Indian Journal of Cancer Medknow Publications on behalf of Indian Cancer Society ISSN: 0019-509X EISSN: 1998-4774 Vol. 48, Num. 2, 2011, pp. 175-180

Indian Journal of Cancer, Vol. 48, No. 2, April-June, 2011, pp. 175-180

Original Article

Prostate biopsy findings in Indian men: A hospital-based study

S Sinha¹, SR Siriguri¹, SK Kanakmedala¹, K Bikkasani²

¹ Department of Urology, Medwin Hospital, Hyderabad, India
² Department of Pathology, Medwin Hospital, Hyderabad, India

Correspondence Address: S Sinha Department of Urology, Medwin Hospital, Hyderabad India drsanjaysinha@hotmail.com

DOI: 10.4103/0019-509X.82879

Abstract

Keywords: Adenocarcinoma prostate, digital rectal examination, histopathology, Indian men, prostate-specific antigen, transrectal ultrasonography

Introduction

Traditionally prostatic biopsy has been recommended for men with a prostate-specific antigen (PSA) level above 4 ng/ml after an assessment of co-morbidity, life expectancy, and a detailed discussion regarding the implications of a biopsy and its result. ^[1] The prostate cancer prevention trial data show that no level of PSA is "normal", and there is a continuum of risk for prostate cancer based on the level of PSA. ^[2] This data have also suggested that the threshold for prostatic biopsy may need to be revised downward since about 25% of American men with a PSA of less than 4 ng/ml have clinically significant malignancy. ^[2] Although centers in India are increasingly adopting these practices, there is little data to support this approach in our patients. Despite India constituting a large proportion of the world′s population, there are a few indexed publications from India regarding the findings of prostatic biopsy in men with elevated PSA. There has been a suspicion that prostate cancer is less common in Asian populations as compared to non-Asians. ^[3] This paper analyses the findings on prostatic biopsy in Indian men with elevated PSA.

Materials and Methods

A total of 119 patients who underwent TRUS-guided prostatic biopsy over the last 3 years were evaluated retrospectively using the archives of the Pathology Department. The archives store the initial biopsy requisition form along with the final histopathology report. Data regarding age, indwelling catheter at presentation, PSA (Vidas Kit®, Biomerieux, immunofluorescence assay), digital rectal examination (DRE), number of cores, and final histology were retrieved from the archived requisition forms. As a departmental policy, PSA testing (and hence biopsy) was not offered to any patient presenting with acute retention or within 1 week of catheterization. The indication for PSA was varied and included patients with lower urinary tract symptoms, suspicion of malignancy, and screening evaluation. Thus, this study includes all PSA estimations that were performed en route to a prostatic biopsy. PSA estimations were not drawn in patients with either a clinical suspicion or laboratory documentation of urinary infection or prostatitis. The indication for biopsy was elevated PSA, abnormal DRE or both. No biopsies were done in patients with PSA less than 4 ng/ml. Biopsies were done by an experienced urologist under transrectal ultrasound guidance. Specimen was sent for histopathology in separate containers. Minimum 10 cores biopsies were performed in 109/119 (92%) and 12 cores in 92/119 (77%). The 12-core biopsy was the target for every patient. Those with small glands underwent lesser number of cores. Also, some patients with clinically obvious extensive malignancy did not undergo a 12-core biopsy. This decision-making process is not captured in the study. Patients were stratified into three groups based on their PSA: 4-10 ng/ml, 10-20 ng/ml, and >20 ng/ml. Unpaired t-test, Chi-square test, and logistic regression were calculated using Excel statistical analysis (Ver 2007) and online calculators from http://home.ubalt.edu/ntsbarsh/Business-stat/otherapplets/Catego.htm and http://statpages.org/logistic.html (P < 0.05 significant).

Not all patients with PSA of above 4 ng/ml underwent a biopsy. Some patients were not offered a biopsy by the treating team. For instance patients with a short life expectancy were not ordinarily offered a biopsy by the treating urologist if they came with a PSA report of less than 10 ng/ml. Also, not all men who were offered a biopsy elected to undergo the procedure. The detailed data of the number of patients who were eligible for a biopsy but did not undergo one are not available.

