Surgical resectability of lung cancer remains a critical factor influencing survival, but approximately 50% of patients present with unresectable stage III disease due to mediastinal disease. With 5 weeks of preoperative neoadjuvant chemo/radiotherapy (PNT) to reduce the mediastinal disease, 67- 85% of Stage III lung cancers become eligible for surgical resection with reported 33% improvement in 5-year survival. The decision for surgical resection depends upon CT evidence of mediastinal disease responding favorably to PNT. However, 10-20% of the patients becoming eligible for surgical resection are later found unresectable intraoperatively with devastating outcome. The ability to identify those patients who may show excellent PNT response as judged by CT but fail in surgical resection can have profound impact to the treatment strategy and outcome. ? ? This imaging-based outcome research is to test the hypothesis that tumor perfusion status during treatment, reflected by the dynamic contrast enhanced (DCE) MR, influences the resectability of stage III lung cancer treated with PNT. This translational research applies a novel approach to in vivo monitoring of PNT response by assessing (1) directly the efficiency in delivery of chemotherapy or oxygen (hypoxia) to the tumor during treatment that critically influence the effectiveness of chemo- or radiation therapy respectively, and (2), indirectly the tumor shrinkage rate that reflects the response of the tumor to the ongoing treatment. ? ? Specific Aim 1: To develop, refine, and test MR imaging protocol and analysis methodologies for assessing response of PNT in lung cancer.
Specific Aim 2 : To apply sequential MR studies and imaging analysis in a clinical population during the course of PNT and assess PNT response with CT, surgical and pathological findings.
Specific Aim 3 : To determine the added predictive value and best timing of the MR predictive assay. ? ? On completion, a MR based early predictive assay for stage III lung cancer treated with PNT will be developed, tested and refined with limited patients in a clinical setting, and the preliminary assessment of the added predictive value of MR parameters will be made by the CT, surgical and pathological findings. We will further discern the differences between favorable and unfavorable responders to PNT and refine the imaging-based predictive algorithm with (1) multi-spectral pixel-by-pixel analysis and cluster analysis by characterizing those low-DCE pixels within a heterogeneous tumor mass which likely represent poorly-perfused tumor regions contributing to treatment failure; and with (2) multi-temporal imaging analysis (sequential MR studies) by identifying the best timing for response assessment.Project Narrative ? ? Even a modest (e.g., 10%) improvement in lung cancer survival rates through this study may result in thousands of lives saved or prolonged every year in the U.S and we believe even with marginal improvement, the overall impact of this proposed study can be enormous considering the high incidence and mortality of lung cancer. ? ? ?
