Physiological Imaging-defined Response-driven Subvolumes

Introduction: It has been hypothesized that dose sculpting/painting of physiological imagingdefined subvolumes of the tumors by intensity-modulated radiotherapy could lead to a better outcome than distributing a uniform dose within a target volume defined by anatomical imaging (1). Our goal is to develop a general method to delineate the subvolumes of a tumor based upon multiple physiological imaging and test their complementary roles for therapy assessment. Methods: Patients. Forty five brain metastases (median volume of 1.65 cm and range of 0.117.6 cm) from 20 patients were obtained in a prospective MRI study. The histology included melanoma (11), non-small cell lung cancer (6), renal cell carcinoma (1), breast cancer (1), and head & neck squamous cell carcinoma (1). All patients received WBRT with a total dose of 30 Gy in 10 fractions (13 patients) or 37.5 Gy in 15 fractions (7 patients). Thirteen patients also received bortezomib during WBRT as a radiation sensitizer as part of a phase II clinical trial. Data Acquisition. All patients had MRI scans (T1-weighted, post-Gd-DTPA T1-weighted, and dynamic contrast enhanced (DCE) images) prior to radiation therapy (Pre-RT), 2 weeks after the start of RT (2W), and 1 month after the completion of RT (1M Post-RT). Each lesion was manually contoured by a physician on the post-Gd T1 weighed images obtained pre-RT, 2W and 1M post-RT. The general Toft model was used to calculate rCBV and K maps (2). Probability Density Functions (PDF) of Physiological Parameters. To analyze a physiological parameter, e.g. rCBV, the PDF of the parameter of a lesion is calculated using a non-parametric PDF estimator. A value of the PDF at a point x, H(rCBV = x), of a lesion is proportional to the number of voxels within |rCBVi-x| < where is a smooth factor set as 4. ⁄ denotes the standard deviation of the rCBV distribution in the tumor. After obtaining the Pre-RT and 2W PDFs for each lesion (HPre(x) and H2W(x)), HPre(x) was normalized to have an area under the PDF curve equal to one ( 1). Then, the normalized HPre(x)s of all the lesions were summed to generate a pooled PDF (Hp) of brain metastases, in which each lesion has an equal contribution regardless of its volume, Fig. 1. Probabilistic Membership Function. Our goal is to find a set of probability membership functions associating with high, intermediate and low rCBV classes. Hence, the fuzzy-c-means (FCM) clustering method is used to partition Hp (rCBV) into three classes. The FCM does not classify a rCBV value into a single class rather generates a probabilistic function of a rCBV value that belongs to a class. The probabilistic membership function, Pj(rCBV), is a new representation of a rCBV value of a tumor voxel (mathematically transfers the data from a rCBV space into a new space) see Fig. 1. A similar computation is applied to K. Physiological-Parameter Defined Tumor Subvolume. Our primary goal is to test if a change in subvolumes of a tumor with high, intermediate or low rCBV values is related to tumor response. We define the subvolumes (SVs) of the tumor using Pj(rCBV), and calculate the percentage change in SV from Pre-RT to 2W as follows: