CROI 2015 Program and Abstracts

Abstract Listing

Poster Abstracts

Results: A total of 155 adults were enrolled with CD4 results for both field POC and lab-based testing. The median age was 32 years old and 63%were females. The median CD4 was 827 (IQR 668, 1058) cells/uL. We observed a mean bias between CD4 measurements by POC and FacsCalibur of -14.1 cells/uL [(95% LOA -327.4 – 299.2), paired t-test p = 0.2740] (Figure 1). Using a threshold of 500 CD4 cells/uL, POC CD4 testing using EDTA capillary sampling into microtube had a sensitivity of 73.3 (95% CI 44.9 – 92.3) cells/uL and specificity of 93.6 (88.1 – 97.0) cells/uL.

Figure 1 : Bland-Altman plot for comparison of Alere PIMA ® CD4 using microtube and FACScalibur Conclusions: The Alere PIMA® POC CD4 testing with EDTA capillary sampling into microtube had one of the lowest non-significant negative mean bias relative to FACScalibur CD4 testing, compared previous findings using finger prick sampling alone. There were fewmeasurements below CD4 threshold of 350 cells/uL and sensitivity was low at threshold of 500 cells/uL, but with good specificity. Coupling HIV testing with POC CD4 in a household setting has potential to close gaps in linkage to HIV care in communities. 631 Zyomyx MyT4 and BD FACSPresto Comparison to the Pima CD4 Assay Katie Tucker ; Sehin Birhanu; LarryWesterman US Centers for Disease Control and Prevention (CDC), Atlanta, GA, US Background: In conjunction with HIV antiretroviral treatment (ART) scale up, CD4 Point-of-Care (POC) assays are being utilized in resource-limiting settings to provide effective access to CD4 testing and determine eligibility for ART. Since 2009, the Alere Pima CD4 assay has been used for POC CD4 counts in resource-limiting setting and having an impact on HIV patient care. Many new POC CD4 assays are being introduced into the market, including the Zyomyx MyT4 and BD FACSPresto Near-Patient CD4 Counter. We assessed the quality, functionality, and performance of the MyT4 and FACSPresto assays in comparison with the established Pima CD4 assay. Methods: In 2014, we evaluated the MyT4 and FACSPresto assays. Whole blood specimens were analyzed for CD4 counts with the Pima CD4 and with the MyT4 (n=141) or FACSPresto (n=101). Scattered-plot and Bland-Altman analysis were done with Pima CD4 as the reference method. Precision of each assay was estimated by analyzing 3 specimens ten times on each platform and calculating the coefficient of variation (CV%). Error rates for each CD4 assay were assessed. Results: Scattered-plots and Bland-Altman analysis indicated that CD4 counts from both the MyT4 (R 2 =0.91, y=1.08x-28, bias +2 (-159, +156)) and FACSPresto (R 2 =0.96, y=1.15x+11, bias +78 (-59, +214)) correlated with the Pima CD4. CD4 counts with the MyT4 were estimated to have slightly more random error associated with the assay, while the FACSPresto CD4 counts were estimated to have a positive bias. Precision studies estimated the MyT4 to have a slightly higher CV% (8.2%) in comparison to FACSPresto (5.4%) and Pima (5.9%). Error rates increased on both MyT4 and FACSPresto with aged specimens (>12h, >20h), while the Pima CD4 was able to consistently analyze specimens up to 48 hours old. Conclusions: The Zyomyx MyT4 and BD FACSPresto correlated well with the Alere Pima CD4 and both prove to be acceptable methods for CD4 analysis. Both assays could potentially be used in point-of-care and resource-limiting settings. 632LB Point-of-Care CD4 (Pima) Impact on Linkage to CareWith Home-Based HIV Testing, Kenya Mitesh A. Desai 1 ; Duncan Okal 2 ; RobertT. Chen 1 ; Richard Ndivo 2 ; Charles Lebaron 1 ;TiffanyWilliams 1 ; Fred Otieno 2 ; Charles Rose 1 ;Taraz Samandari 1 ; Clement Zeh 1 1 US Centers for Disease Control and Prevention, Atlanta, GA, US; 2 Kenya Medical Research Institute, Kisumu, Kenya Background: Referral lab-based CD4 testing has been cited as a key barrier in the HIV care cascade, leading to delays in antiretroviral treatment (ART) initiation. Point-of-care (POC) CD4 testing has been demonstrated to improve linkage to care (LTC) among HIV-seropositive persons, but its impact has not been assessed with a randomized, controlled trial (RCT) in the context of home-based HIV counseling and testing (HBCT). We evaluated whether providing POC CD4 count, integrated with HBCT and referral for HIV care, would improve LTC and ART initiation when compared to standard-of-care (SOC) lab-based CD4 measurement. Methods: RCT participants were enrolled July’13-Feb.’14 in two districts of Western Kenya with ongoing HBCT. Compounds were randomized 1:1 to POC CD4 (Alere PIMA) or SOC CD4 testing at one of three referral labs. Participants in both arms received HBCT, post-test counseling, and referral for HIV care. All HIV+ adults not engaged in care in last 180 days were invited to participate. Intervention arm participants received additional counseling on the POC CD4 result, including ART eligibility if CD4<350cells/ m L. Enrolled participants gave informed consent. Participants were interviewed 180 days after enrollment to ascertain whether and where they sought care, which was verified with charts at any of 23 clinics serving the study area. Verified LTC (a HIV clinic visit) was primary outcome, with protocol-specified analysis of time to ART initiation comparing SOC to intervention using a Cox proportional hazards model. Results: We describe final results from 770 enrolled participants, of whom 692 had chart-verified LTC status, and 78 were lost to followup. Baseline characteristics including sex, age, education, marital and employment status were similar in the two arms. Of 321 SOC participants, 106 (33.0%) had confirmed LTC within 180 days after enrollment, whereas of 371 in intervention arm, 208 (56.1%) had confirmed LTC (HR = 1.996; 95%CI: 1.58-2.52). Kaplan-Meier analysis of time to LTC (Figure) produced a log-rank p<0.0001 for the difference in proportions linked to HIV care in two arms. Among SOC participants, 33 (10.3%) had initiated ART, compared to 61 (16.4%) in the intervention arm (HR = 1.67; 95%CI: 1.09-2.55).

Poster Abstracts

397

CROI 2015