Lung cancers are often detected late when treatment outcomes are poor. Detection of lung cancer while the disease is still preinvasive or minimally invasive may allow more effective treatment. Patients who are at high risk for lung cancer may also have early central lung cancers: radiographically occult central lesions that can be detected at an early stage by bronchoscopy or sputum cytology. Standard white-light bronchoscopy, autofluorescence imaging, and the RP-EBUS can be used to diagnose these lesions. The addition of autofluorescence imaging to standard white-light bronchoscopy improves the sensitivity for detecting such lesions, but its low specificity means that it is often unable to distinguish early neoplastic lesions from inflammation, scarring, or other non-neoplastic localized changes [9, 10]. This can lead to unnecessary biopsies with increased cost and risk to the patient [10]. Autofluorescence bronchoscopy also cannot accurately determine the depth of invasion of early endobronchial lesions which is a major determinant of the most appropriate and effective type of therapy [9]. The application of RP-EBUS to such lesions allows the bronchoscopist to clearly distinguish the five normal layers of the bronchial wall of the trachea and cartilaginous bronchi, making it an ideal tool for evaluating such endobronchial lesions [5, 11] (Fig. 26.1).

Herth et al. prospectively evaluated patients undergoing autofluorescence bronchoscopy at two institutions. Those patients with nonspecific changes (n = 32) or findings suspicious for malignant changes (n = 42) on autofluorescence imaging were also evaluated with RP-EBUS. The RP-EBUS was then used to categorize these patients into those with benign ultrasound features (having a preserved layered bronchial wall structure) or malignant ultrasound features (having thickened wall, destroyed layer structure, or peribronchial infiltration). Then endobronchial biopsies were obtained and the histology was correlated with the findings by RP-EBUS. RP-EBUS was more accurate than autofluorescence in both benign (92% vs. 55%) and malignant lesions (97% vs. 69%). This suggests that RP-EBUS significantly improves the diagnostic specificity of autofluorescence bronchoscopy for such central early lung cancers and may be useful in conjunction with autofluorescence bronchoscopy when available [10].

Several studies have shown RP-EBUS to be useful in assessing the depth of invasion of endobronchial tumors, which can be an important factor in determining therapy for such lesions. Using 24 lung lobes resected for known lung cancer, Kurimoto and colleagues compared RP-EBUS determination of depth of tumor invasion with histologic findings and found a 95.8% correlation between their conclusions [5]. Other authors have also reported similar close correlation between in vivo RP-EBUS findings and subsequent histologic examination of surgically resected specimens [12]. Miyazu et al. have also demonstrated the utility of RP-EBUS in helping clinicians choose the most appropriate therapy for central early lung cancers. Their group first evaluated six patients with central early lung cancer detected by autofluorescence bronchoscopy using RP-EBUS and used the information provided about depth of invasion to determine the suitability of each patient for photodynamic therapy (PDT) . The two patients who had no cartilaginous involvement were deemed candidates for PDT and were treated with that modality with good outcome [13]. In another study, their group used RP-EBUS to evaluate 12 patients with 18 biopsy-proven central squamous cell carcinomas who were thought to be good candidates for PDT based on standard bronchoscopy and high-resolution CT scan. Nine of 18 lesions assessed by RP-EBUS were found to be intracartilaginous and therefore candidates for PDT. All nine underwent PDT and had sustained complete remission at 32 months. The other nine lesions were determined to be extracartilaginous by RP-EBUS. Of these six were resected, and histology confirmed the RP-EBUS estimate of the depth of invasion in all six cases. RP-EBUS successfully identified the nine lesions not amenable to PDT so they could receive more appropriate treatment [14].

