Relationship to Coexistent Adenocarcinoma and Atypical Adenomatous Hyperplasia of the Prostate

Transcription

1 1172 Prostatic Intraepithelial Neoplasia in Surgical Resections Relationship to Coexistent Adenocarcinoma and Atypical Adenomatous Hyperplasia of the Prostate Fredrik J. Skjørten, M.D., Ph.D. 1 BACKGROUND. High grade prostatic intraepithelial neoplasia (PIN) is associated Aasmund Berner, M.D., Ph.D. 2 with coincident prostate carcinoma, and has been considered to be a precursor Sverre Harvei, M.D., M.P.H. 3 of prostate carcinoma. Most studies on PIN have been performed on total prosta- Trude E. Robsahm, M.Sc. 3 tectomy or core needle biopsy specimens. Few reports deal with the occurrence Steinar Tretli, Ph.D. 3 of PIN in consecutive surgical resections, which is the objective of the current study. METHODS. A total of 1135 nonselected surgical specimens from the prostate, exam- 1 Department of Pathology, Ullevaal Hospital, Oslo, Norway. ined during 1974 and 1975, underwent blind review by 2 experienced pathologists. There were 79 core needle biopsies, 731 transurethral resections of the prostate 2 The Norwegian Radium Hospital, Department and 325 transvesical prostatic enucleations. The kappa coefficient for interobserver of Pathology, Oslo, Norway. reproducibility was 0.66 for PIN and 0.86 for carcinoma. 3 The Cancer Registry of Norway, Institute for RESULTS. Grade 1 PIN was found in 9%, Grade 2 PIN in 32%, and Grade 3 PIN in Epidemiological Cancer Research, Oslo, Nor- 20% of the total cases. Atypical adenomatous hyperplasia (AAH) was found in 11% way. and adenocarcinoma in 27% of the cases. Grade 3 PIN was associated with coincident adenocarcinoma in 39% of the cases. The prevalence of carcinoma for cases with Grade 1 and 2 PIN did not exceed that of those with Grade 0 PIN. PIN was most commonly found in association with small carcinomas. The mean age of the entire group of patients was 70.0 years, and was 69.4 years for patients with PIN without coincident carcinoma. Patients with PIN and coincident carcinoma had a mean age of 71.7 years, similar to all PIN grades, but patients with carcinoma without PIN had the highest mean age, 73.3 years. CONCLUSIONS. PIN is a common histologic finding in tissues from prostatic resections. In this study, Grade 3 PIN was strongly associated with coincident carcinoma, but lower grade PIN was not. There was no association between PIN and AAH. Patients with PIN did not appear to have a higher mean age than corresponding patients without PIN. Cancer 1997;79: American Cancer Society. KEYWORDS: prostate intraepithelial neoplasia, prostate carcinoma, grade, age, atypical adenomatous hyperplasia. The concept of intraepithelial neoplasia recognizable in histological sections has long been well established for most types of epithelia. Prostatic intraepithelial neoplasia (PIN) is considered to be a precur- Address for reprints: Sverre Harvei, M.D., sor of prostate carcinoma. 1,2 PIN-related lesions have previously had M.P.H., Cancer Registry of Norway, Institute for different names. 3 The term PIN was introduced by Bostwick and Epidemiological Cancer Research, Montebello, Brawer in 1987, 4 and is now generally accepted. In a study of consecu Oslo, Norway. tive autopsies, Sakr et al. 5 identified high grade PIN in American men Received July 22, 1996; revisions received Ocwith age to a maximum of 63% in men aged years. already in their fourth decade and found the prevalence increased tober 16, 1996, and December 16, 1996; accepted December 16, Atypical adenomatous hyperplasia (AAH) as described by Bost American Cancer Society

2 PIN and Prostate Carcinoma/Skjørten et al wick et al., 6 and accepted in a consensus statement by 20 leading uropathologists, 7 is less common than PIN in histologic material. The frequency with which such lesions have been diagnosed varies from %. 8 AAH has so far been only weakly linked to prostate carcinoma. 9 The tissue removed by transurethral resection of the prostate (TUR-P) mainly originates in the transition zone. The transition zone may show clinical adenocarcinoma; in addition, it often harbors incidental carcinomas, but may also be invaded by peripheral zone carcinomas, which are more frequent and larger than transition zone carcinomas. PIN is also most common in the peripheral zone. 10,11 Helpap 12 found FIGURE 1. Grade 2 prostatic intraepithelial neoplasia. Epithelial crowding PIN-like lesions in 8% of TUR-P specimens, 45% of with disturbed polarity and occasional luminal bridging. Several nuclei which were high grade. Epstein et al. 13 found severe show hyperchromasia and some nucleoli are prominent (1400). dysplasia (PIN 3) in 15.6% of TUR-P specimens, most often adjacent to carcinoma. There are several followup studies of patients with PIN, but follow-up time rarely exceeds an average of 2 years. 