—  LONG COURSE #02  —

The Pathology of Prostate Cancer: From Population Studies to the Molecule
Moderators: Dr. John R. Srigley and Dr. Rodolfo Montironi

Section 8 - Unusual Carcinomas Of The Prostate

Peter A. Humphrey, MD, PhD
Professor of Pathology and Immunology
Washington University School of Medicine
St. Louis, MO


Introduction
The vast majority of prostate cancers are acinar adenocarcinomas. Unusual histologic variants or types of prostatic carcinoma account for about 5-10% of carcinomas that originate in the prostate gland [1, 2, 3]. These are distinguished from variants of acinar adenocarcinoma, which include atrophic adenocarcinoma, pseudohyperplastic adenocarcinoma, foamy gland carcinoma, colloid and signet ring adenocarcinoma, and oncocytic and lymphoepithelioma-like variants [2]. Distinctly uncommon to rare neoplasms in the prostate include adenomas, soft tissue (stromal) tumors, hematolymphoid neoplasms, miscellaneous neoplasms, and secondary neoplasms. Discussed here are unusual histologic variants of prostatic carcinoma.

Variants Of Prostatic Carcinoma

Ductal adenocarcinoma
Squamous cell and adenosquamous carcinoma
Urothelial carcinoma
Small cell carcinoma
Sarcomatoid carcinoma (carcinosarcoma)
Basal cell carcinoma

Ductal Adenocarcinoma
Ductal adenocarcinoma of the prostate [1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31] has also been termed endometrioid, endometrial, papillary, or papillary ductal adenocarcinoma. Recently, the very existence of prostatic ductal adenocarcinoma as a discrete clinico-pathologic entity has been questioned [4]. Here, the authors concluded that the only specific feature of ductal adenocarcinoma is the characteristic growth location, that being large periurethral ducts [4]. Given that most (but not all} series indicate a more aggresive clinical couse for ductal adenocarcinomas, it seems reasonable to diagnostically separate this morphologic variant from pure acinar adenocarcinoma.

Ductal adenocarcinoma is the most common histologic variant of prostatic carcinoma. The incidence of ductal adenocarcinoma, including both pure ductal and mixed ductal-acinar adenocarcinomas, is 3.2 % of all prostatic carcinomas [5, 6, 7, 8, 9]. Mixed ductal-acinar carcinoma is more common than pure ductal carcinoma. In the largest series (by far) of the ductal carcinomas in the literature (262 patients), the incidence of pure ductal carcinoma was 1.3%, while the incidence of mixed ductal-acinar adenocarcinoma was 4.8% [5].

Clinically, men with prostatic ductal adenocarcinoma are typically aged 63-72 years, with a range of 41-89 years. Obstruction and hematuria are common clinical manifestations. The digital rectal examination is usually abnormal and often suspicious for malignancy. Most patients have an elevated serum PSA above 4 ng/ml, with medians of 7.9 ng/ml to 13.8 ng/ml. A substantial minority of men with ductal prostatic adenocarcinoma can present with "metastatic" levels of serum PSA in the hundreds to thousands of ng/ml, bone pain and skeletal metastases. Indeed, clinical stage is more often advanced compared to standard acinar carcinomas [9, 10, 11].

The clinical macroscopic appearance of ductal adenocarcinoma by cystoscopy (urethroscopy) is in many cases that of an exophytic, villous/polypoid growth, with white fronds of "worm-like" tumor protruding into the urethra at or near the verumontanum. The prostatic urethra can also appear narrowed, nodular, or normal.

Grossly, ductal adenocarcinomas are often exophytic, but can be intraductal, and can spread throughout the entire prostate gland. This latter mode of permeation accounts for the high frequency of digital rectal examinations that are suspicious for malignancy. While ductal adenocarcinomas are classically centrally-located, they can be central and peripheral, or even isolated in the peripheral zone.

Microscopically, two forms of ductal adenocarcinoma growth have been noted. These have been defined as adenocarcinoma of the primary prostatic ducts (or type A) and adenocarcinoma of the secondary prostatic ducts (or type B) ,(5,10). These two growth patterns are not mutually exclusive, and can be mixed. Their main importance is in diagnostic recognition of ductal adenocarcinoma. The data are mixed as to whether or not these two patterns are associated with a different prognosis. Moreover, the dividing line between the two is often blurred, since the patterns merge. Thus, it does not appear necessary to distinguish between these two patterns.

Adenocarcinoma of the so-called primary ducts is an exuberant, complex and branching papillary adenocarcinoma that expands, projects into and/or fills large, central, periurethral prostatic ducts. A papillary ductal adenocarcinoma should be suspected if numerous cleft-like spaces and/or complex branching architectural patterns are visualized. The neoplastic cells are high columnar and may be single layered or multilayered. The nuclei are often basally-situated and elongated, with a fairly uniform degree of nucleomegaly, and true macronucleoli. Cytoplasm is abundant and usually eosinophilic to amphophilic, but can be vacuolated. Significant mucin production is not evident and psammoma bodies, which are often detected in papillary adenocarcinomas elsewhere, are infrequent.

Adenocarcinoma of so-called secondary prostatic ducts (type B) display an intraductal papillary proliferation with multicentric, complex glandular, solid and comedo patterns, often with intermediate to small-sized glandular spaces. A particularly distinctive finding is cylinders of tumor with eosinophilic luminal debris. This is not comedo necrosis since necrotic tumor cells are not present in the eosinophilic debris. True comedo carcinoma can be a component of ductal adenocarcinoma. Cribriform-papillary growth arrangements also typify this second form of prostatic ductal adenocarcinoma. An additional clue that one might be dealing with a prostatic ductal adenocarcinoma is the presence of high columnar or cuboidal cells with basal nuclei and abundant, clear cytoplasm. Another diagnostic hint is that the glands of ductal adenocarcinoma are often embedded within a fibrotic stroma, which is unusual for small acinar adenocarcinoma. Cytologically, the nuclear atypia in this adenocarcinoma of secondary ducts can be substantial, with large, pleomorphic, and hyperchromatic nuclei.

Mitotic activity in ductal carcinoma is variable, but is higher than in most acinar adenocarcinomas, where it can be difficult to find any mitoses at all.

Prostatic ductal adenocarcinoma has a significant in-situ component, with extension of the carcinoma inside pre-existing ducts and acini, with a discontinuous to continuous layer of basal cells.

