Neoplasms of the Prostate

Prostatic Cysts

Giant multilocular prostatic cystadenoma is a large tumor composed of acini and cysts lined by prostatic-type epithelium set in a hypocellular fibrous stroma. This rare tumor arises in men between 28 and 80 years old as a large midline prostatic or extraprostatic mass causing urinary obstruction. The epithelial lining displays prostate-specific antigen (PSA) immunoreactivity. One case was associated with high-grade prostatic intraepithelial neoplasia (PIN). Surgical excision is usually curative, although it may recur if incompletely excised. The differential diagnosis includes phyllodes tumor, multilocular peritoneal inclusion cyst, multicystic mesothelioma, müllerian duct cyst, seminal vesicle cyst, lymphangioma, and hemangiopericytoma. The light microscopic appearance of the cyst lining is useful in separating these lesions ( Table 9.1 ). One reported case was successfully treated laparoscopically.

Other benign unilocular cysts that may be sampled by biopsy include seminal vesicle cyst, ejaculatory duct cyst, and müllerian duct cyst. Location is often useful, recognizing that seminal vesicle cyst is typically lateral, whereas müllerian duct cyst is midline. Seminal vesicle cyst may contain ectopic prostatic acini and urothelium. Echinococcal cyst is usually associated with prominent inflammation, and organisms are often demonstrable.

Ejaculatory Duct Adenofibroma

A case of benign adenofibroma of the ejaculatory duct has been reported. This incidental autopsy finding consisted of a small polypoid mass of epithelium and stroma that projected into a cystically dilated duct. Adenomatoid tumor of the ejaculatory duct has also been described.

Epidemiology of Prostatic Intraepithelial Neoplasia

The mean incidence rate of isolated high-grade PIN is 9% (range, 4% to 25%) of prostate biopsies. Given that there are an estimated 1,300,000 prostate biopsies performed annually, a reasonable estimate is that there are about 115,000 new cases annually of high-grade PIN without cancer and a prevalence of more than 16 million ( Table 9.3 ). The prevalence of PIN varies according to the population of men under study, with the lowest likelihood in those participating in PSA screening and early detection studies, with an incidence rate of PIN ranging from 0.7% to 20%.

The relationship between the number of cores sampled and the incidence of PIN on needle biopsy is controversial, although most agree that greater sampling increases the yield of both PIN and cancer. The incidence rate of PIN on 24-core saturation biopsy was 22% to 45%.

Those undergoing TURP have the highest likelihood of PIN, varying from 2.8% to 33%. In such cases, all tissue should be examined, but serial sections of suspicious foci are usually not necessary. Unfortunately, needle biopsies fail to show the suspicious focus on deeper levels in about one-half of cases, often precluding assessment by immunohistochemistry and compounding the diagnostic dilemma.

The prevalence and extent of PIN increase with patient age ( Table 9.3 ). An autopsy study of whole-mount prostates from older men showed that the prevalence of PIN in prostates with cancer increased with age, predating the onset of carcinoma by more than 5 years. A similar study revealed that PIN is first seen in men in their twenties and thirties (9% and 22% frequency, respectively), and preceded the onset of carcinoma by more than 10 years. Most foci of PIN in young men were low grade, with increasing frequency and volume of high-grade PIN with advancing age.

Race and geographic location also appear to influence the incidence of PIN after controlling for patient age. African American men have a greater prevalence of PIN than Caucasians in the 50- to 60-year-old age group, the decade preceding detection of most prostate cancers. African American men also had the highest incidence of cancer (about 50% more than Caucasians). In contrast, Japanese men living in Japan had a lower incidence of PIN than those residing in the United States, and Asians had the lowest clinically detected rate of prostate cancer. Interestingly, Japanese men diagnosed with PIN also had an increased likelihood of development of prostate cancer, indicating that PIN is a precursor of clinical prostate cancer in Asian men. Differences in the frequency of PIN in the 50- to 60-year-old age group across races essentially mirror the rates of clinical prostate cancer observed in the 60- to 70-year-old age group.

The likely causal association of PIN with adenocarcinoma is supported by the observation that the prevalence of both increase with patient age, and that PIN precedes the onset of prostate cancer by less than one decade ( Table 9.4 ). The severity and frequency of PIN in prostates with cancer are greatly increased when compared with prostates without cancer (73% versus 32% frequency, respectively). High-grade PIN in sextant biopsy carries a 50% risk for carcinoma on subsequent biopsy within 3 years, although this risk was lower when more than six cores are obtained; this decline in predictive value is expected given the increased sampling for cancer with a greater number of core biopsies (see later).

Diagnosis of Prostatic Intraepithelial Neoplasia

PIN is characterized by cellular proliferation within preexisting ducts and acini, with cytologic changes mimicking cancer, including nuclear and nucleolar enlargement ( Figs. 9.1 to 9.3 ). There is inversion of the normal orientation of epithelial proliferation from the basal cell compartment to the luminal surface, similar to adenoma in the colon. The requisite nucleolar enlargement must be present in at least 10% of cells within the focus to be considered diagnostic; however, overstaining, hyperchromasia, and nuclear overlap may obscure the nucleolar features. Overexpression of the MYC oncogene is responsible for increased nucleolar number and size in PIN and cancer.

There are four main patterns of high-grade PIN: tufting, micropapillary, cribriform, and flat ( Figs. 9.1 to 9.3 ). The tufting pattern is the most common, present in 97% of cases, although most cases have multiple patterns. There are no known clinically important differences between the architectural patterns, and their recognition appears to be only of diagnostic utility. Sporadic retrospective reports have suggested that the cribriform or micropapillary patterns may indicate higher risk for coexistent cancer, but this has been refuted. Other unusual patterns of PIN include the signet ring cell pattern, small cell pattern, mucinous pattern, microvacuolated (foamy-gland) pattern, inverted (hobnail) pattern, and PIN with squamous differentiation ( Fig. 9.4 ; Table 9.5 ). The small cell pattern is usually negative for neuroendocrine markers and does not appear to be a precursor of small cell carcinoma. The hobnail pattern usually coexists with Gleason score 7 cancer.

The presence of extensive PIN appears to be more predictive of cancer than the more common isolated single acinus with PIN (see later). The presence of intraluminal crystalloids in combination with PIN is a more compelling indication for repeat biopsy than PIN alone.

PIN spreads through prostatic ducts in multiple different patterns, similar to carcinoma. In the first pattern, neoplastic cells replace the normal luminal secretory epithelium, with preservation of the basal cell layer and basement membrane. This pattern often has a cribriform or near-solid appearance. Foci of high-grade PIN may be difficult to distinguish from intraductal/intraacinar spread of carcinoma by routine light microscopy (see later). In the second pattern, there is direct invasion through the ductal or acinar wall, with disruption of the basal cell layer. In the third pattern, neoplastic cells invaginate between the basal cell layer and columnar secretory cell layer (“pagetoid spread”), a rare finding. Proliferative activity according to Ki67 labeling is lower in PIN (mean, 6%; range, 2% to 15%) than in ductal adenocarcinoma; the combination of histologic features and measurements of cellular proliferation help to distinguish these findings in limited tissue samples.

Ultrastructurally, high-grade PIN displays features between those of benign epithelium and adenocarcinoma. These include the presence of cells with a variable number of cytoplasmic secretory vacuoles, luminal apocrine blebs, large nuclei with coarsely clumped chromatin, enlarged nucleoli, prominent apical microvilli, intact or discontinuous basal cell layer, and intact basement membrane. Occasional acini have luminal cells abutting the basement membrane without interposition of basal cells, and other acini with extremely attenuated basal cell cytoplasmic processes contain bundles of intermediate filaments.

Early stromal invasion, the earliest evidence of carcinoma, occurs at sites of acinar outpouching and basal cell disruption in acini with high-grade PIN ( Figs. 9.5 and 9.6 ). Such microinvasion is present in about 2% of high-power microscopic fields of PIN and is seen with equal frequency in all architectural patterns. At the transition from histologically normal epithelium to PIN, there is a surge in phosphorylated Akt (prosurvival protein) and concomitant suppression of downstream apoptosis pathways (antisurvival proteins) that precedes the transition to invasive cancer.

The mean volume of PIN in prostates with cancer is 1.2 to 1.3 mL, and the volume increases with increasing pathologic stage, Gleason grade, positive surgical margins, and perineural invasion (PNI). These findings underscore the close spatial and biologic relationship of PIN and cancer, and may result from an increase in PIN with increasing cancer volume.

