New EMA Pharmacovigilance Recommendations Expected for Long-Acting HIV Regimens
At upcoming regulatory meetings, interim study reports on treatment adherence and resistance to integrase strand transfer inhibitor (INSTI)-based antiretrovirals will be presented.
EMA Rewrites the Rules for Long-Acting HIV Therapy: Why 6.3% Discontinuation in Real-World Practice Is a Catastrophe
Insight that won't make it into official EMA protocols: The regulator is set to revise pharmacovigilance for long-acting HIV regimens not because they are dangerous, but because real-world data show that patients are dropping injections and clinicians don't know what to do with residual drug concentrations that persist in the blood for over a year. The Spanish RELATIVITY cohort (n=3,146) showed a 6.3% discontinuation rate. The Tuscan LAHIV cohort (n=191) reported 9.2%. The Atlanta PrEP clinic saw 19.6% discontinuation of cabotegravir, with 68% occurring after 1-3 doses. When every 10th to 15th patient drops out of a program, EMA cannot pretend that "everything is under control."
[The Core]: What Is Really Happening
In reality, EMA is not just going to "hear reports" but will change the very philosophy of monitoring long-acting regimens. This involves three fundamental issues. First: standardization of monitoring protocols for CD4 T-cells and lymphocytes when developing new drugs. In December 2025, the FDA placed a clinical hold on the combination of GS-1720 and GS-4182 (Gilead candidates) due to unexpected declines in CD4 and absolute lymphocyte count. This echoes the story of islatravir (Merck), which was halted in 2021 due to lymphopenia and never fully returned. EMA requires that future studies of long-acting injectables (LAI) include "predefined CD4 decline thresholds and clear discontinuation criteria," not just "observation."
Second: real-world resistance to second-generation integrase inhibitors has proven not to be "zero" but measurable. The Italian ARCA cohort (n=1,296) showed 9.3% virologic failures, of which 5.7% had INSTI resistance mutations. In a Ukrainian study (Donetsk, Luhansk, Zaporizhzhia), the frequency of INSTI mutations was 3.6%, with high-level resistance to cabotegravir in 42.9% of the 14 patients with mutations. These are not "isolated cases." This is a signal that the genetic barrier of dolutegravir and bictegravir is not absolute. EMA wants pharmacovigilance to track not only the frequency of resistance but also the spectrum of mutations in real time, using standardized algorithms (e.g., ANRS-MIE).
Third, and most importantly: residual concentrations and safety after treatment discontinuation. Cabotegravir and rilpivirine remain in the blood for up to 12 months after the last injection. In the Atlanta PrEP study, 36% of patients who discontinued CAB-LA did not transition to oral PrEP, and only 11% underwent recommended HIV testing. One seroconversion was documented 3 months after discontinuation. EMA requires that new recommendations specify not "recommended" but "mandatory": monitoring every 3 months for one year after discontinuation. This will increase the burden on clinics but will prevent the "window of vulnerability" when drug concentrations are insufficient to suppress the virus but sufficient to select for resistance.
[Timeline and Context]
December 2021: FDA places clinical hold on islatravir (Merck) due to lymphopenia. June 2025: FDA again places a hold—this time on the combination GS-1720 + GS-4182 (Gilead) due to CD4 decline. September 2025: EMA publishes deadlines for PRIME applications in 2026, including accelerated assessment for LAI. December 2025: Publication of a Viewpoint in The Lancet HIV, where authors state: "The clinical hold of GS-1720/4182 is a key safety signal. Future trials should implement standardized monitoring protocols with individual CD4 trajectories and clear discontinuation criteria."
January 2026: Publication by Soulie et al. in JAC: despite an increase in INSTI use from 3.3% to 71.2% over 16 years, resistance to dolutegravir, cabotegravir, and bictegravir in 2024 was only 2.7%. This calmed the market. February 2026: Publication of the Italian ARCA cohort at the STI conference—9.3% virologic failures, but only 5.7% with resistance. March 2026: The Spanish RELATIVITY cohort (n=3,146) in the journal AIDS shows 6.3% discontinuation, with age ≥70 years increasing risk by 5.41 times and foreign nationality by 2.06 times.
April 2026: The Tuscan LAHIV cohort in J Med Virol—9.2% discontinuation, virologic failure rate 2.3/100 person-years. May 2026: Ukrainian study (n=392) in Viruses—3.6% INSTI mutations, with 42.9% showing high-level resistance to cabotegravir. June 2026: At CROI, VH-184 is presented—a third-generation INSTI from ViiV/GSK with an improved resistance profile against bictegravir. And now, June 2026—the EMA meeting where all this data is being reviewed.
Why now? Because lenacapavir (Sunlenca from Gilead) received FDA approval for HIV prevention in June 2025, and the European Commission approval in August 2025. In 2026, widespread implementation is expected. EMA wants to establish rules for monitoring lenacapavir before it is administered to millions. Islatravir failed during development. Lenacapavir is already on the market. The regulator will not allow a repeat of past mistakes.
[Who Wins and Who Loses]
Winner #1: Gilead Sciences (unexpectedly). Their lenacapavir is a capsid inhibitor with a different mechanism of action and safety profile. EMA, spooked by the islatravir and GS-1720 stories, will be more lenient toward Gilead when registering their next combinations (e.g., lenacapavir + broadly neutralizing antibodies). Moreover, Gilead can offer its monitoring protocol as the "gold standard," setting the market trend. Their stock rose 4% after the EMA meeting announcement.
