What We’re Reading

Each issue of the Topics, Trends & Updates email includes a summary of a recently published journal article, written by Christopher J. Hoffmann, MD, MPH, Associate Professor of Medicine,  Johns Hopkins University School of Medicine, Division of Infectious Diseases; Director, JHU-NYSDOH AI Guidelines Program. 

Articles are chosen to highlight topics that are directly relevant to clinical practice in the United States and that are not already widely summarized by other authors. 

All titles are listed below. If you would like to receive the Topics, Trends & Updates email, please subscribe







Weight Gain Associated With TAF (5/21)

Christopher J. Hoffmann, MD, MPH, Johns Hopkins University School of Medicine; Director, JHU-NYSDOH AI Guidelines Program

Weight gain and metabolic complications are increasingly a concern for patients taking antiretroviral (ARV) medications. The connection with a specific class of ARVs was first described for integrase strand transfer inhibitors (INSTI) [Norwood J, Turner M, Bofill C, et al. J Acquir Immune Defic Syndr. 2017; 76:527]; subsequently, an association with tenofovir alafenamide (TAF) was reported [Venter W, Moorhouse M, Sokhela S, et al. N Engl J Med. 2019;381(9):803] was reported. Now there is a recently published analysis of the Swiss HIV Cohort Study that compares weight changes and metabolic outcomes among people on stable antiretroviral therapy (ART) who were switched from tenofovir disoproxil fumarate (TDF) to TAF [Surial B, Mugglin C, Calmy A, et al. Ann Intern Med. 2021 Mar 16; online ahead of print].
The researchers assessed for TAF-associated changes by comparing participants switched from TDF to TAF to those who remained on TDF. Follow-up started upon approval of TAF in Switzerland (January 2016) and continued through July 2019. Participants taking TDF were included if they either continued TDF or switched to TAF. In those who switched, the index visit was the date of the switch to TAF; a random sample of index dates was assigned to those who continued TDF to generate random index dates for those participants. Weight measurements from 2.5 years prior to the index date until the end of follow-up were included. Individuals with any follow-up time following the index date were included in the analysis. If TAF was discontinued, follow-up was censored at that time. Pregnant individuals and those switched to another nucleoside reverse transcriptase inhibitor along with the TDF to TAF switch were excluded.
The analytic population included 4,375 individuals with a median age of 50 years; 25% were women, and 50% were of normal body mass index at the index visit. Overall, 20% continued TDF until the end of observation, and 80% were switched to TAF. Those who were switched to TAF were more likely to be men, identify as gay, have a lower estimated glomerular filtration rate, and less likely to be of African origin. The median follow-up was 17 months with a median of 3 follow-up visits after the index visit.
Crude weight trajectories before the index visit were similar between individuals who continued TDF and those who switched to TAF. The switch to TAF was associated with increases in weight regardless of the third ART agent (INSTI, protease inhibitor, or nonnucleoside reverse transcriptase inhibitor). A switch to TAF was associated with a 1.8 kg increase compared to a 0.7 kg increase in those who remained on TDF 18 months after the index visit. This change was similar when limiting the analysis to those with persistently suppressed viral load. The difference in weight gain between TAF and TDF was most marked for women of African origin (1.5 kg), followed by women of non-African origin (1.4 kg), and men of non-Africa origin (1.1 kg). Men of African origin did not appear to have a difference in weight gain between TDF and TAF. Assessment of metabolic parameters for 18 months following the TAF switch found that TAF was associated with increased total cholesterol, high-density lipoprotein cholesterol, low-density lipoprotein cholesterol, and triglyceride levels; switching to TAF was also associated with new-onset diabetes, with an adjusted incidence rate ratio of 1.3.
This study adds to the growing body of evidence of weight gain with selected ARVs. The strengths of this study are the relatively large sample size and the inclusion of only individuals on stable long-term ART. A limitation, as with all cohort studies, is that unmeasured factors may have influenced both the agent switch and weight change, leading to failure to identify a confounder that is the real reason for the observed differences. Another limitation is the relatively short follow-up up time. The Kaplan-Meier figures from the analysis appear to show that most of the differential in weight gain occurs in the first 6 months following the switch to TAF. This leads to questions of whether greater weight gain persists beyond the initial period on TAF and whether switching from TAF to another ARV following this initial weight gain would affect a patient’s future weight change trajectory.
Notably, individuals in both groups gained weight, developed obesity, and had new onset diabetes, all health conditions that should be prevented and managed among all people living with HIV, regardless of ART regimen. 

Effective Data Sources for Re-Linkage to HIV Care Programs (10/20)

Christopher J. Hoffmann, MD, MPH, Johns Hopkins University School of Medicine; Director, JHU-NYSDOH AI Guidelines Program

U.S. federal government efforts to end the HIV epidemic in the United States aim to reduce new infections by 74% by 2025 and 90% by 2030. Achievement of these goals rests on 4 pillars of targeted intervention:

  1. Diagnose all people with HIV.
  2. Treat HIV rapidly and effectively to achieve sustained viral suppression.
  3. Prevent new infections through use of proven interventions.
  4. Respond quickly to new HIV outbreaks.

Success in treating HIV rapidly and effectively requires linkage to and retention in care and re-linkage when necessary. The U.S. Centers for Disease Control and Prevention’s (CDC) framework for identifying individuals who are not in care (NIC) is a data-driven approach that applies a variety of case management and navigation approaches to re-link people with HIV to medical care. This approach is referred to as “data to care” or “D2C.”

A major challenge with this approach is determining which data are meaningful and how to apply them to guide intervention. The CDC currently recommends use of 3 types of data: public health department surveillance data, reporting data from clinics, and combined surveillance and clinic-level data.

New York State experience: In 2018, researchers from New York State (NYS) published a study comparing use of health department surveillance and clinic-level data to identify individuals NIC for D2C efforts in selected NYS jurisdictions [Hart-Malloy R, et al. AIDS Care. 2018;30(3):391-396]. Between January 1, 2015, and September 1, 2016, health department surveillance identified 1,735 individuals meeting criteria for NIC, 348 (26%) of whom were confirmed to be NIC. Of those confirmed as NIC, 273 (78%) were successfully re-linked to care. During this same period, clinic reporting identified 261 individuals who were potentially NIC; 73 (28%) met the surveillance definition of NIC and were traced for linkage. Of these, 19 were confirmed NIC. Of the 19, 12 (63%) were re-linked to care. The findings from this study highlight the challenge of accurately identifying individuals NIC and suggest that there is little difference in outcomes based on health department surveillance and clinic reporting data, although the small number of individuals reported by clinics limit the value of this comparison.

San Francisco experience: Researchers from the San Francisco Department of Health recently expanded understanding of meaningful data sources for D2C by comparing 3 parallel and simultaneous approaches in place from 2015 to 2017 [Sachdev DD, et al. Open Forum Infect Dis. 2020 Aug 21;7(9):ofaa369]. Surveillance data were used to identify individuals NIC based on no viral load result in the previous 15 months or viral load >1,500 copies/mL in the previous 4 months. Healthcare provider referrals included patients who had no evidence of care post-diagnosis, did not access care, or were believed to be non-adherent to antiretroviral therapy. The third approach used a combination of laboratory surveillance data and electronic medical record data from public clinics. Linkage navigators were assigned to use all available contact information to locate NIC individuals within 30 days and enroll them in a care linkage navigation program.

Results: In the San Francisco study, 954 patients were referred for navigation; 44% were identified with surveillance data, 43% with clinic reporting data, and 13% with combination data, as summarized below.

  • Surveillance data source outcomes:
    • Referred for re-linkage: 422
    • Not located: 92 (22%)
    • Ineligible*: 250 (59%)
    • Enrolled: 38 (9%)
    • Declined: 41 (10%)
    • Suppressed viral load within 12 months of linkage: 24 (6%)
  • Clinic reporting data source outcomes:
    • Referred for re-linkage: 413
    • Not located: 77 (19%)
    • Ineligible*: 111 (27%)
    • Enrolled: 167 (40%)
    • Declined: 58 (14%)
    • Suppressed viral load within 12 months of linkage: 92 (22%)
  • Combination data source outcomes:
    • Referred for re-linkage: 119
    • Not located: 33 (28%)
    • Ineligible*: 45 (38%)
    • Enrolled: 28 (24%)
    • Declined: 13 (11%)
    • Suppressed viral load within 12 months of linkage: 8 (7%)

*406 of the 954 (43%) of NIC individuals were ineligible for referral because they were found to be already enrolled in care, or they had moved away, were incarcerated, had severe medical or psychiatric barriers to navigation, or were deceased. 

Among the NIC individuals enrolled in navigation, the majority were referred by clinics (72%). When compared with those referred through surveillance or combined data sources, a greater proportion of those identified with clinic data were experiencing homelessness, were men who have sex with men, or used injection drugs. A lower proportion of individuals referred through surveillance reported methamphetamine use (16% compared to 55% and 53% for clinic and combination referrals, respectively).

Among individuals enrolled in the navigation program, retention in care and viral suppression increased from 35% to 58% and 18% to 53%, respectively. The improvement in viral suppression was greatest among surveillance referrals (60%) compared to referrals from clinics or combined data referral. This was possibly because of a lower proportion of individuals experiencing homelessness compared to those identified through clinic referrals; those experiencing homelessness also had lower success with viral suppression in this study.

Conclusions: These results demonstrate that D2C programs can successfully re-link individuals with HIV to care and help them to achieve viral load suppression for at least 12 months after re-engagement in care. They also show that, at least in San Francisco, surveillance data are less effective for identifying individuals who are NIC and eligible for linkage activities. Furthermore, only 6% of individuals identified through surveillance ultimately enrolled in linkage navigation and achieved viral suppression; this was partly because many of the individuals identified through surveillance were already in care. Importantly, each of the approaches to identifying NIC individuals reached a slightly different population.

The most successful re-linkage programs may be those that use data from multiple sources and accept substantial inefficiency to achieve ending the HIV epidemic goals.

Current Evidence Suggests Well-Controlled HIV May Not Increase COVID-19 Mortality (8/20)

Christopher J. Hoffmann, MD, MPH, Johns Hopkins University School of Medicine; Director, JHU-NYSDOH AI Guidelines Program

As hope that the COVID-19 pandemic would fade after a few months gives way to the reality that spread is likely to persist for some time, knowledge about risks is steadily accumulating. Two recent cohort studies, one from South Africa and the other from New York City, provide some insight into whether HIV is associated with higher COVID-19 mortality.

South Africa: The largest of the 2 studies is from South Africa and draws from the entire population receiving public health care in a single province, the Western Cape Province [the manuscript is currently available as a pre-print: Davies M. medRxiv. 2020 Jul 3;2020]. This study used public sector health data available in the Western Cape Provincial Health Data Centre to examine 625 COVID-19 related deaths that occurred between March 1 and June 9, 2020. The primary outcome of interest was death. All adult public sector patients with ≥1 health visit in the 3 years prior to March 1, 2020, who were believed to be alive were included. Of the 3,460,932 active patients, 22,308 (0.64%) had been diagnosed with COVID-19, and the rest either had not been tested or had tested negative for SARS-CoV-2.

The prevalence of diabetes was higher in patients with COVID-19 (14%) than in those without (8%), as was the prevalence of hypertension (23% vs 8%). An adjusted analysis included demographics, comorbidities, and general location of health care (as a proxy for sociodemographic status). COVID-19 mortality was associated with male sex, older age, diabetes, hypertension, chronic kidney disease, and HIV, approximately doubling the mortality risk. Mortality risk increased further with lower CD4 count. Notably, older age, diabetes, and chronic kidney disease were more closely associated with risk of death than was HIV.

