Age, prostate volume, prostate-specific antigen and prostate-specific antigen density as predictor factors in results of transrectal ultrasonography-guided prostate biopsy

Introduction

Prostate cancer (PCa) is an important health issue worldwide because of its increasing incidence; it was reported that there are 1.1 million PCa patients worldwide in 2012¹. The incidence of PCa among Indonesian men was estimated to be 14.8 per 100,000 population and the mortality rate was reported as high as 9.8 per 100,000 population in 2012¹.

Transrectal ultrasonography (TRUS) guided biopsy plays an important role for PCa detection and has become a standard diagnostic method for PCa detection. It is recommended for men with elevated prostate specific antigen (PSA) levels or abnormal digital rectal examination (DRE). In Indonesia, 66.5% of PCa was diagnosed with TRUS-guided biopsy during 1994–2005 and 66.7% of them were found in advanced stage during this period^2,3. An appropriate and early detection are needed to detect PCa in its early stages for better treatment results, survival rates, and quality of life⁴.

The detection rate of prostate biopsy in several Asian countries varied between 14.6% and 26.5% in patients with PSA levels ranging from 2–20 ng/ml⁵. This widely used procedure for PCa remains controversial as it may lead to over-diagnosis (23%–42%) and over-treatment⁵. Both are related to unnecessary risk of urinary, sexual, and bowel dysfunction, which can affect patients’ quality of life^4,6. Although TRUS-guided prostate biopsy is a safe procedure, there are still complications causing significant morbidity. The most common complications are mild hematuria (62%), rectal bleeding (2%), urinary tract infections (1.7%–11.3%), acute urinary retention (0.3%–2.6%), hematospermia (45.3%–50.4%), vasovagal response (2.8%), and hospitalization due to infection (1%–4.1%)^7,8). Systematic screening for prostate cancer by determining other factors besides PSA may assist prevention of this circumstance⁵.

The aim of this study was to identify the predictor factors such as PSA, age, prostate volume (PV), and PSA density (PSAD) as indications to perform TRUS guided prostate biopsy. Those variables are thought to be important predictors to reduce unnecessary biopsies and improve detection rate.

Methods

Results

ROC analysis

The ROC curve demonstrated that PSAD was the best predictor factor among those variables (age, PSA, and PV) in TRUS-guided prostate biopsy (Figure 1 and Table 3). The area under the curve (AUC) for PSAD (84.5%) and PSA (78.7%), while the AUC for PV and age were 62.4% and 61.2%, respectively. PSAD and PSA were thus considered better predictors for identifying PCa than PV and age. This study also conducted ROC curve analysis for PSAD in stratified PSA level (Table 3).

Figure 1.

Receiver operating characteristic curve for (A) prostate-specific antigen (PSA), PSA density (PSAD), and age, and (B) prostate volume.

Table 3. Sensitivity, specificity, and cut-off point of each variable.

Variables	AUC	Cut-off point	Sensitivity (%)	Specificity (%)	P value	95% CI (%)
						Lower	Upper
PSA	0.787	32.5 ng/ml	62.0	87.1	<0.001	75.7	81.7
PV	0.674	52.3 ml	55.2	75.5	<0.001	64.3	70.6
PSAD	0.845	0.43 ng/ml/ml	81.4	82.0	<0.001	82.0	87.0
Age	0.607	63.5 years	73.1	57.1	<0.001	57.4	64.0
PSAD of PSA group
PSADc of PSA <4 ng/ml	0.311	0.07 ng/ml/ml	33.0	67.0	0.128	5.0	57.2
PSADc of PSA 4-10 ng/ml	0.737	0.17 ng/ml/ml	75.0	66.5	<0.001	65.8	81.5
PSADc of PSA 10-20 ng/ml	0.741	0.33 ng/ml/ml	62.3	78.1	<0.001	67.2	81.0
PSADc of PSA >20 ng/ml	0.881	1.00 ng/ml/ml	81.9	78.6	<0.001	85.2	91.0
PSA age specific
PSA of age <60 years old	0.753	45.50 ng/ml	55.6	94.4	<0.001	66.8	83.7
PSA of age 60–70 years old	0.768	32.63 ng/ml	61.9	87.0	<0.001	74.5	83.0
PSA of age >70 years old	0.802	43.50 ng/ml	64.6	86.3	<0.001	76.6	83.7

AUC, area under the curve; PSA, prostate-specific antigen; PV, prostate volume; PSAD, PSA density.

