How Might Endotyping Guide COPD Treatment? Current Understanding, Knowledge Gaps, and Future Research Needs

Abstract

Purpose of Review:

This review discusses emerging therapies directed at chronic obstructive pulmonary disease (COPD) endotypes, or discussing pathobiological processes that manifest as the disease.

Recent Findings:

Specific endotypes have been targeted in COPD. These include eosinophilic inflammation, overproduction of interleukin-17, chronic bronchitis and altered nature of mucous, and chronic infection. Therapies exactly directed at the cause of these endotypes or their resultant clinical findings have been assessed. While some intermediate outcomes have seemed promising, there have been no findings that shift the paradigm of COPD therapy.

Summary:

Basic and clinical scientists continue to define endotypes that may be directly addressed with therapeutics. As of the time of this up-to-date review, there is yet to be an endotype-directed therapy to demonstrate great clinical effect.

Introduction

Chronic obstructive pulmonary disease (COPD) is a progressive, preventable, treatable, yet incurable disease marked by daily symptoms, rapid lung function decline and acute exacerbations (AECOPD). COPD occurs commonly in the United States and Worldwide.(1) Over the last several years, growing appreciation for the heterogeneous nature of COPD has led to an increased understanding of the specific immunologic, infectious, and comorbid drivers of COPD symptoms and exacerbations.(2–5) These new investigations have given rise to the concept of COPD ‘endotypes’ which are defined as groups of shared characteristics among patients from a similar biology.(4) As the progress towards personalized medicine continues, there is a growing interest in utilizing identified pathogenic factors in the development and progression of COPD for disease treatment.(6) This review will endeavor to discuss these pathobiological pathways, animal models, early phase, and phase 3 clinical trials directed at endotypes of potential clinical interest due to their present potential to be targets of ‘precision’ therapies (i.e., this will not be an exhaustive review of COPD endotypes). We will focus on endotypes individually, but in any given person there is likely multiple endotypes at play and discrete biological mechanisms discussed may interface with other processes. Also, as alpha-1 antitrypsin deficiency is a well-known and documented endotype deserving of its own review. In this review of the most current literature, we will discuss emerging endotypes and those that more recently have become therapeutic targets.

Eosinophilic Disease

Circulating peripheral eosinophils are thought to reflect eosinophils in the lung compartment and have become important biomarkers for the pharmaceutical approach to COPD.(1, 7) While neutrophilic inflammation has traditionally been considered a major culprit for excess pulmonary inflammation in COPD, more recent studies have addressed the potential of the other major granulocyte - eosinophils. Peripheral eosinophilia has been found to have a moderate-to-strong correlation with airways eosinophilia(8), making therapeutic targeting of eosinophils in COPD putatively feasible. The GALATHEA and TERRANOVA(9) studies investigated the cytolytic agent benralizumab (which targets IL-5 alpha-receptors and binds to natural killer cells which inturn release apoptotic proteins leading to eosinophils desruction(10)) in participants with COPD.

Both trials enrolled persons with peripheral eosinophil counts >220 per cubic mm. In the GALATHEA trial, the exacerbation rates for the treatment group were 1.19 per year (95% CI 1.04-1.36) and 1.03 per year (95% CI 0.90-1.19), for doses of 30 mg and 100 mg respectively. In this trial, the placebo group had an annual exacerbation rate of 1.24 per year (95% CI 1.08-1.42). The rate ratio for annualized exacerbation compared to placebo was 0.96 (P=0.65) and 0.83 (P=0.05), for 30 mg and 100 mg respectively. The results were similar in the TERRANOVA trial. For 10 mg, 30 mg, and 100 mg of benralizumab, respectively, the rate rations for annualized exacerbations when compared to placebo were 0.85 (P=0.06), 1.04 (P=0.66), and 0.93 (P=0.40). The authors concluded that despite depletion of peripheral and sputum eosinophil counts, benralizumab did not reduce annualized AECOPD rates.

Mepolizumab, an interleukin-5 (IL-5) receptor antagonist (which binds to IL-5 neutralizig the effect of this cytokine(10)) was also studied in participants with COPD and peripheral eosinophil counts >300 per cubic millimeter.(11) In intention-to-treat analysis assessing the efficacy of anti-IL-5 therapy to prevent AECOPDwas performed in METREX. In 462 participants, the annualized rate of exacerbations was 1.40 per year in a group receiving 100 mg mepolizumab and 1.71 in the group receiving placebo. The rate ratio was 0.82 (95% CI 0.68-0.98, P=0.04) between the treatment arm and placebo. METREO tested treatment arms of both 100 mg and 300 mg. In the 100 mg arm, the annualized rate of exacerbations was 1.19 with a rate ratio vs. placebo of 0.80 (95% CI 0.65-0.98, P=0.07). In the 300 mg arm the annualized rate of exacerbations was 1.27 and the rate ratio vs. placebo was 0.86 (95% CI 0.70 to 1.15, P=0.14). The authors of this pair of studies concluded that 100 mg of mepolizumab but not 300 mg mepolizumab was associated with a lower annual rate of all cause exacerbations. The authors of this study did not postulate why the lower dose of mepolizumab was more effective or why their results differed from the benralizumab studies. The United States Federal Drug Administration did not approve mepolizumab for the treatment of COPD and cited that the primary endpoint analysis was null in all but one arm of one study, and, more interestingly, that eosinophilic COPD remains too loosely defined.(12)

