Detection of early lung cancer: Autofluorescence bronchoscopy and investigational modalities

INTRODUCTION — 

Lung cancer is a major cause of cancer-related mortality in males and females. Global strategies targeted at lung cancer prevention and screening have been promoted to prevent or detect lung cancer in its early stages.

Modalities including autofluorescence bronchoscopy (AFB) and investigational tools targeted at detecting preinvasive lesions of the airway are discussed in this review. Lung cancer screening and conventional white light bronchoscopy (WLB) are discussed separately. (See "Screening for lung cancer" and "Flexible bronchoscopy in adults: Overview".)

AUTOFLUORESCENCE BRONCHOSCOPY (DETECTING PREINVASIVE SQUAMOUS LESIONS) — 

Technical aspects of AFB and its proposed clinical role in the detection of preinvasive squamous lesions are reviewed in this section.

Rationale — AFB uses the principle that tissue emits fluorescent light when exposed to light of a suitable wavelength to identify preinvasive and invasive lesions of the bronchial tree [1-3]. AFB was originally developed to detect preinvasive lesions (metaplasia, dysplasia, carcinoma in situ [CIS]), with the idea that the identification and treatment of such lesions would improve the outcome from squamous cell carcinoma of the bronchus. However, the falling incidence of squamous cell carcinoma, the rising incidence of adenocarcinoma (which is now the most common lung cancer worldwide and tends not to present as an endobronchial lesion), the low specificity of AFB, and the reluctance of patients to undergo invasive bronchoscopy as part of screening have limited its use mainly to research [4-6].

Principles of fluorescence — Tissues fluoresce when exposed to light of a suitable wavelength. AFB uses a fluorescence-inducing wavelength of light to illuminate endobronchial tissues. The system collects the fluorescent light and subjects it to digital processing or filtering to produce an image for the bronchoscopist. AFB systems are of sufficient sensitivity to use innate tissue autofluorescence to detect abnormalities and do not require the use of an exogenously administered chemical (fluorophore) to enhance the innate tissue fluorescence intensity.

AFB systems are set up to highlight differences in red and green fluorescence from the tissues being studied. Squamous dysplasia, CIS, and microinvasive carcinoma exhibit much weaker green fluorescence and slightly weaker red fluorescence than normal tissues when illuminated by light with a wavelength of 380 to 440 nm (blue spectrum) [7]. The reasons for the differences in autofluorescence from abnormal bronchial epithelium are not well understood, but may be due to increased epithelial thickness, neovascularization, and/or a reduced concentration of fluorophores in abnormal tissue [8,9]. Identification of differences in autofluorescence between normal and abnormal bronchial epithelium increases the sensitivity of detection of bronchial preinvasive lesions.

AFB is used to identify bronchial preinvasive lesions at risk of progression to invasive squamous cell carcinoma, but not the precursor lesions of other lung cancer cell types. Adenocarcinoma originates from alveolar cells, the stem cells of which are not found in the airways and so precursor lesions to adenocarcinoma are not found using AFB.

Natural history of preinvasive squamous lesions — Squamous cell carcinoma is the second most frequent type of lung cancer, representing approximately one-third of pulmonary malignancies. Several lines of evidence suggest that squamous cell carcinoma develops from histopathologically normal but carcinogen-exposed bronchial epithelium through a series of lesions of increasing cellular and morphological atypia, namely hyperplasia, squamous metaplasia, dysplasia (mild, moderate, severe) and CIS to the development of microinvasive carcinoma and subsequently fully invasive carcinoma [10]. Preinvasive lesions may be single or multifocal within any individual and may fluctuate between the various histopathologic stages or skip steps as they evolve [11,12]. Not all lesions are programmed to develop into squamous cell carcinoma and most remain indolent or regress to normal epithelium. The World Health Organization (WHO) classifies preinvasive lesions using the following descriptors [13]:

●Mild dysplasia – Mild cellular atypia that is limited to the lower one-third of the airway epithelium.