New York City: A smaller study included 4,402 patients with confirmed SARS-CoV-2 who were admitted to 5 New York City (NYC) hospitals between mid-March and mid-April 2020 [Sigel K, et al. Clin Infect Dis. ciaa880, https://doi.org/10.1093/cid/ciaa880]. Among those patients, 88 had HIV, the median age was 61 years, 40% were Black, 30% were Latinx, 81% had a viral load <50 copies/mL, and 85% had a CD4 lymphocyte count >200 cells/mm3 prior to admission. People with HIV were matched by age, sex, race/ethnicity, and week of diagnoses of COVID-19 to create a comparator group that included up to 5 people without HIV (n = 405) to each person with HIV (n = 88).

A greater proportion of people with HIV were current smokers (55% vs 23% of comparators) and had chronic obstructive pulmonary disease (10% vs 3% of comparators). Based on need for oxygen supplementation at admission and most laboratory markers, COVID-19 severity was similar in both groups. All people with HIV received antiretroviral therapy during hospitalization, and most patients received hydroxychloroquine and azithromycin. There were no differences in need for mechanical ventilation (18% of people with HIV vs. 23% of comparators) and death (21% of people with HIV vs. 20% of comparators). The cumulative incidence of death was also similar in the 2 groups.

Conclusion: The South Africa and NYC studies provide insights into HIV and mortality among people with HIV and COVID-19. The South Africa study reported an increased risk of mortality among people with HIV and the NYC study reported no difference in risk. The divergent findings may be attributed to a number of factors, including differences in study population, study design, variables used in analyses, and HIV care. For instance, it is possible that the median CD4 lymphocyte count, a variable available for only a fraction of patients in the South Africa study analysis, was lower among that group than among people with HIV in the NYC group. Lower CD4 lymphocyte count was associated with increased mortality risk in the South Africa Cape study, suggesting that a population with low CD4 lymphocyte counts would have an elevated mortality.

These 2 studies suggest that if there is any increase in risk of mortality among people with HIV who have COVID-19, the increase is small. Overall COVID-19 mortality in hospitalized patients is high, and current treatments have limited effectiveness, so prevention of SARS-CoV-2 transmission and infection is critical to stemming the epidemic and reducing mortality among all people.

Long-Acting Cabotegravir and Rilpivirine for Maintenance of HIV-1 Suppression (6/20)

Christopher J. Hoffmann, MD, MPH, Johns Hopkins University School of Medicine; Director, JHU-NYSDOH AI Guidelines Program

Many clinicians provide care for patients who struggle with taking enough of their antiretroviral therapy (ART) to maintain a consistently undetectable viral load. The reasons for this are myriad, and some are more easily addressed than others. For patients who attend clinical care visits as scheduled but struggle with taking pills, long-acting ART preparations have long seemed a dream solution. This dream may be realized once long-acting injectable cabotegravir (CAB) plus rilpivirine (RPV) is available. Long-acting CAB plus RPV is administered every 4 weeks as an intramuscular (IM) injection. (Clinical trials have initiated treatment with 4 weeks of oral CAB plus RPV to ensure tolerability before administration of the long-acting regimen.)

The efficacy and noninferiority of CAB plus RPV for viral suppression have been reported in 2 phase IIb studies: LATTE (Margolis D, et al. Lancet Infect Dis. 2015;15(10):1145-1155) and LATTE-2 (Margolis D, et al. Lancet. 2017;390(10101):1499-1510). Building on the findings of LATTE-2 are 2 phase III noninferiority randomized clinical trials of long-acting CAB plus RPV versus oral therapies: the ATLAS and FLAIR trials. 

The ATLAS trial randomized participants on long-term ART to continue their stable oral regimen or switch to the IM regimen (Swindells S, et al. N Engl J Med. 2020;382(12):1112-1123). In the FLAIR trial, the induction phase for treatment-naive participants was 16 weeks of treatment with oral dolutegravir/abacavir/lamivudine, followed by either continuation on that regimen or a switch to 4 weeks of oral CAB plus RPV followed by monthly injections (Orkin C, et al. N Engl J Med. 2020;382(12): 1124-1135). The primary endpoint in both trials was the percentage of participants with HIV RNA level ≥50 copies/mL at week 48.

In the ATLAS study, of the 618 randomized participants, 33% were female, 32% were nonwhite, 74% had a CD4 count ≥500 /mm3, and the median age was 42 years. At 48 weeks, intention-to-treat analysis found that 92.5% of participants in the IM arm and 95.5% in the oral therapy arm had HIV RNA <50 copies/mL. The 3 participants in the IM arm with confirmed treatment failure all had RPV resistance mutations, and 1 had CAB resistance. Of participants in the IM arm, 10 (3%) withdrew because of adverse events attributed to the study drugs (injection site reactions were the most common). Fatigue was reported in 7% of participants in the IM arm and in 2% of participants in the oral therapy arm. Overall, patient satisfaction was much higher among those receiving the long-acting IM ART.

The FLAIR study enrolled 629 treatment-naive participants and initiated oral induction therapy; 566 remained in the study at the time of randomization. Of those who remained, 22% were female, 26% were nonwhite, 20% had HIV RNA >100,000 copies/mL, and the median age was 34 years. At week 48, 93.6% of participants in the IM arm and 93.3% in the oral therapy arm had HIV RNA <50 copies/mL. Injection site reactions were the primary adverse event in the IM arm. Also reported were back pain in 8% and 5% of participants in the IM and oral medications arms, respectively, and fever (8% vs. 1%).

The ATLAS and FLAIR trials confirm that long-acting CAB plus RPV is safe and efficacious in ART-experienced and ART-naive individuals. The efficacy of long-acting IM-administered regimens in individuals struggling with ART adherence is currently being studied (see ClinicalTrials.gov > NCT03635788). This population stands to gain the most benefit from long-acting ART. Once IM CAB plus RPV* is approved by the U.S. Food and Drug Administration, we will have a powerful tool for helping patients optimize adherence and viral suppression.


*Oral CAB plus RPV will be available for induction and to ensure that patients can continue therapy if there is a problem with access to the IM formulation.

Persistence of Association Between HIV and Risk of Atherosclerotic Cardiovascular Disease (1/20)

Christopher J. Hoffmann, MD, MPH, Johns Hopkins University School of Medicine; Director, JHU-NYSDOH AI Guidelines Program

People with HIV are at increased risk for atherosclerotic cardiovascular disease (ASCVD), including myocardial infarction (MI) and stroke [Schouten J, et al. Clin Infect Dis. 2014;59(12):1787-1797Shah ASV, et al. Circulation. 2018;138(11):1100-1112]. Both untreated HIV and specific HIV treatment regimens have been associated with ASCVD. Earlier initiation of antiretroviral therapy (ART) and a shift away from regimens associated with ASCVD may change the ASCVD risk profile for people with HIV. Rosenson and colleagues assessed the association between HIV and ASCVD, including MI, stroke, and lower-limb artery disease events among a contemporary cohort of people being treated with antiretroviral medications [Rosenson RS, et al. J Am Heart Assoc. 2020;9(1):e013744].

The investigators assembled a retrospective cohort of people with HIV and matched controls using the MarketScan database of administrative and claims data from employer-sponsored healthcare plans and Medicare supplemental plans. From January 1, 2011, to December 31, 2016, individuals with either ≥1 hospitalization with a discharge diagnosis of HIV or ≥2 pharmacy claims for ART were included. Individuals aged ≤19 years living in the United States for <1 year before being identified as having HIV in the database who did not have continuous healthcare coverage were excluded. Four individuals without HIV were frequency matched to each individual with HIV based on age group, sex, and calendar year in which the HIV diagnosis was entered in the database.

In addition to HIV status, age, and sex, the following patient-level data were abstracted from the MarketScan database: history of coronary artery disease, stroke, peripheral artery disease, diabetes, heart failure, chronic kidney disease, liver disease, or depression; receipt of care from a cardiologist; hospitalization; tobacco use; and use of antihypertensive, statin, and nonstatin lipid-lowering medications. Use of ART was defined as >2 prescription fills within 1 year before each individual’s index date for HIV classification. Individuals were followed from the index date until December 31, 2016, for hospitalizations for ASCVD. The investigators used Cox proportional models, including adjusting for potential confounders, to compare ASCVD outcomes between the people with HIV and matched controls.

A total of 82,426 people with HIV and 329,704 matched controls met the inclusion criteria. Slightly more than 40% entered observation in 2011 and were between ages 19 and 44 years (42.6%). A larger proportion of people with HIV had a history of chronic kidney disease, liver disease, depression, antihypertensive medication use, statin use, tobacco use, and polypharmacy. Of people with HIV, 50.2% were receiving nucleoside reverse transcriptase inhibitors, 44.2% were receiving nonnucleoside reverse transcriptase inhibitors, 22.1% were receiving protease inhibitors, and 21.7% were receiving other antiretroviral medications (e.g., integrase strand transferase inhibitors, entry inhibitors, pharmacokinetic enhancers).

During a median follow-up of 1.6 years, 3,287 ASCVD events were identified in the database. The hazard for composite ASCVD was higher among people with HIV (hazard ratio, 1.29; 95% confidence interval, 1.18-1.40) than those without HIV. The hazard was also higher for MI, stroke, and lower-limb arterial disease for people with HIV. The use of statins reduced the hazard of each outcome among people with HIV, although the hazard trend remained higher among people with HIV. When adjusting ASCVD events by characteristics, the hazard was highest among individuals aged 19 to 44 years compared with matched individuals without HIV, suggesting an interaction between age and HIV and the hazard of ASCVD. No characteristics (e.g., statin use) negated the increased hazard for ASCVD among people with HIV.

This study suggests that people with HIV who are in care now or were in the past few years are at higher risk for ASCVD than age- and sex-matched controls. The finding of the highest risk among individuals aged 19 to 44 years is notable because this population is likely to have been diagnosed with HIV recently and have initiated ART soon after diagnosis with the contemporary ART regimens that are less likely to cause atherosclerosis than older regimens. The results of this study suggest the persistence of ASCVD risk even among individuals recently infected with HIV and started on ART.

ASCVD risk reduction includes the use of statins as well as an overall reduction in cholesterol, blood pressure control, smoking cessation, and HIV virologic suppression. The investigators did not include laboratory data, blood pressure measurements, or prior ART history, and controlling for these might have reduced the effect size of the observed association between HIV and ASCVD. In addition, this study was not designed to assess for an association between ASCVD and HIV among people with HIV initiated on rapid ART with regimens not believed to be associated with atherosclerosis. Another limitation is that the causes of stroke and MI were not assessed to differentiate between atherosclerotic and non-atherosclerotic events (i.e., type 2 MI). A higher proportion of non-atherosclerotic events among people with HIV may partly explain the association found by the investigators, and this could be the reason for a larger association effect size among those aged 19 to 44 years. Despite these limitations, this study provides a reminder to optimize ASCVD management among people with HIV.

Vitamin E Therapy: An Option for Treatment of Hepatic Steatosis in People With HIV (12/19)

Christopher J. Hoffmann, MD, MPH, Johns Hopkins University School of Medicine

Among people with HIV, a leading cause of morbidity and mortality is liver disease, most commonly caused by viral hepatitis or non-alcoholic fatty liver disease (NAFLD) [Sherman KE, Peters MG, Thomas DL. Top Antivir Med. 2019;27(3):101]. In the general population, NAFLD, the first step of fatty infiltrate in the liver, is often secondary to obesity and insulin resistance and accounts for the majority of chronic liver disease. In people who are not overweight or obese, NAFLD may be an adverse effect of antiretroviral therapy, genetics, or metabolic syndrome [Verna EC. Lancet Gastroenterol Hepatol. 2017;2(3):211]. NAFLD can progress to non-alcoholic steatohepatitis (NASH), which can lead to liver fibrosis and cirrhosis. NAFLD occurs among 13% to 55% of people with HIV [Verna 2017;  Cai J, Sebastiani G. J Acquir Immune Defic Syndr. 2019;81(1):e23-e25].