Herth et al. prospectively assessed the utility of RP-EBUS in therapeutic bronchoscopy. They evaluated 1174 patients who underwent therapeutic bronchoscopy and RP-EBUS over a 3-year period. It was used in conjunction with tumor debridement (mechanical, laser, APC), in airway stent placement, with endobronchial brachytherapy, in endobronchial foreign body removal, and in endobronchial abscess drainage. The authors reported that RP-EBUS changed the therapy or guided the intervention in 43% of the cases in which it was used. Reported examples included longer stent length for undetected peribronchial tumor spread, aiding decisions about when to stop ablative treatments due to proximity to vital structures, and changes to staging in brachytherapy [15].

Radial probe EBUS can be a useful tool in the diagnostic evaluation of peripheral pulmonary lesions. RP-EBUS provides information about the ultrasound characteristics of these lesions, assists with location verification, and has been used in combination with other guided-bronchoscopy modalities to improve diagnostic yield.

Ultrasound characteristics of the peripheral pulmonary lesions visualized with RP-EBUS may provide helpful information in addition to clinical and radiographic features commonly used to determine the probability that a nodule is malignant [4]. Kurimoto et al. analyzed 124 patients with peripheral pulmonary lesions who had both a confirmed histologic diagnosis and a preoperative RP-EBUS to attempt to identify ultrasound characteristics that could predict tumor type. They identified three major patterns. Type I lesions demonstrated a homogeneous pattern and were mostly benign (92%). Type II lesions (hyperechoic dots and linear arcs) and type III lesions (heterogeneous pattern) were mostly malignant: 98 of 99 lesions [16]. Chao’s group subsequently attempted to develop a simpler predictive model. Peripheral lung lesions of an initial group of 20 patients with known histologic diagnosis were used to identify four ultrasound image patterns: a continuous hyperechoic margin, homogeneous or heterogeneous internal echoes, hyperechoic dots in the lesion, and concentric circles along the echo probe. They then enrolled 131 additional patients to assess these patterns prospectively. Five were excluded because the investigators could not agree on the pattern. Of the remaining 126 patients, 93 had a definitive diagnosis and were included in their analysis. After multivariate analysis, only one of the characteristics—the presence of concentric circles—retained a statistically significant predictor of the nature of the lesion. Eighteen of 19 lesions with concentric circles were benign [17]. Kuo et al. assessed 224 patients with peripheral lung lesions who underwent RP-EBUS and were eventually given a definitive diagnosis. RP-EBUS images were reviewed, and three ultrasound characteristics were selected for analysis: continuous or noncontinuous margin between the lesions and the adjacent lung, presence or absence of an air bronchogram within the lesion, and homogenous or heterogeneous echogenicity of the lesion. The presence of a continuous lung margin, absence of a discrete air bronchogram within the lesion, and a heterogeneous echogenicity of the lesion were all found to be predictors of malignancy. A lesion with none of the three features had a negative predictive value of 93.7% for malignancy, and a lesion with two of the three features had a positive predictive value for malignancy of 89.2% [18].

By employing its unique property of location verification, RP-EBUS is frequently used to confirm the lesion sampling site for the diagnosis of peripheral pulmonary lesions. In the past, minimally invasive options for the diagnosis of these lesions included percutaneous needle biopsy or conventional bronchoscopy with or without fluoroscopy. A recent meta-analysis of CT-guided percutaneous core needle biopsy and fine needle aspiration reported pooled sensitivities of 95% and 90%, respectively [19]. However, the superior sensitivity of percutaneous biopsy is not without complications. Pneumothoraces after percutaneous biopsy of peripheral pulmonary lesions are reported between 15–43% and 4–18% which require chest tube drainage [20]. That risk is increased for those patients whose nodules are further from the pleura, who have emphysema, and who have smaller nodules. Diagnostic accuracy also decreases with greater distance from the pleura and is less reliable for diagnosing nonmalignant lesions than malignant ones. With respect to conventional bronchoscopy , a systematic review reported that the pooled sensitivity for central pulmonary lesions and peripheral pulmonary lesions beyond the level of the segmental bronchi was 88% and 78%, respectively, and depended largely on the size of the lesion (pooled sensitivity of 63% for peripheral lesions >2 cm and 34% for lesions ≤2 cm) [21]. When compared to image-guided percutaneous needle aspiration, RP-EBUS is less sensitive but has a much lower complication rate (Fig. 26.2).

