Morphologic criteria for the diagnosis of prostatic adenocarcinoma in needle biopsy specimens: a study of 250 consecutive cases in a routine surgical pathology practice. (Original Articles).
The diagnosis of prostate cancer is even more difficult if a very limited amount of tissue is available for evaluation. It has been estimated that the contemporary 18-gauge biopsy needle is only about 36% of the width of the traditional 14-gauge Tru-Cut biopsy needle. (9) Moreover, with the current clinical strategy of performing biopsies in patients with elevated serum prostate-specific antigen without abnormal digital rectal examination, the finding of small foci of cancer has become more frequent. Most of the diagnostic criteria and supportive clues for the diagnosis of prostate cancer have been proposed largely on the basis of observations in transurethral resections and radical prostatectomy specimens. (7,10,11) Relatively little information is available regarding these diagnostic histologic features within needle biopsy specimens. The few reports dealing with needle biopsies have been largely confined to analyses of selected cases that were diagnostically difficult. (12-16) To date, the incidence of these diagnostic features in benign glands has not been systematically studied.
In this study, we sought to determine the frequency with which several of these diagnostic or supportive features for prostate cancer occurred in an unselected, consecutively performed series of benign and malignant 18-gauge prostate needle core biopsies. It is well documented that there is an overlap in diagnostic features and supportive clues between prostatic adenocarcinoma and some benign small acinar proliferations, most notably atypical adenomatous hyperplasia, sclerosing adenosis, and postatrophic hyperplasia. (14,17-20) Thus, we decided to focus on observations in benign and malignant glands in general, rather than to restrict the study to small acinar proliferations.
MATERIALS AND METHODS
Eighteen-gauge prostate needle biopsy specimens from 150 consecutive cases diagnosed as cancer and 100 consecutive cases diagnosed as benign were retrieved from the surgical pathology files of Henry Ford Hospital, Detroit, Mich. In this institution, the urologists generally pool needle cores from different areas of the prostate into 2 or 3 parts, usually maintaining only orientation of right or left side. In view of the wide variation in specimen submission employed in clinical practice, we evaluated each case as a whole, without regard to separate parts. All tissue was routinely processed with 10% buffered formalin fixation, paraffin embedding, 4-[micro]m-thick sections, and standard hematoxylin-eosin staining.
The diagnoses were confirmed by review of hematoxylin-eosin--stained sections in all cases. All cancers were graded using the Gleason grading system. The incidence of the following features was analyzed in cancerous and noncancerous epithelium in each case: prominent nucleoli, multiple nucleoli, nucleolar margination, intraluminal crystalloids, intraluminal basophilic mucin, intraluminal amorphous eosinophilic material, collagenous micronodules, glomerulations, retraction clefts, perineural invasion, and glands infiltrating adipose tissue. No specialized techniques, such as morphometry for assessing nucleolar size or special stains for mucin, were employed.
Prominent nucleoli were identified as those readily visible with a x20 objective and by comparing them with nucleoli (if present) of adjacent benign glands. Nucleoli were further characterized as being central or marginated, the latter categorized when they were juxtaposed to the nuclear membrane. The presence of intraluminal, amorphous, eosinophilic material was ascertained with particular care to exclude decapitation secretions or fractured corpora amylacea. (15) Collagenous micronodules and glomerulations were identified as described in the literature. (3,7,13) Collagenous micronodules were recognized as nodular masses of paucicellular, aesinophilic, fibrillar stroma, which impinged on acinar lumens. Glomerulations consisted of rounded epithelial tufts within glands.
We had noted previously that retraction clefting, that is, separation of the acini from the adjacent stroma, occurs more frequently in malignant epithelium. (4) In the present study, retraction clefting was graded as a percentage of gland circumference separated from the stroma using a scale of 0 to 4+ as follows: 0 (absent), 1+ (<25%), 2+ (25%-50%), 3+ (50%-75%), or 4+ (>75%). Clefting at edges of a biopsy was disregarded as at least some of it could potentially be attributable as artifactual due to sectioning. Perineural invasion was noted as being present or absent.
Comparison between groups was made with the [chi square] test, using SPSS statistical software, version 7.5.1 (SPSS Inc, Chicago, Ill). For each variable studied, specificity, sensitivity, and positive and negative predictive values were also calculated.
The distribution of Gleason scores is shown in Table 1. Seventy-four (49.3%) of 150 tumors were moderately differentiated (Gleason 5-6), and 49 (32.7%) of 150 were moderately poorly differentiated (Gleason 7). One case (0.7%) was well differentiated (Gleason 4), and 26 cases (17.3%) were poorly differentiated (Gleason 8-10).
The results of the different histologic features evaluated in the 150 malignant and 100 benign cases are listed in Table 2. Sensitivity, specificity, and positive and negative predictive values are presented in Table 3.
Prominent nucleoli were present in 141 (94%) malignant cases (Figure 1, A) and in 25 (25%) benign cases (Figure 1, B). In the benign cases, they were noted only in occasional nuclei in a few glands and were seen in association with inflammation, basal cell hyperplasia, Paneth cell-like metaplasia, or atrophy, as shown in Table 4. Multiple nucleoli were observed in 96 (64%) malignant cases (Figure 1, A) and in none of the benign biopsies. Nucleolar margination was present in 132 (88%) cases of cancer (Figure 1, A) and 7 (7%) benign cases.
[FIGURE 1 OMITTED]
Crystalloids within malignant glands were present in 61 (40.6%) of 150 biopsy specimens with cancer (Figure 2, A) and were present in benign glands in 13 (9%) cases of cancer, either in the same or different needle cores. Those present in benign glands not associated with cancer were rare, being observed in only 1 case (Figure 2, B). Intraluminal blue-tinged mucinous material was found in 78 (52%) cases of cancer (Figure 3, A) and in none of the benign biopsies. One of the cancerous biopsies contained blue-tinged mucin within a few benign glands, even though no mucin was present in the malignant glands (Figure 3, B). Intraluminal, amorphous, eosinophilic material was a common finding in cancer, being present in 130 (86.7%) of such cases (Figure 4, A). It was present in only 2 (2%) of the benign cases (Figure 4, B). In 4 other cases, such material was observed within benign glands in biopsy specimens with cancer in a separate needle core.
[FIGURES 2-4 OMITTED]
Glomerulations were an infrequent finding, confined to cases with cancer, being present in 23 (15.3%) of such cases (Figure 5). Like glomerulations, collagenous micronodules were not observed in any of the benign biopsies, being found in 3 (2%) cases of cancer (Figure 6). All these cases were associated with obvious intraluminal blue-tinged mucin in malignant glands. Perineural invasion was present in 33 (22%) cancer biopsies and in none of the benign cases. In a single case of cancer, malignant glands were seen infiltrating adipose tissue.
[FIGURES 5-6 OMITTED]
The distribution of the various degrees of retraction clefting in the malignant cases is shown in Table 5. Although minor degrees of clefting (grades 1 and 2) were present with similar frequency in benign and malignant cases, grade 3-4 clefting was more frequent in malignant cases (38.7%) as opposed to benign cases (7%) (Figure 7).
[FIGURE 7 OMITTED]
In contemporary surgical pathology practice, one of the problem areas is making the diagnosis of prostate cancer based on very few atypical glands. Cases falling short of the threshold for the diagnosis of carcinoma are often descriptively diagnosed as atypical small acinar proliferations. (9,21) In evaluating the few atypical acini in question, the surgical pathologist has to critically evaluate all features that are useful in the diagnosis of carcinoma. Since an obvious infiltrative pattern is often lacking in such foci, there is greater reliance on other diagnostic and supportive features.
A number of diagnostic criteria and supportive and/or diagnostic features have been advocated to aid in making the diagnosis of prostate cancer. These have been studied extensively in transurethral resections, cystoprostatectomy and radical prostatectomy specimens, and autopsy material. (7,10,11) These criteria may be present only focally and often infrequently in these specimens, and so may be even less frequent in needle biopsies. Epstein (15) reported a detailed analysis of diagnostic criteria for "limited" adenocarcinoma in prostatic needle biopsies seen in consultation and documented the frequency of a variety of diagnostic features in 300 needle biopsy specimens with prostate cancer. However, due to the highly selected nature of consultation material, the incidence of these features would not be representative of prostate needle biopsies encountered in routine surgical pathology practice. For instance, cases without prominent nucleoli would be overrepresented and other more diagnostic features, such as perineural invasion, would be underrepresented.
The presence of prominent nucleoli was advocated as a diagnostic criterion of prostate cancer by Totten et al in 1953. (1) Nucleolar prominence has been variably defined as nucleolar size greater than 1 [micro]m, (22) greater than 1.6 [micro]m, (23) or even greater than 3 [micro]m. (24) This criterion is useful mainly in routine formalin-fixed sections, as nucleoli may be rendered more prominent in both benign and malignant cells when other fixatives are employed. (25) In routine practice, nucleolar prominence is usually assessed subjectively without resorting to nucleolar size measurement. Since differences in fixation and histologic processing influence the ease of recognition of nucleoli, the importance of comparing the nucleolar size in the suspect glands with benign prostatic epithelial cells elsewhere within the same specimen cannot be overemphasized. Nucleolar enlargement in secretory cells is now generally accepted as a cardinal diagnostic feature of prostatic adenocarcinoma. The presence of prominent nucleoli in at least some of the glands in question has even been considered a mandatory requirement for the diagnosis of microacinar adenocarcinoma if only a very small number of glands is present. Iczkowski and Bostwick, (14) in a study based on consultation material, found prominent nucleoli in 100% of the cases they diagnosed as adenocarcinoma, compared to 55% classified as atypical small acinar proliferation. In our study, performed in consecutive cases, prominent nucleoli were absent in approximately 6% of carcinomas. Hence, the diagnosis of cancer in prostate needle biopsies can occasionally be made in the absence of prominent nucleoli if other architectural and cytologic features of malignancy are present. In the series of prostate needle biopsies seen in consultation by Epstein, (15) as many as 24% of the cancers lacked prominent nucleoli. Some unusual variants of prostate cancer (eg, xanthomatous carcinoma), which were not represented in our series, are also known to be characterized by bland nuclear features without prominent nucleoli. (26) We found prominent nucleoli in 25% of benign cases, but this finding was generally very focal. Most of these prominent nucleoli were in areas of inflammation, basal cell hyperplasia, atrophy, or Paneth cell-like change.
In addition to nucleolar prominence, multiple nucleoli and nucleolar margination have also been suggested as diagnostic criteria for prostate cancer. Our experience is consistent with Helpap's assertion that the presence of multiple nucleoli is specific for cancer. (5) We did not find multiple nucleoli in benign glands. However, this feature lacked sensitivity in our experience and was found in only 64% of cancers. Nucleolar margination was more sensitive (present in 88% of cancers) but less specific, as it was found in 7% of our benign biopsies.
A relatively important, yet underemphasized criterion of malignancy in prostate cancer is the presence of nucleomegaly. The evaluation of this criterion is influenced by factors such as fixation, the thickness of the section, and staining, as well as others. Hence, although we rely on this criterion in a given case, where cancer or atypical glands and benign epithelium are present side by side, we did not formally evaluate nucleomegaly, as the presence of concurrent cancer and benign epithelium are important in judging this parameter and were not present in all cases.
Although nuclear changes in prostate cancer have received the most attention in the literature, intraluminal contents in malignant glands have been the focus of considerable interest in recent years as adjunctive clues to its diagnosis. Prostatic crystalloids are intraluminal, eosinophilic, refractile structures of varying size and shape, which are closely associated with prostate cancer, as first demonstrated by Holmes in 1977. (6) We found crystalloids in 40.6% of adenocarcinomas in 18-gauge needle biopsies. Other studies have noted a widely varying frequency of crystalloids in adenocarcinoma, ranging from 10% to 64.5%. (27,28) The higher values were from reports that used totally embedded cystoprostatectomy or autopsy specimens. (28) With the notable exception of atypical adenomatous hyperplasia or adenosis, crystalloids are rare in benign glands. The association of crystalloids with atypical adenomatous hyperplasia is well documented, with a reported frequency of up to 40%. (29) We found crystalloids in 1% of our benign cases, which did not include any examples of atypical adenomatous hyperplasia, a figure similar to that of Henneberry et al, (30) who reported crystalloids in 1.5% of 473 consecutive benign needle biopsy samples. Anton et al (31) found crystalloids in 5% of 634 benign biopsy specimens, while Hu et al (32) reported presence of crystalloids in 3.5% of 259 consecutive, community-based, benign needle biopsy specimens. We conclude that while the presence of crystalloids in suspicious glands is strongly suggestive of cancer, the morphology of the glands should be carefully assessed, as crystalloids have been reported in atypical adenomatous hyperplasia and sclerosing adenosis, and in rare instances in "ordinary" benign glands. (17,19,30,32)
The presence of intraluminal acidic mucin has been advocated as useful supportive evidence in the diagnosis of prostatic adenocarcinoma. Many earlier studies described positive Alcian blue staining in adenocarcinoma and rarity of such staining in benign glands. (8) However, in recent years acidic mucin has been demonstrated by histochemistry in a high proportion of atypical adenomatous hyperplasia (54%-63%) (33,34) and in sclerosing adenosis, basal cell hyperplasia, and mucinous metaplasia of benign prostatic glands. (20,35,36) Recently, it has been increasingly recognized that the presence of intraluminal mucin in amounts sufficient to be identified as blue-tinged wispy material on hematoxylin-eosin-stained sections is more specific for adenocarcinoma of the prostate. Gaudin and Epstein (17) found such material in only 3% of foci of atypical adenomatous hyperplasia (adenosis). The frequency of this finding in our malignant cases was 52%. Based on our consultation material, we have noted that the incidence and ease of recognition of blue mucin may vary between different pathology laboratories, possibly related to differences in staining technique. In our study, we did not find intraluminal blue wispy mucin in any of the benign biopsies using conventional light microscopy. However, in 1 case of cancer, intraluminal blue-tinged mucinous material was present in a benign gland adjacent to carcinoma. Once other benign small gland proliferations, such as atypical adenomatous hyperplasia and sclerosing adenosis, have been ruled out, we consider intraluminal blue-tinged material to be a very useful criterion for the diagnosis of prostate cancer in needle biopsies since it is specific (100%) and relatively sensitive (52%).
The presence of intraluminal, amorphous, eosinophilic material in malignant prostatic glands has been recognized in recent years. This has been found in 53% to 100% of cancers and only occasionally in benign glands. (15,37) This material is often found in relation to crystalloids, and it has been suggested that it may be the precursor to crystalloids. It is uncertain whether these are true secretions or are a manifestation of individual cell necrosis. (15) Care must be exercised to differentiate these secretions from decapitation secretions and fractured corpora amylacea in benign glands. We found intraluminal amorphous eosinophilic material in 86.7% of malignant cases and only rarely (2%) in benign biopsy specimens. Since such material is easily identified on low-power examination, this finding when visualized during scanning at low power warrants closer scrutiny of the glands containing it. This granular amorphous material must be distinguished from pale pink, eosinophilic, fluidlike secretions. Since 4 (66.6%) of the 6 cases in which intraluminal, eosinophilic, amorphous material was identified in benign glands had carcinoma in other needle cores, its presence in benign glands should initiate careful evaluation of all needle cores from such specimens. This interesting association needs further prospective study.
Collagenous micronodules are another relatively recently described histologic observation in prostate cancer. (3,38) These microscopic nodular aggregates of paucicellular eosinophilic fibrillar stroma are reported as being a specific, but infrequent, diagnostic clue in prostatic adenocarcinoma. (13) Our finding of a 2% frequency of collagenous micronodules in cancer and the absence of these micronodules in benign glands is similar to the experience of Bostwick et al, (3) who found such micronodules in 0.6% of biopsies with cancer and 12.7% of radical prostatectomies.
As in previous studies, (3,38) our cases were associated with mucin-secreting neoplastic glands. However, as noted by Baisden et al, (13) collagenous micronodules are very rarely the key feature in the diagnosis of limited cancer. Collagenous micronodules have not been reported in benign glands, benign prostatic hyperplasia, or prostatic intraepithelial neoplasia, but Iczkowski and Bostwick (14) described them in 2% of the cases they had diagnosed as atypical small acinar proliferations. We conclude that although apparently specific for carcinoma, collagenous micronodules are not very useful in routine practice owing to their extremely low frequency. In our experience, when present, they are invariably associated with conspicuous intraluminal blue mucin.
The presence of glands in a perineural location used to be considered tantamount to a diagnosis of malignancy, (13) but it has been reported with benign acini (39) and in atypical small acinar proliferations as well. (14) We believe that only circumferential growth or intraneural invasion should be regarded as pathognomonic of cancer. (19) The 22% frequency of perineural invasion in our series correlates well with the experience of Bastacky et al, (2) who reported perineural invasion in 20% of needle biopsies with cancer. The 3% frequency in Epstein's series of consultation cases appears to reflect the referral nature of that material. (15) Perineural invasion is generally associated with other histologic features of cancer. In a recent report, Baisden et al (13) noted that perineural invasion was the key diagnostic malignant feature in 0.07% of cancers in routine practice and in 0.6% of consultation cases.
The presence of glomeruloid structures in the prostate, characterized by round or oval epithelial tufts within acini, has been presented recently as a possible useful diagnostic adjunct for cancer, particularly in small specimens. (7) In this study, glomerulations were present in a significantly higher proportion of cancer cases (15.3%) than that reported by Pacelli et al, (7) who found a glomeruloid pattern in 4.5% of radical prostatectomies and in 3% of consecutive malignant needle biopsies. We did not observe this feature in any of the benign cases; thus, we concur that glomerulations may be a useful ancillary finding in prostate cancer.
Our group has observed retraction clefting around malignant glands of prostate cancer, (4) but this subject has otherwise received scant attention in the literature, except for a brief mention in a 1963 autopsy study by Halpert et al. (40) In a pilot study, we found clefting to be more common in malignant glands, compared to benign glands. (4) The mechanism of clefting in malignant glands is unknown, but may be related to an abnormality in the basement membrane or overexpression of collagenases and other enzymes required for invasion. In the present study, we systematically studied clefting and its extent by grading it based on the percentage of gland circumference involved. We found that minor degrees of clefting, while seen more often in cancer, lack specificity, as minor clefting is not uncommonly observed in benign glands. In contrast, more extensive clefting is much more specific for cancer but is rather infrequent, as it was seen in less than 40% of malignant cases.
The presence of suspect glands within adipose tissue is practically diagnostic of extraprostatic extension of prostatic adenocarcinoma. However, this feature is rarely found in needle biopsies and was observed in only 1 (0.67%) of the cases of cancer in this series. The finding of fat in the normal prostate gland is extraordinarily rare. (41)
In small architecturally atypical foci, which are being increasingly encountered in contemporary pathology practice, an infiltrative pattern is usually not evident, and tangentially cut secretory cells and stromal fibroblasts may mimic basal cells. A proposed useful, although not consistently present, feature on screening magnification is the amphophilia of cytoplasm in malignant glands. (15) Cytoplasmic features in malignancy vary from clear to amphophilic to eosinophilic, and while there is a marked overlap with the tinctorial qualities of benign epithelial cytoplasm, the sharp difference, if present, between benign and atypical/cancerous glands is a more useful feature than the cytoplasmic quality of individual atypical/ cancerous glands. In atypical foci, the nuclear features, including prominent nucleoli, especially if multiple or marginated against the nuclear membrane, are strongly indicative of prostate carcinoma. Furthermore, based on our experience, we suggest that while screening a prostate biopsy on low power for carcinoma (in addition to recognizing foci of glands of small caliber that are closely packed or that have different cytoplasmic quality), attention be paid to glands with circumferential stromal retraction or unusual intraglandular contents, such as blue mucin, crystalloids, or amorphous eosinophilic secretions. Table 6 provides a summary and compares the frequency of diagnostic clues and criteria for prostate cancer in different diagnostic or analytic settings as reported in the literature.
Finally, caution is warranted when interpreting statistical data. Although most of the features studied had a statistically significant difference in incidence in benign versus malignant glands, an associated low positive-predictive value or specificity could make this feature of limited utility in the diagnosis of an individual case. In contrast, some features in which statistical significance was not found (ie, fat invasion or collagenous micronodules), when present, are practically diagnostic of carcinoma in an individual case. Thus, the diagnosis of prostatic adenocarcinoma must be based on a constellation of features. Knowledge of the data emerging from this study would be useful in evaluating small foci of cancer in which a cumulative observation of these features weighs more strongly for a diagnosis of carcinoma.
Table 1. Distribution of Gleason Scores No. (%) of Biopsy Specimens Gleason Score With Malignant Glands (n = 150) 4 1 (0.7) 5 2 (1.3) 6 72 (48) 7 49 (32.7) 8 13 (8.7) 9 12 (8) 10 1 (0.7) Table 2. Frequency of Histologic Features Helpful in the Diagnosis of Adenocarcinoma of the Prostate in Needle Biopsy Specimens Frequency in Biopsies Frequency With Malignant in Benign Glands, Biopsies, P Value Histologic Feature No. (%) * No. (%) ([dagger]) Prominent nucleoli 141 (94) 25 (25) <.001 Marginated nucleoli 132 (88) 7 (7) <.001 Multiple nucleoli 96 (64) 0 <.001 Blue-tinged mucinous secretions 78 (52) 0 <.001 Intraluminal crystalloids 61 (40.6) 1 (1) <.001 Intraluminal amorphous eosinophilic material 130 (86.7) 2 (2) <.001 Collagenous micronodules 3 (2) 0 .28 Glomerulations 23 (15.3) 0 <.001 Perineural invasion 33 (22) 0 <.001 3+/4+ retraction clefting 58 (38.6) 7 (7) <.001 Invasion of adipose tissue 1 (0.7) 0 >.99 * Frequency in neoplastic acini only. ([dagger]) P values obtained using [chi square] test. Table 3. Sensitivity, Specificity, and Positive and Negative Predictive Values of Selected Histologic Criteria Sensi- Speci- Positive Negative tivi- fici- Predictive Predictive Histologic Feature ty, % ty, % Value, % Value, % Prominent nucleoli 94 75 85 89 Marginated nucleoli 88 93 95 84 Multiple nucleoli 64 100 100 65 Blue-tinged mucinous secretions 52 100 100 58 Intraluminal crystalloids 41 99 98 53 Intraluminal amorphous eosinophilic material 87 98 99 83 Collagenous micronodules 2 100 100 40 Glomerulations 15 100 100 40 Perineural invasion 22 100 100 46 3+/4+ retraction clefting 38 93 89 50 Invasion of adipose tissue 0.7 100 100 40 Table 4. Conditions Associated With Nucleoli in Benign Glands (n = 25) No. (%) of Cases With Condition Associated Nucleoli With Prominent Nucleoli in Benign Glands Basal cell hyperplasia 8 (32) Inflammation 4 (16) * Atrophy 3 (12) Paneth cell-like change 2 (8) None 10 (40) * One biopsy specimen had nucleoli in a focus of basal cell hyperplasia and inflammation, and another in a focus of Paneth cell-like change and inflammation. Table 5. Distribution of Retraction Clefting in Benign and Malignant Cases Frequency in Frequency in Grade of Malignant Biopsies, * Benign Biopsies, Retraction Clefting No. (%) No. (%) No clefting 25 (16.7) 41 (41) 1+ 18 (12) 12 (12) 2+ 49 (32.7) 40 (40) 3+ 44 (29.3) 7 (7) 4+ 14 (9.3) 0 * Frequency in neoplastic acini only. Table 6. A Comparison of the Frequency of Diagnostic Clues and Criteria for Prostate Cancer in Different Diagnostic or Analytic Settings * Incidence in Consecutive Incidence in Biopsies, % Diagnostically (Previous Studies) Difficult Cases of Feature Analyzed Benign Cancer Cancer, % Prominent nucleoli NA NA 76 (15) Marginated nucleoli NA NA NA Multiple nucleoli NA NA NA Blue-tinged mucinous secretions NA NA 34 (15) Intraluminal crystalloids 1.5 (30)-5 (31) 3.9 (32) 25 (15) Intraluminal amorphous eosinophilic material NA NA 53 (15) Collagenous micronodules NA 0.6 (3) NA Glomerulations NA 3 (7) NA Perineural invasion NA NA 3 (15) 3+/4+ retraction clefting NA NA NA Invasion of fat NA NA NA Incidence in Consecutive Biopsies, % (Present Study) Feature Analyzed Benign Cancer Prominent nucleoli 25 94 Marginated nucleoli 7 88 Multiple nucleoli 0 64 Blue-tinged mucinous secretions 0 52 Intraluminal crystalloids 1 40.6 Intraluminal amorphous eosinophilic material 2 86.7 Collagenous micronodules 0 2 Glomerulations 0 15.3 Perineural invasion 0 22 3+/4+ retraction clefting 7 38.6 Invasion of fat 0 0.7 Incidence Incidence in in Minimal Cancer, Feature Analyzed ASAP, % % Prominent nucleoli 55 (14) 64 (16)-100 (14) Marginated nucleoli NA NA Multiple nucleoli NA NA Blue-tinged mucinous secretions 6 (14) 18 (16)-33 (14) Intraluminal crystalloids 16 (14) 19 (14)-22 (16) Intraluminal amorphous eosinophilic material 66 (14) 72 (16)-73 (14) Collagenous micronodules 2 (14) 2 (16)-5 (14) Glomerulations NA NA Perineural invasion 6 (14) 0 (14)-2 (16) 3+/4+ retraction clefting NA NA Invasion of fat NA NA * NA indicates not available; ASAP, atypical small acinar proliferations.
(1.) Totten RS, Heinemann NW, Hudson PB, Sproul EE, Stout AP. Microscopic differential diagnosis of latent carcinoma of the prostate. Arch Pathol Lab Med. 1953;55:131-141.
(2.) Bastacky SI, Walsh PC, Epstein JI. Relationship between perineural tumor invasion on needle biopsy and radical prostatectomy capsular penetration in clinical stage B adenocarcinoma of the prostate. Am J Surg Pathol. 1993;17:336-341.
(3.) Bostwick DG, Wollan P, Adlakha K. Collagenous micronodules in prostate cancer: a specific but infrequent diagnostic finding. Arch Pathol Lab Med. 1995; 119:444-447.
(4.) Durham JR, Lee MW, Amin MB, Schultz DS, Stetler-Stevenson WG. Clefting in adenocarcinoma of the prostate gland [abstract]. Am J Clin Pathol. 1994;102: 536.
(5.) Helpap B. Observations on the number, size and localization of nucleoli in hyperplastic and neoplastic prostatic disease. Histopathology. 1988;13:203-211.
(6.) Holmes EJ. Crystalloids of prostatic carcinoma: relationship to Bence-Jones crystals. Cancer. 1977;39:2073-2080.
(7.) Pacelli A, Lopez-Beltran A, Egan AJ, Bostwick DG. Prostatic adenocarcinoma with glomeruloid features. Hunt Pathol. 1998;29:543-546.
(8.) Ro JY, Grignon DJ, Troncoso P, Ayala AG. Mucin in prostatic adenocarcinoma. Semin Diagn Pathol. 1988;5:273-283.
(9.) Iczkowski KA, MacLennan GT, Bostwick DG. Atypical small acinar proliferation suspicious for malignancy in prostate needle biopsies: clinical significance in 33 cases. Am J Surg Pathol. 1997;21:1489-1495.
(10.) Franks LM, O'Shea JD, Thomson ER. Mucin in the prostate: a histochemical study in normal glands, latent, clinical and colloid cancers. Cancer. 1964;17: 983-991.
(11.) Ro JY, Ayala AG, Ordonez NG, Cartwright J Jr, Mackay B. Intraluminal crystalloids in prostatic adenocarcinoma: immunohistochemical, electron microscopic, and x-ray microanalytic studies. Cancer. 1986;57:2397-2407.
(12.) Brat DJ, Wills ML, Lecksell KL, Epstein JI. How often are diagnostic features missed with less extensive histologic sampling of prostate needle biopsy specimens? Am J Surg Pathol. 1999;23:257-262.
(13.) Baisden BL, Kahane H, Epstein JI. Perineural invasion, mucinous fibroplasia, and glomerulations: diagnostic features of limited cancer on prostate needle biopsy. Am J Surg Pathol. 1999;23:918-924.
(14.) 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;124:98-107.
(15.) Epstein JI. Diagnostic criteria of limited adenocarcinoma of the prostate on needle biopsy. Hum Pathol. 1995;26:223-229.
(16.) Thorson P, Vollmer RT, Arcangeli C, Keetch DW, Humphrey PA. Minimal carcinoma in prostate needle biopsy specimens: diagnostic features and radical prostatectomy follow-up. Mod Pathol. 1998;11:543-551.
(17.) Gaudin PB, Epstein JI. Adenosis of the prostate: histologic features in needle biopsy specimens. Am J Surg Pathol. 1995;19:737-747.
(18.) Amin MB, Tamboli P, Varma M, Srigley JR. Postatrophic hyperplasia of the prostate gland: a detailed analysis of its morphology in needle biopsy specimens. Am J Surg Pathol. 1999;23:925-931.
(19.) Srigley JR, Bullock N, Amin MB. Small glandular patterns in the prostate: the differential diagnosis of small acinar carcinoma. In: Major Problems in Pathology. Philadelphia, Pa: WB Saunders; 1985:126-156.
(20.) Grignon DJ, Ro JY, Srigley JR, Troncoso P, Raymond AK, Ayala AG. Sclerosing adenosis of the prostate gland: a lesion showing myoepithelial differentiation. Am J Surg Pathol. 1992;16:383-391.
(21.) Cheville JC, Reznicek MJ, Bostwick DG. The focus of "atypical glands, suspicious for malignancy" in prostatic needle biopsy specimens: incidence, histologic features, and clinical follow-up of cases diagnosed in a community practice. Am J Clin Pathol. 1997;108:633-640.
