.png)
Released: November 06, 2020
Expiration: November 05, 2021
Jean-Michel Molina, MD, PhD:
We will begin by discussing the 2-drug regimen DTG plus 3TC, first as an initial option in treatment-naive patients and then as a switch strategy for those already virologically suppressed.
GEMINI-1 and -2 are parallel, international, randomized, double-blind phase III noninferiority studies comparing DTG plus 3TC (n = 716) with DTG plus FTC/TDF (n = 717) for first-line ART.1,2 A few key entry criteria warrant our attention when examining results from these studies. First, patients had no major resistance associated mutations (RAMs), and they were not coinfected with hepatitis B virus (HBV). In addition, baseline HIV-1 RNA level was capped at 500,000 copies/mL, which makes generalizing the data to every PWH a bit more difficult.
Previously, noninferiority of the 2-drug regimen, as measured by the proportion with HIV-1 RNA < 50 copies/mL, was demonstrated at both Week 483 and Week 96.4 New at HIV Glasgow 2020, we saw Week 144 results, which again reinforced the durability of virologic suppression and the noninferiority of DTG plus 3TC compared with DTG plus FTC/TDF.5 In the pooled analysis, 82% vs 84% of patients, respectively, remained virologically suppressed after 3 years of treatment, for an adjusted treatment difference of -1.8% (95% CI: -5.8% to 2.1%), which fell well within the prespecified noninferiority margin of -10%.
Marina Klein, MD, MSc, FRCP(C):
Of note, there seems to be a slight difference between GEMINI-1 and -2 in terms of their adjusted treatment differences at Week 144. GEMINI-2 clearly balanced at zero whereas GEMINI-1 fell left of center at -3.6%. This begs the question, “Were there important differences between the study populations that could explain this shift?”
Jean-Michel Molina, MD, PhD:
It could be interesting to examine the populations separately for their baseline demographics, particularly their CD4+ cell counts and HIV-1 RNA levels.
Jean-Michel Molina, MD, PhD:
In the GEMINI studies, confirmed virologic withdrawal (CVW) was defined as a second and consecutive HIV-1 RNA value meeting any of these criteria: decline from baseline < 1 log10 copies/mL (unless HIV-1 RNA < 200 copies/mL) by Week 12; confirmed plasma HIV-1 RNA ≥ 200 copies/mL at or after Week 24; or confirmed rebound (ie, HIV-1 RNA ≥ 200 copies/mL after confirmed consecutive HIV-1 RNA < 200 copies/mL).3
None of the 21 patients with CVW at Week 144 (1% to 2% per arm) developed treatment-emergent INSTI or NRTI resistance mutations. However, one patient receiving DTG plus 3TC who did not meet criteria for CVW but who reported nonadherence developed M184V at Week 132 (HIV-1 RNA 61,927 copies/mL) and R263R/K at Week 144 (HIV-1 RNA 135 copies/mL). This led to a 1.8-fold change in DTG susceptibility, but the patient regained virologic suppression after a switch to DTG once daily plus darunavir/cobicistat after Week 144.
Three of the 21 patients met criteria for CVW between Week 96 and Week 144—1 in the DTG plus 3TC arm and 2 in the DTG plus FTC/TDF arm. Although we do not yet have details regarding the baseline demographics of these 3 newer CVWs, we do recognize from CROI 2020 that at Week 96, 8 of 11 CVW patients in the DTG plus 3TC arm had baseline CD4+ cell counts > 200 cells/mm3 and 6 of 11 patients had baseline HIV-1 RNA levels ≤ 100,000 copies/mL (Capsule Summary).6 Similarly, for DTG plus FTC/TDF recipients, 5 of 7 and 4 of 7 patients, respectively, also had these advantageous entry characteristics.
Jean-Michel Molina, MD, PhD:
When Week 144 efficacy data (FDA Snapshot) were stratified by baseline HIV-1 RNA level or CD4+ cell count, trends similar to previous subgroup analyses of GEMINI7 were repeated—namely, in the small subset with a baseline CD4+ cell count ≤ 200 cell/mm3, we notice a striking difference where only 67% of DTG plus 3TC recipients achieved virologic suppression vs 76% with DTG plus FTC/TDF.
