MUUCSF Presentations

Conference on Retroviruses and Opportunistic Infections (CROI)
Please click on conference to view abstracts presented.


2013, Atlanta, GA, USA

Comparison of Virologic and Immunologic Outcomes between HIV+ Ugandan Children Randomized to Ritonavir-boosted Lopinavir or NNRTI- based ART
Theodore Ruel*1, A Kakuru2, G Ikilezi2, F Mwangwa2, G Dorsey1, P Rosenthal1, E Charlebois1, D Havlir1, M Kamya3, and J Achan3
1Univ of California, San Francisco, US; 2Infectious Diseases Res Collaboration, Kampala; and 3Makerere Univ, Kampala, Uganda

Background:   In the PROMOTE-pediatrics trial (NCT00978068), HIV+ Ugandan children randomized to receive ritonavir-boosted-lopinavir (LPV/r)-based ART experienced a lower incidence of malaria compared to children receiving NNRTI-based-ART. Here we present results of a planned non-inferiority (NI) analysis of virologic efficacy and a comparison of immunologic outcomes.

Methods: In this open-label randomized trial, ART-naïve and ART-suppressed (HIV RNA <400 copies/mL) HIV+ Ugandan children ages 2 months to <6 years were enrolled from October 2009 to October 2011, and received either LPV/r- or NNRTI-based ART with nevirapine (NVP) or efavirenz (if ≥3 years old). Children <24 months old with perinatal exposure to NVP were excluded. The proportion of children with virological suppression (VS: HIV RNA <400 copies/mL) at 48 weeks was compared by arm; a NI margin of  –11% for LPV/r in per-protocol analysis was pre-specified. Mean change in CD4 number (ΔCD4n) and CD4 percentage (ΔCD4p) since enrollment and the proportion experiencing Grade III/IV adverse events (AEs) were also compared by arm.

Results: A total of 185 children were randomized to LPV/r-based (n = 92) or NNRTI-based (n = 93) ART, and reached a mean (SD) follow-up of 108 (35) weeks. At enrollment, the median age was 3.1 years (range 0.4 to 5.9), 91 (49%) were female and 131 (71%) were ART-naïve. Among the 163 children with available HIV RNA levels at 48 weeks, the proportion with VS was 80% in the LPV/r-arm compared to 76% in the NNRTI-arm, a difference of 3.8% (95% confidence interval [CI] –8.9% to +17). Table 1 summarizes VS and CD4 outcomes stratified by ART-status at enrollment. At least 1 grade III/IV adverse event was experienced by 59% of children in the LPV/r arm vs 52% in the NNRTI arm (p = 0.3). Adverse events led to ART changes in 2 children, both initially receiving NVP: Stevens Johnson Syndrome (n = 1) and hepatitis (n = 1). There were 4 deaths (NNRTI n = 3, LPV/r n = 1).

Conclusions:  In this cohort of ART-naive and ART-suppressed HIV+ Ugandan children, the use of LPV/r-based ART led to a rate of virologic suppression that was not inferior to that with NNRTI-based ART and yielded comparable immunologic responses and adverse event rates. Considering the reduction of malaria incidence associated with LPV/r use in this cohort, these results suggest that wider use of LPV/r to treat HIV+ African children in similar settings could be considered.


Selection of ARV Regimen Impacts Antimalarial Pharmacokinetics and Treatment Outcomes in HIV/Malaria Co-infected Children in Uganda


Norah Mwebaza*1, R Kajubi1, J Ssebuliba1, S Kiconco1, L Huang2, Q Gao2, A Kakuru1, J Achan1, F Aweeka2, and S Parikh2,3
1Makerere Univ, Kampala, Uganda; 2Univ of California, San Francisco, US; and 3Yale Univ Sch of Publ Hlth and Med, New Haven, CT, US

Background: HIV+ children on protease-inhibitor (PI)-based ART regimens are at reduced risk for malaria compared with children on NNRTI-based ART. We conducted an intensive pharmacokinetics/pharmacodynamics (PK/PD) study in children to characterize the disposition of artemether-lumefantrine (AL), a first-line regimen for malaria, in the setting of ART co-administration. We provide the first intensive PK/PD data in HIV-malaria co-infection to inform treatment guidelines.

Methods: HIV/malaria co-infected children aged 6 months to 8 years randomized to either lopinavir/ritonavir (LPV/r) or NNRTI-based ART (nevirapine [NVP] or efavirenz [EFV]) through an open label trial were enrolled for intensive PK/PD evaluations of AL. Exposure was estimated to 21 days (area under the concentration-time curve [AUC]) with 42 days of follow-up. Artemether (AR), its active metabolite dihydroartemisinin (DHA), and the long-acting partner drug, lumefantrine (LR) were quantitated by LC/MS/MS.

Results: PK/PD evaluations were completed and analyzed for 42 of 70 children enrolled since August 2011(26 LPV/r; 10 NVP; 6 EFV). For the artemisinins, higher exposure (AUC(0-8hr)) to AR was seen with concomitant use of LPV/r- vs NPV-based ART (LPV/r:NVP ratio 2.3; p = 0.028) but not with EFV (LPV/r:EFV ratio 1.5; p = 0.331). Exposure to DHA did not differ between ART. However, LR exhibited a 4-fold increase in t1/2and six-fold higher AUC(0-∞) for those on LPV/r vs EFV-based ART (p <0.001), and 2-fold higher AUC on LPV/r compared with NVP (p = 0.005). Comparing children on EFV- to those on NVP-based ART, LR AUC(0-∞) was significantly reduced for those on EFV (EFV:NVP ratio 0.34, p = 0.022). Moreover, mean day 7 LR levels were statistically different between the three arms (908, 315, and 100 ng/mL for those on LPV/r, NVP, and EFV, respectively). We also assessed association of PK exposure and 28 day malaria treatment outcomes, with early results showing that 4/6 children on EFV-, but only 1/10 on NVP- and 2/26 on LPV/r-based ART experienced treatment failure (PCR-unadjusted).

Conclusions: AL co-administration for malaria with LPV/r- compared to NNRTI-based ART results in significant increase in exposure to both AR and LR. Although data are limited, EFV-based ART, with the lowest comparative PK exposure, may impact AL treatment response. These findings suggest that increased exposure to AL in the context of LPV/r-based ART may offer advantages over NNRTI-based regimens in malaria endemic areas.


No Difference in Risk of Preterm Birth among Pregnant Ugandan Women Randomized to Lopinavir/Ritonavir vs Efavirenz-based ART


Deborah Cohan*1, P Natureeba2, A Plenty1, F Luwedde2, J Mwesigwa2, V Ades3, B Nzarubara2, T Clark1, B Osterbauer1, and E Charlebois1
1Univ of California, San Francisco, US; 2Infectious Disease Res Collaboration, Kampala, Uganda; and 3New York Univ, NY, US

Background: As an increasing number of women are started on cART during pregnancy, it is critical to evaluate optimal regimens. There are conflicting observational data on the risk of preterm birth (PTB) among women on protease inhibitor-based ART. The only trial to evaluate this association found a 2-fold increased odds of PTB among women randomized to lopinavir/ritonavir (LPV/r) versus abacavir-based ART between 26 and 34 weeks gestation.

Methods: PROMOTE-Pregnant Women and Infants is an open-label, randomized controlled trial comparing the risk of placental malaria among rural Ugandan women initiated on LPV/r vs efavirenz (EFV)-based cART during pregnancy. ART-naïve women were enrolled at 12-28 weeks’ gestation. Gestational age at enrollment was calculated using last menstrual period and ultrasound biometry. For this analysis we included women who had delivered a live born singleton infant or, if still pregnant, had reached at least 32 weeks gestation by January 11, 2013. We compared the risk of PTB (<37 weeks) and very PTB (<32 weeks) as well as the median gestational age at delivery among women in the 2 arms using c2 and Wilcoxon rank sum tests, as appropriate. Furthermore, we stratified the risk of PTB by gestational age at cART initiation. We also compared median birth weight and the proportion with low birth weight (LBW, <2500 g).

Results: Overall, 391 women enrolled, including 197 on LPV/r and 194 on EFV. As of January 11, 2013, there were 357 singleton deliveries, 346 (96.9%) live born infants, 8 pregnant women who had reached at least 32 weeks gestation and 6 pregnant women who had reached at least 37 weeks gestation. There were no significant differences between the arms at baseline, including gestational age at enrollment. The overall risk of PTB was 14.9% and did not differ by regimen—15.9% in LPV/r and 13.6% in EFV arms, respectively (Table 1). Similarly, the risk of very PTB did not differ between the 2 arms. Median gestational age at delivery among women on LPV/r was 38.7 weeks vs 39.2 weeks for EFV. Moreover, the prevalence of LBW was 16.8% and was not significantly different among women on LPV/r (17.3.0%) vs EFV (16.3%).

Conclusions: There was no difference in either preterm birth or low birth weight among ART-naïve Ugandan women randomized to LPV/r vs EFV combination ART between 12 and 28 weeks gestation.

2012, SEATTLE, WA, USA


Achan J, Kakuru A, Ikilezi G, Ruel T, Clark TD, Charlebois ED, Rosenthal PJ, Dorsey G, Havlir DV, Kamya MR. Significant Reduction in Risk of Malaria among HIV-Infected Children Receiving LPV/r-Based ART Compared to NNRTI-Based ART: A Randomized Open Label Trial. Oral presentation at CROI March 2012

Background: HIV-infected children living in sub-Saharan Africa continue to suffer high morbidity and mortality from malaria. Additional measures to decrease malaria risk in these populations are therefore warranted. HIV protease inhibitors have shown some activity against P.falciparum in vitro. We evaluated the efficacy of LPV/r-based versus NNRTI-based ART regimens for malaria risk reduction in HIV-infected children.

Methods: We conducted a randomized open label trial in Tororo, Uganda. Participants were HIV- infected children aged 2 months to 5 years eligible for ART or currently receiving NNRTI- based ART with virological suppression (HIV RNA<400 copies/ml). Participants were randomized to receive either LPV/r-based or NNRTI-based ART and followed up for 2 years. Uncomplicated malaria was treated with artemether-lumefantrine. We compared the incidence of malaria between the study arms using Poisson regression.

Results: Between September 2009 and July 2011, we enrolled 176 children and randomly assigned 89 to NNRTI-based and 87 to LPV/r-based ART. Median age was 3 years and 67% were ART naïve. Malaria incidence was significantly higher in the NNRTI arm compared to that in the LPV/r arm; 2.25 vs. 1.32 episodes/person year, IRR 0.59, 95% CI 0.36-0.97, p=0.037. The cumulative 28-day risk of treatment failure unadjusted by genotyping was significantly higher in the NNRTI arm compared to the LPV/r arm; 40.8%, 95% CI 33.9 - 48.6 vs. 14.0%, 95% CI 8.7 – 22.2; HR 0.31, 95% CI 0.14 – 0.68 (p= 0.004). Median serum lumefantrine levels on day 7 of follow-up were significantly higher in the LPV/r arm compared to the NNRTI arm. In the NNRTI arm, day 7 lumefantrine levels were higher in those on Nevirapine compared to those on Efavirenz; 388ng/ml vs. 97ng/ml respectively, p < 0.000. In the LPV/r arm only, day 7 lumefantrine levels >300ng/ml were associated with a >85% reduction in the risk of recurrent malaria within 63 days following treatment with a dose dependent reduction in risk compared to levels < 300ng/ml. In-vitro studies showed modest synergy between lopinavir and lumefantrine. No QTc interval prolongation was noted.

