Practice Patterns of Clinicians Following Isolated Diagnoses of Atypical Small Acinar Proliferation on Prostate Biopsy Specimens

During the past decade, prostate cancer mortality rates have declined significantly.1 Notably, there has been an unmistakable downward trend in the incidence of late- stage disease, whereas the reverse is true for local or regional disease.1,2 The observed trends in mortality rates may be attributable to the US Food and Drug Administration's approval of the prostate-specific antigen (PSA) test for screening purposes in 1994, improvements in treatment modalities, previous misclassification of the causes of death in patients with prostate cancer, or other unknown factors.1–4 Although opinions differ on the relative significance of each of the aforementioned parameters, most observers would agree that the proliferation of PSA screening programs has had a significant and probably predominant role in these trends. One consequence of these programs has been the significant increase in the number of prostate biopsies performed in the United States, estimated at 1 million annually in one 1997 report.5 At the authors' center, the number of prostate biopsy specimens received for evaluation increased by 61% between 1994 and 2002. In addition to PSA screening, the increase in the number of biopsies performed has also been attributed to the use of the 18-gauge needle (which significantly simplifies prostate sampling compared to previously used needles), the aging of the US population, and increased public awareness of prostate cancer.5,6 Under these circumstances, biopsies are being performed for biochemical alterations that often precede any palpable prostatic anomaly. Thus, pathologists are expected to accurately classify increasingly minute foci of epithelial proliferations with respect to expected malignant potential, necessitating several publications focused just on morphologic criteria for the so-called limited or minimal volume adenocarcinoma (cancer involving less than 5% of biopsy specimen core tissue7) and minute cancers in general on prostate biopsy specimens.6–9 

Despite all these refinements in the morphologic criteria for cancer, there continues to be a small percentage of prostate biopsy specimens that show acinar proliferations that are suggestive of carcinoma but that fail to attain the requisite architectural or cytologic diagnostic threshold. Throughout the years, pathologists have conveyed their concerns over these cases descriptively based on, apparently, institutional or personal preferences. The diagnostic term atypical small acinar proliferation (ASAP) has been used by some authors to describe these foci,5,10 and similar terms are used at our center. Previous studies have demonstrated that this diagnosis is predictive of malignancy in the subsequent biopsy specimen in 34% to 60% of cases.7 Our objective in this study is to evaluate the effect of this diagnosis on practice patterns of clinicians when any other neoplastic or preneoplastic lesions are absent in a prostate biopsy specimen.

Pathologic reports of all prostate biopsy specimens with a diagnosis of ASAP during a 7-year period (1996–2002) were retrieved from the computerized database of the pathology department at Yale-New Haven Hospital. Cases with concurrent diagnoses of adenocarcinoma and/or prostatic intraepithelial neoplasia were excluded from this study. For the remaining cases, medical and pathology records were evaluated and the following information catalogued: patient age, initial total PSA level, digital rectal examination (DRE) findings, number of sites (bilateral, quadrant, sextant, octant) and cores with each biopsy specimen, and, where applicable, duration to subsequent biopsy (in days), pathologic findings on subsequent biopsy specimens, and pathology report characteristics. For cases with no subsequent biopsy specimens, follow-up was at least 8 months and extended up to 6.5 years in some cases. Ten different pathologists gave an ASAP diagnosis during this period. Statistical significance of the differences between the group means of continuous variables (PSA, age, duration to follow-up biopsies, and number of biopsy sites) was evaluated using the t test (Excel, Microsoft Inc, Redmond, Wash), whereas the χ² test (Statview 5, SAS Institute, Cary, NC) was used for nominal clinicopathologic data (DRE findings and pathology report characteristics); P < .05 was considered statistically significant.

Fifty-five (2.8%) of 1964 prostate biopsies performed during this period provided the diagnosis of ASAP, of which 36 met our study criteria. The average age for these 36 patients was 65 years (median, 66 years; range, 48–83 years) (Table 1). The prostate was entirely normal (n = 26) or only faintly nodular (n = 4) or enlarged (n = 5) on physical examination before initial biopsy in 35 (97%) of 36 cases. Mean total PSA level was 6.41 ng/mL (median, 5.66 ng/mL; range, 3.1–14.64 ng/mL). Biopsies were performed for various reasons, including increasing total PSA level, low free-to-total PSA ratios, familial history of prostate carcinoma with borderline PSA elevations, and miscellaneous other clinical reasons. Initial biopsy specimens were sextant in 33 (92%) of 36 cases, and in the remaining 3 cases, 8, 15, and 16 cores were obtained. However, 29 (88%) of 33 sextant biopsy specimens were only labeled “left” and “right” with 3 cores each in 2 containers.