14,15 The authors are aware of only two long term follow-up studies of patients with PIN-like lesions. 16,17 The goal of the current study was to describe the prevalence of PIN and its relationship to simultaneous carcinoma and AAH in prostatic resections from consecutive specimens representative of a white male, urban population. MATERIAL AND METHODS All specimens from prostatic core needle biopsies (CB) and prostatic resections received in the Department of Pathology, Ullevaal Hospital during 1974 and 1975 FIGURE 2. Prostate gland with Grade 3 prostatic intraepithelial neoplawere included, for a total of 1230 specimens. The masia. Lining epithelium demonstrates luminal trabeculae and bridging. Nuterial was submitted mainly from the Departments of clear size is markedly increased with increased chromatin. Nucleoli are Urology at Ullevaal and Lovisenberg Hospitals, which prominent in lining cells but not in trabeculae (1400). at the time served approximately two-thirds of the population of the city of Oslo. Six total prostatectomies for prostate carcinoma were received during the period 1974 to 1975, they urologists for large, clinically benign glands. In 2.1% were not studied further. Nine cases could not be of patients undergoing TUR-P prostate carcinoma had traced due to insufficient identification, and 19 cases been diagnosed 5.3 years before. Information on prior were excluded due to missing or unsatisfactory histologic treatment was not available. material. Sixty-one patients had 2 or more speci- Before the start of the current study, ten represen- mens taken in the same year. In these cases, only the tative cases of PIN were selected. The slides were stud- first specimen was included in the current study. A ied thoroughly by the participating pathologists (F.J.S total of 1135 specimens from the same number of patients and Aa.B.) applying the diagnostic criteria of Bostwick were studied. These patients were representative and Brawer 4 (Figs. 1 and 2) with one addition: when of the male population of Oslo, Norway. there was coexistent carcinoma, the authors required CB was performed in patients with clinical suspi- a separation of one high-power field (HPF) (0.4 mm) cion of prostate carcinoma, and transurethral resection between a PIN lesion and adjacent carcinoma. After of the prostate (TUR-P) or transvesical prostatic several microscopy sessions a good diagnostic repro- enucleations (TV) were performed in patients with ducibility was obtained. The diagnostic criteria 4 as il- symptoms of urinary obstruction. TUR-P was the most lustrated by color micrographs of the selected cases, common procedure, but TV was preferred by some were collected in a manual that was kept at the micro-

3 1174 CANCER March 15, 1997 / Volume 79 / Number 6 scope and consulted when in doubt. All specimens TABLE 1 were reviewed blindly and independently by the two Summary of Total Material: Frequency of PIN, AAH, Total Carcinoma, Carcinoma Grade, and Carcinoma Involvement, pathologists. The presence and grades of PIN were According to Type of Specimen (Number in parentheses is recorded. AAH as described in a concensus statement 7 percent of total in group) was also noted; high and low grade AAH were recorded according to the presence or absence of enlarged nucleoli. However, in the current study, all AAHs were CB TUR-P TV Total No recorded as one entity. When AAH and adenocarci- Mean age (yrs) noma were found in the same specimen, carcinoma 95% CI was recorded and not AAH, for technical reasons. The No carcinoma presence, grade (World Health Organization [WHO] PIN Grades 1, 2, and 3), and extent (area) of coincident PIN 0 64 (81) 305 (42) 77 (24) 446 (39) PIN 1 3 (4) 60 (8) 34 (10) 97 (9) carcinoma (0, õ 25%, 26 50%, 51 75% and ú75% of PIN 2 4 (5) 216 (29) 146 (45) 366 (32) the section area) were also recorded, as was the type PIN 3 8 (10) 150 (21) 68 (21) 226 (20) of surgery performed, and the