●Moderate dysplasia – More severe cellular atypia that involves the lower two-thirds of the airway epithelium.

●Severe dysplasia – High degree of cellular atypia and minimal cell maturation that extend through the entire airway epithelium, but without reaching the epithelial surface.

●CIS – Extreme cytologic abnormalities (including uneven chromatin, variable nuclear size and shape, dyskaryosis, and other nuclear abnormalities) that extend through the airway epithelium to the epithelial surface, but do not infiltrate the basement membrane.

●Invasive squamous cell carcinoma – Classic histologic features of invasive squamous cell carcinoma that infiltrate through the basement membrane and are capable of metastasis. (See "Pathology of lung malignancies", section on 'Squamous cell carcinoma'.)

Among the preinvasive lesions, squamous metaplasia and mild dysplasia (often termed low-grade preinvasive lesions) progressed to CIS or invasive carcinoma in 8 percent at a median follow-up of 4.7 years [14]. CIS may evolve into squamous cell carcinoma, spread from an established carcinoma, or evolve independently in multiple locations within the bronchial tree [15]. Moderate to severe dysplasia and CIS (often termed high-grade preinvasive lesions) carry the highest risk for progression to invasive squamous cell carcinoma. There are no known markers or histopathologic features of high-grade lesions that reliably predict which will progress to invasive carcinoma or the timescale over which this might occur.

Several studies propose rates of progression of high-grade lesions to invasive carcinoma that range from 39 to 69 percent occurring over several months or longer [16-20]. The presence of preinvasive lesions of the bronchus is known to be a risk factor for the development of invasive carcinoma; however, the invasive carcinoma does not necessarily develop within the index lesion or at its location within the bronchial tree [14,15]. Lesions found to exhibit activity on PET-CT have been shown to have high rates of progression to squamous cell carcinoma [21].

Procedural technique

Devices — Several devices are commercially available. They all use different light excitation and detection wavelengths and have different mechanisms to filter and process the image, meaning different colors are used to detect abnormal tissue fluorescence. The user should be familiar with the differences between devices so that images can be interpreted appropriately.

All modern autofluorescence bronchoscopes have both white light and autofluorescence functions within the single bronchoscope enabling easy switching between white light and fluorescent modes at any point during bronchoscopy.

●LIFE devices – The Light-Induced Fluorescence Endoscopy (LIFE) device was the first to be marketed. While older LIFE devices used blue light and filtered red and green wavelengths to generate fluorescence images [22,23], newer systems (eg, Onco-LIFE device or Pinpoint) use a filtered blue light source to generate green autofluorescence and a red light source to obtain reflected light images [24]. The resulting green autofluorescence and the reflected red light are collected and processed by the autofluorescence bronchoscope that then enhances the contrast between the intensity of the green and red light, with the resulting processed image displayed real-time on a video monitor. Abnormal tissue appears brown-red on a green background.

●Karl Storz device – The Karl Storz device illuminates the bronchial tree with two bands of blue light from a conventional xenon light source to generate autofluorescence [25]. An optical filter collects the red and green autofluorescence from the bronchial epithelium, as well as some of the excitation wavelength of light, which facilitates visualization in areas of low autofluorescence. Abnormal tissue is typically seen as brown-purple on a green background (picture 1 and image 1).

●Pentax devices – Pentax no longer makes an autofluorescence bronchoscope, but it has moved to providing i-scan technology on its standard bronchoscopes.

●Richard Wolf device – This system uses a 300W xenon lamp and filters to induce tissue autofluorescence. The system collects the autofluorescence and optimizes the image by recalibrating the color intensity using the patient's own background autofluorescence [26]. Abnormal tissue appears purple on a blue-grey background, and blood or areas of high vessel density appear black.

●Olympus LUCERA system – The Olympus LUCERA system uses a xenon lamp with a rotating filter wheel to filter blue and green light as the excitation beam. Autofluorescence generated by the blue light is filtered and transmitted on the blue channel of the red-green-blue (RGB) video system [27]. Reflected light from the green filter is transmitted on the red and blue channels of the RGB video system. The red, green, and blue channels are combined within the video system to generate a real-time autofluorescence image. Abnormal tissue appears magenta on a green background.

