Lung cancer remains the leading cause of cancer-related mortality worldwide [1]. The increased use of imaging techniques, such as chest radiography and computed tomography (CT), together with the progressive implementation of lung cancer screening programs, has led to a significant increase in the detection of peripheral pulmonary lesions (PPLs) [2].

The diagnosis of PPLs is a considerable challenge in routine clinical practice for interventional pulmonologists, primarily because of their small size and difficult-to-access anatomical location. The lung periphery, composed of small bronchioles and alveoli, has a complex architecture that requires precise navigation to reach the target lesion. Moreover, pathological conditions such as chronic obstructive pulmonary disease (COPD) may further compromise the structural integrity of peripheral lung tissue [3].

Accurate and timely diagnosis is crucial, because early intervention in malignant disease can significantly improve patient outcomes. The development of minimally invasive procedures has been a milestone, offering high diagnostic accuracy with reduced complication rates. Notably, innovations in imaging modalities and electromagnetic navigation have further enhanced diagnostic precision. Cone-beam computed tomography (CBCT) provides high-resolution, real-time, 3-dimensional imaging, improving visualization of pulmonary anatomy during interventional procedures [4]. Electromagnetic navigation bronchoscopy (ENB) uses electromagnetic fields to guide biopsy tools with high accuracy. The integration of these techniques into a hybrid approach has shown potential to improve diagnostic yield while minimizing procedural risks, representing a significant advance in the management of PPLs [5].

This study aimed to evaluate the diagnostic yield of ENB with CBCT assistance in the assessment of challenging peripheral pulmonary lesions (C-PPLs). The combination of ENB and CBCT has emerged as a promising strategy to improve lesion localization and biopsy precision, potentially overcoming the limitations of conventional methods. Furthermore, this study sought to provide clinically useful data to guide the selection of diagnostic approaches in highly complex cases.

A total of 63 cases of C-PPLs were included in the study. The demographic, clinical, and radiological characteristics of the patients are summarized in Table 1. Radiological assessment showed that the mean nodule diameter was 12.3mm (SD, 3.2mm). A substantial proportion of lesions (66.6%) were located in the upper lobes, and 85.7% were located between the seventh and ninth bronchial generations. The mean distance from the pleura was 13.5mm (SD, 8.1mm). A bronchus sign was observed in 69.8% of cases; of these, 68.2% showed a direct bronchus sign and 31.8% an adjacent pattern. All patients underwent PET–CT before the procedure, and 60.3% showed a maximum standardized uptake value (SUVmax)>2.5 (Table 1).

Regarding procedural metrics, the estimated mean procedure duration was 63.3min (SD, 12.5min). The mean total dose-area product (DAP) recorded for the procedures was 65.7Gycm2, with an estimated mean absorbed dose of approximately 73mGy. The DAP specifically attributable to CBCT acquisitions was 34.4Gycm2, corresponding to an estimated absorbed dose of 38mGy. The calculated effective dose was approximately 11.2mSv for the complete procedure and 5.8mSv for the CBCT component. The mean AF time was 7min. Multiple sampling techniques were used, including transbronchial forceps biopsy (TBFB) in all cases (100%), TBLC in 36.5%, FNA in 46%, and bronchial brushing in 38%.

The diagnostic yield was 65.1% at the time of bronchoscopy, increasing to 73% after follow-up adjudication. The most prevalent malignant diagnosis was non-small cell lung cancer, comprising adenocarcinoma in 38.1% and squamous cell carcinoma in 12.7%, together accounting for 50.8% of all lesions. Additional malignant diagnoses included small cell carcinoma, marginal zone lymphoma, and metastatic prostate cancer. Notably, molecular profiling was successfully performed in 65.6% of diagnosed non-small cell lung cancer cases (Table 2).

Regarding the diagnostic yield of the different sampling techniques, TBFB achieved a diagnostic yield of 60.3% (38/63), whereas TBLC showed a higher yield of 65.6% (21/32). FNA and brushing yielded diagnostic material in 27.6% (8/29) and 16.7% (4/24) of cases, respectively. Notably, in 15.2% of cases with a confirmed diagnosis, TBLC was the sole method that provided the diagnosis. In addition, diagnosis was confirmed in 58% (11/19) of lesions without a bronchus sign.