Conclusions: Use of LPV/r-based ART reduced the incidence of malaria by 41% compared to NNRTI-based ART, largely attributable to a significant reduction in the risk of recurrent malaria following artemether-lumefantrine treatment. LPV/r-based ART could therefore be considered as one of the strategies for malaria prevention in HIV-infected children.


Cohan DL, Young SL, Murray K, Mwesigwa J, Achan J, Ades V, Charlebois ED, Ruel T, Kamya MR, Havlir DV. Maternal Nutritional Status Predicts Adverse Birth Outcomes among HIV-infected Rural Ugandan Women Receiving Combination Antiretroviral Therapy. Abstract presented at CROI March 2012

Background: Maternal nutritional status is an important predictor of birth outcomes, yet data are limited from HIV-infected pregnant women treated with combination antiretroviral therapy (cART). We examined the relationship between markers of maternal nutritional status, including gestational weight gain (GWG) and hemoglobin concentration (Hb), and birth outcomes among a cohort of HIV-infected pregnant women initiating cART in rural Uganda.
Methods: HIV-infected, ART-naïve pregnant women were enrolled between 12 and 28 weeks gestation and treated with either a protease inhibitor or non-nucleoside reverse transcriptase inhibitor-based combination regimen in an on-going, prospective, open-label randomized clinical trial (NCT00993031). Maternal nutritional status was assessed monthly. Neonatal anthropometry was examined at birth. Associations were evaluated using multivariate logistic regression, adjusting for potential risk factors.

Results: At enrollment, 44.9% of the 158 women were anemic. Mean GWG was 0.17kg/week with 14.6% experiencing weight loss during pregnancy. No infants became HIV-infected. Adverse birth outcomes included low birth weight [LBW] (19.6%), preterm delivery [PTD] (17.7%), fetal death (3.9%), stunting (21.1%), small-for-gestational age (15.1%), and head-sparing growth restriction (26.0%). Maternal GWG < 0.1kg/week was associated with LBW (adjusted odds ratio [aOR] 6.08, 95% confidence interval [95% CI] 1.70-20.00, p=.004), PTD (aOR 3.46, 95% CI 1.18-10.15, p=.024), head-sparing growth restriction (aOR 2.24, 95% CI 0.96-5.22, p=0.06), and a composite adverse birth outcome (aOR 2.85, 95% CI 1.32-6.15, p<.01). Each kilogram decrease in maternal weight at 7 months gestation was associated with a nearly 40% increased odds of LBW (aOR 1.38, 95% CI 1.03-1.90, p=0.034). In contrast, for each g/dL higher mean Hb, the odds of small-for-gestational age decreased by 52% (aOR 0.48, 95% CI 0.29-0.80, p=0.004).

Conclusions: Women on cART had grossly inadequate gestational weight gain. Infants whose mothers gained <0.1kg/week were at increased risk for LBW, preterm delivery, and head sparing growth restriction. cART by itself may not be sufficient for decreasing the burden of adverse birth outcomes among HIV-infected women. Strategic nutritional interventions are urgently needed


Chamie G, Kwarisiima D, Kabami J, Clark TD, Jain V, Black D, Geng E, Charlebois ED, Kamya MR, and Havlir DV. Community-based HIV Testing and Point-of-care CD4 in Rural Uganda: Outcomes in a Routine Linkage to Care Strategy and an Enhanced Strategy with Accelerated ART Start. Abstract presented at CROI, March 2012.

Background: Linkage to care rates after HIV diagnosis in sub-Saharan Africa are unacceptably low. Delays and drop-outs can occur before ART start due to multiple steps in care delivery. We sought to improve linkage to care and remove barriers to ART initiation after HIV voluntary counseling and testing (VCT) with rapid referral, point-of-care (POC) CD4 count testing and TB screening.

Methods: We performed HIV VCT at a 5 day, multi-disease health campaign in a rural Ugandan community of 6,000 supported by local staff and the Ministry of Health. We provided HIV+ adults on-site, POC CD4 count (PIMA) testing and TB screening (sputum Xpert MTB/RIF assay), and referred to a local HIV clinic using two strategies. HIV+ adults with CD4>100/µL received routine referral to the clinic within 3 months. HIV+ adults with CD4≤100/µL received an enhanced strategy of rapid referral (<2 weeks post-campaign), expedited pre-ART counseling and immediate ART initiation at the first visit. All participants received a transport stipend and were introduced to a clinic nurse. We examined predictors of linkage to care in the 3 months after the campaign and conducted home visits to evaluate reasons for failing to link.

Results: 140 HIV+ adults participated in the linkage study. We made routine referrals for 132 adults, 75 (57%) of whom linked to care within 3 months. We made enhanced referrals for 7 adults with advanced AIDS and a pregnant woman [CD4=278]. Of these, 6 (75%) linked to care within 10 (median 2.5) days of the campaign and started ART immediately; all 6 screened negative for TB by Xpert assay and clinician assessment. All 6 reported adherence to ART 4 weeks after linking to care. For the 132 adults, new HIV diagnosis was associated with reduced odds of linkage (OR: 0.1 [0.04-0.25], p<0.001), after adjusting for age, sex, CD4, and distance from clinic. Among new diagnoses, median CD4 was higher in adults who did not link vs. linked (513 vs. 315, p=0.007). In home visit interviews, participants who failed to link have reported insufficient transport funds and accessing care at other sites as key reasons for not linking.

Conclusions: An enhanced linkage approach after a community testing drive led to rapid ART initiation in adults with advanced HIV. In contrast, even with robust measures to promote linkage, routine linkage to care efforts were insufficient (57%) in adults with higher CD4 cell counts. Improved linkage strategies across a broad spectrum of CD4 counts are needed



Jain V, Liegler T, Chamie G, Wong J, Geng E, Thirumurthy H, Aweeka F, Charlebois ED, Kamya MR, Havlir DV, and the SEARCH Collaboration Assessment of Community Viral Load in a Rural East African Setting Using a Fingerprick-Based Blood Collection Method. Abstract presented at CROI, March 2012.

Background: Average HIV-1 plasma RNA level in a population (community viral load, CVL) is a proposed metric for ART program effectiveness, and a measure of untreated disease. However, in resource limited settings where ART is scaling up, CVL has not been assessed due to barriers such as lack of routine VL testing and a high degree of undiagnosed HIV infection. We sought to estimate CVL using a field-adapted procedure for HIV RNA measurement in a rural Ugandan community in the midst of rapid ART scale-up for those with CD4<350.

Methods: Fingerprick blood was collected with Safe-T-Fill tubes (RAM Scientific) prior to VL testing (Abbott). Results were validated against phlebotomy based values by Bland-Altman analysis. Efavirenz (EFV) and nevirapine (NVP) levels were measured from dried blood spots via HPLC and LCMS. We deployed these techniques during a 5-day high throughput community-wide health campaign in a Ugandan parish of 5,100 persons. 85% of the community was tested for HIV (Determine HIV 1/2, Inverness). We measured VL and CD4+ count (Alere PIMA) on all HIV+ persons, and assessed predictors of VL via linear regression. We estimated CVL (defined as the mean VL of HIV+ persons) and mapped CVL to geographic units in the community.

Results: VL from fingerprick and standard phlebotomy showed good pairwise agreement (lower cutoff 486 cps/mL); median difference (5th, 95th percentile) was 0.039 log10 cps/mL (-0.24, 0.32). 179/2282 adults (7.8%), and 10/1826 children (0.5%) were HIV+. Fingerprick VL was determined in 174/189 HIV+ persons (92%). The mean CVL was 64,064 cps/mL. Overall, 24% of participants had VL 486-10,000; 25% had VL 10,001-100,000; and 15% had VL>100,000 cps/mL. 37% of participants had an undetectable VL, and 50/57 (88%) of patients with undetectable VL tested had detectable EFV or NVP levels. Of all participants on ART, 83% had undetectable VL. Adjusted associations were seen between VL and ART use (-1.45 log, p<0.0001) and married status (-0.28 log, p=0.017) but not with age or occupation. Regions farther from the local health center did not have higher mean CVL.


Conclusions: We developed and implemented a field-based, fingerprick method for VL measurement and provide the first report of CVL in Africa. In a rural Ugandan community where ART is being rapidly scaled up, we found evidence of population-level ART effectiveness, but found a substantial proportion of HIV positive persons viremic in need of ART, and at risk for transmission.


Geng E, Namusobya J, Semitala F, Kabami J, Chamie G, Charlebois ED, Havlir DV, and Kamya MR. Retention in care and mortality among HIV-infected Patients Entering Care with High CD4 Levels – Understanding outcomes in the pre-ART period using a sampling-based approach: Uganda. Abstract presented at CROI, March 2012.

Background: Understanding outcomes in the growing proportion of HIV-infected patients in Africa who present with CD4 levels above the treatment threshold is needed to assess retention, survival, and effectiveness of care in the pre-ART period. Loss to follow-up (i.e., unknown outcomes) in this group, however, is particularly high and precludes accurate evaluation of this patient population.

Methods: We identified adult patients in a large Ugandan HIV care and treatment program in Kampala with an entry CD4 count >350/µL between October 1, 2008 and April 1, 2011. We used the Kaplan-Meier method to estimate ART initiation in and loss to follow-up from the original clinic site. Among the patients who became lost (defined as at least 6 months late for a return visit), a random sample was identified and tracked in the community to determine updated vital status and clinical care information.

Results: Over 2.5 years, 4318 patients enrolled in care with a CD4 >350/µL. Of these, 3202 (74%) were women, median age was 30 years (IQR 26 to 36), and median entry CD4 level was 550/µL (IQR 439/µL to 715/µL). During follow-up, 1101 patients started ART at the original clinic for a cumulative incidence of 35% (95%CI 32% to 38%). Over this time, 858 patients became lost for a cumulative incidence of 30% (95%CI 27% to 33%). A random sample of 67 of 858 (8%) was tracked in the community, and in 56 of the 67 (84%) outcomes were ascertained. In 34 of 56 (61%), the patient was found in person and in 22 of 56 (39%) outcomes were reported by an informant (e.g., spouse, child, sibling). Overall, 5 of 56 of the lost had died; the cumulative incidence of mortality at 1, 2, and 2.5 years after last clinic visit was 2.0% (95%CI 0.3% to 12.9%), 8.8% (95%CI 3.4% to 22%), and 13.6 % (95%CI 5.5% to 31.6%). Of lost patients, 19 of 34 (56%) were found to be alive and interviewed in person: they reported seeing a doctor or a nurse within the last 90 days, including 8 of 34 (23%) who reported ART initiation at a new clinic site.