Marina Klein, MD, MSc, FRCP(C):
By contrast, at the examined HIV-1 RNA threshold of 100,000 copies/mL, there were no apparent differences in efficacy between treatment arms.
Jean-Michel Molina, MD, PhD:
At this point, it is difficult to make a distinction between the CD4+ cell count signal being real or there being an insufficient number of patients in that subset to draw firm conclusions. To be on the safe side, it is best to mirror your patient selection for this dual therapy after the GEMINI entry criteria. Order a genotype to ensure the patient has no baseline resistance. Confirm that there is no HBV coinfection, that the HIV-1 RNA level remains ≤ 500,000 copies/mL, and that the CD4+ cell count registers ≥ 200 cells/mm3. Current DHHS,8 IAS-USA,9 and EACS10 guidelines recommend first-line DTG/3TC only for patients with baseline HIV-1 RNA < 500,000 copies/mL. The newly updated IAS-USA recommendations mention that “perhaps” CD4+ cell count should be considered.9,10
Marina Klein, MD, MSc, FRCP(C):
I agree. It is clear that this is a small subgroup of the entire study population, but caution is warranted. Generally speaking, treatment-naive persons with low CD4+ cell counts frequently have high HIV-1 RNA levels. We know from IDWeek 2019, that at Week 96 in GEMINI, the higher the baseline HIV-1 RNA level, the more pronounced the difference in efficacy between arms. Virologic suppression at Week 96 occurred in 69% of DTG plus 3TC recipients compared with 80% of DTG plus FTC/TDF recipients who had baseline HIV-1 RNA levels > 500,000 copies/mL (Capsule Summary).7 Now, here again, the subsets were small (n = 13 and n = 15, respectively), but it is not surprising to see this divergence. It would be valuable to have similar data for Week 144, providing a more granular view of outcomes with higher baseline HIV-1 RNA levels.
Jean-Michel Molina, MD, PhD:
In terms of safety, the numbers and profiles of AEs were comparable across arms. Study investigators did call out a statistically reduced risk of drug-related AEs with DTG plus 3TC (20% vs 27% with DTG plus FTC/TDF; relative risk: 0.76; 95% CI: 0.63-0.92) but did not provide clarity on the kinds of toxicities observed by regimen, only that most AEs were grade 1.
Marina Klein, MD, MSc, FRCP(C):
Because we tailor therapy to individual patients, it would be important to know which specific drug-related AEs occurred in each arm to understand if there were regimen-specific trends.
The study investigators also reported a numerically higher mean weight gain at Week 144 with DTG plus 3TC vs DTG plus FTC/TDF (3.7 vs 2.4 kg, respectively). This latter magnitude of weight change contrasts with the phase III ADVANCE trial where, at Week 144, among first-line DTG plus FTC/TDF recipients, we saw mean gains of 7.4 kg for women and 5.5 kg for men (Capsule Summary).11
Jean-Michel Molina, MD, PhD:
In thinking about this difference, we should recall the study population for ADVANCE, which enrolled in South Africa. In the DTG plus FTC/TDF arm of that trial (n = 351), 59% of patients were female, 100% were Black, and patients had a median baseline CD4+ cell count of 322 cells/mm3. Juxtapose with that the pooled analysis of GEMINI where 14% of DTG plus FTC/TDF recipients (n = 717) were female, 70% were White, and patients had a mean baseline CD4+ cell count of 461 cells/mm3. These characteristics—sex, race, and immunologic status—may at least partially account for the comparatively lower weight gain observed in GEMINI.
Marina Klein, MD, MSc, FRCP(C):
As far as potential toxicity concerns during long-term follow-up, the worsened renal and bone parameters with TDF-containing ART vs DTG plus 3TC confirmed what we expected.
