Abstract 944P
Background
Patients who undergo surgery for lung cancer receive postoperative surveillance with regular clinical assessments. However, an optimal surveillance program for the early detection of recurrence is still to be investigated. Here we aimed to optimize surveillance through an investigation of the hazard function of tumor recurrence in patients with completely resected lung cancer.
Methods
We analyzed the records of 12,897 patients of a 2010 Japanese Joint Committee of Lung Cancer Registry who underwent lobectomy to completely resect pathological stages I–III lung cancer. The risk of recurrence after surgery was determined using a cause-specific hazard function. The estimated hazard was obtained using an Epanechnikov kernel smoothing method. The recurrence site was categorized as intrathoracic recurrence (IR) or extrathoracic recurrence (ER). The risk distribution was assessed using clinicopathological profiles.
Results
The hazard function for recurrence exhibited a peak at approximately 9 months after surgery, followed by a tapered plateau-like tail extending to 60 months. The peak risk for IR was approximately two-fold higher compared with that of ER. In particular, the risk for recurrence in bone and brain peaked between 6 and 12 months and decreased sharply until 24 months when it was 5–10 times lower than for IR. The timing of the IR peak was a few months later than the ER peak. Subgroup analysis showed that patients with stage IIIA adenocarcinoma had a continuously higher risk of recurrence compared with patients with earlier stages. However, the risk of recurrence in patients with squamous cell carcinoma was not significantly different compared with that more than 24 months after surgery, regardless of pathological stage.
Conclusions
This study revealed the specified period at the highest risk of postoperative recurrence for each type of disease, suggesting that routine examination with chest CT may be useful even after the second postoperative year due to the high frequency of intrathoracic recurrences. In this manner, this study will help the comprehensive postoperative surveillance for recurrence.
Clinical trial identification
Editorial acknowledgement
Legal entity responsible for the study
The authors.
Funding
Has not received any funding.
Disclosure
All authors have declared no conflicts of interest.