Procedure — WLB is usually performed first since this recreates the sequence described in studies that demonstrated the efficacy of AFB and allows the user to examine the bronchial tree and anatomy using a familiar imaging modality. WLB is used to identify overt carcinoma, visible lesions, or abnormal anatomy and allow clearance of airway secretions, mucus, or blood that can impact the autofluorescence imaging.

The bronchoscope is then switched into AFB mode and the airways inspected in a systematic manner. There is evidence that the order of the procedures (standard white light or AFB) has no effect on the findings, sensitivity, or specificity of the procedure; however, as scope/suction trauma, mucus, and blood can produce mucosal changes that will look abnormal on AFB, some experts recommend performing AFB prior to WLB to ensure that the airways are as clear as possible prior to AFB and that potential confounding mucosal changes are identified.

The airways accessible to the autofluorescence bronchoscope (which are the same as a standard modern high-definition bronchoscope) are systematically examined using both modalities. Easy switching between WLB and AFB modes on most modern devices allows lesions to be imaged with white light and then autofluorescence to allow abnormalities to be characterized as accurately as possible. Care should be taken to avoid unnecessary trauma while maneuvering the bronchoscope, since factors including cough- and suction-induced trauma and bleeding can significantly undermine fluorescent images and increase the false positive rate of AFB.

Abnormalities visible using AFB have color characteristics that depend on the device used and how it is set up (ie, the color of the background normal epithelium and the color assigned to abnormalities) (see 'Devices' above). Autofluorescence bronchoscopes have a standard biopsy channel and so biopsies should be taken from areas that appear abnormal during either WLB or AFB.

The imaging modality that allows the most accurate localization of the lesion is used during biopsy. However, it may be more difficult to accurately place the forceps when using autofluorescence mode in some systems due to a slight time lag in the image display caused by the image processing system. In such circumstances, localization with AFB, then switching to WLB for biopsy is reasonable. The procedural technique performed for bronchoscopy and endobronchial biopsy are described separately. (See "Flexible bronchoscopy in adults: Preparation, procedural technique, and complications" and "Flexible bronchoscopy in adults: Associated diagnostic and therapeutic procedures", section on 'Endobronchial biopsy'.)

Importantly, AFB (and WLB) is limited to the detection of airway epithelial abnormalities and cannot be used to detect changes deep to the endobronchial epithelium as the autofluorescence excitation light does not penetrate beyond the basement membrane. AFB is incapable of predicting the histologic diagnosis, and the AFB appearance of preinvasive lesions does not reliably indicate their histologic grade [24,28].

In one small study, e-cigarette liquid was found to increase autofluorescence signal intensity, which may impact the performance of AFB systems [29].

Proposed indications — There are no universally accepted indications for AFB. Proposed indications by the American College of Chest Physicians suggest AFB as an adjunct to WLB, when available [30]. Data to support the proposed indications are discussed below. (See 'Efficacy' below.)

Sputum atypia and normal radiographic imaging — AFB may be helpful in the evaluation of patients who have high-grade sputum atypia with no lesion visible on radiologic imaging [30]. However, sputum atypia is rarely encountered in clinical practice since sputum cytology is infrequently performed in patients with suspected lung cancer and is not routinely used for lung cancer screening and there are issues with pathologic interpretation. Nonetheless, when an abnormal sputum cytology result occurs in the absence of a culprit lesion on chest imaging, most often chest computed tomography (CT) combined WLB and AFB may be helpful to identify the lesion from which the abnormal cells were exfoliated.