In needle biopsy tissue of prostate papillary and/or cribriform structures are predominant [7]. In one-half of the cases, there is a minor co-existing acinar component, which was most often of a Gleason score of 6 or 7. The percentage of adenocarcinoma that is ductal in needle biopsy tissue correlates with the percentage of ductal cancer in radical prostatectomy tissues: When more than 50% of the adenocarcinoma is ductal in needle core tissue, then 80% of the corresponding radical prostatectomies have greater than 50% ductal adenocarcinoma [7].

In radical prostatectomy specimens, the ductal adenocarcinoma is comprised of confluent masses of papillary and/or cribriform adenocarcinoma, that are admixed with usual acinar adenocarcinoma. The ductal adenocarcinoma is almost always intimately admixed with acinar adenocarcinoma, with the proportion of the tumor showing ductal differentiation ranging from 5-100%. It is striking that in radical prostatectomy specimens ductal adenocarcinoma is predominantly located in the peripheral zone, with focal extension into the transition zone in only a minority of cases. The intraurethral and periurethral duct extension that is characteristic of primary duct or type A tumors is not seen in radical prostatectomy specimens. This is undoubtedly due to selection bias, where these patients were selected for this surgery based on peripheral carcinomas diagnosed by transrectal needle biopsy. Ductal adenocarcinoma in radical prostatectomy specimens tends to be of higher pathologic stage and larger tumor size, compared to pure acinar adenocarcinoma.

Ductal adenocarcinoma spreads outside the prostate gland in the same fashion as pure acinar adenocarcinoma. The papillary and/or cribriform growths can involve periprostatic soft tissue, seminal vesicles, pelvic lymph nodes, and distant sites, including lung and bone. Ductal adenocarcinoma appears to have a propensity to metastasize to testis, penis, and lung [12, 13]. Mixed ductal-acinar adenocarcinomas can penetrate outside the prostate gland and metastasize as pure ductal adenocarcinoma, pure acinar adenocarcinoma, or mixed ductal-acinar adenocarcinoma.

Ductal adenocarcinoma is admixed with usual acinar adenocarcinoma in a majority of cases. This does not represent dual carcinomas of the prostate in the same gland, but rather likely reflects the variability in growth of a single prostatic neoplasm. Although it has been written that ductal adenocarcinoma should be diagnosed when it is predominant, occupying more than 50% [24] or 80% of the tumor, and that adenocarcinoma with ductal differentiation or features should be diagnosed when less than 50% of the tumor is ductal, there is no evidence linking percentage ductal component to patient outcome. An approach this author favors would be to diagnose ductal adenocarcinoma if the neoplasm is pure ductal, and mixed ductal-acinar if both components are present, with a comment describing the percentage of the tumor that is ductal. Rare, reported admixtures of ductal adenocarcinoma with other histologic types of carcinomas include ductal adenocarcinoma with sarcomatoid carcinoma [22, 23], mucinous pools [10] and mucinous carcinoma [19], and urothelial carcinoma [20, 21].

The histologic grade of ductal adenocarcinoma is usually high-grade Gleason pattern 4 [27], but uncommonly pattern 3 and pattern 5 can be seen. For mixed ductal-acinar adenocarcinomas, separate grades should not be given, but rather a combined score, just as for pure acinar adenocarcinomas. The ductal component is usually of higher grade than the acinar proliferation.

The immunophenotype of prostatic ductal adenocarcinoma is similar to that of acinar adenocarcinoma. Although heterogenous in stain distribution and intensity, immunostains for PSA and PSAP are almost always positive, in both primary and metastatic sites. However, there are differences between ductal and acinar adenocarcinomas in the Ki67, 34betaE12, and alpha-methylacyl Coa racemase (AMACR) immunoprofiles. In accord with mitotic counting on H&E-stained sections, the Ki67 labeling index appears to be higher in ductal adenocarcinomas. The 34betaE12 immunohistochemical reaction with basal cells is different since basal cells are, by definition, absent in acinar adenocarcinoma, while some allow basal cells in the intraductal component of ductal adenocarcinoma. AMACR can be detected in ductal carcinoma [28, 29], but at a reduced level [29].

At the DNA level, DNA ploidy profiles of ductal adenocarcinoma do not differ substantially from acinar adenocarcinoma [16], but there is a difference in loss of heterozygosity [30].

The outcome for men with prostatic ductal adenocarcinoma is, in most studies, worse than outcome for men with prostatic acinar adenocarcinoma. Survival and response to therapy appear to be related to stage. Many patients with prostatic ductal adenocarcinoma present with large tumors and advanced stage, including bony metastasis; this may account for the relatively poor prognosis. Some patients respond to radical prostatectomy, hormonal therapy, and radiotherapy. Factors other than stage that predict outcome have not been well-characterized. In needle biopsy tissue only the extent of tumor correlated with time to biochemical failure [7].

Squamous Cell And Adenosquamous Carcinoma
Squamous cell carcinoma and adenosquamous carcinomas that originate in the prostate are rare [32, 33, 34, 35, 36, 37, 38, 39]. The largest series is of 33 cases [32]. In this series men with an average age of 68 presented with bladder outlet obstruction and dysuria. The incidence is less than 1% of all prostate carcinomas. The incidence will be dependent on how the patient population was treated, since many carcinomas in the prostate with squamous differentiation arise after radiation or hormonal therapy. The incidence of adenosquamous carcinoma is lower than that of pure squamous cell carcinoma, with only about 20 cases of adenosquamous carcinoma reported. In addition to admixture with adenocarcinoma, a few patients have had prostatic squamous cell carcinomas mixed with urothelial carcinoma and adenocarcinoma [35] and urothelial carcinoma [36].

The mechanism underlying the development of squamous cell carcinoma or adenosquamous carcinoma following radiation or androgen deprivation therapy is not known. Clearly, this pathway is generated in only very small minority of patients with adenocarcinoma of the prostate treated with hormonal therapy or radiotherapy. These treatment forms include DES, orchiectomy, LHRH analogs, androgen receptor blockers, external beam radiation, and radioactive seed implantation. The time course for the appearance of squamous differentiation in the carcinoma varies from 3 months to many (up to 9) years after therapy. Squamous metplasia and HPV infection do not appear to predispose to this development. It is not possible to completely exclude a second squamous malignancy in these cases, although a more plausible explanation would be clonal evolution/divergence of persistent carcinoma, secondary to the selective pressure of therapy. Finally, other pathways for squamous prostatic carcinogenesis must exist, since in some cases no therapy was administered prior to the diagnosis of squamous carcinoma and prostatic squamous cell carcinoma has been shown to occur in association with prostatic Schistosomiasis. No chemical carcinogen has been associated with prostatic squamous cell carcinoma in men.