number of tissue blocks Sum 79 (100) 731 (100) 325 (100) 1135 (100) available (3.5 blocks per specimen on average). chi-square Å 99.9 df: 6 P õ a The material was comprised of 79 CB, 731 TUR- AAH and total carcinoma No AAH/carcinoma 28 (35) 446 (61) 228 (70) 702 (62) P, and 325 TV specimens that had been processed AAH 0 68 (9) 54 (17) 122 (11) in the following manner: for CB, all tissue had been Total carcinoma 51 (65) 217 (30) 43 (13) 311 (27) embedded. TUR-P chips were mixed well and three to chi-square Å df: 4 P õ a four paraffin blocks were embedded. When carcinoma WHO grade was suspected, all tissue was embedded. The TV mate- WHO 1 3 (6) 42 (19) 20 (47) 65 (21) WHO 2 29 (57) 97 (45) 19 (44) 145 (47) rial was comprised of nodules that were transected; WHO 3 19 (37) 78 (36) 4 (9) 101 (32) three to four paraffin blocks from the largest nodules Total carcinoma 51 (100) 217 (100) 43 (100) 311 (100) were embedded. Histologic examinations were per- chi-square Å 28.9 df: 4 P õ a formed on the original hematoxylin and eosin stained Involvement slides. In cases with diagnostic disagreement between õ25% 6 (12) 87 (40) 34 (79) 127 (41) 25 50% 4 (8) 34 (16) 2 (5) 40 (13) the two pathologists, the slides in question were re % 12 (23) 35 (16) 0 47 (15) viewed in a double microscope and a concensus diag- ú75% 29 (57) 61 (28) 7 (16) 97 (31) nosis was reached. For PIN, the overall agreement was Total carcinoma 51 (100) 217 (100) 43 (100) 311 (100) 773 of 1205 (64%), and the weighted kappa coefficient chi-square Å 54.1 df: 6 P õ a 18,19 was 0.66 (95% confidence interval [CI], 0.63 PIN: prostatic intraepithelial neoplasia; AAH: atypical adenomatous hyperplasia; CB: core needle bi- 0.69). For WHO grade the overall agreement was 997 opsy; TUR-P: transurethral resection of the prostate; TV: transvesical prostatic enucleations; CI: confiof 1205 (83%), and the weighted kappa coefficient was dence interval; PIN 0: no prostatic intraepithelial neoplasia; WHO: World Health Organization; df: 0.86 (95% CI, ). Differences between the degrees of freedom. groups were analyzed by chi-square tests for independence, a Pairwise comparisons between types of material revealed significant differences. and confidence limits for mean age were based a on Student s t distribution. and TUR-P material, but there were few Grade 3 lesions RESULTS in TV specimens. When the material was studied Table 1 shows the main findings for CB, TUR-P, and as a whole (data not shown), it was found that 95% of TV specimens. There was no difference in mean age Grade 1 carcinomas were small (õ25% involvement). of the patients between CB, TUR-P, and TV. The in- Nearly 60% of Grade 2 carcinomas involved õ50% of fluence of age on various parameters was analyzed for the tissue studied. The remaining 40% were equally the material as a whole. PIN was found in 61% of the divided between the two groups of larger carcinomas. specimens, with PIN 2 being most frequently diagnosed. Grade 3 tumors involved ú50% of the tissue in 83% PIN was least frequent in CB, and most com- of the specimens (data not shown). Approximately mon in TV specimens. There was no AAH in CB, 9% 80% of the carcinomas in the CB specimens involved in TUR-P and 17% in TV specimens. The various types ú50% of the section area, and few tumors were small. of material differed with respect to the occurrence of The situation was opposite for TV specimens. In the AAH and carcinoma grade (WHO). CB had 65% carci- TUR-P specimens, small and large carcinomas were noma, whereas the corresponding figures for TUR-P more frequent than intermediate ones. and TV were 30% and 13%, respectively. Approximately 39% of specimens with PIN 3 had WHO Grades 2 and 3 were most frequent in CB coincident carcinoma involvement, versus 31% for PIN