PIN and cancer are usually multicentric. PIN is multicentric in 72% of radical prostatectomies with cancer, including 63% of those involving the nontransition zone and 7% of those involving the transition zone; 2% of cases have concomitant single foci in all zones. The peripheral zone of the prostate, the area in which the majority of cases of prostate cancer occur (≥ 70%), is also the most common location for PIN. Cancer and PIN are frequently multicentric in the peripheral zone, indicating a “field” effect similar to urothelial carcinoma of the bladder. Central zone cancer is more likely to be associated with PIN in the central zone than the peripheral zone. Despite common multifocality, PIN may be a monoclonal process according to similar high-resolution genomewide signatures.

High-grade PIN and prostate cancer are morphometrically and phenotypically similar ( Fig. 9.7 ). PIN occurs primarily in the peripheral zone and is seen in areas that are in continuity with prostate cancer. PIN and prostate cancer are multifocal and heterogeneous. Increasing rates of aneuploidy and angiogenesis as the grade of PIN progresses are further evidence that high-grade PIN is precancerous. Prostate cancer and high-grade PIN have similar proliferative and apoptotic indices.

Swedish investigators claim that PIN cannot be diagnosed by fine needle aspiration (FNA) alone, although this has been refuted. PIN in combination with atypical small acinar proliferation (ASAP) suspicious for but not diagnostic of malignancy is discussed later.

Biopsy remains the definitive method for detecting PIN and early invasive cancer. Serum PSA is not significantly elevated, if at all, and PIN cannot be detected by imaging methods such as ultrasound, magnetic resonance imaging (MRI), or positron emission tomography/computerized tomography (PET/CT) scans. There is a poor correlation of PIN and PSA density. Mean PSA increased from 8 to 12 ng/mL in patients with PIN who experienced development of cancer within 2 years; those with PIN who did not have cancer during this interval had an increase in PSA from 5 to 6 ng/mL. Median PSA velocity is greater in men with PIN who were subsequently diagnosed with cancer. A velocity threshold of 0.75 ng/mL/year predicts which men with PIN experience development of cancer, and velocity was the only significant predictor of subsequent cancer detection on multivariate analysis. The ratio of free to total PSA is the same for patients with high-grade PIN and cancer.

Immunohistochemical Markers for Prostatic Intraepithelial Carcinoma

We routinely use a series of four immunostains that, in combination, are invaluable in separating benign and malignant conditions of the prostate ( Table 9.6 ). Select antibodies, such as antikeratin 34βE12 (high-molecular-weight keratin) and p63 (see later), are used to stain tissue sections for the presence of basal cells, recognizing that PIN retains an intact or fragmented basal cell layer, whereas cancer does not ( Fig. 9.8 ). In addition, racemase and c-Myc are useful for staining of the dysplastic secretory cells of PIN (see later).

We routinely generate unstained intervening sections of all prostate biopsies for possible future immunohistochemical staining, recognizing that small foci of concern are often lost when the tissue block is recut. One study reported loss of the suspicious focus in 31 of 52 cases.

p63

p63 is a nuclear protein that is at least as sensitive and specific for the identification of basal cells in diagnostic prostate specimens as high-molecular-weight CK staining. p63 may be more sensitive than keratin 34βE12 in staining benign basal cells, particularly in TURP specimens, offering slight advantage in diagnostically challenging cases. Basal cell cocktail (the combination of 34βE12 and p63) increases the sensitivity of basal cell detection and reduces staining variability. Triple staining with racemase, high-molecular-weight CK, and p63 is commonly used for the diagnosis of prostate cancer ( Figs. 9.8 through 9.10 ), although we believe that the addition of c-Myc (quadruple stain) optimizes diagnostic yield (discussed later in this chapter).

Aberrant (positive) p63 staining is occasionally observed in primary and metastatic prostate cancer, so diagnosis should not rely exclusively on this marker. Normal stem cells (reserve cells) are maintained by p63, and alteration of p63 expression has an oncogenic role in prostate cancer. In contrast with usual prostatic adenocarcinomas, prostate tumors with p63 expression show a mixed luminal/basal immunophenotype, uniformly lack ERG gene rearrangement, and frequently express GSTP1. Expression of cytoplasmic aberrance of p63 is associated with high ALDH1A1 expression.

Racemase

Racemase (α-methylacyl-coenzyme A [CoA] racemase [AMACR] or P504S) is invaluable for separating benign and neoplastic acini, including evaluation of PIN, ASAP, and separation of cancer from hormonally treated benign acini. This well-characterized enzyme catalyzes the conversion of several (2R)-methyl-branched-chain fatty acyl-CoAs to (S)-stereoisomers. mRNA levels of racemase are upregulated ninefold in prostate cancer. The gene for AMACR is greatly overexpressed in prostate cancer cells. Its advantage over antikeratin 34βE12 is its positive granular cytoplasmic staining in cancer cells, with little or no staining in benign acini. In PIN, monoclonal and polyclonal antibodies to racemase are positive in 77% and 91% of foci, respectively ( Fig. 9.10 ). Because racemase is not specific for prostate cancer and is present in high-grade PIN (> 90% of cases), as well as nephrogenic adenoma, this staining must be interpreted with care. The diagnosis of PIN or prostate cancer should be rendered only in combination with convincing histologic evidence. Moderate to strong racemase expression in PIN is indicative of associated adenocarcinoma. Racemase is negative in up to 27% of cancer cases.

Overexpression of racemase in PIN may be predictive of increased likelihood of cancer on repeat biopsy, but this has not been confirmed.

Molecular Biology of Prostatic Intraepithelial Neoplasia

PIN is associated with progressive abnormalities of phenotype and genotype, which are between normal prostatic epithelium and cancer, indicating impairment of cell differentiation and regulatory control with advancing stages of prostatic carcinogenesis. There is progressive loss of some markers of secretory differentiation, cytoskeletal proteins, and multiple gene products ( Table 9.7 ). Other markers show progressive increase in expression from benign epithelium through PIN to cancer ( Table 9.7 ). A model of prostatic carcinogenesis has been proposed based on the morphologic continuum of PIN and the multistep theory of carcinogenesis.

High-grade PIN and prostate cancer share similar genetic alterations. For example, 8p12-21 allelic loss is commonly found in cancer and microdissected PIN. Other genetic changes found in carcinoma that already exist in PIN include loss of heterozygosity (LOH) at 8p22, 12pter-p12, 10q11.2, and gain of chromosomes 7, 8, 10, and 12, and the 8p24 and PTEN genes. LOH frequencies at 13q (one of the most common chromosomal alterations in high-stage cancer) is 0% in PIN versus 49% in clinical prostate cancer. Alterations in oncogene Bcl-2 expression and RER ⁺ phenotype are similar for PIN and cancer. Up to 64% of patients with PIN have LOH for the mannose 6-phosphate/insulin-like growth factor 2 receptor ( M6P/IGF2R ) gene, a marker for glycolytic metabolism.

Short telomere length in PIN and the surrounding stroma is associated with an increased risk for cancer. Telomere length is also predictive of time from the original biopsy to diagnosis of cancer. Overexpression of p4EBP1 predicted cancer with sensitivity and specificity of 63% and 100%, respectively.

PIN is epigenetically similar to carcinoma according to the percentage of methylated alleles for the APC , GSTP1 , and RARbeta2 genes. Methylation of the apoptosis-associated ASC promoter region is increased in PIN and cancer. TMPRSS2 exon 1 is fused in-frame with ERG exon 4 in 50% of prostate carcinomas and in 19% to 21% of cases of PIN but not in benign controls. PIN and cancer have higher levels of oxidative stress measured by urinary F2-isoprostane than benign prostatic tissue. Another mechanism associated with early prostatic carcinogenesis is deSUMOylation (SUMO refers to small ubiquitin-like modifier proteases), probably through enhanced cell proliferation resulting in PIN. Overexpression of p16INK4A in PIN is reportedly an independent predictor of relapse.

Microvessel density is higher in high-grade PIN than in adjacent benign prostatic tissue, and capillaries are shorter, more widely spaced, have more open lumina and curvaceous external contours. PIN vessels are lined by greater number of endothelial cells. PIN is virtually always accompanied by proliferation of small capillaries in the stroma. It is likely that PIN initially co-opts adjacent vessels, similar to other tumors, and that these vessels soon regress, only to be followed by marked angiogenesis at the cancer’s edge. Microvessel density is higher in cases with PIN associated with prostate cancer than in foci of isolated PIN or benign prostatic hyperplasia (BPH).

