GSK reported a solid second quarter and laid out a sweeping cost-cutting plan, signalling a strategic pivot toward oncology as it prepares for the loss of key HIV patents.

The numbers

Total sales rose 5% year-over-year to £8.4 billion, led by specialty medicines, which jumped 14% to £3.8 billion.

GSK’s dolutegravir HIV franchise brought in £1.4 billion, with the two-drug therapy Dovato at £749 million and the asthma drug Nucala at £610 million.

Why the HIV franchise is so valuable

Dolutegravir is an integrase inhibitor, blocking the enzyme HIV uses to insert its genetic material into a host cell’s DNA. It became a backbone of treatment because it is potent, well tolerated, and has a high barrier to resistance — the virus struggles to develop resistance without also becoming less able to replicate.

Dovato’s significance is the two-drug part. Standard HIV therapy long used three agents, and reducing that to two while maintaining suppression means less cumulative drug exposure over a lifetime of treatment — which matters increasingly as people with HIV live normal lifespans and accumulate the effects of decades of medication.

It also generates predictable, durable revenue: patients take it indefinitely, adherence is high, and switching is uncommon once suppression is achieved.

Cutting to invest

The company is targeting £1.9 billion (about $2.5 billion) in savings by 2029, with the majority redirected into pipeline acceleration and dealmaking rather than pure margin.

That framing distinguishes this from defensive cost-cutting. A company cutting to protect earnings is managing decline; one cutting to fund research is reallocating — and the distinction is verifiable over time by whether R&D spending actually rises.

The plan involves undisclosed job cuts and includes closing operations in Stevenage while opening a £400 million flagship R&D centre at the Cambridge Biomedical Campus.

Why move research to Cambridge

Relocating research is expensive and disruptive, so the rationale is about proximity.

The Cambridge Biomedical Campus concentrates a major teaching hospital, university research institutes, and a dense cluster of biotech companies on one site. Being there provides access to academic collaborators, clinical investigators and patients for early trials, and a labour market of scientists who can move between institutions without moving house.

Clusters of that kind demonstrably produce more research output per unit of investment than isolated sites, largely through informal contact and staff mobility. The cost is losing established teams who will not relocate — a real loss of institutional knowledge that such moves consistently incur.

The move is “a catalyst for faster, bolder medicines discovery,” said Chief Scientific Officer Tony Wood.

The oncology pivot

GSK is racing to build an oncology-heavy pipeline — 11 of 20 Phase 3 studies starting in 2026 are in cancer — before its dolutegravir patents expire between 2028 and 2030.

The concentration is deliberate and carries risk. Oncology is where the most late-stage assets and highest prices are, and it is also the most competitive therapeutic area, with the highest late-stage failure rate.

GSK also exited oncology substantially in 2015, selling its portfolio, then re-entered later — so it is rebuilding capability others never gave up, which is a harder position than it appears from the trial count.

The timing problem

A Phase 3 trial starting in 2026 reads out around 2029 at the earliest, with approval and launch following. The patent cliff begins in 2028.

Why the HIV cliff is unusually steep

Patent expiry affects every pharmaceutical company, and the HIV case has features that make it sharper than most.

These are small-molecule drugs, so what arrives at expiry is generic competition rather than biosimilars. Generics enter quickly, at a fraction of the price, and substitution is automatic in many settings — a pharmacist dispenses the generic unless specifically prevented. Revenue erosion is measured in months.

Prescribing patterns compound it. HIV treatment is guideline-driven and highly standardised, with the same regimens used across large patient populations. Once a generic version of a backbone drug exists, guidelines and payers move to it, and the market shifts almost entirely rather than gradually.

Combination products offer partial protection, since a fixed-dose combination has its own patents and generic manufacturers must replicate the formulation. That defers rather than prevents erosion.

The strategic response available is a next-generation product patients can be moved onto before expiry — which is why long-acting injectable HIV therapy, dosed monthly or less often rather than daily, matters to companies in this position quite apart from its clinical merits.

Which means these programmes largely arrive after the revenue starts falling rather than before. The quarter suggests the strategy is gaining traction, but the real test is whether reinvested savings produce approvals fast enough to offset the coming losses — and on the published timeline, the answer is closer to partly than to yes. Business news, not investment advice.