Presentation is loading. Please wait.

Presentation is loading. Please wait.

Objective of the DAP A) Specify an analysis plan that can be applied to a wide variety of clinical HIV resistance studies. B) Include both Intervention.

Similar presentations


Presentation on theme: "Objective of the DAP A) Specify an analysis plan that can be applied to a wide variety of clinical HIV resistance studies. B) Include both Intervention."— Presentation transcript:

1 Objective of the DAP A) Specify an analysis plan that can be applied to a wide variety of clinical HIV resistance studies. B) Include both Intervention vs. Non-intervention Studies: The objectives of intervention studies (GART and Viradapt) are different from the objectives of non-intervention studies. C) Focus on resistance data in an experienced population. Naïve patients and experienced patients should be analyzed separately.

2 Non-intervention Studies Characterize Study population Summary of Baseline log 10 HIV-1 RNA and CD4 Summary of Prior ART experience – Not exposed or exposure less than 1 week – 1 week – 1 year – greater than 1 year Summary of duration of therapy on a HAART regimen (3 ARTs or more)

3 Endpoints: The primary endpoint is 'virological failure' by week 24 (HIV-1 RNA above threshold, e.g.400 copies/ml at week 24). For studies of duration 8 to 16 weeks, change from baseline HIV-1 RNA may be used. Determination of virological failure – (a) Exclude transient Increases in HIV-1 RNA: Not failure if week 24 value is greater than threshold, but next determination is below despite no change in therapy. – (b) Week 24 window: week 16 to week 32. For studies of duration 16 to 24 weeks, use last week with sufficient data. – (c) HIV-1 RNA failure threshold: As specified in the original protocol. If not specified, use threshold of < 400 copies/ml.

4 – (d) Patients who withdraw early from study medication: Handled in two ways:  In the DAC (‘dropout as censored') analysis, patients who withdraw early from study medication without evidence of virological failure are treated as censored and excluded. –if the last HIV-1 RNA is > threshold, score as a failure if  At least 2 HIV-1 RNA values  baseline,  Reduction from baseline is < 0.5 log10 HIV-1 RNA between weeks,  HIV-1 RNA nadir is below the threshold. –Patients on study regimen for at least 16 weeks who have no HIV-1 RNA values in week 24 window: Exclude from both DAC and DAF analyses.  In the DAF ('dropouts as failures') analysis, patients who withdraw early are counted as failures.

5 Two measures of Genotypic information will be explored. 1) The genotypic sensitivity score (GSS), is based on the number of drugs in study regimen to which the patient has genotypic sensitivity. For each drug in the regimen, the genotypic sensitivity is generally 1 if the patient’s genotype has no resistance mutations to the drug and 0 otherwise. The GSS is defined as the sum of the genotypic sensitivities over all the drugs in the regimen. 2) Three separate variables - the number of mutations in each class (PI, NRTI, NNRTI) for classes of drugs represented in the study regimen (but not limited to the drugs in the patient’s regimen).

6 Phenotypic Measures of resistance are based upon the fold-resistance Two metrics of phenotypic resistance will be explored. 1-Using the Minimum cut off for the assay  S – less than or equal to the minimum cut off for the assay  R – greater than the minimum cut off for the assay 2-Using 10-fold as a cut off  S – less than or equal to 10-fold  R – greater than 10-fold For each drug in the regimen, the phenotypic resistance score is defined as 1 if sensitive (S), or 0 if resistant (R).

7 Two measures of Phenotypic sensitivity will be explored. 1) The phenotypic sensitivity score (PSS), is based on the number of drugs in study regimen to which the patient has phenotypic sensitivity. The PSS is defined as the sum of the phenotypic sensitivities over all the drugs in the regimen. 2) Three separate variables - the number of drugs in each class (PI, NRTI, NNRTI) to which the patient has phenotypic sensitivity - for classes of drugs represented in the study.

8 Confounding variables measured prior to therapy initiation – (a) Baseline log 10 HIV-1 RNA – (b) Potent PI or NNRTI Y/N – Y if naive to PIs and PI in regimen OR if NNRTI-naïve and NNRTI in regimen; N otherwise – (c) Number of new drugs in the regimen  In ddI-naïve subjects, ddI + HU counted as ONE new drug  In ddI-experienced subjects, ddI + HU counted as 0.5 new drug  A "mini-dose" of ritonavir (i.e., 100 or 200 mg BID) should NOT be counted as a new drug

9 Method of Analysis: logistic regression of the binary response virological failure by week 24 with a common set of covariates Models for studies with Genotypic data: 6 models A – Baseline log 10 HIV-1 RNA B – New Drug Covariates (Potent PI or NNRTI, # of new drugs) C – Genotypic Sensitivity Score (GSS) D – # of PI muts, # of NRTI muts, # NNRTI muts E – Baseline log 10 HIV-1 RNA, New Drug Covariates, GSS (A, B, C) F – Baseline log 10 HIV-1 RNA, New Drug Covariates, # of PI muts, # of NRTI muts, # NNRTI muts (A, B, D)

10 Models for studies with Phenotypic data: 10 models A – Baseline log 10 HIV-1 RNA B – New Drug Covariates C – Phenotypic Sensitivity Score (PSS) - assay min cutoff D – Phenotypic Sensitivity Score (PSS) - 10-fold cutoff E – PI PSS; NRTI PSS; NNRTI PSS - assay min cutoff F – PI PSS, NRTI PSS, NNRTI PSS - 10-fold cutoff

11 G – Baseline log 10 HIV-1 RNA, New Drug Covariates, PSS - assay minimum cutoff (A, B, C) H – Baseline log 10 HIV-1 RNA, New Drug Covariates, PSS - 10-fold cutoff (A, B, D) I – Baseline log 10 HIV-1 RNA, New Drug Covariates, PI PSS, NRTI PSS, NNRTI PSS - assay min cutoff (A, B, E) J – Baseline log 10 HIV-1 RNA, New Drug Covariates, PI PSS, NRTI PSS, NNRTI PSS - 10-fold cutoff (A, B, F)

12 Summarizing Results: Summary statistic is the odds-ratio for unit change in each covariate and the 95% confidence interval, for all covariates in each model. P-values for testing each covariate individually also presented. P-values are calculated using either likelihood ratio tests from nested models or from the Wald-statistic.

13 The analysis is similar to that for non-intervention studies except additional models should be run to explore the impact of intervention (Trt) Models: The basic strategy will be to model Trt alone, the models described above, the models described above with the addition of Trt. Intervention Studies


Download ppt "Objective of the DAP A) Specify an analysis plan that can be applied to a wide variety of clinical HIV resistance studies. B) Include both Intervention."

Similar presentations


Ads by Google