Histopathology was evaluated by a pathologist with additional certification training in urinary tract pathology. Inflammation was considered as the histological diagnosis only in men without any cancer or precursor lesions in any of the cores. In such biopsies, inflammation was diagnosed when either lymphoid nodules or large areas of inflammatory infiltrate was noted. Occasional scattered inflammatory cells in the stroma were not categorized as inflammation. The type of inflammatory cell (polymorphs, lymphocyte, monocyte, or plasma cell) was not considered while making the diagnosis. ^[4]

Results

Mean age was 67.6 years (range, 48-90 years). Men with PSA above 20 ng/ml were significantly older than those with PSA less than 4 ng/ml [Table - 1]. Cancer was found in 29/119 biopsies (1/4th). However, cancer was found in only about 7% (5/73) of patients with PSA of less than 20 ng/ml as compared to 52.2% of men with PSA > 20 ng/ml. Patients with PSA > 20 ng/ml were more likely to show cancer (χ² = 31.4546, P = 0) as compared to the other two groups. They were also more likely to be on an indwelling urethral catheter at the time of evaluation [Table - 1]. Precursor lesions were noted in 10/119 (8.4%) including 8 patients with high grade prostatic intraepithelial neoplasia (HGPIN) and 2 with atypia [Table - 2]. Retention was noted in 6/50 (12%), 3/30 (10%), 1/10 (10%), and 6/29 (21%) patients with BPH, inflammation, precursor lesions, and cancer, respectively. There was no statistical difference in the distribution of retention in the different histological groups (χ² = 1.806, P = 0.61). The distribution of high grade (Gleason 8-10) disease was 0/2, 0/3, and 6/24 in the groups with PSA 4-10 ng/ml, 10-20 ng/ml, and >20 ng/ml, respectively, a difference which was not statistically significant (χ² = 1.576, P = 0.455). There was no difference in the overall distribution of Gleason grade based on the PSA level [Table - 3]. The median Gleason grade of 7 was similar in all groups. Additional digital-guided biopsies were performed in six patients of whom four had cancer. All these patients had additional TRUS-guided cores which also showed cancer. Fourteen patients had additional cores taken from hypoechoic areas. Of these five showed cancer. All these patients had at least one other core from an isoechoic area which showed cancer.

On a univariate logistic regression analysis, age, PSA, and DRE all showed significant association with malignancy on the final histology. The odds ratio (range), coefficient, and P-value were 1.05 (1.01-1.10), 0.0500, 0.0238 for age, 1.04 (1.02-1.06), 0.0434, 0.000 for PSA, and 17.94 (6.04-53.27), 2.8869, 0.000 for DRE.

On a multiple logistic regression analysis [Table - 4]a, DRE (OR = 8.07) and PSA (OR = 1.03) were predictive of cancer although the odds ratio was very close to 1 for PSA. When patients with PSA of less than 20 ng/ml were analysed separately, DRE remained the only significant predictor of malignancy with P value of 0.012 [Table - 4]b. PSA was not a significant independent predictor of malignancy in this group on a multiple logistic regression analysis when corrected for the rectal examination findings and age.

Inflammatory pathology was common at all PSA levels and accounted for 30/119 (1/4th) of all biopsy reports. In men with negative DRE and PSA > 10 ng/ml [Table - 5], inflammatory pathology was significantly more likely (χ² = 4.2798, P = 0.039).