In 2002, Herth et al. evaluated 50 consecutive patients with peripheral pulmonary lesions in a crossover study. The patients were randomized to either receive TBLB with RP-EBUS followed by TBLB with fluoroscopy or vice versa. For the RP-EBUS biopsies, the EBUS probe was placed into the bronchi suspected to be the location of the lesion until it was seen with ultrasound. The probe was then removed and the forceps placed into the same bronchus and biopsies taken. The fluoroscopic biopsies were taken in the usual fashion. No significant differences in diagnostic yield were seen between the two methods. Diagnostic accuracy using RP-EBUS TBLB was around 80% [22]. In a randomized controlled trial of 221 patients, Paone and colleagues were able to show improved sensitivity (79% vs. 55%) and diagnostic accuracy (85% vs. 69%) with TBLB using RP-EBUS vs. TBLB without RP-EBUS [23]. Soon thereafter, a guide sheath was introduced to improve the yield of RP-EBUS-guided TBLB for peripheral lung lesions. This technique involves advancing the radial probe through a guide sheath to the lesion and then withdrawing the probe once it has been localized while leaving the guide sheath in place. The biopsy tools are then advanced through the guide sheath to the lesion [24, 25]. In some reports using RP-EBUS with guide sheath to sample peripheral pulmonary nodules has demonstrated impressive diagnostic yield, with Kurimoto and colleagues achieving a yield of 76% in lesions 10 mm or less in size [25]. Many factors have been seen to increase the yield when using RP-EBUS for peripheral pulmonary lesions in different studies. These include lesions >2 cm in size, lesions closer to the hilum, visualization on fluoroscopy, malignant disease (as compared to benign), having the probe within the lesion rather than adjacent to it, and taking at least five biopsy specimens [25–30]. Another factor that may improve yield when diagnosing peripheral pulmonary lesions is the use of transbronchial needle aspiration (TBNA) for parenchymal lesions. Traditionally TBNA has been employed to sample mediastinal and hilar lymph nodes rather than peripheral pulmonary nodules. However, Chao et al. examined this in a randomized trial of 182 patients. They used RP-EBUS without guide sheath or fluoroscopy to locate peripheral lung lesions. The patients were randomized to sampling with conventional techniques (including TBLB and bronchial washings) or conventional techniques with the addition of TBNA. The addition of TBNA to conventional techniques increased the overall diagnostic yield from 60 to 78%. TBLB and bronchial wash demonstrated lower yield when the EBUS probe was located adjacent to the lesion rather than within it, but TBNA did not suffer a decrease in diagnostic yield [31]. Despite its ability to improve the diagnostic yield of bronchoscopy for peripheral lung lesions, TBNA remains underutilized [32]. Unfortunately, only two factors, the lesion seen concentrically around the probe and lesions >2 cm in size, have been consistently seen to improve diagnostic yield with RP-EBUS. This fact emphasizes the clinical variability in the studies reporting these results. Overall, the use of RP-EBUS has been shown to improve the performance characteristics over that of conventional bronchoscopy alone. Two recent meta-analyses report the pooled sensitivity of RP-EBUS to be more than 70% but acknowledge significant heterogeneity between studies [33, 34].

