Patterns of Locoregional Relapses in Patients with Contemporarily Staged Stage III-N2 NSCLC Treated with Induction Chemotherapy and Resection: Implications for Postoperative Radiotherapy Target Volumes.

Objectives: Our aim was to evaluate locoregional relapse (LR) patterns after induction chemotherapy and surgery for stage III-N2 NSCLC staged with current standard methods and their impact on radiation target volumes for postoperative radiotherapy (PORT).

Methods: A total of 150 patients with stage III-N2 NSCLC from a prospective database of patients who underwent surgical resection at the University Hospitals of Leuven or the Oncologic Centre Limburg between 1998 and 2012 were included. Patients were staged with fluorodeoxyglucose F 18 positron emission tomography/computed tomography and brain imaging and treated with induction chemotherapy and surgery. PORT was performed for incomplete resection (R1/R2) and/or persistent nodal disease (ypN2). For the non-PORT group, we created a virtual planning target volume (PTV). In general, the clinical target volume encompassed the bronchial stump, the ipsilateral hilum, the subcarinal region (station 7), and the initially involved mediastinal lymph nodes.

Results: After a mean follow-up time of 49 months, the 5-year overall survival was 35.1% in all patients; disease-free survival was 31.8%. PORT was delivered to 70 patients. LR was seen in 26 patients in the PORT group (37%) and 32 in the non-PORT group (40%). Fifty-eight nodal relapse sites were seen in the PORT group (2.2 sites per patient) versus 113 in the non-PORT group (3.5 sites per patient) (p < 0.01). In the PORT group, the most frequent sites of LR were the ipsilateral hilum (21%), lymph node station 7 (15%), ipsilateral station 4 (9%), ipsilateral station 5 (9%) and ipsilateral station 6 (9%). For the non-PORT group these were station 7 (19%), ipsilateral 4 (16%), and ipsilateral hilum (14%). The dominant pattern of failure was inside (inside or both inside and outside) the PTV. Regarding the out-of-PTV relapses, 47% and 69% of LRs occurred in the contralateral mediastinum for the PORT and non-PORT groups, respectively. Out-of-PTV relapses occurred mostly in initially left-sided tumors.

Conclusions: Despite the limitations of this retrospective study, our data support the role of PORT in decreasing local relapses. Because of the large number of out-of-PTV relapses in the contralateral mediastinum, inclusion of elective contralateral lymph node stations in the PTV could be considered in left-sided tumors. However, prospective randomized trials are needed to verify this.