-
[show abstract]
[hide abstract]
ABSTRACT: To retrospectively evaluate the accuracy of endorectal magnetic resonance (MR) imaging in the detection and local staging of transition zone prostate cancers, with pathologic analysis serving as the reference standard, and to assess MR imaging features of these cancers.
The institutional review board approved this HIPAA-compliant retrospective study and waived the informed consent requirement. An institutional database of 986 patients who underwent MR imaging before radical prostatectomy yielded 148 consecutive patients with at least one transition zone cancer at step-section pathologic analysis. An additional 46 patients without transition zone cancer were randomly selected as a control group. Two readers independently reviewed MR studies to identify patients with transition zone cancers and determine the location and local extent of these cancers. Imaging features that helped in the identification of transition zone cancers were recorded. Descriptive and kappa statistics, as well as receiver operating characteristic and multivariate logistic regression analyses, were used.
For identification of patients with transition zone cancers, sensitivity and specificity were 75% and 87%, respectively, for reader 1 and 80% and 78%, respectively, for reader 2. Interreader agreement was fair. For detection of the location of transition zone cancer, the area under the receiver operating characteristic curve was 0.75 for reader 1 and 0.73 for reader 2. Interreader agreement was fair. The readers' accuracy in detecting transition zone cancer foci increased significantly (P=.001) as tumor volume increased. In the detection of extraprostatic extension of transition zone cancers, sensitivity and specificity were 56% and 94%, respectively, for reader 1 and 28% and 93%, respectively, for reader 2. Homogeneous low T2 signal intensity (P=.001 for reader 1, P<.001 for reader 2) and lenticular shape (P=.017 for reader 1) were significantly associated with the presence of transition zone cancer.
MR imaging can be used to detect, localize, and stage transition zone prostate cancers.
Radiology 06/2006; 239(3):784-92. · 5.73 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: To retrospectively determine the accuracy of endorectal magnetic resonance (MR) imaging in demonstrating seminal vesicle invasion (SVI) and to investigate the MR imaging features that can predict SVI.
The Institutional Review Board granted exempt status for this retrospective study, with waiver of informed consent; patient data were collected and handled in accordance with HIPAA regulations. Fifty-one men (age range, 44-73 years) with SVI and 303 men (age range, 40-76 years) without SVI who underwent endorectal MR imaging before radical prostatectomy between January 2000 and October 2004 were included in the study. Endorectal MR images were retrospectively and independently analyzed by two radiologists for SVI, tumor at prostate base, extracapsular extension, and other features considered indicative of SVI. Areas under the receiver operating characteristic curves (AUCs) were used to assess the accuracy of detecting SVI at endorectal MR imaging. A multiple logistic regression was used to explore the combinations of MR imaging features that might facilitate the detection of SVI.
Readers 1 and 2 had an AUC of 0.93 and 0.81, respectively, for the detection of SVI. For both readers, the features that had the highest sensitivity and specificity were low signal intensity within the seminal vesicle and lack of preservation of seminal vesicle architecture. At multiple regression analysis, tumor at the prostate base that extended beyond the capsule and low signal intensity within a seminal vesicle that has lost its normal architecture were highly predictive of SVI.
Endorectal MR imaging is accurate in demonstrating SVI prior to radical prostatectomy, and recognition of the most predictive features may facilitate the use of this modality.
Radiology 04/2006; 238(3):929-37. · 5.73 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: To evaluate retrospectively the accuracy of endorectal magnetic resonance (MR) imaging for the depiction of tumor, extracapsular extension (ECE), and seminal vesicle invasion (SVI) before salvage prostatectomy in patients with locally recurrent prostate cancer after radiation therapy, by using pathologic analysis as the reference standard.
The Institutional Review Board granted exempt status for this HIPAA-compliant study, with a waiver of informed consent. Forty-five consecutive patients (age range, 43-76 years) were identified who underwent salvage radical prostatectomy for prostate cancer at Memorial Sloan-Kettering Cancer Center between December 1, 1998, and October 31, 2004, and who underwent endorectal MR imaging prior to surgery. Tumor localization and determination of local stage with MR imaging were performed independently by two radiologists. Interpretations were compared to pathologic findings from surgical specimens. Interrater variability was estimated with the kappa statistic. Areas under the receiver operating characteristic curve (AUCs) were used to assess the accuracy of endorectal MR imaging in tumor detection and determination of ECE and SVI.