Conclusions: In HIV-infected patients in Uganda entering care with CD4 levels above treatment thresholds, retention in care is higher than clinic retention, and even includes a notable fraction initiating ART at other sites after loss to follow-up from the original site. Nevertheless, about half of lost patients have disengaged from care and mortality is high. Failure to retain patients entering care with high CD4 represent a missed opportunity to optimize effectiveness.

 

Gabriel Chamie, D Kwarisiima, J Kabami, T Clark, V Jain, D Black, E Geng, E Charlebois, M Kamya, and D Havlir.Outcomes in a Routine Linkage-to-care Strategy and an Enhanced Strategy with Accelerated ART Start; Community-based HIV Testing and Point-of-care CD4: Rural Uganda. Abstract presented at CROI, March 2012.

 

Background:  Linkage to care rates after HIV diagnosis in Sub-Saharan Africa are unacceptably low. Delays and drop-outs can occur before ART start due to multiple steps in care delivery. We sought to improve linkage to care and remove barriers to ART initiation after HIV voluntary counseling and testing (VCT) with rapid referral, point-of-care CD4 count testing, and TB screening.

 

Methods:  We performed HIV VCT at a 5-day, multi-disease health campaign in a rural Ugandan community of 6000, supported by local staff and the Ministry of Health. We provided HIV+ adults with on-site, point-of-care CD4 count (PIMA) testing and TB screening (sputum Xpert MTB/RIF assay), and referred them to a local HIV clinic using 2 strategies. HIV+ adults with CD4 >100/µL received routine referral to the clinic within 3 months. HIV+ adults with CD4 ≤100/µL received an enhanced strategy of rapid referral (<2 weeks post-campaign), expedited pre-ART counseling, and immediate ART initiation at the first visit. All participants received a transport stipend and were introduced to a clinic nurse. We examined predictors of linkage to care in the 3 months after the campaign and conducted home visits to evaluate reasons for failing to link.

 

Results:  The linkage study comprised 140 HIV+ adult participants. We made routine referrals for 132 adults, 75 (57%) of whom linked to care within 3 months. We made enhanced referrals for 7 adults with advanced AIDS and a pregnant woman (CD4 = 278). Of these, 6 (75%) linked to care within 10 (median 2.5) days of the campaign and started ART immediately; all 6 screened negative for TB by Xpert assay and clinician assessment. All 6 reported adherence to ART 4 weeks after linking to care. For the 132 adults, new HIV diagnosis was associated with reduced odds of linkage (OR 0.1, 0.04 to 0.25, p <0.001), after adjusting for age, sex, CD4, and distance from clinic. Among new diagnoses, median CD4 was higher in adults who did not link than those who did (513 vs 315, p = 0.007). In home visit interviews, participants who failed to link reported insufficient transport funds and accessed care at other sites as key reasons for not linking.

 

Conclusions:  An enhanced linkage approach after a community testing drive led to rapid ART initiation in adults with advanced HIV. In contrast, even with robust measures to promote linkage, routine linkage-to-care efforts were insufficient (57%) in adults with higher CD4 cell counts. Improved linkage strategies across a broad spectrum of CD4 counts are needed.


Gabriel Chamie, B Wandera, A Luetkemeyer, R Mugerwa, D Havlir, and E Charlebois. Household Ventilation among TB Index Cases: Kampala, Uganda. Abstract presented at CROI, March 2012.

 

Background:  Household contacts of pulmonary TB cases, particularly HIV+ adults and children, are at increased risk of TB. We sought to test the feasibility of a simple in-home measurement of ventilation using a carbon dioxide (CO2) tracer gas decay technique and to determine whether household ventilation is associated with TB in household contacts.

 

Methods:  We visited homes of adults with pulmonary TB participating in a study of home- vs clinic-based HIV testing for household members of TB suspects in Kampala. We interviewed all available adult household members regarding duration of residence, number of residents, window use, and self-reported co-prevalent TB disease in household contacts over the past 2 years. We measured ventilation in air changes per hour (ACH) in each room with windows and doors closed, and with windows open. We determined ACH by raising CO2 levels using dry ice to >2000 parts per million, well above ambient CO2, removing the dry ice, and measuring changes in CO2 at 5-second intervals until CO2 returned to baseline. Using a well-mixed box model, we regressed the natural log of CO2 over time to estimate ACH. ACH was measured 4 times and averaged in each room. We compared ACH in the sleep rooms of the index cases in homes with and without co-prevalent TB.

 

Results:  We enrolled 61 homes, with a median sleep-room ventilation rate of 7.8 ACH (IQR 5.3 to 11.2), and a median of 5 people (IQR 4 to 6) and 1 room (IQR 1 to 3) per home. Of 61 homes 58 (95%) reported closing all windows and doors at night, and 22 (36%) had ventilation rates <6 ACH. In 12 of the 61 homes (20%), residents reported co-prevalent TB in the past 2 years:  5 of 64 available adult contacts (8%) reported a TB diagnosis (all tested HIV–). Of the 61 homes, 8 (13%) reported a child contact (≤17 years) who had been diagnosed with TB. The 12 homes with co-prevalent TB had a similar mean sleep-room ventilation rate (8.3 ACH) to homes without co-prevalent TB (9.1 ACH, p = 0.29). In the 2 homes reporting TB in a child below school age (<5 years), mean sleep-room ventilation rate was 6.4 ACH.

 

Conclusions:  In this pilot study, measuring ventilation in homes of pulmonary TB cases in urban Uganda was feasible. The majority of ventilation rates fell within recommended levels for TB isolation rooms (6 to 12 ACH), though over a third had low ventilation rates (<6 ACH). Further study is needed to examine the effect of household ventilation on TB transmission and characterize sites of transmission in TB endemic settings.

 

Theodore Ruel, J Achan, E Charlebois, LA Eller, W Huang, P Li, M Kamya, D Havlir, and J Wong. HIV-DNA Levels in Naïve and Memory CD4+ Lymphocytes from ART-naïve Ugandan Children Infected with HIV Subtypes A and D. Abstract presented at CROI, March 2012.

Background:  Infection with HIV subtype D is associated with accelerated CD4 decline and prevalence of CXCR4 tropism compared to subtype A. Children have relatively large populations of naïve T cells that typically express high levels of CXCR4. We hypothesized that infection with subtype D and CXCR4-tropic HIV would be associated with higher HIV-DNA levels in naïve CD4+ T cells compared to infection with subtype A or CCR5-tropic HIV among ART-naïve HIV-infected Ugandan children that might impact pathogenesis.

 

Methods:  HIV subtype was determined from plasma virus using a Taqman assay with subtype-specific, env region probes. Specimens from HIV subtype D (n = 17) and A (n = 18) infected children were selected, including isolates with R5 (n = 19), dual/mixed (DM, n = 5), and unknown (n = 11) tropism (Trofile, Monogram). Peripheral blood mononuclear cells were sorted into naïve (CD27+/CD45RA+) and memory (CD27–/CD45RA–) CD3+/CD4+ T cells. HIV-DNA levels were determined using a Taqman assay targeting the HIV gag and normalized to cellular DNA content (tert, ABI). Plasma HIV RNA (Amplicor, Roche) and total CD4 number/% (FACS, BD) were determined on fresh specimens. HIV subtype and co-receptor tropism were evaluated as predictors of log-transformed HIV DNA levels using Kruskal-Wallis testing and probability weighted multivariate linear regression (MLR).

 

Results:  Children had median (interquartile range, IQR) age of 6.4 (4.8 to 8.6) years, CD4 number of 616 (270 to 875) cells/µL, CD4 % of 17 (14 to 27), and HIV RNA of 5.4 (4.9 to 5.8) log10 (copies/mL). HIV DNA levels could be determined for 25 naïve and 25 memory T cell samples, with median (IQR) log10 (copies/106 cells) levels as shown.

In MLR, younger age (p = 0.011) and lower total CD4 number (p = 0.002) were associated with increased naïve cell HIV DNA level but HIV RNA level (p = 0.3) and HIV subtype (p = 0.3) were not.   No significant predictors of memory HIV DNA levels were identified. Low total CD4 number was also predictive of naïve to memory HIV DNA ratio (p = 0.04).

 

Conclusions:  There were trends toward elevated levels of HIV DNA in naïve CD4+ T cells DNA from children with subtype D and DM tropic virus. Decreased total CD4+ T cell count was associated with increased levels of HIV DNA in naïve and memory CD4+ T cells. Additional study to define the relationships between HIV subtype, co-receptor tropism, naïve T cell infection, and disease course in HIV-infected children are ongoing.

 

Jaco Homsy, G Dorsey, M Muhindo, V Bigira, A Kakuru, E Arinaitwe, E Osilo, T Sandison, M Kamya, and J Tappero. Protective Efficacy of Prolonged TS Prophylaxis against Malaria from Birth to 4 Years of Age among HIV-exposed Children: Rural Uganda, a Randomized Trial. Abstract presented at CROI, March 2012.

 

Background:  WHO recommends that HIV-exposed children be given trimethoprim-sulfamethoxazole (TS) prophylaxis from 6 weeks of age until HIV exposure ends and HIV infection is excluded. However, there are no recommendations for TS prophylaxis among HIV– children beyond the period of exposure to HIV. We have shown that continued TS prophylaxis in HIV– Ugandan children from the end of breast feeding until 2 years of age was associated with a 39% reduction in malaria incidence. We now report on the protective effect of prolonged TS prophylaxis against malaria from the end of breast feeding until 4 years of age among these children.


Methods:  We consecutively enrolled 185 HIV-exposed uninfected breast-feeding children aged 6 weeks to 12 months in Tororo Hospital, Uganda. Each was given an insecticide-treated bed-net, prescribed daily TS prophylaxis, and randomized to 1 of 3 TS prophylaxis duration groups once confirmed HIV–:  group 1 (n = 87) continued TS until breast feeding ended (median age 10 months); group 2 (n = 46) continued TS until 2 years of age; group 3 (n = 45) continued TS until 4 years of age. All children were followed until 4 years of age. Malaria incidence was compared among the 3 groups stratified by age using negative binomial regression adjusted for place of residence and malaria treatment. Malaria was defined as a fever and positive thick blood smear and was treated with 1 of 2 randomly assigned artemisinin combination therapies.

 

Results:  Continued TS prophylaxis until 4 years of age compared to 2 years of age was associated with a 53% reduction in malaria incidence (3.27 vs 7.12 episodes/person-year; IRR 0.47, 95%CI 0.33 to 0.67, p = <0.001). This protective effect was 32% greater than, but not significantly different from (p = 0.34), that previously observed among the same infants who had continued TS until 2 years of age compared to those who stopped after the end of breast feeding (3.29 vs 5.58 episodes/person-year; IRR 0.60, 95%CI 0.40 to 0.80, p = 0.001). There was no difference in the comparative incidences of diarrhea, pneumonia, or rates of hospitalization among any TS duration groups.

