Abstract
Background: African American (AA) men have the highest incidence and mortality rates of prostate cancer (PCa) among all racial groups in the United States. While race is a social construct, for AA men, this overlaps with west African ancestry. Many of the PCa susceptibility variants exhibit distinct allele frequencies and risk estimates across different races and contribute substantially to the large disparities of PCa incidence among races. We previously reported that a SNP in 8q24, rs7824364, was strongly associated with west African ancestry and increased risks of PCa in both AA and Puerto Rican men. In this study, we determined whether this SNP can predict biopsy positivity and detection of clinically significant disease (GS≥7) in a cohort of AA men with suspected PCa. Methods: SNP rs7824364 was genotyped in 199 AA men with elevated total PSA (>2.5 ng/mL) or abnormal digital rectal exam (DRE) and the associations of different genotypes with biopsy positivity and clinically significant disease were analyzed. Results: The variant allele carriers were significantly over-represented in the biopsy-positive group compared to the biopsy-negative group (44% vs 25.7%, P=0.011). In the multivariate logistic regression analyses, variant allele carriers were at a more than a two-fold increased risk of a positive biopsy (OR=2.14, 95% CI=1.06–4.32). Moreover, the variant allele was a predictor (OR=2.26, 95% CI=1.06–4.84) of a positive biopsy in the subgroup of patients with PSA<10 ng/mL and normal DRE. The variant allele carriers were also more prevalent in cases with GS≥7 compared to cases with GS<7 and benign biopsy. Conclusions: This study demonstrated that the west African ancestry-specific SNP rs7824364 on 8q24 independently predicted a positive prostate biopsy in AA men who were candidates for prostate biopsy subsequent to prostate cancer screening.
Introduction
Prostate cancer is the most commonly diagnosed cancer and the second leading cause of cancer death among men in the United States (1). The well-established risk factors for PCa include advanced age, African ancestry, positive family history, and germline gene mutations associated with certain genetic syndromes (2). African American men have the highest incidence and mortality rates of PCa among all racial groups in the United States (3). While race is a social construct, for AA men, this overlaps with west African ancestry.
PCa is one of the most heritable types of cancer (4–10). The most recent and largest genome-wide association study (GWAS) to date identified a total of 451 PCa susceptibility loci across diverse ancestral populations and constructed a genetic risk score (GRS) comprising these 451 single nucleotide polymorphisms (SNPs) (10). The GRS451 is associated with PCa risk, ranging from 1.8 in men of African ancestry to 2.2 in men of European ancestry. Moreover, the GRS451 is associated with a higher risk of aggressive disease in African ancestry men (10). Many of these susceptibility SNPs exhibit distinct allele frequencies and associations across different races. Accumulating evidence has shown that the variations in genetic architecture and susceptibility loci contribute substantially to the large disparities of PCa incidence among races (8–12). Conti et al. (8) developed a GRS using the multi-ancestry weights of 269 GWAS-validated PCa risk variants and found that men of African ancestry exhibited a mean GRS 2.18 times higher than men of European ancestry. In contrast, men of east Asian ancestry had a GRS 0.73 times lower than men of European ancestry. These findings align with the observed population differences in PCa incidence rates and support that germline genetic variations contribute to the population differences in PCa risk.
Inherited genetic susceptibility SNPs hold significant potential in two key clinical applications (11, 12). Firstly, in the pre-PSA screening setting, they can aid in the development of improved risk stratification strategies, allowing for targeted PSA screening in high-risk individuals and improving the specificity of PSA screening. Secondly, in the post-PSA setting, these SNPs can complement clinical variables to improve the selection of suspected patients for prostate biopsy. This could increase the positive biopsy rate and avoid unnecessary biopsies. While considerable efforts have been directed towards the first application (11–15), limited research has focused on assessing the utility of these SNPs in the post-PSA setting (16–19). Notably, research regarding this second application has been particularly scarce in AA men.
The 8q24 region is a major susceptibility region for PCa and contributes to nearly 10% of the familial risk of PCa (20). Moreover, African ancestry is over-represented in 8q24 (21, 22). In a previous retrospective case control study, we assessed nine PCa susceptibility SNPs in 8q24 and found only one SNP, rs7824364, was strongly associated with a west African ancestry and PCa presence in both AA and Puerto Rican (PR) PCa patients versus controls (23). Of note, among PR men with the risk allele of this SNP, there was also an increased prevalence of west African ancestry. In this current study, we aim to determine whether rs7824364 can predict biopsy positivity and clinically significant disease in a prospective cohort of AA men suspected of having PCa.
Materials and Methods
Study population
The study population consisted of 199 AA men between the ages of 43 and 85 who self-identified as being “African-American/Black”subsequent to being identified as potential study particpants. They were prospectively recruited from three different hospitals in Houston from September of 2014- April 2017: The University of Texas MD Anderson Cancer Center (MDACC), Lyndon Baines Johnson General Hospital (LBJ) and the Michael E. DeBakey Veterans Administration Medical Center (MEDVAMC). The inclusion criteria were: 1) candidates for a prostate biopsy based upon an abnormal DRE or serum PSA level ≥ 2.5 ng/mL; and 2) had no more than one prior negative biopsy ≥ 12 months prior enrollment. Exclusion criteria included: 1) a previous diagnosis of PCa; or 2) more than one prior negative prostate biopsy. Each patient had a pre-biopsy total PSA (tPSA) level, free PSA (fPSA) level, [−2]proPSA (p2PSA) level and a calculated Prostate Health Index (PHI) value. Each patient completed a validated questionnaire administered by trained interviewers. Blood specimens were collected from study participants prior to prostate biopsy for DNA extraction and serum PSA and isoform measurements. This study was approved by the Institutional Review Boards of MDACC, LBJ, and MEDVAMC. Informed consent was obtained prior to the sample collection or questionnaire administration.
Prostate Biopsy and Pathologic assessment
Transrectal ultrasound guided prostate biopsies were performed as standard of care and consisted of a minimum of 10 cores obtained from both paramedian and lateral locations. MRI ultrasound targeted fusion biopsies were not routinely performed during the study period. To facilitate uniform grading of prostate biopsy specimens from the different hospitals, the study pathologists involved (P.T, M.C., and M.I.) convened for a consensus meeting at MDACC. The cases reviewed included the entire spectrum of Gleason grades with emphasis on challenging cases to ensure uniform use of ISUP grading criteria and achieve grading consensus (24).
Measurement of serum tPSA, fPSA, and p2PSA concentrations
Serum samples were analyzed for tPSA, fPSA, and p2PSA on an automated Beckman Coulter Access 2 Immunoassay analyzer using the FDA-approved respective Access Hybritech assays (Beckman Coulter, Brea, CA). PHI was then calculated using the formula, (p2PSA/fPSA)×√tPSA.
Genotyping
SNP rs7824364 was genotyped using a pre-designed TaqMan Genotyping Assay with the Applied Biosystems 7900HT Real-Time PCR system according to the manufacturer’s protocol. The genotypes were called automatically using the supplied software. We included two negative controls and 5% of the replicated samples. The concordance of replicated samples was 100%.
Statistical Analysis
Demographic characteristics were summarized using frequency (%) for categorical variables and median (range) for continuous variables. Univariate and multivariate logistic regression models were used to determine the association between outcomes and predictive factors. PHI and %fPSA values were stratified and previously published cut points were used for the analyses (25). The multivariate models included factors with p<0.25 in the univariate analysis or important clinical predictors. Backward elimination methods were used with an alpha of 0.05 to find the final model. The discriminatory ability of the SNP and clinical variables on positive biopsy was determined using the area under the receiver operating characteristic (ROC) curve (AUC). All statistical analyses were conducted in Stata/SE version 17.0 statistical software (Stata Corp. LP, College Station, TX).
Results
Patient characteristics stratified by prostate biopsy results
Table 1 shows the characteristics of the overall patient cohort and stratified by biopsy results. The median age was 63 years (range: 43 to 85 years). Only 6 patients were over 75 years old. Slightly over half (54.3%) were obese and the majority were ever smokers (62.3%) and alcohol drinkers (73.4%). Other common comorbid conditions included elevated cholesterol, diabetes, and hypertension. Moderate to severe lower urinary tract symptoms were observed in approximately 45% of patients. A first-degree family history of PCa was noted in about 18% of the study participants. The vast majority (87.9%) never had a prostate biopsy before.
Table 1.
Patient characteristics stratified by the results of prostate biopsy (positive and negative)
| Characteristic | Total | PCa (−), N (%) | PCa (+), N (%) | P value |
|---|---|---|---|---|
| Total | 199 (100) | 74 (37.2) | 125 (62.8) | |
| Age, years | ||||
| Mean (SD) | 63 (6.7) | 63 (6.4) | 63 (6.9) | |
| Median (min, max) | 63 (43 – 85) | 64 (46 – 77) | 63 (43 – 85) | 0.758 |
| BMI | ||||
| Non-Obese (<30 kg/m2) | 91 (45.7) | 32 (43.2) | 59 (47.2) | |
| Obese (≥30 kg/m2) | 108 (54.3) | 42 (56.8) | 66 (52.8) | 0.588 |
| Smoking | ||||
| Never | 75 (37.7) | 30 (40.5) | 45 (36.0) | |
| Ever | 124 (62.3) | 44 (59.5) | 80 (64.0) | 0.523 |
| Alcohol drinking | ||||
| Never | 53 (26.6) | 23 (31.1) | 30 (24.0) | |
| Ever | 146 (73.4) | 51 (68.9) | 95 (76.0) | 0.276 |
| Diabetes | ||||
| No | 136 (68.3) | 52 (70.3) | 84 (67.2) | |
| Yes | 63 (31.7) | 22 (29.7) | 41 (32.8) | 0.653 |
| Hypertension | ||||
| No | 57 (28.6) | 20 (27) | 37 (29.6) | |
| Yes | 142 (71.4) | 54 (73) | 88 (70.4) | 0.698 |
| High Cholesterol | ||||
| No | 101 (50.8) | 35 (47.3) | 66 (52.8) | |
| Yes | 98 (49.3) | 39 (52.7) | 59 (47.2) | 0.453 |
| Statin Use | ||||
| No | 110 (55.3) | 38 (51.4) | 72 (57.6) | |
| Yes | 89 (44.7) | 36 (48.7) | 53 (42.4) | 0.392 |
| Family history of prostate cancer (first degree relatives) | ||||
| No | 162 (81.4) | 60 (81.1) | 102 (81.6) | |
| Yes | 37 (18.6) | 14 (18.9) | 23 (18.4) | 0.928 |
| AUA Symptom Score | ||||
| 0–7 (mild) | 108 (54.3) | 39 (52.7) | 69 (55.2) | |
| 8–19 (moderate) | 71 (35.7) | 26 (35.1) | 45 (36) | 0.945 |
| ≥20 (severe) | 20 (10.1) | 9 (12.2) | 11 (8.8) | 0.452 |
| Prior prostate biopsy | ||||
| No | 175 (87.9) | 62 (83.8) | 113 (90.4) | |
| Yes | 24 (12.1) | 12 (16.2) | 12 (9.6) | 0.170 |
| Index DRE | ||||
| Normal | 136 (68.3) | 58 (78.4) | 78 (62.4) | |
| Abnormal | 63 (31.7) | 16 (21.6) | 47 (37.6) | 0.021 |
| Total PSA, ng/mL | ||||
| Mean (SD) | 12.4 (19.7) | 6.7 (4.0) | 15.8 (24.0) | 0.002 |
| Median (min, max) | 7.0 (0.9 – 138.2) | 5.5 (0.9 – 18.9) | 8.6 (2.0 – 138.2) | |
| Free PSA | ||||
| Mean (SD) | 1.5 (1.9) | 1.3 (1.5) | 1.5 (2.0) | 0.432 |
| Median (min, max) | 1.0 (0.1 – 16.4) | 1.0 (0.2 – 11.7) | 1.0 (0.1 – 16.4) | |
| %Free PSA | ||||
| Mean (SD) | 15.6 (8.6) | 19.5 (8.3) | 13.3 (7.9) | <0.001 |
| Median (min, max) | 13.6 (0.1 – 47.6) | 17.9 (5.1 – 40.7) | 11.5 (0.1 – 47.6) | |
| p2PSA, pg/mL | ||||
| Mean (SD) | 28.5 (52.1) | 15.7 (11.7) | 36.0 (64.0) | 0.008 |
| Median (min, max) | 15.6 (1.2 – 408.7) | 13.2 (1.2 – 71.3) | 17.5 (2.5 – 408.7) | |
| %Free PSA | ||||
| </=10% | 55 (27.6) | 6 (8.1) | 49 (39.2) | |
| 11–25% | 116 (58.3) | 51 (68.9) | 65 (52.0) | <0.001 |
| >25% | 28 (14.1) | 17 (23) | 11 (8.8) | <0.001 |
| rs7824364 | ||||
| TT | 125 (62.8) | 55 (74.3) | 70 (56.0) | |
| CT | 60 (30.2) | 15 (20.3) | 45 (36.0) | |
| CC | 14 (7.0) | 4 (5.4) | 10 (8.0) | |
| CT+CC | 74 (37.2) | 19 (25.7) | 55 (44.0) | 0.011 |
In the prostate biopsy, 125 (62.8%) tested positive and 74 (37.2%) tested negative. There were no significant differences in the above-mentioned epidemiological and comorbid variables between biopsy positive and negative cases. Total PSA (15.8 vs 6.7 ng/mL, P=0.002) and p2PSA (36.0 vs 15.8 pg/mL, P=0.008) were significantly higher, whereas percent free PSA (%fPSA) (13.3% vs 19.5%, P<0.001) was significantly lower in the biopsy positive than in the negative cases (Table 1). Biopsy-positive cases were more likely to have an abnormal DRE (37.6% vs 21.6 %, P=0.021). The variant C allele frequency was 22.2% and the variant allele carrier (CT+CC) frequency was 37.2%. The variant allele carriers were significantly over-represented in biopsy-positive cases (44% vs 25.7%, P=0.011) (Table 1).
Pathologic outcomes of biopsy
Table 2 shows the pathologic results of the biopsy. One hundred twenty-five were diagnosed with prostate adenocarcinoma. The remainder were benign findings (n=50), prostatic intraepithelial neoplasia (N=22) or atypia (N=2). Among the 125 patients, the distributions of grade group 1, 2, 3, 4, and 5 were 24 (19.2%), 38 (30.4%), 20 (16%), 15 (12%), and 28 (22.4%), respectively. In fact, 101 patients were diagnosed with clinically significant diseases (GS≥7).
Table 2.
Pathologic outcomes of biopsy
| Biopsy Results | N (%) |
|---|---|
| No cancer | 50 (25.1) |
| Atypia without carcinoma | 2 (1.0) |
| PIN without carcinoma | 22 (11.1) |
| Adenocarcinoma | 125 (62.8) |
| Number of cores taken | |
| 10 plus 1 | 6 (3.0) |
| ≥ 12 | 184 (92.5) |
| Side of positive cores | |
| Left | 18 (14.8) |
| Right | 31 (25.4) |
| Both | 73 (59.8) |
| Grade Group | |
| 1 | 24 (19.2) |
| 2 | 38 (30.4) |
| 3 | 20 (16.0) |
| 4 | 15 (12.0) |
| 5 | 28 (22.4) |
| Gleason Score | |
| GS < 7 (3&4) | 24 (19.2) |
| GS ≥ 7 (3&4) | 101 (80.8) |
Patient characteristics stratified by clinically significant diseases
Table 3 shows the characteristics of patient diagnosed with GS≥7 versus those with negative biopsies or GS<7. There were no significant differences in the epidemiological and comorbid variables between groups. Total PSA (17.7 vs 7.0 ng/mL, P=0.001) and p2PSA (40.4 vs 16.2 pg/mL, P=0.002) were significantly higher, whereas %fPSA (12.6% vs 18.8%, P<0.001) was significantly lower in GS≥7 than in GS<7/PCa(–) group. GS≥7 cases were more likely to have an abnormal DRE (39.6% vs 23.5%, P=0.015). The frequency of the variant allele carriers (CT+CC) was higher in GS≥7 cases (42.6% vs 31.6%), which did not reach statistical significance, likely due to limited sample size.
Table 3.
Patient characteristics by the detection of clinically significant disease (GS≥7)
| Characteristic | Total | PCa (−) & GS<7 | GS≥7, N (%) | P value |
|---|---|---|---|---|
| Total | 199 (100) | 98 (49.2) | 101 (50.8) | |
| Age, years | ||||
| Mean (SD) | 63 (6.7) | 63 (6.4) | 63 (7.0) | |
| Median (min, max) | 63 (43 – 85) | 63 (46 – 77) | 64 (43 – 85) | 0.465 |
| BMI | ||||
| Non-Obese (<30 kg/m2) | 91 (45.7) | 44 (44.9) | 47 (46.5) | |
| Obese (≥30 kg/m2) | 108 (54.3) | 54 (55.1) | 54 (53.5) | 0.817 |
| Smoking | ||||
| Never | 75 (37.7) | 43 (43.9) | 32 (31.7) | |
| Ever | 124 (62.3) | 55 (56.1) | 69 (68.3) | 0.077 |
| Alcohol drinking | ||||
| Never | 53 (26.6) | 30 (30.6) | 23 (22.8) | |
| Ever | 146 (73.4) | 68 (69.4) | 78 (77.2) | 0.212 |
| Diabetes | ||||
| No | 136 (68.3) | 68 (69.4) | 68 (67.3) | |
| Yes | 63 (31.7) | 30 (30.6) | 33 (32.7) | 0.755 |
| Hypertension | ||||
| No | 57 (28.6) | 29 (29.6) | 28 (27.7) | |
| Yes | 142 (71.4) | 69 (70.4) | 73 (72.3) | 0.771 |
| High Cholesterol | ||||
| No | 101 (50.8) | 44 (44.9) | 57 (56.4) | |
| Yes | 98 (49.3) | 54 (55.1) | 44 (43.6) | 0.104 |
| Statin Use | ||||
| No | 110 (55.3) | 48 (49) | 62 (61.4) | |
| Yes | 89 (44.7) | 50 (51) | 39 (38.6) | 0.079 |
| Family history of prostate cancer (first degree relatives) | ||||
| No | 162 (81.4) | 82 (83.7) | 80 (79.2) | |
| Yes | 37 (18.6) | 16 (16.3) | 21 (20.8) | 0.419 |
| AUA Symptom Score | ||||
| 0–7 (mild) | 108 (54.3) | 54 (55.1) | 54 (53.5) | |
| 8–19 (moderate) | 71 (35.7) | 34 (34.7) | 37 (36.6) | 0.782 |
| ≥20 (severe) | 20 (10.1) | 10 (10.2) | 10 (9.9) | 0.999 |
| Prior prostate biopsy | ||||
| No | 170 (85.4) | 83 (84.7) | 92 (91.1) | |
| Yes | 29 (14.6) | 15 (15.3) | 9 (8.9) | 0.171 |
| Index DRE | ||||
| Normal | 136 (68.3) | 75 (76.5) | 61 (60.4) | |
| Abnormal | 63 (31.7) | 23 (23.5) | 40 (39.6) | 0.015 |
| Total PSA, ng/mL | ||||
| Mean (SD) | 12.4 (19.7) | 7.0 (4.9) | 17.7 (26.2) | 0.001 |
| Median (min, max) | 7.0 (0.9 – 138.2) | 5.9 (0.9 – 37.3) | 9.2 (2.0 – 138.2) | |
| Free PSA | ||||
| Mean (SD) | 1.5 (1.9) | 1.3 (1.4) | 1.6 (2.2) | 0.189 |
| Median (min, max) | 1.0 (0.1 – 16.4) | 1.0 (0.2 – 11.7) | 1.0 (0.1 – 16.4) | |
| %Free PSA | ||||
| Mean (SD) | 15.6 (8.6) | 18.8 (8.5) | 12.6 (7.5) | <0.001 |
| Median (min, max) | 13.6 (0.1 – 47.6) | 16.1 (5.1 – 40.7) | 10.7 (0.1 – 47.6) | |
| p2PSA pg/ml | ||||
| Mean (SD) | 28.5 (52.1) | 16.2 (11.0) | 40.4 (70.5) | 0.002 |
| Median (min, max) | 15.6 (1.2 – 408.7) | 13.5 (1.2 – 71.3) | 18.5 (2.5 – 408.7) | |
| %Free PSA | ||||
| ≤10% | 55 (27.6) | 10 (10.2) | 45 (44.6) | |
| 11–25% | 116 (58.3) | 65 (66.3) | 51 (50.5) | <0.001 |
| >25% | 28 (14.1) | 23 (23.5) | 5 (5) | <0.001 |
| rs7824364 | ||||
| TT | 125 (62.8) | 67 (68.4) | 58 (57.4) | |
| CT+CC | 74 (37.2) | 31 (31.6) | 43 (42.6) | 0.111 |
Associations of clinical variables and rs7824364 genotypes with positive biopsy and GS≥7 disease
In the multivariate logistic regression analyses, an abnormal DRE was associated with increased risks of being biopsy-positive (OR=2.44, 95% CI=1.19–4.97) and being diagnosed with GS≥7 diseases (OR=2.50, 95% CI=1.26–4.93) (Table 4). Higher %fPSA levels were associated with lower risks of being diagnosed with PCa and GS≥7 diseases. A %fPSA between 11–25% had ORs of 0.16 (95% CI=0.06–0.42) for PCa(+) and 0.16 (95% CI=0.07–0.36) for GS≥7 diseases, respectively, while a %fPSA greater than 25% had ORs of 0.07 (95% CI=0.02–0.23) for PCa(+) and 0.04 (95% CI=0.01–0.16) for GS≥7 diseases, respectively.
Table 4.
Risk factors for positive biopsy and clinically significant disease (GS≥7)
| Variables | PCa (+) | GS≥7 | ||
|---|---|---|---|---|
| OR (95% CI) | P value | OR (95% CI) | P value | |
| DRE | ||||
| Normal | Ref. | Ref. | ||
| Abnormal | 2.44 (1.19–4.97) | 0.014 | 2.50 (1.26–4.93) | 0.008 |
| %Free-PSA | ||||
| </=10% | Ref. | Ref. | ||
| 11–25% | 0.16 (0.06–0.42) | <0.001 | 0.16 (0.07–0.36) | <0.001 |
| >25% | 0.07 (0.02–0.23) | <0.001 | 0.04 (0.01–0.16) | <0.001 |
| Rs7824364 | ||||
| TT | Ref. | Ref. | ||
| CT+CC | 2.14 (1.06–4.32) | 0.033 | 1.60 (0.90–2.86) | 0.111 |
Notably, SNP rs7824364 was a significant independent predictor of biopsy positivity. The variant allele carriers exhibited an over two-fold increased risk of being biopsy-positive (OR=2.14, 95% CI=1.06–4.32). Its association with GS≥7 diseases did not reach statistical significance (OR=1.60, 95% CI=0.90–2.86).
To assess the ability of rs7824364 to predict a positive biopsy, we constructed ROC and calculated AUC (Figure 1). The AUC of SNP alone in predicting a positive biopsy was 0.592 (95% CI, 0.525 to 0.658), higher than abnormal DRE alone (AUC=0.580, 95% CI, 0.644 to 0.679). With the combination of SNP and abnormal DRE, the AUC approached 0.630 (95% CI, 0.557 to 0.704). The AUC for abnormal DRE and low %fPSA was 0.717 (95% CI, 0.650 to 0.783), whereas the addition of SNP to this model increased the AUC to 0.749 (95% CI, 0.682 to 0.816).
Figure 1.
The ability of SNP rs7824364 alone and in combination with clinical variables to predict a positive biopsy in African American men. Receiver Operating Characteristic (ROC) curves were plotted, and the Area Under the ROC Curve (AUC) was calculated. Clinical variables include abornmal Digital Rectal Exam (DRE) test and low percentage of free PSA (%fPSA).
Associations of clinical variables and rs7824364 genotypes with positive biopsy and GS≥7 diseases in patients with PSA<10ng/ml and normal DRE (PHI cohort)
We then evaluated the predictive value of rs7824364 genotypes in the subset of men with PSA<10ng/ml and normal DRE (n=99) in whom the PHI test has shown a significant improvement in PCa detection when compared with tPSA and %fPSA (26). In this sub-cohort, 49 (49.5%) tested biopsy positive and 36 (36.4%) were diagnosed as GS≥7 disease (Supplemental Table 1). Total PSA and p2PSA were both higher in PCa(+) than PC(–) group and in GS≥7 group than GS<7 group, but did not reach statistical significance. Percent fPSA was significantly lower in in PCa(+) than PC(–) group (15.3% vs 19.7%, P=0.019) and in GS≥7 group than GS<7 group (14.8% vs 19.2%, P=0.025). The PHI was significantly higher in PCa(+) than PC(–) group (56.1 vs 31.9, P=0.001) and in GS≥7 group than GS<7 group (57.2 vs 36.3, P=0.01). The frequency of rs7824364 variant allele carriers was higher in PC(+) than in PC(–) group (46.9% vs 26%, P=0.032), and higher in GS≥7 cases than GS<7 group (44.4% vs 31.8%, P=0.208) (Supplemental Table 1).
In the multivariate logistic regression analyses among the PHI cohort, only PHI and the variant genotype were significantly associated with a positive biopsy (Table 5). Compared to individuals in the lowest PHI group (0–26.9), those with PHI between 27.0 and 35.9, between 36.0 and 54.9, and ≥55.0 had increased risks of being biopsy positive with ORs of 1.17 (95% CI=0.27–4.63); 14.00 (95% CI 3.99–49.16), and 11.90 (95% CI=3.34–42.35), respectively. The variant allele carriers exhibited an over two-fold increased risk of being biopsy-positive (OR=2.26, 95% CI=1.06–4.84). Likewise, PHI was an independent predictor of GS≥7 disease in PHI sub-cohort. The variant allele carriers of rs7824364 were associated with an increased risk of being diagnosed with GS≥7 disease (OR=1.72, 95% CI=0.74–4.00).
Table 5.
Risk factors for positive biopsy and GS≥7 disease in PHI cohort
| Variables | PCa (+) | GS≥7 | ||
|---|---|---|---|---|
| OR (95% CI) | P value | OR (95% CI) | P value | |
| PHI | ||||
| 0–26.9 | Ref. | Ref. | ||
| 27.0–35.9 | 1.17 (0.27–4.63) | 0.826 | 1.30 (0.29–5.83) | 0.731 |
| 36.0–54.9 | 14.00 (3.99–49.16) | <0.001 | 4.76 (1.39–16.31) | 0.013 |
| ≥55.0 | 11.90 (3.34–42.35) | <0.001 | 12.53 (3.33–47.21) | <0.001 |
| Rs7824364 | ||||
| TT | Ref. | Ref. | ||
| CT+CC | 2.26 (1.06–4.84) | 0.036 | 1.72 (0.74–4.00) | 0.208 |
We also plotted the ROC curve and calculated the AUC for rs7824364 and PHI in predicting a positive biopsy in the PHI cohort (Supplemental Figure 1). The AUC for the SNP alone was 0.605 (95% CI, 0.511 to 0.698), and for PHI alone, it was 0.784 (95% CI, 0.693 to 0.874). When combining the SNP and PHI, the AUC improved to 0.817 (95% CI, 0.732 to 0.900) in predicting a positive biopsy in this PHI cohort.
Discussion
In this study, we found that the variant genotypes of rs7824364 were predictive of a positive biopsy. The variant genotypes also seemed to be associated with the presence of clinically significant disease. Moreover, the SNP also predicted a positive biopsy in the sub-cohort of AA men with PSA levels below 10 ng/mL and a normal DRE. This specific group typically has a low biopsy detection rate and there is a pressing need for biomarkers that can improve the detection rate in this group. To our knowledge, this is the first report of significant associations between an African ancestry-specific SNP and biopsy positivity in a prospective cohort of AA men with suspected PCa following PSA screening.
Inherited germline PCa susceptibility SNPs may have clinical applications in the pre-PSA screening setting and post-PSA screening setting (11, 12). To date, most of the efforts have been devoted to the pre-PSA screening application (13–15). Shi et al. (13) constructed a GRS using GWAS-identified PCa risk SNPs and compared the performance of this GRS with two currently guideline-recommended inherited risk measures, family history (FH) and rare pathogenic mutations (RPM), for predicting PCa incidence and mortality. The results showed that while FH and RPMs identified 11% of men at a higher PCa risk, incorporation of GRS identified an additional 22 % of men at higher PCa risk. Conti et al. (8) used the GRS to project the lifetime absolute risks of PCa and reported that the absolute risk for men in the top decile of the GRS reached 38% for both men of European and African ancestry. These results support that PCa susceptibility SNPs can be used for risk stratification before PSA testing to increase screening for men with high risk and reduce screening for men with low risk.
The second potential clinical application of PCa susceptibility SNPs is in the post-PSA screening setting. Due to the limited specificity of a single PSA test, various additional parameters have been investigated to enhance patient selection for biopsy, such as the PHI and 4KScore (25, 27, 28). More recently, SNPs have been incorporated into prediction algorithms in post-PSA setting (16, 17, 29–33). The Stockholm3 (STHLM3) test, which combines clinical variables, five plasma protein markers, and a GRS derived from 101 PCa susceptibility SNPs, has been shown to reduce the number of performed prostate biopsies while maintaining sensitivity for clinically significant diseases in European descendants (17, 29–33). However, there has been no prior report using SNPs to enhance the prediction of PCa risk for biopsy referral in AA men with elevated PSA. In this study, we found that carriers of the variant alleles of rs7824364 exhibited significantly increased risks of being biopsy positive in the overall cohort and in the sub-cohort of AA men with PSA levels below 10 ng/mL and a normal DRE. In this specific sub-cohort, the detection rate of PCa through biopsy is approximately one-third, while the remaining two-thirds of biopsies yield benign results (34). Thus, enhancing the prediction of biopsy outcomes has the potential to significantly reduce the number of unnecessary biopsies in this group. Although the effect size of individual SNPs is modest, the collective impact of multiple SNPs, when combined as a GRS, along with FH and RPM, can amplify the predictive capability and enhance the prediction of biopsy positivity in AA men undergoing PSA screening.
The 8q24 region is a major susceptibility region for PCa and at least 12 independent susceptibility SNPs have been identified among European descendants (20, 35). Rs7824364 has a risk allele frequency of 24–25% in African descendants, while it is not polymorphic in Europeans and Asians (Supplemental Table 2). The relatively high frequency of the risk allele in AA and its absence in other racial groups suggests that this SNP may play a role in the disparity of PCa incidence between AA and other races.
From a biological perspective, several 8q24 SNPs have been found to influence transcriptional enhancer activity and long-range regulation of MYC and other genes that are important for cancer development and progression (36–38). Rs7824364 is located near several PCa-associated long non-coding RNAs (lncRNAs) on 8q24.21, including PCAT1, PCAT2, and PRNCR1 (22). PCAT1 is significantly overexpressed in most prostate cancer tissues and acts as an oncogenic lncRNA during prostate carcinogenesis (39). It represses the expression of BRCA2 tumor suppressor, leading to impaired homologous recombination (40). PCAT1 also activates AKT and NF-kappaB signaling in PCa (41). PRNCR1 is another oncogenic lncRNA that is overexpressed in PCa and modulates androgen receptor-mediated gene expression (42). The underlying biological mechanisms linking rs7824364 to PCa development and progression warrant further studies.
This study has several strengths. Firstly, it utilized prospectively collected samples obtained prior to prostate biopsy. Secondly, the inclusion of patients from three different hospitals enhances the generalizability of the findings. Additionally, comprehensive epidemiological, comorbid, and clinicopathological information was collected. The study demonstrates the potential utility of ancestry-specific PCa susceptibility SNPs in improving disease detection rates during prostate biopsy in the post-PSA screening setting. However, there are a few limitations. Firstly, the sample size was relatively small, and the analysis focusing on clinically significant disease did not reach statistical significance. Secondly, only one SNP was studied based on our previous publication. Neither rs7824364 nor its proxy SNPs (R2>0.8) was included in any of previous GRSs (8–12), likely due to its exclusive polymorphism in individuals of African ancestry. Of note, the GRS451 in men of African ancestry was associated with a significantly greater risk of aggressive versus non-aggressive prostate cancer (10). Thus, the inclusion of rs7824364 in future GRS assessing PCa risk/phenotype among men of African ancestry could add predictive accuracy. Thirdly, the overall cancer and high-grade cancer detection rates observed in this study were higher than the rates typically reported in the literature, likely due to higher prevalence of west African ancestry in our population. Finally, among men with an elevated serum PSA level or abnormal DRE, MRI is becoming a standard tool to select men for biopsy and to target abnormal areas more precisely. MRI-guided biopsies were not performed in this current study. In this context, SNPs may assist in determining who should or should not undergo MRI imaging after PSA screening. This could help avoid the morbidity or cost associated with an MRI procedure.
In conclusion, this pilot proof-of-principle study demonstrated that the variant genotypes of the African ancestry-specific SNP rs7824364 on 8q24 independently predicted a positive prostate biopsy in AA men with elevated PSA levels or abnormal DRE during screening. Future research should explore the inclusion of additional SNPs to develop a comprehensive GRS and enhance predictive accuracy for AA men.
Supplementary Material
Acknowledgements
This study was partially supported by grants 8U54MD 007587-03 from the National Institute on Minority Health and Health Disparities of the National Institutes of Health (NIH), U54CA096297/CA096300 (University of Puerto Rico and the MD Anderson Cancer Center Partnership for Excellence in Cancer Research) from NIH, and Prostate Cancer Foundation - via a VALOR Challenge Grant. Beckman Coulter provided reagent kits and partial financial support for the study.
Conflict of Interest
Curtis Pettaway received an unrestricted grant from Beckman Coulter to cover serum processing and technician effort for serum prostate specific antigen[PSA] and PSA isoforms assessed in the study.
References
- 1.Siegel RL, Miller KD, Wagle NS, Jemal A. Cancer statistics, 2023. CA Cancer J Clin. 2023;73(1):17–48. [DOI] [PubMed] [Google Scholar]
- 2.Beebe-Dimmer JL, Kapron AL, Fraser AM, Smith KR, Cooney KA. Risk of Prostate Cancer Associated With Familial and Hereditary Cancer Syndromes. J Clin Oncol. 2020;38(16):1807–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.DeSantis CE, Miller KD, Goding Sauer A, Jemal A, Siegel RL. Cancer statistics for African Americans, 2019. CA Cancer J Clin. 2019;69(3):211–33. [DOI] [PubMed] [Google Scholar]
- 4.Mucci LA, Hjelmborg JB, Harris JR, Czene K, Havelick DJ, Scheike T, et al. Familial Risk and Heritability of Cancer Among Twins in Nordic Countries. JAMA. 2016;315(1):68–76. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Hoffmann TJ, Van Den Eeden SK, Sakoda LC, Jorgenson E, Habel LA, Graff RE, et al. A large multiethnic genome-wide association study of prostate cancer identifies novel risk variants and substantial ethnic differences. Cancer Discov. 2015;5(8):878–91. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Schumacher FR, Al Olama AA, Berndt SI, Benlloch S, Ahmed M, Saunders EJ, et al. Association analyses of more than 140,000 men identify 63 new prostate cancer susceptibility loci. Nat Genet. 2018;50(7):928–36. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Takata R, Takahashi A, Fujita M, Momozawa Y, Saunders EJ, Yamada H, et al. 12 new susceptibility loci for prostate cancer identified by genome-wide association study in Japanese population. Nat Commun. 2019;10(1):4422. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Conti DV, Darst BF, Moss LC, Saunders EJ, Sheng X, Chou A, et al. Trans-ancestry genome-wide association meta-analysis of prostate cancer identifies new susceptibility loci and informs genetic risk prediction. Nat Genet. 2021;53(1):65–75. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Chen F, Madduri RK, Rodriguez AA, Darst BF, Chou A, Sheng X, et al. Evidence of Novel Susceptibility Variants for Prostate Cancer and a Multiancestry Polygenic Risk Score Associated with Aggressive Disease in Men of African Ancestry. Eur Urol. 2023;84(1):13–21. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Wang A, Shen J, Rodriguez AA, Saunders EJ, Chen F, Janivara R, et al. Characterizing prostate cancer risk through multi-ancestry genome-wide discovery of 187 novel risk variants. Nat Genet. 2023;55(12):2065–74. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Kearns JT, Helfand BT, Xu J. Moving Prostate Cancer Polygenic Risk Scores from Research Towards Clinical Practice. Eur Urol Focus. 2022;8(4):913–5. [DOI] [PubMed] [Google Scholar]
- 12.Xu J, Resurreccion WK, Shi Z, Wei J, Wang CH, Zheng SL, et al. Inherited risk assessment and its clinical utility for predicting prostate cancer from diagnostic prostate biopsies. Prostate Cancer Prostatic Dis. 2022;25(3):422–30. [DOI] [PubMed] [Google Scholar]
- 13.Shi Z, Platz EA, Wei J, Na R, Fantus RJ, Wang CH, et al. Performance of Three Inherited Risk Measures for Predicting Prostate Cancer Incidence and Mortality: A Population-based Prospective Analysis. Eur Urol. 2021;79(3):419–26. [DOI] [PubMed] [Google Scholar]
- 14.Huynh-Le MP, Fan CC, Karunamuni R, Thompson WK, Martinez ME, Eeles RA, et al. Polygenic hazard score is associated with prostate cancer in multi-ethnic populations. Nat Commun. 2021;12(1):1236. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Chen F, Madduri RK, Rodriguez AA, Darst BF, Chou A, Sheng X, et al. Evidence of Novel Susceptibility Variants for Prostate Cancer and a Multiancestry Polygenic Risk Score Associated with Aggressive Disease in Men of African Ancestry. Eur Urol. 2023. [DOI] [PMC free article] [PubMed]
- 16.Akamatsu S, Terada N, Takata R, Kinoshita H, Shimatani K, Momozawa Y, et al. Clinical Utility of Germline Genetic Testing in Japanese Men Undergoing Prostate Biopsy. JNCI Cancer Spectr. 2022;6(1). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Gronberg H, Adolfsson J, Aly M, Nordstrom T, Wiklund P, Brandberg Y, et al. Prostate cancer screening in men aged 50–69 years (STHLM3): a prospective population-based diagnostic study. Lancet Oncol. 2015;16(16):1667–76. [DOI] [PubMed] [Google Scholar]
- 18.Kader AK, Sun J, Reck BH, Newcombe PJ, Kim ST, Hsu FC, et al. Potential impact of adding genetic markers to clinical parameters in predicting prostate biopsy outcomes in men following an initial negative biopsy: findings from the REDUCE trial. Eur Urol. 2012;62(6):953–61. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Jiang H, Liu F, Wang Z, Na R, Zhang L, Wu Y, et al. Prediction of prostate cancer from prostate biopsy in Chinese men using a genetic score derived from 24 prostate cancer risk-associated SNPs. Prostate. 2013;73(15):1651–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Matejcic M, Saunders EJ, Dadaev T, Brook MN, Wang K, Sheng X, et al. Germline variation at 8q24 and prostate cancer risk in men of European ancestry. Nat Commun. 2018;9(1):4616. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Freedman ML, Haiman CA, Patterson N, McDonald GJ, Tandon A, Waliszewska A, et al. Admixture mapping identifies 8q24 as a prostate cancer risk locus in African-American men. Proc Natl Acad Sci U S A. 2006;103(38):14068–73. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Han Y, Rand KA, Hazelett DJ, Ingles SA, Kittles RA, Strom SS, et al. Prostate Cancer Susceptibility in Men of African Ancestry at 8q24. J Natl Cancer Inst. 2016;108(7). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Irizarry-Ramirez M, Kittles RA, Wang X, Salgado-Montilla J, Nogueras-Gonzalez GM, Sanchez-Ortiz R, et al. Genetic ancestry and prostate cancer susceptibility SNPs in Puerto Rican and African American men. Prostate. 2017;77(10):1118–27. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Iczkowski KA, van Leenders G, van der Kwast TH. The 2019 International Society of Urological Pathology (ISUP) Consensus Conference on Grading of Prostatic Carcinoma. Am J Surg Pathol. 2021;45(7):1007. [DOI] [PubMed] [Google Scholar]
