Performance of Prostate Health Index in Biopsy Naïve Black Men

Abstract

Purpose:

The Prostate Health Index is validated for prostate cancer detection but has not been well validated for Gleason grade group 2-5 prostate cancer detection in Black men. We hypothesize that the Prostate Health Index has greater accuracy than prostate specific antigen for detection of Gleason grade group 2-5 prostate cancer. We estimated probability of overall and Gleason grade group 2-5 prostate cancer across previously established Prostate Health Index ranges and identified Prostate Health Index cutoffs that maximize specificity for Gleason grade group 2-5 prostate cancer with sensitivity >90%.

Materials and Methods:

We recruited a “cancer-free” Black control cohort (135 patients) and a cohort of biopsy naïve Black men (158) biopsied for elevated prostate specific antigen. Descriptive statistics compared the prostate cancer cases and controls and the frequency of Gleason grade group 2-5 prostate cancer across Prostate Health Index scores. Receiver operating characteristics compared the discrimination of prostate specific antigen, Prostate Health Index and other prostate specific antigen related biomarkers. Sensitivity and specificity for Gleason grade group 2-5 prostate cancer detection were assessed at prostate specific antigen and Prostate Health Index thresholds alone and in series.

Results:

Of biopsied subjects 32.9% had Gleason grade group 2-5 prostate cancer. In Blacks with prostate specific antigen from 4.0–10.0 ng/ml, Prostate Health Index and prostate specific antigen had similar discrimination for Gleason grade group 2-5 prostate cancer (0.63 vs 0.57, p=0.27). In Blacks with prostate specific antigen ≤10.0, a threshold of prostate specific antigen ≥4.0 had 90.4% sensitivity for Gleason grade group 2-5 prostate cancer; a threshold of prostate specific antigen ≥4.0 with Prostate Health Index ≥35.0 in series avoided unnecessary biopsy in 33.0% of men but missed 17.3% of Gleason grade group 2-5 prostate cancer. Prostate specific antigen ≥4.0 with Prostate Health Index ≥28.0 in series spared biopsy in 17.9%, while maintaining 90.4% sensitivity of Gleason grade group 2-5 prostate cancer.

Conclusions:

The Prostate Health Index has moderate accuracy in detecting Gleason grade group 2-5 prostate cancer in Blacks, but Prostate Health Index ≥28.0 can be safely used to avoid some unnecessary biopsies in Blacks.

Prostate specific antigen testing has led to improved detection of prostate cancer and a reduction in prostate cancer specific mortality.¹ However, the poor specificity results in excessive biopsies in over 50% of men (ie negative biopsies and indolent prostate cancer).^2,3 Post-biopsy sepsis and hospitalization occur in 1%–3% of biopsies.⁴ Over detection of indolent prostate cancer causes overtreatment and morbidity, and generates high health care costs.⁵ Given the limitations and risks of prostate specific antigen screening, there is growing interest in new screening tests that accurately detect clinically significant Gleason grade group 2-5 prostate cancer (ie Gleason score ≥3+4) and help men avoid unnecessary biopsies (through improved specificity).

In 2012, the FDA approved the Prostate Health Index, a biomarker with increased specificity for PCa detection in the diagnostic gray zone of PSA from 4.0–10.0 ng/ml with normal digital rectal examination.⁶ The PHI assay uses a composite of total PSA, free PSA and p2PSA.⁷ Several multicenter studies have shown that PHI outperforms PSA in the detection of PCa^7–14 with some studies demonstrating PHI’s increased accuracy for GG 2-5 PCa in men within the diagnostic PSA gray zone.^8,14 It has been determined that PHI should not be used for primary screening in the general population, but instead as a reflex confirmatory test, using a threshold of PHI ≥35.0 for biopsy decisions.¹⁰ Validation studies in diverse populations have suggested varying optimal cut points for PHI across ethnic groups.^15,16 For instance, at a sensitivity of 90%, reported PHI cutoffs include 28.6 for European Americans,¹² 26.5 for Asian Americans¹⁷ and 31.6 for European Whites.¹⁸ Few studies have evaluated the performance of PHI in detecting GG 2-5 PCa in Black cohorts,¹⁹ so its generalizability in this population remains unestablished. Given the high risk of PCa in Black men, PHI should be validated in this population.^20,21