Winner #2: ViiV Healthcare (GSK) with their new INSTI VH-184. At CROI 2026, they showed that VH-184 retains activity against viruses resistant to bictegravir, including mutations G118R, R263K, N155H. This is a third-generation INSTI that could become a "resistance killer." EMA, concerned about rising cabotegravir mutations (42.9% high-level resistance in the Ukrainian study), will accelerate VH-184 review. Forecast: PRIME designation by end of 2026.
Loser: Elderly patients and migrants. RELATIVITY showed: age ≥70 years—HR 5.41 for treatment discontinuation. Foreign nationality—HR 2.06. Reasons: logistical difficulties (injections every 2 months require transportation, language support), drug interactions, comorbidities. New EMA recommendations will likely include "enhanced monitoring" for these groups—meaning more visits, more tests. But this won't solve the access problem. Migrants will still drop out, and EMA will collect statistics.
Silent loser: EU healthcare systems. The Spanish cohort showed: 3.5% discontinuation due to structural reasons (relocation, insurance change, incarceration). The Atlanta clinic: 23% discontinuation due to patient relocation. EMA may require the creation of a "national network of referral clinics for LAI" so that patients can receive injections in any EU city. This will require millions of euros for coordination, staff training, and IT infrastructure. EU country budgets are not infinitely elastic. Who will pay? Patients through taxes, pharma companies through additional fees—or no one, and EMA recommendations will remain on paper.
[What the Media Isn't Saying]
The first dirty secret: INSTI resistance mutation testing is expensive and not always available. In the Italian ARCA cohort, genotyping at virologic failure was available for only 53 of 120 patients (44%). In real-world practice, even fewer. The cost of one test is $300-500 USD. Insurers don't always cover it. Without genotyping, the clinician doesn't know whether to switch the patient to another INSTI (e.g., from cabotegravir to bictegravir) or to a new class. New EMA recommendations will require testing in 100% of virologic failures. But who will pay? No one. And 56% of failures will remain without a genotype.
The second omission concerns cabotegravir resistance in Eastern European countries. The Ukrainian study (Donetsk, Luhansk, Zaporizhzhia) showed 3.6% INSTI mutations, with high-level resistance to cabotegravir in 42.9% of them. The predominant subtype is A6, which dominates the region (97.45%). This means that the spread of Ukrainian refugees with HIV to Europe (estimated 20,000-30,000 people) carries strains with reduced susceptibility to cabotegravir. EMA knows this but does not discuss it publicly to avoid panic. New pharmacovigilance recommendations will include "enhanced monitoring of individuals arriving from regions with high prevalence of INSTI resistance."
And the third layer, which even industry analysts are silent about: the 3.6% INSTI mutation rate in the Ukrainian cohort is an underestimate. Because patients with less than 8 years of therapy were excluded from the study (median cohort duration 101 months). Among "newer" patients, resistance may be higher since they received dolutegravir with less stringent control. If the true resistance rate among new diagnoses in Eastern Europe is 5-7%, then by 2030 we will face an epidemic of INSTI-resistant HIV on a scale comparable to the early non-nucleoside reverse transcriptase inhibitor (NNRTI) epidemic in the 2000s.
[Forecast: Next 30 Days and 90 Days]
Next 30 days:
Expect the publication of the draft EMA recommendations for LAI monitoring. The document will contain three key provisions: (1) mandatory baseline CD4 and lymphocyte determination before starting therapy, repeated every 12 weeks; (2) discontinuation threshold: CD4 decline of 30% from baseline or absolute count <200 cells/µL; (3) requirement for EU member states to create registries of patients who discontinued LAI, with mandatory HIV and RNA testing at 1, 3, 6, and 12 months. This will increase the burden on clinics but provide EMA with real data for future decisions.
Also, ViiV Healthcare will announce the start of a Phase IIIb study of cabotegravir+rilpivirine in patients with A6 subtype mutations. This is a direct response to the Ukrainian data. The study cost is $20-30 million USD.
Next 90 days:
The key event is the meeting of the Committee for Medicinal Products for Human Use (CHMP), which will consider adding a warning to the cabotegravir+rilpivirine label: "Reduced efficacy in patients with subtype A6 and mutations G140R, Y143R, R263K." This is an official acknowledgment that the drug is not universal. GSK shares will drop 2-3% after publication.
And the final, most stringent forecast: EMA and FDA will jointly issue a "Guidance for Industry: Development of Long-Acting Antiretrovirals," which for the first time will mandate "standardized CD4 monitoring protocols" for all LAIs, including preclinical studies. This will be a direct consequence of the GS-1720/4182 clinical hold. Developers of new LAIs (e.g., VH-184 from ViiV) will have to include not only pharmacokinetics but also detailed immunological monitoring in Phase I/II. Development costs will rise by 10-15%. Small biotechs without deep pockets will not be able to compete. The LAI market will consolidate around 3-4 giants by 2028.
The story of LAI pharmacovigilance is a story of regulators catching up with reality. Patients want injections every 2 months. Clinicians want simplicity. Pharma companies want profit. And EMA wants no one to die. To that end, they are ready to rewrite all the rules. Analysts—keep your finger on the pulse of CD4. Patients—don't stop injections without warning. World—get ready for an era where long-acting therapy becomes not an "innovation" but a "new normal" with new risks.
— Editorial Team
