Interleukin-22 is elevated in lavage from patients with lung cancer and other pulmonary diseases

Interleukin-22 (IL-22) is a cytokine from the interleukin-10 family which acts exclusively on IL22-receptor-1 (IL-22-R1) positive epithelial and endothelial cells [1]. In the lung IL-22 has been shown to be expressed by T cells, natural killer-cells, macrophages, epithelial and potentially also by tumour cells [2]. Its effects can be both immunoregulatory and proinflammatory depending on the stage of disease [3, 4]. IL-22 seems to be protective in the acute phases of lung inflammation or injury such as pneumonia, fungal infection, traumatic lung injury, acute lung injury associated with pancreatitis or the initial phase of allergic airway inflammation [4‐8]. In acute inflammation, IL-22 recruits inflammatory cells to clear the infection, probably through the local upregulation of chemokines in the lung, and to rescue lung epithelial cells from cell death [5, 9]. However, if the pathological condition is not cleared and the inflammation becomes chronic, IL-22 seems to sustain inflammation and contribute to the disease phenotype [3, 10]. Recently, we and others have found evidence for IL-22 as a mediator in the interaction between lung cancer cells and the immune environment [11]. In vitro IL-22 promotes tumour growth and chemotherapy resistance of lung cancer cells. Analysis of a large cohort of patients suffering from lung cancer has revealed that IL-22 is frequently expressed in lung cancer tissue, but the clinical significance of these findings has yet to be addressed [12]. In addition, we previously measured IL-22 serum levels in lung cancer patients and matched healthy controls but did not find any difference in spite of strong tissue expression [12]. These observations prompted us to hypothesise that the systemic circulation may not adequately reflect processes in the lung, and that a closer analysis of the pulmonary compartment may help to better understand the role of IL-22 in lung cancer.

In the present study, we analysed lavage specimens from 195 consecutive patients (37 with lung cancer) undergoing clinically indicated bronchoscopy and correlated IL-22 expression with local and systemic cell counts and with serum markers of inflammation.

The present study demonstrates that IL-22 concentration in pulmonary lavage samples can be quantified using ELISA and that the levels of IL-22 in lavage vary between different disease entities. Patients with bacterial pneumonia, lung cancer or pulmonary manifestations of other tumours appear to have higher levels of IL-22 in lavage samples compared with non-lung disease controls. In patients with lung cancer, IL-22 levels in lavage did not correlate with systemic signs of inflammation. We found that within the lung, IL-22 may originate from different cell populations.

The finding that patients with NSCLC show higher levels of IL-22 in pulmonary lavage specimens is in line with the results of a previous study reporting that lung cancer cells may produce IL-22 [11]: Zhang and colleagues found high expression of IL-22 in primary tumour tissue, serum, and malignant pleural effusion of NSCLC patients, as well as expression of the IL-22 receptor (IL-22-R1) on lung cancer cell lines. We recently studied the expression of IL-22 in tissue microarray samples of a large cohort of lung cancer patients and found IL-22 expression mostly in patients with large cell NSCLC and those with small cell lung cancer 22 [12]. The lung cancer patient cohort in the present study is, however, too small to confirm these histological subgroup results. In the present study, we did not detect IL-22 in BAL samples from some of the lung cancer patients studied. While technical reasons may be put forward to explain these results, levels of IL-22 may vary significantly between tumor patients. Low levels of IL-22 may be of prognostic relevance, as IL-22 is thought to promote a more aggressive lung cancer phenotype [12].

To the best of our knowledge, the present study is the first to investigate IL-22 concentrations in pulmonary lavage from cancer patients and show that IL-22 is elevated in lavage samples from lung cancer patients. It thus strengthens the hypothesis that IL-22 has a role in lung cancer. The elevated IL-22 levels found in patients with pulmonary manifestations of extrathoracic tumours support the hypothesis that IL-22 may be a mediator in cancer development and progression [18].

