The normal glandular epithelium of the prostate (Fig. 10.4) is classically defined as having two cell layers, namely a luminal (or secretory) layer and a basal layer. A third cell type is also present, the neuroendocrine cell, which is usually infrequently seen [16]. The secretory epithelial cells make up the majority of the epithelial volume. These cells are tall and columnar, with clear to pale cytoplasm and basally located nuclei. They are situated on the luminal aspect of the gland, as part of the double layer of epithelium. The nuclei are small and round with fine, evenly dispersed chromatin and nucleoli are usually not prominent. Between the secretory cell layer and the basement membrane lies the basal cell layer. The nuclei of the basal cells are small, and hyperchromatic with scant surrounding cytoplasm. These nuclei are flattened and oriented parallel to the basement membrane. The basal layer is often inconspicuous and incomplete, and it may be difficult to distinguish individual basal cells from underlying stromal fibroblasts in benign prostatic tissue. Recognition of the basal cell layer is, however, important, as these cells are absent in prostatic adenocarcinoma, and should be present in benign conditions that mimic cancer [17]. The final epithelial group, the neuroendocrine cells of the prostate, are irregularly distributed throughout the epithelial compartment [18]. They are inconspicuous, but can occasionally be recognised on haematoxylin and eosin (H&E) staining by deeply eosinophilic fine cytoplasmic granules [16]. The intraglandular contents of normal glands must also be examined and include corpora amylacea [19], (inspissated secretions that appear as pink, concentrically lamellated rings) (Fig. 10.5), calculi and degenerating epithelial cells.

Histologically, the non-glandular components of the prostate gland include the pseudocapsule, a fibromuscular layer most prominent along the base and posterior portion of the lateral lobes, the fibrous septa, the periprostatic adipose tissue and the stroma. Carcinoma invading adipose tissue is considered to represent extraprostatic extension [20]. Skeletal muscle fibres can also be seen in sections of prostatic tissue, admixed with the stroma at the distal apex of the gland. The prostatic stroma accounts for about half the volume of the prostate gland. The most abundant normal stromal cell type is the smooth muscle cell, derived from the embryonic urogenital sinus. In addition to being the most abundant cell type, the smooth muscle cell appears to be the most important cell type, with respect to homeostasis. Other stromal cells commonly present include fibroblasts, endothelial cells, nerve fibres with associated ganglia, and various immune cells. The extracellular matrix is rich in collagen fibres that intervene between the glandular epithelium. The benign prostatic stroma is affected by ageing, with inflammatory cells becoming more abundant and stromal fibroblasts becoming senescent with age.

PIN consists of architecturally benign acini lined by cells showing cytologic atypia; namely, nuclear crowding, enlargement, hyperchromasia and prominent nucleolation [29]. There is a surrounding basal cell layer and an intact basement membrane (Fig. 10.7). Most agree that PIN is a precursor of invasive adenocarcinoma, a theory which is supported by the fact that prostates with invasive carcinoma have an increased incidence of PIN, an increase in the number and size of PIN foci, and, on occasion, zones of HGPIN where there appears to be a ‘budding off’ of invasive carcinoma glands. PIN is subdivided into two grades, high and low grades, and the distinction between the two is based on the presence of prominent nucleoli, a diagnostic feature of HGPIN [29]. Most cases of low-grade PIN do not progress, and therefore, many pathologists no longer comment on its presence in biopsy specimens. The presence of multiple foci of HGPIN has a high predictive value for PCa, which develops within 1–2 years in an estimated 30 % of men with multiple cores containing HGPIN [30]. Great variation exists in the literature regarding the incidence of HGPIN on needle biopsy, which is reported to range from 1.5 % to about 16 % [31]. When HGPIN is identified on needle biopsy, a careful search is typically made by the pathologist for invasive cancer in the remainder of the tissue. If no cancer is found, follow-up is unnecessary if only a small amount of HGPIN is present in 1–2 cores. However, if there is a more significant amount of HGPIN, a follow-up protocol is adopted, consisting of serum prostate-specific antigen (PSA) testing, physical examination, and possibly repeat biopsies. However, recommendations for following men with HGPIN have varied widely [32].

PIA refers to foci of epithelial atrophy, usually in the periphery of the gland, which show a high proliferative rate, and the presence of mononuclear and polymorphonuclear inflammatory cells within both epithelial and stromal compartments [26]. PIA is associated with chronic prostatitis [26], and it has been hypothesised that repeated tissue injury and cell loss results in proliferative regeneration of damaged epithelium which may subsequently progress to in situ and/or frank invasive carcinoma [33]. The glutathione-S-transferase Pi 1 (GSTP1) gene is located on chromosome 11q13 [21], and plays an important part in cell detoxification [34]. Hypermethylation of the GSTP1 gene (which results in downregulation of GSTP1 protein expression) [21] is seen in PIA, as well as in >90 % of PCa [34]. Importantly, GSTP1 protein detected in free plasma circulating DNA has proved to be a promising biomarker for detection of primary and recurrent PCa [34].