 

Conclusions:  TS prophylaxis beyond the period of HIV exposure until 4 years of age was associated with a significant reduction in malaria incidence among Ugandan children. These findings suggest that prolonged TS prophylaxis may significantly reduce malaria incidence among HIV– children living in areas of high malaria transmission intensity.

2011, San Francisco, CA USA

Jane Achan, T Ruel, G Ikilezi, A Kakuru, G Dorsey, P Rosenthal, E Charlebois, D Havlir, and M Kamya. HIV-infected Ugandan Children Randomized to Switch from NNRTI- to LPV/r-based ART Maintain Virologic Suppression and Stable CD4 levels. Abstract presented at CROI, March 2011.

 

Background:  While most children in limited resource settings tolerate first-line NNRTI-based ART, some must change to ritonavir-boosted lopinavir (LPV/r) -based therapy because of side effects. The safety of a switch in a randomized comparison not confounded by prior toxicity or tolerance issues had not been evaluated.

 

Methods:  As part of a larger and ongoing clinical trial, HIV+ Ugandan children receiving NNRTI-based ART with virologic suppression (< 400 copies/mL plasma HIV RNA) were randomized to either continue their NNRTI or switch to LPV/r, while maintaining treatment with prior NRTI. The prevalence of virologic suppression, change in CD4 percentage and count, and the frequency of grade 3-4 US NIH Division of HIV/AIDS adverse events after 12 and 24 weeks were compared using non-parametric techniques with available follow-up data.

 

Results:  Of 51 children with median age 2.9 years (range 1 to 6 years) and median prior ART duration (available for 43 children) of 1.7 years (range 0.2 to 4.7), 25 were randomized to switch to LPV/r and 26 to continue their NNRTI (25 neviraprine and 1 efavirenz). The prevalence of virologic suppression was high and comparable in the LPV/r and NNRTI arms at 12 (22 of 22 vs 22 of 22, p = 1.0) and 24 weeks (15 of 16 vs 14 of 16, p = 0.65), respectively. The 3 children without suppression had HIV RNA levels of 431, 532, and 1590 copies/mL. Median change (range) in CD4 percentage for LPV/r and NNRTI arms were similar after 12 weeks (0.5, –9 to 14 vs –1.5, –11 to 12; p = 0.20) and 24 weeks (–2, –14 to 14 vs –1, –10 to 9; p = 0.40). Median change (range) in CD4 counts for LPV/r and NNRTI arms were also similar after 12 weeks (75, –2657 to 797 vs –237, –1506 to 704; p = 0.36) and 24 weeks (–224, –2478 to 1213 vs –150, –1235 to 768; p = 0.63). No grade 3 or 4 adverse events were attributed to either the NNRTI or LPV/r.


Conclusions:  A switch from NNRTI- to LPV/r-based ART in HIV-infected Ugandan children resulted in high rates of continued virologic suppression and no associated grade 3 or 4 adverse events.

 

Theodore Ruel, J Achan, K Murray, M Kamya, D Havlir, and E Charlebois.Rising T Lymphocyte Activation Levels Are Associated with Diarrhea Incidence and CD4 Decline among ART-naïve HIV-infected Ugandan Children.Abstract presented at CROI, March 2011.


Background:   High T lymphocyte activation levels are predictive of disease progression in HIV-infected children. Compared to Europeans, Africans have higher T cell activation levels that are thought to result from increased infections, but longitudinal data are limited. In an observational cohort of 300 HIV-infected Ugandan children we characterized the change in T cell activation levels over time and the association with several incident infectious syndromes.

 

Methods:  CD4+ and CD8+ T cell activation levels were defined as the percentage co-expressing CD38 and HLA-DR and determined by 4-color flow cytometry of fresh peripheral blood mononuclear cells by convenience sampling. Children not receiving ART (per Ugandan guidelines) and with at least 2 values included. Change between first and last T cell activation level were evaluated by sign rank test. Slopes of T cell activation levels over time were generated by simple linear regression using all available T cell activation levels for each subject. Baseline and slope of CD4 count, CD4 percentage, and log plasma HIV RNA copies/mL (dynamic range <2.6 to 5.9), the incidence density of malaria, pneumonia, and diarrhea, and days with fever (of any etiology) were evaluated as predictors. Associations were measured by multivariate linear regression. Ranges are in parentheses.

 

Results:  A median 3 (2 to 6) T cell activation levels over a median 168 (78 to 338) days were examined in 128 children with median age of 6.4 (1 to 11) years. At baseline, CD4+ (median 13%, 4 to 60%) and CD8+ T cell activation levels (median 41%, 16 to 69%) were inversely associated with CD4 percentage (median 24, 3 to 44) (p <0.001 and p = 0.001, respectively), but not CD4 count (median 787, 152 to 1653 cells/mL) or HIV plasma RNA (median 4.9 log10 copies/mL, 2.6 to 5.9). CD4+ T cell activation levels increased in 82 (64%) children (p = 0.0009) with a fitted slope of 0.17 (–7.4 to 3.25) percentage points per month. CD8+ T cell activation levels increased in 88 (69%) children (p <0.0001) with a fitted slope of a median 0.86 (–24 to 7.3) percentage points per month. CD4 percentage slope (coefficient = –0.29, p = 0.008) and incidence density of diarrhea (coefficient = 3.3, p = 0.002) were independently associated with rise in CD4+ T cell activation levels.


Conclusions:  CD4+ and CD8+ T cell activation levels rose in the majority of a cohort of ART-naive Uganda children within 1 year, with increases in CD4+ T cell activation levels independently associated with incidence of diarrhea and declining CD4 percentage slope. While the clinical significance of these absolute changes in T cell activation levels is unclear, these data provide additional evidence of the ongoing and increasing inflammatory state in HIV-infected children not yet receiving ART.

 

J Harris, A Gasasira, T Ruel, E Arinaitwe, Moses Kamya, G Dorsey, D Havlir, J Homsy, J Tappero, and T Sandison. Incidence, Risk Factors, and Management of Severe Neutropenia in HIV-infected and Exposed Children in Uganda.

 

Abstract presented at CROI, March 2011.
Background:  Neutropenia is common among children in Sub-Saharan Africa. Studies have demonstrated an increased risk of neutropenia with trimethoprim-sulfamethoxazole (TS) prophylaxis and ART use among HIV-infected children. However, there are few data regarding the clinical significance of transient neutropenia to guide clinical decision-making for HIV-infected and -exposed children with severe neutropenia (absolute neutrophil count <750 cells/mm3).

 

Methods:  In retrospective analysis, we examined the incidence of severe neutropenia in 2 cohorts of Ugandan children:  a rural (Tororo) cohort of 57 HIV+, 194 HIV-exposed, and 100 HIV– children; and an urban (Kampala) cohort of 300 HIV+ children. We used multivariate analysis to assess severe neutropenia risk factors. We also compared probability of absolute neutrophil count recovery by 100 days and risk of malaria and other infections between children who continued and those who discontinued TS prophylaxis in the setting of severe neutropenia as per the decision of individual providers.

 

Results:  There were 209 episodes of severe neutropenia in 150 (23%, n = 651) children over 1453.5 person-years. In Tororo, children older than 1 year had lower risk of severe neutropenia (for ages 1 to <2, RR 0.46, 95%CI 0.31 to 0.70; for ages 2 to 3, RR 0.37, 95%CI 0.18 to 0.74). ART in Tororo (RR 4.45, 95%CI 2.49 to 7.95) and Kampala (2.63, 95%CI 1.56 to 4.43) and TS prophylaxis in Tororo (RR 1.97, 95%CI 1.17 to 3.30) were also risk factors for severe neutropenia. Over 95% of participants who developed neutropenia had absolute neutrophil counts >750 cells/mm3 by 100 days of recognition, irrespective of TS cessation. In Tororo, an area of high malaria transmission, participants had a 4 times higher incidence of malaria during TS cessation than when they were on TS (4.04 episodes per person-year vs 0.97, p = 0.003).

 

Conclusions:  In 2 cohorts of children in Uganda, severe neutropenia was common and associated with both TS and ART use in the HIV-infected and -exposed populations. However, cessation of TS did not increase the likelihood of recovery, and increased risk of malaria in the cohort from an area of high malaria transmission. Our findings support the WHO recommendation against laboratory monitoring of children taking TS prophylaxis and suggest that with sufficient monitoring the discontinuation of TS in HIV-infected and -exposed children with severe neutropenia may not be indicated.

 

J Homsy, A Kakuru, E Arinaitwe, H Wanzira, V Bigira, T Sandison, P Jagannathan, Moses Kamya, G Dorsey, and J Tappero. No Rebound Incidence of Malaria among HIV-exposed Uninfected Children after Discontinuation of Trimethoprim-Sulfamethoxazole Prophylaxis in Rural Uganda.

 

Abstract presented at CROI, March 2011.
Background:  As per WHO guidelines, HIV-exposed children throughout Africa are placed on trimethoprim-sulfamethoxazole (TS) prophylaxis until HIV infection can be excluded. TS is prescribed to protect from opportunistic infections and we have recently shown that TS prophylaxis yielded a 39% reduction in malaria incidence among HIV-exposed children after cessation of breastfeeding until 2 years of age. There is, however, no evidence as to the risk of rebound incidence of malaria when HIV-exposed uninfected (HEU) children are discontinued from TS.


Methods:  We enrolled a cohort of HIV-unexposed and HIV-exposed infants 1.5 to 9 months of age in an area of high malaria transmission intensity in Uganda. HIV-exposed infants were prescribed TS prophylaxis from enrollment until confirmation of HIV-uninfected status 6 to 8 weeks after breastfeeding cessation. All HEU children were then randomized 1:1 to discontinue or continue TS prophylaxis until 2 years of age. At 2 years of age, those who continued TS were re-randomized 1:1 to discontinue or continue TS prophylaxis until 4 years of age. Malaria was diagnosed when a child presented with a fever and a positive thick blood smear and was treated with artemisinin-combination therapy (ACT). Malaria incidence from 2 years of age was calculated as the number of antimalarial treatments per person-time at risk, and incidence rate ratios were calculated using negative binomial regression.


Results:  At 2 years of age, all HIV-exposed children had stopped breastfeeding for >12 months. Among 80 HIV-exposed infants randomized to stop TS after breastfeeding (group A), there were 613 malaria cases in 101 person-years (6.07/person-year). Among 45 HIV-exposed infants randomized to stop TS at 2 years of age (group B), there were 321 malaria cases in 55 person-years (5.82/person-year). Among 88 HIV-unexposed infants never on TS (group C), there were 713 malaria cases in 88 person-years (6.12/person-year). No difference was found between groups B vs A (IRR 1.05, 95%CI 0.89 to 1.24, p = 0.56) or C vs A (IRR 1.09, 0.92 to 1.28, p = 0.33), adjusting for place of residence and type of ACT.