The optimal cut-off point of PSAD in the PSA group of 4–10 ng/ml was 0.17 ng/ml and its sensitivity and specificity were 75% and 66.5%, respectively. The cut-off point of PSA level was 32.5 ng/ml, giving a sensitivity and specificity for PSA of 62.0% and 87.1%, respectively. Prostate volume and age had lower optimum sensitivity and specificity compared to PSAD and PSA. (Table 3) This study divided PSAD variables into four groups based on PSA level (Table 3).

Discussion

Rashid et al. stated that TRUS-guided biopsy is recommended as the procedure of choice for early detection of PCa. Even though the elevation of PSA and the presence of an abnormal DRE are associated with PCa, those kind of examinations are not appropriate as a standard predictor⁹.

In our study, 33.5% of the 1,232 patients examined had prostate cancer based on the biopsy result of histopathology examination. The overall detection rate (33.5%) of PCa in this study was much lower than in Pakistan (50.3%)¹⁰. Na et al. and Presti et al. also showed differences compared to earlier Chinese and American studies; they reported that the detection rates were 47% and 42%, respectively^11,12.

The same results in positive biopsy rate are also found in other studies (Table 4). In Pakistan, Rashid et al. showed that 48.8% of patients performing TRUS guided biopsy were detected as PCa⁹. In India, Prakash et al. found that 35.2% of PCa were diagnosed from prostate biopsy¹³. Tang et al. in China revealed that 36.4% of patients were positive for PCa after performing prostate biopsy, but this study only investigated PSA levels ranging from 10–50 ng/ml¹⁴. Study from Germany had a lower detection rate than current study, at only 23.5%¹⁵. The low detection rate in this study could be caused by inadequate sampling. In larger prostate volumes, the possibility of error sampling is higher than in smaller prostate volume. Hence, the number of cores should be adjusted based on the prostate volume.

In this study, the median age of PCa patients is 68 years old (range from 35–89 years) and those who were >70 years had the highest positive biopsy rate. This result is supported by other studies in Indonesia and Pakistan^2,10,16. Umbas et al. revealed that patients with PCa in Indonesia were predominantly those >70 years during 1995–2004². In addition, Ariani et al. also reported that the overall age of the PCa patients in their study was 67.12 years¹⁶. The previous two studies mentioned that the rate in Indonesia was not significantly different compared to studies conducted in Nepal Senegal, Pakistan and Taiwan (63.6, 65.5, 65.7, and 67.7 years, respectively)^9,17–19. However, other studies from China (Na et al. and Wu et al.) showed a slightly higher overall age compared to the current study, with mean ages of 71.24 and 73.4 years, respectively^11,20. In addition, the mean age of a group in Korea was found to be lower (63.8 years old) than that in the current study, but its result cannot be compared because the study use tPSA level less than 10 ng/ml²¹.

Age is one of many risk factors affecting PCa because the incidence of the disease increases along with aging. Prostate cancer is mostly (in about 75% of cases) diagnosed after 65 years of age. There only 2% of all PCa cases were diagnosed in males <50 years¹⁹.

The highest positive biopsy rate in PSA group within this study belongs to patients with PSA values >20 ng/ml. This result suggests that severe elevation of PSA will increase the probability of PCa. Prakash et al. and Rashid et al. showed that 73.8–74.6% of patients with PSA >20 ng/ml that underwent TRUS-guided biopsy were diagnosed with PCa^10,13. Another study by Barakazi et al. revealed a similar result to the current study²².