Currently, the options for management of NASH are limited. When indicated, weight loss of at least 10% of body weight can effectively reverse hepatic inflammation and lead to regression of fibrosis, but there are currently no good treatment options for individuals of normal weight. Drugs that affect insulin, including pioglitazone and metformin, have been studied in the past, but clear evidence of benefit in NASH management has not been found. Vitamin E, an antioxidant, has also been studied because oxidative stress appears to contribute to NASH. In 2 studies in individuals without HIV, vitamin E was found to reduce transaminases or improve liver histology [Sanyal AJ, et al. N Engl J Med. 2010;362:1675Levine JE, et al. JAMA. 2011;305(16): 1659].

Now, recently reported results of a single-arm, open-label study of 24 weeks of 800 IU vitamin E for NASH in people with HIV suggest it may be an effective treatment in this population as well [Sebastiani G, et al. AIDS. 2019. Epub ahead of print]. The primary outcome of this study, which was conducted at a single site in Canada from 2015 to 2018, was a decrease in alanine aminotransferase (ALT) at 24 weeks. Inclusion criteria are listed below:

  • Antiretroviral therapy for ≥6 months, with HIV viral load <50 copies/mL
  • Controlled attenuation parameter (CAP) ≥248 dB/m. CAP is a simultaneous measurement of the degree of ultrasound attenuation by hepatic fat and liver stiffness by vibration-controlled transient elastography (e.g. FibroScan). When compared with liver biopsy, CAP correlates more closely with steatosis and steatosis severity.
  • Serum cytokeratin-18 (CK-18) >130.5 U/L. CK-18 is a serum marker generated during apoptosis of liver cells; CK-18 levels are correlated with NAFLD.
  • Patients with hepatitis C virus or significant alcohol use were excluded.

The researchers screened 133 patients and identified 27 who met the inclusion criteria. Of those, 93% were white, 81% were male, 26% were obese, 52% were overweight, and 40% had significant liver fibrosis.

Results: Vitamin E treatment improved participants’ ALT and CAP scores and reduced steatosis. Notably, there were no changes in body mass index (BMI), triglycerides, total cholesterol, or HDL cholesterol. There were no major adverse events from the vitamin E treatment.

Summary of results:
  ALT CAP Severe Steatosis
Baseline Elevated in 74% of participants.
Median: 50 IU/L (IQR, 43-66)

Median: 271 dB/m
(IQR, 274-349)

67% of participants
Week 24
Elevated in 15% of participants.
Median: 30  IU/L (IQR, 25-36)
Median: 277 dB/m
(IQR, 258-310)
41% of participants

These results suggest that vitamin E is a fairly well-tolerated approach to managing NAFLD and NASH among people with HIV. This study does have considerable limitations, including a very small sample size, homogeneity of the study participants, short observation time, and single-arm design. Confirming similar findings in a larger and longer-term study that includes diverse participants would be valuable.

However, given the lack of other effective pharmacologic treatments, clinicians may wish to consider vitamin E therapy for patients with HIV and NAFLD or NASH.

Final Results of the PARTNER 2 Study Confirm Undetectable = Untransmittable (U=U) (5/19)

Christopher J. Hoffmann, MD, MPH, Johns Hopkins University School of Medicine

Dr. Alison Rodger presented the final results of the PARTNER2 study at the 22nd International AIDS Conference in Amsterdam in July 2018. The findings of the study increased the precision of estimates for the risk of HIV transmission during condomless anal sex between serodifferent gay men when the man with HIV has a viral load <200 copies/mL. At the AIDS Conference, Rodger reported that there were zero linked HIV transmissions during approximately 77,000 acts of condomless anal sex.

Rodger and colleagues have now published these results in an article that provides important details about the study population [Rodger A, Cambiano V, Bruun, T, et al. Lancet. 2019 May 04;393(10183)]. The PARTNER2 study recruited serodifferent gay men between September 2010 and July 2017 from 75 clinical sites in 14 European countries. Gay men with HIV were eligible if they were taking antiretroviral therapy (ART) and expected to stay on it, and they had condomless penetrative sex (anal receptive without ejaculation, anal receptive with ejaculation, and insertive anal) in the past month with a partner who did not have HIV. Data on demographics, sexual behavior, ART, viral load, CD4 count, and sexually transmitted infections (STIs) were collected at baseline and every 4 to 6 months. Men with HIV were expected to have viral load testing every 6 to 12 months, and their partners without HIV were expected to have HIV testing every 6 to 12 months. Phylogenetic testing was performed when there was evidence of seroconversion to determine whether the transmission was linked within the couple.

The primary analysis estimated the risk of phylogenetically linked HIV transmission during the follow-up period. Criteria for inclusion of a couple in the follow-up period were as follows:

  • The couple reported condomless anal sex.
  • The man with HIV had viral load testing in the past 12 months.
  • The most recent viral load for the man with HIV was <200 copies/mL.
  • The man who did not have HIV had HIV testing in the past 12 months.
  • The man who did not have HIV did not use pre- or post-exposure prophylaxis (PrEP or PEP).

Of the 972 gay male couples who participated, 88% were white, the median age was about 40 years, and 97% had viral loads <50 copies/mL at enrollment. There were 2,072 couple-years of follow-up and approximately 76,000 episodes of condomless sex during eligible observation periods. Approximately 25% reported an STI during the most recent follow-up period, and 4% of men who did not have HIV reported injection drug use.

There were 15 incident HIV infections detected, none of which were linked phylogenetically to the partner with HIV in the study, resulting in an estimated risk of transmission of zero, with an upper bound of the risk estimate of 0.23 per 100 couple-years for condomless anal sex when the man with HIV has a viral load <200 copies/mL (95% confidence interval). In other words, the researchers reported a possible risk of zero and a maximum risk of one transmission of HIV per 435 years of condomless anal sex; this was interpreted by the authors as follows: “Our findings provide conclusive evidence that the risk of HIV transmission through anal sex when HIV viral load is suppressed is effectively zero.” Without virologic suppression due to ART, 472 within-couple transmissions would have been expected.

Important caveats to consider are that the study population was older (median age 40 years) and the individuals with HIV had been taking ART for several years. Both characteristics are associated with continued viral suppression. Other populations may struggle to maintain a viral load <200 copies/mL and may need greater support to adhere to ART regimens and achieve and maintain the virologic suppression required for U=U.

Relative Contributions of Traditional Modifiable Risk Factors to Morbidity Among People with HIV (4/19)

Christopher J. Hoffmann, MD, MPH, Johns Hopkins University School of Medicine

Compared with people who do not have HIV, people with HIV, even those with well-controlled disease, have an elevated risk of myocardial infarction (MI), end-stage liver disease (ESLD), end-stage renal disease (ESRD), and non-AIDS-defining cancers (NADCs) [Smith CJ, et al. Lancet 2014 Jul 19;384(9939)Weber R, et al. HIV Med. 2013 Apr;14(4)]. Along with traditional modifiable risk factors, such as cigarette smoking, hypertension, diabetes, obesity, and chronic viral hepatitis, HIV infection is an established risk factor for these diseases [Vachiat A, et al. J Am Coll Cardiol. 2017 Jan 3;69(1)Joshi D, et al. Lancet. 2011 Apr 2;37(9772)Cohen SD, et al. N Engl J Med. 2017 Dec 14;377(24)]. A recent study from the North American AIDS Cohort Collaboration on Research and Design (NA-ACCORD) estimates the relative contribution of traditional risk factors to the development of MI, ESLD, ESRD, and NADCs in people with HIV [Althoff KN, et al. Lancet HIV. 2019 Feb;6(2)].
Study Methods
Althoff and colleagues used data from more than 180,000 adults receiving HIV care at 200 NA-ACCORD sites in the United States and Canada; the earliest data are from January 2000. The researchers calculated population attributable fractions (PAFs), which estimate the total burden of a disease attributable to a specific risk factor. Viewed another way, PAF estimates the proportion of a disease that would not occur if a specific risk factor was not present.
In the MI analysis, 347 of 29,515 (1%) of participants had an MI; in the ESLD analysis, 387 of 35,044 (1%) were diagnosed with ESLD; in the ESRD analysis, 255 of 35,620 (1%) developed stage IV chronic kidney disease; and in the NADC analysis, 1,405 of 61,500 (2%) were diagnosed with a malignancy.
For MI, ESLD, ESRD, and NADCs, the researchers calculated the PAFs for the risk factors listed below. Because the study included only people with HIV, PAFs attributable to HIV infection itself were not calculated.
  • Cigarette smoking (ever or never)
  • Elevated total cholesterol
  • Hypertension (clinical diagnosis or anti-hypertensive prescription)
  • Type 2 diabetes
  • Stage IV chronic kidney disease
  • Alcohol use (ever reporting ≥3 drinks per day or ≥7 drinks per week for women and ≥4 drinks per day or ≥14 drinks per week for men)
  • Body mass index
  • Chronic hepatitis C virus (HCV); antibody reactive or detectable HCV RNA
  • Chronic hepatitis B virus (HBV); surface antigen reactive, E antigen reactive, or detectable HBV DNA
PAFs were also calculated for 3 HIV-specific risk factors for increased morbidity and mortality:
  • CD4 count <200 cells/mm3
  • HIV viral load >400 copies/mL
  • History of an AIDS-defining illness
Study Results
Traditional modifiable risk factors made a significant contribution to morbidity among participants receiving routine HIV care at 200 sites in North America. The table below lists the PAFs for MI, ESLD, ESRD, and NADCs for select risk factors.

The results of this study suggest that modification of traditional risk factors could have a significant effect on the health of people with HIV. For example, researchers estimate that 44% of the MI in this population is attributable to elevated total cholesterol. Thus, eliminating that risk factor could greatly reduce MI in this population.

Viral load suppression and CD4 count reconstitution remain fundamental goals of HIV care. However, in the era of well-tolerated and effective antiretroviral therapy, these HIV-associated risk factors appear to make only minor contributions to PAFs for key causes of morbidity among people with HIV.

People with HIV, even well-controlled HIV, are at risk of MI, ESLD, ESRD, and NADCs. Among people with poorly controlled HIV who do not have an AIDS-defining illness, these 4 diseases are also leading causes of death. The findings of this study from Althoff and colleagues reinforce the importance of educating patients about modifying these risk factors and providing aggressive clinical management to reduce morbidity and mortality among people with HIV.

Performance of the ACC/AHA CVD Risk Calculator Among People With HIV (2/19)

Christopher J. Hoffmann, MD, MPH, Johns Hopkins University School of Medicine

In December 2018, researchers reported their evaluation of the utility of 2 guidelines for statin use in predicting subsequent cardiovascular disease (CVD) events in people with HIV in a single-center, Boston-based cohort [Mosepele M, et al. Open Forum Infect Dis. 2018 Dec 13;5(12)]. The guidelines compared were those of the American College of Cardiology/American Heart Association (ACC/AHA) and the National Cholesterol Education Program Adult Treatment Program III (ATPIII). Although known to increase CVD risk, HIV has not been incorporated into standard risk calculation models partly because modeling datasets include a limited number of people with HIV.