The rationale for using AFB as an adjunct to WLB in this setting is based upon the observation that the identification of a culprit lesion(s) may be difficult with conventional WLB alone because abnormalities are often subtle and can occur over a wide area of the tracheobronchial tree (especially among individuals who have smoked heavily or have already had an invasive lung cancer) [31-35]. In addition, studies of patients at high risk for lung cancer that included those with sputum atypia have reported increased sensitivity of AFB when used as an adjunct to WLB for the detection of preinvasive lesions, although these studies used old fiberoptic bronchoscopes rather than the modern high-definition bronchovideoscopes (see 'Efficacy' below). In a prospective randomized trial of high-risk patients, the highest prevalence of preinvasive lesions was reported in the 52 patients who had abnormal sputum cytology and normal radiographic imaging; in this group, combined WLB and AFB detected preinvasive lesions nearly three times more often than conventional WLB alone (11.1 versus 4 percent), although the difference was not statistically significant [36]. In another prospective trial, 188 patients with abnormal sputum cytology underwent AFB [5]. Of these, 40 percent had abnormalities on AFB (38 percent had preinvasive lesions, and 2 percent had invasive carcinoma).

Surveillance of high grade preinvasive lesions — For patients discovered to have high-grade preinvasive lesions (ie, moderate or severe dysplasia, or CIS), the optimal therapy is unknown and there are no clear guidelines on which option provides the best outcome; thus, choosing therapy is often dependent upon clinician- and patient-preferences after careful discussion of the options, including the risks and benefits of each potential management strategy. Options include surveillance bronchoscopy, local ablative endoscopic therapy (eg, endobronchial cryotherapy or photodynamic therapy), stereotactic radiotherapy [15], or surgical resection. We, and other experts, suggest that high-grade preinvasive lesions can be followed with combined WLB and AFB with prompt intervention (endoscopic ablation or surgery) if there is confirmed progression to invasive carcinoma [37]. This approach is supported by clinical practice guidelines from the American College of Chest Physicians [30].

Optimal timing and duration of follow-up, however, are unknown, although the interval between surveillance bronchoscopies should probably be greater than six months based on the low progression rates reported in prior studies and the invasive nature of multiple bronchoscopies [14,16,38]. Such an approach should be combined with serial CT screening given the high rate of development of invasive carcinomas in sites remote from the preinvasive lesion under surveillance [38].

We recommend that in patients who have preinvasive lesions, the strategy of surveillance with WLB combined with AFB (when available), rather than intervention, is an acceptable approach that may prevent unnecessary treatments and their attendant morbidity and cost without a significant increase in risk to the patient. This approach is based upon the following rationale:

●Not all high-grade lesions progress into invasive malignancy and in fact, many spontaneously regress [11,15,17,18,20,39,40]. Although the outcome and potential for malignant development in any given lesion is unknown, one narrative review estimated that only 30 percent of high-grade preinvasive lesions progress to invasive carcinoma, while another third spontaneously regress to normal epithelium [20]. In a long-term study, 5 of 31 CIS lesions regressed and only 2 of 31 progressed to invasive carcinoma.

●With the uncertainty regarding the outcome of individual lesions, the effectiveness of the treatments used for high-grade preinvasive lesions has been difficult to evaluate. In some studies, a significant proportion of lesions have progressed to invasive carcinoma despite endobronchial therapy [17,41]. The outcomes associated with treating high-grade lesions are unknown and, therefore, treatment of an individual lesion is not guaranteed to effect a cure or an improved outcome for the patient [17,18,20,41].

●There is a potential for new lesions to develop elsewhere in the endobronchial tree or in the lung outside the bronchial tree despite therapy [15,38,41]. This means that any intervention for preinvasive lesions should be directed towards sparing available lung and avoiding intervention to lesions that will not develop into invasive carcinoma.

Local bronchoscopic therapies that have been reported in the treatment of early endobronchial cancer as well as curative surgical resection for early lung cancer are discussed separately. (See "Bronchoscopic cryotechniques in adults" and "Bronchoscopic laser in the management of airway disease in adults" and "Bronchoscopic argon plasma coagulation in the management of airway disease in adults", section on 'Early airway lung cancer not amenable to surgery' and "Endobronchial photodynamic therapy in the management of airway disease in adults", section on 'Radiographic occult lung cancer' and "Endobronchial brachytherapy", section on 'Airway tumors without extrabronchial spread' and "Endobronchial electrocautery", section on 'Malignant tumors' and "Management of stage I and stage II non-small cell lung cancer", section on 'Surgical candidates'.)

