Joanna H. Shih

Multiparametric 3T prostate magnetic resonance imaging to detect cancer: histopathological correlation using prostatectomy specimens processed in customized magnetic resonance imaging based molds.

PURPOSE We determined the prostate cancer detection rate of multiparametric magnetic resonance imaging at 3T. Precise one-to-one histopathological correlation with magnetic resonance imaging was possible using prostate magnetic resonance imaging based custom printed specimen molds after radical prostatectomy. MATERIALS AND METHODS This institutional review board approved prospective study included 45 patients (mean age 60.2 years, range 49 to 75) with a mean prostate specific antigen of 6.37 ng/ml (range 2.3 to 23.7) who had biopsy proven prostate cancer (mean Gleason score of 6.7, range 6 to 9). Before prostatectomy all patients underwent prostate magnetic resonance imaging using endorectal and surface coils on a 3T scanner, which included triplane T2-weighted magnetic resonance imaging, apparent diffusion coefficient maps of diffusion weighted magnetic resonance imaging, dynamic contrast enhanced magnetic resonance imaging and spectroscopy. The prostate specimen was whole mount sectioned in a customized mold, allowing geometric alignment to magnetic resonance imaging. Tumors were mapped on magnetic resonance imaging and histopathology. Sensitivity, specificity, positive predictive value and negative predictive value of magnetic resonance imaging for cancer detection were calculated. In addition, the effects of tumor size and Gleason score on the sensitivity of multiparametric magnetic resonance imaging were evaluated. RESULTS The positive predictive value of multiparametric magnetic resonance imaging to detect prostate cancer was 98%, 98% and 100% in the overall prostate, peripheral zone and central gland, respectively. The sensitivity of magnetic resonance imaging sequences was higher for tumors larger than 5 mm in diameter as well as for those with higher Gleason scores (greater than 7, p <0.05). CONCLUSIONS Prostate magnetic resonance imaging at 3T allows for the detection of prostate cancer. A multiparametric approach increases the predictive power of magnetic resonance imaging for diagnosis. In this study accurate correlation between multiparametric magnetic resonance imaging and histopathology was obtained by the patient specific, magnetic resonance imaging based mold technique.

Predicting survival in patients with metastatic kidney cancer by gene-expression profiling in the primary tumor

To identify potential molecular determinants of tumor biology and possible clinical outcomes, global gene-expression patterns were analyzed in the primary tumors of patients with metastatic renal cell cancer by using cDNA microarrays. We used grossly dissected tumor masses that included tumor, blood vessels, connective tissue, and infiltrating immune cells to obtain a gene-expression “profile” from each primary tumor. Two patterns of gene expression were found within this uniformly staged patient population, which correlated with a significant difference in overall survival between the two patient groups. Subsets of genes most significantly associated with survival were defined, and vascular cell adhesion molecule-1 (VCAM-1) was the gene most predictive for survival. Therefore, despite the complex biological nature of metastatic cancer, basic clinical behavior as defined by survival may be determined by the gene-expression patterns expressed within the compilation of primary gross tumor cells. We conclude that survival in patients with metastatic renal cell cancer can be correlated with the expression of various genes based solely on the expression profile in the primary kidney tumor.

Prostate Cancer: Interobserver Agreement and Accuracy with the Revised Prostate Imaging Reporting and Data System at Multiparametric MR Imaging

PURPOSE To evaluate accuracy and interobserver variability with the use of the Prostate Imaging Reporting and Data System (PI-RADS) version 2.0 for detection of prostate cancer at multiparametric magnetic resonance (MR) imaging in a biopsy-naïve patient population. MATERIALS AND METHODS This retrospective HIPAA-compliant study was approved by the local ethics committee, and written informed consent was obtained from all patients for use of their imaging and histopathologic data in future research studies. In 101 biopsy-naïve patients with elevated prostate-specific antigen levels who underwent multiparametric MR imaging of the prostate and subsequent transrectal ultrasonography (US)-MR imaging fusion-guided biopsy, suspicious lesions detected at multiparametric MR imaging were scored by five readers who were blinded to pathologic results by using to the newly revised PI-RADS and the scoring system developed in-house. Interobserver agreement was evaluated by using κ statistics, and the correlation of pathologic results with each of the two scoring systems was evaluated by using the Kendall τ correlation coefficient. RESULTS Specimens of 162 lesions in 94 patients were sampled by means of transrectal US-MR imaging fusion biopsy. Results for 87 (54%) lesions were positive for prostate cancer. Kendall τ values with the PI-RADS and the in-house-developed scoring system, respectively, at T2-weighted MR imaging in the peripheral zone were 0.51 and 0.17 and in the transitional zone, 0.45 and -0.11; at diffusion-weighted MR imaging, 0.42 and 0.28; at dynamic contrast material-enhanced MR imaging, 0.23 and 0.24, and overall suspicion scores were 0.42 and 0.49. Median κ scores among all possible pairs of readers for PI-RADS and the in-house-developed scoring system, respectively, for T2-weighted MR images in the peripheral zone were 0.47 and 0.15; transitional zone, 0.37 and 0.07; diffusion-weighted MR imaging, 0.41 and 0.57; dynamic contrast-enhanced MR imaging, 0.48 and 0.41; and overall suspicion scores, 0.46 and 0.55. CONCLUSION Use of the revised PI-RADS provides moderately reproducible MR imaging scores for detection of clinically relevant disease.

