P504S Immunostaining Boosts Diagnostic Resolution of Suspicious Foci in Prostatic Needle Biopsy Specimens

Transcription

1 Anatomic Pathology / P504S/AMACR IN PROSTATIC BIOPSY SPECIMENS P504S Immunostaining Boosts Diagnostic Resolution of Suspicious Foci in Prostatic Needle Biopsy Specimens Zhong Jiang, MD, 1* Kenneth A. Iczkowski, MD, 2* Bruce A. Woda, MD, 1 Maria Tretiakova, MD, PhD, 3 and Ximing J. Yang, MD, PhD 3* Key Words: Prostate cancer; Atypical small acinar proliferation; P504S; AMACR; α-methylacyl CoA racemase; Biopsy Abstract From 1.5% to 9.0% of prostatic needle biopsy specimens disclose atypical small acinar proliferations (ASAPs) suggestive of malignancy, carrying an approximate 45% predictive value for cancer. We applied keratin 34βE12 and P504S monoclonal immunostains to 93 cases that were judged as ASAP after H&E staining alone. Forty-one ASAP foci survived recutting for both immunostains. Three urologic pathologists independently assigned post keratin 34βE12 diagnoses of cancer, ASAP, highgrade prostatic intraepithelial neoplasia, or benign and then reviewed P504S slides and assigned final diagnoses. Eight foci (20%) were resolved unanimously after keratin 34βE12 staining; 18 (44%) were resolved by 1 or 2 evaluators and 29 (71%) by at least 1. According to whether post keratin 34βE12 ASAP designation was given by 3, 2, or 1 evaluator(s), P504S immunostaining unanimously resolved an additional 5 (12%), 10 (24%), or 23 (56%) of 41 ASAP foci and cumulatively, 31 foci (76%). Among 35 men (excluding 6 with cancer in other cores of the original biopsy), these immunostains could have permitted cancer diagnosis in 11 (31%), without repeated biopsy. Thus, the consensus diagnosis rate improved from poor to good after supplementing 34βE12 immunostaining with P504S. Since the advent of serum prostate-specific antigen screening approximately a decade ago, the frequency of prostatic needle biopsy and number of cores sampled have increased substantially, and biopsy remains the only diagnostic test for prostate cancer. In view of the serious consequences of misdiagnosing prostatic adenocarcinoma, it is prudent to give a definite diagnosis only when histologic evidence is unequivocal. While interpretation of the majority of prostatic needle biopsy specimens is straightforward, cases with only limited atypia provide a major diagnostic challenge. In 1.5% to 9.0% of biopsy specimens among unselected series, 1-6 one finds a localized proliferation of small acini that are suggestive of prostate cancer but lack the complete cytoarchitectural criteria. In such cases, pathologists use terms including atypical small acinar proliferation (ASAP) suggestive but not diagnostic of malignancy and often recommend repeated biopsy. Although it is not a discrete pathologic entity, ASAP is a valid diagnostic category implying absolute uncertainty regarding the diagnosis. ASAP includes a spectrum of subtotally diagnostic atypia, but most cases are classified as ASAP because of the small size of the abnormal focus. 7 Regardless of the rationale for its diagnosis, ASAP conveys a 42% to 45% predictive value for carcinoma on repeated biopsy. 8,9 Basal cells express high-molecular-weight cytokeratins recognized by the antibody clone 34βE12, a common ancillary diagnostic tool for the recognition of prostate cancer because of its absence in most cases. However, 34βE12 requires reliance on the absence of staining for diagnosis and may be limited by false-negative staining and lack of adequate internal positive control. Recently, Jiang et al 10,11 and others have shown that reactivity for P504S, the Am J Clin Pathol 2004;121:

