© 2005 American Thoracic Society doi: 10.1164/rccm.2504002
Can Studies of the Early Bactericidal Activity of Rifapentine Tell Us How to Prevent Acquired Rifamycin-Resistant Relapse?Pharmaceutical Product Development, Washington, DC Rifampin occupies a critical role in tuberculosis (TB) therapy, permitting treatment to be shortened while reducing the risk of relapse (1). Rifapentine differs from its parent compound by virtue of its long plasma half-life and superior in vitro activity, which together increase by severalfold the time that its plasma concentration remains above minimal inhibitory concentration (2). Such characteristics ostensibly should make rifapentine an ideal component of highly intermittent regimens for the continuation phase of TB treatment. Yet, two large studies have failed to realize this promise, with both finding higher rates of drug-sensitive relapse, and one reporting a new phenomenon, acquired rifamycin-monoresistant relapse, in AIDS/TB cases with markedly low isoniazid levels (3–5). Similar relapses have recently been found to occur when rifabutin is administered intermittently, although in that circumstance, they have done so in association with low levels of rifabutin rather than isoniazid (6). In the current issue of the Journal (pp. 128–135), Sirgel and colleagues (7) report a dose-ranging study comparing the early bactericidal activities of rifapentine and rifampin, hoping to elucidate why rifapentine fails and whether higher doses might improve its efficacy. The main findings of the study were that, at equal doses, rifampin and rifapentine do not differ with respect to early bactericidal activity, and that increasing the rifapentine dose increases its early bactericidal activity, although the increment is disproportionately small. What conclusions may we draw from these findings? Relapse of drug-sensitive TB is believed to reflect reactivation of metabolically dormant, nonreplicating foci of infection contained within granulomas or other poorly vascularized tissues. The drug susceptibility of these relapses indicates the relative lack of effect of current TB drugs against this bacillary subpopulation, either because of poor penetration into the tissues in which these dormant bacilli reside, or inadequate molecular expression of the targets of current drugs under the conditions of dormancy. Early bactericidal activity, which reflects the ability of administered chemotherapy to reduce viable sputum counts during the first few days of therapy, is unrelated to the ability to prevent these relapses. Isoniazid and ofloxacin perform very well in studies of early bactericidal activity, whereas pyrazinamide is inactive, and rifampin, intermediate (8, 9). These findings have been interpreted as indicating that early bactericidal activity mainly reflects drug effects against extracellular, metabolically active, replicating bacilli. One is reluctant to assume that measures of early bactericidal activity can adequately guide the dosing of a drug such as rifampin that is used primarily for its effects on dormant bacilli, particularly because so little is known regarding either the biology of this subpopulation or the mechanism of its elimination by current chemotherapy. The absence of data in this regard reflects a major unmet need in TB clinical research—that is, the development of validated surrogate markers to rapidly assess the sterilizing potential of new drugs and the risk of relapse of individual patients. Testing of candidate surrogate markers should be included in all clinical trials of rifapentine and other potentially sterilizing new drugs for TB, so that future trials may be accelerated. However, the main issue at hand is not relapse in general but rather the specific question of acquired rifamycin-resistant relapse. As the authors correctly point out, the reduction in the bacillary burden afforded by 8 weeks of multidrug therapy makes it very unlikely that any rifampin-resistant bacilli remain viable in vivo when rifapentine treatment is started. Rifamycin-resistant relapses must therefore reflect resistance arising de novo because of replication under the selective pressure of antibiotics, in a distinct bacillary microenvironment into which oxygen, nutrients, and anti-TB chemotherapy can penetrate, at least partially. Early bactericidal activity is an appropriate tool to assess drug effects against this subpopulation. Moreover, the concept that rifamycin-resistant relapse may be prevented by increasing the rifapentine dose is an intriguing one. Such relapses certainly would be prevented if one could sufficiently increase the dose to maintain free drug levels above its minimal inhibitory concentration throughout the dosing interval. Unfortunately, the present data indicate that increasing the rifapentine dose to 1,200 mg might extend this period to 66 hours, but that subsequently one would be left depending on postantibiotic effect to cover the time remaining until the next weekly dose. Our ability to measure events in vivo near the end of the dosing interval is limited, because studies of early bactericidal activity presently end 2 days before patients are due to receive their next weekly rifapentine dose. Animal experiments to determine the extent of bacillary regrowth near the end of the dosing interval may help tell us whether trust in postantibiotic effect to contain such regrowth is warranted. Lacking these data, further studies of once-weekly rifapentine in AIDS/TB cases should not be considered, given the potential consequences of acquired resistance. FOOTNOTES Conflict of Interest Statement: R.S.W. does not have a financial relationship with any commercial entity that has an interest in the subject of this manuscript. REFERENCES
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