Conclusions:  In our study, discontinuing TS prophylaxis after prolonging it until 2 years of age after breastfeeding cessation was not followed by rebound incidence of malaria, despite a very high overall incidence of malaria throughout follow-up. This result indicates that it may be safe and efficacious to continue malaria prophylaxis in HIV-exposed children after the end of HIV exposure until 2 years of age.

2010, Boston, MA, USA

Jane Achan, T Ruel, P Li, E Charlebois, T Liegler, M Kamya, D Havlir, and J Wong.Incidence of Early Virological Failure and the Evolution of Antiretroviral Drug Resistance Mutations in Ugandan Children.CROI, February 27th to March 2nd 2010, Boston.

 

Background:  With limited access to virologic testing and ARV-switch guidelines based on CD4/clinical criteria, African children are at increased risk for unrecognized virologic failure and the development of high level ARV resistance. There are limited longitudinal data about the evolution of resistance in non-subtype B HIV+ children. Using banked specimens from an observational cohort of Ugandan children, we sought to determine the incidence of early virologic failure (EVF), the evolution of ART-mutations, and the potential impact on second line therapy options.

 

Methods:  Cases of EVF, defined as HIV RNA >1000 copies/mL 6 to 9 months after ARV initiation were selected from all children with plasma HIV RNA data in that period.  Available banked plasma specimens were used for population sequencing of HIV-1 pol. Composite discrete genotypic susceptibility scores (dGSS) scores were calculated using the standard Ugandan second-line regimen and tested by time with Spearman rank correlation.

 

Results:  Of 126 children with median 746 follow-up days beyond 6 months of ARV, 18 (14%) experienced EVF; all had viremia that persisted throughout follow-up. Genotypic resistance testing was successful in 40 samples from 14 children. First-line ARV regimens for children were zidovudine/lamivudine (53%) or stavudine/lamivudine (47%) plus nevirapine (NVP). With increased time, lower dGSS scores predict worsening compromise of current (r = -0.44, P =0.004) and standard second-line therapies (r =-0.41, P =0.008).

 

Months of failure

 

0-5

6-11

12-17

18-23

24-30

NRTI

 

 

 

 

 

1 TAM

0

0

3

4

3

≥2 TAM

0

0

0

1

1

M184V

9

7

9

8

6

 

 

 

 

 

 

NNRTI

 

 

 

 

 

ETR, 1*

5

5

4

2

2

ETR, 2*

0

1

2

2

3

NVP/EFV

5

4

4

4

5

N=^

(10)

(7)

(9)

(8)

(7)

 

TAM: Thymidine-associated mutation; NVP/EFV:mutations affecting both; ETR:Etravirine; *ETR-related mutations, none had 3.  ^: total # of children with data in time-period.

 

Conclusions:  A significant portion of children starting ARV in Africa experience EVF and quickly develop ARV mutations.  Most have M184V and NNRTI-mutations within 6 months, TAM’s after 12, and 2 ETR-mutations by 30 months.  With persistent and unrecognized viral replication, African children with non-subtype B HIV EVF are likely to require more expensive and less easily accessible second-line drugs.  These data underscore the need for affordable methods to identify viral failure and drug resistance in this population.

 

Theodore Ruel, M Boivin, P Bangirana, H Boal, P Rosenthal, C Akello, M Kamya, D Havlir, and J Wong.HIV Subtype A Is Associated with Impaired Neuropsychological Performance Compared to Subtype D in ART-naïve Ugandan Children. CROI, February 27th to March 2nd 2010, Boston.

 

Background:  Infection with HIV subtype D is associated with increased dementia in Ugandan adults compared to subtype A. There are no data comparing neuropsychological function in African children infected with different subtypes of HIV.

 

Methods:  There were 102 HIV+ ART-naïve children 6 to 12 years old (median 8.6) were enrolled from an observational cohort in Uganda. All completed the Kaufman Assessment Battery for Children, 2nd edition (KABC-2), the Test of Variables of Attention (TOVA), and the Bruininks/Oseretsky Test for Motor Proficiency, 2nd edition (BOT-2). A real time polymerase chain reaction (PCR) -based assay with subtype-specific probes in 5 regions (gag, pol, vpu, env, gp-41) was used to determine env and “total” (≥2 regions, recombinant = discordant regions) subtype. The most recent CD4 (percentage, count) and plasma viral load (log10 copies/mL) were obtained from the study database. ANCOVA was used for multivariate comparison.

 

Results:  Env subtype was determined in 54 (37 A, 16 D, 1 C) children. Children with A and D had comparable age, gender, education, physical development, environment (Caldwell HOME scale), socio-economic status, CD4 status, and WHO Stage. However, children with A had higher viral load (median 5.1 vs 4.6, P =0.02) compared to D. Using age-adjusted scores, children with A performed more poorly than those with D on all measures, especially on KABC-2 Sequential Processing (Memory; P =0.03) and Simultaneous Processing (visual-spatial analysis; P =0.003), with marginal differences on Learning (P =0.06) and TOVA Visual Impulsivity (P =0.06). After adjustment for viral load, Simultaneous Processing remained significantly different (P =0.01) between subtypes. Among 64 children with total subtype determination (37 A, 2 AD, 2 AC, 1 CD, 22 D), those with A performed more poorly on Simultaneous Processing (P =0.02) and Learning (P =0.008), with marginal differences in TOVA Visual Impulsivity (P =0.06). Learning differences remained significant after adjusting for viral load (P =0.01). Recombinant strains had intermediate performance on all measures.

 

Conclusions:  In contrast to recent findings in adults, children infected with HIV subtype A demonstrated poorer neurocognitive performance compared to those with subtype D. Some, but not all, of the difference may be attributable to higher viral load. The discrepancy in subtype association with neurocognitive function may reflect inherent differences in the neuropathogenesis of HIV between children and adults.

 

Theodore Ruel, L A Eller, H Cao, E Charlebois, J Achan, M Kamya, J Wong, and D Havlir. HIV Subtype D Is Associated with Rapid CD4 Decline in ART-naive Ugandan Children. CROI, February 27th to March 2nd 2010, Boston.

 

Background:  HIV subtype D is associated with rapid CD4 loss compared to A in adults, but there have been no prospective data from children who have distinct and developing immune systems.

 

Methods:  HIV+ children not meeting age-specific criteria for ART initiation and having at least 3 CD4 counts participating in an observational cohort in Kampala, Uganda were studied. HIV subtype was determined with a real time PCR-based assay that utilizes subtype-specific probes in 5 different genome regions (env, gag, pol, vpu, gp-41) on banked plasma specimens. Patients received all care in the study clinic; CD4 count and plasma HIV-RNA were obtained every 12 weeks. ART was initiated per Ugandan/WHO guidelines. CD4 and CD8 activation were defined by % HLA-DR and CD38 co-expression on peripheral blood mononuclear cells from within 180 days of enrollment. Predictors of the slope of CD4 decline were examined using Kruskal Wallis and simple linear regression analyses.

 

Results:  In this study, 168 children had a median age of 6.4 years (IQR 5.0 to 8.0), CD4 count of 802 cells/mL (IQR 596 to 1116), plasma HIV RNA of 4.9 (log10 copies/mL, IQR 4.5 to 5.3) and WHO stage distribution (I:68, II:72, III:28). Subtype was determined in at least one region for 140 (83%) children, 50 D-containing (D-Con, includes recombinant strains), 85 pure A (single or multiple regions), or 5 other (C or CA recombinant). In univariate analysis, D-con strains were associated with a mean CD4 decline of 172 vs 72 cells/year for non-D-containing strains (P =0.019). Age less than 5 (P =0.012), low baseline CD4 (P <0.001) and advanced WHO stage (P =0.013) were associated with CD4 decline, while baseline CD4% (P =0.132), plasma HIV RNA (P =0.725), CD4 (P =0.604) and CD8 activation (P =0.792) were not. In multivariate linear regression (coefficient, 95% confidence interval), D-con subtype (-76, -137 to -14), higher baseline CD4 count (-0.19, -0.26 to -0.13), and advanced WHO Stage (48, 6.7 to 89) remained predictive of CD4 decline.

 

Conclusions:  In ART-naive Ugandan children >1 year of age, HIV subtype D was associated with more rapid CD4 decline compared to non-D strains that is not explained by differences in baseline CD4 count, CD4%, HIV RNA, CD4 or CD8 activation. Further study of recombinant strains and the association of subtype with CXCR4/CCR5 co-receptor tropism may yield insight into this subtype-specific increased pathogenicity.

 

Theodore Ruel, H Boal, P Bangirana, C Akello, P Rosenthal, M Kamya, E Charlebois, D Havlir, M Boivin, J Wong, and Children with HIV and Malaria Project (CHAMP). Neurocognitive Impairment in HIV+ African Children Not yet Eligible for ART.CROI, February 27th to March 2nd 2010, Boston.

 

Background:  Studies in the era prior to ART availability demonstrated neurocognitive delay in children with advanced HIV disease. There are few studies in African children evaluating neurocognitive function in early HIV disease. Some HIV+ patients endure high levels of viral replication while maintaining high CD4 counts. Our objective was to determine if HIV+ children not yet eligible to receive ART suffer neurocognitive impairment compared to age-matched HIV– children.

 

Methods:  We enrolled from ongoing Ugandan cohort studies, 86 HIV+ children not yet ART-eligible according to WHO guidelines, and 98 HIV– children. Age-specific and locally validated neurocognitive evaluation tools were utilized. Children aged 2 to 5 years underwent the Mullen Scales of Early Learning and the Color-Object Association Test (COAT). Children aged 6 to 10 years underwent the Kaufman Assessment Battery for Children second edition (KABC2) and the Bruininks and Oseretsky Test of Motor Proficiency, second edition (BOT2). Scores were age stratified and compared using the Mann-Whitney test.

 

Results:  Access to medical care and measures of quality of psychosocial environment were similar between groups. HIV+ children aged 2 to 5 years had a median age of 4.7, 847 CD4 cells/μL, CD4% of 27, and plasma HIV RNA 65,966 copies/mL. HIV+ children aged 6 to 10 had a median age of 8.2 years, 610 CD4 cells/μL, CD4% of 27, and plasma HIV RNA 52,015 copies/mL. The 2 to 5 and 6 to 10 strata of HIV– children had median ages of 4.9 and 8.3 years, respectively. HIV+ children aged 2 to 5 years had lower median scores in all measures of the COAT (2 of 2) and Mullen (5 of 5) modules. HIV+ children aged 6 to 10 had lower median scores in most of the BOT2 (7 of 8) and KABC2 (9 of 15) modules. Significant differences were noted in the 5 measures in the table.