We hypothesize that the accuracy of PHI in GG 2-5 PCa detection is greater than PSA in biopsy naïve Black men. Our secondary objectives were to 1) assess the prevalence of GG 2-5 PCa in Black men across assay reported PHI scores, and 2) assess the overall sensitivity and specificity for GG 2-5 PCa in Black men of PSA and PHI thresholds alone and of PHI when paired with a PSA ≥4.0 ng/ml threshold.^7,12,22,23

METHODS

Study Population

This study utilizes 2 different cohorts recruited in Chicago, Illinois. Research coordinators administered a survey that captured sociodemographic information, substance use, and family and medical history to both cohorts. The first cohort of “cancer-free” controls included 139 self-reported Black men, age 40–79 years. They were recruited from community PCa screening events in Chicago, Illinois from 2018–2019 through the social networks of 8 Black Citizen Scientists who were interested in Black men’s health (Black Citizen Scientists Study). Controls did not undergo assessment of prostate volume or prostate biopsy unless referred by their physician for PB (3). Men with PSA ≥4.0 ng/ml (13) were tracked for 1 year for PSAs or biopsies (1) that occurred after enrollment. Coordinators collected information on repeat PSA, potential biopsies and PCa diagnoses by administering a 1-page survey 12 months after enrollment for the BCS cohort. Four men were excluded for PSA >10.0 ng/ml (3) or subsequent PCa diagnosis (1), leaving 135 for analysis. Men with a history of 5-alpha reductase inhibitor use, prior PB, abnormal DRE, and men with PSA >10.0 ng/ml were excluded from the cohort in order to establish median levels of PHI in “cancer-free” controls (135).

The second cohort, the PHI Biopsy Study, comprised 158 Black men, age 40–79 years, referred to outpatient urology clinics at Northwestern Memorial, John H. Stroger, or University of Illinois at Chicago Hospitals for elevated PSA between February 2017 and January 2020. Coordinators administered the same sociodemographic and medical history survey immediately before the biopsy. Coordinators then reviewed the electronic health records of each participant 1 month and 12 months after the biopsy to collect information on repeat PSAs, PB pathology reports, imaging and clinical stage (if cancer was present). Men underwent PHI testing immediately prior to the initial TRUS guided PB. Men with a history of 5-alpha reductase inhibitor use, prior PB, abnormal DRE, and men with PSA >10.0 ng/ml were excluded from the cohort.

For the study we include all men with PSA ≤10.0 ng/ml but for the comparison of AUC between PHI and PSA, we limited it to the 4.0–10.0 ng/ml range (142), consistent with its FDA approval. Other studies have varied the inclusion criteria to include men with PSA as low as 2.0 ng/ml, and we similarly included PSA levels below 4.0 ng/ml to evaluate the impact of various PSA thresholds on sensitivity and specificity. Coordinators administered the same sociodemographic and medical history survey immediately before the biopsy. Coordinators then reviewed the electronic health records 12 months after the biopsy to collect information on repeat PSAs, PB pathology reports, imaging and clinical stage (if cancer was present). Baseline demographic and clinical data included age, race, family history of PCa, suspicious DRE, referral PSA level and the PSA related biomarkers in the PHI assay along with the PHI score.

The Beckman Coulter Access® 2 immunoassay analyzer (Beckman Coulter, Brea, California, USA) was used to measure serum PSA, fPSA and p2PSA. PHI was calculated by the assay as (p2PSA/fPSA)×(PSA½), and % fPSA was defined as (fPSA/PSA)×100.

As supplementary data, we provide the normal age specific ranges of PHI in community dwelling Black, “cancer-free” controls (supplementary table, https://www.jurology.com). We also include the AUC data for GG 1-5 PCa using PSA, PHI and the PSA related biomarkers.

