Journal of thoracic oncology : official publication of the International Association for the Study of Lung Cancer 2017 01 18() pii S1556-0864(16)33476-1
Variation in treatment and survival outcomes for NSCLC is high among patients with stage III or IV disease, but patients with untreated NSCLC have not been critically analyzed to evaluate for improvable outcomes. We evaluated treatment trends and their association with oncologic outcomes for NSCLC, hypothesizing that there are a substantial number of untreated patients who are similar to patients who undergo treatment.
Linear regression was used to calculate trends in utilization of treatment. Kaplan-Meier and Cox regression modeling were used to determine predictors of receiving treatment. Propensity matching was used to compare survival among subsets of treated versus untreated patients.
Patients with primary NSCLC were identified from the National Cancer Data base from 1998 to 2012, and 21% of patients (190,539) received no treatment. For stage IIIA and IV, the proportion of untreated patients increased over the study period by 0.21% and 0.4%, respectively (p = 0.003 and p < 0.0001). Regardless of stage, untreated patients had significantly shorter overall survival (OS) (p < 0.0001). Propensity-matched analyses of 6144 stage IIIA patient pairs treated with chemoradiation versus no treatment confirmed shorter OS for untreated patients (median 16.5 versus 6.1 months, p < 0.0001). For 19,046 stage IV patient pairs treated with chemotherapy versus no treatment, similar results were obtained (median OS 9.3 versus 2.0 months, p < 0.0001). CONCLUSIONS
The proportion of untreated patients with stage IIIA and IV disease is increasing. Survival outcomes among patients with advanced-stage disease are superior with treatment, independent of selection bias. The benefits and risks of treatment should be carefully assessed before choosing to forego treatment.