As understanding of cellular and molecular mechanisms underlying disease pathogenesis advances, the opportunities increase to identify specific compounds or molecules which are altered by the disease process or appear de novo. These markers of the pathological process have the potential advantage of indices which are indicative of the existing state or change and can be available non-invasively.1

In this issue of Thorax there is a report of the use of Clara cell secretory protein-16 (CC-16, CC-10 or uteroglobulin) as a biomarker for epithelial cell dysfunction (see page 1058).2 CC-16 is a member of the secretoglobin family of secreted disulfide-bridged dimeric proteins.3 It is secreted by non-ciliated Clara cells which reside in respiratory bronchi and by non-ciliated columnar cells of the large and small airways.4 5 CC-16 also occurs in the epithelial cells of the nose and the urogenital tract of men and women.5 There is evidence, however, that serum levels of CC-16 are largely the result of secretion by cells of the respiratory tract rather than the cells of the urogenital tract.6 Serum levels of CC-16 rise following acute exposure to smoke, chlorine and lipopolysaccharide; in patients with asthma, obliterative bronchiolitis and smokers the serum CC-16 levels are low.7 There is an extensive literature on CC-16 levels in serum and bronchoalveolar lavage fluid in normal individuals, experimental animals and individuals exposed to atmospheric pollutants, as well as asthma.7 The exact function of CC-16 is not known, but it may play a role in reducing inflammation in airways.8

The processes which control serum levels of CC-16 are: (1) the rate of synthesis of CC-16 by Clara cells and secretion into the alveolar fluid; (2) the rate of diffusion from alveolar fluid into the capillary blood, which is influenced by leakiness of the pulmonary epithelial barrier; and (3) renal clearance of CC-16. In normal individuals there is variation as a function of gender, age, body mass index, circadian rhythm, ethnicity, temperature, humidity, pulmonary infection and exposure to allergens.7

The ECLIPSE study, a 3-year longitudinal multicentre study of patients with chronic obstructive pulmonary disease (COPD), provided serum for evaluation of the usefulness of CC-16 as a biomarker to identify characterising clinical features of the disease.9 In this trial of 1888 individuals with COPD, 296 smoking controls with no airflow obstruction and 201 non-smoking controls, there were significant differences between the mean CC-16 levels in current and former smokers with no airflow obstruction. There were also significant differences in mean CC-16 levels between current and former smokers with no airflow obstruction and non-smoking controls. The serum CC-16 levels were significantly reduced in 1888 current and former smokers with COPD compared with 296 current and former smokers without airflow obstruction.

A strength of this study is the documentation of serum CC-16 levels in this well-characterised cohort of a large number of patients with COPD, with detailed smoking histories, pulmonary function testing and CT scans of the chest.

A disease biomarker should have: (1) high sensitivity, (2) high specificity, (3) biological relevance to the pathogenesis and (4) sufficient information so that changes in the concentration of the biomarkers can be clinically relevant.10

There are significant limitations to CC-16 as a marker of the disease components of COPD. In the cohorts studied, there was no correlation between the presence or severity of emphysema and the serum CC-16 level. Also, there was no correlation between the serum CC-16 level and the symptoms of chronic bronchitis. The use of inhaled corticosteroids or long-acting β agonists in the ECLIPSE cohort was not reflected in significant differences in serum CC-16 levels. We must be aware that, at present, it is not known if the serum CC-16 level is specific for COPD alone or only to exposure to tobacco smoke or ozone. Testing is required of CC-16 serum levels in other diseases of the lung. Also, studies are required in COPD to determine whether CC-16 can indicate disease progression or regression.

This report of CC-16 from the ECLIPSE study does provide baseline data on several patient cohorts in this large-scale study. Additional data are needed to establish whether CC-16 can reflect short-term or long-term progression or regression of COPD parameters of pulmonary function, clinical state or radiological indices of bronchial structure or, possibly, responses to treatment.

