Prognostic value of tumor hypoxia, as measured by 18F-FMISO Breath Hold PET/CT

Sloan-Kettering Inst Can Research, New York NY

Investigators

Linked publications & trials

Abstract

DESCRIPTION (provided by applicant): In this proposal, we investigate the feasibility of performing compartmental analysis (CA) of 18FM1S0 dynamic PET (dynPET) images in NSCLC patients undergoing neoadjuvant chemotherapy. The 5-yr survival of stage IA NSCLC patients treated with surgery alone is 73%, falling to 58% and 13% for stages IB and IIIA respectively for patients who are surgically staged. Thus, it is necessary to identify prognostic factors determining patient outcome, which will ultimately lead to new treatment regimens. Tumor hypoxia is an independent predictor of poor prognosis in several types of cancer, including NSCLC. Achieving successful hypoxia imaging for prognosis in lung cancer is of greater relevance than for other cancers (e.g. HNC) because the local control in NSCLC is comparatively low. 18FM1S0 PET has been shown to correlate with hypoxia in NSCLC, and that the optimum method to quantitate hypoxia using 18FMIS0 is via compartment analysis of the corresponding dynPET images. However, CA is implausible with PET images of lung due to breathing. We propose a novel technique for motion-correction of 18FMIS0 dynPET images of lung to make CA possible. Our group at MSKCC was the first to establish the methodology for both respiratory gated and Breath-Hold (BH) PET/CT. We will quantitate the uncertainties in CA due to target motion using Monte Carlo simulations of oscillating hypoxic targets in normoxic lung background. We will then demonstrate our method's feasibility in clinical settings. We shall acquire patient clinical 18FMIS0 dynPET images in Free- Breathing (FB) in list mode (LM). We will select end-expiration (EE) events from the LM data using the patient's respiratory signal to produce a BH-like PET data set. Kinetic parameters will be determined from CA of both FB and BH dynPET images and then compared to assess differences due to target motion. We will also examine the feasibility of CA of 18FMIS0 BH dynPET of NSCLC in patients, and will assess the reproducibility of CA in repeat 18FMIS0 dynPET. Finally, we will investigate whether BH 18FMIS0 PET results in an improved correlation with treatment outcome. If this is the case, this approach may set the stage for future studies in which hypoxia imaging findings will be used to modify treatment.