Intraductal carcinoma of the prostate.

Intraductal carcinoma of the prostate (IDC-P) is a diagnostic entity characterized by architecturally or cytologically malignant-appearing prostatic glandular epithelium confined to prostatic ducts. Despite its apparent in situ nature, this lesion is associated with aggressive prostatic adenocarcinoma and is a predictor for poor prognosis when identified on biopsy or radical prostatectomy. This review discusses diagnosis, clinical features, histogenesis, and management of IDC-P, as well as current research and controversies surrounding this entity.

Introduction

Despite its apparent in situ nature, intraductal carcinoma of the prostate (IDC-P) is associated with aggressive behavior and poor outcome. While this may appear counterintuitive, current theories on histogenesis of IDC-P suggest that many of these lesions represent adenocarcinoma with retrograde invasion into the prostatic ducts [1]. Thus, most IDC-P are not biologically or pathologically distinct from conventional prostatic adenocarcinoma but rather a morphologic manifestation of the aggressive nature of the underlying cancer. An overview of the clinical and pathological features of IDC-P will be provided herein with an emphasis on current research and controversies regarding this entity.

Epidemiology and diagnostic criteria

First described in an autopsy study in 1938 [2], IDC-P became more widely studied in the mid-1980’s with the increasing use of immunohistochemistry [3,4]. Diagnostic criteria for IDC-P were first proposed by Guo and Epstein in 2006 (Tab. I, Fig. 1A-1D) and subsequently endorsed by the World Health Organization (WHO) tumor classification, which recognized IDC-P as a distinct entity in 2016 [5,6]. The WHO defines IDC-P as an “intra-acinar and/or intraductal neoplastic epithelial proliferation that has some features of high grade prostatic intraepithelial neoplasia (HGPIN) but exhibits much greater architectural and/or cytological atypia, typically associated with high grade or high stage prostate carcinoma” [6].

Tab. I.

Diagnostic criteria for IDC-P (left panel) and AIP (right panel). Adapted from Guo & Epstein 2006 [5] and Shah et al. 2017 [15].

Intraductal carcinoma of the prostate	Atypical intraductal proliferation
Malignant epithelial cells filling large acini and prostatic ducts Solid or dense cribriform pattern with cellular density > 50% of luminal space Loose cribriform or micropapillary pattern with either Marked nuclear atypia with nuclei at least 6x larger than adjacent benign nuclei or Nonfocal comedonecrosis Preservation of basal cells	Malignant epithelial cells filling large acini and prostatic ducts Solid or dense cribriform pattern incompletely spanning the lumen with cellular density < 50% of luminal space Loose cribriform or micropapillary pattern with both Insufficient nuclear pleomorphism to meet diagnostic criteria for IDC-P Absence of necrosis Preservation of basal cells

Figure 1.

Intraductal carcinoma of the prostate (IDC-P) and atypical intraductal proliferation (AIP): (A) Solid pattern of IDC-P (Original magnification ×40); (B) Nuclei 6x larger than those of adjacent benign cells (Original magnification ×200); (C) Dense cribriform pattern with comedonecrosis (Original magnification ×100); (D) Preserved basal cell staining by immunohistochemistry for p63 (Original magnification ×100). (E) Loose cribriform pattern with lack of pleomorphism and comedonecrosis in AIP (Original magnification ×100); (F) Preserved basal cell staining and AMACR positivity in AIP (Triple stain with p63, HMWCK and AMACR) (Original magnification ×40).

The prevalence of IDC-P is largely dependent on the patient cohort, ranging from 2-3% in low-risk or overall patient cohorts [7,8] to 67% in metastatic/recurrent disease [8-10] for biopsy specimens, and 13-17% in radical prostatectomy (RP) specimens [11,12].

Lesions that do not meet morphologic criteria for IDC-P but appear more complex than HGPIN either architecturally or cytologically have been termed “atypical intraductal proliferations” (AIP) [13-15] (Tab. I, Fig. 1E-F).

Clinical features and reporting recommendations

IDC-P is independently associated with adverse clinical and pathological features on both biopsy and RP. Specifically, RP studies have shown that patients with IDC-P have higher pathologic T stage [10,16-18], higher Gleason score [10,16-20], increased tumor volume [3,4,10,16-19], and increased likelihood of extraprostatic extension [5,10,16,20], seminal vesicle invasion [10,19], and lymph node metastases [10,20]. These patients also have an increased risk of biochemical recurrence [21,22] and metastasis [22] and worse progression-free, cancer-specific, and overall survival [10,20].

Patients with IDC-P detected on biopsy have higher Gleason score and likelihood of seminal vesicle invasion on RP [7] as well as worse progression-free, cancer-specific, and overall survival [23-25]. While the finding of IDC-P on biopsy is highly specific for detection of IDC-P on RP, the sensitivity is relatively low (56.5% overall and 34.1% for active surveillance cohorts), making a positive finding more prognostically valuable than a negative one [26].