 


 Ages 2 to 5

Fine Motor-Mullen (IQR)

Receptive Language-Mullen (IQR)

HIV+ n = 17

35.0 (33.0 to 37.0)

35.0 (28.0 to 42.0)

 

HIV– n = 28

40.0 (36.0 to 44.0)

39.0 (33.0 to 45.0)

 

p value

0.026

0.048

 

Ages 6 to 10

Gestalt Closure to KABC2 (IQR)

Word Order to KABC2 (IQR)

Speed Agility to BOT2 (IQR)

HIV+ n = 69

7.0 (5.0 to 13.0)

14.0 (12.0 to 16.0)

18.0 (15.0 to 20.0)

HIV– n = 70

10.5 (8.0 to 15.0)

15.5 (13.0 to 16.8)

20.0 (18.0 to 23.0

p value

0.0009

0.011

0.0021

 

Conclusions:  Motor and cognitive deficits were identified in HIV+ children 2 to 10 years of age ineligible for ART per current WHO guidelines. Further investigation of the predictors of these deficits, their social and functional effect, and whether early initiation of ART can prevent or reverse them is indicated.

 

Edwin Charlebois, S Mahan, P Srikantiah, C Lancioni, H Boom, C Whalen, A Okwera, R Mugerwa. M Walusimbi-Nanteza, and D Havlir. HIV Viral Dynamics and Immune Activation following Successful TB Treatment.CROI, February 27th to March 2nd 2010, Boston.

 

Background:  Infection with TB has been associated with increased HIV viral load in HIV infected persons. Following TB treatment, studies in the United States have shown decreases in viral load, whereas short-term studies in Africa have not shown reductions in viral load. Viral load dynamics have important implications for HIV disease progression, ART treatment decisions and HIV transmission.

 

Methods:  We measured change in HIV viral load (HIV-1 RNA copy number), CD4 cell count, and CD4 and CD8 cell activation in 61 HIV-infected smear-positive TB patients with CD4 >350 successfully treated for TB in Kampala, Uganda, with 1 to 2 years of follow-up. Change in laboratory measures was calculated within individuals and significance testing was done using the non-parametric Sign-Rank test. Successful TB treatment was defined as TB culture clearance at 2 to 6 months. CD4 and CD8 activation was measured as percent CD4 or CD8 RO+ cells co-expressing HLA-DR and CD38 by four color flow-cytometry.

 

Results:  At baseline (day 0 of TB therapy), median log HIV viral load was 4.4 (range 2.6 to 5.9) and median CD4 cell count was 523 (range 353 to 1247). No significant change in HIV viral load was observed at 6 or 12 months post start of TB therapy (median change in log viral load was +0.06 and +0.09 at 6 and 12 months, respectively, p>0.4 for both). Median change in CD4 cell count at 6 and 12 months was –30 and –77 cells (p = 0.05 and p = 0.01). Similarly, among 34 subjects followed for 24 months, no significant change was seen for log HIV viral load (median change = –0.09, p = 0.32) and CD4 counts were observed to decrease by 122 cells (p = 0.01). Among a subset of subjects with immune activation data, CD4 activation decreased from baseline in 12 of 17 subjects at 3 months of TB therapy (p = 0.01), 10 of 18 at 6 months (p = 0.14) and in 6 of 9 at 12 months (p = 010). CD8 cell activation decreased in 11 of 18 subjects at 3 months (p = 0.06), 14 of 17 at 6 months (p = 0.001), and in 8 of 10 subject at 12 months (p = 0.11).

 

Conclusions:  We found no evidence of significant decreases in HIV viral load following successful TB treatment either in the short term (at the completion of 6 months of TB therapy) or in the long term (1 and 2 years following the start of TB therapy). While TB treatment may have the effect of reducing immune activation, it did not result in reduction in HIV viral load and cessation of CD4 cell loss.

 

Jane Achan, W Pasutti, B Gati, T Ruel, E Charlebois, P Rosenthal, T Liegler, M Kamya, D Havlir, and J Wong. Drug Resistance with Virologic Failure in HIV+ Ugandan Children on ART: Consequences for Second-line Therapy. CROI, February 27th to March 2nd 2010, Boston.

Background:  Substantial progress has been made in deploying ART for HIV+ patients in Africa. However, absence of HIV viral load monitoring and drug resistance testing will likely facilitate the development of drug resistant HIV. Cumulative HIV drug resistance over a lifetime of infection will have particularly profound effects on treatment options for HIV+ children.

 

Methods:  Children receiving ART according to WHO guidelines with viremia after ≥6 months of ART were identified from an ongoing observational cohort study in Uganda. Although CD4 and viral load testing was done every 12 weeks, change to second-line ART was governed by CD4 counts and clinical events by WHO guidelines. Plasma samples from the latest time point with detectable viral load were tested by population sequencing of HIV-1 pol. Discrete genotypic susceptibility scores (dGSS) were calculated based on susceptibility to current ART and to 3 different NRTI backbones of second-line ART + lopinavir/ritonavir (LPVr) inferred by Stanford HIVdb algorithms.

 

Results:  Of 100 children, 22 (22%) receiving ≥6 months of ART had viral load >1000 copies/mL. Median duration of ART prior to testing was 24 months. All subjects were receiving either zidovudine (ZDV)/lamivudine (3TC) (95%) or stavudine (d4T)/3TC (5%) plus nevirapine (NVP). Of 22 subjects, 4 met immunologic criteria for second-line ART at the time of analysis. Genotypic resistance testing was successful in 19 subjects (86%). RT sequence mutations predicted to confer resistance to one or more ART agents was observed in all 19 subjects (100%), 2 class resistance was seen in 16 (84%). The most common mutations were M184V (16 of 19), Y181C/I (7 of 19), and K103N (7 of 19). Thymidine analogue mutations or other multidrug resistance mutations were observed in 8 of 19 cases (42%). Of 19 cases, 15 (79%) had dGSS <2 for the current regimen. The percentage of cases for which dGSS for empiric second-line NRTI + LPVr was predicted to be <2 or <3:  abacavir (ABC)/ddI = 37%, 89%; TDF/FTC = 21%, 89% and ZDV/3TC/TDF = 21%, 32%. All 4 subjects meeting immunologic criteria for second-line therapy with ABC/ddI/LPVr had dGSS score <2 for this second-line regimen.

 

Conclusions:  Among HIV+ Ugandan children with virologic failure, prevalence of drug resistance is high. Only 11% of the cases examined would be fully susceptible to ABC/ddI/LPV/r used in empiric second-line therapy in Uganda. Inability to recognize and respond to virologic failure at an earlier stage and the unavailability of clinical resistance testing may compromise subsequent therapy.


Jane Achan, A Gasasira, T Ruel, F Kateera, J Kalyango, E Charlebois, G Dorsey, P Rosenthal, M Kamya, and D Havlir.A High Risk of Pneumonia in HIV-infected African Children Persists as Long as 6 Months after Initiation of ART. CROI, February 27th to March 2nd 2010, Boston.

Background:  Respiratory tract infections are the most common cause of morbidity and mortality in HIV+ children. ART reduces the risk of invasive bacterial disease including pneumonia in HIV+ children. We previously reported a high rate of pneumonia in HIV+ children in the first 6 months of ART; to determine whether these were immune reconstitution events, we compared the incidence of pneumonia in the 6 months prior to ART initiation to that 12 months after.

 

Methods: Study subjects were 100 HIV+ Ugandan children ages 1 to 10 years in a prospective cohort study; who had been initiated on ART at least 6 months after enrolment. All participants received cotrimoxazole prophylaxis and routine childhood immunizations. A standardized definition for pneumonia was used with the following clinical criteria; fever, cough, tachypnea, and pulmonary infiltrates on chest X-ray. Incidence densities were compared among periods:  6 months prior, 6 months after, and 6 to 12 months after ART initiation.

 

Results:  At ART initiation, median age was 4 years, CD4 count was 340 cells/mm3 (12%) and HIV RNA was 5.4 log copies/mL. Children received either nevirapine (n= 38) or efavirenz (n= 46) ART-based regimens; 6 months after ART initiation, 85 % had <400 copies/mL HIV RNA and median CD4 count was 760 (23%). We observed 65 episodes of pneumonia in 27 participants prior to ART initiation compared to 62 episodes of pneumonia in 23 participants after ART initiation. There was no difference in incidence density of pneumonia 6 months before and 6 months after ART initiation, however the incidence density of pneumonia was lower 6 to 12 months after ART initiation compared to 6 months before ART initiation; p = 0.03. Hospitalization was required for 25 of the 127 episodes of pneumonias (13 pre-ART and 12 post-ART). No deaths occurred as a result of pneumonia. Pulmonary tuberculosis was not diagnosed during this period of follow-up.

 

  

Time before and after ART Initiation

 

≤6 months before

0 to 6 months after

6 to 12 months after

Person-time

192 person-months

212 person-months

202 person-months

# of cases

17

14

7

Incidence density

0.08/person-months

0.07/person-months

0.03/person-months

 

Conclusions:  In this prospective cohort of HIV-infected children receiving septrin, rates of pneumonia were high in the 6 months prior to ART and persisted up to 6 months after ART. Pneumonia rates decreased by 63% 6 to 12 months after the start of ART. Improved clinical management and prevention strategies for pneumonia are needed in this population.

 

Anne Gasasira, D Havlir, J Achan, E Charlebois, T Ruel, P Rosenthal, T Sandison, G Dorsey, and M Kamya. Prophylactic Effect of Trimethoprim-Sulfamethoxazole on Malaria in HIV-infected and -exposed Ugandan Children Living in Settings of High Antifolate Resistance. CROI, February 27th to March 2nd 2010, Boston.

Background:  Trimethoprim-sulfamethoxazole (TS) provides protection against malaria in HIV-infected children, but it is not known whether protective efficacy is maintained in settings where malaria transmission is high and antifolate resistance is extensive. We compared the efficacy of TS prophylaxis against malaria in HIV-infected, -exposed, and -uninfected Ugandan children at 2 sites:  Kampala, a medium malaria transmission setting, and Tororo, a very high malaria transmission setting. We further assessed the prevalence of molecular markers of antifolate resistance at both sites and assessed the change in TS efficacy over time at the Kampala site.

 

Methods:  Subjects were 809 children (517 HIV–, 292 HIV+) from Kampala, and 329 infants (99 HIV–, 190 HIV-exposed, 40 HIV+) from Tororo. Daily TS was prescribed for HIV+ participants and HIV-exposed, breastfeeding participants. HIV– unexposed participants did not receive TS. All participants were provided insecticide-treated bed nets (ITN). Standardized protocols were used to measure malaria incidence and identify 5 key antifolate resistance-mediating mutations. The observation periods were 18 and 9 months for Kampala and Tororo, respectively. The protective efficacy of TS in Kampala was compared for 3 consecutive 6-month periods.