In addition to the data we have presented, IL-22 has been investigated in other non-malignant lung diseases. A recent study in pulmonary lavage samples from patients with bronchial asthma revealed that IL-22 concentrations are elevated, further supporting that IL-22 is a disease-associated cytokine detectable in lavage and associated with lung inflammation, as seen in our study [19]. Studies on the host response towards bacterial or fungal pneumonia have revealed that IL-22 contributes both to the acute phase, where it supports clearance, and to the chronic phase, where is prolongs inflammation. In chronic infections such as tuberculosis, IL-22 seems to play a disease promoting role [20]. Our study detected higher levels of IL-22 in lavage samples from patients with pneumonia compared to controls, supporting the suggestion that IL-22 plays a role in the pulmonary response to infection. This is in line with previously published data which identified IL-22 producing cells in BAL from patients with pneumonia [21]. Recently, IL-22 has been detected in lavage samples from patients with community-acquired pneumonia and correlated with serum levels of IL-22, corroborating our finding that IL-22 is associated with pulmonary inflammation [21].

In the lung, IL-22 is thought to be mainly produced by lymphoid cells, among others by CD4-positive lymphocytes [5, 22, 23]. In previous studies analysing lavage samples in patients with tuberculosis, T helper cells were identified as the major source of IL-22 production in BAL [22]. However, in our current study, we did not find any association between IL-22 production, and lymphocytes found in lavage from patients with pneumonia and the number of patients with lung cancer was too low to draw firm conclusions from the analysis with CD4 to CD8 T cell ratio. While T cells as source of IL-22 in lung cancer have not been proposed prior to our study, our data are supported by evidence from other tumor types such as colon, gastric or hepatic carcinomas where CD4 T cells are thought to be the main sources of IL-22 [24‐26]. In contrast, in lung cancer, the source of IL-22 remains uncertain. One study has proposed that IL-22 is expressed by the tumor cells themselves [11]; however, an earlier study from our group was not able to confirm these results, supporting the notion that IL-22 is expressed in the environment but not in the tumor cells themselves [12]. This idea is promoted by a recent study analysing IL-22 producing cells in malignant pleural effusions which identified CD4 T cells as major source [27]. Our study may point towards a role for different cell populations rather than an individual one as source of IL-22 in the lung, as we found no clear correlation between IL-22 and the analysed cell populations. However, the current study was not powered to prove a lack of association between cell populations and IL-22 in the lung.

A noteworthy finding of the present study is the dissociation between local IL-22 concentrations and systemic parameters of inflammation such as CRP and leukocyte counts in patients with lung cancer. This suggests that IL-22 in the lavage of lung cancer patients reflects local processes in the lung rather than systemic inflammation. The concept of the airway as a distinct biological compartment with cytokine levels differing from those in the systemic circulation is supported by other studies: Hollander et al. [28] found that the concentrations of IL-8 and of other markers of inflammation were significantly higher in BAL samples compared to serum samples in patients with bronchial asthma and COPD.

Lavage is a clinically useful tool in the diagnostic evaluation of many patients with pulmonary disease and malignancy; however, clinicians must consider the limitations of this technique when interpreting results. The sensitivity of lavage cytology for lung cancer is reported to be 48 %, lower than that of brushings or endobronchial biopsy [29]. Biomarkers in lavage have the potential to improve the sensitivity of this minimally invasive method in lung cancer diagnosis. Techniques to increase the concentration of IL-22 in lavage samples, such as the use of small volume lavage or lavage catheters placed near the tumor, may further increase the diagnostic sensitivity.

IL-22 can be measured in lavage samples, and correlates with the presence of lung disease. Lavage IL-22 concentrations are highest in patients with pneumonia and lung cancer. IL-22 in lavage does not significantly correlate with systemic inflammation. IL-22 concentrations were not significantly associated with a particular cell population found in the lavage, indicating that IL-22 production may arise from different sources. Our results suggest that IL-22 in pulmonary lavage may serve as a marker for lung cancer, and, perhaps, for pulmonary metastases of other tumours. Markers expressed in the pulmonary compartment can be sampled using bronchoscopic lavage, and may mirror local disease states.