A critical balance exists between the proangiogenic vascular endothelial growth factor (VEGF) and the angiogenic antagonist angiopoietin-2. Angiogenin is a polypeptide involved in the formation and establishment of new blood vessels necessary for growth and metastasis of cancer. The percentage of cells staining positively for angiogenin in benign epithelium, PIN, and cancer is 17%, 58%, and 60%, respectively, confirming the potential role that angiogenin plays in neoplastic progression.

Treatment Effects in Prostatic Intraepithelial Neoplasia

Differential Diagnosis of Prostatic Intraepithelial Neoplasia

The histologic differential diagnosis of PIN includes a variety of benign conditions including atrophy, postatrophic hyperplasia, atypical basal cell hyperplasia, cribriform hyperplasia, and metaplastic changes associated with radiation, infarction, and prostatitis ( Tables 9.8 and 9.9 ; see Chapter 8 ). Many of these display mild architectural and cytologic atypia, including enlarged nucleoli. The diagnostic difficulty is compounded when these findings appear in small, cauterized, or distorted specimens.

Neoplastic mimics of PIN include IDC (see later), cribriform adenocarcinoma, ductal carcinoma, and urothelial carcinoma. Proliferative activity is higher in ductal carcinoma than in PIN (Ki67 labeling, 33% versus 6%). Stratified epithelium in noncribriform glands of cancer can also resemble high-grade PIN.

Inflammation, Atrophy, and High-Grade Prostatic Intraepithelial Neoplasia

Inflammation may cause prostatic carcinogenesis, but the evidence to date is inconclusive or contradictory, probably because of the ubiquitous presence of inflammation in the prostate (see Chapter 8 ). Chronic inflammation is present in about 8% of biopsies and appears to confer a protective effect from PIN and cancer even after adjusting for age, digital rectal examination findings, serum PSA, and prostate volume. Inflammation is inversely related to cancer risk, and atrophy is marginally related; whereas, PIN is a significant risk factor.

Proliferative changes of the epithelium are linked to carcinogenesis in almost all epithelial malignancies. Proliferative regenerative change, referred to in the prostate as proliferative inflammatory atrophy, may be premalignant. The hypothesis is that cellular injury and regeneration characteristic of inflammatory atrophy is induced by inflammation and release of reactive oxygen species (oxidative stress) resulting from insult caused by chemicals (e.g., dietary carcinogens), physical factors (e.g., arteriosclerosis-induced ischemia), or bacteria. The regenerating cells are at increased risk for mutation, which predisposes them to cancerous initiation, promotion, and progression. The clinical implication of this hypothesis is that antiinflammatory drugs could potentially block procarcinogenic inflammatory processes. Evidence to date is inconclusive as to whether atrophy is a precursor of PIN, direct precursor of cancer that bypasses PIN, or simply an epiphenomenon (see Chapter 8 ).

What is clear is that any association with cancer requires proliferative changes, and these changes in the setting of atrophy are most often associated with inflammation. Substantial inflammation is often linked with infectious and noninfectious prostatitis, and emerging data indicate a causative role for inflammation in prostate cancer development over time, although there are conflicting data (see Chapter 8 ). The presence of CD68-immunoreactive inflammatory cells in the periprostatic white adipose tissue is associated with high-grade cancer. Despite the controversy regarding the role of atrophy (or lack thereof), virtually all authors agree that inflammation-induced oxidative stress is the most plausible explanation for initiation of prostatic carcinogenesis.

Clinical Significance of Prostatic Intraepithelial Neoplasia

Diagnosis of Atypical Small Acinar Proliferation

ASAP represents our inability to render an incontrovertible diagnosis of cancer in a needle biopsy. The focus of concern is invariably no larger than two dozen acini, less than the size of the head of a pin, so the major concern is overdiagnosis of cancer based on insufficient evidence. The diagnostic difficulty with ASAP usually results from one or a combination of the reasons listed in Table 9.12 . All of these may hinder the diagnosis of carcinoma, but in such cases the possibility cannot be definitively excluded. The need for this category is based on our “absolute uncertainty.” That this need exists is manifested by the variety of terms or synonyms currently in use that include the word atypical to describe this diagnosis, although atypical small acinar proliferation is now the preferred term and is most widely used clinically around the world. The diagnosis of ASAP indicates to the clinician that the biopsy in question exhibits histologic features that are neither clearly malignant nor benign, and that follow-up of the patient is warranted.

For pathologists, three questions need to be answered before the diagnosis of ASAP or cancer in a small lesion: (1) Would you be absolutely confident of this biopsy diagnosis if it were followed by a negative prostatectomy? (2) Would another pathologist agree with the diagnosis of cancer? and (3) Can you confidently support the diagnosis of adenocarcinoma based solely on this biopsy? If the answer to any of these questions is no, then we recommend use of the more conservative ASAP diagnosis. In this setting, we believe that “atypical small acinar proliferation” is a valid diagnostic category if it is used judiciously, and that maximum information has been obtained from the available tissue. Other evidence useful in supporting a cancer diagnosis, including patient age, serum PSA concentration, and results of high-molecular-weight CK, p63, racemase, and c-Myc studies, cannot substitute for convincing hematoxylin and eosin (H&E) microscopic findings. To avoid bias, the above information should be considered only in combination with microscopic examination.

The histologic features that most often preclude a definitive diagnosis of malignancy are the small size of the focus (70% of cases), disappearance on step levels (61%), lack of significant cytologic atypia such as nucleolomegaly (55%), and associated inflammation (9%), raising the possibility of one of many mimics of adenocarcinoma ( Fig. 9.14 ). Other causes include negative staining for high-molecular-weight CK or p63, atrophic changes or inflammation accompanying glands lacking a basal cell layer, and the presence of associated PIN.

Immunohistochemical stain for racemase facilitates the diagnostic support provided by antikeratin 34βE12, p63, and c-Myc, particularly in equivocal biopsies such as ASAP ( Fig. 9.15 ). These immunostains resolve 76% of ASAP diagnoses. We routinely use these important techniques in the diagnostic workup of atypical prostate lesions on needle biopsies, thereby decreasing the incidence of ASAP while reducing the risk for false-negative results and the need for additional biopsies. Strand et al. demonstrated that preparing new recut sections and performing immunostains allowed definitive diagnosis of carcinoma in 22% of cases that would otherwise have been signed out as ASAP.

Diagnostic criteria differ between ASAP and minimal cancer ( Table 9.13 ). First, the mean number of acini and length of the focus of concern in ASAP (11 acini) are about one-half of those of minimal cancer (17 acini). The small size of a focus of concern is the commonest source of difficulty in ASAP, accounting for 70% of ASAP diagnoses. Infiltrative growth is a constant feature of prostate cancer, but also occurs in 68% to 75% of ASAP cases. All cancers have at least mild nuclear enlargement; whereas, ASAP may have little or no significant enlargement. Prominent nucleoli may be obscured by nuclear hyperchromasia, and this occurs more commonly in ASAP than in minimal cancer. The greater frequency of mitotic figures in cancer (up to 10% of cases) stands in contrast with that of benign acini or ASAP. However, mitotic figures are too rare to be a reliable diagnostic finding in small foci of cancer. Blue-gray acidic luminal mucin may be a useful discriminator, occurring in 61% of cancers compared with 42% of ASAP cases. Eosinophilic proteinaceous secretions are nonspecific, occurring with similar frequency in cancer and ASAP (73% versus 66% to 74% of cases, respectively). Crystalloids are uncommon and nonspecific, occurring in 19% of cancers and 16% of cases of ASAP. Prior studies reported incidence rates of crystalloids in 6%, 13%, and 22% of ASAP foci in 5% of benign biopsies. These data confirm that crystalloids do not pose an increased risk for cancer in subsequent biopsies. Moderate to severe atrophy accompanies more cases of ASAP than cancer. Small foci of postatrophic hyperplasia (which is invariably associated with typical atrophy and has “moderately enlarged” nuclei in 39% of cases) and typical atrophy are sometimes interpreted as ASAP. Obscuring inflammation is responsible for ASAP diagnoses in 30% of cases; however, 20% of minimal cancers also have associated acute or chronic inflammation.

Interobserver agreement for ASAP among urologic pathologists was higher than that of other pathologists (κ, 0.39 versus 0.21, respectively), but complete (100%) agreement was reached by the experts in only 7 of 20 biopsies. The experts diagnosed adenocarcinoma (49%) more often than the nonexperts (32%), and agreement was particularly poor for foci comprising fewer than six acini. The authors concluded that consultation would be of value with a specialized pathologist for ASAP before rendering the diagnosis of carcinoma.