Effects of pooling mRNA in microarray class comparisons

MOTIVATION In microarray experiments investigators sometimes wish to pool RNA samples before labeling and hybridization due to insufficient RNA from each individual sample or to reduce the number of arrays for the purpose of saving cost. The basic assumption of pooling is that the expression of an mRNA molecule in the pool is close to the average expression from individual samples. Recently, a method for studying the effect of pooling mRNA on statistical power in detecting differentially expressed genes between classes has been proposed, but the different sources of variation arising in microarray experiments were not distinguished. Another paper recently did take different sources of variation into account, but did not address power and sample size for class comparison. In this paper, we study the implication of pooling in detecting differential gene expression taking into account different sources of variation and check the basic assumption of pooling using data from both the cDNA and Affymetrix GeneChip microarray experiments. RESULTS We present formulas for the required number of subjects and arrays to achieve a desired power at a specified significance level. We show that due to the loss of degrees of freedom for a pooled design, a large increase in the number of subjects may be required to achieve a power comparable to that of a non-pooled design. The added expense of additional samples for the pooled design may outweigh the benefit of saving on microarray cost. The microarray data from both platforms show that the major assumption of pooling may not hold. SUPPLEMENTARY INFORMATION Supplementary material referenced in the text is available at http://linus.nci.nih.gov/brb/TechReport.htm.

Accuracy and agreement of PIRADSv2 for prostate cancer mpMRI: A multireader study

Multiparametric MRI (mpMRI) improves the detection of clinically significant prostate cancer, but is limited by interobserver variation. The second version of theProstate Imaging Reporting and Data System (PIRADSv2) was recently proposed as a standard for interpreting mpMRI. To assess the performance and interobserver agreement of PIRADSv2 we performed a multi‐reader study with five radiologists of varying experience.

Statistical issues in the design and analysis of gene expression microarray studies of animal models.

Appropriate statistical design and analysis of gene expression microarray studies is critical in order to draw valid and useful conclusions from expression profiling studies of animal models. In this paper, several aspects of study design are discussed, including the number of animals that need to be studied to ensure sufficiently powered studies, usefulness of replication and pooling, and allocation of samples to arrays. Data preprocessing methods for both cDNA dual-label spotted arrays and Affymetrix-style oligonucleotide arrays are reviewed. High-level analysis strategies are briefly discussed for each of the types of study aims, namely class comparison, class discovery, and class prediction. For class comparison, methods are discussed for identifying genes differentially expressed between classes while guarding against unacceptably high numbers of false positive findings. Various clustering methods are discussed for class discovery aims. Class prediction methods are briefly reviewed, and reference is made to the importance of proper validation of predictors.

Prospective Evaluation of PI-RADS™ Version 2 Using the International Society of Urological Pathology Prostate Cancer Grade Group System

Purpose: The PI‐RADS™ (Prostate Imaging Reporting and Data System), version 2 scoring system, introduced in 2015, is based on expert consensus. In the same time frame ISUP (International Society of Urological Pathology) introduced a new pathological scoring system for prostate cancer. Our goal was to prospectively evaluate the cancer detection rates for each PI‐RADS, version 2 category and compare them to ISUP group scores in patients undergoing systematic biopsy and magnetic resonance imaging‐transrectal ultrasound fusion guided biopsy. Materials and Methods: A total of 339 treatment naïve patients prospectively underwent multiparametric magnetic resonance imaging evaluated with PI‐RADS, version 2 with subsequent systematic and fusion guided biopsy from May 2015 to May 2016. ISUP scores were applied to pathological specimens. An ISUP score of 2 or greater (ie Gleason 3 + 4 or greater) was defined as clinically significant prostate cancer. Cancer detection rates were determined for each PI‐RADS, version 2 category as well as for the T2 weighted PI‐RADS, version 2 categories in the peripheral zone. Results: The cancer detection rate for PI‐RADS, version 2 categories 1, 2, 3, 4 and 5 was 25%, 20.2%, 24.8%, 39.1% and 86.9% for all prostate cancer, and 0%, 9.6%, 12%, 22.1% and 72.4% for clinically significant prostate cancer, respectively. On T2‐weighted magnetic resonance imaging the cancer detection rate in the peripheral zone was significantly higher for PI‐RADS, version 2 category 4 than for overall PI‐RADS, version 2 category 4 in the peripheral zone (all prostate cancer 36.6% vs 48.1%, p = 0.001, and clinically significant prostate cancer 22.9% vs 32.6%, p = 0.002). Conclusions: The cancer detection rate increases with higher PI‐RADS, version 2 categories.