Histologically, when prostatic adenocarcinoma is referred to without further qualification, it can be assumed to be the common, or acinar, variant [21]. The glands are smaller and more crowded than their benign counterparts under the microscope, and have straight luminal borders rather than papillary infoldings. Having lost the outer basal cell layer, the neoplastic glands are lined by a single layer of cuboidal or low columnar epithelium. The cytoplasm of the malignant cells ranges from amphophilic to the pale/clear cytoplasm seen in benign glands. Nuclei are enlarged and often contain one or more prominent nucleoli. Typically, nuclear pleomorphism and mitotic activity are not prominent. Intraluminal crystalloids or mucinous secretions, when present, are observed mostly in carcinomas, but only rarely identified in benign glands [35]. Although there are a few specific features of malignancy on biopsy (e.g. perineural invasion (Fig. 10.8), which is seen in approximately 20 % of prostate biopsies), the diagnosis is typically made on a combination of architectural and cytologic features on routine H&E. The diagnosis of PCa may be difficult because the clues to malignancy are often subtle, and also because there are several benign mimickers of PCa, such as adenosis, atrophy, and basal cell hyperplasia.

The majority of prostate tumours are adenocarcinomas arising from epithelial cells lining the prostatic glands. As a consequence, prostatic epithelial cells have been the major focus of histopathological examination and research studies to date. While the development of an altered stromal microenvironment in response to carcinoma is a common feature of many other tumours [36, 37], emerging evidence suggests that changes occur in the surrounding prostatic stroma also, that may serve to enhance the malignant potential of the nearby epithelium [38]. Tumour-associated stroma is referred to as ‘reactive’ in the prostate, as the phenotypic and genotypic alterations seen are similar to those seen in wound repair, including matrix remodelling and altered expression of growth factors and cytokines [39]. The stroma normally consists predominantly of smooth muscle cells, however, during carcinogenesis, these cells are gradually replaced by the activated form of fibroblasts, termed carcinoma-associated fibroblasts (CAFs). There is some debate as to whether CAFs and myofibroblasts represent different cell types, or whether they are the same cell type with different gene expression profiles [40, 41]. It is also unclear as to whether these CAFs are generated by activation of fibroblasts and perivascular cells already present in the prostatic stroma, or whether they are recruited from other locations, i.e. from the bone marrow. CAFs have been shown to affect cancer cells by secreting growth factors [42], extracellular matrix (ECM) components [43] and proteases [44].

Tumour development is also associated with an influx of macrophages [45], lymphocytes and mast cells into the tumour stroma [34]. These inflammatory cells secrete cytokines, which have stimulatory or inhibitory effects on adjacent CAFs, blood vessels and epithelial cells. While inflammatory cells are generally viewed as beneficial, e.g. in wound repair and in preventing the spread of infection, tumour-associated inflammation appears to be associated with an increased rate of growth and the spread of cancer [46, 47]. This large and varied list of stromal cell types may explain the heterogeneity observed in PCa and further research into stromal–epithelial interactions [48] must be performed.

The most valuable adjunctive study for the diagnosis of adenocarcinoma of the prostate is immunohistochemistry (IHC), which is used currently in up to 20 % of all prostate needle core biopsies. Any IHC marker, however, carries a risk of false-positive and false-negative results, and therefore must be used in conjunction with the H&E morphology. The immunophenotype of prostatic basal cells is distinctive and can be of diagnostic value, as positive staining with antibodies directed against basal epithelial cells (34β[beta]E12 and p63) rules out the presence of invasive carcinoma [21, 35]. Nuclear basal markers can also be used when the differential diagnosis is between atypia, atypical adenomatous hyperplasia, PIN and atrophy. In some cases, a basal cell marker cocktail can be used to further improve detection of basal cells [49]. In contrast, the normal secretory cells of the prostate stain with pan-cytokeratins, cytokeratins 8 and 18, androgen receptor (AR), PSA and prostate-specific acid phosphatase (PSAP). This immunoprofile distinguishes them from the underlying basal cells. Normal neuroendocrine cells are characterised by their immunoreactivity for synaptophysin, chromogranin or CD56. They variably express AR, an important feature in relation to castration-resistant disease in prostatic carcinoma with neuroendocrine differentiation [50, 51].