Findings of histologic examination showed that tumor was present in all patients. For tumor detection, the AUC value for reader 1 was 0.75 (95% confidence interval [CI]: 0.67, 0.84), whereas the AUC value for reader 2 was 0.61 (95% CI: 0.52, 0.71). The AUC values for prediction of ECE were 0.87 (95% CI: 0.80, 0.94) for reader 1 and 0.76 (95% CI: 0.67, 0.85) for reader 2. The AUC values for prediction of SVI were 0.76 (95% CI: 0.62, 0.90) for reader 1 and 0.70 (95% CI: 0.56, 0.85) for reader 2. For all variables, the kappa statistics used to assess interrater agreement between readers were fair (0.45, 0.52, and 0.47 for tumor location, ECE, and SVI, respectively).
Endorectal MR imaging following radiation therapy can help identify tumor sites and depict ECE and SVI with reasonable accuracy in patients with recurrent prostate cancer.
Radiology 02/2006; 238(1):176-83. · 5.73 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: Although statistical methodology is well developed for comparing diagnostic tests in terms of their sensitivities and specificities, comparative inference about predictive values is not.
In this paper we consider the design and analysis of studies comparing the positive and negative predictive values of two diagnostic tests that are measured on all subjects.
We focus on comparing tests using the relative positive and negative predictive values. We discuss directly estimating these quantities from the data and derive analytic variance expressions. Sample size formulas for study design ensue.
We analyze data on patients with cystic fibrosis to illustrate the methodology. This approach is compared and contrasted with an existing regression framework that can also be used for similar analysis purposes and yields similar results.
We have developed a new approach for comparing the predictive values of two tests that gives rise to sample size formulas for study design.
Clinical Trials 02/2006; 3(3):272-9. · 1.92 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: Blood oxygen level-dependent functional MR imaging (BOLD fMRI) is a clinically useful technique for preoperative mapping of eloquent cortices in patients with brain tumors. The purpose of this study was to determine the effect on BOLD fMRI accuracy of susceptibility artifacts caused by prior surgery by comparing volumes of activation in the primary motor cortex (PMC) of patients with and without prior brain surgery.
The volumes of fMRI activation of the PMC were measured for the tumor and nontumor sides in patients with (n = 13) and without (n = 30) prior neurosurgery. Statistical comparisons of the volumes were performed by using paired t tests and linear regression analysis. The location and degree of susceptibility artifact were subjectively assessed.
No significant difference was found between the mean tumor and nontumor volumes of fMRI activations in patients without prior surgery (P = .51). In patients who had prior surgery, the volume of activation was significantly smaller on the side of the prior operation when compared with the contralateral side (P = .001). The volume of activation on the side of the tumor was also significantly smaller in the patients with prior surgery compared with those without prior surgery (P < .001). Nevertheless, the PMC was identified in all cases, and its location was confirmed intraoperatively.
Prior surgery is associated with a decrease in the volume of fMRI activation in patients with prior surgery; however, by examining the T2 images, an astute radiologist can recognize this phenomenon, draw the appropriate conclusions, and correctly identify the PMC.
American Journal of Neuroradiology 10/2005; 26(8):1980-5. · 2.93 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: To prospectively evaluate magnetic resonance (MR) imaging and MR spectroscopy for depiction of local prostate cancer recurrence after external-beam radiation therapy, with step-section pathologic findings as the standard of reference.
