- Email: [email protected]
The Extent and Zonal Location of Prostatic Intraepithelial Neoplasia and Atypical Adenomatous Hyperplasia: Relationship with Carcinoma in Radical Prostatectomy Specimens
Path. Res. Pract. 191,860-867 (1995)
The Extent and Zonal Location of Prostatic Intraepithelial Neoplasia and Atypical Adenomatous Hyperplasia: Relationship with Carcinoma in Radical Prostatectomy Specimens ::Junqi Qian and D. G. Bostwick Department of Pathology and Laboratory Medicine, Mayo Clinic, Rochester, MN, U.S.A.
SUMMARY High-grade prostatic intraepithelial neoplasia (PIN) and atypical adenomatous hyperplasia (AAH) are considered putative precursors of prostatic adenocarcinoma. We determined the extent and zonal distribution of PIN and AAH in totally-embedded radical prostatectomies with prostate cancer, including 195 cases with PIN and 217 with AAH. PIN was identified in 86% of the cases. The mean volume of PIN was 1.32 cc (range, 0-8.12 cc), and was greater for PIN within 2 mm of cancer (mean, 1.0 cc) than for PIN more than 2 mm from cancer (mean, 0.3 cc). PIN was usually multicentric (64.5% of cases) and located in the non-transition zone (63%) or all zones (36%). The volume of PIN was positively correlated with the volume of cancer, patient age, pathologic stage and Gleason score. AAH was identified in 23.0% of the cases, and was more frequent in the transition zone (19.8% of cases) than in the nontransition (peripheral and central) zone (6.0%). The number of foci of AAH in the transition zone was always greater than that in the non-transition zone. AAH was frequently multicentric (46% of cases), especially in the transition zone (47% of transition zone cases) compared with the non-transition zone (23% of non-transition zone cases). The mean volume of AAH was 0.029 cc (range, 0-1.29 cc), and was much higher in the transition zone than in the non-transition zone. AAH was more common in older patients and those with greater prostatic weight, higher prostatic volume, greater percent of nodular hyperplasia, greater volume of cancer, greater percent of Gleason patterns 4 and 5 cancer, higher volume of prostatic intraepithelial neoplasia and higher serum prostate specific antigen concentration. Our results indicate that the extent and zonal distribution of high grade PIN and carcinoma are strongly associated, and that PIN is frequently multicentric; this supports the hypothesis that PIN is a premalignant lesion. AAH and carcinoma show a weak but significant association; if AAH is a premalignant lesion, it probably is associated with a subset of cancers arising in the transition zone.
*Supported in part by National Institutes of Health Grant CA58225-01C. 0344-0338/95/0191-0860$3.50/0
© 1995 by Gustav Fischer Verlag, Stuttgart
PIN and AAH in prostate cancer' 861
Introduction High grade prostatic intraepithelial neoplasia (PIN) and atypical adenomatous hyperplasia (AAH) represent putative precursors of prostatic adenocarcinoma, although the evidence for PIN is stronger l - 18 . Two grades of PIN are recognized (low grade and high grade), and high grade PIN is considered the most likely precursor of invasive carcinoma 19 • The clinical importance of recognizing high grade PIN is based on its strong association with carcinoma 2 - 6 ; its identification in biopsy specimens of the prostate warrants further search for concurrent invasive carcinoma 20 - 23 • The clinical importance of AAH is uncertain, although some data suggest that it is a precursor to adenocarcinoma arising in the transition zone 9 - 11 , 24, 25. Reports to date have provided limited information concerning the quantity of PIN and AAH and their relationship with other clinical and pathologic features of prostate cancer2, 6, 7, 26, 27. In this study, we determined the extent and zonal location of high grade PIN and AAH in a large series of radical retropubic prostatectomy specimens with cancer, and correlated the results with a variety of features, including patient age, volume of cancer, pathologic stage and tumor grade. Results Patients ranged in age from 45 to 77 years (mean 64.4 years). The prostates weighed from 21 to 185 grams (mean, 52.6 grams), and volumes ranged from 8.1 to 189.0 cc (mean, 38.8 cc). Clinical stages included TlaNOMO (2 cases [1 %]), TlbNOMO (4 cases [2%]), TlcNOMO (35 cases [16%]), T2a + bNOMO (48 cases [22%]), T2cNOMO (111 cases [51 %]), T3a + bNOMO (10 cases [5%]) and T3cNOMO (7 cases [3%]). Patho-
logic stages included Tla + bNOMO (6 cases [3%]), T2a + bNOMO (13 cases [6%]), T2cNOMO (116 cases [54%]), T3a + bNOMO (49 cases [22%]), T3cNOMO (21 cases [10%]) and TxNIMO (12 cases [5%]).
Extent and Zonal Location of High Grade PIN PIN was identified in 167 of 195 cases (86%) (Fig. 1). The total mean volume of PIN was 1.32 cc (standard error, 0.10; range 0-8.12 cc), and was greater for PIN within 2 mm of cancer (mean, 1.0 cc; standard error, 0.1 cc; range, 0 - 7.8 cc) than for PIN more than 2 mm from cancer (mean, 0.33 cc; standard error, 0.33 cc; range 0-3.2 cc; R = 0.30, P = 0.00001) (Table 1). High grade PIN was identified in the transition zone in 61 cases (31 % of all cases) and in the non-transition zone (peripheral and central zones) in 166 cases (85%); only 1 case of PIN was exclusively in the transition zone (1.0% of all cases), compared with 106 cases of PIN exclusively in the non-transition zone (54%); 60 other cases of PIN were found in both zones (31 %) (Fig. 2), and 28 cases had no PIN (14%). PIN was multicentric in 72 % of all cases, including 7% multicentricity in the transition zone and 63 % in the non-transition zone; in the remaining 2 % of the cases, multicentricity was characterized by concomitant single foci in the transition zone and non-transition zone. The following number of foci of PIN were found in the transition zone: 0 foci (134 cases; 69%), 1 focus (47 cases; 24%), 2 foci (11 cases; 5.5%), 3 foci (1 case; 0.5%) and 4 foci (2 cases; 1.0%). The following number of foci of PIN were found in the non-transition zone: 0 foci (28 cases; 14%), 1 focus (42 cases; 21.5%), 2 foci (60 cases; 31%), 3 foci (49 cases; 25%), 4 foci (13 cases; 7%) and 5 foci (3 cases; 1.5%) (Fig. 3).
Table 1. Comparison of prostatic intra epithelial neoplasia and atypical adenomatous hyperplasia in a series of totally embedded prostatectomies with cancer
Prevalence (All sites, % of all cases) Prevalence, transition zone only Prevalence, non-transition zone only Prevalence, both zones simultaneously Volume (mean) (range) Multicentricity (% of cases) Multicentricity, transition zone Multicentricity, non-transition zone Correlation with patient age Correlation with prostate weight Correlation with cancer volume Correlation with pathologic stage Correlation with Gleason score < 2 mm from cancer > 2 mm from cancer
High grade PIN
AAH
86% 1% 63% 36% 1.32 cc. (0- 8.12) 72% 7% 63% S" * S" S* 'f" S'·'f" S* S'·"
23% 74% 14% 12% 0.029 cc. 46% 47% 23% S"':· *
S"
-: Not significant; S: significant; *P < 0.01; 'f *P < 0.001; ,. * "P < 0.0001. Note: Correlations for PIN are with volume of PIN; correlations for AAH are with number of foci of AAH.
862 . Junqi Qian and D. G. Bostwick
Fig. 1. Micropapillary Pattern of high grade PIN.
