A comparison of prior and posterior meanings shows what a clinician with these prior opinions would learn from Pim inhibitor these data. He or she would now consider virological failure less likely in older patients and more likely in female patients; higher viral load and higher CD4 cell count when starting darunavir would now be seen as at most slightly increasing and slightly decreasing the
risk of virological failure, respectively; but past poor adherence would still be viewed as probably harmful. He or she would now be less certain that an overall GSS when starting darunavir was predictive of subsequent virological failure. However, under other variants of the FDA’s algorithm, the overall GSS seems more predictive of virological failure (Table 4). Under the first two variants, patients who stop taking darunavir are not considered failures unless the reason given for stopping is treatment failure. Alternatives to the overall GSS suggest that both the number of failed PI regimens and failure on both amprenavir and saquinavir have some value selleck inhibitor as measures of the risk of virological failure, regardless of
the variant used to assess failure. Compared with a model where the potency of therapy is measured by resistance tests (model 2), a model with binary clinical measures (model 3) is as good at predicting the observed data (with 2logBF of –0.1, 1.6 and 3.0 under the three variants, respectively) and a find more model with continuous clinical measures (model 4) is slightly better at predicting the observed data (with 2logBF of 4.4, 9.4 and 3.9 under the three variants, respectively) . The patients receiving darunavir as part of salvage therapy in this study were not dissimilar to the highly treated patients receiving darunavir in the POWER
trials . Our patients were slightly older (mean age 48 years vs. 44 years), had been infected with HIV for longer (mean duration 17 years vs. 12 years) and started darunavir with a more advanced infection (CDC group C 43%vs. 36%), and hepatitis was more prevalent in our patients (chronic hepatitis B or C 23%vs. 11%). Yet our patients started darunavir in a better state of general health, with a lower viral load (mean 3.4 vs. 4.6 log copies/mL) and a higher CD4 cell count (median 250 vs. 150 cells/μL). A similar proportion of patients in our study started darunavir with three or more major PI mutations (57%vs. 54%) and with three or more darunavir-associated mutations (17%vs. 22%). In the POWER trials, 55% of highly treated patients failed to achieve a viral load below 50 copies/mL after 48 weeks of treatment with darunavir . In our study, 61 patients were followed for at least 48 weeks and at 48 weeks, 12 (20%) had experienced virological failure under the third variant of the FDA’s algorithm. In the POWER trials, 21% of patients discontinued darunavir before 48 weeks .