IRB Approval

The BCS study was approved by the institutional review boards at Northwestern University, University of Illinois at Chicago and Northeastern Illinois University (IRB No. STU00206075). The PBS was approved by the IRBs at Northwestern University, University of Illinois at Chicago and Cook County Health (IRB No. STU0020425).

Pathological Review

Pathological assessment of biopsy specimens was performed by expert uropathologists in accordance with the 2016 International Society of Urological Pathology Consensus Conference, ie GG 1: Gleason score ≤6; GG 2: Gleason score 3+4=7; GG 3: Gleason score 4+3=7; GG 4: Gleason score=8; and GG 5: Gleason score=9–10.²⁴ The primary outcome was GG 2-5 PCa.

Statistical Analysis

Statistical comparisons were made using the chi-square test for proportions; comparisons of medians and IQRs were made with medians test for continuous variables. Receiver operating characteristic curves were constructed for nonparametric data to compare the AUC of the referral PSA, PHI and other PSA related biomarkers for GG 2-5 PCa and GG 1-5 PCa (supplementary figure, https://www.jurology.com) in men with PSA from 4.0–10.0 ng/ml using the Kolmogorov-Smirnov test.

For objective 1, the percentages of men with overall and GG 2-5 PCa were reported across previously established PHI score ranges. For comparison, we include the estimated probabilities of overall PCa from the PHI assay report.⁶

For objective 2, thresholds of PSA (from 2.0–6.0) and PHI (from 20–55) were generated, and the specificities for GG 2-5 PCa were compared at ≥90%, and ≥95% sensitivity among biopsy subjects. For brevity, we include only thresholds of PHI where sensitivity was ≥90%. Since PSA ≥4.0 ng/ml is the standard in the U.S.,²² we tested the impact of using various PHI thresholds after serum PSA ≥4.0 ng/m. on sensitivity and specificity for GG 2-5 PCa in the biopsy subjects. To estimate the impact of these thresholds on a Black screening population, we used the PHI and PSA results of the BCS cohort to determine the percentage of “cancer-free” controls that would exceed each threshold and be referred to urologists for a PB. We used the PSA in the PHI assay since these men were recruited in community PCa screening events.

Based on data from Loeb, de la Calle and Nordström et al, we powered this study for comparing the discrimination of PHI and PSA in GG 2-5 PCa detection in Blacks.^8,12,14 Assuming an AUC of 0.60 for PSA, with 142 Black men, and a difference of 0.12 in AUC, we are powered at 82% with a 2-sided alpha of 0.05.

All comparisons were 2-sided and p values <0.05 indicated statistical significance. Statistical analysis was performed using SPSS® 25 and MedCalc® 19.0.5.

RESULTS

Demographics and Distribution

The study included 158 Black men from the PBS cohort and 135 men from the BCS cohort who were included as “cancer-free” controls (table 1). Compared to the controls, men in the PBS cohort were older (median age 60.0 vs 55.0 years), had higher PSA levels and PHI scores, and had greater frequencies of family history of PCa and personal history of benign prostatic hyperplasia/lower urinary tract symptoms (benign prostatic hyperplasia/lower urinary tract symptoms, all p >0.01).

Table 1.

