Aims
Lung cancer continues to remain the leading cause of cancer death in Australia (1). Timely and complete diagnosis (as reflected by satisfactory histopathology and molecular characterization) improves patient outcomes, while with incomplete data many may receive inappropriate initial therapy which can affect outcomes and decrement quality of life (2,3,4) Whilst the two most widely used diagnostic modalities are Computer Tomography guided percutaneous lung biopsy (CT-PLB) and bronchoscopy with endobronchial ultrasound (EBUS), not all facilities have these readily available onsite. Other modalities (e.g., sputum and pleural fluid cytology) have limitations and are less preferred. Our institution has onsite CT-PLB with visiting respiratory physicians, but offsite pathology/cytology and bronchoscopy. With such we undertook to audit the method undertaken to complete a clinically satisfactory diagnosis and the samples completeness regarding molecular analysis.
Methods
Between the period 1 July 2017 and 30 June 2018, all patients with a diagnosis of primary advanced lung cancer at out medium sized regional hospital (210 bed) and who had received treatment were identified via an electronic database. Information was collected regarding diagnostic modality used, time to investigation completion, satisfactory completion of requisite information (histopathology, ALK IHC, ROS-1 IHC, PDL-1 IHC and EGFR mutation) Retrospective data was collected using chart reviews in conjunction with review of biopsy results and external and internal imaging results. Graphical statistical analysis has been used to present our data.
Results
Of 50 patients identified, 37 patient’s diagnostic decisions were initiated locally. Diagnostic information was obtained mostly via bronchoscopy (n=17), CT-PLB (n=10), Other (n=10) [extrathoracic lymph node, n=5; pleural fluid, n=1; sputum, n=1; bone, n=1; liver, n=1; peritoneum, n=1].
From initial radiographic abnormality, the mean time to biopsy was: CT-PLB, 16 days; Bronchoscopy, 20 days; Other, 14 days. Of CT-PLB biopsies 8 /10 were General Physician referred, with bronchoscopy 10/17 were Respiratory Physician referred.
9/10 (90%) CT-PLB were diagnostic with 6/10 (60%) considered sufficient for molecular testing and immunohistochemistry. 15/17 (88%) bronchoscopic biopsies were diagnostic with 12/17 (71%) considered sufficient for molecular testing and immunohistochemistry. PET scan was performed prior to biopsy in 7 cases (CT-PLB, n=1, bronchoscopy, n=6). Histological subtypes included: Adenocarcinoma=23, Squamous cell carcinoma=4, SCLC=7, NET=1, NSCLC-unspecified=2 [CT-PLB, n=1, Bronchoscopy, n=1]. Repeat biopsy was performed in 22% (n=8).
Seven EGFR mutations were detected (CT-PLB=1 [NGS=1], bronchoscopy=5 [NGS=3, PCR=2], Other=1 [NGS=1]).
Conclusions
As expected, broncoscopy was the most preferred diagnostic route. Extra-thoracic biopsies occurred quickest but required repeat biopsy more often. Both CT-PLB and bronchoscopic biopsy had similar performance characteristics in terms of completeness of diagnostic information (i.e., molecular and histopathology), but about 30% still yielded incomplete information. It is uncertain if pre-PET evaluation or MDT review prior to the diagnostic test would help to improve the sample completeness and thus avoid re-biopsy.