Grossly, the tumors are usually large, measuring up to 6.5 mm in greatest dimension and often replacing a substantial portion of the prostate. Cut surfaces reveal a solid, firm, whitish-yellow, white-gray, to gray-tan mass. Central extension, with compression of prostatic urethra, and local invasion into the bladder, rectum, and seminal vesicle was macroscopically apparent in several cases. Microscopically, pure squamous cell carcinoma is typified by infiltrating nests, strands, and sheets of polygonal cells with nuclear atypia, with squamous differentiation manifested as individual cell keratinization, intercellular bridges, and/or keratin pearl formation. In examples of adenosquamous carcinoma, glandular and squamous components could be distinct or could exhibit direct transitions.

The application of a three-tiered grading scheme , with well, moderately, or poorly-differentiated categories, to squamous cell carcinomas of the prostate seems reasonable. The Gleason grading scheme can be used for the glandular component, but not the squamous component, of adenosquamous carcinomas. The adenocarcinoma element is often high-grade, while the grade of the squamous portion is variable.

Prostatic squamous cell carcinoma and adenosquamous carcinoma, like glandular adenocarcinomas, can spread along nerves, extend locally into periprostatic soft tissue, bladder and seminal vesicles, and metastasize to lymph nodes and bone. In bone, however, the metastases were routinely osteolytic rather than osteoblastic [33]. In widely-disseminated disease, metastatic deposits have been detected in glans penis, peritoneum, diaphragm, liver, and lung. Special studies are of limited use in diagnosis. The malignant squamous cells are most often negative for PSA and PSAP immunostains, and this correlates with the normal serum PSA and PSAP found in the vast majority of prostatic squamous cell carcinoma patients. The mean survival for prostatic squamous cell carcinoma is not long, at 6 to 24 months.

Urothelial (Transitional Cell) Carcinoma
Urothelial (also known as transitional cell) carcinoma in the prostate [40, 41, 42, 43, 44, 45, 46] is usually associated with a primary urinary bladder or urethral urothelial carcinoma. This association may be due to intraprostatic extension from the bladder or urethra, implantation from one of these sites, or multifocal disease as a result of a "field effect" of urothelial carcinogenesis, where urothelium normally lining the intraprostatic ducts might be affected. The argument has therefore been made that the terms primary and secondary should not be used for prostatic urothelial carcinoma [40]. Indeed, it is often difficult to know with certainty the primary site of origin and so prostatic urethral and periurethral prostatic ductal sites are often lumped together. Prostatic urothelial carcinomas diagnosed in the absence of a urinary bladder urothelial carcinoma are rare. The mean incidence in 9 series of 11,678 patients was 1.1% (range 0.2 – 3.5%) of all prostatic carcinomas [1].

No reliable gross features of prostatic transitional cell carcinoma have been reported. Microscopically, prostatic urothelial carcinoma has a striking propensity for growth within prostatic ducts and acini. Rounded or elongated cylinders and plugs of solid tumor within ducts are characteristic profiles. Comedo necrosis can also commonly be found, and may undergo dystrophic calcification. Partial involvement of benign prostatic glands is typical. Pagetoid spread can also be found. Intraductal papillary architecture is rarely detected. Prostatic stromal invasion by urothelial carcinoma is typified by irregular solid nests and cords that extend beyond the rounded, smooth outer profile of the urothelial carcinoma-in-situ. Cytologically, urothelial carcinoma in the prostate has a high nuclear grade with substantial nucleomegaly, nuclear pleomorphism, and nuclear hyperchromasia. The cytoplasm often has an eosinophlic "squamoid" appearance. Stromal sclerosis and inflammation are often pronounced.

The immunophenotypic profile of prostatic urothelial carcinoma is the same as for urothelial carcinomas elsewhere and is useful in the differential diagnostic distinction from poorly-differentiated prostatic adenocarcinoma. They are PSA and PSAP negative, and are frequently positive for the cytokeratins CK7, CK20, high molecular weight cytokeratins bound by antibody 34betaE12 (CK903), p63, and thrombmodulin. Prostatic urothelial carcinoma is thought to arise centrally in the large periurethral ducts via a pathway of urothelial dysplasia, with progression to carcinoma-in-situ. Since the neoplasm is more centrally located, in the past tissue diagnosis was mainly achieved by examination of TURP chips. However, urothelial carcinoma in the prostate can be extensive and both urothelial CIS and invasive carcinoma can extend into the peripheral zone, such that transrectal needle biopsy sampling of urothelial carcinoma is possible [45]. Spread of prostatic urothelial carcinoma within ducts and acini can be extensive, without stromal invasion. Prostatic stromal infiltration is associated with lymphvascular space invasion in a substantial percentage (44%) of cases, but of interest, perineural invasion is not frequent (at 14% of cases ) [42]. Many cases exhibit locally aggressive growth, with extension into the urinary bladder and pelvis. Lymph node, lung, and osteolytic (with a few osteoblastic) bone metastases have been identified.

Grading and pathologic staging have been inconsistently performed. The 2004 WHO low and high-grade carcinoma designations should be used. A large majority of cases are high-grade. The Gleason system should not be applied. In the past, the Whitmore-Jewett A-D scheme for adenocarcinoma was utilized to stage prostatic urothelial carcinoma. The current 2002 AJCC/ UICC TNM staging manual should be used, but here category T1 is confusing because the prostate lacks a well-defined subepithelial connective tissue compartment [46]. What is definitive is that the distinction between prostatic urothelial CIS and stromal invasion, with or without an associated bladder carcinoma, is critical for prognosis [44].

The outcome for patients diagnosed with primary urothelial carcinoma of the prostate has been dismal in the past, with an average survival of 17 to 29 months [1]. In contrast, it is abundantly clear that the prognosis for men with pure urothelial CIS in the prostate is excellent [40, 44].

