Interhospital Variability in Localization Techniques for Indeterminate Pulmonary Nodules in Children: A Pediatric Surgical Oncology Research Collaborative Study

Background: Indeterminate pulmonary nodules have various etiologies, including benign, malignant, and infectious. Thoracoscopic biopsy is widely accepted for diagnostic and therapeutic resection of occult pulmonary nodules in children, although subcentimeter lesions that are deep to the pleural surface can be difficult to identify intraoperatively. Various CT-guided techniques have been developed to assist in localization, however, there currently is no gold standard. The purpose of this study is to describe current trends in localization techniques and their performance in indeterminate pulmonary nodules in children.

Methods: A retrospective review using 15 member institutions of the Pediatric Surgical Oncology Research Collaborative was performed on patients under 21 years of age who underwent localization of pulmonary nodules. Patient demographics, lesion characteristics, and localization techniques were described. Localization and resection success, time in interventional radiology (IR), operating room (OR), and total anesthesia time, as well as complication rates were compared between different techniques using χ2 and linear regression. Localization and operative success were defined by the ability to localize or resect the lesion in question, respectively. Technical problems, including unsuccessful localization or malfunction after localization (e.g. wire dislodgement), were also recorded.

Results: 225 primary lesions were identified. In 49 patients, localization was attempted in more than one lesion. The most common preoperative diagnoses were Osteosarcoma (19%), Wilms Tumor (19%), and Ewing Sarcoma (13%). Patients had an average of 1.3 lesions (range 1-5) localized. Mean size and depth were 5.9 mm (range 0.5-35 mm) and 8.8 mm (0-61 mm), respectively. The most common localization techniques were: wire+methylene blue dye (MBD) (28%), MBD only (25%), wire only (14%), technetium 99 (Tc99) only (11%), coil+MBD (7%) and coil only (5%). Localization technique was associated with institution (p < 0.01); some institutions being exclusive users of certain techniques. Localization technique and institution were significantly associated with mean time in IR, OR, and total anesthesia time (all p<0.05) (Table 1). In procedures where only 1 lesion was localized and resected, wire+MBD took longer for all time periods (all p<0.01, Figure 1). Comparing the 6 most used techniques, there was no difference in successful IR localization (range 92-100%, p=0.68), successful resection (92-100%, p=0.30), IR technical problems (p=0.06), or operative complications (p=0.35). Lesion size and depth were not associated with success rate of localization, resection, or complications in IR or OR (all p>0.05). However, lesion depth was associated with technique (p < 0.01), with 70% of lesions deeper than 2cm localized by wire+MBD.

Conclusion: Many IR localization techniques for indeterminate pulmonary nodules in children are successful, but there is wide variability in choice by institution, as well as associated procedure times. Standardization of localization techniques may shorten anesthetic duration and reduce complications.