Clinical characteristics by study population

	Citizen Scientists Study		PHI Biopsy Study Neg Biopsy/GG 1 PCa		PHI Biopsy Study GG 2-5 PCa		p Value
No. pts	135		106		52
Continuous variables (median/IQR):*
Age (yrs)	55.0 (48.0, 60.0)		60.0 (56.0, 64.0)		60.5 (54.3, 64.8)		<0.001
Body mass index (kg/m2)	27.9 (25.4, 31.6)		30.0 (26.6, 34.0)		28.7 (25.2, 31.6)		0.049
PSA (ng/ml)	0.91 (0.45, 2.1)		5.6 (4.7, 7.6)		6.5 (4.6, 7.74)		<0.001
PHI	19.8 (13.3, 28.2)		51.9 (34.7, 70.1)		63.6 (48.7, 94.8)		<0.001
% Free PSA	24.1 (16.7, 35.2)		14.8 (10.4, 20.2)		9.3 (6.0, 14.5)		<0.001
−2proPSA	5.0 (2.9, 8.5)		17.4 (10.3, 27.3)		16.1 (11.5, 23.8)		<0.001
Prostate vol (cm3)		-	42.4 (32.7, 59.5)		29.5 (22.1, 43.0)		<0.001
PSA density (ng/ml/cm3)		-	0.12 (0.09, 0.17)		0.18 (0.13, 0.31)		<0.001
PHI density		-	1.17 (0.66, 1.79)		2.00 (1.38, 3.39)		<0.001
Categorical variables (No./%):†
Family history of PCa	11	(8.1)	21	(19.8)	8	(15.4)	0.03
History of benign prostatic hyperplasia	3	(2.2)	20	(18.9)	8	(15.4)	<0.001
PHI ≥35.0	24	(17.8)	78	(73.6)	47	(90.4)	<0.001
PSA <4 ng/ml	126	(93.3)	11	(10.4)	5	(9.6)	<0.001
PSA 4–10 ng/ml	9	(6.7)	95	(89.6)	47	(90.4)	<0.001
Currently married	39	(28.9)	40	(37.7)	26	(50.0)	0.04
Income <$30,000	69	(51.1)	69	(65.1)	18	(34.6)	0.001
Current smoking	42	(31.1)	35	(33.0)	18	(34.6)	0.85
Heavy drinking	12	(8.9)	6	(5.7)	8	(15.4)	0.12

Bold type indicates p <0.05.

^*Independent sample medians test.

^†Chi-square test.

In the PBS group, using the same PHI categories that are included in the PHI report,⁶ the risk of overall and GG 2-5 PCa for Blacks increases with higher scores (table 2). Black men with PHI scores between 27.0–35.9, 36.0–54.9, and ≥55.0 had a PCa prevalence of 45.5%, 51.3%, and 65.8% and had GG 2-5 PCa in 18.2%, 35.1%, and 39.5%, respectively. No Black men with PHI <27.0 (7) had biopsy proven PCa. Notably, Blacks in the PBS had higher rates of PCa in every category than the population used to validate PHI⁷ for PHI scores above 27.0.

Table 2.

Probability of overall and Gleason grade group 2-5 prostate cancer based on PHI results with PSA 4.0–10.0 ng/ml

PHI Range	Probability of PCa (95% CI)*	Probability of PCa in Blacks (142)	Probability of GG 2-5 PCa in Blacks (142)
0–26.9	9.8% (5.2%–15.4%)	0.0% (0/7)	0.0% (0/7)
27.0–35.9	16.8% (11.3%–22.2%)	45.5% (10/22)	18.2% (4/22)
36.0–54.9	33.3% (26.8%–39.9%)	51.3% (19/37)	35.1% (13/37)
≥55.0	50.1% (39.8%–61.0%)	65.8% (50/76)	39.5% (30/76)

^*Based on Hybritech PHI assay.

Median PHI scores and PSA levels in the BCS cohort by decade of age are presented in the supplementary table (https://www.jurology.com). There was a threefold increase in PSA from men aged 40–49 years (0.69 ng/ml) to men aged 70–79 years (2.11 ng/ml). The PHI score, however, remained relatively constant across all 4 age groups (median range 18.12–22.12).

Discrimination Analysis

For the discrimination analysis of GG 2-5 PCa, we included men in the PBS cohort with PSA from 4.0–10.0 ng/ml and normal DRE, since this is the FDA approved indication. PHI was statistically similar to PSA in the detection of GG 2-5 PCa in Black men (AUC 0.63 vs 0.57, p =0.27). PHI density (AUC 0.74) and PSA density (AUC 0.74) both showed statistically greater accuracy in detecting GG 2-5 PCa relative to PSA and PHI (see figure).