Subsets of Atypical Small Acinar Proliferation

Stratification of ASAP does not increase the predictive accuracy for cancer on repeat biopsy despite multiple attempts. We stratified suspicion in each ASAP case into three levels: ASAPB (ASAP suspicious for but not diagnostic of malignancy, favor benign), ASAPS (ASAP suspicious for but not diagnostic of malignancy), and ASAPH (ASAP highly suspicious for but not diagnostic of malignancy). ASAPB was used for cases in which we deemed the focus of concern unlikely to be cancer but could not with absolute certainty exclude the possibility. Conversely, ASAPH was used for cases in which the focus was almost certainly carcinoma, but a confident diagnosis of cancer could not be rendered. ASAPS was used for cases with intermediate suspicion.

In stratifying ASAP into these levels of suspicion, three criteria emerged as significant. Infiltrative growth was present in just over one-half of ASAPB cases but almost all ASAPH cases. ASAPS and ASAPH had greater degrees of nuclear enlargement than in ASAPB but were confounded more frequently by nuclear hyperchromasia. Hyperchromasia involved more than one-half of cases of ASAPS and ASAPH, often obscuring nuclear detail and nucleoli; nevertheless, we noted a nonsignificant trend toward prominent nucleoli in a higher percent of the cases as suspicion increased. Two other nonsignificant trends were noted with increasing suspicion for malignancy: more frequent coexistent high-grade PIN and less frequent moderate to severe atrophy.

Multiple studies revealed nonsignificant trends for increasing risk for subsequent cancer with increasing suspicion. Stratification of ASAP also did not predict the normalized percent of involvement by cancer on positive repeat biopsy. Thus at present the level of suspicion should not alter follow-up recommendations.

Clinical Significance of Atypical Small Acinar Proliferation

ASAP is predictive of prostate cancer in up 39% to 65% of repeat biopsies. Saturation biopsy “substantially” increases the cancer detection rate with or without the presence of ASAP.

ASAP represents undersampled cancer in at least 40% of cases. Iczkowski et al. observed that some men with ASAP in the first set of biopsies and benign findings or high-grade PIN in the second biopsy may still contain cancer that was not detected.

False-negative results in untreated men with documented adenocarcinoma occurred in 23% of repeat sextant biopsies. These results suggest that the current practice of performing 6 to 12 biopsies per prostate does not lower the frequency of ASAP. A declining volume of cancer at prostatectomy was noted more than a decade ago and is probably reflective of increased screening and multiple sampling. Thus as smaller-volume cancers are detected through increased sampling, many will be undersampled and not resolvable by immunostains, likely leading to an irreducible rate of ASAP diagnosis. Saturation biopsy increases the yield of ASAP.

What prostatic sites should be sampled at repeat biopsy? One study found that sampling only the side or site initially diagnosed as ASAP missed cancer in 39% of patients whose cancer was later detected exclusively at other sites, suggesting that the entire prostate should be rebiopsied.

In a provocative report, the investigators recommended immediate prostatectomy in patients with the biopsy diagnosis of ASAP. They suggested that the risk for subsequent cancer is 100% in prostatectomy specimens. We urge caution in recommending expansion of the indications for prostatectomy to include patients with ASAP. ASAP is best considered as a diagnostic risk category and not a true entity.

Atypical Small Acinar Proliferation + Prostatic Intraepithelial Neoplasia

It is often difficult with small foci in biopsies to separate cancer from suspicious foci (ASAPS) when there is coexistent high-grade PIN; the difficulty is based on the inability to separate tangential cutting of the larger preexisting acini of PIN (that may appear as small separate adjacent acini) from the smaller discrete acini of cancer ( Fig. 9.16 ). In such cases, we prefer the term ASAP + PIN (referring to the coexistence of the two lesions, ASAP and high-grade PIN, in the same high-power microscopic field) to avoid overdiagnosis of tangential cutting of PIN and cancer. ASAP is placed first because of its stronger predictive value for subsequent cancer.

ASAP + PIN, found in up to 16% of all biopsies, has an intermediate predictive value of about 33% to 60% for cancer. Thus it is slightly lower than isolated ASAP (37%) but higher than isolated PIN. In four studies, PIN occurred with ASAP in 17%, 23%, 31%, or 41% of cases with ASAP, but most foci were not adjacent or contiguous. In a study of 12 patients with PIN and adjacent atypical glands (ASAP), 75% had cancer on repeat biopsy. The prostate cancer risk increased in the ASAP subgroups according to the extent of PIN in the initial sample, with ASAP + multifocal PIN carrying a 71% prostate cancer risk.

Differences in predictive values for cancer seen in multiple studies of PIN + ASAP arise from multiple causes. First, contiguous foci might have an intrinsically higher predictive value for cancer than those that include noncontiguous lesions. Second, selection bias present in cases referred for consultation also may have influenced some study results as compared with unselected primary cases in other cohorts. Third, the number of patients reported in some studies was so small that skewing of data in either direction might occur. Finally, the number or method of biopsy may influence results. For example, use of transperineal template saturation biopsies at 5-mm intervals throughout the prostate revealed that ASAP for two or more cores of high-grade PIN on a previous transrectal ultrasound-guided (TRUS) biopsy strongly indicated the presence of cancer on saturation biopsy. Patients who were diagnosed with high-grade PIN, or who had two or more cores with high-grade PIN or ASAP, cancer was detected in 53%, 89%, and 83%, respectively. High-grade cancer rates (Gleason score ≥ 7) in patients with ASAP and two or more cores of high-grade PIN were 20% and 80%, respectively.

The best available evidence today indicates that the presence of either or both lesions in needle biopsies is still a predictor for concurrent/subsequent cancer compared with patients who lack these lesions.

PIN was more than twice as frequently associated with minimal cancer (57%) than ASAP (23%). About one-half of cases of ASAP are probably undersampled cancer, and the smaller mean size of the foci in contemporary specimens decreases the likelihood of sampling accompanying high-grade PIN.

Separation of Intraductal Carcinoma and Prostatic Intraepithelial Neoplasia

Unlike PIN, IDC is strongly associated with aggressive prostate cancer, including high Gleason grade and large tumor volume ( Tables 9.15 and 9.16 ). Most important are the extent of the process (PIN is usually limited in its extent to a small number of acini at most), acinar size (PIN is found within preexisting acini without significant distortion or expansion of the acinar contours, whereas IDC is usually present in greatly expanded acini and ducts), and cytologic features (PIN is rarely if ever frankly anaplastic, whereas noninvasive and invasive ductal carcinoma often exhibit high-grade nuclear abnormalities such as nucleomegaly, anisonucleosis, and marked hyperchromasia); comedonecrosis is never observed in PIN but is occasionally observed in IDC. The incidence of IDC is higher in Japanese patients than in Americans (35% versus 13%, respectively), which is independent of Gleason score and cancer stage.

Unfortunately, uncertainty persists regarding precise criteria for separating IDC from PIN because of the identification of borderline cases with overlapping features or coexistent features, especially those with cribriform or near-solid growth. IDC mixed with invasive carcinoma may occasionally share architectural and cytologic features with cribriform PIN, and all may rarely coexist; in these cases, PIN is identified as an intraductal/intraacinar cribriform proliferation within small, smoothly rounded acini with low-grade nuclei that do not fulfill the typical criteria for IDC. Shah et al. found that the cribriform pattern of IDC was present in 18% of radical prostatectomies, was more common in cancer with Gleason score ≥ 7, ranged in size from 0.2 to 9.0 mm, and contained comedonecrosis in 33% of cases. Pleomorphic nuclei or giant nuclei at least 6 × of the adjacent nuclei were present in 28%. By comparison, isolated cribriform proliferation with malignant cells distant from cancer, referred to by some as isolated atypical cribriform proliferation (we call these IDC when the cells display anaplasia), was usually associated with Gleason score 6 cancers, was smaller (range, 0.2 to 1 mm), and never displayed comedonecrosis or pleomorphic or giant nuclei. Genetic studies may be useful for borderline cases (see later). PTEN and ERG expression may be useful in separating IDC and PIN (incidence rates of 61% versus 0% and 30% versus 0%, respectively).

Clinical Significance of Intraductal Carcinoma

IDC is an aggressive phenotype of prostate cancer that predicts high-grade, high-stage metastases and poor response to androgen deprivation therapy, external beam radiation therapy, and chemotherapy. It should be separated from PIN and, when present, reported in prostate biopsies and prostatectomy specimens, even when there is coexistent adenocarcinoma. Some authors consider IDC to be equivalent to Gleason pattern 4 or 5 adenocarcinoma, and diagnosis on needle biopsy should prompt therapeutic intervention rather than surveillance or repeat biopsy, as is the case for PIN. IDC occasionally coexists with Gleason grade 3 cancer.