4 PIN and Prostate Carcinoma/Skjørten et al TABLE 2 Core Needle Biopsies: Association with PIN and AAH, Total Carcinoma, WHO Grade, and Carcinoma Involvement (Number in parentheses is percent of total in group) PIN 0 PIN 1 PIN 2 PIN 3 Total No AAH and total carcinoma No AAH/no ca 20 (31) 2 (66) 4 (80) 2 (29) 28 (35) AAH 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) Total carcinoma 44 (69) 1 (33) 1 (20) 5 (71) 51 (65) chi-square 6.3 df: 3 P Å WHO grade WHO 1 1 (2) 0 (0) 0 (0) 2 (40) 3 (6) WHO 2 24 (55) 1 (100) 1 (100) 3 (60) 29 (57) WHO 3 19 (43) 0 (0) 0 (0) 0 (0) 19 (37) Sum 44 (100) 1 (100) 1 (100) 5 (100) 51 (100) chi-square: 14.7 df: 6 P õ a Carcinoma involvement õ25% 5 (11) 0 () 0 (0) 1 (20) 6 (11) 26 50% 2 (4) 0 (0) 1 (20) 2 (40) 5 (10) 51 75% 8 (18) 1 (0) 0 (0) 2 (18) 11 (22) ú75% 29 (66) 0 (0) 0 (0) 0 (0) 29 (57) Sum 44 (100) 1 (100) 1 (100) 5 (100) 51 (100) chi-square: 23.5 df: 9 P õ b PIN: prostatic intraepithelial neoplasia; AAH: atypical adenomatous hyperplasia; WHO: World Health Organization; PIN 0: no prostatic intraepithelial neoplasia; ca: carcinoma; df: degrees of freedom. a The significant contribution to the chi-square was mainly from the difference between PIN 0 and PIN 3. b The contributions to the chi-square were mainly from the differences between PIN 0 and PIN 2 and PIN Specimens with PIN 1 and PIN 2 showed a lower noma was present. The presence of coincident PIN frequency of coincident carcinoma than specimens failed to influence the mean age of patients with AAH. with PIN 0. However, in all three types of material the Furthermore, the mean age of patients with WHO frequency of coincident carcinoma was highest for PIN Grade 1 3 carcinoma was not influenced by the pres- 3. Only 23% of specimens with extensive carcinoma ence of coincident PIN. involvement (ú75%) showed coincident PIN lesions. PIN was found in 14% of CB specimens with carci- DISCUSSION noma involvement, and in 29% of CB specimens without The diagnostic criteria for PIN 4 are now well estabimens carcinoma. PIN was present in 62% of TUR-P spec- lished, and were accepted by the Workshop on Pros- with carcinoma involvement, and in 57% of tatic Intraepithelial Neoplasia in However, dif- specimens without carcinoma. Further details regard- ferentiation of benign papillary epithelial proliferation ing CB are shown in Table 2. Table 3 shows the association in prostatic hyperplasia from stratified growth in low of PIN with AAH, WHO grade, and carcinoma grade PIN may be difficult. At a recent international involvement in TUR-P specimens. The association of conference on PIN there was agreement that currently PIN 3 with carcinoma was strong. Small carcinomas there are no accepted criteria for the separation of were associated with PIN Grade 2 and 3 more often intraductal spread of carcinoma and intraepithelial than large carcinomas (ú 75%), which were most com- changes similar to PIN. 9 The authors experience is monly associated with PIN 0. PIN was found in approximately similar. Therefore, they chose to require a separation 75% of TV specimens, with an equal distri- of 1 HPF between PIN and adjacent carcinoma so that bution in tissues with and without carcinoma. Details they would not overestimate PIN in the presence of regarding TV specimens are shown in Table 4. adjacent carcinoma. At this point, their criteria differ The mean patient age for the various categories is from those of Bostwick and Brawer. 4 shown in Table 5. For patients with PIN and coincident Interobserver variability in the diagnosis of PIN carcinoma, the mean age was 2.6 years higher than for has been reported to be lower than for carcinoma. patients with PIN but without coincident carcinoma. Allam et al. 21 obtained a kappa coefficient of 0.48 for Patients without PIN were 3.9 years older when carci- PIN and carcinoma, versus 0.85 for carcinoma only.

5 1176 CANCER March 15, 1997 / Volume 79 / Number 6 TABLE 3 Transuretheral Resections: Association with PIN and AAH, Total Carcinoma, WHO Grade, and Carcinoma Involvement (Number in parentheses is percent of total in group) PIN 0 PIN 1 PIN 2 PIN 3 Total No AAH and total carcinoma No AAH/no ca 199 (65) 46 (77) 139 (64) 62 (14) 446 (61) AAH 24 (8) 5 (7) 22 (10) 17 (11) 68 (9) Total cancer 82 (27) 9 (15) 55 (26) 71 (47) 217 (30) Sum 305 (100) 60 (100) 216 (100) 150 (100) 731 (100) chi-square Å 37.0 df: 6 P õ a WHO grade WHO 1 13 (16) 3 (33) 14 (25) 12 (17) 42 (19) WHO 2 30 (37) 3 (33) 29 (53) 35 (49) 97 (45) WHO 3 39 (47) 3 (33) 12 (22) 24 (34) 78 (36) Sum 82 (100) 9 (100) 55 (100) 71 (100) 217 (100) chi square Å 11.5 df: 6 P Å Carcinoma involvement õ25% 22 (27) 7 (78) 30 (55) 28 (39) 87 (40) 26 50% 9 (11) 1 (11) 13 (24) 11 (15) 34 (16) 51 75% 11 (13) 0 (0) 8 (14) 16 (23) 35 (16) ú75% 40 (49) 1 (11) 4 (7) 16 (23) 61 (28) Sum 82 (100) 9 (100) 55 (100) 71 (100) 217 (100) chi square Å 39.6 df: 9 P õ b PIN: prostatic intraepithelial neoplasia; AAH: atypical adenomatous hyperplasia; WHO: World Health Organization; PIN 0: no prostatic