Diagnosis of lung cancer — AFB, when added to WLB, has been shown to improve bronchoscopic diagnosis rates for invasive carcinoma [42-44]. However, false-positive rates are higher than WLB alone. A meta-analysis of 10 studies reported that AFB was more sensitive than WLB for the detection of invasive carcinoma within the visible airways, with no difference in specificity or negative predictive value [43].

AFB has also been used to demonstrate regression of a carcinoma after pembrolizumab therapy when WLB showed inflammatory-type changes [45].

Planning therapy for early invasive cancer — AFB may be helpful in the evaluation and therapeutic planning of patients with early invasive endobronchial squamous cell carcinoma who are being considered for curative local ablative treatment or surgical resection [30]. The purpose is to evaluate the extent of the tumor, to look for other synchronous lesions within the airway that might alter the therapy chosen, and to identify more precisely the tumor margins to improve the accuracy of local treatment delivery.

Evidence to support this approach is derived from the studies discussed below, some of which included patients with known invasive primary lung cancer undergoing evaluation for surgery (see 'Efficacy' below). As an example, retrospective analyses of patients undergoing preoperative evaluation for lung cancer reported that combined WLB plus AFB detected occult synchronous cancer in 4 percent of patients and occult synchronous preinvasive high-grade lesions in up to 23 percent of the patients studied [46,47]. Although not well documented, the therapeutic plan was altered in a small proportion. (See "Endobronchial photodynamic therapy in the management of airway disease in adults".)

Screening for lung cancer — Screening for lung cancer using low-dose CT scanning has become established in the early detection of lung cancer [48,49]. CT is known to image the airways poorly, and there has been concern that early-stage carcinomas of the bronchial epithelium might be missed using CT screening. A 2007 study of 169 patients who underwent bimodality screening with CT scans and AFB showed a dysplasia rate of 11.5 percent and CIS rate of 1.2 percent. Patients with abnormal nodules were three times more likely to have preinvasive lesions at AFB. In this group, 25 percent of the incident carcinomas were squamous cell histology [50]. A 2016 study has shown that the addition of AFB to a CT scan in a high-grade lung cancer risk cohort detected too few CT occult cancers (0.15 percent) to justify its incorporation into a lung cancer screening program [4]. A trial using abnormal sputum cytology and cytometry to identify high-risk smokers at the highest risk for lung cancer screened participants using annual CT, and AFB showed no stage shift in the lung cancers detected and no improvement in the efficiency of lung cancer screening [5,51]. A Chinese screening trial found that of 403 participants offered AFB screening, only 10 underwent the procedure, reflecting the poor uptake of AFB as a screening tool in previous studies [5]. Screening for lung cancer is discussed in detail separately. (See "Screening for lung cancer".)

Other indications — AFB has been used in a number of different situations, although none are supported by robust data:

●Inspection of the airways prior to surgical resection of carcinoma to more precisely delineate the edge of the tumor with the intention of ensuring a clear resection margin [52].

●Postoperative surveillance of the resection margin stump when microscopic preinvasive disease is found at the resection margin [52].

●Investigation of the airways for malignancy other than lung squamous cell carcinoma. AFB has been used in the diagnosis and evaluation of lymphoma [53] and mucoepidermoid carcinoma.

●Nonmalignant conditions. AFB has been reported to aid in the identification of tuberculous bronchial artery aneurysm [54], tracheobronchial amyloid [55], and tracheoesophageal fistula [56].

●Post-lung transplant. AFB has been used to assess the vascularity of the anastomosis after transplant and identify patients who are at risk from anastomotic bronchial stenosis [57].