IDC with cribriform architecture is present in 2% of autopsies and 1% of cystoprostatectomies with prostate cancer.

IDC displays significant genetic abnormalities. LOH is usually absent in Gleason grade 3 cancer, infrequent in PIN (9%) and Gleason grade 4 cancer (29%), but common in IDC (60%). The cribriform pattern of IDC, when present within 3 mm of invasive carcinoma, displays TMPRSS2-ERG gene fusion in 75% of cases, with 100% concordance with the cancerous component. Cytoplasmic PTEN loss is identified in 84% of cases of IDC and 100% with the cribriform pattern; concordance with invasive carcinoma is greater than 95%. IDC is associated with pathogenic germline DNA-repair gene mutations.

Cribriform IDC has a higher rate of biochemical failure and cancer-specific death than acinar carcinoma with or without cribriform growth. Gleason score 7 cancer with cribriform or IDC is independently associated with poorer biochemical recurrence-free survival after prostatectomy, but not after radiotherapy. IDC using the McNeal criteria is an independent risk factor for progression after prostatectomy. Androgen deprivation therapy induces regressive changes in PIN and prostate cancer but appears to have a minimal effect on IDC.

Epidemiology

Prostate cancer poses a greater risk for American men, especially African American men, than any other nonskin cancer. The veritable epidemic of prostate cancer has resulted in part from successful efforts at early detection with the use of the serum PSA test, thereby narrowing the still-enormous gap between the clinical incidence (8% lifetime risk) and autopsy-based prevalence (80% by age 80 years) ( Fig. 9.18 ). Physicians are unable to accurately stratify patients into those who will have progressive cancer and those who will not (cannot separate the “tigers” from the “pussycats”). An equally great problem is determining which men are at greatest risk for development of clinically apparent prostate cancer.

Prostatic adenocarcinoma is rare in patients younger than 40 years. Fewer than 50 cases of adenocarcinoma have been reported in each of the following groups: children younger than 12 years, adolescents, and young adults between 20 and 25 years old. In all these cases, the cancer was poorly differentiated, clinically aggressive, and unresponsive to hormonal therapy and radiation therapy.

After age 40, the incidence increases quickly. Autopsy studies of thoroughly evaluating prostates from men without clinical evidence of cancer have shown a very high level of latent (clinically occult) cancer. The incidence of prostatic adenocarcinoma is much higher in men of African ancestry (100 per 100,000) than in men of European ancestry (70.1 per 100,000), and men of African ancestry in the United States have the world’s highest mortality rate from prostatic adenocarcinoma. The prevalence of latent cancer is similar in different geographic and ethnic groups despite wide variation in the incidence of clinically apparent adenocarcinoma. The incidence is low in American Indians, Hispanics, and Asians, but high in American men of African and European ancestry.

Prostate cancer mortality varies considerably from country to country. High rates have been reported in the United States, particularly among African Americans, whereas low rates have been found in China and Japan. There are advantages and disadvantages in using mortality data to examine the underlying risk for prostate cancer. Incidence data are often unavailable, so mortality is a commonly used surrogate. However, mortality is a function of incidence, survival, and ascertainment and selection biases. International differences in mortality may reflect differences not only in the underlying risk for development of prostate cancer but also of differences in survival or ascertainment/reporting (death certificate) bias. Remarkably, the incidence of new cases surged in Japan in 2002 after the announcement that the Japanese Emperor, Akihito, had prostate cancer. Comparative autopsy study revealed a similar proportion of cancer in Russian Caucasian and Japanese men. More than 50% of cancers are Gleason score ≥ 7 in Japanese men and nearly 25% in Russian Caucasian men, raising questions about: (1) previous assumptions related to Asian prostate cancer, and (2) the notion of significant versus insignificant cancers.

Latent Carcinoma

The prevalence of carcinoma is most strongly related to age, with the prevalence doubling about every 14 years, and the age-specific prevalence of latent cancer at autopsy is remarkably constant across countries and ethnic groups. Data from the Connecticut Tumor Registry estimated the age-standardized prevalence rate to be 841.6 per 100,000 in 1994, an increase of 126% over the 1982 rate. In contrast, the prevalence in Japan remained largely unchanged from 2005 to 2014 (14% and 12% and Gleason score > 6 of 6% and 7%, respectively). However, another report from that country noted a great increase from 21% to 43% between 1983 to 1987 and 2008 to 2013, respectively. The dramatic differences in prevalence between these autopsy studies from Japan cannot be explained. Systematic literature review concluded that “the prevalence of incidental tumors was relatively low in earlier studies of Japanese men, although more recent study estimates are similar to the rest of the world.”

Incidental prostate cancer is less aggressive than clinically apparent prostate cancer according to stage, surgical margin status, and Gleason score. Totally sampled cystoprostatectomy specimens with bladder cancer also contain clinically undetected incidental prostate cancer in about 42% of cases (range, 15% to 68%) with the highest incidence in older men. Cystoprostatectomies with prostate cancer have lower stage and lower Gleason score than prostatectomy cases. Incidental prostate cancer in cystoprostatectomy cases is usually stage pT2a or pT2b (59% and 29%, respectively). Incidental prostate cancer is usually low grade (Gleason scores 5 and 6 in 25% and 46% of cases, respectively), much lower than for clinical prostate cancer (5% and 21%, respectively). Only 9% of incidental cancers are Gleason scores 8 to 10. The frequency of positive margins is lower than in clinical cancer (7% versus 52%, respectively).

The relationship between incidental, latent, and clinical prostate cancer may be explained by two hypotheses. One contends that incidental and latent carcinomas are identical histologically to lethal cancer, but have never acquired biologically aggressive features. The other hypothesis contends that clinically innocuous cancers are simply the smallest tumors, and cancer acquires the capacity to metastasize as a function of the passage of time, increasing volume, and “biological tumor progression,” a function of the mutational instability of all cancers, which becomes manifest in proportion to the number of mitotic events. We agree with Selman that the term latent carcinoma is a medical misnomer and should not be used.

Risk factors can be classified as endogenous or exogenous, although some factors are not exclusively one or the other (e.g., race, aging, and oxidative stress). There are numerous endogenous risk factors for prostate cancer, including family history, hormones, race, and aging and oxidative stress ( Table 9.17 ). Exogenous risk factors include diet, endocrine disrupting chemicals, and occupation.

Prostate Cancer and Benign Prostatic Hyperplasia

Prostatic hyperplasia is frequently seen in association with prostatic adenocarcinoma ( Fig. 9.19 ). There are several compelling similarities, including increasing incidence and prevalence with age, concordant natural history, hormonal requirements for growth and development, and response to androgen deprivation therapy. However, no causal relationship has been established or seriously suggested.

Tissue Methods of Detection

Needle Core Biopsy

Introduction of the thin needle for transrectal biopsy of the prostate in the 1980s, together with the use of serum PSA, revolutionized early detection efforts for prostate cancer. These two advances were mutually beneficial, feeding off each other to effectively replace the large-bore transperineal needles and exclusive reliance on digital rectal examination for cancer detection. The rate of postbiopsy infection with the thin needle declined from 7% to 39% to 1%, and hemorrhage with urinary clot retention decreased from 3% to less than 1%. The false-negative rate declined from 25% to 11%, and there was an improvement in the quality of the tissue sample obtained, usually with little or no compression artifact at the lateral edges of the specimens. Also, the 18-gauge (18G) needle allows multiple biopsies of the prostate with minimal discomfort, particularly with the use of topical anesthetics such as lidocaine. Today it is hard to imagine practicing urology and urologic pathology without PSA and multiple biopsies.

A greater number of prostate biopsies are obtained currently, and more biopsy cores are submitted than ever before, creating a huge interpretive burden for the pathologist. It is estimated that more than one million biopsies are performed annually in the United States, with each biopsy consisting of on average about 10 to 12 cores, creating an estimated ≥ 10 million prostate tissue samples for the pathologist to interpret.

This burden is compounded by several factors that have increased the difficulty in prostate interpretation. First, many patients now undergo biopsy for elevated serum PSA with no other clinical evidence of cancer, resulting in an enormous number of biopsies that often contain only a small or microscopic suspicious focus. Second, numerous diagnostic pitfalls and mimics of prostate cancer have been described or refined, including postatrophic hyperplasia and IDC (see earlier in this chapter and Chapter 8 ). The great number of prostate biopsies being generated magnifies the risk for encountering rare or unusual lesions and the potential for misinterpretation of small foci. The concept of ASAPS was introduced in the late 1990s, accounting for about 2% of biopsy diagnoses. Finally, ≥ 10 biopsies (≥ 5 from each side) have largely replaced the single bilateral cores of 20 years ago and the sextant biopsies from 10 years ago, providing multiple specimens from each patient.