Discussion

This study includes all patients who underwent a prostatic biopsy irrespective of the decision-making matrix for the individual patient. Hence, this study is a snapshot analysis of Indian patients undergoing prostatic biopsy in clinical practice at a tertiary care hospital. Our analysis of prostatic biopsy shows interesting differences from reports on non-Asian populations. Cancer is less common and inflammatory lesions are more common in the biopsy at all levels of PSA in our patients. On its website, the American Cancer Society (www.cancer.org; how is prostate cancer found?) estimates the risk of prostate cancer at 1 in 4 for men with PSA between 4 and 10 ng/ml and 50% for men with PSA above 10 ng/ml. These values are far higher than what we observed. However, our observations are similar to what was reported by Dublin from an Asian population. ^[3] Dublin found cancer in 10% of men with PSA of 4.1-20 ng/ml. A total of 60.5% of his patients had histological prostatitis and there was no difference in the biopsy outcome between men of Malay, Chinese, or Indian origin. ^[3] In a study from Mumbai, Chavan et al. found cancer rates of 0.6%, 2.3%, 2.5%, 34.1%, and 54.9% in Indian men with PSA values among 0-4 ng/ml, 4-10 ng/ml, 10-20 ng/ml, 20-50 ng/ml, and >50 ng/ml, respectively. These values are remarkably similar to our findings. ^[5] The number of men with cancer was much higher in patients with a PSA of >20 ng/ml (52% versus 7%) as compared to those with a PSA less than 20 ng/ml. This is probably owing to fact that men with a higher PSA were more likely to suffer from cancer. Also, the >20 ng/ml group included men in whom there was an unmistakable clinical suspicion of malignancy and those with very high levels of PSA.

In contrast to our findings, Gupta et al. reported from a North Indian tertiary center on men with PSA 4-10 ng/ml who underwent initial sextant biopsy followed by a more extensive 13-core biopsy in those with continued PSA elevation. ^[6] They found 24% positivity for cancer (34/142) on the initial biopsy as contrasted to the 7% that we found. Patients with negative biopsies were given a course of antibiotics and re-tested. A total of 55% of these men showed a drop in PSA and did not need a repeat biopsy. On follow-up biopsy, an additional 5 patients out of 48 showed cancer giving a total cancer positivity rate of 27% (39/142) for their patients. There are two important messages from these data in the context of our findings. We assume India to be a homogeneous population, but this may not be true. Both our data as well as that from Mumbai ^[5] show lower cancer prevalence in biopsies in contrast to the report from New Delhi. In a country as vast and diverse as India, we must be cautious in extrapolating data obtained from one part of the country to other parts. The second important finding in their paper was that use of an antibiotic led to PSA fall in 55% of their patients and all these patients avoided the second biopsy. It is our contention that by using an initial course of antibiotics (with the caveat of a normal DRE), some of these patients might have avoided the first biopsy too.

The reason for lower rates of cancer on the biopsy could be because the prevalence of prostate cancer is actually lower in our patients. It is important to note that this study is not population-based, and hence one cannot draw direct conclusions regarding the prevalence of cancer in Indian society. However, there are other reasons why this might have happened. Could it be because of our biopsy technique? Unlikely, since we used the standard methodical sampling of the prostate under TRUS guidance. Was the number of cores taken adequate? Eight percent of our patients had less than 10 cores and 23% less than 12 cores. These numbers do not seem to be enough to strongly influence the results. In addition, the European Urology Association Guidelines on Prostate Cancer suggest that 8-12 biopsies are adequate depending on the prostate size. ^[7] The last (and disturbing thought) is that our pathologists might have been under-diagnosing cancer. While this is possible, the chief pathologist has an additional certification training in urinary tract pathology and ~15% of our biopsy reports have been reviewed by a second experienced pathologist without major differences of opinion. Not all patients who were recommended a biopsy, chose to undergo one and not all patients with PSA above 4 ng/ml were offered a biopsy by the treating urologist. For instance, it is conceivable that a 73-year-diabetic man suffering from coronary artery disease with a PSA 4.5 ng/ml might not have been offered a biopsy. These data have not been captured in the study and might have influenced the outcome especially if a larger number of patients with cancer happened not to undergo biopsy. However, there is no reason to suspect that this logic would specifically exclude only those with cancer. Re-biopsy was uncommon in our patients. Most patients with elevated PSA and a non-cancer finding were given a course of antibiotics (prolonged, if there was histological inflammation). These patients were followed up by serial PSAs and re-biopsy suggested only when there was either persistent elevation or a rising trend in PSA. These data are not captured in the study. The lack of data concerning AUA scores of these patients and correlation with the weight of the gland are other important limitations of this study. These issues must be kept in mind while interpreting the data from our study.