In the setting of metastatic disease, patients with IDC-P show worse cancer-specific and overall survival [23,27,28] and increased likelihood of development of castration-resistant prostate cancer (CRPC) [28]. In metastatic CRPC, IDC-P is a predictor of decreased cancer-specific and overall survival [29-31] as well as early relapse following neoadjuvant chemotherapy [32], relapse following androgen deprivation therapy [33] and radiation therapy [9]. IDC-P is also highly prevalent after androgen deprivation therapy or chemotherapy (60%) [8].

There is some heterogeneity in outcomes reported with different patterns of IDC-P. Solid and dense cribriform patterns are prognostically more unfavorable than loose cribriform or micropapillary patterns in localized [19] and metastatic disease [27].

While conventional IDC-P represents retrograde spread of invasive high grade carcinoma and is associated with aggressive features at RP, studies have suggested that isolated IDC-P or IDC-P with low grade adenocarcinoma may represent a precursor, de novo lesion associated with better clinical outcome [34-36]. The former category is termed “regular type” IDC-P, as opposed to “precursor-like” IDC-P, which is considered a precursor to invasive adenocarcinoma [34]. These two entities are currently pathologically indistinguishable. Isolated IDC-P, i.e., IDC-P without concurrent invasive carcinoma is found in 0.1-0.3% of biopsies [7,35].

Lesions diagnosed as AIP have clinical features more similar to IDC-P than to HGPIN. Patients with AIP and IDC-P on needle biopsy do not have a significantly different incidence of associated invasive carcinoma, although invasive carcinoma associated with IDC-P has higher Gleason score and higher percentage of needle core involvement on biopsy [15]. Patients with a diagnosis of AIP without associated prostate cancer on needle core biopsy have a 50% likelihood of a subsequent diagnosis of invasive carcinoma or IDC-P [13]. On RP, patients with invasive carcinoma and AIP or IDC-P have similar clinicopathologic characteristics, including no significant difference in Gleason score, tumor stage, seminal vesicle invasion, and surgical margins [11,12]. Patients with AIP carry an intermediate risk of biochemical recurrence between those with IDC-P and HGPIN [12].

There is significant controversy as to whether IDC-P should be included in the Grade Group system. Based on recent studies showing that IDC-P/cribriform architecture is a poor prognostic factor independent of Grade Group [37,38], some groups advocate its incorporation into the Grade Group system [37,39]. However, in view of the limited evidence on the association of IDC-P with low-grade invasive carcinoma, as well as the overlapping morphology of precursor-like and regular type IDC-P, some authors argue against incorporating IDC-P into the Grade Group system [40]. Furthermore, precursor-like and isolated IDC-P have a different expression pattern of PTEN and ERG from the concurrent low-grade cancer, suggesting that isolated IDC-P is unrelated to the associated low-grade invasive prostate cancer [41]. Currently, the European Association of Urology (EAU) guidelines do not support assigning a grade to IDC-P and recommend reporting IDC-P separately from Grade Group [42].

Differential diagnosis

IDC-P shares morphologic features with several entities. HGPIN is a precursor for prostatic adenocarcinoma and its diagnosis on biopsy (if unifocal, i.e. limited to one core) does not increase the risk of carcinoma detection on repeat biopsy [43,44]. Similarly to IDC-P, HGPIN consists of cytologically malignant cells confined to ducts or acini; however, HGPIN will not demonstrate the same degree of cellular density, architectural and cytologic atypia, or necrosis that are typical of IDC-P (Tab. I, Fig. 1A-1D). In practice, especially in the setting of limited tissue, this distinction is not straightforward, as some IDC-P can have a low-grade appearance [45]. Both IDC-P and HGPIN have preservation of basal cells and both express AMACR by immunohistochemistry. Immunohistochemical ERG expression or PTEN loss may be helpful to support a diagnosis of IDC-P in this context, as these features are not typical of HGPIN [13,46-48] (Fig. 2).

Figure 2.

Molecular alterations in intraductal carcinoma (highlighted in bold) and invasive carcinoma of the prostate.

The distinction between IDC-P and invasive acinar adenocarcinoma (cribriform Gleason pattern 4 or solid Gleason pattern 5) can be challenging on H&E sections, but it can be easily resolved by immunohistochemistry with the former displaying basal cell staining and the latter having complete loss of basal cells. Given that IDC-P is confined by ducts, these will have smoother contours, well circumscribed borders and follow a more organized branching architecture [5]. The presence of comedonecrosis strongly suggests a diagnosis of IDC-P [49], while perineural invasion or extraprostatic extension favor invasive carcinoma [50,51]. As both entities represent high grade malignancy, clinical management is similar [52,53].