Detecting Cancer: Factors That Influence Diagnostic Yield in Biopsies

How can we improve the yield of cancer from prostate needle biopsies? Table 9.18 describes the known variables that influence the diagnostic yield of prostate biopsies. Fixed, uncontrolled factors included patient-related factors and prostate-related factors; however, biopsy method-related factors are controllable by the urologist and pathologist to increase the diagnostic yield of cancer and are thus deserving of additional consideration.

Table 9.18

Factors That Influence the Detection Rate of Cancer in Contemporary Prostate Needle Biopsies

Uncontrolled Factors

• Patient risk factors

  • Patient population (e.g., screening population versus urologic practice)

  • Patient symptoms

  • Serum prostate-specific antigen

  • Clinical stage

  • Patient age

  • Patient race

  • Prior biopsy findings (e.g., prostatic intraepithelial neoplasia, atypical small acinar proliferation)

• Prostate-related factors

  • Prostate volume

  • Transrectal ultrasound and other imaging findings

Controlled Factors

• Urologist-controlled factors

  • Number of needle cores obtained

  • Method of biopsy (e.g., random, ultrasound-guided, magnetic resonance imaging–targeted, etc.)

  • Location of biopsy (e.g., laterally directed biopsies versus midline, etc.)

  • Amount of tissue obtained (e.g., biopsy “gun” used; operator skill)

• Pathologist-controlled factors

  • Histotechnologist’s skill in processing and cutting prostate biopsies

  • Number of needle cores embedded per cassette

  • Number of tissue cuts obtained per specimen

  • Pathologist’s skill in prostate biopsy interpretation

Number, Length, and Location of Needle Cores Obtained

The increase from 6 to 12 cores improved prostate cancer detection by 29%, and all were greater than 0.5 cm ³ in volume. By obtaining 10, 12, and 13 core biopsies, cancer detection rates increased by 26%, 22%, and 35%, respectively. The number of biopsy cores correlates with probability of prostate cancer detection ( Fig. 9.20 ). Saturation biopsy (often defined as > 12 or 16 cores) may be of greatest value in men with persistent suspicion of cancer after negative initial biopsy and in those with ASAP or multifocal PIN.

Cancer detection rate with the routine 18G biopsy needle (40%) is similar to that with the narrower 20G needle (35%), but pain is significantly less with 20G. The diagnostic yield of biopsy is improved by ultrasound-targeted or MRI-fusion–targeted biopsies, but the magnitude of the increase in accuracy remains controversial. Use of a 29-mm cutting length increases cancer yield 18% above that of a 19-mm cutting length, although the correlation of length and cancer yield has been contested. The combination of six transperineal and six transrectal biopsies resulted in cancer detection of 49%, and the detection rate increased 7% and 9% compared with the transperineal and transrectal groups alone, respectively. Increased core length per container improves cancer detection rate and Gleason score accuracy, and 12-mm cores have optimal sensitivity (42%) and specificity (62%) for cancer detection. Transperineal mapping biopsies are more accurate in determining clinical risk than transrectal biopsies and are less likely to induce sepsis, but transrectal biopsies are more popular in Europe than the United States. Template-guided transperineal saturation biopsy detects 64% of clinically significant cancers. Most of the undetected lesions are those with small volume. Approximately 20% to 30% of patients have clinically significant undetected lesions in a different lobe or different quadrant from the detected lesions in the biopsy.

Quantitation of cancer length in biopsies is discussed in the following paragraphs.

The detection rate of cancer in biopsies is higher with longer cores, particularly at the apex. For example, a 20-mm core from the right apex has 27% probability of cancer detection versus 18% for 10 mm. Only 1% of single-core biopsies are longer than 20 mm. The amount of tissue obtained by biopsy varies widely, and cumulative core sample (likely to be an inadequate sample) is less than 50 mm in 4% of biopsies. The dependence of diagnosis on tissue length is strongest at the apex. Nonglandular sampling is directly associated with shorter core lengths, with a 6-mm minimum threshold. Most urologists now routinely obtain at least 10 to 12 specimen vials per biopsy for prostate biopsies.

Lateral mid-gland and lateral base biopsy cores have the highest cancer detection rates for all prostate volumes, perhaps because of extensive sampling of the peripheral zone by lateral biopsies. Mid-gland and base biopsy cores have a relatively low yield, especially in small prostates, because of sampling of the central zone, where the prostate cancer incidence is known to be low.

Overall cancer detection rate with saturation biopsy is 44%, increasing to 52% with additional biopsies of the anterior apical region. Cancer detection rate increases 23% with a modified sextant protocol by directing the needles only into the more lateral aspect. Conversely, ultrasound-directed biopsies may be omitted when using 10-core biopsy protocols because the yield of these biopsies was less than 2%. Rebiopsy detection rate is higher with 20 rather than 12 cores.

Magnetic Resonance Imaging–Transrectal Ultrasound-Guided Fusion–Targeted Biopsies

Multiparametric magnetic resonance–transrectal ultrasound fusion–targeted biopsy may improve accuracy of: (1) detection rate of clinically significant cancer previously missed by TRUS biopsy; (2) identification of extraprostatic extension (EPE); and (3) categorization of aggressiveness based on diffusion-weighted imaging. Accuracy is influenced by interobserver variability in imaging interpretation. As a triage test, there may be a substantial number of false-negative cases with clinically significant cancer, suggesting that transperineal systematic biopsy may be superior for triage. Combining multiparametric magnetic resonance–transrectal ultrasound fusion–targeted biopsy with systematic biopsy improves the diagnostic accuracy of either single modality without increasing the rate of insignificant prostate cancer.

Histotechnologist’s Skill in Processing Biopsies

Prostate biopsies are particularly difficult to embed and cut because of their small size and tendency to fragment and curve. Flat embedding of the biopsy cores enhances the amount of tissue that is examined by the pathologist. Laboratories that process prostate biopsies with other tissues of differing density and consistency (e.g., breast biopsies with abundant fatty tissue) usually handle all specimens the same way, optimizing results for some tissues but often resulting in prostate biopsies that are too thick to interpret or are overstained. Excessively thick tissue specimens are two or three cells in thickness rather than the optimal one to two cells in thickness, precluding adequate assessment of nuclear and cytoplasmic details in foci of concern. Similarly, overstained sections (the most common problem in our consultation practice) contain obscured nuclear chromatin without recognizable nucleoli. These problems are compounded in biopsies with small foci that are suspicious for malignancy and in younger patients (those in their 40’s and 50’s) who have abundant proliferative epithelium that may mimic malignancy. The problems noted here apply doubly to interpretation of high-grade PIN and ASAP, accounting for both overdiagnosis (the most common mistake in current practice, in our experience) and underdiagnosis. Separate processing of the delicate prostate needle cores is recommended by the European Society of Uropathology.

Number of Needle Cores Embedded per Cassette

Multiple needle biopsies submitted in one or two containers tend to entangle and fragment and are difficult to embed in a single plane during processing. The resulting loss of tissue surface area makes a definitive diagnosis difficult in many cases, resulting in equivocal pathology reports. If multiple cores are embedded in one cassette, it is necessary to take care that all are separated. We carefully embed up to six cores in parallel arrays per cassette after differential inking and find the cancer yield equivalent to one core per cassette with significant reduction in labor cost and effort. In a study from Finland, there was no significant difference in cancer detection rate when up to nine cores were embedded per cassette.

Number of Tissue Cuts Obtained per Specimen

There is variation between laboratories in the number of serial tissue cuts obtained from each needle core for routine examination. To avoid the serious problem of undersampling, we routinely obtain six separate cuts (two adjacent sections from three separate levels) from the paraffin block for H&E staining; additional intervening sections are placed on another slide and saved for immunohistochemical stains or special studies ( Fig. 9.21 ). We consider the recommendation of the European Randomized Study of Screening for Prostate Cancer to be inadequate (they recommend only two cuts in total), probably missing up to 3% of cancers with such limited sampling. In our experience, recutting the block for additional levels with small suspicious foci is useful in about one-half of cases, with usually no more than four additional slides before the tissue specimen is exhausted. Rogatsch et al. compared biopsy cores submitted floating free in formalin with those that were stretched and oriented at biopsy and before formalin fixation, and found that the diagnostic rate of cancer increased from 24% to 31%.

