Background: Locoregional tumor failure (LRF) after definitive chemoradiation for patients with stage III NSCLC remains unacceptably high. This analysis sought to further define where LRF occurs relative to radiation dose received and pre-treatment PET scan-defined maximum standard uptake value (SUVmax).
Methods: This was a retrospective study analyzing patients with stage III NSCLC treated with definitive radiation between 2006 and 2011. LRF was defined as failure within the ipsilateral lung, hilum or mediastinum. The CT simulation scan with the radiation dose distribution was registered to the CT or PET/CT documenting LRF. The region of LRF was contoured, and the dose to 95% of the volume (D95) of LRF was extracted. The pre-treatment SUVmax was also extracted for the anatomic region of LRF.
Results: Sixty-one patients were identified. Median follow-up time was 19.1 months (range 2.37-76.33). Seventy four percent of patients were treated with 3-D conformal technique (3DCRT), 15% were treated with Intensity Modulated Radiotherapy (IMRT), and 11% were treated with a combination of 3DCRT and IMRT. Median prescribed radiation dose for all patients was 66 Gy (39.6-74). Concurrent chemotherapy was delivered in 90% of patients. Twenty-two patients (36%) developed a LRF, with a total of 39 anatomic regions of LRF identified. Median time to LRF was 11.4 months (3.5-44.6). Failures were distributed as follows: 36% were in-field failures, 27% were out-of-field failures, 18% were in-field and out-of-field failures, and 18% were in-field and marginal (recurrences within the field edge) failures. There were no isolated marginal failures. Of the patients that developed a LRF, 73% developed a LRF with an in-field component. Sixty-two percent of LRFs were nodal. The median pre-treatment SUVmax for the anatomic region of LRF for patients with an in-field failure was 13. The median D95 of in-field LRF was 63 Gy.
Conclusions: LRF after definitive chemoradiation are comprised primarily of in-field failures, though out-of field failures are not insignificant. Marginal failures are rare, indicating field margins are appropriate. Although radiation dose escalation to standard radiation fields has not yielded success, using PET parameters to define high-risk regions remains worthy of further investigation.