COPD Pharmacotherapy in the GOLD 2026 Era: The Eosinophil-Guided Algorithm and the New Agents

COPD pharmacotherapy follows a stepwise, individualized approach guided by symptom burden, exacerbation risk, and blood eosinophil count (GOLD 2026; Nowalk et al., JAMA 2025). The goals: reduce dyspnea and exercise limitation, reduce the frequency and severity of exacerbations, improve health status — and, with triple therapy, potentially reduce mortality.

Initial therapy by GOLD group

Group A (low symptoms, low exacerbation risk): a bronchodilator — SABA, SAMA, LABA, or LAMA.

Group B (higher symptoms, low exacerbation risk): LABA + LAMA preferred.

Group E (high exacerbation risk): LABA + LAMA preferred; consider initial triple therapy (LABA + LAMA + ICS) when blood eosinophils are ≥300 cells/μL.

When initiating long-acting monotherapy, LAMA is generally preferred over LABA for its superior exacerbation reduction — LAMAs also reduce hospitalizations. A network meta-analysis of 99 studies (n=101,311) found LABA + LAMA associated with fewer exacerbations than LABA alone (HR 0.70), LAMA alone (HR 0.87), or ICS + LABA (HR 0.86). And one rule worth repeating: LABA + ICS dual therapy is not encouraged in COPD — if an ICS is indicated, triple therapy is preferred.

The eosinophil rule for ICS

ICS is never monotherapy in COPD; it is added to long-acting bronchodilators based on blood eosinophil count and exacerbation history. Eosinophils ≥300 cells/μL: strong indication — consider initial triple therapy. 100–299 cells/μL: ICS may be beneficial, especially with continued exacerbations on LABA + LAMA. Below 100 cells/μL: ICS is unlikely to benefit and may increase pneumonia risk without reducing exacerbations.

The trade-off is real: ICS in COPD carries a pneumonia signal (RR ~1.43, dose-dependent, higher with fluticasone than budesonide), plus oral candidiasis (RR 2.96), dysphonia, and skin bruising. Consider de-escalating ICS after 12 months of stability — particularly with eosinophils <100 cells/μL, no exacerbation history, or a pneumonia event or other adverse effect.

Triple therapy and mortality

Two landmark trials anchor triple therapy: IMPACT (reduced exacerbations versus dual bronchodilation, with a mortality benefit) and ETHOS (8,509 patients — all-cause mortality 46% lower with LABA + LAMA + ICS versus LABA + LAMA; HR 0.54). Initiate as first-line for Group E with eosinophils ≥300 cells/μL, for continued exacerbations on LABA + LAMA with eosinophils ≥100, and whenever concomitant asthma is present (ICS is mandatory there).

Escalation beyond triple therapy

For persistent exacerbations, let the eosinophil count route the add-on. Eosinophils <100 cells/μL: roflumilast (chronic bronchitis phenotype, FEV1 <50%, frequent exacerbations; start 250 μg daily for 4 weeks, then 500 μg) or azithromycin (250 mg daily or 500 mg three times weekly; works best in former smokers; 27–42% exacerbation reduction; baseline ECG to exclude QTc >450 ms). Eosinophils ≥300 despite triple therapy: dupilumab or mepolizumab — both now FDA-approved for COPD with an eosinophilic phenotype. In BOREAS, dupilumab cut moderate/severe exacerbations 30% (rate ratio 0.70), with an NNT of 3 in pooled analysis. Benralizumab, by contrast, is not approved for COPD — its phase III trials did not reduce exacerbations.

The newest agent

Ensifentrine (Ohtuvayre) is a first-in-class dual PDE3/PDE4 inhibitor — bronchodilation plus anti-inflammatory activity in one nebulized molecule (3 mg twice daily). In the ENHANCE-1 and ENHANCE-2 trials it improved FEV1 (87–94 mL) and reduced moderate/severe exacerbations (rate ratio 0.59), with a safety profile comparable to placebo — no significant GI effects, unlike oral roflumilast. GOLD 2026 notes its positioning remains unsettled: the trials were not designed to test it on top of LABA + LAMA or triple therapy.

Before any escalation: the device check

More than two-thirds of patients make at least one error using their inhaler. Verify technique with the teach-back method before declaring therapy failure, reassess at every visit, and match the device to the patient: MDI with a spacer when hand-breath coordination is poor, avoid DPIs when peak inspiratory flow is inadequate (switch to MDI plus spacer, or SMI), and nebulizers when cognition or dexterity is limited. Simplify regimens when possible — single-inhaler triple therapy improves adherence versus multiple devices.

The bottom line

Check blood eosinophils before adding or removing ICS. Check inhaler technique before escalating anything. And retire LABA + ICS without a LAMA in COPD.