intraepithelial neoplasia; ca: carcinoma; df: degrees of freedom. a The significant contribution to the chi-square was mainly from the differences between PIN 3 and PIN 0, and PIN 1 and PIN 2. b The contributions to the chi-square were mainly from the differences between PIN 0 and PIN 1, PIN 2 and PIN 3. TABLE 4 Transvesical Resections: Association with PIN and AAH, Total Carcinoma, WHO Grade, and Carcinoma Involvement (Number in parentheses is percent of total in group) PIN 0 PIN 1 PIN 2 PIN 3 Total No AAH and total ca No AAH/ca 56 (73) 28 (82) 98 (67) 46 (68) 228 (70) AAH 10 (13) 3 (9) 30 (21) 11 (16) 54 (17) Total cancer 11 (14) 3 (9) 18 (12) 11 (16) 43 (13) Sum 77 (100) 34 (100) 146 (100) 68 (100) 325 (100) chi-square Å 5.36 df: 6 P Å WHO grade WHO 1 3 (28) 3 (100) 9 (50) 5 (45) 20 (47) WHO 2 4 (36) 0 (0) 9 (50) 6 (55) 19 (44) WHO 3 4 (36) 0 (0) 0 (0) 0 (0) 4 (9) Sum 11 (11) 3 (100) 18 (100) 11 (100) 43 (100) chi-square df: 6 P Å a Carcinoma involvement õ25% 5 (45) 3 (100) 17 (94) 9 (82) 34 (79) 26 50% 0 (0) 0 (0) 0 (0) 2 (18) 2 (5) 51 75% 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) ú75% 6 (55) 0 (0) 1 (6) 0 (0) 7 (16) Sum 11 (100) 3 (100) 18 (100) 11 (100) 43 (100) chi-square Å df: 6 P Å a PIN: prostatic intraepithelial neoplasia; AAH: atypical adenomatous hyperplasia; WHO: World Health Organization; PIN 0: no prostatic intraepithelial neoplasia; ca: carcinoma; df: degrees of freedom. a No special patterns in the contribution to the chi-square were observed.

6 PIN and Prostate Carcinoma/Skjørten et al TABLE 5 PIN, AAH, and carcinoma in three types of material: Mean Age and 95% Confidence Intervals for PIN with and without CB, TUR-P, and TV. PIN was found in approximately Coincident Carcinoma, and for AAH, WHO Grade, and Carcinoma 60% of the specimens, but with a very different preva- Involvement with and without Coincident Carcinoma lence, showing a low frequency in the CB specimens, Mean age (CI) Mean age (CI) but 3 to 4 times higher in TUR-P and TV specimens. No. (yrs) No. (yrs) The authors believe that the very low frequency of PIN in CB specimens is related to the small amount of No carcinoma Carcinoma tissue present, and to the fact that the majority of PIN PIN ( ) ( ) carcinoma in CB specimens were large. Epstein et al. 13 PIN ( ) ( ) in a small group of selected TUR-P material from Stage PIN ( ) ( ) A incidental carcinomas, found 15.6% with severe dys- PIN ( ) ( ) plasia (PIN 3). In the current study, TUR-P material PIN ( ) ( ) with small carcinomas (õ25% of section area involved, PIN 0 PIN 1 3 stage unknown) showed association with PIN 3 in 32%, AAH and total carcinoma which was not significantly different (P Å 0.07) from No AAH/no ca ( ) ( ) that of Epstein et al. 13 However, the authors diagnostic AAH ( ) ( ) criteria were more restrictive than those of Bostwick Total carcinoma ( ) ( ) and Brawer. 4 Therefore, their prevalence figures for PIN 0 PIN 1 3 PIN in the TUR-P zone were probably actually higher WHO grade than those of others. 13,24 A part of this discrepancy WHO ( ) ( ) might be explained by differences in mean age; the WHO ( ) ( ) mean age of the authors patients was 70 years versus WHO ( ) ( ) 63 years for the patients providing the material studied PIN 0 PIN 1 3 by Troncoso et al. 24 Sakr et al. 5 demonstrated that the Carcinoma involvement frequency of high grade PIN increases with age. An- No carcinoma ( ) ( ) other factor that might have contributed to a higher õ25% ( ) ( ) prevalence of PIN could be that most of the authors 26 50% ( ) ( ) patients underwent surgery for the relief of urinary 51 75% ( ) ( ) ú75% ( ) ( ) obstruction, and therefore had a more extensive pro- Total carcinoma ( ) ( ) cedure with removal of tissue from the peripheral as well as from the transition zone. However, the authors PIN: prostatic intraepithelial neoplasia; AAH: atypical adenomatous hyperplasia; WHO: World Health files do not contain information that would permit Organizaiton; CI: confidence interval; PIN 0: no prostatic intraepithelial neoplasia; ca: carcinoma. clarification of this point. In a similar study, Epstein et al. 22 reported a kappa The current study did show that PIN 3 was associated with coincident carcinoma in all types of material. No such association was found for PIN 2. Approxicoefficient of 0.61 for PIN and carcinoma. The authors mately 77% of large carcinomas (ú75% of the section findings