resistance.

Our modeling analysis of the emergence of praziquantel resistance took as its base-case the 8-year experience with treatment outcomes in an area of Kenya that had not previously been exposed to praziquantel. It was not, in fact, "a relatively small part of the population" that was treated, but rather the greater majority (75%-95% per year) of all school-aged children in the Msambweni area. Based on the uneven age distribution of S. haematobium infection (2), we estimated that 50%-75% of worms in the community were exposed to the drug during the treatment period. Sensitivity analysis allowed our model to address the implications of greater or lesser worm exposure and of greater or lesser prevalence of resistance genes.

Clearly, untreated worms in refugia would have played an important role in delaying emergence of resistance during the study period; our analysis suggests that attempts to increase community treatment coverage to 100% would have accelerated the emergence of clinically significant resistance. Similarly, a higher initial prevalence of resistance gene(s) or a faster genetic mutation rate would be predicted to hasten the onset of substantial levels of resistance. Still, on the basis of known features of parasite transmission dynamics, the effects of obligate sexual parasite reproduction and of worm clustering within human hosts were predicted to slow the emergence of resistance (on a population basis) by several years. We agree that Dr. Van Wyk's recent review on "Refugia" (3) provides a thought-provoking discussion of the effects of mass treatment of helminthic infections in a setting where drugs are not 100% effective in eradicating infection, where transmission quickly resumes, and where reinfection with resistant parasites is favored.

The title of our paper was not meant to cast doubt on the likelihood of praziquantel resistance. Instead, it was meant to point out that, under the conditions of our study, we observed no substantial praziquantel resistance and its emergence was not as "rapid" as might have been predicted. We concur that the spread of resistance will be accelerated by widespread drug usage, and we emphasize that targeted treatment has the potential advantage of prolonging the useful lifespan of a drug such as praziquantel.

The conclusion of our modeling analysis is that there may be only a 7- to 10-year period during which control projects will consistent, drug-mediated reductions in worm burden. It is essential, therefore, that planners anticipate eventual drug failure and incorporate, as part of an integrated infection-management system, nondrug interventions that will prolong drug usefulness. Prevention of transmission and not just development of newer drugs will finally provide the best form of "therapy."