Fig. 2. Whole-mount section of the prostate showing two foci of PIN arising in the transition zone and peripheral zone (nontransition zone) (solid lines outline cancer; dotted lines outline PIN). 100
Zone Transition III Nontransition ~
80
£' 0
69
60
Q)
III
40
20 0
o 0
2
3
Foci of PIN (no.)
4
1.5
5
Fig. 3. Multicentricity of PIN according to zonal location.
PIN and AAH in prostate cancer . 863
Correlation of Volume of PIN with Patient Age and Prostate Weight The total volume of PIN was positively correlated with age (R = 0.23, P = 0.001), and this correlation was significant for PIN within 2 mm of cancer and for PIN greater than 2 mm from cancer (R = 0.18 and 0.24, respectively). There was also a positive correhition of total volume of PIN and prostatic weight (R = 0.20, P = 0.004), but this correlation was only significant for PIN within 2 mm of cancer (R = 0.22, P = 0.002) and not for PIN greater than 2 mm from cancer (p > 0.05).
Correlation of Volume of PIN with Volume of Prostate Cancer The mean prostate cancer volume was 9.19 cc (standard error, 0.6 cc), with a range from 0.01 to 56.9 cc. There was a strong positive correlation of total volume of PIN with cancer volume (R = 0.44, P = 0.0001) but this correlation was only seen in PIN within 2 mm of cancer (R = 0.56, P < 0.001) and not in PIN greater than 2 mm from cancer (p > 0.05).
Correlation of Volume of PIN with Pathologic Stage of Cancer The total volume of PIN increased with increasing pathologic stage of cancer (R = 0.28, P = 0.0001).
Correlation of Volume of PIN with Gleason Grade The mean Gleason score was 6.2 (range, 3-9). The mean percent of Gleason patterns 4 and 5 was 28.5%, with a range of 0 to 100%. The total volume of PIN increased with increasing Gleason score (R = 0.16, P = 0.02). The volume of PIN decreased as the percentage of each of the following measures of poorly differentiated adenocarcinoma increased: Gleason primary pattern 4 (R = 0.16, P = 0.03), primary pattern 5 (R = 0.18, P = 0.01) and combined primary patterns 4 and 5 (R = 0.15, P = 0.03); this trend was found only for PIN within 2 mm of cancer (R = 0.24, P = 0.001), and not for PIN greater than 2 mm from cancer (p > 0.05).
Light Microscopic and Immunohistochemical Features of AAH AAH in the transition zone was always found in association with nodular hyperplasia, and the majority of foci occurred within or adjacent to hyperplastic epithelial nodules. AAH usually consisted of a cohesive proliferation of small glands with pale finely-granular cytoplasm at the edge of a nodule, often accompanied by a mildly cellular stroma. Cytologically, AAH was similar to the adjacent uninvolved prostatic glands, with no evidence of significant nuclear or nucleolar enlargement. The gland lumens occasionally contained small corpora amylacea, basophilic mucinous material, and crystalloids, but these were not common features (Fig. 4).
AAH in the non-transition zone (peripheral and central zones) occasionally occurred in association with rare peripheral zone nodules, but was otherwise randomly distributed in areas with normal and hyperplastic epithelium; areas of atrophy rarely showed AAH. Rare foci of AAH appeared as discrete circumscribed proliferations in the stroma without other features of peripheral zone nodule, possibly representing incipient hyperplastic epithelial nodules at this site. The architectural and cytologic features in these foci were similar to AAH in the transition zone. One focus of adenomatous hyperplasia without atypia (small gland variant of nodular hyperplasia, characterized by a uniform circumscribed proliferation of small acini) was excluded from study. Also, two foci of atypical small glandular proliferation of uncertain significance were excluded from study, including one each in the transition zone and non-transition zone. Immunohistochemical studies for basal cell-specific anti-keratin monoclonal antibody 34~E12 in 28 cases invariably revealed fragmented and discontinuous staining in AAH and no staining in adenocarcinoma. There was no staining for S-100 protein in any of the glands.
Zonal Location, Frequency and Multicentricity of AAH and Spatial Association with Cancer AAH was identified in 50 of 217 cases (23.0 % ), and was more common in the transition zone than in the non-transition zone (43 of 217 cases [19.8% of all cases] vs. 13 of 217 cases [6.0%]) (R=O.72; p < 0.0001). Only 7 cases of AAH were exclusively in the non-transition zone (3.0% of all cases; 14% of cases of AAH), compared with 37 cases of AAH exclusively in the transition zone (17% of all cases; 74% of cases of AAH); 6 other cases of AAH were found in both zones (3 % of all cases; 12 % of all cases of AAH). AAH was found within 2 mm of cancer in 17 of 50 cases of AAH (34% of the cases of AAH), including 13 of 43 cases of AAH in the transition zone (30%) and 4 of 13 cases of AAH in the non-transition zone (31 %). The number of foci of AAH in the transition zone was always greater than that in the non-transition zone, regardless of whether it was within 2 mm of cancer (R = 0.89; P < 0.0001) or more than 2 mm from cancer (R = 0.75; P < 0.0001). AAH was multicentric in 23 cases (10.6% of all cases; 46% of the cases of AAH), including 20 cases in the transition zone (9.2% of all cases; 40% of the cases of AAH; 47% of the cases of AAH in the transition zone) and 3 cases in the non-transition zone (1.4% of all cases; 6 % of the cases of AAH; 23 % of the cases of AAH in the non-transition zone). The following number of foci of AAH were found in the transition zone: 0 foci (174 cases; 80.1%), 1 focus (23 cases; 10.6%), 2 foci (7 cases; 3.2%), 3 foci (4 cases; 1.8%), 4 foci (3 cases; 1.4%), 5 foci (2 cases; 0.9%), 6 foci (2 cases; 0.9%), 7 foci (1 case; 0.5%) and 10 foci (1 case; 0.5%). The following number of foci of AAH were found in the non-transition zone:
864 . Junqi Qian and D. G. Bostwick
ofoci (204 cases; 94.0%),1 focus (10 cases; 4.6%) and 2 foci (3 cases; 1.4%). Correlation of Number of Foci of AAH with Volume of Cancer The mean prostate cancer volume was 9.6 cc (standard error, 0.6 cc) with a range of 0.01 to 56.9 cc. There was a weak positive correlation of cancer volume with number of foci of AAH in the transition zone (R = 0.17, P < 0.01) and non-transition zone
(R = 0.39, p < 0.0001), but this correlation was not significant when all foci of AAH were considered (p > 0.05); a similar trend was seen with volume of AAH.
Correlation of Number of Foci of AAH with Other Variables There was a positive correlation of patient age with total number of foci of AAH and number of foci in the transition zone and non-transition zone (all
Fig. 4. Atypical adenomatous hyperplasia in the transition zone. A: At the periphery of this hyperplasia nodule, the glands are small and tightly packed. B: At high magnification, the nuclei are small and the nucleoli are inconspicuous.
PIN and AAH in prostate cancer . 865
p < 0.0001). Also, there was a positive correlation of prostatic weight and prostatic volume with total number of foci of AAH regardless of zonal location (all p < 0.0001). However, the strongest positive correlations were with the percent nodular hyperplasia and total number of foci of AAH (R = 0.48), number of foci in the transition zone (R = 0.44) and number of foci in the non-transition zone (R = 0.45) (all p < 0.0001). Among the measures of tumor grade, only the percent of Gleason patterns 4 and 5 (poorly differentiated cancer) correlated with total number of foci of AAH and number of foci in the transition and non-transition zones. The volume of high grade PIN and preoperative serum PSA concentration showed a positive correlation with number of foci of AAH in the transition and non-transition zones, but only PSA also correlated with total foci of AAH, although weakly. There was no correlation of pathologic stage with number of foci of AAH (all p > 0.05).
