The Prostate 28359-363 (I996) Microvascular Invasion of the Seminal Vesicles in Adenocarcinoma of the Prostate Sam D. Graham, Jr., Pave1 Napalkov, Lynda Watts, Diva Salomao, and David G. Bostwick Department of Surgery Division of Urology, Emory University School of Medicine, Atlanta, Georgia (S.D.G., P.N., L W.); Department of Pathology, Mayo Clinic, Rochester, Minnesota (D.S., D.G.B.) ABSTRACT: The objective of this article is to determine the relationship between microvascular invasion and seminal vesicle invasion in prostatic adenocarcinoma.Radical prostatectomies with seminal vesicle involvement were examined histologically and immunohistochemicallywith antibodies directed against S-100protein and factor WI. Microvascular invasion of the seminal vesicles showed a positive correlation with microvascular and capsular invasion of the prostate (P= 0.006 and 0.048,respectively)and lymph node metastases. Tumor progression was found in 8 of 14 (57%) patients with microvascular invasion of the seminal vesicles, compared with 3 of 22 (14%) without microvascular invasion (P=O.OOl). Microvascular invasion of the seminal vesicles is predictive of tumor progression and lymph node metastases in prostatic adenocarcinoma. 0 1996 Wiley-Liss, Inc. KEY WORDS: Microvascular invasion, seminal vesicle, radical prostatectomy INTRODUCTION The prognosis of prostatic adenocarcinoma is determined by the extent of tumor at radical prostatectomy [1,2]. Seminal vesicle invasion carries a risk of recurrence and/or progression of more than 50% [3-51 and usually appears either as tumor surrounding the seminal vesicles or as stromal and muscular invasion [6,7. The tissue around the seminal vesicles is predominantly fat with a rich anastomosing network of small vessels . This study was performed to determine whether seminal vesicle invasion is chiefly due to microvascular invasion and to investigate the relationship between microvascular or perineural invasion in the prostate and microvascular invasion of the periseminal vesicular tissue. METHODS Specimen Preparation Prostates from radical prostatectomies were painted with India ink and Bouin's solution upon receipt and then fixed in formalin. Sampling of the prostate included sections of the distal urethral margin; then, representative sections were taken every 0 1996 Wiley-Liss, Inc. 2-3 mm, on both the right and left sides. Sections were also taken of any suspiciously area. All specimens were examined by a single reference pathologist. Seminal vesicle invasion was defined as the presence of adenocarcinoma within the muscular wall of the seminal vesicle. Microvascular invasion of the seminal vesicles included foci within the seminal vesicles and the adjacent periseminal vesicular soft tissue, which is in intimate association. case selection The study group consisted of 36 patients with prostatic adenocarcinoma involving the seminal vesicles with a minimum of 2 years of follow-up selected from 201 radical prostatectomy specimens for clinically localized cancer performed at Emory University between 1979 and 1991. Patient medical records were reviewed for evidence of tumor progression and re- Received for publication November 18, 1994; accepted May 17, 1995. Address reprint requests to Dr. Sam D. Graham, Jr., 1365 Clifton Rd., NE, Atlanta, GA 30320. 360 Grahamet al. TABLE 1. Patterns of Invasion of Prostatic AdenouKinorna in the Prostate and Seminal Vesicles Prostate N % Microvascular and perineural 10 28 Microvascular only 0 0 Perineural only 23 64 Fibromuscular stromal only 3 8 Total 36 loo Seminal vesicle N % 9 5 18 25 14 50 11 100 4 36 currence based upon followup digital rectal examination (DRE), serum prostate-specific antigen (PSA) level, radionuclide bone scan, and/or repeat biopsies when indicated. Fig. I. Microvascular invasion (arrowhead) of prostatic adenocarcinoma in the seminal vesicle (immunoperoxidase staining with antibody to factor VIII. x 100). lmmunohistochemistry Tissue sections were obtained from formali-fixed paraffin-embedded blocks of the prostate and seminal vesicles, cut at 7 pm, and placed on silanized or poly-l-lysine coated slides. Routine indirect immunoperoxidase staining was performed on each specimen. Briefly, sections were deparaffinized in xylene and rehydrated in alcohol. Endogenous peroxidase activity was quenched with 3% hydrogen peroxide, and the sections were incubated in stable 0.1% pepsin solution with normal serum. Slides were incubated with primary mouse monoclonal antibody (S-100, Dako, Carpenteria, CA; or factor VIII-related antigen Biogenix, San Ramon, CA) and processed with biotinylated streptavidin kit (Dako). Aminoethylcarbazole or diaminobenzidine were used as chromogenic peroxidase substrates; the slides were counterstained with Mayer's hematoxylin. Known positive and negative controls for each antibody were run in parallel with each staining and gave appropriate results. cular arteries, arterioles, venules, and small capillarylike spaces. The capillary-like structures were uniformly distributed throughout the smooth muscle layer and in the mucosal folds of the normal seminal vesicle, and represent either lymphatic channels, capillaries, and venules. In some sections, the number of immunoreactive capillary-like structures was increased in areas with malignant glands in comparison with normal tissue, although this was not quantified. Microvascular invasion was defined as transmural invasion of vascular structures by malignant glands (Fig. 1). Frequently, immunoreactive spaces were compressed by adjacent malignant glands; when there was no definite luminal invasion, this was not considered evidence of invasion (Fig. 2). S-100 protein immunoreactivity was identified in large and small nerves throughout the prostate. Perineural invasion varied from rare neoplastic glands abutting nerves to massive circumferential and intraneural invasion (Fig. 3). RESULTS Seminal Vesicle Involvement Seminal vesicle involvement was microvascular in 14 cases (39%),perineural in 27 (82%),and confined to the fibromuscular stroma in 4 (11%).Nine cases (25%)showed both perineural and microvascular invasion, and 23 (64%)had stromal invasion associated with either microvascular (5)or perineural (18) invasion (Table I). lmmunohistochemical Studies There was intense cytoplasmic reactivity for factor VIII-related antigen in endothelial cells of large mus- Correlation of Seminal Vesicle and Prostatic Microvascular Invasion Ten patients (28%) had microvascular invasion of the prostate; 8 of these had concurrent microvascular invasion of the seminal vesicles (Table 11, P = 0.003); 33 of the 36 patients (92%) in the study group had perineural invasion in the prostate confirmed by S-100 staining, and 27 (82%) also had perineural invasion of the seminal vesicles (Table 111, P = 0.01). Of 10 patients who had both perineural and microvascular invasion of the prostate, 5 had both in the seminal vesicles. Microvascular Invasion of Seminal Vesicles in Prostatic Adenocarcinoma 36I TABLE II. Relationship of Microvascular Invasion in the Prostate to Microvascular Invasion in the Seminal Vesicles in Adenocarcinoma of the Prostate* Prostate Positive Negative Total Seminal vesicles N % N % N % Positive Negative Total 8 2 10 80 20 100 6 20 26 23.1 76.9 14 22 36 38.9 61.1 100 100 *P= 0.0029; Fisher's exact test. Fig. 2. lmmunoreactive microvascular space in the seminal vesicle compressed by the malignant glands of prostatic adenocarcinoma (immunoperoxidase staining with antibody to kctor VIII, x 100.) TABLE 111. Relationship of Perineural Invasion in the Prostate to Perineural Invasion in the Seminal Vesicles in Adenocarcinoma of the Prostate* Prostate Positive Negative Total Seminalvesicles N % N % N % Positive Negative 27 6 33 81.8 18.2 100 0 3 3 0 100 100 27 9 36 75 25 100 Total *P= 0.0117; Fisher's exact test. Fig. 3. Perineural invasion (arrowheads) of prostatic adenocarcinoma in the seminal vesicle (immunoperoxidase staining with antibody to S-I OO, X I OO). Correlation of Microvascular Invasion and Tumor Progression Eight of the 36 (22%) patients with seminal vesicle involvement had lymph node metastases at radical prostatectomy. Of these, 5 (63%)had microvascular involvement of the seminal vesicle, 4 (50%) had microvascular invasion of the prostate, and l had neither microvascular nor perineural invasion. In patients with microvascular invasion of the seminal vesicles, 5 (36%) had lymph node metastases, compared with 8 (22%)in the entire study group. Median follow-up was 43 months, with a range of 24-136 months. 