Department of Medical Statistics, Leiden University Medical Center, Leiden,1 National Institute of Public Health and the Environment, Bilthoven,2 Royal Netherlands Tuberculosis Association (KNCV), The Hague, The Netherlands,5 Pham Ngoc Thach Tuberculosis and Lung Disease Centre, Ho Chi Minh City, Vietnam,3 MRC Centre for Molecular and Cellular Biology, Department of Medical Biochemistry, University of Stellenbosch, Tygerberg, South Africa4
Received 11 September 2003/ Returned for modification 5 November 2003/ Accepted 18 February 2004
To determine the rate at which IS6110 restriction fragment length polymorphism (RFLP) patterns in Mycobacterium tuberculosis change over time, we applied a smooth nonparametric survival model to several data sets, including data from previous publications on the rate of change. The results strongly suggest a simple parametric model, with an instantaneous change at time zero and essentially a zero rate of change thereafter. Our interpretation of the results is that at the time of collection of the first isolate, more than one strain is present. We speculate that the selection of mutant strains is most likely during rapid growth, revival of the dormant bacteria, and/or adaptation to a new host. The parameter most accurately describing changing RFLP patterns is the proportion of isolates with band changes, rather than the half-life or the rate of change.
This article has been cited by other articles:
| Antimicrob. Agents Chemother. | Clin. Microbiol. Rev. |
|---|---|
| Clin. Vaccine Immunol. | ALL ASM JOURNALS |
|---|