Correlation of Volume of AAH with Other Variables The total mean volume of AAH was 0.029 cc (standard error, 0.009; range, 0-1.29 cc), and was much higher in the transition zone than in the non-transition zone (total mean volumes of AAH, 0.03 cc vs. 0.003 cc, respectively); this was true regardless of whether it was within 2 mm of cancer or more than 2 mm from cancer (all p < 0.01). The volume of AAH within 2 mm of cancer was lower than that for AAH more than 2 mm from cancer, and this was true for AAH in the transition zone (R = 0.77, p < 0.0001) and in the non-transition zone (R = 0.92, p < 0.0001). Discussion In this investigation, we measured the volume of high grade PIN, AAH and cancer in a series of totally sectioned radical prostatectomy specimens. PIN was found in 86% of the cases. The mean volume of PIN was 1.32 cc, similar to the mean of about 1.2 cc reported by de la Torre et al. 30 . These results support the close association of high grade PIN and cancer. AAH was found in 23 % of the cases, similar to the frequency in transurethral resection specimens (19.6%36.9%)17,31 and autopsy specimens (15_40%)10,32. The mean volume of AAH was 0.029 cc (range, 01.29 cc) and was much higher in the transition zone than in the non-transition zone.
Zonal Location Most foci of PIN were found exclusively in the nontransition zone (63% of cases of PIN) or simultaneously in the transition and non-transition zones (36%); only 1 % of the cases were exclusively in the transition zone. This high frequency of non-transition zone involvement by PIN is similar to previous reports2, 3, 6, 23, 26. Conversely, most foci of AAH were
found exclusively in the transition zone (74% of cases of AAH) or simultaneously in the transition and nontransition zones (12%); only 14% of the cases of AAH were exclusively in the non-transition zone. This high frequency of transition zone involvement by AAH is similar to previous reports in transurethral resection specimens and autopsy cases 10,16,25,33. These results have important implications for the origin of prostatic carcinoma in the transition zone; if PIN or AAH is the precursor for many cases of prostate cancer at this site, as has been proposed, the level of involvement of the transition zone which we observed for either lesion could account for the relative frequency (25%) of cancer at this site. The high frequency of involvement of the transition zone by PIN which we observed was similar to that seen in prostatectomy and autopsy cases by Kovi et al. (31.2 %)4, but higher than in other studies which utilized transurethral resections or only a small number of patients 6, 23.
Multicentricity PIN was multicentric in 72 % of all cases, including 63% of those involving the non-transition zone and 7% of those involving the transition zone; 2 % of the cases had concomitant single foci in all zones. AAH was multicentric in 46% of cases of AAH, including 47% of those involving the transition zone and 23 % of those involving the non-transition zone. These findings are in agreement with previous reports2-4, 10, 16,23,26.
Spatial Association with Cancer Foci of PIN were usually within 2 mm of cancer. There was a strong positive correlation of total volume of PIN with cancer volume (R = 0.44, p = 0.0001) but this correlation was only seen in PIN within 2 mm of cancer (R = 0.56, P < 0.001) and not in PIN greater than 2 mm from cancer. There have been conflicting reports concerning the relation between the volume of PIN and volume of cancer, perhaps due to differences in methods of measurement. The positive correlation that we found was only significant for PIN within 2 mm of cancer. A negative correlation of PIN and cancer was obtained by others, and these results were attributed to overgrowth and replacement of PIN by cancer6, 23, 30-34. These differences may also be due to difficulty in identifying PIN in fields with adenocarcinoma. Numerous reports have shown that the highest grades of PIN are spatially closest to carcinoma 1,30. AAH was spatially associated with cancer (within 2 mm) in 34% of the cases, with an almost equal frequency of association in the transition zone and nontransition zone (30% vs. 31 %, respectively). However, because AAH was more common in the transition zone, the absolute number of cases near cancer was higher at that site. Despite the frequent spatial association of AAH and cancer, there were no morphologic features suggesting a transition from one to another.
866 . Junqi Qian and D. G. Bostwick
Patient Age We found that the volume of high grade PIN increased as patient age increased. Other studies have compared the prevalence of high grade PIN according to patient age. Kovi et a1. found that the prevalence of PIN in prostates with cancer increased with age 3. In an autopsy series, Sakr et a1. identified an increase in the frequency of low grade PIN in men in their twenties and thirties (9% and 22%, respectively), with increasing frequency of high grade PIN with age 35 . However, Humphrey et a1. found no relationship between estimated extent of PIN and patient age in a series of 81 patients treated by radical prostatectomy34. Similarly, McNeal and Bostwick did not find a significant difference in the frequency of PIN in cancer patients with advancing age 2 • The frequency of AAH increased with patient age. Brawn identified "mild adenosis" in 20% of 20-29 year-aIds and 24% of 30-40 year-olds at autopsy9; similarily, Melhorn noted an increase from 23 % in 40-49 year-olds to 40% in men over age 80 years33. This increase in AAH may be due to an increase in prostate volume, prostatic weight and percent of nodular hyperplasia which we observed in the current series; these increases were observed in foci of AAH involving the transition zone and non-transition zone when evaluated separately. The increase in frequency of AAH with increasing serum PSA level probably reflects the increase in cancer volume more than any other factor 19 .
Conclusion Our findings indicate that the extent and anatomic location of high grade PIN and carcinoma are strongly associated, and that PIN is frequently multicentric. Because PIN has a high predictive value as a marker for adenocarcinoma, its identification in biopsy specimens of the prostate warrants further search for concurrent invasive carcinoma 2o - 23 ,36. AAH and cancer are also spatially associated, albeit to a lesser extent than PIN; whether this is coincidental or evidence of causation is unsettled. Follow-up studies of AAH are needed to determine the predictive value of this lesion for carcinoma. Although the biologic significance of AAH is uncertain, its light microscopic appearance and immunophenotype allow it to be separated from carcinoma in most cases.
Material and Methods Paraffin-embedded serially sectioned radical retropubic prostatectomy specimens from 1991-1993 were retrieved from the files of the Mayo Clinic. For the study of PIN, only patients with clinically localized prostate cancer were included; the final PIN study group included 195 cases. For the study of AAH, 17 additional patients without clinically localized prostate cancer and 5 additional localized cases were included; the final AAH study group included 217 cases. None had received preoperative therapy.
All foci of high grade PIN, AAH and andenocarcinoma were mapped directly on the slides. The volume of each was determined by the grid-counting method?; the sum of the area of each slice was multipled by the average slice thickness, and the sum of these volumes was multi pled by 1.25 to account for formalin shrinkage. The spatial proximity of PIN and AAH to cancer was recorded (less than or equal to 2 mm from cancer, and greater than 2 mm from cancer). We arbitrarily chose 2 mm because it is easy to measure by micrometer and has been used previously by other investigators. The zonal location (transition zone and non-transition [peripheral and central] zone) for each focus of PIN and AAH was noted, and the number of foci was recorded; foci were considered separate when 2 mm or more apart. In 28 cases with AAH, immunohistochemical studies for basal cell-specific anti-keratin antibody 34~E12 and S-100 protein (Dako Corp., Carpinteria, CAl were performed using the avidin-biotin peroxidase method and citrate buffer microwave antigen retrieval, as previously described!, 11. Substitution of non-immune serum served as negative control. For each case, the following was recorded: patient age, prostate weight, prostate volume (according to the formula for a prolate ellipsoid: length x height x width x 0.532), percent benign prostatic hyperplasia (BPH) estimated in 10% increments, Gleason score (sum of primary and secondary patterns), percent Gleason primary patterns 4 and 5 (indicating the amount of poorly differentiated carcinoma), preoperative serum prostate specific antigen (PSA) level (monoclonal)28 and pathologic stage (TNM classification, 1992 revision)29.
