AbstractNowadays, in many different fields, massive data are available and for several reasons, it might be convenient to analyze just a subset of the data. The application of the D-optimality criterion can be helpful to optimally select a subsample of observations. However, it is well known that D-optimal support points lie on the boundary of the design space and if they go hand in hand with extreme response values, they can have a severe influence on the estimated linear model (leverage points with high influence). To overcome this problem, firstly, we propose a non-informative "exchange" procedure that enables us to select a "nearly" D-optimal subset of observations without high leverage values. Then, we provide an informative version of this exchange procedure, where besides high leverage points also the outliers in the responses (that are not necessarily associated to high leverage points) are avoided. This is possible because, unlike other design situations, in subsampling from big datasets the response values may be available. Finally, both the non-informative and informative selection procedures are adapted to I-optimality, with the goal of getting accurate predictions.
AbstractMaxi-min efficiency criteria are a kind of multi-objective criteria, since they enable us to take into consideration several tasks expressed by different component-wise criteria. However, they are difficult to manage because of their lack of differentiability. As a consequence, maxi-min efficiency designs are frequently built through heuristic and ad hoc algorithms, without the possibility of checking for their optimality. The main contribution of this study is to prove that the maxi-min efficiency optimality is equivalent to a Bayesian criterion, which is differentiable. In addition, we provide an analytic method to find the prior probability associated with a maxi-min efficient design, making feasible the application of the equivalence theorem. Two illustrative examples show how the proposed theory works.
IntroductionPI/r monotherapy has been suggested as an attainable maintenance strategy in patients achieving stable HIV suppression in plasma. The objective of trial was to compare the virological outcome of two different PI/r QD monotherapy strategies (LPV/r or DRV/r) with maintaining a triple PI/r‐based ARV regimen.Material and MethodsPhase III, open‐label, non‐inferiority (−12% margin), randomized trial of HIV adults with HIV‐RNA <50 cp/mL for at least 48 weeks while on PI/r‐based cART, CD4 nadir >100 cell/mm3, without previous PIs virological failure. Eligible patients were randomized to continue PI/r+2NRTIs (Arm A), to switch to LPV/r 800/200 mg QD monotherapy (Arm B), or to switch to DRV/r 800/100 mg QD monotherapy (Arm C). Primary endpoint was proportion of patients with plasma HIV‐1 RNA <50 cp/mL (TLOVR) at 48w by intent to treat (ITT) analysis (missing/re‐induction=failure). FDA snapshot and ITT switch‐included analysis (ITT‐SI) were also used. In ITT‐SI, patients who had <50 copies/mL at 96w were counted as successes even if they had confirmed HIV‐RNA elevations and had subsequently successfully intensified by NRTI.ResultsDue to slow recruitment, only 103 patients were included. No differences were observed between the three arms with respect to gender, age, HIV transmission, CD4 nadir and at screening. At randomization, 61 patients were receiving TDF/FTC (60%), 19 ZDV/3TC (18%), 8 ABV/3TC (8%), 75 LPV/r (73%), 13 ATV/r (13%), 4 DRV/r (4%). Differences in proportion of virological success by groups using Arm A as comparator according to FDA TLOVR were reported in Figure 1. Similar results were obtained by Snapshot analysis. Of 14 patients with virological failure, 8 patients restarted triple therapy with 2NRTI and 7/8 regained a VL <50 cp/mL over time. According to ITT‐SI analysis, 96 week differences [95% CI] were −5.7 [−29.6; +18.2] in Arm B, and +19.6 [−1.6; +40.8] in Arm C. A GRT was performed in 6/14 patients (one not amplifiable; four without mutations; one showed E138A).ConclusionsCompared to maintaining a PI/r‐based triple ARV regimen, LPV/r QD monotherapy tended to have higher rate of virological failure and of discontinuation due to adverse event. In contrast, the response rate at week 96 during DRV/r QD mono‐therapy was non‐inferior to that of triple PI/r‐based ARV therapy. A re‐induction with 2NRTI was adequate to obtain an undetectable viremia in most of patients with virological failure.
BackgroundTo assess efficacy and safety of treatment simplification to co‐formulated Rilpivirine/Emtricitabine/Tenofovir (RPV/FTC/TDF) in virologically suppressed patients.Materials and MethodsEndpoints of the analysis were: (a) treatment discontinuation of RPV/FTC/TDF for any reasons and (b) occurrence of virological failure (VF) defined as confirmed HIV‐RNA >50 cp/mL).ResultsOverall, 508 patients from five Italian reference centres were included: male gender 71.9%; median age 47 years (IQR 40–52); IVDU as HIV risk 17.7%; HCV‐AB positive 23.4%; CDC‐C stage 17.5%; median CD4 cells/µL at switch 655 (IQR: 487–843); median number of previous regimens three (IQR 2–7). In a median follow‐up of 196 days (IQR: 84–287), 31 patients discontinued RPV/FTC/TDF (virological failure n=5, hypersensitivity reaction n=2, GI‐toxicity n=6, liver toxicity n=1, CNS‐toxicity n=4, kidney toxicity n=5, patient's decision/lost in follow‐up n=10). Moreover, VF occurred in eight patients (five discontinued regimen, while three remained on RPV/FTC/TDF). At survival analysis, the probabilities of treatment discontinuation or VF were 5.5% and 1.2% at 6 months, 13.2% and 2.8% at 12 months, respectively (Figure 1). At adjusted Cox model, factors associated with discontinuation were: <200 CD4 cells/µL at switch (HR 5.3, 95% CI 1.1–25.9, p=0.038), number of pre‐switch regimens (for each extra regimen: HR 1.05, 95% CI 1.01–1.10, p=0.024), male gender (HR 0.4, 95% CI 0.2–0.9, p=0.032). Only the number of pre‐switch regimens was associated with VF (HR 1.13, 95% CI 1.06–1.21, p=0.001). Type of pre‐switch regimen was not associated with discontinuation or failure, but no VF was observed if switching from co‐formulated Efavirenz/FTC/TDF or from Raltegravir containing regimens. Switching to RPV/FTC/TDF led to significant improvement in fasting lipids levels: the decrease in cholesterol, LDL and triglycerides was observed switching from any regimen, but was more marked from boosted PI. In contrast, a moderate increase in transaminase (switching from all regimens except NNRTI‐containing) and creatinine (except from TDF‐containing regimens) were observed.ConclusionsOur data suggest that switching to RPV/FTC/TDF in virologically suppressed patients could be a good strategy with low risk of virological failure or treatment discontinuation; the switch is also associated with significant improvement in lipid profile.