Small Cell Carcinoma
Small cell carcinoma primary in the prostate [47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63] is a rare, extremely aggressive tumor that often presents with disseminated disease. Incidence estimates range from 0.3 % [47] to 1% [48] of all prostatic carcinomas. In about one-half of cases the carcinoma is pure small cell carcinoma, and in the other half there is a mixture with prostatic acinar adenocarcinoma. An intriguing and important aspect of prostatic small cell carcinoma diagnosis is that in one-third of patients there was an initial diagnosis of adenocarcinoma, followed by therapy, usually hormonal, in turn followed [at median of 18-49 months, range 7 months to 8 years [48, 49, 50] by a subsequent diagnosis of small cell carcinoma. This could relate to tissue sampling error and/or emergence/progression of an aggressive androgen-insensitive neuroendocrine clone related to therapeutic selection presence.

There are significant differences in the clinical features of prostatic small cell carcinoma vs. prostatic acinar adenocarcinoma. Distinctive clinical features include a lower percentage of men who present with an elevated serum PSA, despite presentation at advanced stages, the propensity for rapid and widespread metastasis, with visceral involvement, lack of hormonal responsiveness, and a short survival [48, 49, 51, 52, 53]. This aggressive clinical behavior is similar to small cell carcinomas arising at other anatomic sites.

Most patients are men aged 65-72 years of age [range 24 to 89]. The most frequent presenting symptoms are related to bladder outlet obstruction and disseminated disease [54]. In a few cases paraneoplastic syndromes have been reported; these include Cushing's syndrome (most common— due to tumoral ACTH production), hypercalcemia, SIADH, hyperparathyroidism, thyrotoxicosis, hyperglucagonemia, Eaton-Lambert syndrome, and limbic encephalitis [54]. The presence of such paraneoplastic syndromes in a patient with prostatic carcinoma should prompt a search, at the histological level, for a small cell carcinoma element, but these syndromes are not specific for the small cell histological type. One-third to two-thirds of patients with prostatic small cell carcinoma present with an elevated serum PSA; this elevation could be due to an admixed adenocarcinomatous component. The digital rectal examination is often suspicious for malignancy, with an enlarged, irregular, and stony-hard prostate. Another dramatic difference compared to prostatic acinar adenocarcinoma is that 92% of prostatic small cell carcinoma patients have advanced disease at presentation, and three-quarters of patients have metastatic disease [54]. The sites of clinically evident metastasis are also different, with spread to viscera, including liver and lung, being much more common in prostatic small cell carcinomas. Also, widespread dissemination and metastatic deposits in unusual locations [such as omentum, vocal cord, temporal bone, axillary lymph node, epididymis, pericardium and soft tissue] are more often seen with prostatic small cell carcinomas compared to prostatic acinar adenocarcinoma. Finally, the relatively low volume of osseous disease in the presence of visceral metastases should lead one to suspect small cell carcinoma [49].

Grossly, there is usually extensive involvement of the prostate, sometimes with replacement of the entire prostate by tumor. At autopsy, cut surfaces are gray-whitish and nodular and extraprostatic extension into seminal vesicles, periprostatic soft tissue, and bladder can be visualized. Microscopically, prostatic small cell carcinoma is identical to its more common counterpart in the lung, except that admixture with adenocarcinoma is much more frequent in the prostate. Histologically, prostatic small cell carcinoma grows as sheets, with ribbons, nesting, palisading along fibrous bands, and rosette-like structures occasionally noted. The polygonal, round, or spindled tumor cells exhibit only scanty cytoplasm, with hyperchromatic nuclei, "salt and pepper" stippled chromatin and inconspicuous nucleoli. Nuclear molding can be evident. Both individual cell necrosis, manifested by karryorhectic debris, and large areas of geographic necrosis can be seen. Encrustation of vessels with DNA (the Azzopardi phenomenon) can also be appreciated. Crush artifact is present in most cases. Mitoses are readily found. In one study [55], an average of 1-2 mitotic figures were detected per high-power field. Cytologically, prostatic small cell carcinoma has been diagnosed in pleural fluid and urine specimens.

The acinar adenocarcinoma component in mixed small cell-adenocarcinoma in the prostate is variable in grade and extent. The glandular component varies from focal to 70% [48] of the tumor volume. In most cases, particularly those where adenocarcinoma preceded a diagnosis of small cell carcinoma [50], the histologic grade of the adenocarcinoma portion was low-grade but high-grade Gleason pattern 4 was reported in about one-third of cases. In two-thirds of cases the malignant epithelium of the adenocarcinoma and small cell carcinoma merges directly into each other. In one-third of cases, there is physical separation of the two components. In metastatic deposits derived from primary mixed small cell carcinoma-adenocarcinoma, the small cell carcinoma element is often found, although both components can be seen. Rare examples of admixture of small cell carcinoma with malignant cells other than adenocarcinoma include co-existence with Paneth cell-like change, squamous cells, and a spindle cell sarcomatoid component.

The immunophenotypic profile of prostatic small cell carcinoma can be valuable in confirmation of the diagnosis. Epithelial and neuroendocrine differentiation is best demonstrated by applying a panel of antibodies, since no single marker is 100% sensitive. The prostatic markers PSA and PSAP are identified in only a small percentage (or 0% – as in this author's experience) of small cell carcinoma cases and so these immunostains are most often non-contributory is establishing a prostatic origin for a rare metastatic small cell carcinoma of unknown primary origin. The adenocarcinomatous component of mixed small cell carcinoma-adenocarcinoma is positive for PSA and PSAP in almost all cases. Immunohistochemical evidence of biosynthesis of a variety of hormones can be obtained, including ACTH, calcitonin-serotonin, parathormone, antidiuretic hormone, gastrin, vasoactive intestinal peptide, bombesin, somatostatin, and thyrotropin [54]. Expression of these bioactive molecules by prostatic small cell carcinoma can result in elevated serum levels in some cases, and production of a paraneoplastic syndrome in a few cases. One should note that prostatic small cell carcinoma expresses TTF-1, such that this marker is not useful in the distinction of prostatic and lung small cell carcinoma. Indeed, gene expression profiling of prostatic small cell carcinoma demonstrates a similar pattern of gene expression as small cell carcinoma of the lung.

The outcome for patients with small cell carcinoma of the prostate is dismal, with median average survival of 9 to 17 months. These are highly proliferative tumors [with a mean MIB-1 proliferation labeling index of 83%! [56]], which generate early, widespread metastases. The neoplasms are not responsive to hormonal therapy, but there may be an initial response to chemotherapy. Patient age, sex, primary tumor size, and stage were not determinants of survival [57].

The spectrum of neuroendocrine tumors in the prostate includes carcinoid tumors (extraordinarily rare), usual acinar adenocarcinoma with neuroendocrine differentiation defined by immunohistochemical staining, for chromogranin, for example (very common) [63], and a recently-characterized entity, large cell neuroendocrine carcinoma (rare) [58].

Sarcomatoid Carcinoma (Carcinosarcoma)
Sarcomatoid carcinoma is a rare biphasic malignancy in the prostate. Fewer than 60 cases have been reported in the literature [1]. Over one-half of the cases are found in the two largest series [64, 65]. The two elements of sarcomatoid carcinoma are a malignant epithelial (carcinomatous) component and a malignant mesenchymal or mesenchymal-like (sarcomatous) component. Sarcomatoid carcinoma may be either homologous, where the mesenchymal-like areas have the appearance of a undifferentiated sarcoma, or heterologous, where the sarcoma exhibits differentiation along the lines of specific mesenchymal cells such as bone and cartilage. Carcinosarcoma is a term that is synonymous with heterologous sarcomatoid carcinoma. Other terms that are equivalent to sarcomatoid carcinoma include metaplastic carcinoma, spindle cell carcinoma, and malignant mixed mesodermal tumor.

Clinically, most patients are around 70 years of age (range 50-89), and present with urinary tract obstruction and symptoms of frequency, urgency and nocturia. On digital rectal examination, the palpable prostate was enlarged, nodular, and hard. An equal number of men had normal and elevated serum PSA levels. Of note, in about half of all cases the initial diagnosis was a usual acinar adenocarcinoma, followed by hormonal and/or radiation therapy, with a subsequent diagnosis of sarcomatoid carcinoma.

Grossly, prostatic sarcomatoid carcinomas at pelvic exenteration, prostatectomy, and autopsy are large (5.5 to 18 cm), gray-white to pink masses with hemorrhage, necrosis and local extension into surrounding structures such as seminal vesicle and rectal wall.

Microscopically, the carcinomatous and sarcomatous components are admixed, with blending of the two in some areas. The carcinomatous element is almost always glandular and acinar. Two cases were ductal adenocarcinoma [67], three cases demonstrated squamous differentiation or adenosquamous carcinoma [66], and one recent case had urothelial and squamous components [68]. One case was mixed I have seen a case of mixed acinar, ductal, and squamous differentiation in a sarcomatoid carcinoma. The adenocarcinoma is typically high-grade, with a mean Gleason score of 9, and range of 7-10. A Gleason grade can be assigned only to the glandular component of sarcomatoid carcinoma. In most cases, the adenocarcinoma component was a minor component, but was dominant in one-third of cases. The sarcomatoid component often exhibits large areas of undifferentiated spindled and pleomorphic cells arranged in sheets, pinwheels, or fascicles. These patterns often resemble malignant fibrous histiocytoma. Cellularity is variable, from higher degrees to lower degrees in more hyalinized areas. Osteosarcoma and chondrosarcoma were by far the most frequently identified heterologous elements. Areas of fibrosarcoma, leiomyosarcoma, angiosarcoma, and rhabdomyosarcoma are rarely found. Different types of sarcoma can be found together in the same case. Cytologically, nuclear pleomorphism is moderate to marked, with numerous mitotic figures, including atypical ones, readily identified. Tumor giant cells, including osteoclast-like giant cells, can be seen.

Immunohistochemical stains and electron microscopy have demonstrated mesenchymal or epithelial differentiation in the sarcomatoid component. Epithelial markers (cytokeratin, PSA, PSAP) and muscle markers can be detected by immunohistochemistry in the malignant spindle cells. Vimentin immunostains are uniformly positive, and S-100 protein is consistently found in chondrosarcomatous regions. In one case, skeletal muscle and vascular differentiation was substantiated by positivity for myoglobin and Ulex Europeus I agglutinin, respectively [69].

The outcome of patients with prostatic sarcomatoid carcinoma is poor, with a median survival of 3 years and seven-year survival of 14%. The disease can be locally aggressive, with local recurrences and formation of large pelvic masses. Sites of metastasis include, in order of frequency, lung, bone, lymph nodes, and brain, with rare cases of metastatic spread to skin, liver, peritoneum, adrenal, pleura, and kidney. Both epithelial and/or mesenchymal components can be seen in metastatic deposits. Treatment has varied; nonsurgical therapy is ineffective. The histologic pattern does not predict progression and survival.

The development of sarcomatoid carcinoma after treatment of an acinar adenocarcinoma with androgen deprivation therapy or radiotherapy is provocative. In some of these cases, particularly those where the sarcomatoid carcinoma was diagnosed shortly (less than one year) after the pure adenocarcinoma, it is possible that the sarcomatoid part was just not sampled intially. However, there are at least 25 cases where sarcomatoid carcinoma was diagnosed one year or greater after acinar adenocarcinoma, and during the time interval there was hormonal treatment or radiotherapy. Indeed, the average time for development of sarcomatoid carcinoma after acinar adenocarcinoma is about three years. One could then argue that this might be treatment selection pressure, where evolution and progression of an aggressive sarcomatoid clone from the adenocarcinoma occurred. In support of this notion are the findings of morphologic transitions between the adenocarcinoma and sarcomatoid elements, the mixed epithelial and mesenchymal phenotype in the spindle cell population, and a recent case with clonal malignant epithelial and mesenchymal components [70]. The molecular mechanism(s) underlying such selection has not been elucidated, although a role of abnormal p53 protein accumulation has been suggested [71]. Certainly, treatment is not the sole trigger since sarcomatoid carcinomas can occur de novo. Moreover, it appears to be a rare stimulus since saromatoid carcinoma itself is so rare.