Statistical Analysis The relationship between volume of PIN and volume of cancer was determined by linear regression analysis. Spearman rank correlations were used to evaluate the relationship between the volume of PIN and other variables; these correlations were also used to evaluate the relationship between number of foci of AAH, volume of AAH and other variables. For ordinal categorical variables, the Kruskal-Wallis test was used. Significance level was 0.05.
References ) Bostwick DG, Brawer MK (1987) Prostatic Intra-Epithelial Neoplasia and Early Invasion in Prostate Cancer. Cancer 59: 788-794 2 McNeal JE, Bostwick DG (1986) Intraductal Dysplasia: A Premalignant Lesion of the Prostate. Hum Pathol17: 6471 3 Kovi J, Mostofi FK, Heshmat MY, Enterline JP (1988) Large Acinar Atypical Hyperplasia and Carcinoma of the Prostate. Cancer 61: 555 - 561 4 McNeal JE, Alroy J, Leav I, Redwine EA, Freiha FS, Stamey TA (1988) Immunohistochemical Evidence for Impaired Cell Differentiation in the Premalignant Phase of Prostate Carcinogenesis. Am J Clin Path 90: 23-32 5 McNeaIJE, Villers A, Redwine EA, Freiha FS, StameyTA (1991) Microcarcinoma in the Prostate: Its Association with Duct-Acinar Dysplasia. Hum Pathol 22: 644-652 6 Quinn BD, Cho KR, Epstein JI (1990) Relationship of Severe Dysplasia to Stage B Adenocarcinoma of Prostate. Cancer 65: 2328-2337
PIN and AAH in prostate cancer· 867 7 Humphrey PA, Vollmer RT (1990) Intraglandular Tumor Extent and Prognosis in Prostate Carcinoma: Application of Grid Method to Prostatectomy Specimens. Hum Pathol 21: 799-804 8 Bostwick DG, Amin MB, Dundore P, Marsh W, Schultz DS (1993) Architectural Patterns of High-Grade Prostatic Intraepithelial Neoplasia. Hum Pathol 24: 298-310 9 Brawn PN (1982) Adenosis of the prostate: A dysplastic lesion that can be confused with prostate adenocarcinoma. Cancer 49: 826-833 10 Brawn PN, Speights VO, Contin JU, Bayardo RJ, Kuhl DL (1989) Atypical hyperplasia in prostates of 20 to 40 year old men. J Clin Pathol 42: 383-386 11 Bostwick DG, Srigley J, Grignon D, Maksem J, Humphrey P, van der Kwast T, Bose D, Harrison J, Young R (1993) Atypical Adenomatous Hyperplasia of the prostate: Morphologic Criteria for its Distinction from Well-Differentiated Carcinoma. Hum Pathol 24: 819-832 12 Bostwick DG, Srigley J (1990) Premalignant Lesions. In: Bostwick DG (Ed). Pathology of the Prostate, ChurchillLivingstone, New York, pp 37-59.19. 13 Gleason DF (1985) Atypical hyperplasia, benign hyperplasia, and well-differentiated adenocarcinoma of the prostate. Am J Surg Pathol 9: 53-67 14 Goldstein NS, Qian J, Bostwick DG (1995) Mucin expression in atypical adenomatous hyperplasia of the prostate Hum Pathol (in press) 15 Grignon D], Ro JY, Srigley JR, Troncoso P, Raymond AK, Ayala AG (1992) Sclerosing adenosis of the prostate gland. A lesion showing myoepithelial differentiation. Am J Surg Pathol16: 383-391 16 Helpap B (1980) The biological significance of atypical hyperplasia of the prostate. Virchows Arch A Path Histol3 87: 307-317 17 Strigley JR (1988) Small-acinar patterns in the prostate gland with emphasis on atypical adenomatous hyperplasia and small-acinar carcinoma. Semin Diag Pathol 5: 254-272 18 Young RH, Clement PB (1987) Sclerosing adenosis of the prostate. Arch Pathol Lab Med 11: 363 - 366 19 Drago JE, Mostofi FK, Lee F (1989) Introductory Remarks and workshop summary. Urol 34 (Suppl): 2-3 20 Brawer MK, Bigler SA, Sohlberg OE, Nagle RB, Lange PH (1991) Significance of Prostatic Intraepithelial Neoplasia on Prostate Needle Biopsy. Urology 38: 103-107 21 Weinstein MH, Epstein JI (1993) Significance of High Grade Prostatic Intraepithelial Neoplasia on Needle Biopsy. Hum Pathol24: 624-629
22 Davidson D, Bostwick DG, Qian J, Siroky M, Rudders R, Osterling J, Stilmant M (1995) Prostatic Intraepithelial Neoplasia is predictive of Adenocarcinoma. J Urol (in press) 23 Epstein JI, Cho KR, Quinn BD (1990) Relationship of Severe Dysplasia to Stage A [Incidental] Adenocarcinoma of the Prostate. Cancer 65: 2321-2327 24 Helpap B (1988) Observations on the number, size, and localization of nucleoli in hyperplastic and neoplastic prostatic disease. Histopathol13: 203-211 25 McNeal JE (1969) Origin and development of carcinoma of the prostate. Cancer 28: 24-34 26 Troncosco P, Babaian RJ, Ro JY, Grignon DJ, von Eschenbach AC, Ayala AC (1989) Prostatic Intraepithelial Neoplasia and Invasive Prostatic Adenocarcinoma in Cystoprotatectomy Specimens. Urology (Suppl) 24: 52-56 27 Kovi J (1985) Microscopic differential diagnosis of small acinar adenocarcinoma of prostate. Pathol Annu 20: 157-179 28 Blackwell KL, Bostwick DG, Zincke H, et al. (1994) Prostate specific antigen: Its use as a practical preoperative marker of tumor burden and pathologic stage in adenocarcinoma of the prostate. J Urol15: 1565-1570 29 Bostwick DG, Myers RP, Oesterling JE (1994) Staging of rcrostate cancer. Sem Surg Oncol 10: 60-72. o De La Torre, Haggman M, Brandstedt S, Busch C (1993) Prostatic Intraepithelial Neoplasia and Invasive Carcinoma in Total Prostatectomy Specimens: Distribution, Volume and DNA Ploidy. Bri J Urol 72: 207-213 31 Srigley J, Toth P, Hartwick RWJ (1989) Atypical histologic patterns in cases of benign prostatic hyperplasia. Lab Invest 60: 90A 32 Murphy WM (1989) Urological Pathology. W. B. Saunders Company, Philadelphia, PA, pp 184 33 Melhorn J (1987) Problems of so-called atypical hyperplasia of the prostate. Z Urol Nephrol 80: 379-385 34 Humphrey PA, Frazier HA, Paulson DF, Vollmer RT (1992) Extent of Severe Dysplasia in the Prostate is Inversely Related to Pathologic Stage. Modern Pathol5: 54A (abstract) 35 Sakr WA, Haas GP, Cassin BF, Pontes JE, Crissman JD (1993) The Frequency of Carcinoma and Intraepithelial Neoplasia of the Prostate in Young Male Patients. J Urol 150: 379-385 36 Bostwick DG (1992) Prostatic Intraepithelial Neoplasia (PIN): Current Concepts. J Cell Biochem (SuppI16H): 10-19
Received March 8, 1995 . Accepted June 10, 1995
Key words: Prostatic intraepithelial neoplasia - Atypical adenomatous hyperplasia - Adenocarcinoma - Tumor stage - multicentricity - Anatomy - Tumor volume David G. Bostwick, M. D., Division of Anatomic Pathology, Mayo Clinic, 200 First Street, SW, Rochester, MN 55905, USA, Phone (507) 284-1870, Fax (507) 284-1599
The Extent and Zonal Location of Prostatic Intraepithelial Neoplasia and Atypical Adenomatous Hyperplasia: Relationship with Carcinoma in Radical Prostatectomy Specimens ::Junqi Qian and D. G. Bostwick Department of Pathology and Laboratory Medicine, Mayo Clinic, Rochester, MN, U.S.A.