The study by Mosepele and colleagues calculated CVD risk scores and compared risk classification with subsequent CVD outcomes among patients from the Partners HIV cohort and a matched control comparison cohort from the Partners HealthCare System. Individuals with known CVD at the start of the observation period were excluded.

Participants were followed for 3 years, with the earliest observation starting in 2006 and the latest ending in 2011. Abstracted data for risk factor calculations included: diabetes, hypertension, dyslipidemia, smoking, CD4 cell count, HIV viral load, and use of statins, antihypertensive medications, and antiretroviral medications. CVD outcomes were ascertained from the Partners Healthcare System electronic medical record. Participants were primarily assessed for whether they met statin prescribing recommendations of either ACC/AHA or ATPIII guidelines. Risk stratification and subsequent CVD events were also compared.

The study included 1,394 people (median age, 46 years) with HIV, and 6,141 controls (median age, 47 years). The median duration of follow-up was 3.9 years. Hypertension was less prevalent in the group with HIV (41%) than in the control group (44%), and smoking was more prevalent among those with HIV (45%) than among those without HIV (28%). Of note, 39% of all participants with HIV met ACC/AHA criteria and 20% met ATPIII criteria for statin therapy.

A high proportion (5%) of people with HIV experienced CVD events during the approximately 3-year observation period of the study (see Table 1, below).

Table 1: Comparison of Risk Calculators for People With or Without HIV
Cohort % of Participants
With CVD Events
% of Participants with CVD Events
Who Met Statin Prescribing Criteria
People with HIV 71 (5%) 42 (59%) 25 (35%)
Controls 197 (3%) 141 (72%) 85 (43%)
Abbreviations: ACC/AHA, American College of Cardiology/American Heart Association; ATPIII, National Cholesterol Education Program Adult Treatment Program III; CVD, cardiovascular disease.

The results highlight the limitations of the ACC/AHA risk assessment calculation for predicting CVD risk among people with HIV. Notably, 40% of those with HIV who experienced a CVD event were classified as low risk (not recommended for statin therapy) by ACC/AHA calculation (estimated 10-year CVD risk, <7.5%). Improved risk stratification models are needed to provide more accurate eligibility criteria for statin use to improve CVD prevention among people with HIV.

Real-World, Event-Based PrEP at a Clinic in France (10/18)

Christopher J. Hoffmann, MD, MPH, Johns Hopkins University School of Medicine

Pre-exposure prophylaxis (PrEP) is effective and important for HIV prevention among individuals at higher risk of HIV acquisition. In the United States, the Centers for Disease Control and Prevention (CDC), the New York State Department of Health AIDS Institute, and other groups recommend daily use of tenofovir disoproxil/emtricitabine (TDF/FTC) for PrEP. Elsewhere in the world, event-based dosing of PrEP (2 pills, 2 to 24 hours before sex, 1 pill 24 hours later, and 1 pill 48 hours after the first dose) is accepted based on results from the IPERGAY study*, which compared event-based PrEP with placebo in HIV-negative men who have sex with men (MSM). To date, there have been no randomized clinical trials comparing daily and event-based PrEP, and there are limited data regarding the use of event-based PrEP in routine clinical settings.

Observational Study of Event-Based vs. Daily PrEP

Noret and colleagues published data from a single-arm observational study describing event-based and daily PrEP at a clinic in France that participated in the IPERGAY study [Noret M, et al. AIDS. 2018 Sept 24; 32(15)]. Participants were recruited between November 2015 and April 2017 and could opt for event-based or daily PrEP with TDF/FTC if they met study criteria for PrEP initiation (noted in the table below, along with actual participant characteristics). Participants were scheduled for follow-up every 3 months, at which time HIV testing and screening for sexually transmitted infections (STIs) were performed, and risk reduction counseling was provided.

Table: Summary of Noret Study Participant Criteria and Characteristics
Criteria for PrEP Initiation Participant Characteristics
  • HIV-negative
  • MSM or a transgender man or women who engaged in condomless anal sex with at least 2 partners in the previous 6 months
  • Diagnosis of an STI in the past 12 months
  • Receipt of multiple courses of post-exposure prophylaxis (PEP) over the past 12 months
  • Engagement in “chemsex” (sex involving recreational drug use)
  • Estimated glomerular filtration rate >50 mL/min 
  • Median age: 36 years
  • Percent MSM: 99.4%
  • Median number of sex partners reported for the previous 3 months: 10
  • Reported condomless sex for most recent sexual encounter: 53.3%
  • Positive for at least 1 bacterial STI at PrEP initiation: 14.6%

PrEP was prescribed to 1,049 individuals, of whom 793 (75.6 %) opted for event-based PrEP. Participants who opted for event-based PrEP reported fewer sex partners and less frequent condomless sex than those who did not. The median follow-up period was 5.6 months. Among a subset of participants with tenofovir levels measured 6 months after enrollment, 87.3% had levels consistent with intake within the prior week, and 62% had levels consistent with intake the prior day.

 Study Results

  • Loss from care was high:
    • 162 participants (15.5%) discontinued care at the clinic
    • 107 had unknown PrEP status (no follow-up data were available)
    • 23 had discontinued PrEP due to being in a stable relationship
  • There were 4 diagnoses of new HIV infections after PrEP prescription:
    • 2 in participants who had not started PrEP at the time of HIV acquisition
    • 1 in a participant who reported poor adherence to PrEP
    • 1 in a participant who was lost from care and then found to be HIV-positive upon return to re-initiate PrEP
  • There was a modest increase in STI rate, from 14.6% at baseline to 19.6% at 12-month follow-up, associated with an increase in cases of rectal chlamydia.

These results support the use of event-based PrEP in a “real-world” setting. The majority of participants opted for event-based PrEP, and HIV acquisition was confined to those who were not taking TDF/FTC around the time of acquisition.

This study has several limitations, including relatively short follow-up time no description of the frequency of use of event-based PrEP. Despite these limitations, there are a number of valuable lessons, including the preference for event-based PrEP among the study participants. Unfortunately, attrition from care, even with this simple-to-use, event-based approach, was high. This result, which is consistent with other studies of “real-world” PrEP, raises the critical question of how to ensure long-term engagement with PrEP among individuals with ongoing high risk for HIV infection. Further work is needed to understand specific barriers to continued PrEP engagement, such as understanding the effect of asking PrEP users to commit to ongoing medical care and diagnostic testing and to take pills regularly. Effective approaches to support clinical follow-up and PrEP use are also needed.

*In the IPERGAY study, 2 of the 199 MSM who received TDF/FTC seroconverted to HIV-positive, compared with 14 of the 201 men who received placebo. A median of 15 doses of TDF/FTC or placebo were taken monthly (interquartile range, 11-21) [Molina JM, et al. N Engl J Med. 2015 Dec 3; 373(23].

Is High-Sensitivity CRP a Useful Cardiovascular Risk Predictor for Women with HIV? (5/18)

Christopher J. Hoffmann, MD, MPH, Johns Hopkins University School of Medicine

Risk of cardiovascular disease (CVD) and ischemic stroke in women with HIV appears to be greater than in men with HIV and in women who do not have HIV [Hanna DB, et al. Clin Infect Dis.2016;63(8)Womack JA, et al. J Am Heart Assoc. 2014;3(5)]. Identifying women at the highest risk of CVD and estimating their absolute risk is important to optimize risk reduction through strategies that include lifestyle modification (exercise, diet, weight management, smoking cessation) and medical management with statins, aspirin, or other cardiovascular-specific medications.

CVD risk calculators (e.g., FraminghamACC/AHA) factor data such as age, sex, Black or non-Black race, total cholesterol, HDL cholesterol, blood pressure, diabetes, tobacco smoking, and treatment for hypertension to estimate risk. However, such calculations underestimate CVD risk among some populations, notably women and people with HIV [Losina E, et al. Clin Infect Dis. 2017;65(8)Thompson-Paul AM, et al. Clin Infect Dis. 2016;63(911)]. Identification of additional, accurate biomarkers would be valuable in stratifying risk among people with HIV. The American Heart Association (AHA) guideline suggests the use of high-sensitivity C-reactive protein (hsCRP) and coronary artery calcium scores in intermediate-risk patients.

In a recently published article, Moran and colleagues report results of a study assessing the accuracy of hsCRP in predicting CVD and its progression among a prospective cohort of women with HIV or at increased risk for HIV [Moran CA, et al. AIDS. 2018;32(8)]. This sub-study of the Women’s Interagency HIV Study (WIHS) used carotid ultrasound to examine carotid intimal-media thickness (CIMT) and carotid plaque formation as markers of CVD risk and progression (both are associated with subsequent CVD events). 

The carotid ultrasound sub-study, which included 1,870 women, began in 2004. All baseline study visits occurred between 2004 and 2006, and final sub-study visits took place between 2010 and 2013. The median time between baseline and follow-up ultrasound was 6.6 years.

Participants with ≥2 carotid ultrasounds and hsCRP measured at the baseline sub-study visit were included in the study (N= 788). The analysis ultimately included 783 women after 5 were excluded due to HIV seroconversion; 62% were Black; median age at first visit was 41 years; and the average 10-year Framingham risk calculation for all was 1%.

Median hsCRP was higher among women who did not have HIV (3.2 mg/L) than among those who did (2.2 mg/L). Baseline focal plaques and CIMT were similar in both populations. Change in CIMT was defined as the difference in thickness between final and baseline measurements. Focal plaque progression was defined as the number of new plaques formed between the baseline and final visits. The team assessed for associations between hsCRP and CIMT and focal plaques using hsCRP as a categorical variable, with a cut-off of ≥3 mg/L and with hsCRP as a continuous natural log-transformed variable. Analysis revealed the following:

  • No association between baseline CIMT and hsCRP in women with or without HIV.
  • Non-significant trend toward an association between high baseline hsCRP and focal plaque prevalence.
  • Non-significant trend in CIMT progression and high baseline hsCRP among women both with and without HIV.
  • In a sub-group analysis by HIV status, high baseline hsCRP was associated with focal plaque progression among women who did not have HIV, but not among women with HIV. This lack of association held even in analysis of only the 143 women with persistent virological suppression.

A potential limitation of this study is the use of carotid artery ultrasound results rather than actual cardiovascular events. Nonetheless, the results provide valuable evidence that hsCRP is not likely to be a useful tool for stratifying CVD risk among women with HIV. The authors comment that CIMT and carotid plaque are strongly associated with CVD outcomes, suggesting that the study results likely do represent a true reflection of the lack of utility of using hsCRP for estimating risk in this population. More expensive tests, such as computed tomography-based coronary artery calcium scores may be something to consider for a subgroup of women with HIV. Further work by this research team and others may provide evidence to help determine when such testing is most appropriate.

Subscribe to receive Topics, Trends & Updates by email.

Chronic Comorbidities among People Living with HIV (1/18)

Christopher J. Hoffmann, MD, MPH, Johns Hopkins University School of Medicine

As a result of potent and tolerable antiretroviral therapy (ART), people with HIV infection (PLWH) are living longer. However, rates of chronic, non-communicable end-organ diseases are higher in PLWH than in people of similar ages who do not have HIV infection. Even when adjusted for traditional risk factors, rates of myocardial infarction and other comorbidities, including liver disease, kidney disease, and several cancers, are increased among PLWH [Drozd DR, Kitahata MM, Althoff KN, et al. J Acquir Immune Defic Syndr. 2017 Aug 15;75(5); Hogg RS, Eyawo O, Collins AB, et al. Lancet HIV. 2017 Jun;4(6)]. The increased prevalence of these diseases in PLWH has been attributed to HIV-related chronic inflammation and immunosuppression, treatment with specific antiretroviral agents, and coinfections, such as hepatitis C. In addition, PLWH may have traditional, non-HIV-related risk factors for these diseases that require management beyond traditional HIV care. 