compare favorably with these figures. area) were PIN negative, whereas 68% of smaller carci- Most studies on the prevalence of PIN have been nomas (õ 50% of the section area) showed either coincident performed on selected material, obtained either after PIN 2 or PIN 3. A similar proportion was obperformed total prostatectomy or at autopsy. McNeal and Bost- served for PIN and WHO grade; 61% of WHO Grade wick 23 studied 100 benign and 100 malignant prostate 3 carcinomas were without PIN. These findings may glands obtained at autopsy. They found intraductal represent the same phenomenon that was found for dysplasia (PIN) in 43% of benign and 82% of malignant CB specimens, large carcinomas reduce the chance prostates. Troncoso et al. 24 found PIN lesions in 72% of finding PIN. When PIN is present in high grade of benign prostate glands obtained at surgery for carcinoma carcinomas, PIN 3 predominates. The authors finddominantly of the bladder. PIN lesions were located pre- ings agree with those of De la Torre et al. 25 in the peripheral zone. Few such lesions The authors found a clear difference in the preva- were found in the central and transition zones. Ba- lence of carcinoma in three types of material studied; baian et al., 11 in a further study of the same material, 65% in CB, 30% in TUR-P, and 13% in TV. These differ- found that 70% of patients with transition zone carci- ences reflect the clinical situation leading to the choice noma had PIN 3 lesions but did not indicate the zonal of diagnostic or therapeutic procedure. In location of PIN 3. transrectal ultrasound was not available in the study The current report provides prevalence figures for groups collaborating hospitals. CB was performed

7 1178 CANCER March 15, 1997 / Volume 79 / Number 6 with digital guidance in patients in whom digital rectal the prostate glands in question. Mapping studies of examination had revealed suspicious nodules, or when total prostatectomy specimens have shown that the skeletal metastases had been found. CB yields tissue transition zone harbors only a part of the total number mainly from the peripheral part of the prostate gland. of foci of PIN and carcinoma present. 11,25 It is to be Large prostate carcinomas are frequent in the periph- expected that the remaining part of the gland also may eral zone. 10 When the biopsy needle is directed toward harbor a number of small, latent carcinomas that were a palpable lesion suspicious for carcinoma, it is to be undetected at the time of surgery. expected that most of the tissue core will be involved In the current study, the authors used the WHO by carcinoma tissue. In the current study, 80% of CB grading system, which in their experience has provided specimens with carcinoma involved ú50% of the area high interobsever reproducibility. 29,30 Greater of the tissue cores. No foci of atypical hyperplasia were than 80% of Grade 3 carcinomas involved 50% of found in the CB specimens. the sections studied, whereas only 3% of Grade 1 carci- TV specimens comprised approximately 25% of nomas were that large. In fact 95% involved õ25% of the material. TV had been chosen for patents with the sections. This was to be expected because large large glands without suspicion of malignancy. In addi- prostate carcinomas frequently are high grade. 26,28 tion to the low percentage of carcinoma in these specimens, Grade 3 carcinomas were frequent in CB and TUR-P as much as 79% of these carcinomas were small specimens (29% and 36%, respectively), whereas TV and low grade, in agreement with Bauer et al. 26 TUR- specimens had only 9% Grade 3 carcinoma, probably P was the procedure of choice at that time for most because TV was reserved mostly for glands with clinically patients with urinary obstruction caused either by benign enlargement. clinically benign or malignant prostatic enlargement. Several reports have shown that patient age is an TUR-P specimens showed a cancer frequency of 30%, independent risk factor for carcinoma of the prosof which 40% were small (õ 25%), and 28% were large tate. 29,31 In the current study, the mean age of patients carcinomas (ú75%). The frequency of small carcinomas without carcinoma was 69.1 years, versus 72.4 years was similar to the figure given for incidental tran- for patients with carcinoma. The mean age for patients sition zone cancers in TUR-P specimens. 