SUMMARY High-grade prostatic intraepithelial neoplasia (PIN) and atypical adenomatous hyperplasia (AAH) are considered putative precursors of prostatic adenocarcinoma. We determined the extent and zonal distribution of PIN and AAH in totally-embedded radical prostatectomies with prostate cancer, including 195 cases with PIN and 217 with AAH. PIN was identified in 86% of the cases. The mean volume of PIN was 1.32 cc (range, 0-8.12 cc), and was greater for PIN within 2 mm of cancer (mean, 1.0 cc) than for PIN more than 2 mm from cancer (mean, 0.3 cc). PIN was usually multicentric (64.5% of cases) and located in the non-transition zone (63%) or all zones (36%). The volume of PIN was positively correlated with the volume of cancer, patient age, pathologic stage and Gleason score. AAH was identified in 23.0% of the cases, and was more frequent in the transition zone (19.8% of cases) than in the nontransition (peripheral and central) zone (6.0%). The number of foci of AAH in the transition zone was always greater than that in the non-transition zone. AAH was frequently multicentric (46% of cases), especially in the transition zone (47% of transition zone cases) compared with the non-transition zone (23% of non-transition zone cases). The mean volume of AAH was 0.029 cc (range, 0-1.29 cc), and was much higher in the transition zone than in the non-transition zone. AAH was more common in older patients and those with greater prostatic weight, higher prostatic volume, greater percent of nodular hyperplasia, greater volume of cancer, greater percent of Gleason patterns 4 and 5 cancer, higher volume of prostatic intraepithelial neoplasia and higher serum prostate specific antigen concentration. Our results indicate that the extent and zonal distribution of high grade PIN and carcinoma are strongly associated, and that PIN is frequently multicentric; this supports the hypothesis that PIN is a premalignant lesion. AAH and carcinoma show a weak but significant association; if AAH is a premalignant lesion, it probably is associated with a subset of cancers arising in the transition zone.
*Supported in part by National Institutes of Health Grant CA58225-01C. 0344-0338/95/0191-0860$3.50/0
© 1995 by Gustav Fischer Verlag, Stuttgart
PIN and AAH in prostate cancer' 861
Introduction High grade prostatic intraepithelial neoplasia (PIN) and atypical adenomatous hyperplasia (AAH) represent putative precursors of prostatic adenocarcinoma, although the evidence for PIN is stronger l - 18 . Two grades of PIN are recognized (low grade and high grade), and high grade PIN is considered the most likely precursor of invasive carcinoma 19 • The clinical importance of recognizing high grade PIN is based on its strong association with carcinoma 2 - 6 ; its identification in biopsy specimens of the prostate warrants further search for concurrent invasive carcinoma 20 - 23 • The clinical importance of AAH is uncertain, although some data suggest that it is a precursor to adenocarcinoma arising in the transition zone 9 - 11 , 24, 25. Reports to date have provided limited information concerning the quantity of PIN and AAH and their relationship with other clinical and pathologic features of prostate cancer2, 6, 7, 26, 27. In this study, we determined the extent and zonal location of high grade PIN and AAH in a large series of radical retropubic prostatectomy specimens with cancer, and correlated the results with a variety of features, including patient age, volume of cancer, pathologic stage and tumor grade. Results Patients ranged in age from 45 to 77 years (mean 64.4 years). The prostates weighed from 21 to 185 grams (mean, 52.6 grams), and volumes ranged from 8.1 to 189.0 cc (mean, 38.8 cc). Clinical stages included TlaNOMO (2 cases [1 %]), TlbNOMO (4 cases [2%]), TlcNOMO (35 cases [16%]), T2a + bNOMO (48 cases [22%]), T2cNOMO (111 cases [51 %]), T3a + bNOMO (10 cases [5%]) and T3cNOMO (7 cases [3%]). Patho-
logic stages included Tla + bNOMO (6 cases [3%]), T2a + bNOMO (13 cases [6%]), T2cNOMO (116 cases [54%]), T3a + bNOMO (49 cases [22%]), T3cNOMO (21 cases [10%]) and TxNIMO (12 cases [5%]).
Extent and Zonal Location of High Grade PIN PIN was identified in 167 of 195 cases (86%) (Fig. 1). The total mean volume of PIN was 1.32 cc (standard error, 0.10; range 0-8.12 cc), and was greater for PIN within 2 mm of cancer (mean, 1.0 cc; standard error, 0.1 cc; range, 0 - 7.8 cc) than for PIN more than 2 mm from cancer (mean, 0.33 cc; standard error, 0.33 cc; range 0-3.2 cc; R = 0.30, P = 0.00001) (Table 1). High grade PIN was identified in the transition zone in 61 cases (31 % of all cases) and in the non-transition zone (peripheral and central zones) in 166 cases (85%); only 1 case of PIN was exclusively in the transition zone (1.0% of all cases), compared with 106 cases of PIN exclusively in the non-transition zone (54%); 60 other cases of PIN were found in both zones (31 %) (Fig. 2), and 28 cases had no PIN (14%). PIN was multicentric in 72 % of all cases, including 7% multicentricity in the transition zone and 63 % in the non-transition zone; in the remaining 2 % of the cases, multicentricity was characterized by concomitant single foci in the transition zone and non-transition zone. The following number of foci of PIN were found in the transition zone: 0 foci (134 cases; 69%), 1 focus (47 cases; 24%), 2 foci (11 cases; 5.5%), 3 foci (1 case; 0.5%) and 4 foci (2 cases; 1.0%). The following number of foci of PIN were found in the non-transition zone: 0 foci (28 cases; 14%), 1 focus (42 cases; 21.5%), 2 foci (60 cases; 31%), 3 foci (49 cases; 25%), 4 foci (13 cases; 7%) and 5 foci (3 cases; 1.5%) (Fig. 3).
Table 1. Comparison of prostatic intra epithelial neoplasia and atypical adenomatous hyperplasia in a series of totally embedded prostatectomies with cancer
Prevalence (All sites, % of all cases) Prevalence, transition zone only Prevalence, non-transition zone only Prevalence, both zones simultaneously Volume (mean) (range) Multicentricity (% of cases) Multicentricity, transition zone Multicentricity, non-transition zone Correlation with patient age Correlation with prostate weight Correlation with cancer volume Correlation with pathologic stage Correlation with Gleason score < 2 mm from cancer > 2 mm from cancer
High grade PIN
AAH
86% 1% 63% 36% 1.32 cc. (0- 8.12) 72% 7% 63% S" * S" S* 'f" S'·'f" S* S'·"
23% 74% 14% 12% 0.029 cc. 46% 47% 23% S"':· *
S"
-: Not significant; S: significant; *P < 0.01; 'f *P < 0.001; ,. * "P < 0.0001. Note: Correlations for PIN are with volume of PIN; correlations for AAH are with number of foci of AAH.
862 . Junqi Qian and D. G. Bostwick
Fig. 1. Micropapillary Pattern of high grade PIN.
Fig. 2. Whole-mount section of the prostate showing two foci of PIN arising in the transition zone and peripheral zone (nontransition zone) (solid lines outline cancer; dotted lines outline PIN). 100
Zone Transition III Nontransition ~
80
£' 0
69
60
Q)
III
40
20 0
o 0
2
3
Foci of PIN (no.)