Prospective Study Evaluating Na18F PET/CT in Predicting Clinical Outcomes and Survival in Advanced Prostate Cancer

This prospective pilot study evaluated the ability of Na18F PET/CT to detect and monitor bone metastases over time and its correlation with clinical outcomes and survival in advanced prostate cancer. Methods: Sixty prostate cancer patients, including 30 with and 30 without known bone metastases by conventional imaging, underwent Na18F PET/CT at baseline, 6 mo, and 12 mo. Positive lesions were verified on follow-up scans. Changes in SUVs and lesion number were correlated with prostate-specific antigen change, clinical impression, and overall survival. Results: Significant associations included the following: SUV and prostate-specific antigen percentage change at 6 mo (P = 0.014) and 12 mo (P = 0.0005); SUV maximal percentage change from baseline and clinical impression at 6 mo (P = 0.0147) and 6–12 mo (P = 0.0053); SUV change at 6 mo and overall survival (P = 0.018); number of lesions on Na18F PET/CT and clinical impression at baseline (P < 0.0001), 6 mo (P = 0.0078), and 12 mo (P = 0.0029); and number of lesions on Na18F PET/CT per patient at baseline and overall survival (P = 0.017). In an exploratory analysis, paired 99mTc-methylene diphosphonate bone scans (99mTc-BS) were available for 35 patients at baseline, 19 at 6 mo, and 14 at 12 mo (68 scans). Malignant lesions on Na18F PET/CT (n = 57) were classified on 99mTc-BS as malignant 65% of the time, indeterminate 25% of the time, and negative 10% of the time. Additionally, 69% of paired scans showed more lesions on Na18F PET/CT than on 99mTc-BS. Conclusion: The baseline number of malignant lesions and changes in SUV on follow-up Na18F PET/CT significantly correlate with clinical impression and overall survival. Na18F PET/CT detects more bone metastases earlier than 99mTc-BS and enhances detection of new bone disease in high-risk patients.

On Estimating the Relationship between Longitudinal Measurements and Time‐to‐Event Data Using a Simple Two‐Stage Procedure

Ye, Lin, and Taylor (2008, Biometrics 64, 1238-1246) proposed a joint model for longitudinal measurements and time-to-event data in which the longitudinal measurements are modeled with a semiparametric mixed model to allow for the complex patterns in longitudinal biomarker data. They proposed a two-stage regression calibration approach that is simpler to implement than a joint modeling approach. In the first stage of their approach, the mixed model is fit without regard to the time-to-event data. In the second stage, the posterior expectation of an individuals random effects from the mixed-model are included as covariates in a Cox model. Although Ye et al. (2008) acknowledged that their regression calibration approach may cause a bias due to the problem of informative dropout and measurement error, they argued that the bias is small relative to alternative methods. In this article, we show that this bias may be substantial. We show how to alleviate much of this bias with an alternative regression calibration approach that can be applied for both discrete and continuous time-to-event data. Through simulations, the proposed approach is shown to have substantially less bias than the regression calibration approach proposed by Ye et al. (2008). In agreement with the methodology proposed by Ye et al. (2008), an advantage of our proposed approach over joint modeling is that it can be implemented with standard statistical software and does not require complex estimation techniques.

AN APPROACH FOR JOINTLY MODELING MULTIVARIATE LONGITUDINAL MEASUREMENTS AND DISCRETE TIME-TO-EVENT DATA

In many medical studies, patients are followed longitudinally and interest is on assessing the relationship between longitudinal measurements and time to an event. Recently, various authors have proposed joint modeling approaches for longitudinal and time-to-event data for a single longitudinal variable. These joint modeling approaches become intractable with even a few longitudinal variables. In this paper we propose a regression calibration approach for jointly modeling multiple longitudinal measurements and discrete time-to-event data. Ideally, a two-stage modeling approach could be applied in which the multiple longitudinal measurements are modeled in the first stage and the longitudinal model is related to the time-to-event data in the second stage. Biased parameter estimation due to informative dropout makes this direct two-stage modeling approach problematic. We propose a regression calibration approach which appropriately accounts for informative dropout. We approximate the conditional distribution of the multiple longitudinal measurements given the event time by modeling all pairwise combinations of the longitudinal measurements using a bivariate linear mixed model which conditions on the event time. Complete data are then simulated based on estimates from these pairwise conditional models, and regression calibration is used to estimate the relationship between longitudinal data and time-to-event data using the complete data. We show that this approach performs well in estimating the relationship between multivariate longitudinal measurements and the time-to-event data and in estimating the parameters of the multiple longitudinal process subject to informative dropout. We illustrate this methodology with simulations and with an analysis of primary biliary cirrhosis (PBC) data.