2 Jiang et al / P504S/AMACR IN PROSTATIC BIOPSY SPECIMENS monoclonal antibody to α-methylacyl CoA racemase (AMACR), serves as a useful positive immunostain for prostatic adenocarcinoma. This racemase catalyzes peroxisomal β-oxidation of branched chain fatty acids. By using a series of prostatic biopsy specimens from our institutions diagnosed as suggestive of malignancy, we provide the first assessment of the added value of using P504S, together with keratin 34βE12, for reaching a definite diagnosis, with concurrence by 3 genitourinary specialist pathologists (Z.J., K.A.I., and X.J.Y.). Materials and Methods Case Selection and Immunohistochemical Analysis With institutional review board approval, we compiled 93 cases from February 1992 through July 2002 from the Gainesville VA Medical Center, Gainesville, FL (n = 33), University of Chicago Hospital, Chicago, IL (n = 37), and University of Massachusetts Memorial Hospital, Worcester (n = 23), that were classified as ASAP or suggestive of cancer. Diagnoses had been made by at least 1 general pathologist, without the use of immunostains. For patients who underwent repeated biopsy, results were noted through March Immunostaining was performed at the University of Massachusetts Medical School. Formalin-fixed, paraffinembedded tissue blocks were cut into 5-µm sections, transferred to glass slides, and treated with a 0.1-mol/L concentration of citrate buffer, ph 6.0, in an 800-W microwave oven for 15 minutes for antigen retrieval before immunostaining. The slides were stained on a TechMate 1000 (Ventana Medical Systems, Tucson, AZ) automated immunostainer using the avidin-biotin complex staining procedure. The sections were incubated with a rabbit monoclonal antibody (P504S, Corixa, Seattle, WA) to AMACR at a 0.5-µg/mL dilution or monoclonal antibody 34βE12 (DAKO, Carpinteria, CA) to high-molecular-weight cytokeratin at a dilution of 1:50 for 45 minutes followed by brief buffer washes. Sections then were incubated with a cocktail of biotinylated antirabbit IgG and antimouse IgG/IgM (Ventana) for 30 minutes. The sections were washed, incubated in avidin-peroxidase complex (Ventana) for 30 minutes, washed, and then reacted with diaminobenzidine and hydrogen peroxide to visualize the end product. The sections were counterstained with hematoxylin. Analysis of Immunostaining After the recutting process, the atypical focus or foci remained visible in both immunostained sections in only 43 cases (46%) because of the small size of the foci. These 43 cases disclosed 47 foci of ASAP, but 6 had to be excluded because all 3 pathologists considered 2 definitely benign and 4 definitely malignant. All remaining 41 ASAP foci were agreed on as ASAP by pathologists 1 and 3. These 41 foci constituted the group for which the diagnosis was reassessed by the evaluators after immunostain review. Pathologist 2 ventured definite diagnoses more often, leaving 17 as ASAP and assigning definite diagnoses to 24. The influence of immunostaining as a function of the H&E stain based diagnoses of pathologist 2 was subjected to a separate analysis (see the Results section). For each focus, the same H&E-stained slide(s) originally diagnosed as ASAP by a general pathologist were mailed, along with the immunostains, to the 3 urologic pathologists. Each pathologist evaluated the slides independently. The preliminary diagnosis based on H&E stains alone was recorded. Subsequently, the 34βE12 immunostains were scored as 2 (continuous basal cell staining pattern), 1 (discontinuous and scant basal cell staining pattern), or 0 (absent) and a post keratin 34βE12 diagnosis was recorded. P504S reactivity then was scored as +2 (strong continuous dark cytoplasmic staining), +1 (discontinuous dark cytoplasmic or weak diffuse cytoplasmic staining), or 0 (absent at 100, although higher magnification might disclose scant fine granular background staining). Evaluators recorded a final diagnosis for each focus based on both immunostains. Scores from the 2 immunostains were summed, yielding a range of sums from 2 (favor benign) to +2 (favor cancer). Immunostaining intensity scores were averaged from 3 sets of numbers to yield consensus scores. The number of post keratin 34βE12 diagnoses in each category was tabulated, to show whether ASAP diagnoses were resolved for only 1 or 2 or all 3 evaluators. After P504S evaluations, resolved foci were broken down into those for which the post keratin 34βE12 diagnosis had been ASAP for 1, 2, or 3 evaluators. The final diagnostic category was tabulated for each case, indicating which foci were resolved by all 3 evaluators. The no agreement category was applied if final diagnoses included 1 cancer diagnosis and 1 benign diagnosis. The κ statistic (κ = proportion of observed agreement minus the proportion of expected agreement divided by 1 minus the proportion of expected agreement) was calculated for each of 3 pairings of pathologists based on the final diagnoses of cancer, ASAP, high-grade prostatic intraepithelial neoplasia (PIN), and benign. The Cramer V coefficient for association between the score and diagnosis for each evaluator was calculated. The Mantel-Haenszel test for correlation was performed to determine whether the association between each evaluator s score and consensus diagnoses was significantly greater than zero. 100 Am J Clin Pathol 2004;121: Downloaded 100 from

3 Anatomic Pathology / ORIGINAL ARTICLE Results The 3 pathologists achieved good agreement on postimmunostaining diagnoses (for pairwise combinations, κ = 0.78, 0.72, and 0.63). Cases Resolved After Keratin 34βE12 Immunostain Alone After reviewing the keratin 34βE12 immunostaining of the 41 foci Table 1, all 3 pathologists reached definite diagnoses in 8 foci (20%), including benign prostate (4/41 [10%]), high-grade PIN (2/41 [5%]), and adenocarcinoma (2/41 [5%]). ASAP foci were resolved to the satisfaction of 2 of 3 evaluators in 10 additional cases (24%). Thus, as many as 18 (44%) of ASAP foci were resolved for a subset of at least 2 evaluators, although this subset varied from case to case. By individual evaluators, resolution to a non-asap diagnosis was achieved in 19 (46%), 19 (46%), and 23 (56%) of foci. Eleven cases were still considered ASAP by all 3 pathologists, and there was no agreement on 1 focus. Addition of P504S Immunostain Resolved More Cases Ten additional foci (24%) that were considered ASAP after keratin 34βE12 staining by 2 (5 foci) or 3 (5 foci) evaluators were resolved after consideration of the P504S result. Thirteen more foci (32%), which only 1 of 3 evaluators still considered ASAP, were resolved. Adding this 56% to the 8 foci resolved by keratin 34βE12 staining (20%), 31 (76%) of 41 foci were resolved. Thus, according to whether a post keratin 34βE12 ASAP opinion was given by 3, 2, or 1 evaluator, P504S immunostaining was responsible for resolution of an additional 5 (12%), 10 (24%), or 23 (56%) ASAP foci. Separate Analysis of Diagnoses by Pathologist 2 Pathologist 2 ventured definite diagnoses based on H&E-stained slides for 24 of 41 ASAP foci. While his own preimmunostaining diagnoses correctly predicted his final diagnosis (all of which agreed with group consensus) in 18 (75%) of these cases, the immunostains changed his final diagnosis in 6 cases (25%) Table 2. Two cancer diagnoses were changed to ASAP and 3 to benign; 1 benign diagnosis was changed to cancer. Resolution after immunostaining of the 17 foci that pathologist 2 initially considered ASAP is shown Table 3. Of the 17 ASAP foci, 11 (65%) were resolved unanimously to a definite diagnosis, 10 of which required a P504S result. If the 6 foci in which the diagnosis of pathologist 2 changed after immunostaining were added to the aforementioned 17 foci, 13 (57%) of 23 cases would have been resolved. Table 1 Group Resolution of Diagnoses of Suspicious ASAPs by Keratin 34βE12 Immunostain Alone and With Added Use of P504S for 41 H&E-Stained Biopsy Specimens * Resolved for 3/3 Evaluators by Adding Resolved After Keratin 34βE12 P504S When ASAP Was Diagnosed After Alone for These Evaluators Keratin 34βE12 by These Evaluators Unresolved Final Diagnosis 1/3 2/3 All 3 All 3 2/3 1/3 by Both Stains Total Benign (29) High-grade PIN (10) Cancer (37) ASAP 9 9 (22) No agreement 1 1 (2) Subtotal 11 (27) 10 (24) 8 (20) 5 (12) 5 (12) 13 (32) Concurrence 8 (20) 23 (56); 10 (24) 41 (100) cumulative 76% ASAP, atypical small acinar proliferation suggestive but not diagnostic of carcinoma ; PIN, prostatic intraepithelial neoplasia. * Values in parentheses are percentages. Table 2 Pathologist 2 Predicted the Diagnoses of 24 of 41 ASAP Cases on H&E Stain Alone * Final Consensus Diagnosis (for All Pathologists and Pathologist 2) Predicted Cancer High-Grade PIN ASAP Benign Cancer High-grade PIN Benign ASAP, atypical small acinar proliferation suggestive but not diagnostic of carcinoma ; PIN, prostatic intraepithelial neoplasia. * Numbers in bold indicate 6 diagnoses changed after immunostaining. Am J Clin Pathol 2004;121:

4 Jiang et al / P504S/AMACR IN PROSTATIC BIOPSY SPECIMENS Table 3 Immunostaining Resolution of 17 Suspicious ASAPs Based on Opinion of Pathologist 2 for H&E-Stained Slides * Resolved After Unresolved Final Diagnosis Keratin 34βE12 Alone Resolved After P504S by Both Stains Total No. (%) Benign (18) High grade PIN (18) Cancer (29) ASAP 6 6 (35) No agreement 0 0 (0) No. (%) resolved subtotal 1 (6) 10 (59) No. (%) unresolved 6 (35) 17 (100) ASAP, atypical small acinar proliferation suspicious for but not diagnostic of carcinoma ; PIN, prostatic intraepithelial neoplasia. * Resolution, 65% (11/17). Relation of Immunostaining Score to Diagnosis Table 4 shows the correlation of the immunostaining score ( 2 to +2) with the final consensus diagnosis. Cancer diagnoses Image 1 were correlated strongly with a positive score, with a mean of 1.4 (Cramer V coefficient, ; P <.001, all observers). Benign diagnoses Image 2 were correlated strongly with a negative score, with a mean of 1.25 (Cramer V coefficient, ; P <.001, all observers). Intermediate scores were seen for 4 cases with highgrade PIN (mean, 0.67). The distribution of immunostaining scores for foci with a final consensus diagnosis of high-grade PIN is shown (Table 4). In these 4 foci, the 34βE12 result usually was strong (in 3 foci), but the P504S immunostain was positive (in all 4 cases). One case of high-grade PIN was accompanied by adjacent ASAP and was counted as ASAP. Intermediate staining scores characterized 9 cases with final ASAP suggestive diagnoses (one accompanied by Table 4 Distributions of Immunostaining Scores in 40 Foci as a Function of Final Diagnosis * P504S Keratin 34βE foci diagnosed as benign foci diagnosed as high-grade PIN NA foci diagnosed as cancer foci diagnosed ASAP ASAP, atypical small acinar proliferation suggestive but not diagnostic of carcinoma ; NA, not available; PIN, prostatic intraepithelial neoplasia. * One ASAP focus was omitted because of no agreement on final diagnosis. high-grade PIN; mean, 0.22). Table 4 shows the distribution of immunostaining scores for 9 ASAP foci. The results for the 34βE12 immunostain were weak or patchy in 4 and negative in 5. The P504S stain was negative in 4 and weak in 4. Strong P504S reactivity in 1 case was confounded by a weak, equivocal 34βE12 result. The presence of acute inflammation was a diagnostically confounding factor in another case. In the illustrated ASAP case Image 3, the 34βE12 result was negative, P504S reactivity was weak and not circumferential around the lumen, and the focus of concern had ambiguous cytologic features. In the 1 case in which the pathologists reached no consensus Image 4, the 34βE12 result was negative but with very weak P504S reactivity (0 + 1 = 1). Repeated Biopsy Correlation With Postimmunostaining Diagnoses Findings on repeated biopsy or concomitant cores of the original biopsy specimen are summarized in Table 5. Six men had cancer present in other cores of the original specimen (first row of Table 5) and did not undergo repeated biopsy. All but 3 other patients underwent repeated biopsy before our study. Eleven men had been given postimmunostaining diagnoses of cancer, and 9 underwent repeated biopsy that revealed cancer. The other 2 men required revised reports because repeated biopsy results were benign in 1 case and unavailable in 1 case. Thus, among 35 men without cancer in other original cores, 11 (31%) could have had cancer diagnosed on the original biopsy specimen. Of 11 men with benign postimmunostaining diagnoses, repeated biopsy results were benign in only 5 (45%) and cancer in 4 (36%). Six men with ASAP underwent repeated biopsy, disclosing cancer in 2 (33%) and benign tissue in 4 (67%). Discussion We have demonstrated that addition of the P504S immunostain for AMACR resolved substantially more atypical 102 Am J Clin Pathol 2004;121: Downloaded 102 from