Receiver operating curves for PSA vs PSA related biomarkers for detection of Gleason grade group 2-5 prostate cancer in men with PSA 4.0–10.0 ng/ml. A, receiver operating characteristic curves are drawn for PHI, PSA, % free PSA, ⁻²proPSA, PSA density and PHI density in men with PSA 4.0–10.0 ng/ml. B, area under receiver operating characteristic curve for detection of GG 2-5 prostate cancer is shown for each biomarker in men with PSA 4.0–10.0 ng/ml. Biomarkers were log-transformed. p Values were compared between biomarkers and PSA. Bold font indicates AUC comparisons with p values less than 0.05.

A secondary discrimination analysis of overall PCa was performed using the PBS population. For overall PCa as well, PHI was statistically similar to PSA in the detection of GG 1-5 PCa (AUC 0.66 vs 0.64, p=0.75). PSA density (AUC 0.75) was statistically more accurate than PSA in detecting overall PCa, however, while PHI density had a higher AUC than PSA (AUC 0.73) it was not a statistically significant difference (p=0.11, supplementary figure, https://www.jurology.com).

Sensitivity and Specificity Analysis

Sensitivity and specificity were calculated for detecting GG 2-5 PCa using varied thresholds of PSA and PHI alone and in series (table 3) to achieve ≥90% and ≥95% sensitivity for GG 2-5 PCa. To estimate the impact on screening healthy Black men, we calculated the percentage of the BCS “cancer-free” controls that would exceed each threshold (ie have a false-positive test) and undergo a theoretically unnecessary biopsy. A sensitivity of 90.4% is achieved with a PHI threshold of ≥35.0, a PSA threshold of ≥4.0 ng/ml, and a PSA ≥4.0 ng/ml with PHI ≥28.0 in series. These correspond to specificities of 26.2%, 10.4% and 17.9%, respectively, which are proxies for avoidable biopsies. However, these thresholds would lead to biopsies in 17.8%, 6.7%, and 5.9% of “cancer-free” controls, respectively. The current standard of care combines the screening threshold of PSA ≥4.0 and PHI ≥35.0 and captures only 82.7% of GG 2-5 PCa, but spares 35 (33.0%) men an unnecessary biopsy. This also causes 8 (5.9%) unnecessary biopsies among “cancer-free” controls. Meanwhile, using the screening threshold of PSA ≥4.0 ng/ml with PHI ≥28.0 maintains the same sensitivity as PSA ≥4.0 ng/ml alone, 90.4% for GG 2-5 PCa, and spares 28 (17.9%) men an unnecessary biopsy.

Table 3.

Sensitivity and specificity for Gleason grade group 2-5 prostate cancer of PSA and PHI alone and in series with PSA ≤10.0 ng/ml

Biopsy Strategy	Sensitivity-Blacks in Biopsy Cohort	Specificity-Blacks in Biopsy Cohort	“Cancer-Free” Controls Exceeding Threshold
No. pts	158	158	135
PSA alone (%):
≥3.0	100.0	4.7	11.1
≥3.5	96.2	6.6	8.1
≥4.0	90.4	10.4	6.7
≥5.0	71.2	33.0	4.4
≥6.0	51.9	33.0	1.5
PHI alone (%):
≥27.0	100.0	9.4	31.1
≥28.0	100.0	10.4	25.9
≥28.6	98.1	10.4	23.7
≥29.0	96.2	10.4	23.7
≥30.0	94.2	15.1	22.2
≥31.0	92.3	18.9	21.5
≥32.0	92.3	19.8	21.5
≥33.0	90.4	21.7	20.0
≥34.0	90.4	22.6	18.5
≥35.0	90.4	26.2	17.8
PSA+PHI in series (%):
PSA ≥4.0, PHI ≥27.0	90.4	17.0	5.9
PSA ≥4.0, PHI ≥28.0	90.4	17.9	5.9
PSA ≥4.0, PHI ≥28.6	88.5	17.9	5.9
PSA ≥4.0, PHI ≥29.0	88.5	17.9	5.9
PSA ≥4.0, PHI ≥30.0	86.5	21.7	5.9
PSA ≥4.0, PHI ≥31.0	84.6	25.5	5.9
PSA ≥4.0, PHI ≥32.0	84.6	26.4	5.9
PSA ≥4.0, PHI ≥33.0	82.7	28.3	5.9
PSA ≥4.0, PHI ≥34.0	82.7	29.2	5.9
PSA ≥4.0, PHI 35.0	82.7	33.0	5.9

