Radiologic-pathologic correlation of response to chemoradiation in resectable locally advanced NSCLC

- Early response assessment of chemoradiation can guide clinical-decision making.
- Volume change after chemoradiation is associated with pathological response for NSCLC.
- Volume change is further associated with prediction of locoregional control.
- Association with pathologic response is significant in the IIIa subgroup.
- RECIST response not associated with pathologic response or locoregional control.

ObjectivesAccurate assessment of tumor response to chemoradiation has the potential to guide clinical decision-making regarding surgical resection and/or dose escalation for patients. Early assessment has implications for Optimal local therapy for operable locally advanced non-small cell lung cancer (LA-NSCLC) is controversial. This study evaluated quantitative CT-based tumor measurements to predict pathologic response.Materials and methodsPatients with operable LA-NSCLC treated with chemoradiation followed by surgical resection were assessed. Tumor diameter and volume were quantified from CT imaging obtained prior to chemoradiation and post-chemoradiation prior to surgical resection. Univariate and multivariate logistic regression were used to determine association with the primary endpoint of pathologic complete response (pCR). Overall survival, locoregional recurrence, and distant metastasis were assessed as secondary endpoints.Results101 LA-NSCLC patients were identified and treated with preoperative chemoradiation and surgical resection. The median RT dose was 54 Gy (range, 46-70) and 98% of patients received concurrent chemoradiation as part of their preoperative treatment. Reduction of CT-defined tumor volume was associated with pCR (OR 1.06 [1.02-1.09], p = 0.002) and LRR (HR 1.01 [1.00-1.02], p = 0.048). Conventional response assessment determined by RECIST (p = 0.213) was not associated with pCR or any secondary endpoints.ConclusionCT-measured reductions in tumor volume after chemoradiation are associated with pCR and provide greater clinical information about tumor response than conventional response assessment (RECIST) or absolute tumor sizes or volumes. This study demonstrates that change in tumor volumes provides better radiologic-pathologic correlation and is thus an additional tool to assess tumor response following chemoradiation.