Equipoise surrounding the role of peripheral eosinophils as either a biomarker to predict exacerbations or a biological process to target exacerbations remains.(13) Further, in COPD patients eosinophil levels tend to fluctuate in individual with COPD suggesting that targeting a single eosinophil level may not be the most efficient means to detect who would benefit from eosinophil directed therapy.(14) The relationship between IL-5, eosinophils, and COPD is complicated further by potential mechanisms for the development of eosinophilia in lung tissue other than the IL-5 pathway.(15) While some modest effects of addressing the IL-5-driven pathway have been noted in clinical trials, future studies should focus on better defining the reflection of peripheral eosinophils to those in the lung compartment and means to identify which COPD patients have maladaptive IL-5 activity rather than high levels of pulmonary inflammation, of which other mediators (e.g., IL-3, GM-CSF, eotaxins, and leukotrienes(15)) can lead to eosinophilia outside of the IL-5 pathway.

Interleuken-17 (IL-17)

IL-17 is released by multiple immune cells including natural killer cells, T-cells, B-cells, and lymphoid tissue inducer cells and receptors for IL-17 are ubiquitous.(16) IL-17 is induced by IL-6 and IL-23 and is important in the host defense against infectious pathogens and potentially in the development of autoimmunity.(16, 17)IL-17 may also play a role in the propagation of airway inflammation by amplifying immune cell activation in response to the generation of airway elastin peptides from injured airways.(18) Thought to contribute to the parenchymal destruction of emphysema, mouse studies have shown that knocking out the IL-17 gene attenuates the development of parenchymal lung destruction in mice exposed to tobacco smoke and may also lessen bone mineral loss associated with smoking-related lung disease.(19, 20) A single phase 2 trial of Anti-IL-17 therapy was reported in 187 participants.(21) The main endpoint in this study was lung function decline at 24 weeks. The difference in lung function decline between an anti-IL-17 agent (CNTO 6785) and placebo at 24 weeks was −0.49 %predicted FEV1 (P=0.599). Further, while the medication itself was reported to be well tolerated, there was an increase in infection in the treatment cohort. Recent assessment of infections in IL-17 deficient humans suggests that IL-17 does play a notable role in epithelial immune defense against certain pathogens and that persons with IL-17 deficiency have the propensity to develop infectious bronchitis as well as lobar pneumonia.(22) At the time of this writing there are no ongoing clinical trials in the role of IL-17 for the treatment of COPD, but study of this cytokine in human subjects persists.(23) Because of the putative role of IL-17 in parenchymal lung destruction, anti-IL-17 agents may remain attractive. However, those performing future human studies will need to be cognizant of the likely infectious risks carried by inhibiting this cytokine.

Infection and Infection Control

Understanding the interplay between the microbiome of the lung and COPD outcomes may identify a treatable phenotype in COPD. The persistent presence of bacteria in the airways are able to neutralize the inflammatory response to their presence and cause direct airways damage through toxin (e.g., exotoxins, lipopolysaccharide, alkaline protease, hemolysin) release.(24) The normal human lung contains a balanced mix of populations of symbiotic and pathological bacteria which becomes unbalanced in COPD.(25) As discussed in the introduction, dysbiosis in COPD interfaces with other potential COPD endotypes. IL-17 is a major factor in airway bacterial control(25) highlighting the issues with leveraging IL-17 reduction as a therapy. Still, addressing dysbiosis directly does remain and attractive area for the development of novel COPD treatments. A recent review of translational techniques for the utilization of the microbiome in COPD clinical research stresses the importance and difficulty of proper sample collection and storage as well as accurate genomic analysis.(26) Both the populations of bacteria and bacterial activity in the airway will need to be known to effectively direct therapy. An interesting study utilizing elevated markers of inflammation in individuals 14 days after AECOPD examined whether retreatment with antibiotics confers clinical benefit.(27) In a multicenter, randomized, placebo-controlled study of 144 individuals, participants with ongoing symptoms or a c-reactive protein ≥8 mg/L were given a seven day course of ciprofloxacin or placebo. No additional benefit was noted when analyzing repeat exacerbations over the ensuing 90-day time horizon. Employing antibiotics at ‘high-risk’ times to specifically treat bacterial infection (i.e., not azithromycin for exacerbation prevention(1)) would be a low cost, low risk intervention and it is likely that defining patients of benefit and the appropriate stage of their COPD disease remains the most pressing endeavors.

