Gleason Grading of Prostatic Carcinoma

Problem Areas

• Undergrading
• Overdiagnosis of Gleason Score 2-4
• Focal Cancer on Biopsy
• Cribriform Patterns
• Fused Glands
• Borderline Histology

Undergrading

The single most important problem in Gleason grading is "undergrading", and the major reason for "undergrading", especially in needle biopsies, is the overdiagnosis of Gleason score 2-4. In addition, focal carcinoma on needle biopsy, cribriform and fused gland patterns, and "borderline" histology are often undergraded. This section will discuss overdiagnosis of Gleason patterns 2-4, including undergrading of "focal" carcinoma on needle biopsies. Cribriform and fused glands patterns and other "borderline" areas will be discussed subsequently.

Overdiagnosis of Gleason Score 2-4

Assignment of a Gleason score 2-4 essentially always means that pattern 1 and/or 2, as opposed to pattern 3, has been diagnosed. Pattern 1 and 2 (see "Gleason Patterns") are mostly seen in transition zone tumors. Both patterns are composed of circumscribed nodules of well defined, round to oval, usually moderate to large-sized glands with abundant pale cytoplasm. In pattern 3, the glands are obviously infiltrating, including extensively infiltrating adjacent benign prostatic tissue. The glands are generally smaller than pattern 1 and 2, and often have amphophilic cytoplasm. Pattern 2 glands may focally infiltrate adjacent benign prostatic tissue, but the infiltration is only minimal.

It is usually difficult or impossible to evaluate circumscription and minimal invasion on needle biopsies. Further, separating pattern 2 and 3 on the basis of the degree of separation of the glands and the morphology of glands is, at times, very difficult. It is not surprising, therefore, the the interobserver reproducibility of the diagnosis of Gleason score 2-4 in needle biopsies is poor, even amongst urologic pathologists.

It has recently been recommended that a diagnosis of Gleason score 2-4 should not be made on needle biopsy. The recommendation was made for several reasons. First, in a study of men coming to The Johns Hopkins Hospital for radical prostatectomy, there was concurrence with the outside needle biopsy grade of Gleason score 2-4 in only 4 of 87 cases; 68 were graded at Hopkins as Gleason 5-6, 13 as Gleason score 7, and 2 as Gleason score 8-10. Second, assigning Gleason score 2-4 on needle biopsies can adversely impact patient care, as clinicians may assume that low grade carcinoma on needle biopsy does not need definitive therapy. Some clinicians have concluded that since low volume Gleason score 2-4 tumors on transurethral resection (TURP) specimens have a relatively indolent course, all tumors with a Gleason core 2-4, whether on biopsy or TURP, have a low risk of progression. In the series noted above, 48 of the 87 cases graded as Gleason score 2-4 on needle biopsy at outside institutions showed extraprostatic extension at radical prostatectomy, including four cases with invasion of either seminal vesicles or lymph nodes. Even the four cases in which the diagnosis of Gleason score 2-4 was supported at Hopkins had higher grade tumor at radical prostatectomy. Third, as previously noted, interobserver reproducibility of the diagnosis of Gleason score 2-4 on needle biopsies, even among urologic pathology experts, is poor. Fourth, needle biopsies, unless specifically directed, usually do not sample transition zone tissue, the site of most Gleason score 2-4 carcinomas.

For the rare truly low grade carcinoma sampled on needle biopsy, there are three alternate grading approaches that can be taken. The first is to diagnose Gleason score 2-4 and add a note stating that "Tumors with a low grade appearance on needle biopsy almost always have higher grade tumor in the prostate and in up to 50% of cases are associated with extra-prostatic extension." Second, little harm will be done by assigning it a Gleason score 5 (Gleason score 2+3=5). Finally, these cases could be signed out as well to moderately differentiated prostatic carcinoma (Gleason score 2-6). by adopting one of these alternatives, one would not have to worry about the possibility, or implications of "undergrading". At the same time, the needle biopsy grade would more closely correlate with that found at radical prostatectomy and provide prognostically distinct groups to guide therapy.

Focal Cancer on Biopsy

A surprising number of pathologists believe that a microscopic focus of prostatic carcinoma on needle biopsy means that the carcinoma is low grade. This is not true. In fact, the presence of a microscopic focus in a needle biopsy essentially precludes the diagnosis of low Gleason score because low grade tumors consist of a fairly well circumscribed nodule numerous crowded glands with, in the case of pattern 2, only minimal invasion of adjacent prostatic tissue. Sampling such a nodule would yield multiple crowded cancer glands without admixed benign glands, not a small focus of cancer in between benign glands.