Study received institutional approval, and written informed consent was obtained. Study was compliant with Health Insurance Portability and Accountability Act. Sextant biopsy, digital rectal examination, MR imaging, MR spectroscopy, and salvage radical prostatectomy with step-section pathologic examination were performed in nine patients with increasing prostate-specific antigen levels after external-beam radiation therapy. MR imaging criterion for tumor was a focal nodular region of reduced signal intensity at T2-weighted imaging. MR spectroscopic criteria for tumor were voxels with choline (Cho) plus creatine (Cr) to citrate (Cit) ratio ([Cho + Cr]/Cit) of at least 0.5 or voxels with detectable Cho and no Cit in the peripheral zone. Sensitivity and specificity of sextant biopsy, digital rectal examination, MR imaging, and MR spectroscopy were determined by using a prostate sextant as the unit of analysis. For feature analysis, MR imaging and MR spectroscopic findings were correlated with step-section pathologic findings.
MR imaging and MR spectroscopy showed estimated sensitivities of 68% and 77%, respectively, while sensitivities of biopsy and digital rectal examination were 48% and 16%, respectively. MR spectroscopy appears to be less specific (78%) than the other three tests, each of which had a specificity higher than 90%. MR spectroscopic feature analysis showed that a metabolically altered benign gland could be falsely identified as tumor by using MR spectroscopic criteria; further analysis of MR spectroscopic features did not lead to improved MR spectroscopic criteria for recurrent tumor.
In summary, MR imaging and MR spectroscopy may be more sensitive than sextant biopsy and digital rectal examination for sextant localization of cancer recurrence after external-beam radiation therapy.
Radiology 09/2005; 236(2):545-53. · 5.73 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: Increased levels of cyclooxygenase-2 and prostaglandin E2 (PGE2) have been observed in tobacco-related malignancies of the upper aerodigestive tract. Moreover, exposure to tobacco smoke can stimulate the synthesis of PGE2. Recent evidence suggests that urinary PGE metabolite (PGE-M) can be used as an index of systemic PGE2 production. In this study, we investigated whether levels of urinary PGE-M were increased in smokers and in patients with head and neck squamous cell carcinoma (HNSCC).
Fifty-eight HNSCC cases and 29 age- and gender-matched healthy controls were prospectively enrolled in the study. A detailed smoking history and single void urine specimen were obtained from each participant. Levels of urinary PGE-M were quantified in a blinded fashion using mass spectrometry and compared with smoking history and tumor status.
Adjusted for case-control matching, median urinary PGE-M levels were significantly higher in ever smokers (15.7 ng/mg creatinine) compared with never smokers (9.9 ng/mg creatinine) for the entire study population (n = 87, P = 0.005). Concentrations of urinary PGE-M were nearly doubled in ever smokers (15.2 ng/mg creatinine) versus never smokers (7.8 ng/mg creatinine) among healthy controls (P = 0.001). Higher PGE-M levels were observed in current versus former smokers and in those with greater pack-year exposure. A significant difference in amounts of PGE-M was not observed in patients with HNSCC versus healthy controls.
Increased levels of urinary PGE-M were observed in smokers. Urinary PGE-M may have use as a noninvasive biomarker of the effects of tobacco smoke exposure.
Clinical Cancer Research 09/2005; 11(16):6087-93. · 7.74 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: To assess the effect of commonly used computed tomographic (CT) section thicknesses on metastatic tumor measurements calculated with unidimensional, bidimensional, area, and volumetric methods.
Analysis and data collection were approved by the Institutional Review Board, with waived informed patient consent. Forty-two pulmonary metastases in 10 patients (three men and seven women; age range, 43-83 years; mean age, 65.4 years) were analyzed on CT scans obtained with 3.75-, 5.0-, and 7.5-mm section thicknesses. The lesions were automatically delineated by using a three-dimensional multicriteria segmentation algorithm. Unidimensional (the largest diameter), bidimensional (the product of the two maximal perpendicular diameters), maximal cross-sectional area, and volumetric measurements were automatically obtained for each pulmonary lesion on each section thickness. Means and variances were calculated, and the differences across the three section thicknesses for each of the four measurements were studied by using linear mixed-effects models. The Levene test was used to study the equality of variances.
