[Skip to Content]
Sign In
Individual Sign In
Create an Account
Institutional Sign In
OpenAthens Shibboleth
Purchase Options:
[Skip to Content Landing]
Letters
October 25, 2000

Viral Load in Treatment With Antiretroviral Therapy and Interleukin 2—Reply

Author Affiliations
 

Stephen J.LurieMD, PhD, Senior EditorIndividualAuthorPhil B.FontanarosaMD, Executive Deputy EditorIndividualAuthor

JAMA. 2000;284(16):2055-2056. doi:10.1001/jama.284.16.2053

In Reply: We agree with Dr Aboulker that potential biases introduced by differences in treatment adherence can confound the statistical analysis of any study. However, short of serial measurements of plasma antiretroviral drug levels, there are no well-accepted surrogates for adherence beyond patient self-reporting as determined through interval history taking. In our study, we had no evidence that ART adherence was superior in the IL-2 plus ART treatment arm. Indeed, one could postulate that patients in the IL-2 plus ART arm, by virtue of their marked CD4 cell count increase, might have had less motivation to adhere to their medication schedule than the control group. Table 2 in our article presents the viral load outcomes when the analysis is restricted to just those patients reported to be taking protease inhibitors during the 90 days prior to the final branched DNA (bDNA) determination.

To directly address the concern that excluding the 11 patients for whom bDNA version 3.0 samples were not available introduced a bias, we performed the analysis of the bDNA version 2.0 data restricted to the 66 patients for whom bDNA version 3.0 data are available for comparison with the bDNA version 2.0 data from 76 patients. The resulting P value for the comparison was .09 (Table 1). The trend favoring the IL-2 plus ART arm and the P values are nearly identical to the published bDNA version 2.0 analysis. Furthermore, the relationship between unavailable frozen samples and final bDNA version 2.0 results appears to be the same in each group: 3 (43%) of 7 IL-2 plus ART patients and 2 (50%) of 4 control patients with missing bDNA version 3.0 samples had final bDNA version 2.0 results of more than 500 copies/mL. Thus, we doubt that a bias was introduced to the analysis of the version 2.0 data by excluding the 11 patients without bDNA version 3.0 samples; by extension, it is unlikely that bias was introduced into the analysis of the bDNA version 3.0 data by restricting the analysis to the same subsample.

Table. Analysis of Branched DNA Version 2.0
Data for Patients With Branched DNA Version 3.0 Samples (n = 66)*
Table. Analysis of Branched DNA Version 2.0 Data for Patients With Branched DNA Version 3.0 Samples (n = 66)*

As we noted in our article, the small viral load differences we observed between the treatment arms were statistically significant but of unknown clinical significance. These results notwithstanding, the 2 larger phase 2 trials of IL-2 (AIDS Clinical Trial Group [ACTG] 328 and Community Programs for Clinical Research on AIDS [CPCRA] 059) currently undergoing final analysis also compared virologic outcomes directly and should provide a better assessment in this regard.

×