Efficacy — Data to support the proposed indications (see 'Proposed indications' above) are derived from small prospective studies or small randomized trials of patients who were at high risk of lung cancer or were suspected to have or had lung cancer. The studies involved biopsies being taken from areas that appeared abnormal by either white light or autofluorescence modalities. Control biopsies were taken from areas that appeared bronchoscopically normal and were compared with biopsies taken from areas that appeared abnormal. The ability of each bronchoscopic modality to distinguish abnormal epithelium (ie, dysplasia, CIS, microinvasive carcinoma, or invasive carcinoma) from normal epithelium was determined by histologic examination of the biopsy specimens.

●Sensitivity – Several studies and three meta-analyses have confirmed that combined WLB and AFB is more sensitive than WLB alone for the detection of airway preinvasive lesions, particularly high-grade lesions [22-28,30,36,43,58-65] regardless of the device used. One meta-analysis of 21 studies with a total of 3266 patients reported a pooled sensitivity of 85 percent for WLB plus AFB compared with 43 percent for WLB alone for combined high-grade preinvasive lesions and invasive carcinoma [61]. However, sensitivity rates varied between studies, ranging from 43 to 100 percent for WLB plus AFB compared with 0 to 85 percent for WLB alone.

In contrast, both WLB alone and combined WLB plus AFB detect low-grade lesions with a similar sensitivity [60,66]. As an example, one study found a total of 23 low-grade lesions in 391 biopsies from 55 patients. As many low-grade lesions showed no abnormality bronchoscopically as were visible using WLB or AFB [60].

●Specificity – The major pitfall of AFB is that is it has a high false-positive rate, leading to poor specificity and a low positive predictive value for the detection of preinvasive lesions [24,63]. In one meta-analysis of 3266 patients in 21 studies, specificity of WLB plus AFB was lower than that for WLB alone (61 percent versus 80 percent) [61]. In a study that determined the proportion of biopsies from areas of abnormal autofluorescence that contained abnormal histology, the positive predictive value was only 33 percent [28], although a limitation of predictive values is that they are highly influenced by the prevalence of the disorder in the population studied. (See "Evaluating diagnostic tests".)

Examples of false-positive appearances under AFB include altered autofluorescence from areas of infection, inflammation, or trauma that give the appearance of being abnormal on AFB but on biopsy, no abnormal lesion is detected. Suction trauma and secretions can cause abnormal autofluorescence appearances but are easy to distinguish from abnormal lesions using white light imaging. It is possible that some areas of mucosa that appear abnormal on autofluorescence but have normal epithelium on histology, which are thus considered false-positive results, actually represent abnormal epithelium that is so early in its development that it still appears histologically normal. Supporting this theory, one study found that 50 percent of the lesions that exhibited abnormal autofluorescence (but normal histology) had molecular abnormalities that were similar to preinvasive lesions and invasive carcinoma [67]. Further studies are needed to determine whether AFB can be used to detect lesions at the molecular level at risk of progressing to invasive squamous cell carcinoma.

Studies that evaluate AFB have been criticized due to several limitations:

●Over estimation of the sensitivity of combined WLB plus AFB – Estimates of the sensitivity of autofluorescence may be lower than the reported rates because the true number of preinvasive lesions within the visible bronchial tree cannot be accurately determined by the methods used. The study design where control biopsies are those taken from normal-appearing areas may have underestimated the true number of preinvasive lesions, thereby increasing the false-negative rate and sensitivity.

In a study that addressed this possibility, AFB was performed prior to lobectomy [68]. The bronchial tree was dissected from the lobectomy specimen and the number of preinvasive lesions was determined. AFB detected approximately 50 percent of the preinvasive lesions, suggesting that the true sensitivity is lower than that estimated by studies that used control biopsies.

In addition, estimates of the sensitivity of AFB may have been exaggerated in studies that calculated sensitivity from the number of biopsies that revealed abnormal histology, rather than the number of discrete lesions detected, since more than one biopsy may have been taken from a lesion [28,64].