Future Trends in Biopsies

Select future trends in biopsy handling and clinical significance are presented in Table 9.19 . Precise localization of cancer by site-specific labeling and three-dimensional mapping of extended saturation biopsies enables the use of targeted focal therapy such as cryosurgery. Quality-assurance measures should focus on the quantity and quality of biopsy samples. The ultimate goal of cancer detection—prediction of outcome for the individual patient—can be augmented by advanced methods of database analysis, such as artificial intelligence. Furthermore, there is continuing improvement in the definition and treatment of clinically insignificant cancer. Finally, molecular biology is beginning to revolutionize the field of diagnostic pathology.

Table 9.19

Trends in Biopsy Reporting and Clinical Application

Location of Prostate Cancer

• Site-specific labeling

• 3D mapping

• Focal therapy

Quality Assurance in Urology

• Quality of biopsies

• Number of biopsies

• Quality assurance in urologic pathology

Personal Outcome Predictions

• Use of neural networks and advanced measures of outcome

• Improved understanding of “clinically insignificant” cancer

Molecular Diagnostics From Needle Biopsies

Fine Needle Aspiration

Interest in FNA in the United States is minimal because of the ease of acquisition and interpretation of the 18G needle core biopsy. Both techniques have similar sensitivity in the diagnosis of prostatic adenocarcinoma, and both are limited by small sample size; they are best considered as complementary techniques. Complications of FNA occur in less than 2% of cases and are like those with needle core biopsy including epididymitis, transient hematuria, hemospermia, fever, and sepsis.

FNA produces clusters and small sheets of epithelial cells without stroma. This enrichment for epithelium allows evaluation of single-cell morphology and the relationship between cells. Benign and hyperplastic prostatic epithelium consists of orderly sheets of cells with distinct margins creating a honeycomb-like pattern. Benign nuclei are uniform with finely granular chromatin and indistinct nucleoli; basal cells are often present at the edge. Prostatic carcinoma is distinguished from benign epithelium by increased cellularity, loss of cell adhesion, variation in nuclear size and shape, and nucleolar enlargement ( Fig. 9.22 ). A recent case of prostatic synovial sarcoma was diagnosed by FNA and confirmed by fluorescent in situ hybridization (FISH) analysis for SYT rearrangement on the cell block.

Transurethral Resection

The regions of the prostate sampled by TURP and transrectal needle biopsy tend to be different. TURP specimens usually consist of tissue from the transition zone, urethra, periurethral area, bladder neck, and anterior fibromuscular stroma. Occasionally, TURP specimens may also contain small portions of seminal vesicle tissue. Radical prostatectomies performed after TURP show that the resection does not usually include tissue from the central or peripheral zones, and not all of the transition zone is removed. Most needle biopsies consist only of tissue from the peripheral zone, seldom including central or transition zones.

The role of TURP in dealing with urinary obstructive symptoms has declined in the past decade because of the introduction of ablative techniques such as the yttrium-aluminum-garnet (YAG) laser, as well as prostate-reducing medications such as 5α reductase inhibitors (e.g., finasteride, dutasteride).

The incidence rate of cancer in TURP specimens for nodular hyperplasia is about 5%. Well-differentiated adenocarcinoma found incidentally in TURP chips usually has arisen in the transition zone ( Fig. 9.23 ). These tumors are frequently small and may be completely resected by TURP. Transition zone cancer location is associated with better biochemical recurrence-free survival. Poorly differentiated adenocarcinoma in TURP chips usually represents part of a larger tumor that has invaded the transition zone after arising in the peripheral zone.

TMPRSS2:ERG and ETV1 are upregulated in acini in the peripheral zone compared with the transition zone, and these fusions account for 50% to 80% and 20% of prostate cancers, respectively. These results indicate that the benign and neoplastic glands of the two zones display distinct molecular differences and zonal-specific expression, and that ERG and ETV1 may play a role in the development and progression of cancer.

The optimal number of chips to submit for histologic evaluation is a minimum of six cassettes for the first 30 g of tissue and one cassette for every 10 g thereafter.

Radical Prostatectomy

There are two main surgical approaches to prostatectomy. The first, retropubic prostatectomy, is the most popular approach in the United States, allowing staging of lymph node biopsies with frozen section evaluation before removal of the prostate when desired. The second surgical approach, perineal prostatectomy, does not allow lymph node biopsy during the same operation because of the anatomic approach used. Refinements in technique include nerve-sparing prostatectomy, robotic prostatectomy, and laparoscopic prostatectomy, all of which are gaining greatly in popularity.

The completeness of pathologic examination of prostatectomy specimens affects the determination of pathologic stage ( Tables 9.20 and 9.21 ). Partial sampling results in missing 29% of cases with positive margins and 20% of those with EPE. There is an increase in positive surgical margin rate (12% versus 59%, respectively) and pathologic stage with complete sectioning compared with limited sampling (sections of palpable tumor and two random sections of apex and base). Complete sectioning shows a higher detection rate of EPE (34% versus 55%, respectively) and seminal vesicle invasion (8% versus 15%, respectively) than lesser methods of sampling, even in prostates removed as part of a cystoprostatectomy. Complete sectioning also correlates with greater cancer-free survival in organ-confined and specimen-confined cases. Sehdev et al. compared 10 different methods of prostatectomy sampling using complete sectioning as the gold standard and favored two partial sampling methods ( Fig. 9.24 ) that balanced the extra time and expense of additional sections versus the risk of missing important predictive information. The first method was submission of every posterior section plus one midanterior section from right and left sides; if either of these anterior sections shows sizable tumor, all ipsilateral anterior slides are examined. This method detects 98% of tumors with Gleason score ≥ 7, 100% of positive margins, and 96% of cases with EPE (mean, 27 slides). The second method is the same as the first but with submission only of sections ipsilateral to the previous positive needle biopsy. This method detects 92% of tumors with Gleason score ≥ 7, 93% of positive margins, and 85% of cases with EPE (mean, 17 slides).

Table 9.20

ISUP , International Society of Urological Pathology.

Sample Surgical Pathology Report

Tissue Description

• Prostate (5.5 × 3.8 × 3.5 cm) and seminal vesicles (4 × 3.2 × 1 cm) are submitted and weigh 40 and 15 g, respectively. Tumor is identified grossly involving both sides of the prostate extensively, chiefly on the right.

• Pelvic lymphadenectomy tissue (right, 5.5 × 3 × 1 cm and 3 × 1 × 1 cm; left, 4.5 × 2 × 1 cm and 3 × 0.5 × 0.5 cm) submitted separately.

Diagnosis

• Radical retropubic prostatoseminovesiculectomy:

  • ADENOCARCINOMA (GLEASON GRADE 4 + 3 = 7). ISUP 2014 GRADE GROUPING 3. GLEASON PATTERN 4 COMPRISES 60% OF THE CANCER. TERTIARY PATTERN 5 COMPRISES 5% OF THE CANCER.

• Size: about 27.72 cm3

• Location: bilateral peripheral zone and transition zone

• Resection margins: negative

• Perineural invasion: extensive

  • Capsule: bilateral invasion and extensive multifocal right-sided extraprostatic extension (cumulative 1.1-cm linear length)

• Premalignant change: patchy high-grade prostatic intraepithelial neoplasia

• Pelvic lymph nodes: metastases to 2 of 9 right and 1 of 6 left pelvic lymph nodes

• Apex: involvement of the right anterior and posterior and left posterior quadrants without extension to the margin

• Bladder base: negative

• Seminal vesicles: positive on the right side

• Vascular/lymphatic invasion: extensive

• Other: nodular hyperplasia

• DNA content (flow cytometry): tetraploid (block C8; 60% cancer)

• TNM (2017 revision) stage: T3bN1M: not applicable

Table 9.21

EPE , Extraprostatic extension; ISUP , International Society of Urological Pathology.

Protocol for Processing Radical Prostatectomies

Frozen Sections of Lymph Nodes During Surgery

This is optional and at the discretion of the treating surgeon. If frozen sections are requested, then all lymph nodes and perinodal adipose tissue are submitted for evaluation. Permanent sections of all frozen tissue should be obtained.

Radical Prostatoseminovesiculectomy

Each prostate is weighed, measured in three dimensions, and inked. Inking is performed by coating with India ink (or another preferred color) and rapid (1-2 seconds) immersion in acetone to create ink adherence to the tissue. Some use more than one ink to mark different anatomic sites (e.g., left and right sides).