In people who do not have HIV infection, cardiovascular, liver, and kidney diseases are associated with clear risk factors, including hypertension, hyperlipidemia, and diabetes. Characterizing the prevalence of such chronic conditions among PLWH has important implications for care: If cardiovascular, liver, and kidney disease are primarily a result of HIV infection, then adjustments in HIV-specific management may abate or prolong end-organ disease. If traditional risk factors are also highly prevalent, then optimal management of these conditions will be an essential aspect of care for all PLWH.

Gallant and colleagues recently published results of a study that examined the prevalence of chronic, non-communicable, and end-organ diseases among people with and without HIV infection [Gallant J, Hsue PY, Shreay S, Meyer N. J Infect Dis. 2017 Dec 19;216(12)]. The researchers used claims data from private and public health insurers to determine the prevalence of specific non-communicable diseases among PLWH in 2013 compared with the prevalence among PLWH in 2003. In addition, they matched prevalence with a general population sample and compared prevalence between the two groups (PLWH and people without HIV infection) by year from 2003 to 2013. 

The research team used ICD-9-CM claims coding to describe co-morbidities and non-communicable diseases. Data for all individuals who were 18 years or older with at least 6 months of documented care were acquired from the private Truven Health Market Scan Commercial Claims Encounters database and from public Medicare and Medicaid claims. ICD-9-CM coding for HIV was used to identify PLWH: 36,298 patients with private insurance; 26,246 patients covered by Medicaid; and 1,854 patients covered by Medicare. PLWH were matched approximately 1:2 with non-HIV-infected controls.

Among PLWH, between 2003 and 2013, there was a substantial increase in the prevalence of multiple comorbidities, as illustrated in the table below.

These increases in the prevalence of comorbidities between 2003 and 2013 indicate that multiple health conditions are common and are increasing among PLWH. The high prevalence of hypertension, hyperlipidemia, and diabetes substantially increases the risk of cardiovascular, kidney, and other diseases for PLWH.

When the researchers compared the matched populations (with and without HIV) by age and sex they found that the prevalence of hypertension, hyperlipidemia, and diabetes was roughly the same. Disease prevalence did vary slightly by type of insurance coverage. For example, the prevalence of diabetes was higher among Medicare patients without HIV (33.9%) than it was among those with HIV (29.5%). However, diabetes prevalence was the same for patients with private insurance, irrespective of HIV status. The prevalence of end-organ disease (cardiovascular, venous, liver, and kidney) was higher among PLWH, regardless of insurance type.

A yearly comparison of comorbidities between people with and without HIV infection revealed increases in disease prevalence in both groups (see the published article for interesting graphics). This suggests that the current high prevalence of some non-communicable diseases among PLWH likely represents an overall trend in morbidity and diagnosis or claims for those conditions. It does not appear that the increase in prevalence of non-HIV age-related comorbidities from 2003 to 2013 among PLWH is unique to HIV.

Optimal management of hyperlipidemia, hypertension, and diabetes begins with evidence-based behavioral (lifestyle) interventions and medical management. Clinicians caring for people who are living and aging with HIV infection in 2018 will have to be able to manage hypertension, hyperlipidemia, and diabetes (or refer patients for care) as well as they manage ART to improve and maintain the health of their patients.

Subscribe to receive Topics, Trends & Updates by email.

PrEP: Progress and Challenges in the Elimination of HIV Transmission (11/17)

Christopher J. Hoffmann, MD, MPH, Johns Hopkins University School of Medicine

The elimination of HIV transmission may be achievable if increases can be realized in both the proportion of people with HIV infection who maintain an undetectable viral load and the proportion of people at the highest risk of acquiring HIV who are using pre-exposure prophylaxis (PrEP). Several studies (PROUD, IPERGAY, IPREX) have demonstrated the efficacy of tenofovir disoproxil fumarate and emtricitabine (TDF/FTC; Truvada) PrEP to prevent HIV infection. However, 5 years after FDA approval of PrEP in the United States, many of those at the highest risk of HIV infection still are not aware of or are not using PrEP. As of the first quarter of 2017, an estimated 136,000 people were being prescribed PrEP in the U.S. However, groups in critical need-younger men who have sex with men (MSM), particularly African American MSM-are not receiving PrEP. 

Two recently published studies, discussed below, highlight the promise, progress, and challenges associated with PrEP and any effort to scale-up to a level that could result in the elimination of HIV transmission. The first provides evidence that changes in the risk of acquiring sexually transmitted infections (STIs) do occur among men on PrEP, but any increase in STIs appears to be minimal. The second is a small study of offering PrEP to participants in an observational cohort of MSM at higher risk of HIV. That study highlights barriers to PrEP uptake that extend beyond a lack of knowledge and willingness among clinicians and beyond drug cost and lack of insurance coverage among potential users of PrEP. 

PrEP Use and STIs 

Molina and colleagues examined behavioral change in an analysis of an open-label PrEP continuation phase of the randomized double-blind IPERGAY PrEP trial [Molina JM, Charreau I, Spire B, et al. Lancet HIV. 2017 Sep;4(9)]. The researchers compared reported behavior and STI incidence during the randomized study period, in which participants received either placebo or PrEP, to STI incidence during the open-label PrEP continuation phase, when all participants received PrEP and knew they were getting it. In this study, the number of reported sex partners and sex acts remained stable, with a modest increase in condomless anal sex (77% to 86%). A numerical increase in bacterial STIs (from 59 to 69 per 100 person-years) was also identified; though the incidence remained high, the increase was not statistically significant. 

Results of this study suggest that (1) there is a population at high risk of acquiring HIV and STIs that can benefit from PrEP and (2) PrEP markedly reduces HIV incidence without leading to a substantial increase in STIs or other potential harm. Educating people about STI risks and encouraging condom use remain important components of sexual health care in general. However, the incidence of STIs in this study demonstrates the valuable role of PrEP in preventing HIV infection among a population at persistently high risk. 

Complex Behavioral Barriers to PrEP 

Despite positive results on PrEP benefits, many of those at the highest risk of HIV infection in some populations are not using PrEP. Awareness, marketing, and access have been identified as barriers; in a just-published observational study of HIV risk and incidence, the researchers describe the interest in and uptake of PrEP among a cohort of young black MSM (YBMSM) participating in the EleMENt longitudinal cohort study [Rolle CP, Rosenberg ES, Siegler AJ, et al. J Acquir Immune Defic Syndr. 2017 Nov 1;76 (3)]. The YBMSM study participants were recruited in Atlanta, GA., between July and December 2016 through social media, venue sampling, and advertisements in public transport. The box below summarizes the characteristics of participants.

At enrollment, participants were screened for HIV, other STIs, and substance use. The researchers subsequently provided education about PrEP and offered it to participants who tested negative for HIV infection. PrEP costs were not directly covered by the study, but the study team sought to provide PrEP at no cost through the use of private insurance and medication assistance programs; no incentives were provided for PrEP-specific visits. 

Study participants (N = 192) and characteristics:

  • Age: 16 to 29 years; median age 24 years
  • Sex: Male
  • Sexual identity: Most identified as gay; all reported ≥1 male sex partner in the previous 3 months.
  • Race: Black, non-Hispanic
  • HIV status: confirmed uninfected
  • Self-reported HIV risk factors:
    • Substance use: ~75%
    • Condomless anal sex within the prior 6 months: ~75%
    • STI diagnosis within the previous 12 months: 25%
  • Using PrEP at start of the study: 4%
  • Heard of PrEP before study: 50%

At the time of study entry, 4% (8) of participants were already using PrEP, which left 184 participants eligible for PrEP initiation. Of those, approximately 50% reported that they had heard of PrEP previously. When PrEP was offered, 10% of participants were not interested, 27% wanted to discuss PrEP further at the next study visit, and 63% (116) were interested and scheduled for an initiation visit. However, only 37 attended a PrEP initiation visit despite repeated attempts by the study staff to schedule a visit. By the end of the observation period, only 34% (63) of participants had initiated PrEP.

The only statistically significant difference between those who started PrEP and those who did not was a higher proportion reporting an STI in the past 12 months: 33% of those starting PrEP and 20% of those not starting PrEP. There was no statistical difference in age, but the proportion of participants over the age of 24 years was greater among those starting PrEP: 67% compared to 54%. The proportion who reported having previously heard of PrEP was also higher among those who started PrEP: 60% versus 49%. Income, sexual identity, relationship status, substance use, and having had an HIV test in the past 12 months were not correlated with choosing to use PrEP. The team also explored reasons for interest or disinterest in PrEP and found that 24% of PrEP users vs. 37% of non-users agreed with the following statement: “I always use condoms,” and 3% of PrEP users vs. 22% of non-users agreed with the following statement: “I would be unhappy about taking a pill every day.”

Important behavioral factors beyond the scope of this study appear to mediate the use of PrEP. This study has important limitations: It included a very small sample size and was geographically limited. Clearly, additional behavioral research is needed to understand how to communicate effectively with diverse groups at higher risk of HIV. 

Challenges in PrEP Scale-Up

These two studies highlight important considerations regarding HIV prevention through increased use of PrEP. The first consideration is that individuals who are most likely to benefit from PrEP have a high incidence of STIs. The STI incidence in Molina’s study confirms that this is a group also at high risk of HIV infection. It also validates the importance of getting PrEP to YBMSM and others who, for a variety of reasons, including number of sex partners and lack of condom use, are at highest risk of acquiring HIV and other STIs. We need to reach this population with PrEP. The second consideration is the on-the-ground challenges to increasing PrEP uptake among a higher risk group in a setting without barriers to access. Public health professionals, community activists, and primary care doctors all have important roles in developing communication strategies and providing effective health communication about PrEP, especially for those at higher risk of HIV infection. 

Subscribe to receive Topics, Trends & Updates by email.

Does HIV Accelerate Age-Related Neurocognitive Decline? (9/17)

By Christopher J. Hoffmann, MD, MPH, Johns Hopkins University School of Medicine

Advanced HIV disease and advanced age are both associated with impaired neurocognitive function. The emergent population of people living and now aging with HIV infection and being treated with potent antiretrovirals has little risk of developing classic HIV-related opportunistic illnesses. However, this population does appear to be at increased risk of conditions associated with aging, such as cardiovascular disease and frailty.

The authors of a recently published study [Goodkin K, Miller EN, Cox C, et al. Lancet HIV. 2017;4(9)] sought to evaluate whether HIV disease accelerates age-related neurocognitive decline. The authors analyzed a subset of data from the MACS (Multicenter AIDS Cohort Study) that included people with and without HIV infection, who had undergone repeated, extensive neurocognitive testing to assess functioning in five domains: information processing speed, episodic memory, executive function, motor function, and working memory. The researchers included medical and demographic data in their analysis to control for multiple potential confounders, such as education, income, alcohol use, other drug use, comorbid diseases, antiretroviral therapy era, HIV stage (using CDC staging), and duration of infection since HIV seroconversion.

Advanced age (>50 years) and advanced HIV stage were each associated with lower neurocognitive scores. Increasing age was associated with lower scores across all domains. Participants with CDC Stage A or B HIV (asymptomatic or early symptomatic) had neurocognitive scores similar to those without HIV. However, those with Stage C HIV (AIDS) had lower scores for motor function and working memory. Furthermore, there was an interaction between aging and advanced HIV leading to a multiplicatively greater decline in motor function and episodic memory.