 

Results:  TS was associated with a protective efficacy of 80% (0.12 vs 0.54 episodes PPY, p<0.001) and 73% (0.68 vs 2.74 episodes PPY, p<0.001) in Kampala and Tororo, respectively, despite prevalence in 87% and 96% of parasites of all 5 antifolate resistance-mediating mutations (dhfr 51I, 108N, C59R, and dhps 437G, 540E). TS use was not significantly associated with frequency of mutations at either site. The prophylactic effect of TS was similar among HIV+ and -exposed children. In Kampala, TS efficacy did not change significantly over the 3 consecutive 6-months periods (IRR 0.16, 0.09 to 0.27; 0.28, 0.16 to 0.46; and 0.24, 0.13 to 0.46).
Conclusions:  TS was highly effective for malaria prevention across very different malaria transmission intensities in regions with high-level antifolate resistance. TS efficacy was not observed to decline over an 18-month period. However, the malaria burden remains inadequately controlled by TS and ITN in high-transmission settings such as Tororo, Uganda.

 

Moses Kamya, A Gasasira, J Achan, T Ruel, E Charlebois, S Staedke, A Kekitiinwa, P Rosenthal, G Dorsey, and D Havlir. high Risk of Neutropenia in HIV-infected Children following Treatment with Artesunate + Amodiaquine for Uncomplicated Malaria in Uganda. CROI, February 27th to March 2nd 2010, Boston

Background:  Artemisinin-based combination therapies are rapidly being adopted for the treatment of malaria in Africa, yet data are limited on safety and efficacy among HIV-infected populations.

 

Methods:  We compared malaria treatment outcomes between cohorts of HIV-infected and HIV-uninfected children in Uganda followed for 29 and 18 months, respectively. Malaria was treated with artesunate + amodiaquine (AS/AQ) and outcomes assessed using standardized guidelines. HIV-infected children received trimethoprim-sulfamethoxazole prophylaxis and ART according to current guidelines.

Results:  Were included 35 malaria episodes among 26 HIV-infected participants, and 258 malaria episodes among 134 HIV-uninfected children; 12 HIV-infected children were receiving ART, 11 of whom were receiving zidovudine. Malaria treatment was highly efficacious in both the HIV-infected and HIV-uninfected cohorts (28-day risk of recrudescence 0% and 3.6%, respectively) however, there was a trend towards increased risk of recurrent malaria in the HIV-uninfected children (2.9% vs 13.2% respectively, p = 0.08). Importantly, the risk of neutropenia 14 days after treatment with artesunate plus amodiaquine was higher in HIV-infected compared to HIV-uninfected children (45% vs 6%, respectively, p<0.001). All neutropenia episodes in HIV-uninfected children were of mild to moderate severity while 16% of neutropenia episodes in the HIV-infected cohort were severe or life-threatening (<750/mm3). Among HIV-infected children, the risk of neutropenia was significantly higher in those receiving ART (75% vs 26%, p = 0.001). We compared the risk of significant clinical events in HIV-infected children during the period of neutropenia to HIV-infected controls not treated with AS/AQ over the same time. There was a trend towards a higher risk of any significant clinical event in subjects treated with AS/AQ compared to those not treated with AS/AQ (57% vs 35%, p = 0.16)

 

Conclusions: AS/AQ was highly efficacious for malaria treatment in HIV-infected children, but associated with a high risk of neutropenia, especially in the setting of concurrent ART. Our findings highlight an urgent need for evaluation of alternative antimalarial therapies in HIV-infected individuals.

 

Isaac Ssewanyana, M Elrefaei, G Dorsey, T Ruel, E Charlebois, A Gasasira, M Kamya, J Achan, D Havlir, and H Cao.Profile of T Cell Immune Responses in HIV-infected Children from Uganda.CROI, February 27th to March 2nd 2010, Boston

Background:  HIV immunopathogenesis in children remains poorly understood. We assessed T cell immune activation and HIV-specific responses in ART-naïve children in Uganda .

 

Method:  Samples were obtained from the study cohort Children with HIV and Malaria Project (CHAMP), an ongoing prospective observational study investigating interactions between HIV and malaria in co-infected children. Only ART-naïve volunteers were evaluated for this immunologic study (n = 154). Demographic information, CD4+ T cell status (percentage and absolute count) and HIV viral load were obtained at the time of blood draw. For comparison, samples from HIV-1+ adult volunteers (n = 81) obtained from an existing cross-sectional cohort study in Uganda. CD8 and CD4 activation was defined as the percentage of CD3+CD8+CD38+HLA-DR+ or CD3+CD8–CD38+HLA-DR+ lymphocytes, respectively. HIV-specific T cell responses were determined by the level of surface CD107 expression and the intracellular production of interferon-gamma following Gag peptide stimulation

 

Results:  Increased CD4+ and CD8+ T cell activation strongly correlated with decreased CD4 cell percentage (r = –0.617, p<0.001, r = 0.279, p = 0.004, respectively. Interestingly, no correlation between plasma HIV RNA and T cell activation was observed after controlling for CD4 count (p>0.05), in contrast to findings observed in HIV positive adults.  HIV-specific T cells were readily detectable and the presence of Gag-specific CD4+ T help in children was associated with increased HIV-specific effector CD8+ T cells.

 

Conclusions:  Our study is the first to report the T cell profile in HIV-infected children in Uganda. Understanding the balance between immune activation and T cell immunity in HIV infected children may provide further insights into the mechanisms leading to effective immune control.

 

Theodore Ruel, I Ssewanyana, J Achan, A Gasasira, M Kamya,A Kekitiinwa, J Wong, H Cao, D Havlir, and E Charlebois. Dynamics of T Cell Activation Accompanying CD4 Recovery in Antiretroviral-treated HIV-infected Ugandan Children.CROI, February 27th to March 2nd 2010, Boston

 

Background:  Little is known about the dynamics of T cell activation in HIV-infected African children. In a cross-sectional and prospective study, we evaluated the change in CD4 and CD8 activation with ART in the Children with HIV and Malaria Project (CHAMP), an observational cohort of HIV-infected Ugandan children, comparing to levels in age-matched HIV-uninfected children.

 

Methods:  Levels ofCD4 and CD8 cell activation, defined as the percent co-expressing CD38 and HLADR by 4-color flow cytometry, were compared among 4 subgroups of HIV+ children (based on ART and CD4% status) and with HIV– children using the non-parametric Wilcoxon rank sum test. ART was initiated per World Health Organization (WHO) guidelines and change in activation was prospectively evaluated in a subgroup from baseline to 10 to 14 weeks after ART initiation by Wilcoxon signed rank and Spearman rank correlation testing.

 

Results:  Levels of CD4 and CD8 cell activation in ART-treated children with CD4 recovery (CD4% ³25) were lower than in ART-naïve children (p<0.03 for CD4 or CD8, CD4% <25 or ³25) but higher than in HIV– children (p<0.001, CD4 or CD8). The same pattern was observed when stratified by age (1 to 6, 6 to 12 years) and when the 13 ART-treated children with persistent detectable plasma HIV RNA were excluded. In prospective study of 8 participants, level of CD4 activation decreased by median 8.7 percentage points (p = 0.01) and CD8 activation decreased by 17.2 percentage points (p = 0.01); the median increase in CD4% was 6.5 percentage points and was strongly correlated with lower post-treatment CD8 activation level (ρ = 0.73, p = 0.008).

 


  

HIV+, ART-naïve

HIV+, ART ³24 weeks*

HIV–

 

CD4% <25

CD4%³25

CD4% <25

CD4%³25

 

N

113

87

26

32

30

CD4 38+DR+

15.2
(11.8 to 19.9)

10.4
(7.1 to 13.2)

11.8
(9.5 to 14.8)

7.7
(5.6 to 11.2)

4.6
(3.1 to 9.1)

CD8 38+DR+

47.7
(39.0 to 56.7

38.9
(30.2 to 47.0)

29.5
(25.0 to 43.9)

27.8
(20.6 to 35.4)

17.7
(8.8 to 26.1)

Values are median percentage co-expression (IQR); *(median: 37.9 weeks, IQR 32.0 to 60.1)

 

 

Conclusions:  Even in the setting of high T cell activation, African children can respond to ART with robust CD4 recovery and rapid decreases in immune activation. The degree of CD4 recovery is associated with post-treatment CD8 activation levels.

 

Lisa Bebell, C Pilcher, G Dorsey, D Havlir, M Kamya, M Busch, P Rosenthal, C Nugent, C Bentsen, and E Charlebois. Acute HIV-1 Infection Is Highly Prevalent in Ugandan Outpatients Suspected of Malaria.CROI, February 27th to March 2nd 2010, Boston

 

Background:  Malaria is the most common clinical diagnosis at health clinics in sub-Saharan Africa, accounting for 30 to 50% of outpatient visits. The symptoms of acute HIV infection can closely mimic malaria. We investigated whether acute HIV infections could be identified among adult patients with suspected malaria in rural Uganda.

 

Methods:  A survey was conducted at 7 rural government health centers in Uganda of patients suspected to have malaria. HIV antibody testing was performed on whole dried blood spot (DBS) samples using antibody enzyme immunoassay (EIA) confirmed by Western blot and DNA polymerase chain reaction (PCR). EIA non-reactive and EIA reactive-but-unconfirmed specimens were pooled 9:1 and tested for HIV RNA by NAAT for detection of acute infections. The BED assay was performed on EIA/Western blot positive specimens for additional staging of confirmed antibody positive results. Staging of cases as acute HIV infection used a serological classification system of Fiebig et al.). Predictors of having serology consistent with acute HIV infection were identified using logistic regression.

 

Results:  Between May 2006 and February 2007, 2893 adults (≥13 years of age) were evaluated. Of 2893 adults, 324were confirmed as HIV+ (overall HIV prevalence 11.2%). Of HIV+ adults, 30 (9.3%), had evidence of early acute HIV infection (HIV RNA positive and a non-reactive/negative or indeterminate EIA or Western blot, Fiebig stages I-IV). An additional 47 (15%) of HIV+ cases had evolving Western blot patterns or BED ODn values <0.8, indicating recent infection (stage V). High HIV prevalence >10% at the site (OR = 5.4, p = 0.0018), lower malaria prevalence at the site (OR = 0.12, p = 0.034) and male sex (OR = 2.0, p = 0.05) were significantly associated with having early acute HIV infection (stages I-IV). Predictors of acute and recent HIV infection (stages I-V) were nearly identical to those for stage I-IV infections. Use of a more stringent BED cutoff for stage V (e.g., ODn <0.5) gave similar results.

 

Conclusions:  In Ugandan health centers, we found that an average of 1% of all adults with suspect malaria had serologic evidence of early acute HIV infection. At some sites, as many as one in 15 symptomatic adults had some serologic evidence of ongoing HIV seroconversion. These findings suggest opportunities for recognition of acute HIV infections in Africa on a scale that has not previously been considered possible.