5 Anatomic Pathology / ORIGINAL ARTICLE A B C Image 1 Cancer was the final consensus diagnosis of 3 pathologists. A, Atypical small acinar proliferation suggestive but not diagnostic of carcinoma (ASAP). Three small atypical acini contain blue mucin and more amphophilic cytoplasm (H&E, 200). B, Nonreactive in the atypical acini; note the good internal positive control (keratin 34βE12, 200). C, Strong positivity in the atypical acini, in contrast with neighboring reactive, benign acini P504S, 200). A B C Image 2 Benign prostate was the final consensus diagnosis of 3 pathologists. A, Atypical small acinar proliferation suggestive but not diagnostic of carcinoma (ASAP). A basal cell layer is not apparent in these acini, and nuclei are slightly enlarged. However, corpora amylacea are present, favoring a benign diagnosis (H&E, 200). B, Patchy reactivity in most of the acini (keratin 34βE12, 200). C, Negative staining. Sparse grains of signal are noted focally, but this finding may be seen in benign acini (P504S, 200). small acinar proliferations suggestive of malignancy (ASAP) than did keratin 34βE12 immunostaining alone, based on achievement of consensus by 3 urologic pathologists. With keratin 34βE12, 46% to 56% (19/41-23/41) of the ASAP foci were resolved by individual pathologists, but all 3 evaluators reached a definite diagnosis in only 20% of cases (8/41). The addition of P504S raised the unanimous concurrence to 76% (31/41). Thus, the reproducibility of a definite diagnosis among evaluators was poor after 34βE12 alone; addition of a P504S result broadened the diagnostic consensus. ASAP diagnosis can be subjective, and our analysis was complicated by the preference of pathologist 2 to give definitive diagnoses on H&E-stained slides in 24 of 41 ASAP cases. By this approach, there was 75% accuracy (18/24 cases) for postimmunostaining consensus diagnosis, Am J Clin Pathol 2004;121:

6 Jiang et al / P504S/AMACR IN PROSTATIC BIOPSY SPECIMENS A B C Image 3 Suggestive of malignancy ( atypical small acinar proliferation suggestive but not diagnostic of carcinoma [ASAP]) was the final consensus diagnosis of 3 pathologists. A, ASAP. The small acini have a single cell layer and ambiguous cytologic features. Atrophy cannot be ruled out (H&E, 200). B, Negative staining in the atypical acini with the exception of a few acini on the edge of the core (keratin 34βE12, 200). C, Equivocal reactivity in some acini (P504S, 200). A B C Image 4 No consensus reached by 3 pathologists. A, Atypical small acinar proliferation suggestive but not diagnostic of carcinoma (ASAP). Small acini with ambiguous cytologic features suggest cancer or atypical adenomatous hyperplasia (H&E, 100). B, Negative staining in this area (keratin 34βE12, 200). C, Staining is very weak, leaving the diagnosis equivocal (P504S, 200). but 25% (6/24) of the definite diagnoses had to be modified after examination of the stains. If the starting cases were limited to the 17 that pathologist 2 considered ASAP, the percentage of poststaining resolution would fall from 76% (31/41) to 65% (11/17). This probably is because the 17 foci included as ASAP by this evaluator were particularly ambiguous at the outset. The change in 25% of the diagnoses by pathologist 2 demonstrated that even when the pathologist is fairly sure of an H&E stain based diagnosis, the ordering of both immunostains might sometimes change the interpretation. In general, positive reactivity with P504S and absent reactivity with 34βE12 was sufficient to convert ASAP foci to a cancer diagnosis; the inverse pattern supported a benign diagnosis, and reactivity for both was seen in high-grade PIN. With 1 exception, the 9 cases with final ASAP diagnoses and the 1 with no consensus all had in common absent or weak staining with both immunostains in the focus of 104 Am J Clin Pathol 2004;121: Downloaded 104 from