While it has been postulated that the sequence of WLB before AFB may have biased efficacy in favor of autofluorescence, one trial where the sequence of procedures was subjected to randomization reported similar outcomes regardless of whether WLB or AFB was performed first [60].

Intuitively, since the incidence of squamous cell carcinoma has fallen over the last 40 years, it is likely that the incidence and prevalence of preinvasive lesions has also fallen. Modern studies of the airways of smokers are therefore more likely show a lower prevalence of preinvasive lesions than studies performed 20 years ago [4,50].

●Technical heterogeneity – Most clinical studies have used the LIFE autofluorescence bronchoscope [22,23,28]. Since then, the Karl Storz device has been used with similar results [59,64,65]. However, newer WLB devices with improved resolution, in particular high-definition bronchoscopes with more modern and sensitive imaging technology, are now available. Data suggest that AFB is more sensitive than high-definition WLB for the detection of preinvasive lesions and invasive carcinomas [42-44,69].

●Poor generalizability – Participants in these studies included individuals at high risk for lung cancer due to extensive smoking history and/or chronic obstructive lung disease, abnormal sputum cytology, and prior history of known or suspected lung cancer or other smoking-related malignancies. In addition, most studies primarily included a predominantly male population. Thus, generalizability to other populations may be limited.

●Unknown effect on survival – Although it appears that AFB may detect more high-grade preinvasive lesions than WLB, the impact of AFB on clinically impactful outcomes such as survival is unknown because it has not been studied as a primary endpoint. A number of studies have followed high-grade lesions that were subsequently treated. The uncertainty regarding the outcome of identified preinvasive lesions (progression, regression, or indolence) makes the interpretation of treatment studies difficult [20]. There are no features of a given preinvasive lesion that reliably predict the outcome of the lesion [20]. In patients with preinvasive lesions, there are increased rates of development of lesions elsewhere in the bronchial tree [70] and an increased incidence of the development of remote carcinomas [38]. With these issues, the survival effect of AFB is difficult to estimate. The treatment applied in a number of studies did not prevent progression to squamous cell carcinoma, and thus, in the absence of a clearly effective treatment, the impact of AFB on survival in patients is unclear [17,41].

INVESTIGATIONAL MODALITIES (DETECTING EARLY LUNG CANCER) — 

Developments in conventional white light bronchoscopy (WLB; eg, high-definition video, optical coherence tomography, i-scan imaging, narrow band imaging [NBI], high magnification bronchoscopy) combined with technologically advanced autofluorescence imaging or with screening computed tomography (CT) of the chest may have the potential to improve the detection of endobronchial preinvasive lesions and early invasive carcinomas.

Bronchoscopic methods

i-scan — i-scan technology uses a high-definition camera built into a white light bronchoscope that is used to obtain images of very high quality. The image undergoes post-processing and filtering to generate a real-time image. There are three modes:

●In i-scan1, image contrast enhancement is used to accentuate the edges of lesions and vessel patterns within the bronchial mucosa.

●In i-scan2, image color tone enhancement improves the definition of vessel patterns and further highlights the margins of lesions.

●In i-scan3, a color mode enhances the vascular pattern in a manner similar to narrow band imaging. (See 'Narrow band imaging' below.)

Although there are no major studies of this technology, there is the potential for improved identification of preinvasive lesions and invasive carcinomas within the airways [71]. The device and system require careful calibration and settings optimized to the airways to provide clinically usable imaging of the airways and bronchial epithelium.

Narrow band imaging — NBI uses two narrow bands of light (400 to 430 nm and 525 to 550 nm, blue and green) to image the bronchial epithelium during bronchoscopy. The light is taken up by hemoglobin and allows visualization of surface blood vessels and the surface structure of lesions. The technology relies on the high vessel density of preinvasive lesions to facilitate their detection within the bronchial tree. Abnormalities include three or more of capillary loops, dotted vessels, complex vascular networks of tortuous vessels, or abrupt-ending vessels [72]. Preliminary evidence suggests that the vascular pattern on NBI does not reliably predict the histology of the underlying mucosa [73] but may have similar sensitivity than WLB for detection of preinvasive lesions [72,74-76]. However, imaging relies on a detailed analysis of the entire visible bronchial epithelium, and abnormalities/differences in vessel density, especially in patients with inflammatory airways disease (eg, chronic obstructive pulmonary disease [COPD]), can be difficult to identify.