For the remaining 17 nondiagnostic lesions, the final diagnosis was established by surgical resection in 64.7% of cases, percutaneous biopsy in 23.5%, and radiological follow-up for at least 12 months in 11.8%. No factors were identified as significantly associated with diagnostic yield.

A previous diagnostic attempt had been performed unsuccessfully in 25.4% of cases: rEBUS alone in 56.2%, ENB combined with rEBUS in 31.3%, and bronchoalveolar lavage in the remaining 12.5%. Diagnosis was confirmed using CBCT-assisted ENB in 68.8% of these lesions. Among confirmed cases, 54.5% were diagnosed as NSCLC, including adenocarcinoma in 36% and squamous cell carcinoma in 18%.

Regarding safety outcomes, pneumothorax occurred in 4.7% of procedures (3 of 63), none of which required invasive management. No cases of clinically significant bleeding, respiratory failure, or other major complications were observed.

This prospective study shows that CBCT-assisted ENB achieves a high diagnostic yield in the evaluation of C-PPLs, reaching 65.1% at the time of bronchoscopy and 73% after longitudinal follow-up. These findings are consistent with previous studies that have highlighted the value of integrating advanced navigation technologies with intraprocedural imaging to improve lesion localization and sampling precision [5,9].

It is important to emphasize that all lesions included in the present cohort measured <20mm, with a mean size of 12.3mm, which is substantially smaller than in previously published series [10,11]. A well-established correlation exists between lesion size and diagnostic yield, with larger lesions being more likely to be successfully diagnosed [12,13]. The results observed in this study underscore the usefulness of this hybrid approach for the evaluation of C-PPLs. In addition, this technique confirmed the diagnosis in 68.8% of lesions that had previously been evaluated using other diagnostic methods.

Another key aspect in the diagnosis of PPLs is the method used to verify correct positioning of the navigation catheter before tissue sampling. In this study, CBCT was used after navigation to confirm placement of the catheter within or adjacent to the target lesion. This technique provides high-resolution, 3-dimensional anatomical imaging, improving precision and reducing sampling errors [14]. As shown in previous studies, CBCT can improve diagnostic accuracy by 10–20% compared with ENB alone [10,15]. Moreover, augmented fluoroscopy (AF) was used in dynamic mode throughout tissue acquisition. This provided continuous visual feedback, allowing intraprocedural adjustments during the sampling phase [16].

A noteworthy finding is that 30.2% of lesions did not show a bronchus sign on preprocedural CT. Despite this anatomical limitation, diagnosis was achieved in 58% of these cases. In this context, CBCT may compensate for the absence of a discernible bronchial pathway and the limited effectiveness of endoscopic verification using rEBUS.

The present study did not include rEBUS to verify lesion proximity endoscopically. rEBUS involves the use of a small ultrasound probe to obtain real-time images of the bronchial wall and surrounding lung tissue. When a lesion is in contact with the airway, the ultrasound signal generates a characteristic image, typically appearing as concentric or eccentric hypoechoic patterns [17]. However, the diagnostic value of rEBUS is especially notable in lesions with a positive bronchus sign on CT imaging. Furthermore, rEBUS does not involve ionizing radiation [18]. By contrast, CBCT and AF are extracorporeal imaging modalities that provide high anatomical precision independently of bronchial anatomy. Although these techniques involve exposure to ionizing radiation, the levels observed in this study (5.8mSv for CBCT and 11.2mSv for the complete procedure) remain within previously reported acceptable ranges [19,20]. As noted above, CBCT is particularly valuable in challenging cases involving small or complexly located lesions and in cases in which the preprocedural CT scan suggests that rEBUS would be unsuccessful or unfeasible.

Robotic-assisted bronchoscopy (RAB) has recently emerged as an alternative navigation platform for PPLs, offering enhanced navigation success and catheter stability during sampling. A recent comparative study reported that RAB combined with cone-beam CT guidance may provide a higher diagnostic yield than ENB combined with CBCT for pulmonary nodule evaluation [21]. Furthermore, a multicenter study highlighted the complementary role of mobile CBCT in optimizing lesion localization during robotic procedures [22]. Notably, the randomized RELIANT trial showed that the diagnostic yield of RAB was noninferior to that of ENB for the evaluation of PPLs. These findings suggest that differences between navigation technologies may be less important than their use in combination with CBCT during the procedure [23].