Yale-New Haven Hospital is a 944-bed academic medical center located in the northeastern United States. Most prostate biopsy results are communicated to the clinician within 24 to 48 hours after biopsy specimens are submitted. Fourteen clinicians received pathology reports for these 36 patients during this period. A note explaining the biopsy findings, in addition to the principal diagnostic line, was included in 26 (72%) of 36 reports (Table 2). A second biopsy was explicitly recommended only in 11 (31%) of 36 cases. The suffix “suspicious for carcinoma” (or similar variations) was present in the main diagnostic line in half the cases and was embedded in the accompanying note in the other half.

The rate of biopsy subsequent to an ASAP diagnosis was 67% (24/36), and the mean duration to subsequent biopsy was 246 days (median, 182 days; range, 71–728 days). Of those 24 cases who underwent follow-up biopsies, carcinoma was diagnosed in 9 (38%). The Gleason score distribution was as follows: 3 + 3 = 6/10 (n = 7), 3 + 4 = 7/10 (n = 1), and 4 + 3 = 7/10 (n = 1). A subsequent diagnosis of ASAP was rendered in 3 (12%) of 24 cases, and the remaining 12 (50%) were benign. From the latter group, one patient underwent a second biopsy 73 days after an ASAP diagnosis. The second biopsy showed only benign prostatic tissue. A subsequent biopsy 6 years later showed adenocarcinoma (Gleason score, 4 + 5 = 9/10). All 24 subsequent biopsy specimens were sextant, and an ASAP did not significantly increase the number of sites sampled (7.18 average cores on initial biopsy specimen vs 6 average cores on second biopsy specimens, P = .64). Twenty of the 24 second biopsies were only labeled bilaterally (ie, 6 total cores: 3 cores on each side in 2 separate containers labeled left and right). Thus, it was impossible to determine whether there was any specific site correlation between the original ASAP diagnoses and the subsequent carcinoma or ASAP in the 12 cases where they were identified. However, the laterality (left or right) matched in 10 (83%) of 12 cases. No significant differences (P > .05) were identified in mean age, duration to subsequent biopsy, or average total serum PSA level when cases with benign and malignant subsequent diagnoses were compared. Average age and total PSA level also were not significantly different among the group of patients with no further biopsy specimens compared with those in each of the aforementioned groups. Finally, a comparison of cases with and without certain pathology report characteristics (inclusion of a descriptive note, explicit recommendation of a subsequent biopsy, and whether or not the suspicious for carcinoma suffix or similar phrases were in the main diagnostic line or embedded within the accompanying note) showed no statistically significant differences between these 2 groups (Table 2).

Atypical small acinar proliferation of uncertain significance was a term originally used by Bostwick et al10 in 1993 to describe atypical glandular prostate proliferations, which 7 independent observers from 3 different countries considered neither atypical adenomatous hyperplasia nor carcinoma. Four years later, the clinical significance of this diagnosis (modified to ASAP suspicious for, but not diagnostic of, adenocarcinoma) was outlined in a study from Mayo Clinic.5 ASAP is not characterized by distinctive morphologic criteria; indeed, its establishment as a diagnostic phrase is a consequence of the equivocation that needs to be expressed when a given acinar proliferation fails to attain the diagnostic threshold for carcinoma but is nonetheless sufficiently worrisome to warrant a comment. Although no studies, to our knowledge, have been performed that specifically addressed this issue, it is likely that the interobserver variability in cases diagnosed as ASAP, as with borderline lesions in other organs,11–13 would be high, and whether or not a given prostatic acinus fulfills the morphologic criteria for malignancy6–9 may be subject to some debate. It has been a rather common occurrence in the discussion over ASAP14–24 for a published photomicrograph purportedly depicting an ASAP or “atypical” focus to be diagnosed as carcinoma in the opinion of another observer.20,23,24 A preliminary inference may be drawn from one study in which pathologists at Mayo Clinic disagreed with 22% of the ASAP diagnoses rendered at a large urologic reference laboratory in Oklahoma.25 Thus, in this study we sought not to reevaluate the morphologic reasons why a case may or may not be considered suggestive of or diagnostic of carcinoma. Rather, we investigated the actions of clinicians once the diagnosis of ASAP is rendered to evaluate the relative importance that clinicians affiliated with our institution place on this diagnosis.

Our study shows that the diagnosis of ASAP on a prostate biopsy specimen generates a follow-up biopsy in 67% of cases following an average duration of 246 days. ASAP predicted carcinoma in subsequent biopsy specimens in 38% of the cases and was typically unassociated with an abnormal DRE result. Neither initial PSA level nor patient age significantly stratified patients with respect to probability of malignancy on a subsequent biopsy specimen. Pathology report characteristics, such as the inclusion of a descriptive note at the end of the main diagnosis line, did not significantly increase the likelihood of a second biopsy following an ASAP diagnosis.