Differences in the means for unidimensional, bidimensional, and area measurements were significant between a section thickness of 3.75 and 5.0 mm (unidimensional, P=.05; bidimensional, P=.05; area, P=.01) and 3.75 and 7.5 mm (unidimensional, P=.06; bidimensional, P=.03; area, P=.02), but not 5.0 and 7.5 mm. There was a significant difference in volumetric measurement as section thickness decreased from 7.5 to 5.0 mm (P <.001) and from 7.5 to 3.75 mm (P <.001). Although there was a slight trend for differences in the variances across section thickness for each measurement, none of the differences were significant.
Volumetric tumor measurements change with a reduction in section thickness from 7.5 to 5.0 and 3.75 mm. For unidimensional measurement, no change was found when thickness decreased from 7.5 to 5.0 mm.
Radiology 04/2005; 234(3):934-9. · 5.73 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: OBJECTIVE: Our aim was to determine whether pelvic CT scans reveal clinically relevant information that would change treatment in the initial or follow-up radiologic examination of patients with esophageal cancer. CONCLUSION: We observed that the addition of pelvic CT to 201 examinations of the chest and abdomen had a minimal effect on patient treatment. No pelvic examination changed the cancer stage, but three pelvic CT scans in three patients (3%) altered treatment.
American Journal of Roentgenology 03/2005; 184(2):487-90. · 2.78 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: The positive and negative predictive values are standard measures used to quantify the predictive accuracy of binary biomarkers when the outcome being predicted is also binary. When the biomarkers are instead being used to predict a failure time outcome, there is no standard way of quantifying predictive accuracy. We propose a natural extension of the traditional predictive values to accommodate censored survival data. We discuss not only quantifying predictive accuracy using these extended predictive values, but also rigorously comparing the accuracy of two biomarkers in terms of their predictive values. Using a marginal regression framework, we describe how to estimate differences in predictive accuracy and how to test whether the observed difference is statistically significant.
Statistics in Medicine 06/2004; 23(10):1555-70. · 1.88 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: Comparing the accuracy of two screening tests is ideally achieved by administering both tests as well as a gold standard test to all study subjects. In practice, a more ethical screen positive study design is often used, one that requires gold standard determination only for subjects that screen positive on either test under investigation. Although it is not possible to quantify the absolute accuracy of each test with such a design,the relative accuracy of the tests can be estimated. Since relative accuracy estimation has poor small sample properties, adjusted estimators based on adding constants to the observed data have been proposed. The adjusted estimators have the advantage that they yield point and variance estimates of relative accuracy in all settings. However, we show through both theory and numerical examples that adding constants to the data can be beneficial or detrimental to small sample performance. Furthermore, the performance of the adjusted estimator depends not only on the magnitude of the constant but also on parameters that cannot be estimated with data from a screen positive study, making selection of an optimal constant difficult in practice. We also examine the performance of the adjusted estimator using data from a study comparing the accuracy of two screening tests for cervical cancer.
Statistics in Medicine 02/2004; 23(1):21-34. · 1.88 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: The positive and negative predictive values are standard ways of quantifying predictive accuracy when both the outcome and the prognostic factor are binary. Methods for comparing the predictive values of two or more binary factors have been discussed previously (Leisenring et al., 2000, Biometrics 56, 345-351). We propose extending the standard definitions of the predictive values to accommodate prognostic factors that are measured on a continuous scale and suggest a corresponding graphical method to summarize predictive accuracy. Drawing on the work of Leisenring et al. we make use of a marginal regression framework and discuss methods for estimating these predictive value functions and their differences within this framework. The methods presented in this paper have the potential to be useful in a number of areas including the design of clinical trials and health policy analysis.
Biostatistics 02/2004; 5(1):113-27. · 2.14 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: Technologic advances give rise to new tests for detecting disease in many fields, including cancer and sexually transmitted disease. Before a new disease screening test is approved for public use, its accuracy should be shown to be better than or at least not inferior to an existing test. Standards do not yet exist for designing and analysing studies to address this issue. Established principles for the design of therapeutic studies can be adapted for studies of screening tests. In particular, drawing upon methods for superiority and non-inferiority studies of therapeutic agents, we propose that confidence intervals for the relative accuracy of dichotomous tests drive the design of comparative studies of disease screening tests. We derive sample size formulae for a variety of designs, including studies where patients undergo several tests and studies where patients receive only one of the tests under evaluation. Both cohort and case-control study designs are considered. Modifications to the confidence intervals and sample size formulae are discussed to accommodate studies where, because of the invasive nature of definitive testing, true disease status can only be obtained for subjects who are positive on one or more of the screening tests. The methods proposed are applied to a study comparing a modified pap test to the conventional pap for cervical cancer screening. The impact of error in the gold standard reference test on the design and evaluation of comparative screening test studies is also discussed.
