Classifying disease according to treatable traits offers a new way forward in successful management of COPD an expert has told a respiratory insights forum.

Professor Sanjay Sethi, Chief of Pulmonary, Critical Care and Sleep Medicine at the University at Buffalo in Buffalo, New York, told delegates it was now possible to distinguish four new clear phenotypes in COPD: patients with bronchiectasis; younger patients with severe disease; patients with comorbidities; and patients with increased blood eosinophils.

“These are phenotypes you can identify readily based on easily obtainable data,” Professor Sethi told the respiratory insights forum hosted by Menarini.

“Our treatment choices in COPD are relatively limited, so this enables us to move towards ‘treatable traits’…. It means we need to pay more attention to identify younger more advanced disease patients and older comorbid obese ones – these are things we can do.”

Why phenotype?

Traditional classification of COPD relies on spirometry, but there is mounting recognition that this fails to take into account the heterogeneity of the disease.

In fact, most respiratory physicians are already using some degree of phenotyping in their clinical practice to choose a treatment for their patients, Professor Sethi said.

For instance, the current GOLD classification advocates clinical sub-groups of COPD, graded A to D based on lung function, symptoms and exacerbations. The classic definition of COPD by the American Thoracic Society had also identified two phenotypes, chronic bronchitis with or without obstruction, and emphysema with or without obstruction, both of which may overlap with asthma.

According to professor Sethi it would be more clinically relevant to group patients related to outcomes such as symptoms, exacerbations, response to therapy and rate of disease progression.

Phenotyping COPD into clinically meaningful sub-types could promote personalised medicine, provide new biological insights by identifying the right groups for the right agent, and enrich clinical trials, he said.

For example, viewing COPD patients with concomitant bronchiectasis as a phenotype experiencing more exacerbations and requiring more hospitalisations could be useful in the future to tailor treatments to a sub-group likely to do worse.

“Clinically there is little difference in their management today, but down the track as new treatments are developed this group will become more relevant phenotypically,” Professor Sethi said.

Evidence highlights clear phenotypes

A systematic review¹ into studies conducted in tertiary care settings that used unbiased cluster analysis identified two new distinct phenotypes of COPD in patients with poor health outcomes.

The first consisted of young patients with severe respiratory disease, few cardiovascular co-morbidities, and poor nutritional and health status. This group, said Professor Sethi, were possibly rapid progressors who were more susceptible to COPD, or people whose lung function had never fully developed. Patients with this phenotype required aggressive treatment and would be good candidates for lung transplantation, he said.

The second phenotype consisted of older patients with moderate respiratory disease, obesity, and cardiovascular and metabolic co-morbidities. For this group, the therapeutic implication was that more attention needed to be paid to their co-morbidities, Professor Sethi said.

A cluster analysis of the ECLIPSE cohort² distinguished five sub-groups with different prognoses including patients with: milder disease, fewer deaths and hospitalisations; less systemic inflammation at baseline but with notable progressions in health status and emphysema extent; many comorbidities, evidence of systemic inflammation, and the highest mortality; low FEV₁, severe emphysema, and the highest exacerbation and COPD hospitalisation rate; and intermediate for most variables, representing a mixed group that may include further clusters.

“None of these studies give cut offs to define the phenotype, so clinically they are still difficult to use, but I find them useful in terms of thinking about the younger person with bad disease and then the older comorbid person who I need to pay attention to their comorbidities,” Professor Sethi said.

COPD and bronchiectasis

Phenotyping patients based on high resolution CT scan results was unlikely to be of much clinical benefit, apart from in defining another possible COPD phenotype: patients with comorbid bronchiectasis, Professor Sethi said.

A meta-analysis³ including six observational studies found a mean prevalence of radiological bronchiectasis in patients with COPD of 54.3%. These patients were older, more often males with a smoking history, and had worse lung function.

The clinically useful findings of this study were that these patients had greater daily sputum production, more frequent exacerbations, poorer lung function, higher level of inflammatory biomarkers, more chronic colonisation by potentially pathogenic microorganisms (PPMs), and a higher rate of Pseudomonas aeruginosa isolation compared with COPD patients without bronchiectasis.

This meant that patients with daily sputum should probably receive a CT scan and in future may benefit from inhaled antibiotics, Professor Sethi said.

Patients with eosinophilic inflammation

There was currently great confusion around asthma-COPD overlap syndrome (ACOS), Professor Sethi noted. A plethora of studies showed that patients with both asthma and COPD did worse in terms of burden of disease, more symptoms and more exacerbations, but the longitudinal outcomes were often better.

“If patients with adult asthma and substantial noxious particle exposure present with COPD, to me that’s ACOS, and I like to use ICS upfront,” Professor Sethi said.

“In the future we may define a group who have higher levels of significant TH2 inflammation, and that can have therapeutic implications in use of anti-TH2 biological drugs,” he said.

Are endotypes of more use than phenotypes?

Current thinking is that categorising patients by endotypes – sub-types defined by a distinct physiological or pathophysiological mechanism that allows for detailed prognosis- could be even more useful in clinical practice, Professor Sethi said.

A review in The Lancet⁴ published by a team of international experts defined seven endotypes in COPD:

1. Alpha 1 anti-trypsin deficiency
2. Increased systemic inflammation
3. Eosinophilic/TH2 high
4. COPD with bacterial colonisation
5. Exacerbation endotypes
6. Comorbidities
7. Lung cancer.

While phenotyping and endotyping of COPD was still in its early stages, this new data would eventually progress the GOLD classifications, giving respiratory physicians greater insights into how to manage patients and drive the development of new, targeted treatments, Professor Sethi concluded.

References

1 Pinto et al. Derivation and validation of clinical phenotypes for COPD: a systematic review. Resp Res 2015; 16: 50

2 Rennard S et al. Identification of five chronic obstructive pulmonary disease subgroups with different prognoses in the ECLIPSE cohort using cluster analysis. Ann Am Thorac Soc 2015; 12(3): 303–312.

3 Ni Y et al. Clinical characteristics of patients with chronic obstructive pulmonary disease with comorbid bronchiectasis: a systemic review and meta-analysis. Int J Chron Obstruct Pulmon Dis. 2015; 10: 1465–1475.

4 Woodruff et al. Current concepts in targeting chronic obstructive pulmonary disease pharmacotherapy: making progress towards personalised management. Lancet 2015; 385:1789–1798