Alpha-methylacyl-CoA racemase (AMACR, also known as ‘P504S’ or ‘racemase’) [35] is a commonly used cytoplasmic marker which is selective and very sensitive for carcinoma. While some benign glands (e.g. atrophic glands) and glands showing PIN can also stain positively for AMACR [52], the staining is usually more focal and weaker in comparison to neoplastic glands. Some laboratories combine antibodies in a cocktail also, using a basal marker together with AMACR for definitive diagnosis [53]. A ‘triple stain’ of p63, 34β[beta]E12, and AMACR in prostatic adenocarcinoma can be seen in Fig. 10.9. Stromal tumour markers are not used as routinely in the diagnosis of PCa, however, it is known that decreased expression of smooth muscle cell markers (desmin, smooth muscle actin) and increased fibroblast markers (vimentin) are seen [54]. As no therapeutic targets have currently been identified in PCa, IHC for proteins such as AR, HER2, CD117 (c-Kit), EGFR, etc., are not routinely employed as they are not predictive of response to treatment. Therefore, there is an urgent need to identify molecular signatures to discriminate lethal and indolent PCa and to identify predictive and prognostic biomarkers. While most studies in the past have concentrated on epithelial-based signatures, the stromal microenvironment may hold the answer to this question.

Histologic grading is the most useful tissue-based prognostic predictor in PCa [55]. The quintessential grading scheme for PCa, the Gleason grading system, is based solely on the architectural pattern of the tumour. Both the predominant (‘primary’) and the second most prevalent (‘secondary’) patterns are identified [56]. Each is assigned a numeric value, and these are added together to yield the ‘Gleason Score’. The grading system originally devised in 1966 by Donald Gleason categorised prostate cancer into five separate grades ranging from 1–5, representing increasingly more aggressive and less differentiated morphologic patterns. In practice, however, grades 1 and 2 are no longer in use, and the primary and secondary patterns of the tumour are generally assigned a value from 3 (well differentiated, with individual, discrete, well-formed glands) to 5 (poorly differentiated/undifferentiated tumour, or tumour with comedonecrosis). A Gleason grade of 4 indicates that the tumour displays an intermediate level of differentiation; poorly formed glands, fused glands, or glands with cribriform architecture [56]. In practice, therefore, the Gleason score on prostatic specimens usually falls within the range of 6–10 [56]. Examples of Gleason patterns (i.e. grades 3+3, 3+4, 4+4 and 5+4) are shown in Figs. 10.10, 10.11, 10.12 and 10.13. In addition, it is advisable to report the percentage of Gleason pattern 4 tumour in a biopsy specimen with an overall Gleason Score of 7; this is in order that clinicians are aware of the approximate extent of the less differentiated (and therefore more aggressive) tumour present in the biopsy material. Increasingly, pathologists also comment on the presence of a ‘tertiary’ component—this situation arises where a third, typically more aggressive pattern is identified. Studies have shown that the presence of a tertiary higher grade component is associated with an increased risk of biochemical recurrence, and it is currently recommended that tertiary grade patterns be recorded in the pathology report in order to accurately reflect the overall grade [57]. The approach to tertiary grading differs between biopsy and prostatectomy specimens. For biopsies, the primary pattern and the pattern of highest grade should be added together to formulate the Gleason score. In the case of a higher grade, tertiary pattern, this would result in the exclusion of the secondary pattern from the overall Gleason score. Conversely, in radical prostatectomy specimens, the Gleason score should be based on primary and secondary patterns, with a comment added on the tertiary pattern [58]. Some special rules apply to Gleason grading of uncommon histologic variants of PCa [55], but these are beyond the scope of this discussion.

