Biomarkers aid Type-2 inflammation phenotype selection
medwireNews: Researchers have identified a panel of clinically accessible biomarkers that may help select asthma patients with an airway-mucosal T-helper 2 (Type-2) inflammation phenotype.
“As the majority of effective anti-inflammatory agents work best in a Type-2 high populations e.g. [inhaled corticosteroids], and anti-IL [interleukin]-13 antibodies, defining Type-2 status in patients using clinically accessible biomarker[s] will be important,” comment Philip Silkoff (Janssen Research & Development LLC, Spring House, Pennsylvania, USA) and colleagues.
Using the ADEPT dataset, they defined 83 patients with asthma as being Type-2-inflammation high or low based on the airway mucosa expression of three genes driven by the Type-2 cytokine IL-13, compared with 25 non-atopic healthy individuals.
All three genes – CC-motif chemokine ligand (CCL)-26, periostin, and multi-gene IL-13 in vitro signature (IVS) – provided segregation, but this was optimal with CCL26, leading to it being the “primary anchor” for Type-2 status, the researchers report in The Journal of Allergy and Clinical Immunology.
This airway-mucosa CCL26 Type-2 phenotype also had the best concordance with airway eosinophilic inflammation, with high sputum eosinophilia levels of 3% or above detected in 88% of patients. The association was not limited to a CCL26-high status, however, with high sputum eosinophilia levels also detected in 45% of 31 patients with CCL26-low status.
While airway-CCL26-high status was not associated with disease characteristics, such as airflow obstruction, it was strongly associated with increased levels of the Type-2 clinical markers fractional exhaled nitric oxide (FeNO), blood eosinophils, and serum CCL26 and CCL17.
All 13 patients with airway-mucosa-CCL26-high status and moderate-to-severe-asthma had high FeNO (≥35 ppb) or high blood eosinophils (≥300 cells/mm3) compared with a minority (36%) of the 42 patients with airway-mucosal-CCL26-low status.
The researchers note that while routine measurement of FeNO and blood eosinophils in asthma patients means it could be readily implemented as a co-diagnostic test to screen for the Type-2-high phenotype, a poor positive predictive value (PPV) of 46% hampers its use as such, despite a perfect negative predictive value (NPV) of 100%.
Adding high serum CCL26 levels (75th percentile of the healthy individuals) to the high FeNO or high blood eosinophilia classification improved the PPV to 88%, while maintaining a good NPV of 89%. But the full panel of markers, including high serum CCL17 levels, provided the best PPV of 100%, with an NPV of 87%. The findings were similar for periostin and IL-13-IVS but less pronounced.
“These last 2 classification models may be useful to prospectively select patients for Type-2-targeted therapeutic clinical studies, highly enriching for patients more likely to respond,” suggests the team. They add that “Type-2-low asthma, a group with high unmet need and few therapeutic options, can be reliably identified using the same biomarkers.”
The researchers caution, however, that “these models have low sensitivities that would result in about half of Type-2-high patients being called Type-2-low, which may be tolerable for proof-of-concept studies but not for application as a co-diagnostic test.”
They call for confirmation of these classification models in independent study populations and validation from interventional clinical trials.
By Lucy Piper
medwireNews is an independent medical news service provided by Springer Healthcare Limited. © Springer Healthcare Ltd; 2017