Apex and Base

Conization or shave margins are acceptable at the discretion of the pathologist; however, each pathologist is encouraged to use only one of these methods for all of their cases. Cancer involving the apex or base is not considered EPE even if the margins are positive (exception: when adipose tissue is present and cancer is in contact).

Conization Margins

• After fixation, the apex and base are amputated at a thickness of about 4 mm. The apical slice is divided into four quadrants, and each is serially sectioned at 3-mm intervals in the vertical parasagittal plane, similar to a cervical conization specimen, and submitted separately. The section from the bladder base is sectioned in a similar manner, although the amount of tissue was usually less, with divisions as hemispheres rather than quadrants. Cancer touching ink is considered a positive margin. For conization the apex usually requires quadrant sectioning, and we routinely use abbreviations for the right anterior apex (RAX), left anterior apex (LAX), right posterior apex (RPX), and left posterior apex (LPX). Similarly, the base is sampled, usually into left and right halves as left bladder base (LBB) and right bladder base (RBB), respectively. Advantages include greater localization of the cancer and determination of proximity to the margin when the margin is negative.

Shave Margins

• A thin, translucent shaving of the entire face of the apex and face of the bladder should be taken. Any evidence of cancer in these specimens indicates a positive margin. Advantages include encompassing the entire surface of the apex and base for analysis.

Seminal Vesicles

The seminal vesicles are amputated from the prostate at the junction of the two organs without incising the prostate itself; a slice encompassing each seminal vesicle at this junction should be submitted for routine histologic examination. If cancer is identified in the seminal vesicles or adjacent soft tissues, then the seminal vesicles should be serially sectioned in a manner similar to the prostate and submitted entirely for histologic review.

Prostate

The prostate is serially sectioned as thinly as possible (about 4- to 5-mm-thick sections) by knife in a coronal plane perpendicular to the long axis of the gland from the apex of the prostate to the site of the amputated seminal vesicles. Orientation and ordering of the slices are maintained throughout to allow spatial reconstruction later of the prostate. The transverse sections are submitted in total for routine processing through neutral-buffered formalin (or equivalent alternative fixative) and sectioning as whole-mount sections (preferred) or after subdivision into two parts, four parts, or more, depending on the size of each slice. A record should be maintained so that the prostate can be reconstructed at histologic review. Routine sections should be stained with hematoxylin and eosin.

Protocol for Results Reporting of Radical Prostatectomies

Volume of Prostate

The volume of the prostate is calculated using the formula for a prolate ellipsoid, defined as length × height × width × 0.532 (correction factor for a prolate ellipsoid); no shrinkage correction factor is necessary because these measurements are made in fresh specimens.

Cancer and Ablated Tissue Volume, Location, and Extent

On each slide the exact outlines of carcinoma and ablated tissue are dotted in different color inks. The area of EPE is measured by an ink line drawn along the prostatic surface, and sites of positive surgical margins are indicated with a plus (+) sign.

Cancer volume is calculated by the grid method. In brief, a transparent grid of premeasured squares is placed over the slides, and the number of squares overlying carcinoma is counted; the total number of squares per case is multiplied by the area of each square, and the sum is multiplied by the thickness of each slice of the prostatectomy. Slice thickness is calculated by dividing the fresh tissue measurement of the long axis of the prostate minus 4 mm for conization (accounts for apical section amputation) (2 mm for shave margins) by the total number of slices of the prostate. To include the volume of cancer in the apical and basal sections, the number of grids overlying cancer from these sections is multiplied by 0.09 cm ² and then by 0.5 cm (section thickness). Final cancer volume is multiplied by 1.25, to account for tissue shrinkage caused by fixation.

The area of EPE for each case is calculated as the sum of measured line lengths on traced slides multiplied by the section thickness. The volume of tumor in the seminal vesicles is calculated by multiplying the number of grids overlying tumor in the muscular wall of the seminal vesicles by the section thickness. Positive surgical margins are defined as ink touching tumor cells (or foci of ablation); the number of separate foci with positive surgical margins is evaluated. Determinants of pathologic staging include evaluation of seminal vesicle involvement (unilateral or bilateral, as well as volume of tumor), EPE (unilateral or bilateral, as well as sites of perforation and area of EPE), and lymph node involvement (unilateral or bilateral, as well as number of nodes involved). Do not use focal versus established terminology.

The orientation of the long axis of the tumor is determined, as well as the greatest diameter of the predominant tumor nodule. Tumor location and ablated tissue location are recorded as transition zone and/or nontransition zone, with no attempt to determine precise site of tumor origin because of significant overlap of tumor and multifocality. The number of tumor foci and ablated tissue foci are counted; tumor > 2 mm from a tumor nodule is considered a separate focus by convention. Perineural invasion and vascular/lymphatic invasion are considered focal (present in less than three separate foci when evaluated by × 400 microscopic fields) or extensive (three or more foci involved).

Cancer Grade

The following variables are evaluated: Gleason primary pattern, Gleason secondary pattern, Gleason score (sum of the primary and secondary patterns), and ISUP grade grouping. The percentages of Gleason primary patterns 4 and 5, and the sum of primary patterns 4 and 5 are estimated in 10% increments. The Gleason grading scheme is noted (e.g., classic Gleason grading, ISUP 2005 modification, ISUP 2014 modification).

The presence and extent of EPE in clinical stage T2 adenocarcinoma (and hence clinical staging error) is also related to the number of blocks processed. Current guidelines for the evaluation of radical prostatectomy specimens emphasize information that should be included in the pathology report but leave the decision regarding partial or complete sampling to the pathologist. The 2009 ISUP consensus conference recommended standardization of pathology reporting of prostatectomy specimens. In response to the question relating to how much of the prostate should be blocked, participants considered both partial and complete embedding of prostates to be acceptable if the method of partial embedding is stated. Pathologists have to balance the extra costs and time involved in processing entire specimens against the risk of missing important prognostic information.

All methods begin with weighing the specimen and measuring in three dimensions: apical to basal (vertical), left to right (transverse), and anterior to posterior (sagittal), with the maximum length of each dimension recorded. Prostate weight with and without the seminal vesicles should be recorded. For ultrasonographic measurements, radiologists often describe the shape of the prostate as a prolate ellipsoid (length × height × width × 0.532), but this is only a rough estimate that shows considerable variability. Separate measurements are made of the seminal vesicles.

The gland should be fully fixed before sectioning. Several inking systems are available. It is recommended that a minimum of two different colors be used to secure correct left-side and right-side identification. The number of colors may be increased to three or four to facilitate identification of anterior and posterior. The adhesion of the ink is subsequently improved by immersing the specimen in acetone or Bouin fixative.

The apical and bladder neck margins are a common location for positive margins and EPE. The cone method is preferred for blocking the apex and the base. A thick section is amputated from the apex and the base, then cut and embedded sagittally. Sagittal slicing offers the advantage over radial slicing of producing uniform thickness tissue blocks. The shave method of the apical margin is no longer recommended because it causes either underdiagnosis or overdiagnosis of positive margins. Assessment of EPE is also more difficult with shave margin blocks. However, thin shave margins at the base are not as undesirable as apical shave margins. The bladder neck margin consists of thick muscle bundles outside of the prostate, so a positive thin shave margin at this site indicates EPE.

The remaining specimen is serially sectioned at 4 to 5 mm thickness by knife to create transverse sections perpendicular to the long axis of the prostate from the apex to the tip of the seminal vesicles. Partial and complete sampling differ by the amount of prostate tissue submitted.

The Bostwick Laboratories protocol using whole mount sections for preparing and reporting prostatectomy specimens is illustrated with multiple examples in Fig. 9.25 . Complete and careful submission of tissue for histologic evaluation allows the following: unequivocal orientation of specimen and tumor (left, right; transition zone, peripheral zone; anterior, mid, posterior; apex, base, etc.); evaluation of the extent and location of positive surgical margins; assessment and quantitation of the extent and location of EPE and seminal vesicle(s) invasion; quality-control data for the surgeon, particularly in regard to surgical margins in nerve-sparing or robotic prostatectomy. This method also facilitates postoperative measurement of tumor volume for correlation with imaging studies, as desired; evaluation of tumor grade (percentage of poorly differentiated adenocarcinoma); fulfillment of all recommendations by the Cancer Committee of the College of American Pathologists; and comparison of results with published studies.

Standard template protocols are useful because of the frequent multifocality of prostatic adenocarcinoma, the inability to fully identify the location and extent of tumor by examining randomly chosen tissue slices, and the inability to grossly identify positive surgical margins and EPE. Despite our personal preference for complete submission, most pathologists undertake partial submission, and this is practical for most cases, does not require special processing or large cassettes, and provides all necessary clinical information.