High-Grade Prostatic Intraepithelial Neoplasia

It has been more than a decade since specific criteria for the recognition and grading of prostatic intraepithelial neoplasia (PIN) were first published.1,2 Prostatic intraepithelial neoplasia is now divided into 2 grades: low grade (PIN 1) and high grade (PIN 2 and 3). The architectural spectrum characterizing high-grade PIN (HGPIN) has been clearly delineated,3 and there has been growing recognition among pathologists of PIN as a diagnostic entity distinct from infiltrating adenocarcinoma, as well as other mimics of adenocarcinoma.In recent years, the incidence of HGPIN found on needle biopsy has risen. Isolated HGPIN in needle biopsy specimens has been reported with a highly variable frequency of 0.7% to 16.5%, reflecting differences in the type of patient population studied.4–6The clinical significance of HGPIN is its strong association with carcinoma, and it is recognized as a putative precursor of peripheral zone carcinomas. The identification of isolated HGPIN in needle biopsy specimens warrants a recommendation for close follow-up with repeat biopsy. Carcinoma has been found in 22.6% to 58% of subsequent biopsies.5,7–10Case 99-20 was selected by the Surgical Pathology Committee of the College of American Pathologists as an example of HGPIN. The case was specifically chosen to emphasize the features that distinguish HGPIN from other benign and malignant lesions that mimic PIN.A 58-year-old man was found to have an elevated serum prostate-specific antigen (PSA) level of 7 ng/mL. Digital rectal and transrectal ultrasound examinations were normal. Thin-core sextant biopsies were performed. Six specimen containers labeled left and right base, mid and apex, respectively, were received. Each contained 3 tan tissue cores measuring 18 mm in length by less than 1 mm in diameter.In addition to this clinical information, participants were sent 7 selected kodachrome slides of hematoxylin-eosin–stained sections illustrating histologic features in one of the cores from the left base and photographed at varying magnifications. Three of these corresponding microscopic images are shown in Figures 1 through 3.Of the 3016 pathologists who returned the questionnaire in the second cycle of the 1999 Performance Improvement Program (1999 PIP B), 2990 (99%) included a diagnosis for case 99-20. A compilation of participant diagnoses is presented in the Table.Of the 2329 participants who evaluated the technical quality of the case, 704 (30.2%) considered the kodachromes excellent, 1488 (63.9%) felt they were satisfactory, 110 (4.7%) judged them less than optimal, and 27 (1.2%) considered them unsatisfactory.The lowest number of participants responding to each of 3 supplemental questions based on the master list of diagnostic possibilities was 2853. The largest number of responses (2880) came from participants asked to choose “a putative precursor lesion of prostatic adenocarcinoma,” and 98.5% (2837) correctly selected HGPIN. Minority choices included basal cell hyperplasia (0.7%) and clear cell cribriform hyperplasia (0.6%).Prostatic intraepithelial neoplasia is characterized by glands that are architecturally benign but have cytologic atypia. On low power, foci of HGPIN are usually identified as complex medium-sized glands that have increased basophilia (Figure 1). The intraglandular proliferation of atypical secretory cells in HGPIN has many cytologic features mimicking adenocarcinoma, but without infiltrative growth. There are 4 characteristic architectural patterns of HGPIN.3 The tufting and micropapillary patterns, as illustrated in PIP case 99-20 (Figure 2), are most frequent, while the cribriform and flat patterns are less common. The atypical secretory cells have enlarged hyperchromatic nuclei with relatively high nuclear-cytoplasmic ratios, parachromatin clearing, and focally prominent nucleoli (Figure 3). Luminal cytoplasmic apical blebs are a constant feature of HGPIN regardless of the architectural pattern. Cytoplasmic amphophilia frequently contributes to the basophilia of the glands. Residual basal cells are present, although they characteristically have a discontinuous pattern.In PIP case 99-20, the overwhelming majority of participants (91%) correctly identified HGPIN, reinforcing the fact that this distinct entity is a reproducible diagnosis. This diagnostic accuracy was achieved in spite of the diverse experience and practice settings of the participants and the limitations of the artificial kodachrome format.A study by Epstein et al11 evaluated interobserver reproducibility among 7 experienced urologic pathologists in the diagnosis of PIN. Twenty-five needle biopsy cases were selected to represent the full spectrum of diagnostic issues, and each pathologist analyzed exactly the same focus. Lesions were combined into 3 groups for data analysis: (a) benign/PIN 1, (b) PIN 2/PIN 3/HGPIN cannot rule out cancer, and (c) HGPIN plus cancer. The level of agreement for the 3 groups was substantial (κ = 0.61; P < .00001). In general, there was good distinction between low-grade PIN (PIN 1) and HGPIN (PIN 2–3). The best interobserver agreement was for the category benign/PIN 1 (κ = 0.77), and the next best level of interobserver agreement was for the combined group HGPIN/HGPIN cannot rule out cancer (κ = 0.57).In an interobserver variability study by Allam et al,12 8 pathologists with various levels of interest in prostatic pathology reviewed 321 prostate biopsy specimens. The level of agreement for HGPIN was moderate (mean κ = 0.451). These results indicated that there is considerable variability in the diagnosis of HGPIN as made by practicing pathologists. However, a higher level of agreement may be achieved by pathologists with expertise in prostatic pathology. Causes of interobserver disagreement in the diagnosis of HGPIN were attributed mainly to differences in the visual estimation of the nucleoar size, but also to the minuteness or focal distribution of HGPIN, partial glandular involvement, the flat variant of HGPIN missed on low- or intermediate-power examination, and confusion with basal cell hyperplasia and/or inflammatory changes in which nucleoli are increased in size.For PIP case 99-20, false-positive diagnoses of carcinoma were rendered by 2% of participants, including a small minority (0.2%) who favored transitional cell carcinoma. Two-dimensional static images may have generated diagnostic uncertainty, since multiple levels are commonly needed to distinguish out-pouchings or tangential sections of PIN glands from a few atypical microacini budding off HGPIN and representing an apparent early phase of invasive carcinoma. Also, the ability to focus a microscopic slide enhances diagnostic recognition of cytologic features. However, only 10 participants (0.3%) diagnosed prostatic adenocarcinoma, not otherwise specified. A cribriform architecture and glands with necrosis were not present in the PIP case, but can cause difficulties in distinguishing HGPIN from prostatic acinar or ductal carcinomas.Finally, 7% of the participants favored a benign mimic of PIN, including basal cell hyperplasia (3.7%), clear cell cribriform hyperplasia (1.2%), normal seminal vesicle (1.4%), and transitional cell metaplasia (0.7%). Although these choices are clearly incorrect, the clinical impact of this type of misdiagnosis is less than that of a false-positive diagnosis of carcinoma. An underdiagnosis of HGPIN, however, can have clinical implications in that the patient may not undergo repeat biopsy, particularly if the serum PSA level is not elevated.For patients with isolated HGPIN, the incidence of subsequent carcinoma in repeat biopsies has been shown to be high, regardless of the serum PSA levels (0–4 ng/mL, 33.3%; 4.1–10 ng/mL, 38.1%; and >10 ng/mL, 61.9%).13 In another study, HGPIN provided the highest risk ratio (14.93) of all known predictive factors, including patient age, digital rectal findings, and serum PSA concentration.9 When identified in biopsies from patients with a serum PSA level greater than 4 ng/mL, the predictive value of HGPIN was even higher, with 90% of patients having cancer on follow-up biopsy within 2 years, although they tended to be older.The original critique writer for Performance Improvement Program case 99-20 was John R. Srigley, MD, The Credit Valley Hospital, Toronto, Ontario, Canada.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.