Our patient is a 61-year-old Caucasian man who originally presented with urinary obstructive symptoms and a serum prostate specific antigen (PSA) level of 1.5 ng/mL six years before prostatectomy. The obstructive process was treated conservatively, with observation and symptomatics. Over the following three years, PSA of our patient continued to rise, prompting multiple sets of prostatic biopsies that were negative for carcinoma. Ultrasound showed a moderately enlarged nodular prostate consistent with benign prostatic hyperplasia. A few weeks before prostatectomy, his serum PSA increased up to 4.7 ng/mL, at which time additional core biopsies were performed. A single focus of Gleason grade 3+3 (score = 6) prostatic adenocarcinoma was identified, involving 5% of only one of the tissue cores submitted. Our patient was subject to undergo a radical prostatectomy. A conventional suprapubic, radical prostatectomy on our patient was performed. The prostate was submitted for whole-mount processing. Gross examination revealed a 32 gm prostate measuring 4.2 cm apex-base, 4 cm wide, and 3.5 cm antero-posterior. The attached seminal vesicles were up to 2.4 cm in length and 0.4 cm average diameter. Also submitted were dissections of the right and left pelvic lymph nodes. The prostate was fixed in formalin, serially sectioned in 5 mm intervals from apex to the base and paraffin embedded. Histological sections, 5 μm-thick, were cut from the anterior face of each paraffin block, mounted on 2 × 3-inch glass slides and stained with conventional hematoxylin and eosin. Immunoperoxidase staining specific for chromogranin A, neuron specific enolase, synaptophysin, pan-cytokeratin and PSA, and a special combined staining for racemase [α-methyl CoA] antigen and p63 antigen were performed.
Digital microscopic pictures were obtained during different magnifications using an Olympus BX51 microscope (Olympus, Japan) and a Macrofire digital camera (Optronics, Goleta, CA) connected though a firewire to a Dell Optiplex 745 desktop computer (Dell, USA). Low power images from whole mount slides were digitally scanned using an Aperio ScanScope Model T3 (Aperio Technologies, Inc., Vista, CA, USA) and captured using the ImageScope software version 9.x (Aperio Technologies, Inc., Vista, CA, USA). All images were optimized using Adobe Photoshop CS2 software (San Diego, CA).
A single focus of invasive, variably spaced, small atypical acinar glands without basal cells was identified in the periphery of the mid-left, posterior quadrant of the prostate, involving less than 5% of the prostatic tissue (Figure ). The cytologic features were consistent with prostatic adenocarcinoma and included enlarged and overlapping nuclei with prominent and peripherally marginated nucleoli (Figures and ). The tumor from our patient was organ confined, without angiolymphatic, perineural, extracapsular or seminal vesicle invasion, and with negative surgical margins. Additional histopathology included focal high grade prostatic intraepithelial neoplasia, benign prostatic hyperplasia, focal acute and chronic inflammation, focal glandular atrophy and basal cell hyperplasia. The two right and two left pelvic lymph nodes were without evidence of malignancy.
Figure 1 Location of (A) the carcinoid tumor in the verumontanum of the prostatic urethra; and (B) the prostatic adenocarcinoma within the prostate at the right posterior lobe of the prostate. These consecutive whole prostatectomy slices were sectioned 5 mm apart (more ...)
(A) Carcinoid tumor (40× objective), (B) and C) prostatic adenocarcinoma Gleason grade 3 + 3 (sum = 6) (10× & 40× objectives). Hematoxylin-eosin staining.
Approximately 17 mm from the invasive adenocarcinoma, within the verumontanum of the prostatic urethra of our patient, there was a 3 mm maximal dimension tumor with markedly different morphology. From low microscopic power, the tumor of our patient appeared well circumscribed with pushing, rather than invasive, margins (Figure ). Higher power examination revealed cells forming nested, acinar and trabecular architecture with small uniform nuclei, a granular "salt and pepper" chromatin pattern, and moderate amounts of granular cytoplasm (Figure ). Mitotic figures, necrosis, and angiolymphatic invasion were not identified. Immunoperoxidase staining for chromogranin A, neuron specific enolase, and synaptophysin were positive in nearly 100% of the urethral tumor cells, showing a granular cytoplasmic distribution (Figure ). These cells were negative for pan-cytokeratin and PSA. Immunoperoxidase staining for PSA was positive in the prostatic adenocarcinoma cells, which also were positive for racemase (α-methyl CoA) antigen and negative for p63 antigen. Based on this pattern of reactivity, we concluded that the two tumors were distinct and a diagnosis of carcinoid tumor was made on the urethral tumor of our patient. An electron microscopy was attempted on our patient for neurosecretory granules in the urethral tumor, but it was unsuccessful due to formalin fixation artifact.
Carcinoid tumor of the veramontanum (colliculus seminalis) of the prostatic urethra, hematoxylin-eosin staining (2× objective).
Immunoperoxidase staining for chromogranin A (20×) (A), neuron specific enolase (B) and synaptophysin (C) of the carcinoid tumor cells (40× objective).