7 Anatomic Pathology / ORIGINAL ARTICLE Table 5 Correlation of 41 Diagnoses by Repeated Biopsy or Other Core(s) of Original Biopsy With Diagnosis After Keratin 34βE12 and P504S Immunostaining of ASAP Foci * Our Postimmunostaining Diagnoses Correlation Cancer High-Grade PIN Benign ASAP No Agreement Cancer in other core(s) Repeated biopsy Cancer High-grade PIN Benign ASAP No repeated biopsy or result not available Total ASAP, atypical small acinar proliferation suggestive but not diagnostic of carcinoma ; PIN, prostatic intraepithelial neoplasia. * Numbers in bold indicate identical diagnoses. Original pathology report revised. concern. Caution is in order, and thorough histologic evaluation of prostatic biopsy specimens with H&E staining using multiple levels cannot be replaced by P504S and 34βE12 immunostains because both have limitations. Cytokeratin 34βE12 has been reported to be useful for obtaining a definite diagnosis in ASAP cases. However, using 34βE12 alone as a negative marker for prostate cancer has several caveats. First, false-negative staining can be caused by poor preservation of tissue cores or technical difficulty in immunostaining. Second, acini with inflammatory atypia or postatrophic hyperplasia can be falsely negative, 15 and, rarely, cancer acini show false-positive reactivity. 16 It thus is well-established that a negative 34βE12 immunostaining result cannot alone support the diagnosis of cancer, 17 and, indeed, a negative result in a previous study of 295 ASAP cases was not predictive of cancer on repeated biopsy. 8 Other investigators have cautioned against using the 34βE12 result as a sole criterion for diagnosis 2,7,8,18 and stressed the need for a complementary, positive cancer marker. 15 P504S/AMACR, a new prostate cancer marker, first gained attention for its overexpression in malignant but not benign prostatic acini. 10,12-14 Jiang et al 11 previously reported that P504S is a sensitive and specific marker for small foci of prostate carcinoma. Our present study performed the first crucial test of P504S for use in prostatic needle biopsy specimens with an ASAP diagnosis. We validated the usefulness of P504S for foci of ASAP. Although our findings support P504S as a contributory marker for prostate carcinoma, critical interpretation is required because P504S immunostaining is not 100% sensitive for prostate cancer Negative P504S staining in small suspicious glands does not necessarily establish a benign diagnosis. Particular carcinoma variants have higher P504S false-negative reactivity rates than ordinary carcinoma: 27.6% of glands for the foamy gland variant 19 and 33% of cases for the atrophic variant. 20 A false-positive diagnosis also must be avoided, since 16% 11 to 21% 13 of benign specimens were reactive, although reactivity was weak and not circumferential. However, when focal atypical glands are observed and basal cells are not detectable, our findings suggest that positive staining for P504S can provide the pathologist with the additional confidence to establish a definite malignant diagnosis. Loss of the ASAP focus in the levels used for immunostaining was, as reported earlier, 9,21 the greatest limitation of immunostaining use. We recovered diagnostic tissue in only 46% of ASAP cases (43/93) for this study, owing to loss of the suspicious foci during the recutting process. The use of intervening unstained sections may improve the recovery rate, 21 but not to 100%. We do not expect that recovering diagnostic tissue in a higher percentage of cases would alter the approximate 24% of ASAP foci that could not be resolved after both P504S and 34βE12 immunostains. A second limitation of our study is that we did not evaluate the effect of emergence of a definite diagnosis on H&Estained levels 21 separately from the use of immunostains. However, given the frequent loss of ASAP foci on levels and the relatively small study group, this was not feasible. Third, we relied on expert genitourinary pathologists to evaluate the