Sensitivity reflects percentage of Gleason grade group 2-5 prostate cancers detected at each threshold. Specificity reflects percentage of men with negative biopsy or indolent cancer who would avoid biopsy at each threshold. Proportion of controls exceeding threshold reflects percentage of “cancer-free” controls who would have a false-positive test and be referred for biopsy at each threshold. Serum PSA for “cancer-free” controls is from Hybritech PHI assay at community screening. PSA for PHI Biopsy cohort is from referral PSA, which uses Beckman Coulter PSA assay in over 95% of cases.

DISCUSSION

Black men have higher rates of overall PCa compared to the general population across several PHI risk groups relative to the data that is provided by the PHI assay. In Black men, PHI is moderately accurate at detection of GG 2-5 PCa (AUC 0.63) and overall PCa (AUC 0.66, supplementary figure, https://www.jurology.com,), but is not superior to PSA in either outcome (both p >0.25). When used alone, PSA ≥4.0 ng/ml threshold only has 90.4% sensitivity, while PHI ≥28.0 has 100% sensitivity; use of PHI as an initial screening test would subject >25% of “cancer-free” controls to unnecessary biopsies. When combined with the standard PSA ≥4.0 ng/ml threshold, PHI ≥28.0 only increases specificity to 17.9% while maintaining 90.4% sensitivity.

Currently, PHI scores are reported in 4 distinct “risk categories”: 0–26.9, 27.0–35.9, 36.0–54.9, and >55.0. The PHI report lists the probability of PCa on biopsy as 9.8%, 16.8%, 33.3%, and 50.1%, respectively.⁶ Our data show that Black men have a higher prevalence of PCa within each risk category above 26.9 (table 2). In response to the higher prevalence of GG 2-5 PCa in Blacks, a lower PHI threshold in this group is needed to reduce false negatives.

PHI performed poorly in Black men compared to the external literature on GG 2-5 PCa. In a manuscript by de la Calle et al,⁸ the AUC of PHI for GG 2-5 PCa in a mostly nonBlack cohort was 0.81 in their primary cohort (561) and AUC of 0.78 in their validation cohort (395). A paper by Loeb et al showed an AUC of 0.70 for PHI with GG 2-5 PCa, but 21% of men had prior negative biopsies.¹² Another manuscript by Tosoian et al also showed an AUC of 0.77 for GG 2-5 PCa, but included men with prior biopsies.²² In Swedish biopsy naïve men, they also found a lower AUC of 0.71 for GG 2-5 PCa relative to studies including prior negative biopsies.¹⁴ In our study, the AUC for PHI was also not statistically different than the AUC for PSA (AUC 0.63 vs 0.57, p=0.27). Notably, our lower AUCs relates to the fairly modest increase in specificity of PHI in men with elevated PSA ≥4.0 ng/ml. This renders PHI somewhat less useful as a reflex test in this high risk population.

PHI density and PSA density have higher AUC and are more accurate than PSA and PHI for predicting GG 2-5 PCa (p <0.05 for both comparisons with PHI, data not shown). However, many Black men do not have access to prostate MRI²⁵ and will likely not undergo TRUS for volume measurement without a concomitant PB. Hence, PHI seems to be the most practical reflex test.

Using PSA ≥4.0 ng/ml alone as a screening test missed 9.6% of GG 2-5 PCa in Black men in the study population and spared 10.4% of men an unnecessary PB. The current standard of care in the NCCN Guidelines® recommends reflex testing using thresholds of PSA ≥4.0 ng/ml and PHI ≥35.0.²³ This missed 17.3% of GG 2-5 PCa in Black men (82.7% sensitivity), but potentially allowed 33.0% of men to avoid an unnecessary PB. To maintain sensitivity above 90%, using a threshold of PHI ≥28.0 following a threshold of referral PSA ≥4.0 ng/ml avoids unnecessary biopsy in an additional 7.5% of men compared to PSA ≥4.0 ng/ml alone (17.9%) without missing any additional cases of GG 2-5 PCa (90.4% sensitivity). We highlight this 28.0 threshold as a preliminary suggestion for PCa screening in biopsy naïve Black men as they have a higher PCa mortality rate and lower rate of insurance and may not return for a repeat PSA in a year.²⁶ Hence, increasing sensitivity with PHI ≥28.0 may be of higher import than the gains in specificity using the 35.0 threshold.

The percentage of avoided unnecessary biopsies with PHI in Black men appears to be notably lower than previously reported values in other population groups based on GG 2-5 PCa detection.¹² For instance, in a mostly nonBlack cohort, Loeb et al reported that a PHI threshold of 28.6 leads to 30% of unnecessary biopsies avoided.¹² Another paper by Chiu et al reported that, maintaining a sensitivity of 90%, PHI as a reflex test avoided unnecessary biopsies in 40% of men.¹⁶ However, in our Black biopsy patients, the specificity of PHI was only 17.9%. A PHI threshold of 27.0 has also been suggested to trigger a multiparametric magnetic resonance imaging of the prostate, but only 7.8% (25) of the entire PHI Biopsy Cohort had a PHI below 27.0.²² In centers that integrate multiparametric MRI in their practice, such low PHI thresholds might not meaningfully reduce unnecessary MRIs for Black men with elevated PSA in the diagnostic gray zone. It is likely that the high sensitivity of multiparametric MRI and moderate specificity would be a useful adjunct to PHI in helping men avoid unnecessary biopsies, but insurance constraints limit its widespread use.^27,28

As previously established, median PSA increases with age, while PHI does not need age adjustment in Blacks (supplementary table, https://www.jurology.com).²⁹

Several limitations in this study should be noted. The study population was selected for PB based on a screening PSA, which potentially introduces a selection bias. We use a PSA ≥4.0 ng/ml threshold for our basis of comparison since this is considered standard in the U.S. The PSA ≥4.0 ng/ml threshold misses 9.6% of GG 2-5 PCa in Black men.²³ The population in this study was relatively small, recruited from Chicago communities and tertiary and publicly funded medical centers, and included exclusively Black men, which limits the generalizability of our results. Our findings should be validated in a larger cohort. The controls did not undergo rectal exam or PB as they were recruited from community sites, but were followed for 1 year and all men with elevated PSA were linked to care. Also, standard TRUS biopsies were performed, which has lower sensitivity than saturation or MRI-guided biopsies, which could lead to misclassification bias.^27,28 Furthermore, because the median age of the controls was 5 years younger than men in the PBS cohort, the percentage of controls exceeding screening thresholds may be underestimated.

CONCLUSIONS

Compared to the PHI assay report for PCa prevalence by PHI risk group, Black men have a much higher overall prevalence of PCa. PHI does not appear to be more accurate than PSA for detection of GG 2-5 PCa in Black men and only avoids 7.5% more unnecessary PBs. We propose a PHI ≥28.0 cut point that maintains high sensitivity and can avoid a modest proportion of unnecessary biopsies in Black men.

Abbreviations and Acronyms

BCS

Black Citizen Scientists

DRE

digital rectal examination of the prostate

FDA

U.S. Food and Drug Administration

fPSA

free PSA

GG

Gleason grade group

MRI

magnetic resonance imaging

p2PSA

⁻²proPSA

PB

prostate biopsy

PBS

PHI Biopsy Study

PCa

prostate cancer

PHI

Prostate Health Index

PSA

prostate specific antigen

TRUS

transrectal ultrasound