Basal Cell Carcinoma
Basal cell carcinoma of the prostate includes malignant basaloid proliferations (basaloid carcinomas) and also neoplasms that resemble, to a certain degree, adenoid cystic carcinomas of the salivary glands [73, 74, 75, 76, 77, 78, 79, 80]. A large number of terms have been utilized for these noeplasms and related growths such as adenoid basal cell tumor, adenoid cystic tumor, adenoid cystic-like tumor, basal cell carcinoma, and adenoid basal proliferation of uncertain significance. Some of these cases most likely represent adenoid cystic-like hyperplasia. The difficulty in classification of these proliferations resides in the facts that they are rare, there is not agreement on histologic criteria, and follow-up is available for only a few cases. Basal cell carcinomas are quite rare, with only about 50 reported cases. The gross features have been reported for fewer than 10 cases. In the largest series [73], the tumors were white and solid, sometimes with microcysts. Microscopically, several growth arrangements may be evident including large basaloid nests with peripheral palisading and necrosis, a florid basal cell hyperplasia-like pattern, and an adenoid basal cell hyperplasia-like pattern (adenoid cystic carcinoma pattern). Infiltrative permeation, extraprostatic extension, perineural invasion, necrosis, and stromal desmoplasia are characteristics of basal cell carcinoma that can help in the differential diagnosis distinction from basal cell hyperplasia. Immunohistochemical marker studies for bcl2 and Ki67 (for proliferation index) may also be of value in this separation [77]. Histologic grading of basal cell carcinoma is generally not done. Limited data on patient outcome have revealed a few cancer-specific deaths, indicating that basal cell carcinoma of the prostate is a potentially aggressive neoplasm.

Exceedingly Rare Variants
There are single case reports of renal-type clear cell carcinoma [81] and tubulocystic clear cell adenocarcinoma of the prostate [82].

References:
Ductal Adenocarcinoma
  1. Humphrey PA : Variants of Prostatic Carcinoma (Chapter 17) and Unusual Prostatic Neoplasms (Chapter 18). In: Prostate Pathology (ASCP Press, Chicago, 2003) 390-354.

  2. Tumours of the Prostate. In: Eble JN, Sauter G, Epstein JI, Sesterhenn IA, eds. Tumours of the Urinary System and Male Genital Organs (IARC Press: Lyon, France, 2004) 159-214.

  3. Young RH, Srigley JR, Amin MB, Ulbright TM, Cubilla AL :Variants of Prostatic Adenocarcinoma, Other Primary Carcinomas of Prostate, and Secondary Carcinomas (Chapter 5) and Miscellaneous Tumors of the Prostate (Chapter 6) In: Tumors of the Prostate Gland, Seminal Vesicles, Male Urethra, and Penis, Third Series, Fascicle 28 (Armed Forces Institute of Pathology, Washington DC, 2000) 217-288.

  4. Bock BJ, Bostwick DG : Does prostatic ductal adenocarcinoma exist? Am J Surg Pathol 23:781-785, 1999.

  5. Dube VE, Farrow GM, Greene LF: Prostatic adenocarcinoma of ductal origin. Cancer 32:402-409, 1973.

  6. Bostwick DG, Kindrachuk RW, Rouse RV: Prostatic adenocarcinoma with endometrioid features. Clinical, pathologic, and ultrastructural findings. Am J Surg Path 9:595-609,1985.

  7. Brinker DA, Potter SR, Epstein JI : Ductal adenocarcinom of the prostate diagnosed on needle biopsy. Correlation with clinical and radical prostatectomy findings and progression. Am J Surg Pathol 23:1471-1479, 1999.

  8. Epstein JI, Woodruff JM: Adenocarcinoma of the prostate with endometrioid features. A light microscopic and immunnohistochemical study of ten cases. Cancer 57:111-119,1986.

  9. Lemberger RJ, Bishop MC, Bates CP, Blundell W, Ansell ID: Carcinoma of the prostate of ductal origin. Br J Urol 56:706-709,1984.

  10. Ro JY, Ayala AG, Wishnow KI, Ordońez NG: Prostatic duct adenocarcinoma with endometrioid features: Immunohistochemical and electron microscopic study. Semin Diagn Pathol 5:301-311,1988.

  11. Tannenbaum M: Histopathology of the prostate gland. In: Tannenbaum M, ed: Urologic Pathology: The Prostate. Philadelphia: Lea and Febiger, 1977, pp.312-315.

  12. Copeland JN, Amin MB, Humphrey PA, et al : The morphologic spectrum of metastatic prostatic adenocarcinoma to the lung : special emphasis on histologic features overlapping with other pulmonary neoplasms. Am J Clin Pathol 117:552-557, 2002.

  13. Tu SM, Reyes A, Maa A, et al : Prostate carcinoma with testicular or penile metastases. Clinical, pathologic, and immunohistochemical features. Cancer 94:2610-2617, 2002.

  14. Carney JA, Kelalis PP: Endometrial carcinoma of the prostatic utricle. Am J Clin pathol 60:565-573,1973.

  15. Zaloudek C, Williams JW, Kempson RL: "Endometrial" adenocarcinoma of the prostate. A distinctive tumor of probably prostatic duct origin. Cancer 37:2255-2262,1976.

  16. Christensen WN, Steinberg G, Walsh PC, Epstein JI: Prostatic duct adenocarcinoma . Findings at radical prostatectomy. Cancer 67:2118-2124, 1991.

  17. Walker AN, Mills SE, Feihner RE, Perry JM: Endometrial adenocarcinoma of the prostatic urethra arising in a villous polyp. Arch Pathol Lab Med 106:624-627,1982.

  18. Sogbein SK, Steele AA: Papillary prostatic epithelial hyperplasia of the urethra: a cause of hematuria in young men. J Urol 142:1218-1220,1989.

  19. Nishimura K, Higashino M, Hara T, Oka T: [Prostatic adenocarcinoma showing features of endometrioid and mucinous carcinomas: a case report] [Japanese] Acta Urol Jpn 41:805-807,1995.

  20. Kopelson G, Harisiadis L, Romas NA, Veenema RJ, Tannenbaum M: Periurethral prostatic duct carcinoma. Clinical features and treatment results. Cancer 42:2894-2902,1978.

  21. Pacchioni D, Casetta G, Piovano M, et al : Prostatic duct carcinoma with combined prostatic duct adenocarcinoma and urothelial carcinoma features : report of a case. Int J Surg Pathol 12: 293-297, 2004.

  22. Millar EKA, Sharma NK, Lessels AM: Ductal (endometrioid) adenocarcinoma of the prostate: a clinicopathological study of 16 cases. Histopathology 29:11-19,1996.