When the model included time since seroconversion, people living with Stage C HIV (AIDS) also had significantly lower scores on motor function and episodic memory and a greater decline in both with aging, when compared with those with less advanced HIV or those uninfected with HIV. The researchers described an anatomical correlate that may explain these motor function and episodic memory results: The basal ganglia and hippocampus are both affected by HIV, often with early involvement. Motor function depends on the basal ganglia, and the hippocampus is critical for episodic memory.

This study highlights the value of neurocognitive testing to evaluate potential cognitive decline in people living with HIV. Results can help to quantify the level of impairment and help guide diagnoses. The study results suggest that in a patient with a history of advanced HIV infection, the finding of a greater decline in motor function and episodic memory than in other domains is consistent with primary HIV-related brain disease. If other patterns of neurocognitive decline–in executive function, for example–are identified, then a non-HIV-related cause may be more likely and further evaluation may be warranted. 

Questions still to be answered include: (1) What is the long-term effect of HIV infection on someone started on antiretroviral therapy soon after seroconversion who maintained an undetectable viral load long-term? (2) What is the effect of HIV infection among individuals >70 years of age? (3) What is the potential effect on neurocognitive performance over time of HIV medications that are neurotoxic or that have greater central nervous system penetration? I look forward to future publications from this and other groups to further explain the long-term neurocognitive implications of HIV infection and successful HIV treatment.

Subscribe to receive Topics, Trends & Updates by email.

Risk of Frailty Among People with HIV (7/17)

By Christopher J. Hoffmann, MD, MPH, Johns Hopkins University School of Medicine

Effective antiretroviral therapy (ART) is enabling people with HIV infection to live longer. As a result, the HIV-related morbidity experienced by this aging population is shifting from illness caused by opportunistic infections to illness caused by chronic, non-communicable cardiovascular, kidney, and liver diseases. This population is also subject to frailty, an age-related cognitive and functional decline that can decrease a person’s physiological reserve and increase vulnerability to external stressors.

Characteristics associated with frailty: In the HIV-related peer-reviewed literature, frailty is often defined as the presence of 3 or more of the following 5 characteristics: unintentional weight loss, exhaustion, weakness, low physical activity, and slowed walking speed. Frailty increases the risk of falls, hospitalizations, and death; thus, it is the subject of increasing research and clinical attention.

A recently published study by Erlandson and colleagues [J Infect Dis, 2017 Mar 15;215(6)] sought to identify modifiable risk factors for frailty in patients taking ART. The researchers analyzed the frailty assessments of 1,016 participants from the ACTG A5322 observational cohort study, which had recruited participants from prior ACTG studies. All participants were aged ≥40 years, and 85% were aged <60 years; 19% were females; and 48% were White non-Hispanic, 30% were Black non-Hispanic, and 20% were Hispanic. Upon study entry, participants’ frailty was assessed as follows:

  • Walking speed and grip strength: Observation
  • Weight loss: Self-report
  • Low physical activity and exhaustion: Quality of life scales
  • Demographics, economic status, and medical characteristics: Historical study data

Of all study participants, 6% met criteria for frailty, and 38% were pre-frail (had 1 or 2 of the 5 frailty components). Risk factors associated with frailty included the following:

  • Female sex
  • Non-white
  • Receiving Medicaid or Medicare benefits
  • Older age
  • Smoking
  • Less education
  • Lower physical activity
  • HCV antibody positivity
  • Neurocognitive impairment
  • Obesity

ART associated with frailty? Efavirenz use appeared to increase the risk of frailty in the overall analysis, but not in an analysis limited to participants whose initial regimen had been randomized to efavirenz or atazanavir in a prior study (ACTG 5202). Integrase strand transfer inhibitors (INSTIs) had a stronger association with a non-frail state. The authors argue, and I agree, that the findings regarding specific ART agents and frailty should be interpreted with caution because this is not a randomized clinical trial. It is an observational study, and multiple factors may have influenced ART selection. In addition, INSTIs are a newer class of antiretroviral drugs and may reflect changes in care or participant characteristics over time. 

Risks: What was most striking in this study is that many of the factors associated with increased risk of frailty represented structural inequality, such as being a recipient of public health benefits or having less than a high school education. Moreover, factors that are often associated with structural inequality were also associated with increased frailty: female sex, race other than White, poverty, smoking, and obesity. 

These results suggest that preventing frailty may require far more than choosing ART regimens carefully. That said, it’s clear that more studies are needed to better understand how to address frailty and pre-frailty once those conditions are identified. 

Subscribe to receive Topics, Trends & Updates by email.

High-Grade Anal Intraepithelial Neoplasia Recurrence (6/17)

By Christopher J. Hoffmann, MD, MPH, Johns Hopkins University School of Medicine

HIV-infected men who have sex with men (MSM) have a higher risk of anal cancer than men in the general population (65 to 130 per 100,000 person-years versus 5 to 15 per 100,000 person-years). High-grade anal intraepithelial neoplasia (AIN) is considered the precursor to anal cancer, which has high morbidity and mortality rates. High-grade AIN can be suspected on the basis of an abnormal anal cytology result (generally, high-grade squamous intraepithelial lesion or SIL). Although limited data are available to support a specific screening and management approach, abnormal anal cytology is generally followed-up with anal biopsy via high-resolution anoscopy (HRA). Histologically proven high-grade AIN lesions can be ablated using electrocautery, cryotherapy, or chemical means. However, recurrence after treatment persists, as shown in a recently published study that examined factors associated with high-grade AIN recurrence after successful ablation.

The study authors [Burgos J, et al. AIDS 2017 Jun 1;32(9)] used data collected from an observational cohort of MSM with HIV infection whose care was managed in an anal dysplasia clinic at a hospital in Spain. All patients were diagnosed with high-grade AIN and had a biopsy-proven response to electrocautery 6 to 8 weeks after their last treatment session. Follow-up with HRA continued every 3 to 6 months. Recurrent high-grade AIN was defined by biopsy either at the previously treated site or an untreated site.

The study included 141 participants with high-grade AIN who accepted treatment. Of these, 100 (71%) had successful treatment. The median age of successfully treated patients 43 years; the mean CD4 count was 629 cells/mm3, and 70% had an undetectable viral load at the time of treatment. The participants were followed for a median of 13.6 months. During the follow-up period, 39 (39%) developed recurrent high-grade AIN; no patients progressed to invasive anal cancer.

Of those with recurrent high-grade AIN, 37/39 (95%) had a larger initial lesion (lesion affecting two or more octants of the anal canal circumference) than did those without recurrence (45/61 [74%]). Those with recurrence also had a lower nadir CD4 count (232 vs. 326 cells/mm3) and were more likely to have HCV antibodies present (16% vs. 2%). 

Factors not associated with recurrence of high-grade AIN included age, undetectable viral load, a current and stable sex partner, and routine use of condoms.

After 1 year of treatment, 57% of participants had high-grade neoplasia (41 patients did not respond to initial treatment, and 39 patients experienced recurrence). This appears to be a remarkable failure rate for any therapy. This result raises several questions regarding the most effective management of anal neoplasia: What is achieved through electrocautery ablation? Is the risk of anal cancer being reduced with this treatment? How frequently should follow-up occur? What risk factors predict progression to anal cancer?

The take-home message for me is that high-grade anal neoplasia is common and the success with current treatment approaches is limited. There remains a need for improved data and guidance on optimal approaches to managing and preventing anal cancer among MSM.

Subscribe to receive Topics, Trends & Updates by email.

Vitamin D Deficiency Associated with Statin-Induced Muscle Toxicity in Patients Living with HIV (3/17)

By Christopher J. Hoffmann, MD, MPH, Johns Hopkins University School of Medicine

A recently published study supports the association between statin-related myalgia and vitamin D deficiency among people treated with antiretroviral (ARV) agents, with an impressive effect size of 2.3 [Calza L, et al. AIDS 2017 Mar 13;31(5)].

Several antiretroviral (ARV) agents are associated with hypercholesterolemia. In addition, some ARVs, and a pro-inflammatory state caused by HIV itself, can increase the risk of cardiovascular disease (CVD). For these reasons, both hypercholesterolemia and CVD appear to be increased among people living with HIV. Statins play an important role in managing hypercholesterolemia and in reducing the risk of CVD, thereby reducing mortality, among patients living with HIV. The American College of Cardiology and the American Heart Association recommend statins for all individuals at high risk of cardiovascular disease [Stone NJ, et al. J Am Coll Cardiol. 2014 Jul 63(25 Pt B)]. Whether a role for statins exists among people living with HIV who are at low to moderate risk of cardiovascular disease is being evaluated in the REPRIEVE study.

Adherence to statin treatment can be compromised by muscle toxicity. In the general population, between 10% and 25% of individuals taking statins report having myalgia, and many discontinue the therapy as a result. Further, vitamin D deficiency has been associated with statin-induced myalgia in the general population. Given the frequency of vitamin D deficiency among people living with HIV (70%-84%), Calza and colleagues [AIDS 2017 Mar 13;31(5)] performed a retrospective analysis in Italy on a cohort of 545 patients taking ART and either atorvastatin or rosuvastatin. The primary study goal was to evaluate the association between vitamin D level and muscle toxicity using the following markers: myalgia, elevated creatinine kinase level in the absence of myalgia, or myalgia and elevated creatinine kinase level.

The majority of patients were male (80%) and white (91%); 95% had an undetectable viral load. Muscle toxicity was identified among 100 patients (18%), evidenced by myalgia in 42, elevated creatinine kinase without myalgia in 33, and elevated creatinine kinase with myalgia in 25. There was no statistically significant difference in muscle toxicity between patients taking atorvastatin or rosuvastatin. The median time to muscle toxicity was 412 days, with 10% of patients having toxicity within 34 days of starting a statin. Vitamin D deficiency was associated with muscle toxicity. Among patients with muscle toxicity, the vitamin D level was 19 to 22 ng/mL (depending on the presence of myalgia or elevated creatinine kinase level), compared with a level of 32 ng/mL among patients without muscle toxicity. Among the 40 patients with a vitamin D level available at the time of myalgia or elevated creatinine kinase diagnosis, the level was similar to that at baseline. In an adjusted analysis, factors associated with muscle toxicity were age older than 60 years, history of myalgia, vitamin D levels lower than 30 ng/mL (odds ratio, 2.3), and longer duration on statin therapy. Notably, therapy with a boosted protease inhibitor was not associated with an increased myalgia risk.

Two observational studies in the general population suggest that supplementing vitamin D when the level is low may reduce the occurrence of statin-related myalgia [Glueck CJ, et al. Curr Med Res Opin. 2011 Sep;27(9); Khayznikov M, et al. N Am J Med Sci. 2015 Mar;7(3)]. Randomized trials of vitamin D supplementation to reduce statin myalgia have not been completed in the general population or among people living with HIV. However, given the frequency of vitamin D deficiency in patients living with HIV and the potential benefits of a normal vitamin D level on bone health and immune activation [Overton ET, et al. Ann Intern Med. 2015 Jun16;162(12); Fabre-Mersseman V, et al. AIDS. 2014 Nov 28;28(18)], treating vitamin D deficiency with oral vitamin D3 supplementation is reasonable. Achieving a normal vitamin D level may also reduce statin-induced myalgia, increasing the chance of success with statin therapy and ultimately reducing cardiovascular disease.  

Subscribe to receive Topics, Trends & Updates by email.