8 with large carcinomas was higher than for patients Foci of AAH were found in 9% of TUR-P speci- with small carcinomas. The presence of PIN did not mens. Gaudin and Epstein 8 found 6% AAH in benign lead to higher mean age either in the presence or ab- TUR-Ps, whereas Qian and Bostwick 27 found AAH in sence of coincident carcinoma. These findings are 23% of total prostatectomies for carcinoma. As pointed similar to those of Lee et al. 32 The authors found a out by Bostwick et al., 6 the distinction between AAH clear difference between the mean age of patients with and small foci of low grade adenocarcinoma can at PIN and patients with carcinoma, patients with carci- times be difficult, particularly in TUR-P specimens if noma being 3 years older than patients with PIN. Lee the whole lesion is not present in the chip. The current et al. 32 found an age difference of 4 years, and Kovi et study is based on the criteria of a consensus state- al. 33 reported a 5-year age difference between patients ment. 7 When AAH was present in specimens with coincident with carcinoma and patients with PIN. These findings carcinoma, carcinoma was recorded and not indicate that PIN precedes carcinoma by several years. AAH. Therefore, the authors were unable to present A Workshop on Prostatic Intraepithelial Neoplafigures on the association of AAH and coincident carci- sia 20 recommended that PIN 2 and PIN 3 were combined noma. The data shown in the present report are figures to form high grade PIN, which was considered for the prevalence of AAH without coincident carci- to be a marker for coincident carcinoma. PIN 1 was noma. PIN 2 and PIN 3 were associated with AAH. called low grade PIN. In the current study, no sig- However, the authors intend to report a 20-year fol- nificant association was found between PIN 1, PIN 2, low-up for patients with PIN and AAH. and carcinoma. The current findings are in agreement TUR-P and TV specimens are mostly derived from with those of Aboseif et al., 14 who found a 94% positive the transition zone, into which peripheral zone carcinomas, predictive value for PIN 3 but only 50% for PIN 2 in when present, may extend after an unknown the diagnosis of coincident carcinoma. time interval. McNeal et al. 28 state that carcinomas In a recent review, Bostwick 1 examined the possi- that arise in the transition zone are frequently small bility of PIN being a precursor of invasive carcinoma, and incidental findings in TUR-P specimens. Because and cited strong evidence for such a development; PIN the TV material also comes mainly from the transition is associated with progressive abnormalities of phenotype zone, what has been said for TUR-P should also be and genotype, loss of markers of secretory differzone, valid for TV specimens. Neither CB, TUR-P, nor TV entiation, and an increase in markers usually associated specimens reflect the true prevalence of carcinoma in with malignant change. In spite of this, the evi-

8 PIN and Prostate Carcinoma/Skjørten et al dence that high grade PIN is a precursor of prostate 16. Garnett JE, Oyasu R. Urologic evaluation of atypical prostatic carcinoma needs to be proven by long term follow-up hyperplasia. Urology 1989;34(6 Suppl): Berner Aa, Danielsen HE, Pettersen EO, FossaM SD, Reith A, studies. In a article to be published, the authors will Nesland JM. DNA distribution in the prostate. Normal gland, report the results of a 20-year follow-up of the patient benign and premalignant lesions, and subsequent adenocarcinoma. material presented in this study. Anal Quant Cytol Histol 1993;15: Cohen J. A coefficient of agreement for nominal scales. Educ Psychol Meas 1960;20: REFERENCES 1. Bostwick DG. High grade prostatic intraepithelial neoplasia. 19. Fleiss JL. Statistical methods for rates and proportions. New The most likely precursor of prostate cancer. Cancer York: John Wiley and Sons, 1972: ;75: Drago JR, Mostofi FK, Lee F. Introductory remarks and work- 2. Davidson D, Bostwick DG, Quian J, Wollan PC, Oesterling shop summary. Urology 1989;34(6 Suppl):2 3. JE, Rudders RA, et al. Prostatic intraepithelial neoplasia is a 21. Allam CK, Bostwick DG, Hayes JA, Upton MP, Wade GG, risk factor for adenocarcinoma: predictive accuracy in neenosis of high grade prostatic intraepithelial neoplasia. Lab Domanowski GF, et al. Interobserver variability in the diag- dle biopsies. J Urol 1995;154: McNeal JE. Morphogenesis of