  23. Ma TKF, Chapman WB, McLean M, Srigley J: Prostatic carcinosarcoma consisting of the unusual combination of ductal adenocarcinoma with osteogenic sarcoma: a report of a case and review of the literature. J Urol Pathol 8:111-120,1998.

  24. Samaratunga H, Singh M: Distribution pattern of basal cells detected by cytokeratin 34 Beta E12 in primary prostatic duct adenocarcinoma. Am J Surg Pathol 21:435-440,1997.

  25. Colpaert C, Gentens P, Van Marck E: Ductal ("endometrioid") adenocarcinoma of the prostate. Acta Urol Belg 66:29-32,1998.

  26. Oxley JD, Abbott CD, Gillatt DA, MacIver AG: Ductal carcinomas of the prostate: a clinicopathological and immunohistochemical study. Br J Urol 81:109-115,1998.

  27. Epstein JI, Allsbrook WC Jr, Amin MB, et al : The 2005 International Society of Urological Pathology (ISUP) Consensus Conference on Gleason grading of prostatic carcinoma. Am J Surg Pathol 29: 1228-1242, 2005.

  28. Beach R, Gown AM, DePeralta-Venturina MN, et al : P504S immunohistochemical detection in 405 prostatic specimens including 376 18-gauge needle biopsies. Am J Surg Pathol 26: 1588-1596, 2002.

  29. Kelemen K, Adley B, Yang XJ, Wang HL, Humphrey PA : Ductal adenocarcinoma of the prostate. A clinicopathologic study of 50 cases. Lab Invest 86:144A, 2006.

  30. Dawkins HJ, Sellner LN, Turbett GR, et al : Distinction between intraductal carcinoma of the prostate (IDC-P), high grade dysplasia (PIN), and invasive prostatic adenocarcinoma, using molecular markers of cancer progression. Prostate 44:265-270, 2000.

  31. Bosland MC, Ford H, Horton L: Induction at high incidence of ductal prostate adenocarcinomas in NBL/Cr and Sprague-Dawley Hsd: SD rats treated with a combination of testosterone and estadiol-17beta or diethylstilbestrol. Carcinogenesis 16:1311-1317,1995.
Squamous and adenosquamous carcinoma
  1. Parwani AV, Kronz JD, Genega EM, Gaudin P, Chang S, Epstein JI : Prostate carcinoma with squamous differentiation: an analysis of 33 cases. Am J Surg Pathol 28:651-657, 2004.

  2. Little NA, Wiener JS, Walther PJ, Paulson DF, Anderson EE : Squamous cell carcinoma of the prostate: 2 cases of a rare malignancy and review of the literature. J Urol 149:137-139, 1993

  3. Mohan H, Bal A, Punia RPS, et al : Squamous cell carcinoma of the prostate. Int J Urol 10: 114-116, 2003.

  4. Bennett RS, Edgerton EO, Mixed prostatic carcinoma. J Urol 110:561-563, 1973.

  5. Wernert N,Goebbels R, Bonkhoff H, Dhom G : Squamous cell carcinoma of the prostate. Histopathology 17:339-344, 1990.

  6. Wang Q, Lin C-S, Unger PP : Squamous cell carcinoma arising in hormonally treated adenocarcinoma of the prostate. Int J Surg Pathol 14:13-16, 1996.

  7. Miller VA, Reuter V, Scher HL : Primary squamous cell carcinoma of the prostate after radiation seed implantation for adenocarcinoma. Urology 46:111-113, 1995.

  8. Al Adnani MS : Schistosomiasis, metaplasia and squamous cell carcinoma of the prostate: Histogenesis of the squamous cancer cells determined by localization of specific markers. Neoplasma 32:613-622, 1985.
Urothelial (transitional cell) carcinoma
  1. Cheville JC : Urothelial Carcinoma of the Prostate. An immunohistochemical comparison with high grade prostatic adenocarcicoma and review of the literature. J Urol Pathol 9:141-154, 1998.

  2. Greene LF, O'Dea MJ, Dockerty MG : Primary transitional cell carcinoma of the prostate. J Urol 116:761-763, 1976.

  3. Goebbels R, Amberger L,Wernert N, Dhom G : Urothelial carcinoma of the prostate. Appl Pathol 3:242-254, 1965.

  4. Nicolaisen GS, Williams RD : Primary transitional cell carcinoma of prostate. Urology 24:544-549, 1984.

  5. Cheville JC, Dundore PA, Bostwick DG, et al : Transitional cell carcinoma of the prostate. Clinicopathologic study of 50 cases. Cancer 82:703-707, 1998.

  6. Oliai BR, Kahane H, Epstein JI: A clinicopathologic analysis of urothelial carcinomas diagnosed in prostate needle biopsy. Am J Surg Pathol 25:794-801, 2001.

  7. Greene FL, Page DL, Fleming ID, et al, Urethra. In: AJCC Cancer Staging Manual, 6th edition. (Springer Verlag, New York, 2002) 341-343.
Small cell and neuroendocrine carcinoma
  1. Losi L, Di Gregori C,Brausi M, Fante R, Hurlimann J: Expression of p53 protein in prostate cancers of different histologic types. Pathol Res Pract 190:384-388, 1994.

  2. Tętu B, Ro JY, Ayala AB, Johnson DE, Logothetis CJ, Ordonez NG: Small cell carcinoma of the prostate. Part 1. A clinicopathologic study of 20 cases. Cancer 59:1803-1809, 1987.

  3. Amato RJ, Logothetis CJ, Hallinan R, Ro JY, Sella A, Dexeus FH: Chemotherapy for small cell carcinoma of prostate origin. J Urol 147:935-937, 1992.

  4. Hindson DA, Knight LL, Ocker JM: Small-cell carcinoma of prostate. Transient complete remission with chemotherapy. Urology 26:182-184, 1985.

  5. Schwartz LH, Latrenta LR, Bonaccio E, Kelly WK, Scher HI, Panicek DM: Small cell and anaplastic prostate cancer-correlation between CT findings and prostate-specific antigen level. Radiology 208:735-738, 1998.

  6. Oesterling JE, Hanzeur CG, Farrow GM : Small cell anaplastic carcinoma in clinical, pathological, and immunological study of 27 patients. J Urol 147:804-807, 1992.

  7. Papandreou CN, Daliani DD, Thall PF, et al : Results of a phase II study with doxorubicin, etoposide, and cisplatin in patients with fully characterized small cell carcinoma of the prostate. Am J Surg Pathol 24: 1433-1436, 2000.