Quality of Life Improvement with Early ART Initiation (2/17)

By Christopher J. Hoffmann, MD, MPH, Johns Hopkins University School of Medicine 

Results of a just-published study suggest that, in addition to reducing all-cause mortality and improving health overall for patients with HIV infection, early initiation of ART also improves quality of life (QOL). Using data collected at baseline and during follow-up visits for the START trial, Lifson and colleagues assessed whether timing of ART initiation affected patients’ self-assessed QOL (Lifson, AR et al. AIDS. 2017 Apr 24;31(7)).

The START trial, which was conducted in 35 countries, randomized 4,685 participants with CD4 counts >500 cells/mm3 to one of 2 arms: immediate ART initiation or ART initiation at CD4 count <350 cells/mm3. Roughly 50% of participants lived in high-income countries and 50% in low- and middle-income countries. Approximately 75% received an initial ART regimen containing efavirenz, an agent well known to cause CNS and mood side effects.

At each study visit, patients were asked to assess their QOL using 2 methods–a visual analog scale to record their perceived current state of health (scored 0-100) and the SF-12, a widely used survey that includes physical and mental health components. 

QOL data were collected for 4,561 START trial participants, distributed evenly between the immediate and deferred ART study arms. At baseline, physical and mental health scores were similar between the two arms. By 4 months they diverged, with scores improving in the immediate ART arm and remaining relatively unchanged in the delayed ART arm. Scores of deferred ART patients were censored from analysis upon ART initiation.

The difference in QOL scores persisted: throughout 5 years of follow-up; when individuals with severe events were excluded; and through multiple subgroup analyses, including age, gender, race, geographic region, baseline CD4 count, efavirenz ART, and baseline visual analog scale.

This analysis provides further support for universal access to ART upon diagnosis of HIV infection: ART reduces severe health events, reduces HIV transmission, and improves quality of life for otherwise healthy people living with HIV.  

Subscribe to receive Topics, Trends & Updates by email.

A High Rate of Dolutegravir Side Effects in "Real-World" Settings (1/17)

By Christopher J. Hoffmann, MD, MPH, Johns Hopkins University School of Medicine 

In my practice, 5% to 15% of the patients I start on dolutegravir (DTG) experience intolerable headaches or insomnia that necessitate a switch to another agent. This level of intolerance is higher than reported from clinical trials. I always wonder whether the patients of other clinicians have had the same “real-world” experience with DTG side effects and now several recent studies shed light on this issue. 

In one study, researchers from the Netherlands [de Boer MG, et al. AIDS. 2016 Nov 28;30(18)] reported on the tolerability of DTG among 556 patients initiated on DTG between August 2014 and March 2016. DTG was stopped in 76 patients (13.7%) due to adverse drug reactions after a median of 73 days (range 5-327). The combination of DTG with abacavir (ABC) led to a higher overall discontinuation rate (16.3%). The majority of side effects were neuropsychological or psychiatric (insomnia, headaches, mood alterations, and malaise) and gastrointestinal. Discontinuation for neuropsychiatric side effects was 2.3 times more likely among patients taking DTG plus ABC.

Other researchers have also reported neuropsychiatric side effects with DTG. Researchers in Germany [Hoffmann, C et al. HIV Med. 2017 Jan;18(1)] evaluated a cohort of 1,704 patients who initiated an ART regimen containing an integrase strand transfer inhibitor (INSTI) between January 2007 and April 2016. The rates of adverse events leading to discontinuation within 12 months were 7.6% for elvitegravir (EVG), 7.6% for DTG, and 3.3% for raltegravir (RAL). Renal side effects were the leading cause of EVG discontinuation (3.5%); neuropsychiatric side effects were most commonly reported with DTG (5.0%). The most common neuropsychiatric symptoms among the 49 patients who discontinued DTG were insomnia and other sleep disturbances, dizziness, headache, poor concentration or slowed thinking, depression, and painful paresthesia.

An older (2015) case series from France that includes four patients [Kheloufi, F et al. AIDS. 2015 Aug 24;29(13)] described the following symptoms post-DTG-initiation: fatigue and “drunk feeling” in patient 1, headaches and depression in patient 2, worsening psychiatric symptoms in patient 3, and intense headaches (which spontaneously resolved) followed by suicidal ideation in patient 4. 

DTG is a highly potent agent with a high barrier to resistance and unparalleled outcomes in randomized trials [Cahn, P et al. Lancet. 2013 Aug 24;382(9893); Clotet, B et al. Lancet. 2014 June 28;383(9936); Walmsley, SL et al. N Engl J Med. 2013 Nov 7;369(19)], and it appears to be very well tolerated in most patients who do not experience side effects. For these reasons, it has an important role in current antiretroviral therapy (ART). However, assessing for side effects and correctly identifying DTG as a potential cause of new and unexpected symptoms in the days to weeks after initiating this drug is essential to optimal care. My experience with the side effects with DTG has taken a little of the shine off this potent agent and has led me to consider the potential for DTG-associated side effects when selecting a regimen for an ART-naïve patient or planning a switch for failure or convenience.    

Subscribe to receive Topics, Trends & Updates by email.

Is Memory Really the Problem in ART Adherence? (10/16)

By Christopher J. Hoffmann, MD, MPH, Johns Hopkins University

An undetectable HIV viral load is the sine qua non of effective antiretroviral therapy (ART). Once the correct ART regimen is prescribed, achievement of an undetectable viral load then depends on sustained adherence. The authors of a recent study in AIDS Care (Kalichman SC, et al. AIDS Care. 2017 Apr;29(4).) sought to better understand and differentiate between factors associated with moderate and severe non-adherence. Moderate non-adherence was defined as 75% to 95% of medication doses taken as prescribed; severe non-adherence was defined as <75% of doses taken as prescribed.

The authors used computerized interviews to collect patient self-reported data on symptoms, side effects, use of common reminder strategies, depression, life stress, and alcohol consumption. The authors also used urine testing to assess for recent illicit drug use. Over a 1-month period, participants were contacted three times for telephone-based pill counts. Participants who had missed doses were then asked about cognitive, organizational, mental health, substance use, and structural barriers to adherence. 

The team followed 214 severely non-adherent and 342 moderately non-adherent participants for 1 month. A large and equal proportion of patients in each group used memory aids to support adherence, including pillboxes, alarms, routine timing of pills, journals, reminders from family members, or posted reminder signs. Although a larger proportion of the severely non-adherent participants reported having forgotten or being too busy to take medications, the number was not statistically significant. Of the severely non-adherent, 54% reported forgetting compared with 41% of the moderately non-adherent. A larger and statistically significant difference in adherence was related to side effects, running out of pills, and consumption of alcohol (see table below). Depression and stress were also higher among severely non-adherent participants.

Factors Affecting ART Adherence Participants Reporting this Effect,          by Level of Non-Adherence
Severe Moderate
Medication side-effects 12% 4%
Running out of pills 21% 9%
Consumption of alcohol 21% 13%

The results of this study suggest that even when adherence reminder strategies are used, they are not effective in helping patients overcome factors related to both severe and moderate non-adherence, including, in this group, alcohol use, access to medications, and mental health challenges. 

This study did have some limitations. The researchers followed participants for just 1 month, and the results were not strong: 68% of the severely non-adherent had undetectable viral loads, and less than 50% of non-adherence among participants was associated with the factors named above.

Nonetheless, the results highlight the roles that mental health, substance use, and structural barriers may play in reducing adherence to ART. The findings also suggest that other, yet-to-be-identified factors are at play as well. Perhaps most importantly, the standard focus on memory, reminders, and memory aids may not be helpful in significantly improving adherence among those with demonstrated non-adherence, and we need to dig deeper to help our patients take the medications that will suppress their viral loads.

Subscribe to receive Topics, Trends & Updates by email.

Successful Same-Day ART Initiation with a Complex Patient Population (8/16)

By Christopher J. Hoffmann, MD, MPH, Johns Hopkins University

A move to test and initiate treatment on the same day will require changes in traditional approaches to HIV management. Those changes may include restructuring clinic services and identifying individuals who are the best candidates for immediate antiretroviral therapy (ART) initiation. A recent article (Pilcher et al. J Acquir Immune Defic Syndr. 2017 Jan 1;74(1)) describing immediate ART initiation by the San Francisco General Hospital outpatient HIV clinic provides insight into both issues.

Beginning in July 2013, the clinic began to offer same-day ART initiation with a protocol called RAPID. During the entire reported period (6/2013 to 12/2014), only patients lacking private insurance were eligible. In the first 6 months, eligibility for RAPID care was also limited to patients with acute or recent HIV infection, but criteria were liberalized after January 2014. Patients were referred from outpatient clinics and HIV testing centers in San Francisco, and taxi vouchers were available for patients to be transported from across the city. 

The RAPID protocol provided same day HIV education, blood draws for laboratory testing, accelerated insurance approval efforts, and a directly observed first ART dose. Patients for whom insurance approval was not obtained were given a 5-day starter-pack. This process took 3 to 4 hours. Within the 7 days after the initial visit, a nurse followed up by phone.

During the 18-month study period, 39 patients who would be considered “complicated” in most treatment settings were initiated on the RAPID protocol (2 patients per month). Patients’ CD4 counts ranged from 3 to 1391 cells/mm3, 27% were considered homeless, 42% had a “major” mental health disorder, and 42% reported use of any illicit substance. Despite these challenges, 37 patients (95%) initiated ART within 24 hours of their clinic visit. Approximately 90% had a viral load <200 copies/mL within 6 months after clinic referral. Only 4 (10%) were lost from care (no visit in a 6-month period); 8 (20%) transferred care to another clinic.

Although limited by its very small sample size and low volume of patients during the observation period (2 patients per month), this report is valuable for its description of a same-day ART initiation program that succeeded with complex patients. With the right clinical resources, these results suggest that most people living with HIV can initiate ART rapidly with success. This report also underscores that ART should not be delayed automatically because of mental illness, substance abuse, and/or unstable housing. 

Subscribe to receive Topics, Trends & Updates by email.

Blood Pressure Control and Medication Adherence among People Living with HIV (7/16)

By Christopher J. Hoffmann, MD, MPH, Johns Hopkins University

Hypertension (HTN) is a common comorbidity among people living with HIV (PLH), affecting 20% to 43%. This proportion is likely to grow with the aging of the PLH population (see Okeke NL, et al. Clin Infect Dis 2016 Jul 15;63(2), for time trends in HTN diagnosis among PLH). Recent studies suggest the need to educate patients about the risks of uncontrolled hypertension.

The pivotal 2015 SPRINT study [SPRINT Research Group, et al. NEJM 2015 Nov 26;373(22)] quantified a reduction in mortality and cardiovascular outcomes with tight versus routine HTN control among individuals who were not HIV-infected. This large, randomized, controlled trial involved 9,361 individuals 50 years of age and older with systolic blood pressure (SBP) between 130 and 180 mmHg (median 140 mmHg) and cardiovascular risk factors other than diabetes. Participants were randomized to one of two arms, with results reported after a median of 3.3 years of follow-up. The results are summarized in the table below.

  Treatment Arm
Participants Standard Intensive
Number 4,683 4,678
Target SBP mmHg <140 <120
Median no. HTN meds 1.8 2.8
Serious adverse events 2.5% 4.7%
No. reaching primary clinical outcome
(hazard ratio = 0.75, p < 0.001)
319 243

The primary clinical outcome was defined as one of the following: myocardial infarction, acute coronary syndrome, stroke, heart failure, or cardiovascular death. All-cause mortality was also substantially lower (hazard ratio = 0.73, p = 0.003) in the intensive treatment arm. Approximately 38% of participants in each arm experienced side effects, although serious adverse events, such as hypotension, syncope, electrolyte abnormalities, and acute kidney injury, were higher in the intensive treatment arm. In all, the SPRINT study demonstrated that a goal of SBP<120 mmHg reduced mortality and cardiovascular outcomes among individuals over the age of 50.