Synergistic combinations of different bronchoscopic modalities with varying properties of maneuverability (e.g., thin and ultrathin bronchoscopy), navigation (e.g., virtual bronchoscopy and electromagnetic navigation), and location verification (e.g., RP-EBUS) may improve diagnostic yield for peripheral pulmonary nodules. Asahina and colleagues combined virtual bronchoscopy with RP-EBUS and guide sheath in 29 patients with small peripheral pulmonary lesions. Their sensitivity was 92% for lesions between 20 and 30 mm in size, but only 44% for those less than 20 mm in size [35]. Combining a technique that improves maneuverability and navigation (such as electromagnetic navigational bronchoscopy) with one that confirms the location (such as RP-EBUS) is another appealing option [36]. Eberhardt et al. conducted a prospective randomized trial to precisely this approach. They randomized 120 patients (118 of which had a definitive diagnosis in their final analysis) to electromagnetic navigational bronchoscopy (ENB) , EBUS, or a combination of both techniques. Diagnostic yield was higher with combined EBUS/ENB (88%) than with EBUS or ENB alone (69% and 59%, respectively) [37]. Ishida’s group showed similar results in a trial of 199 patients with small peripheral lung lesions who were randomized to EBUS with virtual bronchoscopic navigation or bronchoscopy with RP-EBUS but no virtual bronchoscopic navigation. They reported a diagnostic yield of 80% in the combined modality group against 67% in the group with EBUS alone [38]. The evidence suggests that using RP-EBUS in conjunction with other modalities—such as a guide sheath, peripheral TBNA, and navigational bronchoscopy—may significantly improve diagnostic yield for small peripheral lung lesions.

Other clinical applications of RP-EBUS in both malignant and benign disease have been reported. Although CP-EBUS has supplanted RP-EBUS in systematic mediastinal staging for non-small cell lung cancer (NSCLC) , RP-EBUS, due to its property to visualize airway wall infiltration, is still used to accurately assess the distance of the endobronchial tumor from the carina, which is an important element of cancer staging [4]. RP-EBUS can also be used to differentiate external compression of a bronchus by tumor from direct tumor invasion of the airway wall which affects the stage of cancer. Herth et al. examined this prospectively in 105 patients who presented with central airway lesions. CT scan was first performed, followed by bronchoscopy with EBUS. The 105 patients who were analyzed went on to surgical procedures for treatment or staging which also involved sampling the airway so histologic confirmation of the bronchoscopic findings was available. EBUS was far superior to CT scan in predicting tumor invasion of the airway wall, with sensitivity, specificity, and diagnostic accuracy of 89%, 100%, and 94% for EBUS compared with 75%, 28%, and 51% for CT scan [39]. RP-EBUS, with and without navigational bronchoscopy as an adjunct, has also been used to aid in placing fiducial markers in order to guide stereotactic radiosurgery for lung tumors [40].

Applications of RP-EBUS in benign diseases such as lung transplantation and asthma have also been reported, although their general use has not yet been widely adopted. In one study of ten patients who underwent lung transplantation, RP-EBUS was used to measure the thickness of the layers of the autologous and allogenic parts of the central bronchi. In patients with evidence of acute graft rejection on transbronchial biopsies, the relative area of the second submucosal layer of the autologous airways was smaller than in those without graft rejection. Additionally, the relative area of the second layer of the autologous airways was thicker in those patients with evidence of infection on bronchoalveolar lavage [41]. Asthma is a disease process characterized by airway wall remodeling, and measurements of the total bronchial wall thickness by RP-EBUS have been shown to be comparable to that made by high-resolution computerized tomography scans [42]. In the same study including 35 asthmatics and 23 controls, the thickness of the first two layers of the bronchial wall measured with RP-EBUS was significantly larger in asthmatics and negatively correlated with forced expiratory volume in 1 s [42].

In the past, conventional TBNA was one of the minimally invasive options to obtain a tissue diagnosis of intrathoracic lymphadenopathy, but it had a variable diagnostic yield. In one study comparing conventional TBNA with RP-EBUS-guided TBNA, the diagnostic yield for lymphadenopathy in sites other than the subcarinal area was seen to be higher with RP-EBUS (84% for RP-EBUS vs. 58% for conventional TBNA) [43]. The disadvantage of RP-EBUS guidance for TBNA of mediastinal lymph nodes is the inability to perform the needle aspiration under direct, real-time guidance which led to the development of the specialized bronchoscope with an integrated convex probe and working channel for needle aspiration [7].

The indications for CP-EBUS are summarized in Table 26.2.