TBLC has emerged as a promising alternative method for tissue acquisition in C-PPLs, especially given the limitations of conventional TBFB, such as small sample size, crush artifacts, and suboptimal diagnostic yield. TBLC provides larger, better-preserved tissue specimens through a 360-degree freezing effect, enabling improved histopathological, immunohistochemical, and molecular analysis. This is a fundamental component in the diagnosis and management of NSCLC [24,25]. Several studies have reported diagnostic yields ranging from 74% to 94% using ENB combined with rEBUS, particularly in lesions smaller than 2cm and in those with eccentric or adjacent orientations, which typically reduce the accuracy of conventional TBFB [26]. In the present study, TBLC achieved a diagnostic yield of 65.6%. Although TBLC was performed in only 50.7% of PPLs, it is noteworthy that in up to 15.2% of lesions with a confirmed diagnosis, the diagnosis was obtained exclusively through this sampling technique. TBLC was performed using a 1.1-mm disposable cryoprobe (Erbe, Germany). The use of the 1.1-mm ultrathin cryoprobe provides enhanced flexibility and compatibility with ultrathin bronchoscopes and guide sheaths. Furthermore, this innovation may expand access to peripheral ground-glass opacities (GGOs), thereby improving their diagnosis [27,28].

The ENB-CBCT hybrid approach has considerable potential not only to improve diagnostic yield but also to enable bronchoscopic therapeutic interventions in PPLs. CBCT-guided transbronchial microwave ablation (MWA) has been reported to be a safe and effective technique [29,30]. The NAVABLATE study demonstrated the usefulness of CBCT for confirming navigation catheter placement and evaluating ablation margins [31]. At the same time, RAB platforms are being investigated for bronchoscopic therapeutic applications, given their potential to enhance catheter stability during energy delivery. A recent multicenter prospective study evaluated the feasibility and safety of RAB-guided bronchoscopic microwave ablation, further reflecting this technological evolution [32]. However, regardless of the navigation platform used, intraprocedural 3D confirmation with CBCT remains fundamental for accurate targeting and margin verification.

Pulsed electric fields (PEF) have recently been evaluated as a novel bronchoscopic ablation modality. A prospective first-in-human study demonstrated the feasibility and safety of endobronchial PEF application in patients with PPLs confirmed as NSCLC, supported by CBCT for precise energy delivery [33,34]. This integrated procedure has the potential to facilitate a seamless transition from biopsy to localized treatment in a single procedure. This could pave the way for future minimally invasive approaches, including different ablation techniques and drug delivery methods.

This study has several limitations. First, it was conducted at a single, high-expertise center with experienced operators, which may limit the generalizability of the findings to centers with fewer resources or lower procedural volume. Second, rEBUS was not used as part of the lesion confirmation process, although it may have enhanced the multimodal approach and potentially improved targeting in selected cases. Third, although radiation exposure remained within acceptable safety thresholds, it remains an important consideration when selecting patients for this type of procedure. Finally, although, to the best of our knowledge, this is the largest prospective series evaluating the hybrid ENB-CBCT workflow in C-PPLs, larger multicenter studies are required to confirm these results and characterize the performance of this approach in broader anatomical and clinical scenarios.

This prospective study demonstrates that a multimodal approach combining ENB with CBCT is safe and highly effective for diagnosing C-PPLs. The high diagnostic yield achieved, even in the absence of a bronchus sign or rEBUS confirmation, underscores the value of integrating high-resolution 3-dimensional imaging technologies during the procedure. Furthermore, TBLC was shown to be a valuable sampling method that can improve tissue quality without compromising patient safety. These findings support the clinical usefulness of a personalized diagnostic strategy adapted to lesion-specific characteristics, ultimately enabling earlier and more accurate detection of pulmonary malignancies through minimally invasive procedures.

The authors declare that they obtained the patient's informed consent in accordance with the ethical guidelines of the institution. This document is held by the corresponding author.

The material was produced by the authors, without partial or total assistance from artificial intelligence software or tools.

None declared.

None declared.