For the cholesterol markers, outcomes statistically favored DTG plus FTC/TDF at Week 144, but I suspect that some of the P values overestimated the differences between arms in the sense that these are relatively small changes overall. As an example, I am not confident that the observed difference in total cholesterol–to–high-density lipoprotein cholesterol ratio is a clinically significant deviation.
Marina Klein, MD, MSc, FRCP(C):
We will now move from discussing DTG/3TC as a first-line regimen to examining its efficacy and safety as a switch option for virologically suppressed patients. Based on results from TANGO, in August 2020, the FDA expanded the indication for DTG/3TC to include use as a switch regimen for adults with stable virologic suppression on their current regimen, no history of treatment failure, and no known resistance to DTG or 3TC.12
TANGO is an international, open-label phase III noninferiority study that randomized 741 adults with stable viral suppression on tenofovir alafenamide (TAF)–based ART to either switch to DTG/3TC or continue their baseline regimen.13 Enrollees had no previous virologic failure, no current HBV coinfection, and no resistance to INSTIs or NRTIs. The study population was predominantly White (79%) and male (92%) with median baseline CD4+ cell counts of 682-720 cells/mm3 across arms. At trial entry, the median duration of TAF use was 17.7-18.2 months, and most patients (66%) had been receiving cobicistat-boosted elvitegravir as their third agent. Previously, at the Week 48 analysis, the switch to DTG/3TC proved noninferior to continuing TAF-based ART with ≤ 0.5% of patients per arm having HIV-1 RNA ≥ 50 copies/mL (Capsule Summary).14
At HIV Glasgow 2020, Week 96 results were presented, reinforcing the noninferior virologic efficacy of a switch to DTG/3TC in a Snapshot analysis of the intention-to-treat, exposed (ITT-E) population.15 The proportion with HIV-1 RNA ≥ 50 copies/mL had crept up to ≤ 1% per arm, but the adjusted treatment difference of -0.8% (95% CI: -2.0% to 0.4%) still fell comfortably below the noninferiority margin of 4%. Likewise, DTG/3TC was noninferior to continued TAF-based ART for HIV-1 RNA < 50 copies/mL with 86% vs 79% of patients, respectively, achieving this secondary endpoint. There, the adjusted treatment difference was 6.8% (95% CI: 1.4% to 12.3%), again falling within the noninferiority margin, in this case of -8%.
Marina Klein, MD, MSc, FRCP(C):
The COVID-19 pandemic has undoubtedly affected clinical trial follow-up. Where TANGO was concerned, 16 DTG/3TC recipients and 28 TAF-based ART recipients lacked Week 96 virologic data for this reason. With that in mind, in addition to the standard ITT-E analysis, study investigators provided an “efficacy data evaluable” analysis, which excluded these participants with pandemic-related missing values.
Jean-Michel Molina, MD, PhD:
Yes, in ITT-E analysis, the ORRs were a bit lower than would be expected for participants stably suppressed at baseline. From other switch studies, we are accustomed to seeing numbers in the 90% range, if not 95% or 96%. Here, the ITT-E numbers were 79% to 86%. However, when patients lacking virologic data because of the COVID-19 outbreak were omitted, suppression rates rose to 90% for the switch to DTG/3TC and 85% for continued TAF-based ART.
Marina Klein, MD, MSc, FRCP(C):
There were no CVWs in the DTG/3TC arm, but even among the 3 cases with TAF-based ART, no resistance developed. This is always reassuring.
Jean-Michel Molina, MD, PhD:
It is important to note that the study population here was not just any patient suppressed on TAF-based ART. Participants were eligible only if they were receiving a first-line regimen of TAF/FTC plus a PI, NNRTI, or INSTI, or if they had switched TDF to TAF ≥ 3 months before screening with no other regimen changes. As Dr. Klein mentioned when introducing the trial, these patients had no history of virologic failure, and both of these points are critical considerations when switching patients in the absence of genotypic data.