11 patients experienced tumor progression as defined by elevation of PSA >0.5 n g / d (n = ll), positive DRE (n = 2), abnormal radionuclide bone scan (n=6), and/or positive biopsy (n=4). Of the 17patients who progressed, 8 had only microvas- cular invasion of the seminal vesicle, 6 had both microvascular and perineural invasion of the seminal vesicle, and 5 had lymph node metastases. Tumor progression was found in 8 of 14 (57%)patients with microvascular invasion compared with 3 of 22 (14%) without microvascular invasion of the seminal vesicles (Fig. 4). Chi-square analysis of variance was used to test the predictive value of microvascular invasion in the seminal vesicle, the combination of microvascular and perineural invasion in the seminal vesicle, and lymph node metastases for tumor progression. In all groups, the predictive values were significant (P=O.O09, P=O.OlO, and P=O.O40, respectively). Microvascular invasion in the seminal vesicles alone was the best single predictor (P= 0.03, standard error (SE) = 0.866), as compared with lymph node metastases (P= 0.02, SE = 0.956). The combination of microvascular invasion and perineural invasion of the seminal vesicles was a stronger predictor of tumor progression than lymph node metastases (P= 0.01, SE = 0.934 and P=0.05, SE -0.934, respectively), but its predictive value was not as sigruficant as that of microvascular invasion alone (Fig. 4). No correlation was found between perineural invasion alone and tumor 362 Graham et al. Fig. 4. Correlation between progression rate and invasion of the seminal vesicles. There was a significant correlation between progression rate compared to microvascular invasion (P=O.Ol). microvaxular and perineural invasion (P=O.OI), and lymph node metastases (P=O.OI). Of 36 patients in the study group, I I progressed and 25 were free of tumor (median follow-up, 43 months). All P-values based on Fisher’s exact test. progression, probably because perineural invasion of the prostate and seminal vesicles were found in almost all of the patients in this study. DISCUSSION Seminal vesicle invasion is one of the most sigruficant predictors of failure of radical prostatectomy in prostatic adenocarcinoma, accounting for progression rates of 51-64% [4,5,9]. It usually involves the fibromuscular stroma or the adjacent soft tissue [6,7l, which are rich in microvascular channels and may be a pathway for tumor spread to the obturator lymph nodes [ l q . We found a high incidence of microvascular and perineural invasion in the seminal vesicles, similar to that seen in the prostate, and there was a positive trend with nodal metastases (63% vs. 32%). Five cases showed stromal invasion and microvascular invasion of the seminal vesicles, but we could not determine whether the microvascular invasion was secondary to stromal invasion. Since the seminal vesicles were step sectioned at 3mm levels, our findings may underestimate the incidence of microvascular and perineural invasion due to sampling error. Microvascular invasion by tumors in other organs is predictive of tumor progression and lower cancerspecific survival, including transitional cell carcinoma of the bladder and testicular carcinoma, probably due to access to the systemic vasculature for metastasis [lo-121. If tumor cells are to gain access to the systemic lymphatics, as is currently evaluated surgically by obturator node sampling, microvascular invasion of the seminal vesicle may represent the first level of lymphatic invasion. Perineural invasion may also play a role in metastatic spread [13,14]. Vessels and nerves are closely aligned anatomically, and it is likely that there is crossover invasion. In mapping studies, McNeal and colleagues [7,15,16] showed that the neurovascular pedicles of the prostate at the apex and base provide paths of egress for carcinoma to the periprostatic soft tissues; however, these investigators did not evaluate the incidence of microvascular invasion. Rare studies have looked at the intraprostatic or periprostatic microlymphatic architecture [17-191. Using light and electron microscopy, Furusato and Mostofi  demonstrated lymphatics in the stroma and small capillaries in the stroma and perineural spaces. We identified factor VIII-related antigen-immunoreactive capillary-like structures in the stroma of the prostate, seminal vesicles, and periseminal vesicular fat. In most of the sections of the prostate and seminal vesicles stained for factor VIII-related antigen, the number of reactive capillaries appeared to be increased in foci of invasive carcinoma, as compared with the adjacent benign tissue, indicative of tumor neo-angiogenesis. Recent studies indicate that angiogenesis plays a role in prostate cancer [21,22]. Our findings indicate that microvascular invasion of the seminal vesicles by prostatic adenocarcinoma is predictive of tumor progression. Perineural invasion of the prostate was present in virtually