Since its original description, high-grade prostatic intraepithelial neoplasia (HGPIN) as a finding on prostate biopsy has been a source of ongoing confusion. Original reports identified very high rates of cancer detection when patients with HGPIN underwent repeat biopsy, so it was regarded as a marker of missed prostate cancer that necessitated immediate and potentially frequent repeat biopsy.
Subsequent studies suggested that it actually might be a precursor lesion that can lead to prostate cancer, so surveillance alone with interval biopsy might be more appropriate. More recently, some studies have suggested that HGPIN might actually be relatively unimportant and repeat biopsy might be of no value at all.
Multiple biopsies magnify risk
Is HGPIN a marker of unidentified prostate cancer, a precursor to the malignancy, or an insignificant finding? The answer is all three, with a few critical caveats, of course. The explanation for this seeming contradiction lies in some critical details that are easy to overlook in this discussion.
For example, in the mid-1990s, several experts recommended that patients with HGPIN should undergo repeat biopsy every three months for some duration. This was based on reports showing that up to half of these patients would be found to have prostate cancer with this progressive biopsy strategy. The caveat is that these patients had undergone initial sextant biopsy, which in retrospect has been shown to identify only about half of all tumors. Multiple repeat biopsies in these patients clearly would have found cancer in about half of the cases based not on coexistent HGPIN, but on the aggressive biopsy strategy.
As Ian Thompson, MD, of the University of Texas Health Science Center at San Antonio, has most eloquently expressed, the greatest risk factor for being diagnosed with prostate cancer is to have a biopsy. Multiple biopsies clearly magnify that risk regardless of whether HGPIN is present. As importantly, none of my patients were enthusiastic by a recommendation for serial biopsy in this era before periprostatic block took the torture factor out of the experience, so this strategy was doomed from the start.
Further studies by Taneja and Lepor at New York University (NYU) demonstrated this principle of under sampling by sextant biopsy, finding only one case of prostate cancer in men who underwent relatively immediate repeat biopsy if HGPIN was identified on extended biopsy. Thus, a reasonable position became not to repeat biopsy solely for the indication of HGPIN (Urology. 2001;58:999-1003).
HGPIN—a true precursor
The NYU group also generated the most significant data suggesting that HGPIN appears to be a true precursor to the development of prostate cancer. They performed scheduled “delayed interval biopsies” approximately every three years and found prostate cancer in 22.3% and 23.4% of patients at three- and six-year intervals, respectively (Taneja et al. Urology, in press).
The authors acknowledged that their patients with truly negative initial biopsy might have also been found to have the same risk of cancer on interval biopsy had such been performed. To answer this, we recently compared the risk of cancer on repeat prostate biopsy in men with or without HGPIN. We found that 42.2% more patients with HGPIN had prostate cancer on delayed interval biopsy compared with those who had an initial completely benign diagnosis (35.67% vs. 25.07%). This translated into a nearly twofold increased risk of prostate cancer in the HGPIN group.
Thus, the risk of future prostate cancer predicted by HGPIN as a precursor lesion appears to persist at least through a six-year time frame, and theoretically beyond that point. For this reason, prevention strategies become appealing, potentially using toremifene, dietary manipulations, or 5-alpha reductase inhibitors, although none of these has achieved broad acceptance to date.
Should we ignore HGPIN?
However, it is well recognized that a number of physicians have questioned whether HGPIN is even a risk factor for future prostate cancer. I have observed that many urologists are currently ignoring the presence of HGPIN on biopsy completely. We recently reviewed the concept and found that, indeed, a majority of patients with HGPIN had not undergone repeat biopsy within three years. I believe this nonchalant attitude towards HGPIN is based on inconsistent findings reported in the literature and on personal experience.
To clarify the issue, we performed sub-analyses based on the extent of HGPIN in the original biopsy. Critically, the impact of HGPIN on cancer detection was almost exclusively in patients with multifocal disease. Our Kaplan-Meier curve estimated cancer rates at one, three, and five years of 3.6%, 12.5%, and 22.4%, respectively, for patients with a benign diagnosis on initial biopsy, 4.4%, 14.7% and 26.1% for patients with unifocal HGPIN, and 9.1%, 29.0% and 47.8% for patients with multifocal HGPIN. Thus, our data confirm that immediate repeat biopsy solely for the indication of HGPIN is rarely if ever necessary if unifocal, although multifocal HGPIN still suggests at least a 10% chance that cancer was missed.
More importantly, our data provide important information on the natural history of HGPIN. The risk of de novo cancer development is minimally increased in men with unifocal HGPIN compared to those with initial benign biopsy (by approximately 2% at three years and 4% at five years), but multifocal HGPIN more than doubles the risk of de novo cancer development at both three years (12.5% vs. 29.0%) and five years (22.4% vs. 47.8%).
By reporting HGPIN not as a homogeneous population, but segregated by unifocal versus multifocal disease, it appears that unifocal disease may truly have limited significance. However, multifocal HGPIN is a significant precursor to the development of prostate cancer.
A conservative approach
Consequently, it appears that ongoing controversy over whether HGPIN is associated with cancer risk relates to two concepts. First, instead of indicating a “missed” cancer based on under sampling, multifocal HGPIN more commonly is a precursor lesion that leads to a significant risk of developing prostate cancer. This risk seems to persist or even increase over time.
Second, although unifocal HGPIN may be relatively meaningless, multifocal or bilateral HGPIN is a real entity, as are its implications for future prostate cancer risk. Therefore, we suggest a conservative approach to patients with unifocal HGPIN, reserving repeat biopsy for those that have independent indication such as rising PSA or low percent-free PSA. In contrast, we recommend repeat “delayed interval” biopsy as described by Taneja et al to be performed approximately every three years with multifocal HGPIN for as long as the patient’s overall health would warrant treatment for prostate cancer.
Finally, and back to the original misconception regarding HGPIN, if inadequate biopsy—defined as fewer than 12 laterally and apically based cores—was performed, a repeat biopsy is probably indicated regardless of whether HGPIN was identified.
In summary, if inadequate biopsy was performed, HGPIN has a high chance of signifying coexistent unidentified prostate cancer. If adequate biopsy was performed and multifocal HGPIN was identified, it is a precursor lesion associated with cancer development in approximately half of patients within the next five to six years. If adequate biopsy was performed and unifocal HGPIN was identified, it may be insignificant and the decision to perform a repeat biopsy should be based on factors other than HGPIN, such as percent-free PSA, persistently elevated or rising PSA, or nomograms predicting a high risk of cancer.
J. Stephen Jones, MD, FACS, is Chairman of the Cleveland Clinic Department of Regional Urology and Professor of Surgery (Urology) at the Cleveland Clinic Lerner College of Medicine at Case Western Reserve University in Cleveland. Dr. Jones also is a member of the Renal & Urology News Editorial Advisory Board.