The subsequent biopsy rate of 67% noted in this study is higher than the 35% to 57% rate noted in 3 previous studies5,26,27 and is similar to the 63.9% rate in another study (28%). There was an 8-month average duration between the ASAP diagnoses and follow-up biopsies in this study, which is not inconsistent with the published experience of others. In a study of atypical prostate biopsy specimens,28 only 39% of patients receiving this diagnosis underwent an additional biopsy within 3 months and duration to subsequent biopsy ranged from 2 weeks to 3 years. In another study by Iczkowski et al,5 median interval from ASAP diagnoses to subsequent biopsy was 3 months in cases subsequently diagnosed as having carcinoma and 9 months in cases with benign diagnoses at subsequent biopsy, suggesting that clinical parameters played a significant role in determining the interval to second biopsy. In the current group of patients, to further characterize the reasons for the 8-month average interval, we investigated the 14 of 24 cases in which the duration to subsequent biopsy was greater than 90 days, and no specific patterns emerged: in 2 cases, the delay was related to the logistics of patients changing health care providers, 4 patients missed multiple appointments, 3 patients developed other serious medical conditions in the interim, and the reasons for the delay were unclear in the remaining 5. For the 12 patients on whom no further follow-up biopsies were performed, the reasons were largely similar: lost to follow-up (n = 3), refused further biopsies (n = 1), death from other medical conditions (n = 1), changed health care providers more than 1 year after biopsy (n = 2), and reasons for no further biopsy unclear (n = 5). Drawing definitive conclusions from these findings is hampered by the relatively small sample, and the presence or absence of a sense of urgency on the part of the urologist following this diagnosis is not readily quantifiable; however, the data suggest that the practice patterns are not attributable to a lack of understanding by urologists of the significance of an ASAP diagnosis. Additionally, we are unaware of any studies that have specifically stratified patients with an ASAP diagnosis by interval to subsequent biopsy and demonstrated clinical outcome differences.

Significant criticisms have been leveled against the use of the term ASAP, related primarily to its given designation and concerns that a specific designation might cause nonspecific morphologic alterations to be misconstrued as a specific entity. Murphy20 objected to the “tendency to legitimize uncertainty by expressing it in terms that read like diagnoses,” whereas Epstein16 notes that, based on the subsequent biopsy rate of 60%, the ASAP designation “does not fully convey the seriousness of the condition to the urologist.” It was suggested that urologists be educated to make them aware of the significance of this diagnosis,16,21 and subsequently, there have been numerous publications of ASAP-related studies in the urology literature.27,29,30 Interestingly, the 63.9% subsequent biopsy rate in cases signed out as atypical with a comment26 is nearly identical to the subsequent biopsy rate of 67% associated with an ASAP diagnosis, as noted in this study. However, as previously noted, others5,27 have reported lesser subsequent biopsy rates.

When the expression of uncertainty is necessary, it is unclear whether standardization of diagnostic terms, a generally laudable goal in other areas of surgical pathology, is appropriate. Examples of successful application of standardized terms in these instances include indefinite for dysplasia in the gastrointestinal system and atypical squamous cells of undetermined significance in cytology. The main advantage of such uniformity is that it allows management guidelines to be formulated; the main disadvantage is that it tends to homogenize biologically discordant entities. Given the myriad ways in which an equivocal diagnosis may be rendered by different pathologists, what is not known is what effect on the subsequent biopsy rate a specific alternative to ASAP will have (ie, constant and specific way of expressing equivocal diagnoses at a given institution). The aforementioned 63.9% subsequent biopsy rate on atypical biopsy specimens was generated from an institution in which specific macros (canned texts) are used.16 Other studies evaluating atypical prostate biopsy specimens have not indicated specifically how this term was used in the pathology reports. It is thus impossible to ascertain whether the standardization of terms expressing uncertainty is actually detrimental by decreasing the subsequent biopsy rate. In our own database, there were insufficient numbers of alternative equivocal designations for a valid comparison.

Irrespective of how equivocal diagnoses are stated, the most important factor is that a clear understanding exist between the pathologist and clinician on the impact of the diagnosis on patient care. Thus, if all parties understand the significance of an ASAP diagnosis and recognize that it does not represent a specific pathologic entity, its use seems reasonable. Alternatively, if the pathology practice settings are such that pathologists and urologists are not in frequent communication, other ways of expressing the uncertainty may be warranted.

In conclusion, our study is concordant with previous findings showing that ASAP is diagnosed in less than 5% of prostate biopsy specimens in routine practice31 and is predictive of carcinoma in subsequent biopsy specimens in 34% to 60% of cases.7 Although closer follow-up may be recommended based on the high rate of association with carcinoma on subsequent biopsy specimens, we found no evidence that any delays in or lack of a subsequent biopsy is attributable to a lack of understanding on the part of urologists of the significance of the diagnosis.