There are two additional comments relating to focal carcinoma on needle biopsies. First, we recommend assigning a Gleason score to even tiny (< 1mm) foci of prostatic carcinoma on needle biopsies because the scores of tiny foci are essentially as accurate as scores on needle biopsies with more extensive carcinoma in predicting Gleason score in subsequent radical prostatectomy specimens. A more practical and important reason for providing a Gleason score on focal carcinomas on needle biopsies is that we have seen some cases signed out as "Gleason grade 4" where the pathologist meant to convey that the tumor was high grade (i.e. Gleason pattern 4). However, the clinician interpreted it to mean a Gleason score of 4 (Gleason score 2+2=4). Assigning a Gleason score, including the primary and secondary pattern, even in cases with a limited amount of cancer, insures that the grade of the tumor is clear to the clinician.

Cribriform Patterns

Cribriform patterns 3 and 4 must be separated from one another. Further, the distinction of infiltrating cribriform carcinoma from intraductal cribriform proliferations may, at times, be difficult. This latter distinction is a diagnostic, rather than grading, issue and is not discussed within this web site.

Infiltrating cribriform carcinoma may be Gleason pattern 3 or 4. In cribriform pattern 3, the cribriform proliferations are small, rounded, and not crowded together. They are reminiscent of intraductal cribriform carcinoma of the breast. Cribriform pattern 4 proliferations may also be small and relatively rounded but they are usually larger, often in large sheets. The smaller relatively rounded lesions of pattern 4 are more crowded than pattern 3 and at least in areas, have irregular infiltrating borders. In needle biopsies, fragments of cribriform carcinoma are classified as pattern 4 because they almost certainly arise from large proliferations of cribriform carcinoma. Finally, there is a trend toward interpreting most infiltrating cribriform patterns as pattern 4. Indeed, in most cases with cribriform carcinoma, overt pattern 4 is also found elsewhere.

Fused Glands

Borderline Histology

Interpreting histology that is "borderline" between Gleason pattern is highly subjective, even for urologic pathologists. "Borderline" histology between Gleason pattern 1-2 and 3 and between cribriform Gleason pattern 3 and 4 has been discussed. (see "Undergrading" and "Cribriform Patterns.") The "borderline" between the small acini of Gleason pattern 3 and poorly formed small acini of Gleason pattern 4 is a major problem. Undergrading and overgrading is common. This "borderline" is particularly important because the presence of pattern 4 usually means the difference between assigning a Gleason score of 6 or 7, the latter having a significantly worse prognosis than the former. (see "Prognosis") Furthermore, the presence of pattern 4, and also pattern 5, have been shown to be a poor prognostic finding, even when present as a tertiary component. (see "Tertiary Gleason Patterns") In small acinar pattern 3, one should be able to mentally draw a circle around the small, but still well-formed, glands. Occasionally, a tumor will appear to be predominantly pattern 3 but will have a few glands with poorly-formed or no lumens. If one can convince oneself that these can be explained on the basis of tangential cutting, then a diagnosis of Gleason pattern 3 should be made. If on the other hand, there are "too many" glands with poorly-formed or no lumens where the lack of well-defined gland formation is visible at low magnification, a diagnosis of pattern 4 should be made. Occasionally, on needle biopsies, one will see a tumor that is predominantly pattern 4, with only two or three glands that are classical pattern 3. In that setting, we recommend that a diagnosis of Gleason score 4+3=7 be made because the prostatectomy specimen will usually contain significant areas of pattern 3. The "borderline" between patterns 4 and 5 is another problem. However, the clinical implications for this distinction are usually much less because it usually leads to the separation of Gleason score 4+4=8, 4+5=9, or 5+5=10, all of which are poorly differentiated and have similar behavior. As in the borderlines between 3 and 4, one must evaluate whether the lack of glandular differentiation in pattern 5 could be the result of tangential cutting of poorly defined glands of pattern 4. At times, nests of cells contain "vacuoles" as opposed to poorly defined lumens and these should be designated pattern 5. This latter separation is also highly subjective. In all of these "borderline" lesions, there will always be problems with interobserver and even intraobserver variability. Top | Pathology Consultation | Pathology Homepage | Gleason Grading Homepage Copyright © 2004 Johns Hopkins University All rights reserved. No part of this website may be reproduced in any form or by any means or incorporated into any information retrieval system, electronic or mechanical, without the written permission of JHU.