 

Jaco Homsy, T Sandison, E Arinaitwe, N Vora, A Kakuru, H Wanzira, V Bigira, M Kamya, G Dorsey, and J Tappero. Efficacy of Trimethoprim-sulfamethoxazole Prophylaxis against Malaria in HIV-exposed Uninfected Children until 2 Years of Age: A Randomized Clinical Trial. CROI, February 27th to March 2nd 2010, Boston

 

Background:  Trimethoprim-sulfamethoxazole (TS) prophylaxis is recommended by WHO and used throughout Africa to prevent opportunistic infections in HIV-infected and HIV-exposed uninfected children. However, there are no recommendations and no randomized data regarding TS prophylaxis in HIV-exposed children beyond the period of HIV exposure. In an area of high malaria transmission in Uganda, we are conducting the first randomized clinical trial of TS protective efficacy among HIV-exposed children. 

 

Methods:  We enrolled 203 HIV-exposed infants 1.5 to 9 months of age and, in addition to giving them insecticide-treated bed nets, prescribed per WHO recommendations daily TS prophylaxis for each child until confirmation of negative HIV status by PCR 6 to 8 weeks after breastfeeding cessation. Children were then randomized to discontinue or continue TS prophylaxis through age 2 years. Adherence to insecticide-treated bed nets and TS was assessed by nurses at a dedicated study clinic that provided comprehensive care to the children. Malaria was diagnosed when a child presented with fever and a positive thick blood smear and treated with artemisinin-combination therapy. The association between TS use and malaria incidence was estimated as an incidence rate ratio (IRR) using negative binomial regression.

 

Results:  Median age at enrolment was 3.7 months (95%CI 2.4 to 6.6). Adherence to insecticide-treated bed nets was 98%. Adherence to TS was >98% among children and their mothers. Among 203 HIV-exposed infants enrolled, 185 were randomized to continue or discontinue TS prophylaxis. Among 98 infants who continued TS, there were 291 malaria cases in 101.92 person-years (2.86 cases/person-years). Among 87 infants who discontinued TS, there were 397 malaria cases in 86.01 person-years (4.62 cases/person-years). TS prophylaxis yielded a 38% reduction in malaria incidence (IRR = 0.62, 95%CI 0.49 to 0.78, <0.001).

 

Conclusion:  Our data suggest that TS prophylaxis is protective against malaria in HIV-exposed children when continued beyond the period of HIV exposure. Previous observational data have reported a higher degree of protection possibly due to selection bias, differences in antifolate resistance and transmission intensity, or other unmeasured confounders. Preliminary data will be presented on further randomization of the children who continued TS to continue or discontinue TS for another 2 years in order to assess the efficacy of continued prophylaxis and the possibility of rebound effect following TS cessation.

Jaco Homsy, N Vora, E Arinaitwe, T Sandison, A Kakuru, H Wanzira, J Kalamya, M Kamya, J Tappero, and G Dorsey.Breast Feeding Reduces the Risk of Malaria in Children Born to HIV-infected Mothers. CROI, February 27th to March 2nd 2010, Boston

Background:  Malaria is the leading cause of death among African children under 5. Breast feeding protects infants against major infectious diseases, but data on its effect on malaria are limited.

 

Methods: We assessed whether breast feeding was protective against malaria in a prospective cohort of children born to HIV-infected mothers in an area of high malaria transmission in Uganda. Infants were tested for HIV at 6 weeks of age and 6 weeks after cessation of breast feeding as per national guidelines. Mothers of HIV-uninfected infants were counseled to exclusively breast feed for 6 months. Mothers of HIV-infected children were counseled to breast feed for as long as possible. Children were enrolled between 6 and 9 months of age and followed to 15 months of age. All children were given cotrimoxazole prophylaxis throughout the study and insecticide-treated bed nets at enrollment. Malaria was diagnosed when a child presented with a fever and a positive blood smear. Date of breast feeding cessation was determined using monthly questionnaires. Kaplan-Meier survival and Cox proportional hazard were used to model the association between breast feeding and the daily risk of malaria adjusting for age and repeated measures.

 

Results:  Of 200 HIV-exposed uninfected children, 175 (88%) had stopped breast feeding by 15 months of age, as had 10 (22%) of 45 HIV-infected children. The median age at breast feeding cessation for these 175 and 10 children was 7.2 months (range 3.9 to 14.8) and 5.9 months (2.4 to 13.9), respectively (see the first table). Due to significant effect modification by age, data were stratified into 2 age groups (6 to <9 months and 9 to 15 months). Among children aged 6 to <9 months, breast feeding was not associated with a significant malaria risk reduction (HR = 086; p = 0.57). However, breast feeding was associated with a significant malaria risk reduction among children aged 9 to 15 months (HR = 0.51; p = 0.003) (see the second table). These results remained unchanged when stratifying by HIV status.

 


 p = 0.007

http://www.retroconference.org/2009/Abstracts/33618_files/image001.gifhttp://www.retroconference.org/2009/Abstracts/33618_files/image002.gif
 Number of children at risk


Age

6
months

9 months

12
months

15
months

Not breast feeding

28

106

82

42

Breast feeding

139

77

45

17


 


Age group (months)

Breast-feeding status

Person-time (years)

Malaria incidence (per person-years)

Hazard ratio (95%CI)

p-value

6 to <9

No breast feeding

22.0

1.64

0.86
(0.51 to 1.45)

0.57

Any breast feeding

23.8

1.13

9 to 15

No breast feeding

40.3

2.70

0.51
(0.33 to 0.79)

0.003

Any breast feeding

22.9

1.35

 

 Conclusions:  Breast feeding reduced the risk of malaria by 50% in HIV-exposed Ugandan children aged 9 to 15 months. This finding adds to the body of evidence suggesting that in resource-limited settings, HIV-exposed children may need to breast feed for >6 months in order to avoid the morbidity and mortality risks associated with early weaning.

 

Theodore Ruel, J Achan, A Gasasira, E Charlebois, P Rosenthal, G Dorsey, M Kamya, A Kekitiinwa, J Wong, and D Havlir. Dramatic Reductions in HIV RNA among HIV-infected Children with Acute Measles in Uganda.CROI, February 27th to March 2nd 2010, Boston

 

Background:  Many HIV-related co-infections cause increases in plasma levels of HIV RNA, but several agents have been reported to cause transient decreases. During the ongoing measles epidemic in Uganda, we evaluated HIV viral dynamics in the Children with HIV and Malaria Project (CHAMP), an observational cohort of 300 HIV-infected children between 1 and 10 years of age. 

 

Methods:  We identified all children in the cohort who met the WHO clinical case definition of measles between June and September 2006. HIV plasma RNA viral load, absolute lymphocyte count, absolute CD4 count, and CD4 percentage and were evaluated before, on the day of diagnosis, and after clinical measles resolution. The log of plasma viral load and other parameters were compared using a paired 2-tailed Student’s t-test.

 

Results:  Results are shown in the table for the9 children who met the measles case definition; 6 were ART-naïve. HIV plasma viral load declined significantly (p<0.0001) during clinical measles by a median of 1.4 log in the ART-naïve children. Absolute CD4 count (p = 0.06) and absolute lymphocyte count (p = 0.04) also declined, but CD4 percentage remained stable (p = 0.46). Notably, the plasma viral load of subject 5 dropped 1.4 log despite a rise in absolute lymphocyte count and absolute CD4 count.  Similar changes were observed in the children on ART, although recent commencement of therapy in subject 9 limits interpretation. Post-measles assessments are available in 2 children to date. HIV plasma viral load of subject 2 returned to 49,227 copies/mL 46 days after diagnosis; plasma viral load of subject 4 was 248,646 copies/mL at 28 days. 

 

Conclusions:  Acute measles is associated with significant and transient reductions in HIV RNA plasma levels. These data argue against target cell depletion as the exclusive explanation, suggesting instead direct viral and cytokine-mediated interference with replication as potential mechanisms.

2009, MONTREAL, CANADA


Ruel T, Boal HE, Bangirana P, Akello C, Rosenthal PJ, Kamya MR, Charlebois ED, Havlir, DV, Boivin, MJ, Wong, JK, and Children with HIV and Malaria Project (CHAMP). Neurocognitive Impairment in HIV+ African Children Not Yet Eligible for Antiretroviral Therapy (ART). CROI, February 8-11, 2009, Montreal.


Achan J, Gasasira A, Ruel T, Kateera F, Kalyango J, Akello Caroline, Charlebois E, Dorsey G, Rosenthal P, Kamya M, Havlir D. A high risk of pneumonia in HIV infected African children persists up to six months after initiation of antiretroviral therapy (ART). CROI, February 8-11, 2009, Montreal.


Gasasira A, Diane Havlir, Jane Achan, Edwin Charlebois, Theodore Ruel, Philip Rosenthal, Taylor Sandison, Grant Dorsey and Moses Kamya. Prophylactic effect of trimethoprim-sulfamethoxazole against malaria in HIV-infected and exposed Ugandan children living in settings of high antifolate resistance. CROI, February 8-11, 2009, Montreal.

2008, BOSTON, MA, USA


German, P. Parikh S, Lawrence J, Lindegårdh, N, Rosenthal P, Havlir D, Charlebois, E. Dorsey G, Aweeka, F. Drug Interaction between Antimalarial Drugs and Lopinavir/r. CROI 2008, Boston.


Ruel T, Ssewanyana I, Achan J, Gasasira A, Kamya M, Adeodata Kekitiinwa A, Wong J, Cao H, Havlir D, Charlebois E. Dynamics of T-cell activation accompanying CD4 recovery in antiretroviral treated HIV-infected Ugandan children. Accepted by 15th CROI, Boston, February 3-6, 2008.


Bebel L, Pilcher C, Dorsey G, Havlir D, Kamya M, Busch M, Nugent C, Bentsen C, Charlebois E. Acute HIV-1 infection is highly prevalent in Ugandan Outpatients Suspected of Malaria. CROI 2008, Boston.

2007, LOS ANGELES, CA, USA


Gasasira AF, Kamya MR, Achan J, Mebrahtu T, Theodore R, Kekitiinwa A, Charlebois ED, Rosenthal PJ, Havlir D, Dorsey G . "The effect of co-trimoxazole (CTX) prophylaxis and insecticide-treated bednets (ITN) on the risk of malaria among HIV infected Ugandan children. Presented at CROI," Los Angeles, February 25-28, 2007.


Theodore RD, Achan J, Gasasira AF, Charlebois ED, Rosenthal PJ, Dorsey G, Kamya MR, Kekitiinwa A, Wong J, Havlir D, and the CHAMP Team. "Abstract #R123: Dramatic reductions in HIV RNA among HIV-infected children with acute measles in Uganda." Presented at CROI, Los Angeles, February 25-28, 2007.

2006, DENVER, CO, USA

Padmini Srikantiah, R Lin, M Walusimbi, A Okwera, H Luzze, C Whalen, W Boom, D Havlir, and E Charlebois. Elevated HIV Seroprevalence and Risk Behaviors among TB Suspects in Uganda: Implications for HIV Transmission and Testing. Conference of Retroviruses and Opportunistic Infections. Abs.#798