This study shows that patients with PSA level 4–10 ng/ml and 10–20 ng/ml had lower detection rate compared to those with PSA level below 4 ng/ml. Prakash et al. also generated similar results to those of our study, showing that PCa in patients with PSA level <4 ng/ml was could be detected more easily compared to patients with PSA levels of 4–10 ng/ml and 10–20 ng/ml, with the addition of DRE examination¹³. These results suggested that DRE may still have an important role in assisting or predicting PCa in suspicious patients²³. Previous studies conducted in Taiwan and The Netherlands also revealed that DRE is an independent predictor of PCa^19,24.

In this study, the lowest detection rate of PCa is in those with a PSA level of 4–10 ng/ml. Therefore, PSA alone is not an effective screening method to diagnose PCa. PSA level alone could cause the high rate of false positive or negative biopsy rate in patients that underwent TRUS-guided prostate biopsy. PSA is a protein produced by the prostatic epithelium. It is organ-specific, but not cancer-specific. PSA can be manipulated by other disorders of the prostate, such as prostatitis, urinary retention, and BPH. Research into a large population regarding PSA screening in ERSPC revealed that PSA screening was able to reduce a number of 20% mortality in patients with PCa. Another study reported that PSA had no benefit in reducing mortality in PCa. PSA alone is not optimal to determine the risk of PCa¹¹. Other factors such as DRE, TRUS to asses PV, and PSAD must be considered when performing prostate biopsy²⁴.

The role of PV as a predictor factor is still being debated. In this study, the mean PV in patients with PCa is much lower than in those with negative biopsy results. A previous study in Indonesia conducted by Ariani et al. produced the same result as our study¹⁶. The study revealed that the incidence of PCa increased with decreasing PV¹⁶. Wu et al. reported that lower PV was able to indicate a higher risk of PCa detection²⁰. Tang et al. reported that decreasing PV was associated with higher risk of PCa in patient undergoing TRUS biopsy¹⁴. The high rate of negative biopsies resulting due to larger PV might be due to more sampling errors in TRUS biopsy. Therefore, assessing more cores from larger prostates could increase detection rate and reduce bias^5,14,25. Pietzak et al. revealed that patients with larger PV (50 ml) who had prior negative biopsy require multiple biopsies to increase the chance of gaining a positive biopsy²⁶.

This study revealed that PSAD had the best predictor in detecting PCa. A study from China showed a similar ROC cut-off point (0.2 ng/ml/ml) and AUC (0.664) in intermediate PSA level (4–10 ng/ml) to this study²⁷. The previous study in Indonesia also reported that the optimum cut-off point was 0.19 in patients with an intermediate PSA level²⁸. The cut-off point of PSAD is similar in our study and is higher compared to the reference values of Western countries (0.15 ng/ml/ml) for prostate biopsy²⁹. This examination could be used to reduce unnecessary biopsy because the specificity of PSAD was much higher than specificity of PSA level 4 ng/ml (as indicated for prostate biopsy) in this study.

The limitation of our study is that the TRUS-guided biopsy is of limited value. It can only show small number of malignancies. The use of contrast agent combined with TRUS-guided prostate biopsy can increase the accuracy of prostate cancer diagnosis²⁹. We also noticed that according to Prakash et al., the number of core biopsies have to be increased up to 16 cores in order to significantly increase the rate of prostate cancer detection¹³.

Another limitation of this study is that it is a retrospective study conducted in a single institution. In this study, PSA level was not defined adequately, therefore the cut-off point and mean variables were difficult to obtain. Although this study is conducted in a single institution, it can represent the races from Indonesian men because the samples are recruited from several provinces in Indonesia.

Conclusion

In conclusion, the overall detection rate of PCa by biopsy in our study is 33.5%, much lower than other countries in Asia. The incidence of PCa increased with higher PSA level, older age and lower PV. Utilization of PSAD 0.17 ng/ml/ml as a cut-off point in patients with PSA level between 4–10 ng/ml is recommended to improve PCa detection in Indonesian men.

Data availability