  8. Abbas F, Civantos F, Benedetto P, Soloway MS: Small cell carcinoma of the bladder and prostate. Urology 46:617-630, 1995.

  9. Christopher ME, Seftel AD, Sorenson K, Resnick MI: Small cell carcinoma of the genitourinary tract: An immunohistochemical, electron microscopic and clinicopathological study. J Urol 146:382-388, 1991.

  10. Helpap B, Kollermann J: Undifferentiated carcinoma of the prostate with small cell features: immunohistochemical subtyping and reflections on histogenesis. Virchows Archiv 434:385-391, 1999.

  11. Mackey JR, Au H-J, Hugh J, Venner P: Genitourinary small cell caricnoma: Determination of clinical and therapeutic factors associated with survival. J Urol 159:1624-1629, 1998.

  12. Evans AJ, Humphrey PA, Belani J, van der Kwast TH, Srigley JR : Large cell neuroendocrine carcinoma of the prostate. A clinicopathologic summary of 7 cases of a rare manifestation of advance prostate cancer. Am J Surg Pathol 2006, in press.

  13. Ro JY, Tętu B, Ayala AG, Ordóńez NG: Small cell carcinoma of the prostate. II. Immunohistochemical and electron microscopic studies of 18 cases. Cancer 59:977-982, 1987.

  14. Weaver MG, Abdul-Karim FW, Srigley JR: Paneth cell-like change and small cell carcinoma of the prostate. Two divergent forms of prostatic neuroendocrine differentiation. Am J Surg Pathol 16:1013-1016, 1992.

  15. Nguyen-Ho P, Nguyen G-K, Villaneuva RR; Small cell anaplastic carcinoma of the prostate: Report of a case with positive urine cytology. Diagn Cytopathol 10:159-161, 1994.

  16. Shah RB, Mehra R, Chinnaiyan AR, et al : Androgen-independent prostate cancer is a heterogeneous group of diseases : Lessons from a rapid autopsy program. Cancer Res 64: 9209-9216, 2004.

  17. Di Sant'Agnese PA: Divergent neuroendocrine differentiation in prostatic carcinoma. Semin Diagn Pathol 17:149-161, 2000.
Sarcomatoid carcinoma
  1. Dundore PA, Cheville CJ, Nascimento AG, Farrow GM, Bostwick DG: Carcinosacoma of the prostate. Report of 21 cases. Cancer 76:1035-1042, 1995.

  2. Shannon RL, Ro JY, Grignon DJ, Ordóńez NG, Johnson DE, Mackay B, Tętu B, Ayala AG: Sarcomatoid carcinoma of the prostate. A clinicopathologic study of 12 patients. Cancer 69:2676-2682, 1992.

  3. Berney DM, Ravi R, Baithun SI: Prostatic carcinosarcoma with squamous cell differentiation: A consequence of hormonal therapy: Report of two cases and review of the literature. J Urol Pathol 11:123-132, 1979.

  4. Poblet E, Gomez-Tierno A, Alfaro L: Prostatic carcinosarcoma: a case originating in a previous ductal adenocarcinoma of the prostate. Pathol Res Pract 196:569-572, 2000.

  5. Rogers CG, Parwani A, Tekes A, Schoenberg MP, Epstein JI : Carcinosarcoma of the prostate with urothelial and squamous components. J Urol 173: 439-440, 2005.

  6. Wick MR, Young RH, Malvesta R, Beebe DS, Hansen JJ, Dehner LP: Prostatic carcinosarcomas. Clinical, histologic, and immunohistochemical data on two cases, with a review of the literature. Am J Clin Pathol 92:131-139, 1989.

  7. Ray ME, Wojno KJ, Goldstein KJ, et al : Clonality of sarcomatous and carcinomatous elements in sarcomatoid carcinoma of the prostate. Urology 67:423.e5-423.e8, 2006.

  8. Delahunt B, Eble JN, Nacey JN, Grebe SK: Sarcomatoid carcinoma of the prostate: progression from adenocarcinoma associated with p53 over-expression. Anticancer Res 19:4279

  9. Reuter VE: Sarcomatoid lesions of the urogenital tract. Semin Diagn Pathol 10:188-201, 1993.
Basal cell carcinoma
  1. Iczkowski KA, Ferguson KL, Grier DD, et al : Adenoid cystic/basal cell carcinoma of the prostate. Am J Surg Pathol 27:1523-1529, 2003.

  2. Mastropasqua MG, Pruneri G, Renne G, De Cobelli O, Viale G : Basaloid cell carcinoma of the prostate. Virchows Archiv 443:787-791, 2003.

  3. Young RH, Frierson HF Jr, Mills SE, Kaiser JS, Talbot WH, Bhan AK, Adenoid cystic-like tumor of the prostate gland. Am J Clin Pathol 89:4956, 1998.

  4. Denholm SW, Webb JN, Howard G-CW, Chisholm GD : Basaloid carcinoma of the prostate gland: Histogenesis and review of the literature. Histopathology 20:151-155, 1992.

  5. Yang XJ, McEntee M, Epstein JI: Distinction of basaloid carcinoma of the prostate from benign basal cell lesions by using immunohistochemistry for bcl-2 and Ki-67. Hum Pathol 29:1447-1450, 1998.

  6. Grignon DJ, Ro JY, Ordóńez 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 19:1425-1433, 1988.

  7. McKenney JK, Amin MB, Srigley JR, et al : Basal cell proliferations other than usual basal cell hyperplasia : A clinicopathologic study of 23 cases, including four carcinomas, with a proposed classification. Am J Surg Pathol 28:1289-1298, 2004.

  8. Montironi R, Mazzucchelli R, Stramazzotti D, et al : Basal cell hyperplasia and basal cell carcinoma of the prostate : a comprehensive review and discussion of a case with c-erbB2 expression. J Clin Pathol 58:290-296, 2005.
Exceedingly rare variants
  1. Singh H, Flores-Sandoval N, Abrams J : Renal-type clear cell carcinoma occurring in the prostate. Am J Surg Pathol 27:4070410, 2003.

  2. Pan C-C, Chiang H, Chang Y-H, et al : Tubulocystic clear cell adenocarcinoma arising within the prostate. Am J Surg Pathol 24:1433-1436, 2000.