Patterns in two proteins found in airway tissue were associated with how well people with COPD responded to inhaled corticosteroids, according to a secondary analysis of 172 patients. The reported associations involved HDAC2 and HDAC8 in airway smooth muscle and other tissue compartments, but the findings do not show that either protein causes a better response.
Patients counted as responders had higher nuclear HDAC2 in airway smooth muscle cells than nonresponders. The adjusted average was 69.86% positive nuclei among responders, compared with 59.39% among nonresponders, a difference reported with a p-value of 0.0219.
The clearest differences were found in airway tissue
The study defined an inhaled-corticosteroid responder by improvement in FEV1 over 12 months. Responders also had higher HDAC8 expression than nonresponders across multiple non-epithelial compartments and in glandular epithelium. In the multiple-compartment comparison, the adjusted averages were 32.67 for responders and 23.81 for nonresponders, with a p-value of 0.0228. In glandular epithelium, the corresponding averages were 57.29 and 45.67, with a p-value of 0.0466.
A separate response measure pointed in the same direction. Among patients whose COPD Assessment Test score fell by at least two points over 12 months, glandular epithelial HDAC8 averaged 4.21% positive tissue, compared with 2.5% among those who did not meet that response definition. The reported p-value was 0.0316.
The protein patterns also varied with disease severity. Compared with patients at GOLD stage 1 to 2, those at stage 3 to 4 had lower nuclear HDAC2 in glandular epithelium, with adjusted averages of 62.7% and 73.7% positive nuclei, respectively. They also had lower HDAC2 in airway smooth muscle cells, at 58.03% versus 64.78%.
The same severity comparison showed lower nuclear HDAC8 in severe disease. In glandular epithelium, the adjusted averages were 43.8% positive nuclei for GOLD stage 3 to 4 disease and 53.33% for stage 1 to 2. In airway smooth muscle cells, they were 20.24% and 28.47%, respectively.
Baseline treatment was linked to lower expression
Patients who were already receiving inhaled corticosteroids at baseline had lower nuclear HDAC2 than patients who were not receiving them in stromal tissue, glandular epithelium and airway smooth muscle cells. They also had lower HDAC8 in airway smooth muscle cells. After adjustment for GOLD stage, all of the reported associations remained significant except the stromal HDAC2 comparison.
These baseline comparisons cannot be read as evidence that inhaled corticosteroids lowered protein expression. Baseline treatment was not assigned within this analysis, so other differences between the groups could have contributed to the pattern.
A smaller follow-up group showed a change in HDAC8
The investigators also examined repeat biopsies from 44 people at baseline and after 12 months. No significant change in HDAC2 was observed. Responders, however, had a greater reduction in HDAC8 than nonresponders across all tissue compartments, in stromal tissue and in stromal-cell cytoplasm.
Across all compartments, the adjusted mean HDAC8 change was minus 0.149 among responders and 0.056 among nonresponders, with a p-value of 0.0478. The reported difference was also significant in stromal tissue and stromal-cell cytoplasm. The authors say the pattern needs mechanistic confirmation.
What the analysis measured
The work was a secondary analysis of endobronchial biopsy tissue from patients enrolled in the HISTORIC randomised, double-blind, placebo-controlled trial. The analysis related tissue expression to disease severity, baseline ICS use and response classifications. It does not specify how the parent trial's treatment allocation or placebo control was used in the HDAC comparisons.
Biopsy slides were digitised at 40 times magnification. Pathologists reviewed the HALO artificial-intelligence tissue classification, and the investigators quantified protein staining at the pixel or cell level, including separate nuclear and cytoplasmic compartments. Both proteins were found predominantly in the nucleus, with weaker staining in the cytoplasm.
The findings are not yet a clinical test
Most of the analyses were predominantly cross-sectional, so they mainly compared tissue measurements between response groups rather than establishing a cause-and-effect pathway. The authors say this limits causal inference and call for single-cell transcriptomics and functional assays to determine the roles of HDAC2 and HDAC8 in specific cell types.
The researchers interpret the compartment-specific patterns, especially in airway smooth muscle, stromal tissue and glandular compartments, as potentially linked to differences in inhaled-corticosteroid responsiveness. They suggest the proteins could be considered as biomarker or therapeutic-target candidates, while the evidence does not yet validate them for choosing treatment.
The authors present these proteins as candidates for further study, not as a validated test or established treatment target. The analysis does not establish that changing either protein would improve clinical outcomes or guide treatment selection.
The article is presented as a Letter to the Editor and was supported by the Berta-Ottenstein-Program, with funding information deposited in the Open Funder Registry.
Paper data and sources
Original title: Expression of HDAC2 and HDAC8 in the airways of COPD patients is associated with responsiveness to ICS: insights from the HISTORIC study.
Authors: Eleni Papakonstantinou, Marta Nowak, Spasenija Savic Prince et al.
Journal/Repository: The European respiratory journal
Status: Peer-reviewed
First online: 2026-10-09
DOI: 10.1183/13993003.00950-2026
Original paper · Full text