Assessing components of the immune system that play a primary role in the control of infection may aidin a patient-centered approach to therapy. Plasma cathelicidin is a peptide in the innate immune system that has been associated with poor COPD outcomes.(28–30) However, findings in humanand murine models are yet to determine the utility of this peptide as a marker for therapy. Mouse models suggest that elevated cathelicidin levels are associated with increased responsiveness to budesonide in mice exposed to cigarette smoke(31), whereas other mouse models suggest that inhaled corticosteroids suppress cathelicidin activity.(32) Taken together, in murine models, it appears that high levels of cathelicidin are required for effectiveness of inhaled corticosteroids in mice and this effect may not be due to a synergistic action, but rather higher levels and more antimicrobial activity are needed to overcome the negative effect of inhaled corticosteroids on cathelicidin activity. With a focus on the utilization of inhaled corticosteroids in COPD therapy regimens becoming more prevalent, assessments of the relationships between cathelicidin levels, inhaled corticosteroid use, and COPD outcomes should be investigated.

Mucopathy

The hypersecretion of mucous potentially manifesting as symptoms of chronic bronchitis(1) is a potential target for future therapies. The ratio of the mucin proteins MUC5B and MUC5AC is related to clinical outcomes in large cohorts of COPD participants.(33) At the time of this writing, there are no therapies to specifically change the mucous ratio of a COPD patient to favor more MUC5AC, which is more easily cleared. However, there are studies in methods to hydrate mucous. A small study of 22 participants with a diagnosis of chronic bronchitis and 33 healthy volunteers were conducted to assess the safety and potential clinical benefit of hypertonic saline.(34) This study used both 7% and 12% hypertonic saline with a two-to-four week washout period between two week treatment doses. No difference in lung function after two weeks of either hypertonic saline concentration was observed. Further, any marginal gain in airway clearance potential was lost after the two-week washout.

N-acetylcysteine (NAC) has also been employed to target breaking of the disulfide bonds in mucous, and clinical utility data is generally limited to systemic reviews of small studies.(35) Taken as a whole, studies of NAC for mucous clearance in persons with COPD, both hospitalized and as outpatients, are highly variable . The authors(35) of the most recent review conclude that there is not enough available evidence to make a conclusion on the clinical utility of NAC at this time. Another small study evaluated inhaled liquid nitrogen in 35 participants and suggested improvements in patient-reported outcomes, but clinical endpoints were not reported.(36)

Both hypertonic saline and NAC have predominantly been studied in small trials without the necessary enrollment to detect small effects or to allow for subset analysis. Cough and phlegm production are major clinical symptoms that significant impair quality of life. Also, there is concern for bronchitis symptoms in smokers, both with and without COPD.(37) Due to the general tolerability of these medications, they may have therapeutic potential. However, variability in clinical symptom definitions, participant selection, and study endpoints are significant hurdles that will need to be overcome to evaluate these agents. The GOLD criteria define chronic bronchitis as a productive cough that lasts at least three months with recurring bouts occurring for at least two consecutive years.(1) Quantifying cough and phlegm production can be difficult and are not easily defined study endpoints. Perhaps, clinical factors concomitant with chronic bronchitis (exacerbation history, rapid lung function decline, increased daily symptoms) might be used in the recruitment process and endpoints should reflect these concomitant factors. Ultimately, addressing altered mucous production, viscosity, and clearance remains attractive but these studies will require large cohorts and more elegant means of quantifying effects and identifying participants likely to benefit from the intervention.

Conclusion

This review covers currently identified COPD endotypes that have the potential to be treated directly. Eosinophilia in COPD has been addressed in four large, randomized, clinical trials whereas the other endotypes have only generated smaller trials and mouse model studies. Despite these efforts, there seems to be no imminently emerging therapy to address a specific COPD endotype. Also, these are not an exhaustive listing of COPD endotypes. While progress is being made toward understanding the pathogenesis of COPD, the clinical importance of many of these endotypes are not currently well defined. Further, as the concept of endotypes in COPD is relatively new, it is likely that continued basic and clinical investigations will lead to the identification of further COPD endotypes and determination of which endotypes might benefit from specific treatments. It is apparent from clinical and animal studies of inflammatory pathways and phenotypes that COPD is indeed very heterogeneous and the clinical course is affected by interplay between multiple pathobiologic processes. Another potential area for utilizing the growing knowledge of COPD pathogenesis may be to assess treatable traits in at-risk smokers and those with early airflow limitation prior to the development of COPD symptoms. Further studies are needed and COPD endotypes will need to be refined or new endotypes discovered before endotype-guided treatment is a common practice in COPD management.

Key Points.

• Multiple therapeutically attractive endotypes have been described In COPD without trial data showing clinical efficacy of targeting these endotypes

• Pathologic endotypes in COPD likely interface with one another to a degree that targeting a single pathway may not provide clinical benefit

• Cell-signaling pathways and cytokines upstream of the already assessed interleukins may be a natural next step in investigations (i.e., effectors of lymphoid progenitor cells)

• Future trials may benefit from more strict selection of participants based on inflammatory phenotype and well-defined assessment of intermediate outcomes