Optical coherence tomography — Optical coherence tomography (OCT) uses a probe inserted through the working channel of the bronchoscope to deliver a radial beam to the bronchial epithelium. A radial transluminal image is obtained of the bronchus, with high-resolution images of cellular and extracellular structures, and the structure and depth of lesions. Preliminary studies suggest that AFB-guided OCT may be a potential tool for the identification of preneoplastic lesions of the airway [77].

INFORMATION FOR PATIENTS — 

UpToDate offers two types of patient education materials, "The Basics" and "Beyond the Basics." The Basics patient education pieces are written in plain language, at the 5^th to 6^th grade reading level, and they answer the four or five key questions a patient might have about a given condition. These articles are best for patients who want a general overview and who prefer short, easy-to-read materials. Beyond the Basics patient education pieces are longer, more sophisticated, and more detailed. These articles are written at the 10^th to 12^th grade reading level and are best for patients who want in-depth information and are comfortable with some medical jargon.

Here are the patient education articles that are relevant to this topic. We encourage you to print or email these topics to your patients. (You can also locate patient education articles on a variety of subjects by searching on "patient education" and the keyword(s) of interest.)

●Basics topic (see "Patient education: Diagnostic bronchoscopy (The Basics)")

SUMMARY AND RECOMMENDATIONS

●Rationale – Autofluorescence bronchoscopy (AFB) was developed to detect precancerous lesions with the hope that the identification and treatment of such lesions would improve the outcome from squamous cell carcinoma of the bronchus. However, squamous cell carcinoma is now less common, and data have shown limited clinical usefulness of AFB at this time. (See 'Autofluorescence bronchoscopy (detecting preinvasive squamous lesions)' above.)

●Natural history squamous cell carcinoma – Squamous cell carcinoma develops through a series of progressive steps from normal epithelium to preinvasive lesions (hyperplasia, squamous metaplasia, dysplasia [mild, moderate, severe] and carcinoma in situ [CIS]) before developing into invasive carcinoma. Among the preinvasive lesions, moderate or severe dysplasia and CIS (ie, high-grade preinvasive lesions) are the most worrisome for progressing to invasive squamous cell carcinoma. (See 'Natural history of preinvasive squamous lesions' above.)

●Procedure – Commercially available AFB devices use different light excitation and detection wavelengths and have different mechanisms to filter and process the image, resulting in different colors to detect abnormal tissue fluorescence (picture 1 and image 1). The airways are systematically examined using both conventional white light bronchoscopy (WLB) and by AFB during the same session, although the sequence varies among operators. Biopsies are taken from areas that appear abnormal under either modality. (See 'Procedural technique' above.)

●Proposed indications – There are no universally accepted indications for AFB. However, AFB may be used as an adjunct to WLB in patients with high-grade sputum atypia and normal chest imaging, patients with high-grade preinvasive lesions without evidence of invasive carcinoma, or patients with early invasive endobronchial lung cancer who require therapeutic planning. Data to support these indications are derived from small prospective studies or small randomized trials that demonstrated that combined WLB and AFB detected preinvasive lesions with greater sensitivity than WLB alone. However, this increased sensitivity comes at the expense of poor specificity due to a high rate of false-positives. The impact of AFB on clinical outcomes, such as mortality, is unknown because it has not been well studied. (See 'Proposed indications' above.)

●Other investigational modalities – Developments in WLB (eg, high-definition video, optical coherence tomography [OCT], i-scan imaging, narrow band imaging [NBI], high magnification bronchoscopy) may have the potential to improve the detection of endobronchial preinvasive lesions and early invasive carcinomas. However, these modalities need further study. (See 'Investigational modalities (Detecting early lung cancer)' above.)