The Gleason grade remains the most reliable prognostic factor in PCa [59], however, there has been a gradual shift in how the Gleason grade is applied in practice, with a general trend towards upgrading. In addition, the Gleason grading system has been adapted over time in order to accommodate changing clinical practices [55]. In particular, a consensus conference which took place in 2005 [60] was organised in order to standardise the perception of the various patterns, as well as the manner in which the information relating to tumour grading was reported. The resultant ‘Modified Gleason System’ reflected this effort at standardisation, and one of its effects was a shift towards upgrading [55]. Studies have shown that needle biopsies using the Modified Gleason System correlated with progression after radical prostatectomy [61] and predicted biochemical recurrence after radical prostatectomy [62]. More recently (November 2014), a second major consensus conference was held with the aim of further updating the Gleason System. Among the parameters examined, an effort was made to standardise the grading of particular morphologic patterns and variants of prostate cancer, including those commonly encountered (e.g. cribriform glands), in addition to relatively uncommon patterns/variants (e.g. glomeruloid glands, mucinous adenocarcinoma, intraductal carcinoma). With regard to the grading of cribriform glands in PCa, according to the updated (2014) guidelines, all cribriform glands are now assigned a Gleason grade of 4, in contrast to the previous iteration of the Gleason System, which designated a Gleason Grade of 3 to small and relatively round/regular cribriform glands. [63] Interestingly, further prognostic information can be gleaned from examining the relative contributions of the individual patterns that ultimately make up the Gleason score. The most striking example of this is the dramatic difference seen between the clinical stages of tumours with a Gleason score of 7, depending on whether there was a primary pattern of 3 and a secondary pattern of 4, and vice versa (3 + 4 = 7, and 4 + 3 = 7, respectively). In one study, 95 % of the 3 + 4 = 7 tumours were clinically staged as pT2, while 79 % of 4 + 3 = 7 cancers were pT3 or pT4 [64]. The prognostic implications of both the overall Gleason score and the relative proportions of the individual component patterns were also discussed at the 2014 consensus meeting, and a novel grading system was proposed. The practice of combining the most common and second most common Gleason patterns to obtain an overall Gleason Score was retained, but the new system stratifies tumours with Gleason Scores of 2–10 into prognostically distinct ‘Grade Groups’ numbered 1–5. Grade Group 1 tumours are composed of only discrete, well-formed glands (therefore with a Gleason Score of 6 or lower), while Grade Groups 2 and 3 comprise tumours of Gleason Score 7 whose component patterns are 3 + 4 = 7 and 4 + 3 = 7, respectively. Grade Group 4 contains tumours of Gleason Score 8, while the most aggressive tumours (Gleason Score of 9 or 10) make up Grade Group 5. These ‘Grade Groups’ are felt to carry greater prognostic value and to more accurately reflect the biology of prostate cancer than the previous Gleason system and various algorithms which had been used for prognostic ‘grouping’ based on Gleason Score. Furthermore, a perceived flaw of the older system was that although 6 was the lowest score assigned to a tumour in practice, a score of 6 out of 10 nevertheless might contribute to patient fear by implying an intermediate rather than an excellent prognosis. It is thought that the new Grade Groups would—for the foreseeable future—be used in conjunction with the Gleason System. The new system and its terminology ‘Grade Groups 1–5’ have been accepted by the World Health Organisation for the 2016 edition of Pathology and Genetics: Tumours of the Urinary System and Male Genital Organs [63].

Lastly, because of heterogeneity and sampling error, some studies report that up to 28 % of cases are upgraded in radical prostatectomies compared to biopsies [65, 66]. This may have a significant impact on clinical decision-making, and can constitute a confusing and problematic factor with regard to the choice of therapeutic approach for a given patient (e.g. active surveillance versus surgery or radiotherapy) [59, 60, 67].

Local extension of prostatic carcinoma is usually into periprostatic tissue, seminal vesicles, and bladder base [21]. Cancer staging takes into account whether the cancer is unilateral or bilateral and the presence or absence of invasion beyond the prostate or into surrounding structures. [24] Metastases initially spread via the lymphatics to the obturator nodes, and eventually to the paraaortic nodes. Haematogenous spread is chiefly to the bones, in particular the axial skeleton, proximal femurs and ribs. However, rare cancers can spread widely to the viscera [21]. The reader should refer back to Chap. 3 for an overview of the principles of cancer staging. Organ-specific TNM staging information can be obtained from the current American Joint Committee on Cancer (AJCC) staging manual [68].

Aside from the common, acinar variant, several distinct morphologic subtypes of prostatic carcinoma exist. These rare subtypes are beyond the scope of this book chapter, but they are described in detail in the current World Health Organisation/International Agency for Research on Cancer’s (WHO/IARC) Classification of Tumours publication relating to the urinary system and male genital organs [69]. Of the uncommon variants of PCa, carcinoma with neuroendocrine (NE) features merits a brief discussion, as it is increasingly recognised as an aggressive variant that is associated with castration-resistant disease [51]. The NE cells of the prostate lack ARs, and NE differentiation increases in the late stages of castration-resistant PCa and in response to androgen deprivation therapy. Transformation to a predominantly AR-negative PCa with increasing NE differentiation may be an important resistance mechanism in castration-resistant disease, and indeed may be more common than was previously recognised. NE tumours of the prostate comprise a heterogeneous group, and the current WHO classification of these cancers includes the following subtypes: Focal NE differentiation in otherwise conventional prostatic adenocarcinoma, carcinoid tumour (well-differentiated NE tumour) and small cell carcinoma (poorly differentiated NE tumour). It has been suggested, however, that this basic classification should be broadened to include three further variants; namely, adenocarcinoma with paneth cell-like NE differentiation, large cell NE carcinoma and mixed NE-acinar carcinoma. Diagnosis of carcinoma with NE features is important, as it is considered an aggressive phenotype of advanced PCa with treatment strategies that differ from those of conventional adenocarcinoma (notably, the use of platinum-based therapies as opposed to AR-targeted treatment). The diagnosis generally depends on histology, as no reliable serum markers are currently available for identification of patients who are transforming to an NE phenotype. However, the detection and molecular characterisation of circulating tumour cells (CTCs) has been suggested for this purpose.