Despite similarities in nomenclature, IDC-P and ductal adenocarcinoma have distinct morphologies. Ductal adenocarcinoma is characterized by tall pseudostratified columnar cells commonly forming true papillae with fibrovascular cores and sometimes displaying cribriform architecture. IDC-P has cuboidal (acinar) cells, can display micropapillary architecture without fibrovascular cores, and has a greater degree of cytologic atypia. Notably, ductal adenocarcinoma can occasionally involve the ducts and demonstrate persistence of basal cell markers. In these cases, a diagnosis of “intraductal carcinoma with ductal features” is rendered. Ductal adenocarcinoma is graded as Gleason pattern 4 [6].

Urothelial carcinoma involving prostatic ducts can sometimes be challenging to distinguish from the solid pattern of IDC-P on H&E. Extreme pleomorphism, high mitotic activity or apoptosis, and dense eosinophilic cytoplasm are morphologic features that favor urothelial carcinoma, while any glandular differentiation favors IDC-P. Fortunately, these entities are easily distinguishable by immunohistochemical markers as IDC-P will stain with prostatic markers, e.g. PSA, PSAP, p501s, and NKX3.1, while urothelial carcinoma will stain with HMWCK, p63, and GATA3. This distinction is important for management, as urothelial carcinoma involving prostatic ducts will be managed with radical cystoprostatectomy [43,52].

Molecular features and histogenesis

The most common early molecular aberration in prostatic adenocarcinoma is the fusion between the androgen regulated serine protease TMPRSS2 and members of the ETS transcription factor family, most commonly ERG, driving overexpression of the ERG oncogene [54-56]. This rearrangement is mutually exclusive with other early driver mutations such as those involving SPOP. The prognostic significance of the TMPRSS2-ERG rearrangement is unclear [57-61]. Additional second line genetic alterations, all associated with more aggressive disease, include PTEN, TP53, RB1, MYC, and BRCA2 [54,62,63]. Of these, PTEN loss has been the most studied and has been shown to be significantly associated with high Gleason grade, advanced tumor stage, and decreased time to metastasis [64,65]. Large scale genomic alterations, including copy number alterations, deletions, gains, and fusions comprise the majority of genetic alterations in prostate cancer.

IDC-P predominantly shows TMPRSS2-ERG fusion as an early driver mutation and has numerous second line genetic alterations associated with aggressive and metastatic disease (Fig. 2). Of note, HGPIN harbors the TMPRSS2-ERG fusion as well as PTEN loss much less frequently [10,66-69], while prevalence of these alterations in AIP is more similar to that seen in IDC-P [48,70]. Loss of heterozygosity (LOH), chromosomal imbalances, and percent genomic alteration have been found to be significantly higher in IDC-P compared with invasive prostate carcinoma and much higher than those seen in HGPIN [22,71-73].

Patients with germline mutations in genes associated with DNA repair (most commonly BRCA2, ATM, CHEK2, BRCA1) and BRCA2 specifically have been found to have a higher incidence of IDC-P [74-76], prompting recommendations for germline genetic testing in patients with this diagnosis even in the absence of family history [77].

Molecular findings also provide evidence for the histogenesis of IDC-P. Intraductal carcinoma and adjacent acinar adenocarcinoma share the same TMPRSS2-ERG genomic breakpoints, supporting the idea of a common clonal origin [1,78]. Additionally, invasive carcinoma adjacent to IDC-P with PTEN loss also shows similar PTEN abnormality, suggesting that IDC-P represents retrograde colonization of ducts by a subclone of an invasive carcinoma. The aggressiveness of this subclone is highlighted by the presence of similar copy number alteration patterns in IDC-P and metastasis [79].

Molecular findings also provide evidence for the distinction between regular type and precursor-like IDC-P. A subset of isolated IDC-P showed molecular alterations in the MAPK/PI3K pathway that are distinct from those of regular type IDC-P and conventional acinar adenocarcinoma [41]. This heterogeneity in IDC-P makes its categorization and clinical ramifications difficult and requires further research.

Management

There is no consensus on the management of IDC-P. Given the association of IDC-P with high grade and high volume invasive carcinoma on RP, the majority of authors recommend definitive therapy or suggest immediate re-biopsy when IDC-P is diagnosed. This is true even in the absence of invasive carcinoma [5,18,35,52] or with low grade, Gleason 3 + 3 = 6 carcinoma [36,52]. Because AIP on biopsy is associated with an increased risk of invasive carcinoma or IDC-P [13,48], this diagnosis warrants immediate re-biopsy [15]. Reporting IDC-P on radical prostatectomy is useful for prognostic purposes and for recommendation of germline testing; otherwise, clinical management is decided on a case by case basis [52,77].

Conclusions

IDC-P is associated with aggressive prostate cancer, early biochemical recurrence, and decreased survival. Diagnosis and early reporting are paramount for patient management due to significant therapeutic and prognostic implications. Further studies are needed to address current unresolved issues and fully understand the biological nature of this entity.