Obtaining pathologic diagnosis for patients with early stage lung cancer is often difficult due to medical comorbidities, tumour location, or patient preference. Far fewer patients over an 8-year period received a clinical diagnosis of early-stage lung cancer than a pathological diagnosis, and improved rates of cancer specific survival (CSS) but not overall survival were linked to this diagnostic mode.
Dr. Talha Shaikh, Department of Radiation Oncology, Fox Chase Cancer Centre, Philadelphia, USA presented findings during the ‘Multidisciplinary Management of Thoracic Malignancies’ session at the European Lung Cancer Conference (ELCC 2016), held in Geneva, Switzerland, 13 - 16 April 2016.
Dr. Shaikh and colleagues used information from a large database to assess trends in clinical diagnosis of early-stage lung cancer and to determine the association between diagnosis and treatment outcomes. Data contained in the Surveillance, Epidemiology, and End Results registry over an 8-year period beginning in 2004 were reviewed. This analysis included data from 7050 patients aged 18 years and older that received only radiation therapy for stage 1, clinical T1a-T2a lung cancer. The association between patient outcomes of overall survival (OS) were determined by Cox proportional hazards model and competing risk regression analysis was used to assess CSS with each type of diagnosis and subsequent treatment with radiotherapy.
Clinical diagnosis associates with improved cancer specific but not overall survival
The investigators saw no significant change over the course of the study between the diagnostic methods used (p = 0.172): Pathological diagnosis was made in 6399 (90.8%) patients compared with 651 (9.2%) patients who were clinically diagnosed.
Patients with clinically-diagnosed disease demonstrated improved CSS outcomes that were significant in some patient subgroups.
Significant differences were observed between CSS and the mode of diagnosis. By multivariable analysis, improved CSS associated with clinical diagnosis, HR 0.82; 95% CI 0.71, 0.96. In patients stratified by clinical tumour stage, improved CSS was observed in clinically diagnosed T1a patients, HR 0.75; 95% CI 0.58, 0.96 (p = 0.022) and a trend towards improved CSS was observed in patients with clinically diagnosed T1b tumours, HR 0.74; 95% CI 0.55,1.00 (p = 0.052).
The investigators also reported a stepwise increase in the hazard ratio (HR) for CSS according to T-stage and tumour size when comparing clinically-diagnosed versus pathologically-diagnosed tumours.
When the outcome of the patients in the clinical diagnosis cohort was evaluated according to interval quartile tumour size, improved CSS was seen in patients with smaller tumours ranging between 0 to 1.9 cm, HR 0.74; 95% CI 0.58, 0.99; (p = 0.040) and a trend towards improved CSS was seen in patients with tumours sized 2.0 to 2.7 cm, HR 0.78; 95% CI 0.58, 1.03 (p = 0.083).
However, clinically-diagnosed early-stage lung cancer did not associate with overall survival, HR 1.01; 95% CI 0.90,1.13.
Dr Gaetano Rocco of the Department of Thoracic Surgical and Medical Oncology, Istituto Nazionale Tumori, Fondazione Pascale, IRCCS, Naples, Italy, who discussed the study findings, said that in recent literature, pre-SABR (stereotactic ablative radiotherapy) histological diagnosis is missing in 8-100% of the patients depending on cardiopulmonary function. Furthermore, overlapping outcomes were reported between biopsy-proven and not histologically-diagnosed lung cancer SABR-treated patients. He further eleaborated drawbacks such as significant acute and chronic toxicities in compromised patients especially for SABR, retrospective analysis of non homogenous patient cohorts, overall (3 year) rather than cancer specific survival report, and relatively short median follow-up.
By elaborating the study conclusions, Dr Rocco emphasized that the SEER registry does not include data on comorbidities, performance status, margin status, radiation dose, or chemotherapy use. Serious propensity score analysis in this context is difficult. He questioned what were, if any, the acute and chronic toxicities reported in non-histologically diagnosed patients undergoing SABR, could a stratification of CSS between SABR treated peripheral vs central tumours be performed, is the 17-month median follow-up acceptable to draw definitive conclusions, a need for critical revision of the manuscripts reporting outcomes of SABR in mixed population (diagnosed and non-diagnosed) and possibly focusing on histologically proven patients from now on, should we abandon reporting OS in favor of CSS and the implications in clinical practice.
The authors concluded that the improved CSS seen in clinically-diagnosed patients suggests possible overtreatment of benign disease, particularly in patients with smaller tumours. They advised that prudent patient selection is necessary to reduce the potential for overtreatment.