HIV and ID Observations

The International AIDS Conference — or “AIDS 2018” — returned to Amsterdam for the first time since 1992.

It’s worth pausing, with gratitude, to remember that 26 years ago antiretroviral therapy consisted of three available drugs — zidovudine (AZT), didanosine (ddI), and zalcitabine (ddC).

All were marginally effective, with limited durability and significant toxicity. In fact, the big debate in the early 1990s was whether treatment of HIV before symptom onset was worthwhile at all.

We had no NNRTIs, PIs, or integrase inhibitors. Aside from recommending condom use and abstinence, we had no strategies to prevent HIV among sexually active adults.

Fast-forward to today, where the goal is to have everyone with HIV on treatment for both their individual benefit and to end the epidemic.

Gratitude indeed!

Here are some highlights from the meeting which took place last week, a Really Rapid Review© of important (or just interesting) studies that caught my eye. Apologies ahead of time for missing your favorite — as always, let me know in the comments section what I missed!

• As initial therapy, dolutegravir plus lamivudine was non-inferior to dolutegravir plus TDF/FTC at week 48. Additional favorable results from the much-awaited GEMINI studies included no treatment-emergent resistance, comparable activity in patients with HIV RNA >100,000, and (not surprisingly) less effect on renal and bone markers. The only caveat was a lower response rate in those with baseline CD4 <200 due to miscellaneous reasons (not virologic failures). Unclear what caused this imbalance.
• Zero HIV transmissions among serodiscordant MSM couples if the infected partner had virologic suppression. This follow-up to the PARTNER study of serodiscordant couples — all not using condom — enrolled additional gay men to provide greater assurance about this remarkable benefit of HIV therapy in this subgroup. Those who are quantitatively minded (!) might be reassured that there were no transmissions with an estimated 76,991 episodes of condomless anal intercourse. Results also published here. The audience hearing the presentation burst into spontaneous applause with presentation of these results. Yes, as I have noted here numerous times, the Swiss were right a decade ago, which allows me to use that pretentious (but appropriate) word “prescient.”
• Investigators updated the preliminary data on dolutegravir during conception and the risk of neural tube defects. Publication of their initial report is here in the NEJM; reported at the conference were additional newborn outcomes for women who conceived while on DTG, with no additional cases through July 15. This brought the estimated incidence down to 4/596 (0.67%, 95% CI 0.26%, 1.7%) — still higher than the approximately 0.1% incidence seen with non-DTG ART (mostly EFV). Further observations are expected through March 2019, which may help clarify this important issue.
• More women initiating ART with DTG rather than EFV-based regimens in late pregnancy achieved an undetectable viral load. Note that there is no evidence that DTG treatment started during pregnancy is associated with any adverse outcomes in the newborn, though it does seem that babies metabolize this drug slowly.
• On-demand PrEP with TDF/FTC in France appears to be working as well as daily treatment. This is an analysis of effectiveness in the community, not a clinical trial, among approximately 1500 individuals (mostly MSM) choosing between on-demand or daily PrEP. Both interventions had zero seroconversions with 7 months of follow-up, with an estimated 85 infections averted based on historical data. Note that the recently released IAS-USA Guidelines also endorse on-demand therapy as an option.
• Melbourne and San Francisco demonstrated impressive declines in new HIV diagnoses. The rapid initiation of ART for newly diagnosed individuals is strongly correlated with this improvement. Here’s an amazing statistic — in 2016, there were only 233 people newly diagnosed with HIV in all of San Francisco, less than half the number (523) in 2008.
• When divided by quintile, states with the highest PrEP uptake had a decline in new HIV infections, while those in the lowest quintile had stable to increased HIV incidence. This association was made stronger by the observation that estimated virologic suppression rates were similar in each of the quintiles. I can’t quite understand how they made this last estimate, but it does support that PrEP reduces HIV acquisition in communities with higher use.
• Dolutegravir plus rilpivirine maintained virologic suppression at 96 weeks. This two-drug regimen is approved for patients currently suppressed who have no resistance to these drugs. There were three patients with virologic rebound and RPV resistance, none with DTG resistance. While the study did not give follow-up on these 3, presumably they could be successfully treated with NRTIs plus DTG or BIC, so the consequences of this resistance are small.
• In treatment-naive studies, bictegravir-FTC-TAF demonstrated high efficacy in patients with high baseline HIV RNA or low CD4 cell counts. The same was true for the comparator arms — dolutegravir plus TAF/FTC or dolutegravir/ABC/3TC. These per-protocol analyses show that with vanishingly few exceptions, patients stable enough to enter our clinical trials of current ART will achieve virologic suppression if they take their medications.
• The single tablet of DCF-TAF was effective in a prospective “rapid start” study. Impressively, 29% of the 107 patients were enrolled within 48 hours of HIV diagnosis. This combination tablet (just approved) does not require results of resistance, HLA-B5701, or hepatitis B testing.
• Should we routinely be doing resistance testing before starting therapy today if the first regimen is INSTI-based? This analysis (disclosure: I’m a co-author) suggests that such testing is no longer cost-effective. Remember, our current transmitted drug resistance does not impact outcomes in INSTI-based regimens.
• One patient developed DTG resistance when started on TDF/FTC plus DTG as initial therapy. As with the previously reported case, the patient had a low CD4 cell count and high viral load (CD4 39, HIV RNA 457,000), and an active infection. She was also receiving rifampin as adjunctive treatment for Staph aureus, with appropriate doubling the dose of DTG. Resistance occurred via the R263K pathway, with rapid virologic rebound. The case is a reminder both how rarely this occurs, and that clinical trials do not include our most difficult to treat patients — they need especially close follow-up after starting ART.
• As second-line therapy after failure of NNRTI-based regimens, DTG plus NRTIs is superior to LPV/r plus NRTIs. These 48-week results confirm the practice-changing data presented last year, and were consistent across most baseline characteristics favoring DTG. A separate presentation provided some resistance data, though hopefully more are forthcoming. Two individuals experienced virologic failure with DTG resistance with an unusual G118R mutation. Surprisingly, neither of them had much NRTI resistance.
• Resistance to the second-generation integrase inhibitors dolutegravir, bictegravir, and cabotegravir is highly correlated. The G140S + Q148H mutations plus any other integrase mutation can lead to high-level resistance. In other words, it does not appear that resistance to one of these agents could be salvaged with different one — fortunately this resistance remains quite rare. The incidence of resistance to cabotegravir in the upcoming phase 3 studies of long-acting cabotegravir-rilpivirine will be a critical secondary endpoint.
• Doravirine was superior to darunavir/ritonavir at 96 weeks. This difference was driven by a higher discontinuation rate in the darunavir/ritonavir arm. Approval of doravirine is anticipated in October 2018.
• Approximately 1% of doravirine-treated patients in treatment-naive studies developed resistance. For those of you who like to memorize resistance mutations, V106I and F227C were commonly identified. Two interesting aspects of doravirine resistance are 1) most of the isolates retained susceptibility in vitro to certain other NNRTIs (especially etravirine); 2) the F227C containing viruses were hypersusceptible to MK-8591, and these two drugs are now in Phase 2 studies.
• Two studies evaluated dolutegravir monotherapy as a maintenance strategy — and both provide further evidence that we shouldn’t do it. The first was in patients initially treated during primary infection, who may have a lower HIV reservoir. In this small study, it appeared to work — but of course we can’t measure viral reservoirs in practice, so results are of little practical value. Furthermore, one patient with chronic infection was enrolled erroneously and failed. The second study (“MONCAY”) in chronic HIV infection had excessive virologic failures after week 24, with 2/5 patients developing DTG resistance. Time to stop MONCAYING around, folks!
• Heart failure appears to more common in people with HIV than in uninfected controls. This did not appear to be mediated by atherosclerotic disease. It’s possible this is an effect of viral replication, since at the start of the study most of the patients were not receiving ART.
• In acute hepatitis A, the clinical illness with HIV was milder but more prolonged compared to those who were HIV negative. These data come from a large outbreak in Taiwan among MSM, 63% of whom were HIV posititive. Hepatitis A vaccination appeared to be  critical in controlling the outbreak.
• Vedolizumab had no effect on time to virologic rebound after ART interruption in people with HIV. (Results reported within the linked plenary session.) Remember hearing about the remarkable effect of antibodies to Alpha4Beta7 integrin in SIV infected macaques? A refresher — it appeared to block virologic rebound after ART treatment discontinuation. Not only did the intervention not work in humans, but the effect was not reproduced in monkeys either. Oh well.
• A minority population of CXCR4 virus pre-stem cell transplant led to virologic rebound in a patient who received cells from a CCR5-negative donor. This case report highlights the unique nature of the one case of HIV cure, and is a reminder that it will take more than host engraftment of CCR5-negative cells to have the same result.
• HIV therapy augmented with both vorinostat and HIV immunization did not reduce the viral reservoir. The study was performed in patients treated initially during acute HIV, the group widely believed to be most amenable to eradication strategies. This is the so-called “kick and kill” approach of stimulating the latent reservoir and then striving to eliminate it with an augmented immune response. Both the vorinostat and the vaccines did their jobs, but the reservoir remained stable. Oh well (again).