The PLH population already on antiretroviral therapy (ART) has experience with multiple medications and with regimens that require adherence as strict as those for HTN. 

Weiss and colleagues [J Acquir Immune Defic Syndr. 2016 Dec 1;73(4)] recently studied the effects of adding both medications and exhortations for strict adherence to the treatment regimens of HIV-infected participants. The researchers recruited 117 participants with HIV and HTN and 37 participants with HIV and chronic kidney disease (CKD) and followed them with health belief surveys and electronic monitoring of adherence to HIV, HTN, and CKD medications. At the 10-week follow-up, 88% of participants had virologic control and 60% had HTN control (SBP<140 mmHg). However, there was no difference in adherence to ART and HTN medications.

ART adherence was associated with participants’ reports of greater understanding of HIV, the perception of greater severity of HIV, and the view that ART medications are more necessary than medications for other conditions. This study suggests that HTN control is worse than virologic control among PLH, even though adherence to the two types of medication may be similar.

Taken together, the results of these two studies suggest that clinicians caring for PLH may need to intensify efforts to educate patients about HTN and its consequences and the potential benefit of a blood pressure even lower than the standard <140 mmHg. 

Subscribe to receive Topics, Trends & Updates by email.

Cardiovascular Risk Scores and CVD Risk Reduction among People Living with HIV (6/16)

By Christopher J. Hoffmann, MD, MPH, Johns Hopkins University

Primary methods of preventing cardiovascular disease (CVD) with smoking cessation, control of hypertension, and use of aspirin and HMG-CoA reductase inhibitors can substantially reduce the risk of myocardial infarction. Guidelines from the American College of Cardiology/American Heart Association (ACC/AHA) recommend risk-based prescription of HMG-CoA reductase inhibitors. The ACC/AHA guidelines use a specific population-based risk calculator; however, other scores are also available, including the Framingham Risk Score and the D:A:D cardiac risk score for people living with HIV (PLH). Among PLH, it remains unclear how well these scores predict the risk of myocardial infarction caused by plaque rupture.

Monroe and colleagues [Monroe AK, et al. AIDS. 2016 Aug 24;30(13)] used data from the Multicenter AIDS Cohort Study (MACS), a multi-site longitudinal cohort of HIV-infected and -uninfected men who have sex with men, to assess the performance of risk calculators for predicting coronary atherosclerosis, coronary calcium, or coronary stenosis of greater than 50%, as measured using cardiac computed tomography (CT). The authors reported a poor association between cardiac risk score and the presence of any plaque (area under the curve is 0.65 for PLH; an area under the curve of 0.5 indicates no correlation). Among PLH who were at low risk according to the ACC/AHA formula, 36% had a coronary artery calcium score greater than 0, 63% had any plaque (calcified and non-calcified), and 11% had greater than 50% stenosis. Performance was marginally better for those without HIV infection; however, 53% with any plaque and 5.4% with greater than 50% stenosis were in the low-risk category.

There are several potential interpretations of these findings:

  • The MACS population is different from everyone else, and risk calculators do not apply.
  • Compared with the general population, PLH have different needs and risk profiles for some aspects of primary care, such as CVD risk management, and calculators may not be the best approach to CVD risk management in PLH.

Limitations of this study include its use of a cohort that is not representative of the general PLH population and its use of coronary calcium score and atherosclerosis on CT as opposed to hard endpoints, such as CVD events.

Studies such as this raise questions regarding the use of HMG-CoA reductase inhibitors among PLH. In other words, are we treating the right people? The ongoing REPRIEVE clinical trial is aimed at helping to answer this question.

Subscribe to receive Topics, Trends & Updates by email.

Predicting Kidney Disease Risk to Guide ART Selection (4/16)

By Christopher J. Hoffmann, MD, MPH, Johns Hopkins University

Kidney disease is common among people living with HIV. Multiple factors, including HIV-associated nephropathy, antiretroviral therapy (ART), hepatitis B (HBV) or C (HCV) virus co-infection, diabetes, hypertension, and primary kidney disease can cause or contribute to kidney dysfunction. More common factors, though, are transient fluctuations in serum creatinine and estimated glomerular filtration rate (eGFR). Reducing the occurrence of comorbidities associated with kidney disease and limiting the diagnostic uncertainty of eGFR fluctuations may influence ART-prescribing decisions.

In a recently published article [AIDS 2016 Jun 1;30(9)], Flandre and colleagues report on the relationship between various initial ART regimens and the risk of chronic kidney disease (CKD). The researchers used a predictive score, developed and validated through the D:A:D cohort, that defined CKD as a persistent eGFR <60 mL per min per 1.73 m2.

The study cohort included 6,301 patients receiving care at 12 reference centers in France, all of whom initiated ART after January 1, 2004. All patients had at least one pre-ART and two on-ART creatinine measures, and eGFR was determined with the MDRD equation. Patients were stratified into low, medium, and high risk for CKD using the D:A:D score for CKD. This score accounts for a wide variety of factors, including type of HIV exposure (injection drug use [IDU] vs non-IDU), HCV co-infection, age, pre-ART eGFR, sex, nadir CD4 count, hypertension, cardiovascular disease, and diabetes. Incident CKD occurred in 211 (3.4%) of participants and was associated with older age, lower pre-ART eGFR, and lower nadir CD4 count. The probability of CKD by risk score was as follows: low risk, 0.65%; medium risk, 4.6%; and high risk, 15.9%.

CKD risk was highest for patients receiving a boosted protease inhibitor (PI) with tenofovir disoproxil fumarate (TDF). The risk was lower for patients receiving a non-nucleoside reverse transcriptase inhibitor (NNRTI) with or without TDF. Notably, among low-risk patients, the choice of an NNRTI vs a PI or TDF vs no TDF made no apparent difference in CKD risk.

This study is helpful in assessing the appropriateness of TDF or boosted-PI regimens for specific patients. For instance, avoidance of an initial ART regimen that includes TDF or a boosted-PI may be best for patients with a high D:A:D score for CKD. Patients with a low CKD risk score do not appear at increased risk for CKD from TDF or boosted PIs. These types of data may also be useful in securing insurer authorization to prescribe specific medications, such as tenofovir alafenamide (TAF)-containing combination tablets, and in discussing ART regimen side effect profiles with patients.

Taking a risk-based approach when choosing an ART regimen does not require actual calculation, but taking into account a patient’s D:A:D score and risk factors for CKD is encouraged. Such an approach may reduce the number of patients requiring switches to less nephrotoxic regimens due to elevated eGFR and may reduce the incidence of CKD among people living with HIV.

Subscribe to receive Topics, Trends & Updates by email.

Life Expectancy after an HIV Diagnosis (3/16)

By Christopher J. Hoffmann, MD, MPH, Johns Hopkins University

One of the first questions a patient newly diagnosed with HIV may ask is, “How long will I live?” A new study from CDC researchers, which reports data on life expectancy from the time of diagnosis and average years of life lost (AYLL)*, may help in answering that question [Siddiqi, et al. J Acquir Immune Defic Syndr 2016 Jun 1;72(2)].

Using CDC HIV Surveillance Report data from all 50 US states and the District of Columbia, the researchers estimated life expectancy by year of HIV diagnosis for individuals reported to have been diagnosed during the four years from 2008 to 2011.  Life expectancy was further stratified by HIV exposure category, sex, and race/ethnicity.  The authors did not consider post-diagnosis factors such as engagement in care or virologic suppression.

Overall, between 2008 and 2011, the average life expectancy at the time of HIV diagnosis increased by 3.5 years, from 25.4 to 28.9 years. By transmission category, life expectancy at the time of diagnosis in 2011 was highest (32.3 years) among men having sex with men (MSM); it was lowest (20.2 years) among men injecting drugs. Analysis by race indicated that life expectancy was highest among Hispanics/Latinos, followed by blacks, whites, and other races (range: 26.1 to 30.8 years). Average years of life lost were not similarly stratified. With regard to sex, for all men diagnosed at age 25, the AYLL was 16.8 years; for women diagnosed at age 25, AYLL was 23.2 years.

This analysis demonstrates continued overall improvement in HIV survival, which may reflect increasing efforts to engage people living with HIV in care and to initiate antiretroviral therapy (ART). However, the analysis also highlights discrepancies in survival by population group, which most likely reflects better engagement in care among MSM than among injection drug users. And while the study reinforces the value of ART for achieving longevity, its findings suggest that nearly normal life expectancy with HIV infection depends not only on access to ART for all patients but also on support to overcome systemic social, economic, and substance use challenges and maintain engagement in care and adherence to ART.

*AYLL is calculated by subtracting life expectancy with HIV from life expectancy for the general population. 

Subscribe to receive Topics, Trends & Updates by email.

Detection of Invasive Anal Cancer in HIV-Infected Patients (1/16)

By Christopher J. Hoffmann, MD, MPH, Johns Hopkins University School of Medicine

Although invasive anal cancer is relatively rare, the disease can be devastating. The incidence of this cancer is increased among people living with HIV, especially among those who engage in receptive anal sex. For someone living with HIV in the United States or Canada, the cumulative incidence for anal cancer by the age of 65 has been estimated to be 1.3% compared with 0.03% for someone without HIV [Silverberg et al., Ann Int Med 2015 Oct 6;163(7)]. This cumulative incidence exceeds those for colorectal, liver, and pharyngeal cancer.

Clinical guidelines for the prevention of anal cancer borrow heavily from the cervical cancer literature, with limited data to support specific screening and management practices. Routine screening generally includes anal cytology without HPV typing. A recent article by D’Souza and colleagues adds to the knowledge base of anal cancer. They studied the incidence and clearance of squamous intraepithelial lesions, as well as high-risk HPV infection, among HIV-infected and uninfected men who have sex with men (MSM) in the United States [J Acquir Immune Defic Syndr. 2016 Apr 15;71(5)]. The authors describe longitudinal follow-up for 1511 men with repeat anal cytology separated by 2 years; 723 were living with HIV; 788 did not have HIV. Most of those with HIV had undetectable viral loads (78%), and the median CD4 cell count was 583 cells/mm3. Abnormal cytology (of any kind) was identified among 40% of the men living with HIV and in 25% of those without HIV. Only 1.5% of men had high-grade squamous intraepithelial lesions. However, 2.4% had atypical squamous cells for which high-grade lesions could not be excluded. Among HIV-infected men who received repeat cytologic testing after initial ASCUS (atypical squamous cells of undetermined significance) or low-grade squamous intraepithelial lesions, 56% regressed to normal cytology; 81% of those with high-grade cytologic findings regressed to a lower grade. High-risk HPV 16 was identified among 25% of men living with HIV and 16% of those without HIV. Among 220 men who received follow-up high-resolution anoscopy with biopsy after cytologic findings of high-grade lesions, 17% had histologically confirmed disease.

This study adds to the data on the high prevalence of anal HPV infection and cellular changes among MSM with HIV. It also highlights the limitation of cytology to identify individuals who should receive high-resolution anoscopy and biopsy, given the frequency of regression of lower-grade lesions to normal. Additional studies are underway that could shed more light on the prevention of anal cancer and the management of high- and low-grade lesions. Until results are available from those studies, the common practice of routine cytologic testing and referral for high-grade lesions, as well as for atypical squamous cells for which high-grade lesions cannot be excluded, remains a reasonable approach.

Subscribe to receive Topics, Trends & Updates by email.