Best IPF Drug Combinations in 2026: What AI Research Reveals

Idiopathic pulmonary fibrosis treatment has historically relied on single-drug therapy. Nintedanib and pirfenidone, both approved in 2014, remain the standard of care. In October 2025, the FDA approved nerandomilast (formerly BI 1015550), the first new IPF drug in over a decade. But the question that drives modern research is whether combinations of drugs can outperform any single agent.

Why Drug Combinations Matter

IPF involves multiple biological pathways simultaneously. TGF-beta drives fibroblast activation. Senescent cells accumulate and secrete inflammatory factors. The extracellular matrix stiffens, creating a feedback loop that accelerates disease. No single drug targets all of these mechanisms.

The rationale for combination therapy is straightforward: attack multiple pathways at the same time. This approach has proven effective in oncology, HIV treatment, and tuberculosis. In IPF, clinical evidence for combinations is still emerging, but computational modeling provides strong theoretical support.

Standard Monotherapy Results

In clinical trials, nintedanib reduces FVC decline by approximately 50% compared to placebo. Pirfenidone achieves a similar reduction. Nerandomilast, in the FIBRONEER-IPF trial, showed a reduction in FVC decline of approximately 63% at 52 weeks. These are meaningful results, but none of these drugs stop progression entirely, and none reverse existing fibrosis.

Top Combinations from the PF-Atlas Solver

The PF-Atlas genetic algorithm has tested over 2 million drug combinations against a biological model of lung fibrosis. Each combination is scored on FVC improvement, toxicity, pill burden, drug-drug interactions, and cost. The top results fall into three categories.

Best 2-Drug Combination

Nintedanib + nerandomilast emerges as the strongest dual therapy. In simulation, this combination slows FVC decline by approximately 75% compared to no treatment. The two drugs have complementary mechanisms: nintedanib inhibits tyrosine kinases (FGFR, PDGFR, VEGFR) while nerandomilast inhibits PDE4B, reducing inflammation and TGF-beta signaling. Side effect overlap is minimal.

Best 3-Drug Combination

Adding pamrevlumab (anti-CTGF antibody) to the nintedanib + nerandomilast pair produces the top 3-drug result. Pamrevlumab targets connective tissue growth factor, a downstream mediator of TGF-beta that directly promotes collagen production. The Phase 3 ZEPHYRUS trials for pamrevlumab were discontinued in 2023 due to insufficient efficacy as monotherapy, but the PF-Atlas solver suggests it may be more effective as part of a combination.

The 7-Drug Regenerative Protocol

The most ambitious result from the solver is a 7-drug sequential protocol that achieves FVC improvement from 50% to 69.5% in 52-week simulation. This suggests not merely slowing disease but partially reversing it. The protocol uses:

1. Nintedanib (anti-fibrotic baseline)
2. Pirfenidone (anti-inflammatory + anti-fibrotic)
3. Nerandomilast (PDE4B inhibition)
4. Pamrevlumab (anti-CTGF)
5. PRM-151 / pentraxin-2 (macrophage modulation, ECM repair)
6. Saracatinib (Src kinase inhibition, anti-fibrotic)
7. GLPG1690 / ziritaxestat (autotaxin inhibition, LPA pathway)

The drugs are introduced sequentially over 6 phases, not all at once. This reduces toxicity risk and allows dose titration. The protocol details, including dosing schedules and monitoring targets, are available in the Treatment Sequencer panel of PF-Atlas.

Nintedanib + Pirfenidone: The Existing Evidence

The INJOURNEY trial (2018) studied nintedanib + pirfenidone and found the combination was tolerable, with GI side effects being the primary concern. The study was not powered for efficacy, but PF-Atlas modeling suggests a modest benefit over either drug alone. For patients already stable on one drug, adding the second may provide incremental improvement.

Limitations

All combination results from PF-Atlas are computational. The biological model, while detailed, is a simplification of real human physiology. Clinical validation is essential before any combination protocol is used in practice. Patients should not alter their treatment without consulting their physician.

The PF-Atlas solver results are designed to guide research priorities and help physicians understand which combinations have the strongest theoretical basis. They are not medical advice.