all cases, limiting its usefulness. Patients with microvascular involvement of the prostate had similar involvement in the seminal vesicles; this correlated with tumor progression. We found a sigruficant positive correlation between lymph node metastases, microvascularinvasion, and microvascular with perineural invasion and tumor progression; microvascular invasion in the seminal vesicles was the strongest predictor of progression. The freely anastomosing microvascular channels in the fibromuscular stroma and fat adjacent to the seminal vesicles may represent a sigruficant pathway by which prostatic adenocarcinoma disseminates. Identification of microvascular and perineural invasion at this site appears to be prognostically useful. ACKNOWLEDGMENTS The authors acknowledge the help of Helen M. Hamson, Division of Urology, Emory University, and of Ted Gansler, Department of Pathology, Emory University, for their assistance in this paper. REFERENCES 1. Middleton RG, Smith JAJr, Melzer RB: Patient survival and local recurrence rate following radical prostatectomy for prostatic carcinoma. J Urol 136:422-424, 1986. Microvascular Invasion of Seminal Vesicles in Prostatic Adenocarcinoma 2. Anscher MS, Prosnitz LR: Multivariate analysis of factors predicting local relapse after radical prostatectomy-possible indications for postoperative radiotherapy. Int J Radiat Oncol Biol Phys 21:941-947, 1991. 3. Hering F, Rist M, Roth J, Mihatsch M, Rutishauser G: Does microinvasion of the capsule and/or micrometastases in regional lymph nodes influence disease-free survival after radical prostatectomy? Br J Urol 66:177181, 1990. 4. Mukamel E, deKernion JB, Dorey F, Hannah J: Significance of histological prognostic indicators in patients with carcinoma of the prostate. Br J Urol 6546-50, 1990. 5. Stein A, deKernion JB, Smith RB, Dorey F, Pate1 H: Prostate specific antigen levels after radical prostatectomy in patients with organ confined and locally extensive prostate cancer. J Urol 147942-946, 1992. 6. Jewett HL, Eggleston JC, Yawn DH: Radical prostatectomy in the management of carcinoma of the prostate: Probable causes of some therapeutic failures. J UrollO7: 1034-1040, 1972. 7. Villers AA, McNeal JE, Redwine EA, Freiha FS, Stamey TA: Pathogenesis and biological sigruficance of seminal vesicle invasion in prostatic adenocarcinoma. J Urol 143~1183-1187,1990. 8. Setchell BP, Brooks DE: Anatomy, vasculature, innervation and fluids of the male reproductive tract. In Knobit E, Neill J (eds): "The Physiology of Reproduction." New York: Raven Press, pp 753-836. 9. Elder JS, Jewett HJ,Walsh PS: Radical perineal prostatectomy for clinical state B2 carcinoma of the prostate. J Ur01 127704-706, 1982. 10. Reuter VE: Pathology of bladder cancer: Assessment of prognostic variables and response to therapy. Semin Oncol 17524-532, 1990. 11. Javadpour N, Young JD Jr: Prognostic factors in non- 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 363 seminomatous testicular cancer. J Urol 135:497-499, 1986. Sesterhenn IA, Falkson G, Rivkind SE, Vogelzang NJ: Prognosis and other clinical correlates a pathologic review in stage I and I1 testicular carcinoma: A report from the Testicular Cancer Intergroup Study. J Clin OnC O ~10:69-78, 1992. Rodin AE, Larson LD, Roberts DK: Nature of perineural spaces invaded by prostatic carcinoma. Cancer 20: 1772-1779, 1967. Mostofi FK, Sesterhenn IA, Davis CJ: Prostatic carcinoma: Problems in the interpretation of prostatic biopsies. Hum Pathol 23223-241, 1992. McNeal JE, Bostwick DG, Kindrachuk RA, Redwine EA, Freiha FS, Stamey TA: Patterns of progression in prostate cancer. Lancet 1:60-63, 1986. McNeal JE, Villers AA, Redwine EA, Freiha FS, Stamey TA Capsular penetration in prostate cancer. Significance for natural history and treatment. Am J Surg Pathol14:240-247, 1990. Smith MJV: The lymphatics of the prostate. Invest Urol 3:439-444, 1966. Connolly JG, Thomson A, Jewett MAS, Hartman N, Webber M: Intraprostatic lymphatics. Invest Urol 5~371-378,1968. Raghavaiah NV, Jordan WP Jr: Prostatic lyphography. J Urol 121:178-181, 1979. Furusato M, Mostofi FK: Intraprostatic lymphatics in man: Light and ultrastructural observations. Prostate 1~15-23, 1980. Brawer MK, Bigler SA, Deering RE: Quantitative morphometric analysis of the microcirculation in prostate carcinoma. J Cell Biochem Suppl16H:64-64, 1992. Bigler SA, Deering RE, Brawer MK: Comparison of microscopic vascularity in benign and malignant prostate tissue. Hum Pathol 24220-226, 1993.