Amsterdam remains a gem, a marvelous city with distinctive architecture, great food, nice people, picturesque canals, and incredible museums. It was even better than the last time I was here — seemed cleaner, bustling but less congested (at least with car traffic), and decidedly less seedy. The fact that it was unbelievably hot didn’t deter many of us (including me) from biking around town, joining the throngs who regularly take advantage of the flat terrain and the extensive bike lanes.

But visitors beware — when it’s time to go home, the airport is a chaotic, crowded mess. The warnings to arrive 3 hours before departure time should not be taken lightly. I’d call the airport a zoo, but that might be insulting to some zoos!

.@KLM: Worst airline & cust svc experience ever. Long lines @Schiphol, checked in, then >45 min line for baggage drop. Still 30 min pre-takeoff, told to rebook bcs too late to check baggage, but no seats avail x2days! So, to LHR on BA, hotel, int’l change fee to ATL=$1800!

— Melanie Thompson (@drmt) July 28, 2018

Poor Melanie! Hartsfield will seem downright peaceful by comparison.

Next year’s meeting will be in Mexico City, July 21-24, 2019. It too is a wonderful, vibrant city (and where I’ve had some of the best restaurant meals in my life) — but based on the traffic the last time we were there, bike transport seems highly unlikely!

So what did I miss?