prostatic carcinoma. Cancer Invest 1995;72:71A. 1965;18: Epstein JI, Grignon DJ, Humphrey PA, McNeal JE, Sester- 4. Bostwick DG, Brawer MK. Prostatic intra-epithelial neoplathe diagnosis of prostatic intraepithelial neoplasia (PIN). Lab henn IA, Troncoso P, et al. Interobserver reproducibility in sia and early invasion in prostate cancer. Cancer 1987; 59: Invest 1995;72:75A. 5. Sakr WA, Grignon DJ, Crissman JD, Heilbrun LK, Cassin BJ, 23. McNeal JE, Bostwick DG. Intraductal dysplasia: a premalig- Pontes JJE, et al. High grade prostatic intraepithelial neoplanant lesion of the prostate. Hum Pathol 1986;17: sia (HGPIN) and prostatic adenocarcinoma between the 24. Troncoso P, Babaian RJ, Ro JY, Grignon DJ, Eschenbach ACv, ages of 20 69: an autopsy study of 249 cases. In Vivo Ayala AG. Prostatic intraepithelial neoplasia and invasive 1994;8: prostatic adenocarcinoma in cystoprostatectomy speci- 6. Bostwick DG, Srigley J, Grignon D, Maksem J, Humphrey P, mens. Urology 1989;34(6 Suppl):52 6. Kwast THvd, et al. Atypical adenomatous hyperplasia of the 25. De la Torre M, Häggman M, Brändstedt S, Busch C. Prostatic prostate: morphologic criteria for its distinction from well intraepithelial neoplasia and invasive carcinoma in total differentiated carcinoma. Hum Pathol 1993;24: prostatectomy specimens: distribution, volumes and DNA 7. Bostwick DG, and collaborating pathologists. Consensus ploidy. Br J Urol 1993;72: statement on terminology: recommendation to use atypical 26. Bauer WC, McGavran MH, Carlin RM. Unsuspected carciadenomatous hyperplasia in place of adenosis of the prosmens. A clinocopathological study of 55 consecutive cases. noma of the prostate in suprapubic prostatectomy speci- tate [letter]. Am J Surg Pathol 1994;18: Gaudin PB, Epstein JI. Adenosis of the prostate. Histologic Cancer 1960;13: features in transurethral resection specimens. Am J Surg Patatic intraepithelial neoplasia and atypical adenomatous hy- 27. Qian J, Bostwick DG. The extent and zonal location of prosthol 1994;18: Montironi R, Bostwick DG, Bonkhoff H, Cockett ATK, Heltomy perplasia: relationship with carcinoma in radical prostatecpap B, Troncoso B, et al. International consultation on prostatic specimens. Pathol Res Pract 1995;191: intraepithelial neoplasia and pathologic staging of 28. McNeal JE, Redwine EA, Freiha FS, Stamey TA. Zonal distriprostatic carcinoma. Workgroup 1. Origins of prostate canlogic bution of prostatic adenocarcinoma. Correlation with histo- cer. Cancer 1996;78: pattern and direction of spread. Am J Surg Pathol 10. McNeal JE, Price HM, Redwine EA, Freiha FS, Stamey TA. 1988;12: Stage A versus stage B adenocarcinoma of the prostate: morcarcinoma: 29. Berner Aa, Harvei S, Tretli S, FossaM SD, Nesland JM. Prostatic phological comparison and biological significance. J Urol a multivariate analysis of prognostic factors. Br 1988;139:61 5. J Cancer 1994;69: Babaian RJ, Troncoso P, Ayala A. Transuretheral-resection 30. Jørgensen T, Yogesan K, Skjørten F, Berner A, Tveter KJ, zone prostate cancer detected at cystoprostatectomy. A deprognostic Danielsen HE. Histopathological grading and DNA ploidy as tailed histologic analysis and clinical implications. Cancer markers in metastatic prostatic carcinoma. Br J 1991;67: Cancer 1995;71: Helpap B. The biological significance of atypical hyperplasia 31. Partin AW, Steinberg GD, Pitcock RV, Lee W, Piantadosi S, of the prostate. Virchows Arch A Pathol Anat Histopathol Coffey DS, et al. Use of nuclear morphometry, Gleason histo- 1980;387: logic scoring, clinical stage and age to predict disease free 13. Epstein JI, Cho KR, Quinn BD. Relationship of severe dyspla- survival among patients with prostate cancer. Cancer sia to stage A (incidental) adenocarcinoma of the prostate. 1992;70: Cancer 1990;65: Lee F, Siders DB, Torp-Pedersen ST, Christensen-Day C, 14. Aboseif S, Shinohara K, Weidner N, Narayan P, Carroll PR. Carroll JT, Mitchell AE. Use of transrectal ultrasound and The significance of prostatic intra-epithelial neoplasia. Br J prostate-specific antigen in diagnosis of prostatic intraepi- Urol 1995;76: thelial neoplasia. Urology 1989;34(6 Suppl): Weinstein MH, Epstein JI. Significance of high-grade prosatypical 33. Kovi J, Mostofi FK, Heshmat MY, Enterline JP. Large acinar tatic intraepithelial neoplasia on needle biopsy. Hum Pathol 1993;24: hyperplasia and carcinoma of the prostate. Cancer 1988;61: