icon-folder.gif   Conference Reports for NATAP  
  ID Week
Oct 8-12 2014
Back grey_arrow_rt.gif
HIV at ID Week 2014
  October 8-12, Philadelphia, PA
David H Shepp, MD
Associate Professor of Medicine
Hofstra-North Shore LIJ School of Medicine
Antiretroviral Therapy (ART). For many years, efavirenz (EFV) has been one of the most useful and widely prescribed antiretroviral agents (ARV) because of its potency, favorable pharmacokinetics, and availability in a co-formulated single-tablet regimen. It has been used successfully by many patients despite bothersome central nervous system adverse effects that occur frequently, but usually resolve within several weeks. Controversy exists about the role EFV plays in exacerbations of depression, a common comorbidity in HIV-infected patients. A recent analysis of patients randomized in clinical trials to EFV-containing or EFV-free ART found an increased risk of suicidality associated with EFV [1]. To further evaluate the risk of suicidality in EFV recipients, Nkhoma et al. analyzed a large health insurance claims database from the years 2006-2013 for suicidal ideation or attempted suicide events among HIV-infected individuals [2]. Approximately 20,000 with commercial insurance and 5,000 with Medicaid who were prescribed ARVs were included in the analysis. For the combined end-point of either event, there was no association between EFV prescription, and in fact, within the Medicaid subset rates were slightly lower with EFV use. However, there was evidence of channelling bias leading to significantly less use of EFV in those with psychiatric comorbid diagnoses or using psychoactive medication. A multivariable analysis with propensity score adjustment still found no association. Because very low numbers of suicide attempts were identified (0.6-6/1000 person-years) an expanded definition that captures various codes suggesting self-injury was also used, but again no relation with EFV use was found. An important limitation of this analysis is that completed suicides were not captured, since they do not result in health care claims. The authors also stated that the claims codes used for suicide attempt have been validated in studies using medical record review, but not those for suicidal ideation. This study does not confirm the association between EFV use and suicidality reported by in randomized trials. The findings are reassuring that when EFV use is avoided in those with psychiatric comorbidities, an increased risk of suicidality is not seen, and support continuing this pattern of EFV prescription.
IDSA: Using Real World Data to Assess the Risk of Suicidality among Patients Initiating an Efavirenz versus an Efavirenz-Free Antiretroviral Regimen - (10/15/14)
Selection of ART in Clinical Practice. The current DHHS guidelines offer a choice of 10 recommended regimens for initial of ART. All have been well studied and have similar efficacy, tolerability, and safety in appropriately selected patients. To understand how well-informed clinicians choose among the many good options available, Saag et al reviewed the choice of ART anchor drug among 1215 patients initiating an ART regimen at 8 academic clinical practices [3]. The study period lasted from July, 2009 to December 2012. The analysis focused on the influence on ART choice of demographics, insurance type, baseline CD4, baseline viral load and co-morbid diagnoses. Multivariable models were used to adjust for confounding. Women and those with either depression or liver disease were 40-50% less likely to receive non-nucleoside RT inhibitors (NNRTI). Women were twice as likely to receive boosted protease inhibitors (PI), while those with depression were more likely to get either boosted PIs or integrase inhibitors (INI). Patients with cardiovascular co-morbidities were 2.7 times as likely to get INIs. The presence of diabetes or hypertension, important risk factors for cardiovascular disease, did not significantly alter regimen choice, but there was a trend toward more INI use. The authors did not collect information on the clinical reasoning behind the observed differences. However, it is likely that lower NNRTI and greater PI use in women reflects concerns about drug exposure during potential pregnancy, since the NNRTI EFV should be avoided and the boosted PIs lopinavir and atazanavir are preferred. Lower NNRTI use among patients with depression also reflects concerns about EFV causing exacerbations, while the preference for INIs in cardiovascular disease likely reflects the neutral lipid effects of the INI raltegravir and the lack of drug-drug interactions between this medication and many commonly used drugs for cardiovascular disease. This study provides valuable insight into prescribing decisions for initial ART regimens. Not all of the regimens recommended in the most recent revision of the DHHS ART guidelines were available during the study period and choice of nucleoside backbone was not analyzed. An updated analysis showing how ART regimen choice evolves with the recent introduction of several NNRTI- and INI-based options, and how backbone choice is influenced would be of interest.
IDSA: Factors associated with the selection of initial antiretroviral therapy: Real-world channeling - (10/15/14)
Investigational ARVs. BMS 6633068 (068) is an investigational ARV with a novel mechanism of action. It is a prodrug for the active moiety which blocks HIV entry by binding to the HIV envelope protein gp120, thereby preventing HIV attachment and entry. The drug has intrinsic activity against most but not all HIV strains. Brinson et al presented a reanalysis of a phase 2b study of 068 [4] to look for patterns of response among various patient subsets [5]. The trial compared 4 doses of 068 to atazanavir/ritonavir, each given with tenofovir DF plus raltegravir. Eligible participants were treatment-experienced, on or off ART, with HIV-RNA levels >1,000 and HIV isolates susceptible to all study drugs. Based on earlier studies, susceptibility criteria included an IC50 <100nM in a phenotypic assay using the active form of 068. At 24 weeks, 69-80% achieved a viral load <50 with no apparent 068 dose-response relationship, while the response in the atazanavir control arm was 74%. Analysis of viral and CD4 response by age, gender, race did not appear to vary greatly but since each treatment arm contained only about 50 patients, the numbers in each subset were very small. As has been seen in many ART trials, there was a trend toward lower viral responses in those with a baseline viral load >100,000 or a baseline CD4 <200. Similar subset results were seen for the CD4 response. Safety of 068 was addressed in a separate poster [6]. The drug was generally well tolerated with was a low rate of discontinuation for adverse events, no discontinuations attributed to 068 and no dose-related pattern of adverse events. This novel attachment inhibitor has promising activity and safety. Further development is planned. An important unanswered question is whether clinical use of this drug should be preceded by a specialized phenotypic assay to document susceptibility, a requirement that has blunted the clinical use of CCR-5 inhibitors.
IDSA: HIV-1 Attachment Inhibitor Prodrug BMS-663068 in Antiretroviral-Experienced Subjects: Week 24 Subgroup Analysis - (10/10/14)
Compared to twice-daily dosing, once-daily dosing of ART is preferred by patients and leads to modestly improved adherence, although treatment outcome is not clearly affected. Of the 10 regimens currently recommended for initial ART in the DHHS guidelines, all are administered once daily except for the regimen containing raltegravir, which is dosed twice daily. A previous trial of once-daily dosing of raltegravir using the currently approved dose (800 mg) and formulation did not establish non-inferiority to twice daily dosing. To offer a once daily dosing option, the manufacturer is now studying the efficacy and safety of a new formulation of raltegravir at the dose of 1200 mg daily. This dose was reported previously to result in 24 hour trough drug levels about 60% of those achieved at 12 hours with twice daily dosing [7]. To explore the safety and efficacy of higher doses, Krishna et al gave raltegravir 1800 mg daily for 28 days to healthy volunteers [8]. Trough levels were similar to those previously reported after dosing for 5 days with 1200 mg. Two volunteers developed transient ALT and AST elevations which did not result in treatment discontinuation. Therefore, it appears the approval of a once-daily dosing option for raltegravir will depend on the outcome of the current trial using 1200 mg. Higher doses would not appear to offer any advantage.
IDSA: A 28-Day High-Dose Safety and Pharmacokinetics Study of Raltegravir in Healthy Subjects - (10/15/14)
HIV Care Cascade. With contemporary ART, viral suppression rates in clinical trials reach 90%, but in the real world, management of HIV is less successful. Many patients do not achieve viral suppression because they are undiagnosed, have never been linked to HIV care, are not retained in care, are not prescribed ART or fail treatment. The CDC estimates as few as 25% of all people infected with HIV in the US have been successful at each of these steps. About 18% of HIV infections are undiagnosed. The traditional approach to diagnosis requires self-referral for counseling, testing and referral for further care (CTR). An alternate social network strategy (SNS) uses peers to recruit high risk individuals. Members of a high risk social network are offered a small financial incentive (typically $10-$20/referral) to recruit other network members for testing. Compared to CTR, SNS has been shown to increase the proportion of those testing positive for HIV. However, use of a financial incentive could lead to lower risk individuals being referred over time as the pool of high risk individuals within the network becomes exhausted. For this reason, some SNS programs limit the number of individuals referred by a single recruiter. To examine this concern and to confirm the efficacy of SNS, Westergard et al. reviewed results from 48 testing sites in Wisconsin [9]; 14 used SNS and 34 CTR. The rate of HIV-positivity at the SNS sites was significantly higher (64/2620; 2.4%) than at the CTR sites (580/63,073; 0.9%). The investigators developed an HIV risk a score to assess the appropriateness of each referral and studied the relationship between the risk score and the number of individuals recruited and the time since first referral. The risk score of referred individuals did not decline with either the number of referrals or the length of time that recruiters continued to refer. Larger referral networks were more likely to include HIV-positive individuals. When analyzed by days elapsed until referral, after an initial modest decline, the risk score of referred individuals increased starting around 200 days from initial recruitment. This study shows the SNS strategy can enrich the testing pool for HIV-positive individuals. However, it is worth noting that while 29% of sites used SNS, they performed far fewer tests, contributing only 10% of the HIV diagnoses, and to achieve this result utilized 536 peer recruiters. Since the HIV risk of referred clients did not decline with larger numbers and actually increased in those who continued to recruit for long periods, there does not appear to be a need to limit the scope or duration of activity of peer recruiters. Those individuals who continued to recruit became more effective over time.