Jean-Michel Molina, MD, PhD:
The number of drug-related AEs resulting in study withdrawal was higher in the DTG/3TC arm (n = 14) than with TAF-based ART (n = 3). This is not completely unexpected as patients in the latter arm have been stably suppressed and are not changing therapy whereas switch patients are likely being exposed to new drugs and reacting accordingly. Among the AEs causing withdrawal with DTG/3TC were key neuropsychiatric and metabolic events we have previously seen associated with DTG use including depression, anxiety, insomnia, fatigue, irritability, disturbance in attention, and increased weight.
Marina Klein, MD, MSc, FRCP(C):
It is also worthwhile to discuss the observed biomarker changes at Week 96. Many of the small numerical differences that were statistically significant are another example of where P values may be misleading. The investigators tested numerous renal and lipid parameters with most differences being nonsignificant or just bordering on statistical significance (ie, P < .05). One notable instance of this borderline significance was estimated glomerular filtration rate (eGFR) based on cystatin C. It was unexpected to see a greater reduction with TAF-based ART.
Jean-Michel Molina, MD, PhD:
Right. That is not easy to understand or to explain and is inconsistent with other data. I would expect to see no difference in eGFR by cystatin C. The greater reduction in creatinine-based eGFR with the switch to DTG/3TC was anticipated based on known inhibition of tubular creatinine secretion with DTG. This phenomenon does not necessarily translate to a true reduction in eGFR.
Marina Klein, MD, MSc, FRCP(C):
I agree. Overall, there does not seem to be a major renal signal with TAF-based ART, and what we do see with DTG/3TC is expected based on inhibition of creatinine secretion. The numerical difference for cystatin C is relatively small, and we cannot be confident that it has any clinical relevance.
Marina Klein, MD, MSc, FRCP(C):
By now, based on overwhelming evidence, we should agree that ART can reduce HIV-1 RNA to a threshold at which the risk of sexual transmission is effectively zero. Clinical trials such as HPTN 052,16 PARTNER,17 Opposites Attract,18 and PARTNER219 underpin this “U=U” message where undetectable equals untransmittable.
To query awareness and acceptance of the U=U message, investigators in Italy performed a nationwide web-based survey of 90 ID physicians, 397 PWH, and 634 persons at risk of HIV infection based on their sexual behaviors.20 This design is valuable for its inclusion of both community members and providers. In examining these 3 diverse subgroups, we gain a sense of U=U from all populations for whom it could have impact.
Among at-risk persons, only 46% were aware of U=U, a proportion that rose to 74% for PWH. Although 92% of physicians acknowledged awareness of U=U, most striking to me were the metrics that 21% of physicians did not rank the U=U concept as highly accurate, and only 34% of PWH reported definitive positive U=U–related messaging from their providers. Until we adequately educate all stakeholders, the idea of HIV treatment as a valid and important means of HIV prevention cannot fully permeate society.
Jean-Michel Molina, MD, PhD:
Among PWH, factors associated with U=U awareness included longer receipt of ART, higher education, and being bisexual patients or men who have sex with men vs heterosexual individuals. These latter 2 characteristics similarly predicted PWH belief of U=U.
Marina Klein, MD, MSc, FRCP(C):
It is not surprising to see educational level affecting U=U endorsement. We may need to tailor our messaging in populations less likely to read on the Internet or keep up with the literature. We would also expect men who have sex with men—and bisexual individuals, to a lesser extent—to be more in tune to U=U than heterosexuals based on targeted efforts in prevention and treatment for these populations.
Belief in U=U among providers and at-risk persons seemed to hinge, in part, on exposure to PWH. For physicians, managing care for PWH strongly predicted belief. For HIV-negative respondents, having a partner with HIV infection similarly increased the likelihood of belief. Conversely, never being tested for HIV predicted an at-risk person’s reduced belief in U=U. This may be a particularly key population in which focused efforts could modify risk perception, reduce stigma, and shift individuals toward seeking pre-exposure prophylaxis.
Ultimately, there are quite a few messages about how we might adapt our educational campaign across different groups to disseminate U=U more broadly. Doing so could help increase, among physicians, earlier treatment of PWH and more ready adoption of the test and treat strategy.
Back