cytoarchitectural and immunohistochemical findings. Novis et al 22 studied 15,753 prostatic needle biopsy specimens and showed rates of diagnostic uncertainty ranging from 0% at the 10th percentile to 14% at the 90th percentile among 332 institutions. The usefulness of the 2 immunostains in permitting general pathologists to convert ASAP diagnoses to definite ones would be of interest to study. On the other hand, diagnostic rates of ASAP by general pathologists might be higher than those of expert urologic pathologists, which would generate different rates of conversion from uncertain to definite diagnosis. Nevertheless, further studies of a larger series of prostate biopsy specimens suggestive of malignancy might be warranted. Am J Clin Pathol 2004;121:

8 Jiang et al / P504S/AMACR IN PROSTATIC BIOPSY SPECIMENS In other series, 1-6 the 1.5% to 9.0% range in the rate of ASAP diagnosis might reflect some objective factors such as intensity of patient screening and biopsy sampling technique, but much of the variation might depend on the pathologist s judgment. P504S immunostain results substantially boost the percentage of definite diagnoses attainable, and addition of P504S to keratin 34βE12 in the workup of ASAP might spare men unnecessary repeated biopsy or delayed cancer treatment. In our study, 35 men did not have cancer in concomitant cores of the original biopsy; 11 (31%) of 35 men with postimmunostaining diagnoses of cancer would have received final diagnoses sooner without needing a repeated biopsy. An additional 5 men (14%) with benign postimmunostaining diagnoses might have been spared repeated biopsy that resulted in a benign diagnosis, although the serum prostatespecific antigen level and clinical suspicion could have warranted a repeated biopsy for some patients. At the University of Massachusetts Medical Center, the immunostains cost $200, whereas a repeated biopsy costs $900 ($300 to the urologist and $600 to the pathologist). Approximately 1 million prostatic needle biopsies are performed annually in the United States, so a decrease of 1% to 2% in the ASAP diagnosis rate would spare 10,000 to 20,000 men a second painful procedure. An ASAP diagnosis places men at risk for prostate cancer that is equally significant, by pathologic measures, as cancer diagnosed without a preceding suspicious ASAP. 23 If the diagnosis of ASAP persists after attempted use of both immunostains, we encourage repeated biopsy. Even a negative result on a repeated biopsy warrants follow-up, since the false-negative rate of repeated sextant biopsies is 23% 17 and the rate after repeating 8 to 12 biopsies, although probably smaller, is still not negligible. By using a combination of P504S/AMACR specific for prostatic adenocarcinoma and keratin 34βE12 specific for prostatic basal cells, a definite consensus diagnosis could be achieved in 76% (31/41) of prostate biopsy specimens suggestive of malignancy. For the 1 pathologist who applied ASAP more sparingly, 65% (11/17) of the resulting smaller group of ASAP cases could be resolved by the use of both stains. Therefore, more than half of diagnostically uncertain ASAP foci are resolvable by using P504S and keratin 34βE12, with potential not only to make the underdiagnosis of carcinoma less frequent but also to obviate the need for some repeat biopsies. From the Departments of 1 Pathology, University of Massachusetts Medical School, Worcester; 2 Pathology and Laboratory Medicine, Veterans Affairs Medical Center, Gainesville, FL, and Pathology, Immunology, and Laboratory Medicine, University of Florida, Gainesville; and 3 Pathology, University of Chicago, Chicago, IL. Address reprint requests to Dr Iczkowski: Pathology and Laboratory Medicine Service (113), VA Medical Center, 1601 SW Archer Rd, Gainesville, FL * The first 2 authors contributed equally to this work. Dr Yang is currently with the Department of Pathology, Northwestern Memorial Hospital, Chicago, IL. References 1. Bostwick DG, Qian J, Frankel K. The incidence of high grade prostatic intraepithelial neoplasia in needle biopsies. J Urol. 1995;154: Cheville JC, Reznicek MJ, Bostwick DG. The focus of atypical glands suspicious for malignancy in prostate needle biopsy specimens: incidence, histologic features, and clinical follow-up of cases diagnosed in a community practice. Am J Clin Pathol. 