The third critical step in the HIV care cascade is retention in care. To predict patients who are at risk to miss appointments, Woodward et al. analyzed a patient database and clinical records at a Vanderbilt University clinic in Nashville [10]. From August, 2013 to March, 2014, they identified 510 patients with an upcoming scheduled appointment and a viral load >200 on last determination. Using a previously described system, each patient was scored 1 point for each of 7 characteristics; poor adherence (self report of ≥1 missed dose/week) in the past year, missed clinic appointments (>1 in the past year), substance abuse (in the past year), CD4 <100 (in the past year), history of 3-class ART exposure, history of viral failure with genotype, and current VL >200. Those with 0-1 points were considered low risk, 2-4 medium risk, and ≥4 high risk. During follow-up, 193 patients missed their appointment. Multivariable analysis adjusted for age, race, sex, HIV risk group, baseline CD4 and viral load identified 3 factors significantly associated with missed appointments: black, non-hispanic race, medium risk score and high risk score (aOR 2.5, 3.9 and 9.5 respectively). This study validates a simple, clinically accessible tool to identify patients at risk for missed appointment. Effective interventions targeted to these high risk patients need to be tested.
The care cascade estimates published by the CDC probably provides a worse case scenario, it they may not capture all data or account for patients who remain in care outside of the CDC surveillance system. Other HIV care systems and certain subsets of HIV-infected individuals may do substantially better. The Kaiser-Permanente Health System (K-P) is the second largest provider of health care to HIV-infected individuals in the US with over 22,000 HIV-infected members. Compared to CDC data, previous reports demonstrated patients within K-P system achieved substantially higher success rates at multiple steps in the care cascade. To understand the HIV care cascade outcomes for subsets of individuals within the K-P system, Horberg et al. analyzed their database by age and gender from 2010 to 2012 [11]. The investigators used definitions of linkage to care, retention in care, prescription of ART and viral suppression similar to those used by CDC, but could not estimate the proportion with undiagnosed HIV, so their care cascade begins with all diagnosed patients. For 2010 and 2011, men were more likely than women to begin ART and to achieve viral suppression. However, by 2012 these differences were no longer apparent, as 65% of women and 67% of men had a viral load <200. Women were more likely than men to be retained in care in all 3 years. More significant differences were seen in the analysis by age. Although rates of success increased over the 3 years of study for all age groups, younger patients (<35) had lower levels of retention in care, ART initiation and viral suppression than older patients. For 2012 the success rates for viral suppression were 57%, 65%, and 72% for ages <35, 35-54, and ≥55, respectively. This analysis used the CDC method, in which success at each step is dependent on success at the last step. However, the K-P investigators have found that analyzing each step separately gives somewhat higher success rates. For example, some patients in long-term care may not visit their health care provider twice yearly (and therefore would not be considered retained in care), but still remain on ART with viral load suppression. Analyzing each step independently added 13-16% to viral suppression rates. This study shows that HIV care outcomes are different within different care systems and among different subsets of HIV-infected individuals. It is not surprising that a system that includes a large number of privately insured patients fares better than the HIV population at large. It also highlights the problem with lower rates of treatment success with younger patients, even within a system with more insured patient and many resources. These younger patients should be targeted for interventions to improve retention in care, ART initiation and treatment adherence.