- 25.Loeb S, Sanda MG, Broyles DL, Shin SS, Bangma CH, Wei JT, et al. The prostate health index selectively identifies clinically significant prostate cancer. J Urol. 2015;193(4):1163–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Catalona WJ, Partin AW, Sanda MG, Wei JT, Klee GG, Bangma CH, et al. A multicenter study of [−2]pro-prostate specific antigen combined with prostate specific antigen and free prostate specific antigen for prostate cancer detection in the 2.0 to 10.0 ng/ml prostate specific antigen range. J Urol. 2011;185(5):1650–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Bryant RJ, Sjoberg DD, Vickers AJ, Robinson MC, Kumar R, Marsden L, et al. Predicting high-grade cancer at ten-core prostate biopsy using four kallikrein markers measured in blood in the ProtecT study. J Natl Cancer Inst. 2015;107(7). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Nordstrom T, Vickers A, Assel M, Lilja H, Gronberg H, Eklund M. Comparison Between the Four-kallikrein Panel and Prostate Health Index for Predicting Prostate Cancer. Eur Urol. 2015;68(1):139–46. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Gronberg H, Eklund M, Picker W, Aly M, Jaderling F, Adolfsson J, et al. Prostate Cancer Diagnostics Using a Combination of the Stockholm3 Blood Test and Multiparametric Magnetic Resonance Imaging. Eur Urol. 2018;74(6):722–8. [DOI] [PubMed] [Google Scholar]
- 30.Eklund M, Nordstrom T, Aly M, Adolfsson J, Wiklund P, Brandberg Y, et al. The Stockholm-3 (STHLM3) Model can Improve Prostate Cancer Diagnostics in Men Aged 50–69 yr Compared with Current Prostate Cancer Testing. Eur Urol Focus. 2018;4(5):707–10. [DOI] [PubMed] [Google Scholar]
- 31.Moller A, Olsson H, Gronberg H, Eklund M, Aly M, Nordstrom T. The Stockholm3 blood-test predicts clinically-significant cancer on biopsy: independent validation in a multicenter community cohort. Prostate Cancer Prostatic Dis. 2019;22(1):137–42. [DOI] [PubMed] [Google Scholar]
- 32.Vigneswaran HT, Discacciati A, Gann PH, Gronberg H, Eklund M, Abern MR. Ethnic variation in prostate cancer detection: a feasibility study for use of the Stockholm3 test in a multiethnic U.S. cohort. Prostate Cancer Prostatic Dis. 2021;24(1):120–7. [DOI] [PubMed] [Google Scholar]
- 33.Nordstrom T, Discacciati A, Bergman M, Clements M, Aly M, Annerstedt M, et al. Prostate cancer screening using a combination of risk-prediction, MRI, and targeted prostate biopsies (STHLM3-MRI): a prospective, population-based, randomised, open-label, non-inferiority trial. Lancet Oncol. 2021;22(9):1240–9. [DOI] [PubMed] [Google Scholar]
- 34.Xu J. The Xu’s chart for prostate biopsy: a visual presentation of the added value of biomarkers to prostate-specific antigen for estimating detection rates of prostate cancer. Asian J Androl. 2014;16(4):536–40. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Al Olama AA, Kote-Jarai Z, Giles GG, Guy M, Morrison J, Severi G, et al. Multiple loci on 8q24 associated with prostate cancer susceptibility. Nat Genet. 2009;41(10):1058–60. [DOI] [PubMed] [Google Scholar]
- 36.Jia L, Landan G, Pomerantz M, Jaschek R, Herman P, Reich D, et al. Functional enhancers at the gene-poor 8q24 cancer-linked locus. PLoS Genet. 2009;5(8):e1000597. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Wasserman NF, Aneas I, Nobrega MA. An 8q24 gene desert variant associated with prostate cancer risk confers differential in vivo activity to a MYC enhancer. Genome Res. 2010;20(9):1191–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Pomerantz MM, Beckwith CA, Regan MM, Wyman SK, Petrovics G, Chen Y, et al. Evaluation of the 8q24 prostate cancer risk locus and MYC expression. Cancer Res. 2009;69(13):5568–74. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Xiong T, Li J, Chen F, Zhang F. PCAT-1: A Novel Oncogenic Long Non-Coding RNA in Human Cancers. Int J Biol Sci. 2019;15(4):847–56. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Prensner JR, Chen W, Iyer MK, Cao Q, Ma T, Han S, et al. PCAT-1, a long noncoding RNA, regulates BRCA2 and controls homologous recombination in cancer. Cancer Res. 2014;74(6):1651–60. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Shang Z, Yu J, Sun L, Tian J, Zhu S, Zhang B, et al. LncRNA PCAT1 activates AKT and NF-kappaB signaling in castration-resistant prostate cancer by regulating the PHLPP/FKBP51/IKKalpha complex. Nucleic Acids Res. 2019;47(8):4211–25. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Yang L, Lin C, Jin C, Yang JC, Tanasa B, Li W, et al. lncRNA-dependent mechanisms of androgen-receptor-regulated gene activation programs. Nature. 2013;500(7464):598–602. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.