Lung cancer remains the leading cause of death due to cancer in both men and women in several countries. Most patients with lung cancer present with non-specific symptoms such as cough, dyspnoea or haemoptysis and have late-stage disease at the time of diagnosis. Treatment strategies are thus often palliative rather than curative. Early detection trials using routine chest roentgenograms and sputum cytology have shown that cancers can be detected at earlier stages but the overall mortality from lung cancer was not improved. With new technology developments, such as low-dose computed tomography and fluorescence bronchoscopy, comes new hope that early detection may in fact change this pattern.

Squamous cell carcinoma, although no longer the most common type of lung cancer, remains a major problem, accounting for 17–29% of lung cancers.¹ Studies on sputum cytology showed that 11% of individuals with moderate dysplasia and 19–46% with severe dysplasia progressed to squamous cell cancers. Once detected, localisation of these early lesions was not easy using conventional white-light bronchoscopy. Thus, fluorescence bronchoscopy systems were developed to improve the detection of preinvasive and invasive lesions of the central airways. Several studies showed the improved detection of both preinvasive and invasive cancers using fluorescence bronchoscopy compared with white-light bronchoscopy alone.^2–5 As often happens, however, many new questions have surfaced as more patients are found to have these preinvasive lesions. For example, what should be done with a patient who has an abnormal area showing carcinoma in situ on a biopsy specimen? Several studies have suggested that the carcinoma in situ, if left alone, will progress to invasive cancer and should be treated.^6,7 What about a patient with severe dysplasia? Here the data suggest that the rate of progression to invasive cancer is much less but sufficiently high for these patients to be followed-up.^8,9 But do we really know the natural history of these lesions?

In this issue of Thorax, George et al¹⁰(see p 43) have presented their work on 22 patients with either previous lung cancer, asbestos exposure or known chronic obstructive pulmonary disease. Fifty one lesions were identified using fluorescence bronchoscopy. High-grade lesions included seven classified as severe dysplasia, 28 as carcinoma in situ and 16 as low-grade lesions of mild or moderate dysplasia. These patients were followed up for a median of 23 months (range 12–85 months). Indolence or regression was observed in 83% of the high-grade lesions and no progression was seen in any of the low-grade lesions. However, five patients with high-grade lesions subsequently developed cancers that were detected by computed tomography during their follow-up.

What can be learned from the work by George et al?¹⁰ Their data do not change the current practice that patients with carcinoma in situ should receive treatment. Surgical resection remains the standard treatment, but local treatment such as photodynamic therapy or argon plasma coagulation could be an option if carried out in a clinical trial.¹¹ For patients who are not surgical candidates either because of poor pulmonary function or other comorbidities, local treatment has been shown to be a reasonable option.^12–14 What about those patients with severe dysplasia? The experience presented by George et al¹⁰ shows that these patients are at high risk for developing an invasive cancer. In this population, surveillance is important with both fluorescence bronchoscopy and with computed tomography of the chest.

As we look to the future many gaps remain in our understanding of the natural history of squamous cell carcinoma, and lung cancer in general. Studies such as the one by George et al¹⁰ have again shown that there are subsets of patients with high-grade lesions that progress to invasive cancers and others that do not. This shows the need for reliable markers in tumour biology. George et al¹⁰ point out that surveillance programmes need to include tissue archives of blood, sputum and bronchial specimens, thus enabling studies on histologically identical lesions with different clinical outcomes. This work will require the cooperation and collaboration of many centres involved in surveillance programmes.