4
1.5
5
Fig. 3. Multicentricity of PIN according to zonal location.
PIN and AAH in prostate cancer . 863
Correlation of Volume of PIN with Patient Age and Prostate Weight The total volume of PIN was positively correlated with age (R = 0.23, P = 0.001), and this correlation was significant for PIN within 2 mm of cancer and for PIN greater than 2 mm from cancer (R = 0.18 and 0.24, respectively). There was also a positive correhition of total volume of PIN and prostatic weight (R = 0.20, P = 0.004), but this correlation was only significant for PIN within 2 mm of cancer (R = 0.22, P = 0.002) and not for PIN greater than 2 mm from cancer (p > 0.05).
Correlation of Volume of PIN with Volume of Prostate Cancer The mean prostate cancer volume was 9.19 cc (standard error, 0.6 cc), with a range from 0.01 to 56.9 cc. There was a strong positive correlation of total volume of PIN with cancer volume (R = 0.44, P = 0.0001) but this correlation was only seen in PIN within 2 mm of cancer (R = 0.56, P < 0.001) and not in PIN greater than 2 mm from cancer (p > 0.05).
Correlation of Volume of PIN with Pathologic Stage of Cancer The total volume of PIN increased with increasing pathologic stage of cancer (R = 0.28, P = 0.0001).
Correlation of Volume of PIN with Gleason Grade The mean Gleason score was 6.2 (range, 3-9). The mean percent of Gleason patterns 4 and 5 was 28.5%, with a range of 0 to 100%. The total volume of PIN increased with increasing Gleason score (R = 0.16, P = 0.02). The volume of PIN decreased as the percentage of each of the following measures of poorly differentiated adenocarcinoma increased: Gleason primary pattern 4 (R = 0.16, P = 0.03), primary pattern 5 (R = 0.18, P = 0.01) and combined primary patterns 4 and 5 (R = 0.15, P = 0.03); this trend was found only for PIN within 2 mm of cancer (R = 0.24, P = 0.001), and not for PIN greater than 2 mm from cancer (p > 0.05).
Light Microscopic and Immunohistochemical Features of AAH AAH in the transition zone was always found in association with nodular hyperplasia, and the majority of foci occurred within or adjacent to hyperplastic epithelial nodules. AAH usually consisted of a cohesive proliferation of small glands with pale finely-granular cytoplasm at the edge of a nodule, often accompanied by a mildly cellular stroma. Cytologically, AAH was similar to the adjacent uninvolved prostatic glands, with no evidence of significant nuclear or nucleolar enlargement. The gland lumens occasionally contained small corpora amylacea, basophilic mucinous material, and crystalloids, but these were not common features (Fig. 4).
AAH in the non-transition zone (peripheral and central zones) occasionally occurred in association with rare peripheral zone nodules, but was otherwise randomly distributed in areas with normal and hyperplastic epithelium; areas of atrophy rarely showed AAH. Rare foci of AAH appeared as discrete circumscribed proliferations in the stroma without other features of peripheral zone nodule, possibly representing incipient hyperplastic epithelial nodules at this site. The architectural and cytologic features in these foci were similar to AAH in the transition zone. One focus of adenomatous hyperplasia without atypia (small gland variant of nodular hyperplasia, characterized by a uniform circumscribed proliferation of small acini) was excluded from study. Also, two foci of atypical small glandular proliferation of uncertain significance were excluded from study, including one each in the transition zone and non-transition zone. Immunohistochemical studies for basal cell-specific anti-keratin monoclonal antibody 34~E12 in 28 cases invariably revealed fragmented and discontinuous staining in AAH and no staining in adenocarcinoma. There was no staining for S-100 protein in any of the glands.
Zonal Location, Frequency and Multicentricity of AAH and Spatial Association with Cancer AAH was identified in 50 of 217 cases (23.0 % ), and was more common in the transition zone than in the non-transition zone (43 of 217 cases [19.8% of all cases] vs. 13 of 217 cases [6.0%]) (R=O.72; p < 0.0001). Only 7 cases of AAH were exclusively in the non-transition zone (3.0% of all cases; 14% of cases of AAH), compared with 37 cases of AAH exclusively in the transition zone (17% of all cases; 74% of cases of AAH); 6 other cases of AAH were found in both zones (3 % of all cases; 12 % of all cases of AAH). AAH was found within 2 mm of cancer in 17 of 50 cases of AAH (34% of the cases of AAH), including 13 of 43 cases of AAH in the transition zone (30%) and 4 of 13 cases of AAH in the non-transition zone (31 %). The number of foci of AAH in the transition zone was always greater than that in the non-transition zone, regardless of whether it was within 2 mm of cancer (R = 0.89; P < 0.0001) or more than 2 mm from cancer (R = 0.75; P < 0.0001). AAH was multicentric in 23 cases (10.6% of all cases; 46% of the cases of AAH), including 20 cases in the transition zone (9.2% of all cases; 40% of the cases of AAH; 47% of the cases of AAH in the transition zone) and 3 cases in the non-transition zone (1.4% of all cases; 6 % of the cases of AAH; 23 % of the cases of AAH in the non-transition zone). The following number of foci of AAH were found in the transition zone: 0 foci (174 cases; 80.1%), 1 focus (23 cases; 10.6%), 2 foci (7 cases; 3.2%), 3 foci (4 cases; 1.8%), 4 foci (3 cases; 1.4%), 5 foci (2 cases; 0.9%), 6 foci (2 cases; 0.9%), 7 foci (1 case; 0.5%) and 10 foci (1 case; 0.5%). The following number of foci of AAH were found in the non-transition zone:
864 . Junqi Qian and D. G. Bostwick
ofoci (204 cases; 94.0%),1 focus (10 cases; 4.6%) and 2 foci (3 cases; 1.4%). Correlation of Number of Foci of AAH with Volume of Cancer The mean prostate cancer volume was 9.6 cc (standard error, 0.6 cc) with a range of 0.01 to 56.9 cc. There was a weak positive correlation of cancer volume with number of foci of AAH in the transition zone (R = 0.17, P < 0.01) and non-transition zone
(R = 0.39, p < 0.0001), but this correlation was not significant when all foci of AAH were considered (p > 0.05); a similar trend was seen with volume of AAH.
Correlation of Number of Foci of AAH with Other Variables There was a positive correlation of patient age with total number of foci of AAH and number of foci in the transition zone and non-transition zone (all
Fig. 4. Atypical adenomatous hyperplasia in the transition zone. A: At the periphery of this hyperplasia nodule, the glands are small and tightly packed. B: At high magnification, the nuclei are small and the nucleoli are inconspicuous.
PIN and AAH in prostate cancer . 865
p < 0.0001). Also, there was a positive correlation of prostatic weight and prostatic volume with total number of foci of AAH regardless of zonal location (all p < 0.0001). However, the strongest positive correlations were with the percent nodular hyperplasia and total number of foci of AAH (R = 0.48), number of foci in the transition zone (R = 0.44) and number of foci in the non-transition zone (R = 0.45) (all p < 0.0001). Among the measures of tumor grade, only the percent of Gleason patterns 4 and 5 (poorly differentiated cancer) correlated with total number of foci of AAH and number of foci in the transition and non-transition zones. The volume of high grade PIN and preoperative serum PSA concentration showed a positive correlation with number of foci of AAH in the transition and non-transition zones, but only PSA also correlated with total foci of AAH, although weakly. There was no correlation of pathologic stage with number of foci of AAH (all p > 0.05).
