The origin, progression and dissemination of solid tumours are strongly regulated by stromal, endothelial or defence-related cells 1. Tumour cells do not merely grow into an inert interstitial tissue, but they actively interact with their microenvironment establishing cell-to-cell contacts, releasing soluble factors and responding to soluble factors released by neighbouring cells. In recent years, evidence has been accumulated to support a dual role for macrophages in the regulation of tumour cell proliferation, invasion, angiogenesis or immune control 2, 3. On one hand, the “resident” tissue macrophages have a cytotoxic function against tumour development as they are a major component of the inflammatory infiltrate frequently seen in primary tumours. On the other hand, a different phenotype of monocyte-derived “tumour-associated macrophages” (TAMs) shows opposite behaviour with a commonly pro-tumoral phenotype. The ability of macrophages to promote or inhibit neoplasia seems to depend on their differentiation and activation states 4, 5. At least two major functional states have been described for macrophages and named as M1 and M2. Classically activated M1 macrophages are induced by interferon-γ alone or in combination with lipopolysaccharide and tumour necrosis factor-α. M1 macrophages produce high levels of inducible nitric oxide synthase and pro-inflammatory cytokines, such as interleukin (IL)-12, and are generally considered as part of the inflammatory response against cancer progression. In contrast, M2 macrophages express arginase I, promote angiogenesis, favour tissue remodelling and invasion, and are able to inhibit the inflammatory response through the secretion of IL-10. Thus M2 macrophages are thought to exert a tumour-promoting activity. In most of the solid tumour types that have been analysed so far, TAMs show an M2 differentiation profile, correlate with microvessel density and are associated with reduced patient survival. Redente et al. 6 have recently published a detailed analysis of the changes in activation state of macrophages during neoplastic progression in a well-known experimental lung cancer mouse model. Previous studies on the association of tumour infiltration by macrophages and the survival of patients with surgically resected nonsmall cell lung cancer (NSCLC) have rendered conflicting results. The prognostic value of macrophages in NSCLC seems to depend on their microanatomical distribution (within tumour islets or in the stroma) 7. Macrophage activation states have not been examined yet in human lung carcinogenesis.
The study by Zeni et al. 8 in the present issue of the European Respiratory Journal studies the expression of IL-10 in human NSCLC focusing on both TAMs and tumour cells. IL-10 has been previously shown in NSCLC tissue homogenates 9 and is elevated in the serum of NSCLC patients 10. Zeni et al. 8 examine whether the expression of IL-10 in NSCLC cells and TAMs can be correlated with clinicopathological features of the patients. In total, 50 NSCLC patients (34 adenocarcinomas and 16 squamous cell carcinomas) were included; approximately half of the patients were classified as stage I and the remaining as stage II–IV. Samples from surgically resected tumours were analysed for IL-10 expression by immunocytochemistry. Although the density of macrophages was not different between early and late stages, the authors observed a significant increase of IL-10 positive TAMs in tumour specimens from patients with stage II–IV when compared with patients with stage I. Interestingly, IL-10 expression in TAMs was associated with shorter overall patient survival, although it did not prove to be an independent prognosis factor. Previous studies have already suggested the prognostic value of IL-10 expressed by TAMs in other malignancies 11.
In contrast to what Zeni et al. 8 found in TAMs, no difference in IL-10 expression was detected in tumour cells among stages. The prognostic significance of IL-10 expression in lung tumour tissues has rendered controversial results in the past. Hatanaka et al. 12 reported that NSCLC patients with high IL-10 expressing tumours showed poorer prognosis than those without IL-10 expression. Previously, Soria et al. 13 had reported that stage-I NSCLC patients with tumours lacking IL-10 expression had worse prognosis than those with IL-10 expression. Major differences in the design of the three studies may explain this inconsistency. Hatanaka et al. 12 evaluated IL-10 expression by reverse transcriptase-PCR in 82 patients with stage I–III NSCLC. Zeni et al. 8, however, applied immunohistochemistry to 50 stage I–IV cases. Finally, Soria et al. 13 used a different antibody for immunocytochemical localisation of IL-10 and focused on a more homogeneous population of 138 stage I patients. Therefore, the strikingly contrasting conclusions are probably due to both technical differences and, very likely, to the specific biology of stage I early lung tumours 13 as compared with more advanced II–IV stages, also studied by Hatanaka et al. 12 and Zeni et al. 8.