Statistics in Medicine 04/2002; 21(6):835-52. · 1.88 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: To investigate whether the presence and degree of peritumoral vascularity may provide any diagnostic information regarding the histological subtypes of renal tumors.
Between January 2004 and March 2006, 214 patients (age, 19-94 years; 120 men, 94 women) with renal masses underwent total or partial nephrectomy and preoperative renal protocol computed tomography. Two radiologists retrospectively reviewed the computed tomographic exams in an independent and blinded fashion. The radiologists evaluated the presence of peritumoral vascularity and measured the largest vessel caliber. Fisher exact tests, multivariate logistic regression, and Wilcoxon rank-sum tests were performed for statistical analysis.
Of 219 renal tumors (median size, 3.5 cm; range, 0.9-20.0 cm) included in this study, 112 (51%) were clear cell renal cell carcinoma, 34 (16%) were papillary, 32 (15%) were chromophobe, 17 (8%) were oncocytomas, 6 (3%) were lipid-poor angiomyolipomas, and 18 (8%) were other or unclassified renal tumors. The presence of peritumoral vascularity was significantly associated with tumor size within each subtype. For both readers, peritumoral vascularity was more frequently identified in clear cell carcinomas than in papillary renal carcinomas of similar size (P = 0.019 and 0.008, respectively). For one of the readers, chromophobe carcinomas were also significantly less frequently associated with peritumoral vascularity than clear cell carcinomas of similar size (P = 0.014).
Clear cell carcinomas demonstrate peritumoral vascularity significantly more frequently than other malignant renal tumors of similar size. The presence of peritumoral vascularity may provide additional diagnostic information and improve therapeutic planning in some cases.
Journal of Computer Assisted Tomography 31(5):717-23. · 1.22 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: The objective of this study was to determine the sensitivity of the CT "whirl sign" for the diagnosis of small bowel volvulus in patients who present with a clinical suspicion of intestinal obstruction. Between January 2002 and September 2004, 1,493 CT scans performed in 1,213 patients suspected of having small bowel obstruction were retrospectively reviewed by one attending radiologist with gastrointestinal subspecialization and one senior radiology resident. Multislice helical CT scans were performed after oral and intravenous contrast administration. All CT scans showing a combined vessel and bowel whirl appearance were identified. Other features recorded included the number of degrees of whirl rotation, direction of rotation, presence of bowel obstruction, and signs of ischemia. Diagnoses were determined at either surgery or clinicoradiographic follow-up. Surgical follow-up was available in 174 of the 1,213 patients. There were 460 males and 753 females ranging in age from 1 to 95 years (mean 59 years). A whirl sign was found in 33 of the 1,493 CT scans by reader 1 and in 13 of the 1,493 CT scans by reader 2. In 11 patients, surgery revealed small bowel volvulus (0.9%). Reader 1 detected 7 of the 11 volvuli (sensitivity 64%, specificity 98%, positive predictive value 21%, negative predictive value 99.7%). Reader 2 detected 3 of the 11 volvuli (sensitivity 27%, specificity 99%, positive predictive value 23%, negative predictive value 99.5%). The CT scans of the four remaining patients with volvulus not initially recognized by either reader were re-reviewed and were felt to contain whirl signs. Most patients with small bowel volvulus can be identified on CT through detection of a whirl sign. However, most whirl signs detected on CT will not prove to be indicative of small bowel volvulus.