While PSA was a predictor of cancer in our patients, this association was tenuous on a multiple logistic regression analysis when age and DRE findings were taken into account. The rectal examination was a much better predictor of the biopsy report. This can be explained by the fact that men with all levels of PSA were included. In men with markedly elevated PSAs, the DRE would be expected to perform well and it did. What was more interesting was that these findings remained essentially unchanged when men with PSA less than 20 ng/ml were analyzed separately. DRE remained a significant predictor while PSA lost even the semblance of tenuous association that was noted with the data as a whole. Does this imply that there is no point obtaining a PSA for our patients? We do not think so. These data are skewed in terms of the distribution of men in the various PSA subgroups and the number of men with PSA below 20 ng/ml with cancer was too small to draw any meaningful conclusion.

Other authors have recommended a higher cutoff of 20 ng/ml in Indian patients based on their findings. ^[5] While our data too show that the pick-up rate for cancer was low in men with PSA of less than 20 ng/ml, we believe that there is not enough evidence to raise our cutoff for biopsy. Such a recommendation would need data showing that raising the cutoff was not significantly impacting the stage at diagnosis and (ultimately) the disease-specific mortality outcome, data that are not available currently for our patients. Also, it would need far more robust (preferably) prospective data that this (and other available publications) from India cannot currently provide. If the stage-at-diagnosis data were to show that our patients are, on average, of higher stage at diagnosis, it would suggest that we are not identifying the early cancer patients in the biopsies done at lower PSA levels. The impact of these data on policies regarding screening in the Indian population is difficult to judge since the study is not population-based, is cross-sectional rather than longitudinal, and does not include stage data.

The higher incidence of inflammatory pathology in men with elevated PSA without any clinical suspicion of prostatitis is interesting. Prostate inflammation has been suggested as a factor in the development and progression of prostate cancer. ^[8] In an experimental animal model, induced chronic prostatic inflammation has been shown to lead to focal atypia. ^[9] Inflammation may induce cancer by cell and genome damage, increasing cellular turnover, and creating an environment that enables cell replication and angiogenesis. ^[10] Prostate inflammation has also been shown to have a link to the development of benign prostatic hyperplasia. Interleukin 8 may represent a link between chronic inflammation and autocrine or paracrine stomal cell proliferation. ^[11] Like cancer, inflammation is known to be a cause for elevation of PSA levels. However, our study showed a much higher incidence of unsuspected prostatic inflammation, but lower incidence of cancer on the biopsy. An association between inflammation and cancer should have led to increased prevalence of cancer on biopsies done in populations where prostatic inflammation is more prevalent, which we did not find.

There is no literature documenting a higher prevalence of clinical prostatitis in Indian men. In view of such a high prevalence of prostatitis in the biopsy specimen, it would seem logical to use a course of antibiotics in men with elevated PSA who have normal DRE findings and repeat the test before taking a decision for biopsy. The use of antibiotics has been shown to reduce PSA values in Chinese men and has the potential to reduce unnecessary biopsies in some men with elevated PSA. ^[12] Another approach has been to examine prostatic secretions following a prostatic massage and treating those with increased leukocyte count. ^[13] Thirty percent of Japanese men with more than 10 wbc in the prostatic massage secretion showed reduction in PSA after a course antibiotics. ^[13] The empirical use of antibiotics in men with an elevated PSA is not without risks. Besides the delay in diagnosis in those men who are actually suffering from cancer, there is the possibility of inducing drug resistance to the commonly used antibiotics in the urological armamentarium. Drug resistance is a constant menace in the treatment of urinary infections and prostatitis and there is recent Asian data suggesting that E. coli is developing resistance to ciprofloxacin, the commonly recommended agent against prostatitis. ^[14],[15] Our findings suggest the need for a study to analyze the effect of antibiotics on PSA values, biopsy findings, and microbial flora. The findings of this study also emphasize the continued relevance of DRE in our practice.

References

Copyright 2011 - Indian Journal of Cancer

The following images related to this document are available:

Photo images