Besides lung tumours, overexpression of IL-10 in serum or peritumoral stroma has been reported in several other malignancies: melanoma 14; cutaneous basal and squamous cell carcinomas 15; ovarian carcinoma 16; head and neck squamous cell carcinoma 17; brain tumours 18; and colorectal carcinoma 19. IL-10 is thought to suppress immune and inflammatory responses; it inhibits the tumoricidal capacity of macrophages, tampers with the cytotoxicity and cytokine production of tumour-specific T-cells and blocks the presentation of antigens by antigen-presenting cells 20. Production of IL-10 may provide a mechanism for the tumour to evade the local T-cell-mediated immune response 15. In lung carcinomas, IL-10 production can inhibit tumour cell susceptibility to cytotoxic T-lymphocyte-mediated killing 21. IL-10 transgenic mice injected with Lewis lung carcinoma cells developed larger tumours than control mice, suggesting that the production of IL-10 prevents the development of an effective immune response against the tumour cells 22. In summary, the classical view is that IL-10-mediated downregulation of the host's immune system may allow immunogenic tumours to escape immunosurveillance and promote tumorigenesis. From the clinical point of view, this role of IL-10 in cancer leads to the assumption that high IL-10 production by a tumour may limit the success of lung cancer immunotherapy. However, the situation seems to be more complex, and the real role of IL-10 in cancer is far from being completely elucidated. An increasing number of studies show that IL-10 has pleiotropic biological activities, including immunostimulatory and antineoplastic effects 23. Some in vivo studies have suggested that IL-10 might inhibit tumour growth and metastasis in murine models 24–28. Even IL-10 knockout models have yielded opposite results in different experimental situations: B-cell tumours in IL-10 null mice grow slower than controls 29 but IL-10-deficient mice develop colitis-associated colorectal carcinoma 30. In fact, although it is commonly regarded as an anti-inflammatory, immunosuppressive cytokine, IL-10 also possesses immunostimulating properties 23. These conflicting biological data may explain the contradictory results found on the prognostic relevance of IL-10 expressed by lung cancer cells 8, 12, 13, 31.
It is evident that further efforts should be made to clarify the implication of IL-10 in lung tumour development, first at the basic carcinogenesis level with cell and animal models and eventually in the clinical setting. In this sense, the association between IL-10 gene promoter polymorphisms and cancer susceptibility and progression is also an interesting field of research 32. Shih et al. 33 have reported that certain genotypes associated with higher IL-10 production can increase the susceptibility to NSCLC. This association must be confirmed in independent and larger cohorts of NSCLC patients. To date, no study on the association between IL-10 genotype and prognosis of lung cancer has been reported. There are also many questions left open by the data provided by Zeni et al. 8. It is striking that they have found differences among stages in the expression of IL-10 by macrophages but not by tumoural cells. Does it mean that the potential regulation by IL-10 depends more on the macrophages than on the tumour cells? The field of TAMs is still in its infancy. Unfortunately, most of the studies have been carried out in tumours other than lung cancer. Data generated in other organs need to be validated in the lung, and specific lung-related questions need to be addressed. Two examples are as follows. 1) Do alveolar macrophages change their phenotype upon stimulation by peripheral tumour cells or are they monocyte-derived cells recruited from the circulation? 2) Would the IL-10 levels in macrophages from bronchoalveolar lavage be a useful prognostic tool for NSCLC patients?
Understanding the interplay between inflammatory and tumour cells in lung cancer will permit the development of new cancer therapeutic strategies aimed to modulate the interaction between tumour and stromal cells. If the immunosuppressive effect of interleukin-10 in nonsmall cell lung cancer is confirmed, strategies to inhibit its effect should be evaluated. The elucidation of the mechanisms by which tumour-associated macrophages cooperate with tumour growth will also aid in the design of approaches to boost anti-tumour immunity. To date, there are still fragmentary and even contradictory results on the role of interleukin-10 and tumour-associated macrophages in lung cancer progression. Studies, such as the one by Zeni et al. 8, are needed to progress knowledge about the association between inflammation and lung cancer.
- © ERS Journals Ltd