Journal of Computer Assisted Tomography 30(1):25-32. · 1.22 Impact Factor
-
-
[show abstract]
[hide abstract]
ABSTRACT: To examine the accuracy of the cardiothoracic ratio (CTR) at routine computed tomography (CT) of the chest to diagnose cardiomegaly by using echocardiography (ECHO) as a criterion standard, and secondarily, to search for CT cardiac measurements correlating with left ventricular hypertrophy (LVH) at ECHO.
We searched 2 databases: cancer patients who had chest CT who also underwent routine ECHO by one experienced cardiologist, from January to March 2008. Consecutive patients were enrolled who had chest radiography (CXR) within 1 month and ECHO within 6 months of CT. Cardiothoracic ratio was defined as the transverse greatest cardiac diameter from outer to outer myocardium on axial images divided by the transverse greatest thoracic diameter from inner to inner chest wall on axial images, as measured by 2 board-certified radiologists. Left ventricular short diameter on CT was measured from inner to inner myocardium. Left ventricular hypertrophy at ECHO was defined using the standard American Society of Echocardiography criteria. Other data captured included cardiac and chemotherapy history, and secondary signs of heart failure. The Pearson correlation coefficient (r) and Wilcoxon rank sum tests and receiver operating characteristic (ROC) curves were used for statistical analysis.
101 patients, 52 men and 49 women, with a mean age of 58 years met the entry criteria. Cardiothoracic ratio at CXR and CT were highly correlated (r = 0.802) (P < 0.001). There was moderate ability of CT CTR to identify LVH (area under the receiver operating characteristic curve AUC = 0.70; 95% CI, 0.51-0.90). The CT left ventricular short diameter showed moderate correlation with the ECHO left ventricular internal diameter (r = 0.49) and left ventricular mass (r = 0.37).
In patients with cancer undergoing routine ECHO, the cardiothoracic ratio at routine CT scans was highly correlated with that at CXR. Preliminary estimates in this small study indicate a low likelihood of LVH when the CTR is less than 0.49.
Journal of computer assisted tomography 36(1):67-71. · 1.38 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: To determine the computed tomographic imaging features predictive of adnexal malignancy and evaluate the accuracy of contrast-enhanced helical computed tomography (CT) in the characterization of adnexal masses.
The institutional review board waived the informed consent requirement for this retrospective study, which was Health Insurance Portability and Accountability Act compliant. For 143 consecutive patients who underwent preoperative contrast-enhanced CT of the abdomen and pelvis and had adnexal masses found at surgical pathology, preoperative contrast-enhanced computed tomographic scans were retrospectively and independently reviewed by 2 radiologists. Receiver operating characteristic analysis was used to assess the value of contrast-enhanced CT in detecting and characterizing adnexal masses. Feature analysis was performed to select the findings with the highest sensitivities and specificities for predicting malignant lesions. Interobserver variability was assessed using kappa statistics.
At surgical pathology, 234 adnexal masses were found (165 were malignant). Readers 1 and 2 detected 215 and 216 (92%) adnexal masses, respectively. For both readers, the features most predictive of malignancy were heterogeneity for a solid lesion, multilocularity (>3 locules), irregular and thickened cystic septations or walls, and the presence of internal vegetations for a cystic lesion. Irregular lesion contour and ancillary findings, including ascites, peritoneal implants, lymphadenopathy, and pleural effusion, were predictive of malignancy in both solid and cystic lesions. In diagnosing malignancy in all patients and in the subgroup without ancillary findings, areas under the receiver operating characteristic curves were 0.88 (95% confidence interval [CI], 0.84-0.93) and 0.89 (95% CI, 0.83-0.96) for reader 1, respectively, and 0.90 (95% CI, 0.86-0.94) and 0.88 (95% CI, 0.80-0.97) for reader 2, respectively. Interobserver agreement (kappa = 0.71) was good.
Contrast-enhanced helical CT is highly accurate in characterizing adnexal masses as malignant. Recognition of the computed tomographic features most often associated with adnexal malignancy will assist in more confident use of this modality and may potentially obviate the need for additional imaging studies before treatment selection.
Journal of computer assisted tomography 32(4):533-40. · 1.38 Impact Factor