IDSA: HIV Care Cascade Differs by Gender and Age in Large US Healthcare System (10/10/14)
Perinatal HIV. Transmission research has yielded the puzzling observation that HIV-infected women who exclusively breast-feed their infants transmit less often than those who practice mixed feeding. This finding seems counter-intuitive, as one might expect a larger exposure to infected breast milk in infants given no other source of nutrition. To help understand this paradox, Wood et al. investigated the hypothesis that early introduction of food facilitates HIV transmission through immune activation in the oral cavity [12]. The study population included 156 breast-feed infants of HIV-uninfected mothers in South Africa who were studied for evidence of immune activation in oral mucosal swabs and peripheral blood. Non-breast milk food was introduced at the discretion of the mothers. At 14 weeks of age, infants receiving mixed feeding (about 3/4 of the study population) had a significantly higher level of activated CD4+ lymphocytes in peripheral blood. At the same time, oral mucosal swabs from mixed fed infants also showed more than a 2-fold greater level of activated CD4+ lymphocytes that had a gene expression pattern consistent with Langerhans cells, an antigen-presenting cell susceptible to HIV infection. These cells produced high levels of CCL22 and CCL5 (RANTES), two chemokines that help recruit lymphocytes to skin and mucosa. These findings support the hypothesis that introduction of non-breast milk foods into the diet of young infants triggers an immune response that increases both the number of HIV susceptible target cells in the oral mucosa and the recruitment of activated CD4+ lymphocytes into the submucosa. Assuming the same pattern occurs in the infants of HIV-infected mothers, more efficient HIV transmission would results despite the probability of lower exposure to infected breast milk in mixed-fed infants.
Bone Disease in HCV/HIV. Osteopenia and osteoporosis are common problems in HIV-infected individuals. Conventional risk factors such as low body weight, hypogonadism, smoking, alcohol and drug use account for some of the bone mineral density (BMD) loss in HIV, but HIV itself may play a role. Other chronic inflammatory diseases, such as Crohn's disease and rheumatoid arthritis are associated with low BMD characterized by loss thinning of cortical bone. Patients with HIV/HCV co-infection have lower BMD than HIV or HCV mono-infected patients but DEXA scanning, the tool used clinically to diagnose osteoporosis, does not distinguish the pattern of bone loss. To determine if the pattern of bone loss in HIV/HCV co-infection fits the pattern seen in other inflammatory diseases, Lo Re et al. used peripheral quantitative CT scanning of the tibia in women with HCV/HIV, HCV mono-infection or HIV mono-infection [13]. Compared to a reference group of healthy women previously scanned in a nutrition study, co-infected women had significantly lower trabecular volumetric bone mineral density (vBMD), cortical thickness, cortical area and higher endosteal circumference, a pattern of cortical thinning as seen in other inflammatory states. The observed differences persisted after statistical adjustment for several important difference between the groups, including age race and smoking. Co-infected women also had significantly lower trabecular vBMD and higher plasma TNF-alpha levels, than either HIV or HCV mono-infected women but cortical bone dimensions, vitamin D, IL-1 and IL-6 levels did not differ among these groups. Among all HCV-infected women, those with more advanced liver fibrosis (metavir F3-4) had significantly lower vBMD, cortical thickness and cortical area than those with lesser stages of fibrosis. These findings are consistent with the hypothesis that lower BMD in HCV/HIV women are due to cortical thinning and may result from higher levels of chronic inflammation. Data comparing the mono-infected groups to health controls was not presented, so it's not clear if this picture is unique to co-infection or simply more severe, or if the same pattern of bone loss occurs in men. In this study, all HCV-infected women were untreated. Further studies should help determine if HCV treatment can improve bone health in HIV/HCV co-infection.
IDSA: Structural Bone Deficits in HIV/HCV, HCV-Monoinfected, and HIV-Monoinfected Women - (10/14/14)
ID Week (HIV & HCV)
Oct 8-12 2014
1. Mollan KR, et al. Ann Intern Med. 2014;161:1-10.
2. Nkhoma et al, ID Week 2014 Philadelphia PA, abstract 646.
3. Saag M, et al. ID Week 2014 Philadelphia PA, abstract 1555.
4. Lalezari J, et al. CROI 2014, Boston, MA, March 3-6, abstract 86.
5. Brinson C, et al. ID Week 2014 Philadelphia PA, abstract 540.
6. Lalezari J, et al. ID Week 2014 Philadelphia PA, abstract 1574.
7. Rizk ML, et al. CROI 2014, Boston, MA, March 3-6, abstract 523.
8. Krishna R, et al. ID Week 2014 Philadelphia PA, abstract 1575.
9. Westergard R, et al. ID Week 2014 Philadelphia PA, abstract 84.
10. Woodward B, et al. ID Week 2014 Philadelphia PA, abstract 86.
11. Horberg M, et al. ID Week 2014 Philadelphia PA, abstract 88.
12. Wood et al, ID Week 2014 Philadelphia PA, abstract 642.
13. Lo Re V, et al. ID Week 2014 Philadelphia PA, abstract 645.