(22.) Gleason DF. Atypical hyperplasia, benign hyperplasia and well differentiated adenocarcinoma of the prostate. Am J Surg Pathol. 1985;1985:53-67.
(23.) Kramer CE, Epstein JI. Nucleoli in low-grade prostate adenocarcinoma and adenosis. Hum Pathol. 1993;24:618-623.
(24.) Kelemen PR, Buschmann RJ, Weisz-Carrington P. Nucleolar prominence as a diagnostic variable in prostatic carcinoma. Cancer. 1990;65:1017-1020.
(25.) Bostwick DG. Evaluating prostate needle biopsy: therapeutic and prognostic importance. CA Cancer J Clin. 1997;47:297-319.
(26.) Nelson RS, Epstein JI. Prostatic carcinoma with abundant xanthomatous cytoplasm: foamy gland carcinoma. Am J Surg Pathol. 1996;20:419-426.
(27.) Jensen PE, Gardner WA Jr, Piserchia PV. Prostatic crystalloids: association with adenocarcinoma. Prostate. 1980;1:25-30.
(28.) Ro JY, Grignon DJ, Troncoso P, Ayala AG. Intraluminal crystalloids in whole-organ sections of prostate. Prostate. 1988;13:233-239.
(29.) Gaudin PB, Epstein JI. Adenosis of the prostate: histologic features in transurethral resection specimens. Am J Surg Pathol. 1994;18:863-870.
(30.) Henneberry JM, Kahane H, Humphrey PA, Keetch DW, Epstein JI. The significance of intraluminal crystalloids in benign prostatic glands on needle biopsy. Am J Surg Pathol. 1997;21:725-728.
(31.) Anton RC, Chakraborty S, Wheeler TM. The significance of intraluminal prostatic crystalloids in benign needle biopsies. Am J Surg Pathol. 1998;22:446-449.
(32.) Hu JC, Palapattu GS, Kattan MW, Scardino PT, Wheeler TM. The association of selected pathological features with prostate cancer in a single-needle biopsy accession. Hum Pathol. 1998;29:1536-1538.
(33.) Epstein JI, Fynheer J. Acidic mucin in the prostate: can it differentiate adenosis from adenocarcinoma? Hum Pathol. 1992;23:1321-1325.
(34.) Goldstein NS, Qian J, Bostwick DG. Mucin expression in atypical adenomatous hyperplasia of the prostate. Hum Pathol. 1995;26:887-891.
(35.) Grignon DJ, O'Malley FP. Mucinous metaplasia in the prostate gland. Am J Surg Pathol. 1993;17:287-290.
(36.) Grignon DJ, Ro JY, Ordonez NG, Ayala AG, Cleary KR. Basal cell hyperplasia, adenoid basal cell tumor, and adenoid cystic carcinoma of the prostate gland: an immunohistochemical study. Hum Pathol. 1988;19:1425-1433.
(37.) Egan AJ, Lopez-Beltran A, Bostwick DG. Prostatic adenocarcinoma with atrophic features: malignancy mimicking a benign process. Am J Surg Pathol. 1997;21:931-935.
(38.) McNeal JE, Alroy J, Villers A, Redwine EA, Freiha FS, Stamey TA. Mucinous differentiation in prostatic adenocarcinoma. Hum Pathol. 1991;22:979-988.
(39.) Carstens PH. Perineural glands in normal and hyperplastic prostates. J Urol. 1980;123:686-688.
(40.) Halpert B, Sheehan EA, Schmalhorst WR, Scott RJ. Carcinoma of the prostate: a survey of 5000 autopsies. Cancer. 1965;16:737-742.
(41.) Cohen RJ, Stables S. Intraprostatic fat [letter]. Hum Pathol. 1998;29:424-425.
Accepted for publication December 19, 2001.
From the Departments of Pathology, Henry Ford Hospital, Detroit, Mich (Drs Varma, Lee, Tamboli, and Zarbo), and Emory University Hospital, Atlanta, Ga (Drs Jimenez, Salles, and Amin).
Presented in part at the 87th Annual Meeting of the United States and Canadian Academy of Pathology, Boston, Mass, March 2-8, 1998.
Corresponding author: Mahul B. Amin, MD, Emory University Hospital, Department of Pathology, Room G169, 1364 Clifton Rd, Atlanta, GA 30322 (e-mail: email@example.com).
Reprints not available from the author.