Correlation of Volume of AAH with Other Variables The total mean volume of AAH was 0.029 cc (standard error, 0.009; range, 0-1.29 cc), and was much higher in the transition zone than in the non-transition zone (total mean volumes of AAH, 0.03 cc vs. 0.003 cc, respectively); this was true regardless of whether it was within 2 mm of cancer or more than 2 mm from cancer (all p < 0.01). The volume of AAH within 2 mm of cancer was lower than that for AAH more than 2 mm from cancer, and this was true for AAH in the transition zone (R = 0.77, p < 0.0001) and in the non-transition zone (R = 0.92, p < 0.0001). Discussion In this investigation, we measured the volume of high grade PIN, AAH and cancer in a series of totally sectioned radical prostatectomy specimens. PIN was found in 86% of the cases. The mean volume of PIN was 1.32 cc, similar to the mean of about 1.2 cc reported by de la Torre et al. 30 . These results support the close association of high grade PIN and cancer. AAH was found in 23 % of the cases, similar to the frequency in transurethral resection specimens (19.6%36.9%)17,31 and autopsy specimens (15_40%)10,32. The mean volume of AAH was 0.029 cc (range, 01.29 cc) and was much higher in the transition zone than in the non-transition zone.
Zonal Location Most foci of PIN were found exclusively in the nontransition zone (63% of cases of PIN) or simultaneously in the transition and non-transition zones (36%); only 1 % of the cases were exclusively in the transition zone. This high frequency of non-transition zone involvement by PIN is similar to previous reports2, 3, 6, 23, 26. Conversely, most foci of AAH were
found exclusively in the transition zone (74% of cases of AAH) or simultaneously in the transition and nontransition zones (12%); only 14% of the cases of AAH were exclusively in the non-transition zone. This high frequency of transition zone involvement by AAH is similar to previous reports in transurethral resection specimens and autopsy cases 10,16,25,33. These results have important implications for the origin of prostatic carcinoma in the transition zone; if PIN or AAH is the precursor for many cases of prostate cancer at this site, as has been proposed, the level of involvement of the transition zone which we observed for either lesion could account for the relative frequency (25%) of cancer at this site. The high frequency of involvement of the transition zone by PIN which we observed was similar to that seen in prostatectomy and autopsy cases by Kovi et al. (31.2 %)4, but higher than in other studies which utilized transurethral resections or only a small number of patients 6, 23.
Multicentricity PIN was multicentric in 72 % of all cases, including 63% of those involving the non-transition zone and 7% of those involving the transition zone; 2 % of the cases had concomitant single foci in all zones. AAH was multicentric in 46% of cases of AAH, including 47% of those involving the transition zone and 23 % of those involving the non-transition zone. These findings are in agreement with previous reports2-4, 10, 16,23,26.
Spatial Association with Cancer Foci of PIN were usually within 2 mm of cancer. There was a strong positive correlation of total volume of PIN with cancer volume (R = 0.44, p = 0.0001) but this correlation was only seen in PIN within 2 mm of cancer (R = 0.56, P < 0.001) and not in PIN greater than 2 mm from cancer. There have been conflicting reports concerning the relation between the volume of PIN and volume of cancer, perhaps due to differences in methods of measurement. The positive correlation that we found was only significant for PIN within 2 mm of cancer. A negative correlation of PIN and cancer was obtained by others, and these results were attributed to overgrowth and replacement of PIN by cancer6, 23, 30-34. These differences may also be due to difficulty in identifying PIN in fields with adenocarcinoma. Numerous reports have shown that the highest grades of PIN are spatially closest to carcinoma 1,30. AAH was spatially associated with cancer (within 2 mm) in 34% of the cases, with an almost equal frequency of association in the transition zone and nontransition zone (30% vs. 31 %, respectively). However, because AAH was more common in the transition zone, the absolute number of cases near cancer was higher at that site. Despite the frequent spatial association of AAH and cancer, there were no morphologic features suggesting a transition from one to another.
866 . Junqi Qian and D. G. Bostwick
Patient Age We found that the volume of high grade PIN increased as patient age increased. Other studies have compared the prevalence of high grade PIN according to patient age. Kovi et a1. found that the prevalence of PIN in prostates with cancer increased with age 3. In an autopsy series, Sakr et a1. identified an increase in the frequency of low grade PIN in men in their twenties and thirties (9% and 22%, respectively), with increasing frequency of high grade PIN with age 35 . However, Humphrey et a1. found no relationship between estimated extent of PIN and patient age in a series of 81 patients treated by radical prostatectomy34. Similarly, McNeal and Bostwick did not find a significant difference in the frequency of PIN in cancer patients with advancing age 2 • The frequency of AAH increased with patient age. Brawn identified "mild adenosis" in 20% of 20-29 year-aIds and 24% of 30-40 year-olds at autopsy9; similarily, Melhorn noted an increase from 23 % in 40-49 year-olds to 40% in men over age 80 years33. This increase in AAH may be due to an increase in prostate volume, prostatic weight and percent of nodular hyperplasia which we observed in the current series; these increases were observed in foci of AAH involving the transition zone and non-transition zone when evaluated separately. The increase in frequency of AAH with increasing serum PSA level probably reflects the increase in cancer volume more than any other factor 19 .
Conclusion Our findings indicate that the extent and anatomic location of high grade PIN and carcinoma are strongly associated, and that PIN is frequently multicentric. Because PIN has a high predictive value as a marker for adenocarcinoma, its identification in biopsy specimens of the prostate warrants further search for concurrent invasive carcinoma 2o - 23 ,36. AAH and cancer are also spatially associated, albeit to a lesser extent than PIN; whether this is coincidental or evidence of causation is unsettled. Follow-up studies of AAH are needed to determine the predictive value of this lesion for carcinoma. Although the biologic significance of AAH is uncertain, its light microscopic appearance and immunophenotype allow it to be separated from carcinoma in most cases.
Material and Methods Paraffin-embedded serially sectioned radical retropubic prostatectomy specimens from 1991-1993 were retrieved from the files of the Mayo Clinic. For the study of PIN, only patients with clinically localized prostate cancer were included; the final PIN study group included 195 cases. For the study of AAH, 17 additional patients without clinically localized prostate cancer and 5 additional localized cases were included; the final AAH study group included 217 cases. None had received preoperative therapy.
All foci of high grade PIN, AAH and andenocarcinoma were mapped directly on the slides. The volume of each was determined by the grid-counting method?; the sum of the area of each slice was multipled by the average slice thickness, and the sum of these volumes was multi pled by 1.25 to account for formalin shrinkage. The spatial proximity of PIN and AAH to cancer was recorded (less than or equal to 2 mm from cancer, and greater than 2 mm from cancer). We arbitrarily chose 2 mm because it is easy to measure by micrometer and has been used previously by other investigators. The zonal location (transition zone and non-transition [peripheral and central] zone) for each focus of PIN and AAH was noted, and the number of foci was recorded; foci were considered separate when 2 mm or more apart. In 28 cases with AAH, immunohistochemical studies for basal cell-specific anti-keratin antibody 34~E12 and S-100 protein (Dako Corp., Carpinteria, CAl were performed using the avidin-biotin peroxidase method and citrate buffer microwave antigen retrieval, as previously described!, 11. Substitution of non-immune serum served as negative control. For each case, the following was recorded: patient age, prostate weight, prostate volume (according to the formula for a prolate ellipsoid: length x height x width x 0.532), percent benign prostatic hyperplasia (BPH) estimated in 10% increments, Gleason score (sum of primary and secondary patterns), percent Gleason primary patterns 4 and 5 (indicating the amount of poorly differentiated carcinoma), preoperative serum prostate specific antigen (PSA) level (monoclonal)28 and pathologic stage (TNM classification, 1992 revision)29.
Statistical Analysis The relationship between volume of PIN and volume of cancer was determined by linear regression analysis. Spearman rank correlations were used to evaluate the relationship between the volume of PIN and other variables; these correlations were also used to evaluate the relationship between number of foci of AAH, volume of AAH and other variables. For ordinal categorical variables, the Kruskal-Wallis test was used. Significance level was 0.05.
References ) Bostwick DG, Brawer MK (1987) Prostatic Intra-Epithelial Neoplasia and Early Invasion in Prostate Cancer. Cancer 59: 788-794 2 McNeal JE, Bostwick DG (1986) Intraductal Dysplasia: A Premalignant Lesion of the Prostate. Hum Pathol17: 6471 3 Kovi J, Mostofi FK, Heshmat MY, Enterline JP (1988) Large Acinar Atypical Hyperplasia and Carcinoma of the Prostate. Cancer 61: 555 - 561 4 McNeal JE, Alroy J, Leav I, Redwine EA, Freiha FS, Stamey TA (1988) Immunohistochemical Evidence for Impaired Cell Differentiation in the Premalignant Phase of Prostate Carcinogenesis. Am J Clin Path 90: 23-32 5 McNeaIJE, Villers A, Redwine EA, Freiha FS, StameyTA (1991) Microcarcinoma in the Prostate: Its Association with Duct-Acinar Dysplasia. Hum Pathol 22: 644-652 6 Quinn BD, Cho KR, Epstein JI (1990) Relationship of Severe Dysplasia to Stage B Adenocarcinoma of Prostate. Cancer 65: 2328-2337
PIN and AAH in prostate cancer· 867 7 Humphrey PA, Vollmer RT (1990) Intraglandular Tumor Extent and Prognosis in Prostate Carcinoma: Application of Grid Method to Prostatectomy Specimens. Hum Pathol 21: 799-804 8 Bostwick DG, Amin MB, Dundore P, Marsh W, Schultz DS (1993) Architectural Patterns of High-Grade Prostatic Intraepithelial Neoplasia. Hum Pathol 24: 298-310 9 Brawn PN (1982) Adenosis of the prostate: A dysplastic lesion that can be confused with prostate adenocarcinoma. Cancer 49: 826-833 10 Brawn PN, Speights VO, Contin JU, Bayardo RJ, Kuhl DL (1989) Atypical hyperplasia in prostates of 20 to 40 year old men. J Clin Pathol 42: 383-386 11 Bostwick DG, Srigley J, Grignon D, Maksem J, Humphrey P, van der Kwast T, Bose D, Harrison J, Young R (1993) Atypical Adenomatous Hyperplasia of the prostate: Morphologic Criteria for its Distinction from Well-Differentiated Carcinoma. Hum Pathol 24: 819-832 12 Bostwick DG, Srigley J (1990) Premalignant Lesions. In: Bostwick DG (Ed). Pathology of the Prostate, ChurchillLivingstone, New York, pp 37-59.19. 13 Gleason DF (1985) Atypical hyperplasia, benign hyperplasia, and well-differentiated adenocarcinoma of the prostate. Am J Surg Pathol 9: 53-67 14 Goldstein NS, Qian J, Bostwick DG (1995) Mucin expression in atypical adenomatous hyperplasia of the prostate Hum Pathol (in press) 15 Grignon D], Ro JY, Srigley JR, Troncoso P, Raymond AK, Ayala AG (1992) Sclerosing adenosis of the prostate gland. A lesion showing myoepithelial differentiation. Am J Surg Pathol16: 383-391 16 Helpap B (1980) The biological significance of atypical hyperplasia of the prostate. Virchows Arch A Path Histol3 87: 307-317 17 Strigley JR (1988) Small-acinar patterns in the prostate gland with emphasis on atypical adenomatous hyperplasia and small-acinar carcinoma. Semin Diag Pathol 5: 254-272 18 Young RH, Clement PB (1987) Sclerosing adenosis of the prostate. Arch Pathol Lab Med 11: 363 - 366 19 Drago JE, Mostofi FK, Lee F (1989) Introductory Remarks and workshop summary. Urol 34 (Suppl): 2-3 20 Brawer MK, Bigler SA, Sohlberg OE, Nagle RB, Lange PH (1991) Significance of Prostatic Intraepithelial Neoplasia on Prostate Needle Biopsy. Urology 38: 103-107 21 Weinstein MH, Epstein JI (1993) Significance of High Grade Prostatic Intraepithelial Neoplasia on Needle Biopsy. Hum Pathol24: 624-629
22 Davidson D, Bostwick DG, Qian J, Siroky M, Rudders R, Osterling J, Stilmant M (1995) Prostatic Intraepithelial Neoplasia is predictive of Adenocarcinoma. J Urol (in press) 23 Epstein JI, Cho KR, Quinn BD (1990) Relationship of Severe Dysplasia to Stage A [Incidental] Adenocarcinoma of the Prostate. Cancer 65: 2321-2327 24 Helpap B (1988) Observations on the number, size, and localization of nucleoli in hyperplastic and neoplastic prostatic disease. Histopathol13: 203-211 25 McNeal JE (1969) Origin and development of carcinoma of the prostate. Cancer 28: 24-34 26 Troncosco P, Babaian RJ, Ro JY, Grignon DJ, von Eschenbach AC, Ayala AC (1989) Prostatic Intraepithelial Neoplasia and Invasive Prostatic Adenocarcinoma in Cystoprotatectomy Specimens. Urology (Suppl) 24: 52-56 27 Kovi J (1985) Microscopic differential diagnosis of small acinar adenocarcinoma of prostate. Pathol Annu 20: 157-179 28 Blackwell KL, Bostwick DG, Zincke H, et al. (1994) Prostate specific antigen: Its use as a practical preoperative marker of tumor burden and pathologic stage in adenocarcinoma of the prostate. J Urol15: 1565-1570 29 Bostwick DG, Myers RP, Oesterling JE (1994) Staging of rcrostate cancer. Sem Surg Oncol 10: 60-72. o De La Torre, Haggman M, Brandstedt S, Busch C (1993) Prostatic Intraepithelial Neoplasia and Invasive Carcinoma in Total Prostatectomy Specimens: Distribution, Volume and DNA Ploidy. Bri J Urol 72: 207-213 31 Srigley J, Toth P, Hartwick RWJ (1989) Atypical histologic patterns in cases of benign prostatic hyperplasia. Lab Invest 60: 90A 32 Murphy WM (1989) Urological Pathology. W. B. Saunders Company, Philadelphia, PA, pp 184 33 Melhorn J (1987) Problems of so-called atypical hyperplasia of the prostate. Z Urol Nephrol 80: 379-385 34 Humphrey PA, Frazier HA, Paulson DF, Vollmer RT (1992) Extent of Severe Dysplasia in the Prostate is Inversely Related to Pathologic Stage. Modern Pathol5: 54A (abstract) 35 Sakr WA, Haas GP, Cassin BF, Pontes JE, Crissman JD (1993) The Frequency of Carcinoma and Intraepithelial Neoplasia of the Prostate in Young Male Patients. J Urol 150: 379-385 36 Bostwick DG (1992) Prostatic Intraepithelial Neoplasia (PIN): Current Concepts. J Cell Biochem (SuppI16H): 10-19
Received March 8, 1995 . Accepted June 10, 1995
Key words: Prostatic intraepithelial neoplasia - Atypical adenomatous hyperplasia - Adenocarcinoma - Tumor stage - multicentricity - Anatomy - Tumor volume David G. Bostwick, M. D., Division of Anatomic Pathology, Mayo Clinic, 200 First Street, SW, Rochester, MN 55905, USA, Phone (507) 284-1870, Fax (507) 284-1599