We need to establish where CC-16 might be considered in the spectrum of possible biomarkers of COPD. COPD is a systemic disease beyond the lung with inflammation as a significant contributor,11 and markers of the inflammatory state such as C-reactive protein (CRP)12 13 and interleukin 8 (IL-8)14 have been found to be raised in COPD; significant increases in tumour necrosis factor α (TNFα) in COPD have also been reported.15 However, these reflect augmentation of the systemic inflammatory state in COPD which can be influenced by co-morbid conditions in the cardiovascular system or metabolic co-morbidities. Thus, biomarkers of the pathological state of patients with COPD may be indicators of extrapulmonary processes as well as pulmonary pathology per se. CC-16 has the advantage that it is an indicator anchored to the Clara cells of the bronchial epithelium, which gives it potential relevance to the effects on the diseased lung. We need to separate biomarkers of COPD which reflect abnormalities in specific anatomical structures or biological functions of the lung per se from those such as CRP, IL-8 and TNFα which may reflect systemic pathology such as a heightened reactive state of inflammatory cells.

For decades the forced expiratory volume in 1 s (FEV₁) has been used as the indicator of severity of COPD, predictor of longevity and index of functional response to potential treatments.16 However, as is well recognised, the variability of this measure in any single individual under study requires prolonged periods of observation and multiple measurements to establish significant results and therefore has limited usefulness as a timely indicator.

In several studies the plasma and urine levels of desmosine and isodesmosine have shown elevations as markers of above-normal elastin degradation in COPD.17 18 Recent advances in the techniques of measurement of desmosine and isodesmosine have increased the specificity and sensitivity of quantification so that measurements can be made in sputum as well as plasma and urine.19^–21 Sputum measurements reflect changes in the lung matrix elastin rather than in non-pulmonary sources. Additional studies have shown a reduction in desmosine and isodesmosine levels in experimental animals exposed to smoke22 and in patients with COPD treated with tiotropium.23

Any biomarker should have a relationship with the pathological process ongoing in the lung or bear some relationship to the physiological functions of the lung (such as air flow, gas exchange or pulmonary circulation) or to disease progression, regression and the impact of these changes on the patient’s clinical state and quality of life.

Useful biomarkers in COPD which should not be overlooked are the cellular and cytokine components of sputum. In a study of 56 patients with chronic bronchitis studied over 4 years using physiological measures and CT densitometry, the results support a causative role for neutrophil inflammation and a predictive role in clinical practice.24 A faster decline in lung function has also been correlated with sputum IL-6 levels, neutrophil count and plasma fibrinogen.25 26

When considering possible biomarkers for COPD, the presence of alpha-1 antitrypsin deficiency (ATTD) should be included since this neutrophil protease inhibitor deficiency and COPD are so closely linked. This association extends to heterozygote intermediate deficiencies as well as homozygous severe deficiencies and should continue to be ruled out routinely.27 The development of COPD in the setting of severe ATTD is highly variable. It seems likely that modifier genes may interact with environmental factors to determine an individual’s manifestations of lung disease. A study which focused on identifying risk factors for severe COPD found that sex, smoking, pneumonia and chronic bronchitis all had risk ratios of >2.28 Such predictors could identify significant pathways for genetic modifiers of COPD in ATTD and possible non-ATTD COPD. More recent work has shown that IL-10 polymorphisms are associated with airflow obstruction in severe ATTD, attesting to the validity of searching for genetic determinants as distinguishing markers in COPD.29

Given the clinical complexity of COPD—which is really a syndrome with elements of bronchitis, airway hyperreactivity, pulmonary emphysema and an inflammatory state in variable proportions—it seems likely that multiple biomarkers will be required to characterise pathogenetic factors and their course over time. Biomarkers may be selected or designed to address specific questions of pathogenesis or treatment. Within the spectrum of biomarkers, CC-16 has the potential to indicate abnormal function of the Clara cell or Clara cell abnormalities in airway structure and function. However, its usefulness will be limited unless the measured levels of CC-16 can be shown to reflect the presence and severity of airway dysfunction, normalcy or injury to Clara cells and changes in these states with reasonable speed. To date, this has not been done. Biomarkers of the pathological state of patients with COPD may be indicators of extrapulmonary processes as well as pulmonary pathology per se. CC-16 has the advantage of being an indicator anchored to the Clara cells of the bronchial epithelium, which gives it a potential relevance to effects on the diseased lung.