Harold Tsai

EFFECT OF NEOADJUVANT HORMONAL THERAPY ON PROSTATIC INTRAEPITHELIAL NEOPLASIA AND ITS PROGNOSTIC SIGNIFICANCE

PURPOSE We studied the effect of neoadjuvant hormonal therapy on prostatic intraepithelial neoplasia in patients undergoing radical prostatectomy and assessed the effect of prostatic intraepithelial neoplasia on disease recurrence as measured by serum prostate specific antigen (PSA). MATERIALS AND METHODS A total of 278 patients with clinically localized prostate cancer were included in phase II and III studies evaluating radical prostatectomy alone versus radical prostatectomy following neoadjuvant hormonal therapy at Memorial Sloan-Kettering Cancer Center between October 1991 and August 1996. Patient data related to prostatic intraepithelial neoplasia were analyzed. RESULTS Of 275 evaluable patients 145 (52.7%) had prostatic intraepithelial neoplasia. Of 50 patients treated with neoadjuvant hormonal therapy (hormone group) 22 (44%) had a lower incidence of prostatic intraepithelial neoplasia compared to 69 of 80 controls (86.3%) (chi-square test p<0.0001). Of 262 patients (95.3%) with followup PSA 44 (16.8%) had PSA recurrence at a median followup of 32 months, with a median time to recurrence of 30 months. PSA recurrence was noted in 23 of 145 patients with compared to 21 of 130 without prostatic intraepithelial neoplasia (chi-square test p = 0.95), and did not significantly differ between the hormone group (25 of 142, 17.6%) and controls (19 of 130, 14.6%) (chi-square test p = 0.45). CONCLUSIONS While patients treated with neoadjuvant hormonal therapy had significantly lower incidence of prostatic intraepithelial neoplasia, neither prostatic intraepithelial neoplasia nor neoadjuvant hormonal therapy significantly affected PSA recurrence at a median followup of 32 months.

ROLE OF ADVANCED 2 AND 3-DIMENSIONAL ULTRASOUND FOR DETECTING PROSTATE CANCER

PURPOSE We explored the clinical usefulness of spectrum analysis and neural networks for classifying prostate tissue and identifying prostate cancer in patients undergoing transrectal ultrasound for diagnostic or therapeutic reasons. MATERIALS AND METHODS Data on a cohort of 215 patients who underwent transrectal ultrasound guided prostate biopsies at Memorial-Sloan Kettering Cancer Center, New York, New York were included in this study. Radio frequency data necessary for 2 and 3-dimensional (D) computer reconstruction of the prostate were digitally recorded at transrectal ultrasound and prostate biopsy. The data were spectrally processed and 2-D tissue typing images were generated based on a pre-trained neural network classification. We used manually masked 2-D tissue images as building blocks for generating 3-D tissue images and the images were tissue type color coded using custom software. Radio frequency data on the study cohort were analyzed for cancer probability using the data set pre-trained by neural network methods and compared with conventional B-mode imaging. ROC curves were generated for the 2 methods using biopsy results as the gold standard. RESULTS The mean area under the ROC curve plus or minus SEM for detecting prostate cancer for the conventional B-mode and neural network methods was 0.66 +/- 0.03 and 0.80 +/- 0.05, respectively. Sensitivity and specificity for detecting prostate cancer by the neural network method were significantly increased compared with conventional B-mode imaging. In addition, the 2 and 3-D prostate images provided excellent visual identification of areas with a higher likelihood of cancer. CONCLUSIONS Spectrum analysis could significantly improve the detection and evaluation of prostate cancer. Routine real-time application of spectrum analysis may significantly decrease the number of false-negative biopsies and improve the detection of prostate cancer at transrectal ultrasound guided prostate biopsy. It may also provide improved identification of prostate cancer foci during therapeutic intervention, such as brachytherapy, external beam radiotherapy or cryotherapy. In addition, 2 and 3-D images with prostate cancer foci specifically identified can help surgical planning and may in the distant future be an additional reliable noninvasive method of selecting patients for prostate biopsy.

Two-dimensional and three-dimensional tissue-type imaging of the prostate based on ultrasonic spectrum analysis and neural network classification

Spectrum analysis of ultrasonic radio-frequency echo signals has proven to be an effective means of characterizing tissues of the eye and liver, thrombi, plaque, etc. Such characterization can be of value in detecting, differentiating, and monitoring disease. In some clinical applications, linear methods of tissue classification cannot adequately differentiate among the various manifestations of cancerous and non-cancerous tissue; in these cases, non-linear methods, such as neural-networks, are required for tissue typing. Combining spectrum-analysis methods for quantitatively characterizing tissue properties with neural-network methods for classifying tissue, a powerful new means of guiding biopsies, targeting therapy, and monitoring treatment may be available. Current studies are investigating potential applications of these methods that use novel tissue-typing images presented in two and three dimensions. Results to date show significant sensitivity improvements of possible benefit in cancer detection and effective tissue-type imaging that promise improved means of planning and monitoring treatment of prostate cancer.

2-D images for biopsy guidance and 3-D images for treatment planning and monitoring of prostate cancer based upon spectrum analysis and neural-network classification

Spectrum analysis of ultrasonic echo signals has been showing potential for distinguishing cancerous from non-cancerous prostate tissues. Recently, using neural networks to classify tissue from spectrum analysis results has provided a powerful basis for imaging, guiding biopsies, and planning, executing, and monitoring therapy. ROC curves derived from leave-one-out evaluations of neural-network classifier performance have an area of 0.87/spl plusmn/0.04 compared to an area of 0.64/spl plusmn/0.04 for B-mode methods, which implies significantly superior differentiation of cancerous from non-cancerous prostate tissue.