1997;108: Hoedemaeker RF, Kraanse R, Rietbergen JB, et al. Evaluation of prostate needle biopsies in a population-based screening study: the impact of borderline lesions. Cancer. 1999;85: Orozco R, O Dowd G, Kunnel B, et al. Observations on pathology trends in 62,537 prostate biopsies obtained from urology practices in the United States. Urology. 1998;51: Renshaw AA, Santis WF, Ritchie JP. Clinicopathologic characteristics of prostatic adenocarcinoma in men with atypical prostate needle biopsies. J Urol. 1998;159: Rogatsch H, Mairinger T, Gschwendtner A, et al. Optimized preembedding method improves the histologic yield of prostatic core needle biopsies. Prostate. 2000;42: Iczkowski KA, Bostwick DG. Criteria for biopsy diagnosis of minimal volume prostatic adenocarcinoma: analytic comparison with nondiagnostic but suspicious, atypical small acinar proliferation. Arch Pathol Lab Med. 2000;124: Iczkowski KA, Bassler TJ, Schwob VS, et al. Diagnosis of suspicious for malignancy in prostate biopsies: predictive value for cancer. Urology. 1998;51: Iczkowski KA, MacLennan GT, Bostwick DG. Atypical small acinar proliferation of the prostate suspicious for malignancy in needle biopsies: histologic features and clinical significance in 33 cases. Am J Surg Pathol. 1997;21: Jiang Z, Woda BA, Rock KL, et al. P504S: a new molecular marker for the detection of prostate carcinoma. Am J Surg Pathol. 2001;25: Jiang Z, Wu CL, Woda BA, et al. P504S/α-methylacyl-CoA racemase: a useful marker for diagnosis of small foci of prostatic carcinoma on needle biopsy. Am J Surg Pathol. 2002;26: Rubin MA, Zhou M, Dhanasekaran SM, et al. α-methyl coenzyme A racemase as a tissue biomarker for prostate cancer. JAMA. 2002;287: Beach R, Gown AM, de Peralta-Venturina MN, et al. P504S immunohistochemical detection in 405 prostatic specimens including gauge needle biopsies. Am J Surg Pathol. 2002;26: Zhou M, Chinnaiyan AM, Kleer CG, et al. Alpha methylacyl-coa racemase: a novel tumor marker overexpressed in several human cancers and their precursor lesions. Am J Surg Pathol. 2002;26: Ramnani DM, Bostwick DG. Basal cell specific antikeratin antibody 34βE12: optimizing its use in distinguishing benign prostate and cancer [editorial]. Mod Pathol. 1999;12: Am J Clin Pathol 2004;121: Downloaded 106 from

9 Anatomic Pathology / ORIGINAL ARTICLE 16. Varma M, Linden MD, Amin MB. Effect of formalin fixation and epitope retrieval on antibody 34βE12 immunostaining of prostatic tissues. Mod Pathol. 1999;12: Rabbani F, Stroumbakis N, Kava BR, et al. Incidence and clinical significance of false-negative sextant prostate biopsies. J Urol. 1998;159: Wojno KJ, Epstein JI. The utility of basal cell specific anticytokeratin antibody 34βE12 in the diagnosis of prostate cancer: a review of 228 cases. Am J Surg Pathol. 1995;19: Zhou M, Jiang Z, Epstein JI. Expression and diagnostic utility of alpha-methylacyl-coa-racemase (P504S) in foamy gland and pseudohyperplastic prostate cancer. Am J Surg Pathol. 2003;27: Farinola MA, Epstein JI. Utility of immunohistochemistry for alpha-methylacyl CoA racemase (AMACR) in distinguishing atrophic prostate cancer from benign atrophy [abstract]. Mod Pathol. 2003;16:149A. 21. Reyes AO, Humphrey PA. Diagnostic effect of complete histologic sampling of prostate needle biopsies. Am J Clin Pathol. 1998;109: Novis DA, Zarbo RJ, Valenstein PA. Diagnostic uncertainty expressed in prostate needle biopsies: a College of American Pathologists Q-Probes study of 15,753 prostate needle biopsies in 332 institutions. Arch Pathol Lab Med. 1999;123: Iczkowski KA, Chen HM, Yang XJ, et al. Prostate cancer diagnosed after initial biopsy with atypical small acinar proliferation suspicious for malignancy is similar to cancer found on initial biopsy. Urology. 2002;60: Am J Clin Pathol 2004;121: