Initial Prostate Specific Antigen 1.5 ng/ml or Greater in Men 50 Years Old or Younger Predicts Higher Prostate Cancer Risk

Results 

PCa by Age, Initial PSA and Followup Stratified by Race 

At ages less than 30, 30 to 39 and 40 to 50 years 0, 0 and 28 black men (0.9%), and 0, 1 (0.1%) and 39 white men (0.7%) had PCa, respectively. In black and white men with initial PSA less than 1.5 ng/ml the overall PCa rate was less than 0.4%. Of those with initial PSA between 1.5 and 2.5 ng/ml the PCa rate was greater in black than in white men but the difference was not significant (4.0% vs 2.2%, p = 0.174). At PSA 2.5 to 4.0 ng/ml black men had a lower PCa incidence than white men but again the difference was not significant (7.8% vs 9.0%, p = 0.739, table 1).

Table 1. Study population characteristics

		Black		White		p Value
		PCa	No PCa	PCa	No PCa
	No. pts (%)	28 (0.8)	3,502 (99.2)	40 (0.7)	6,078 (99.3)	0.430 (chi-squaretest)
	PSA (ng/ml):
	No. 0–0.6 (%)	7 (0.4)	1,891 (99.6)	10 (0.4)	3,355 (99.7)	0.660 (chi-squaretest)
	No. 0.7–1.4 (%)	3 (0.2)	1,261 (99.8)	7 (0.3)	2,152 (99.7)	0.754 (Fisher'sexacttest)
	No. 1.5–2.4 (%)	11 (4.0)	267 (96.0)	10 (2.2)	440 (97.8)	0.174 (chi-squaretest)
	No. 2.5–4.0 (%)	7 (7.8)	83 (92.2)	13 (9.0)	131 (91.0)	0.739 (chi-squaretest)
	Median (IQR)	0.7 (0.5–1.1)		0.7 (0.5–1.0)		0.211 (Mann-WhitneyUtest)
	Mean ± SD	0.86 ± 0.59		0.84 ± 0.57
	Age:
	No. less than 30 (%)	0	43 (100)	0	135 (100)
	No. 30–39 (%)	0	445 (100)	1 (0.1)	673 (99.9)
	No. 40–50 (%)	28 (0.9)	3,014 (99.1)	39 (0.7)	5,270 (99.3)
	Median (IQR)	45.7 (42.3–48.2)		44.6 (41.4–47.5)		<0.001 (Mann-WhitneyUtest)
	Mean ± SD	44.0 ± 4.6		44.5 ± 5.1
	Yrs to PCa diagnosis:
	No. 0.0 (Immediate diagnosis)	9 (0.5)	1,989 (99.5)	16 (0.4)	3,855 (99.6)	0.836 (chi-squaretest)
	No. 0.1–2.0 (%)	5 (0.9)	582 (99.1)	8 (0.9)	910 (99.1)	0.968 (chi-squaretest)
	No. 2.1–4.0 (%)	3 (0.7)	414 (99.3)	5 (0.9)	544 (99.1)	1.000 (Fisher'sexacttest)
	No. greater than 4.0 (%)	11 (2.1)	517 (97.9)	11 (1.4)	769 (98.6)	0.353 (chi-squaretest)
	Median (IQR)	0.0 (0.0–1.5)		0.0 (0.0–2.2)		<0.001 (Mann-WhitneyUtest)
	Mean ± SD (range)	1.4 ± 2.2 (0–9.1)		1.2 ± 2.1 (0–9.2)

Of the 28 black men with PCa the diagnosis was made in 9 at time 0.0 (immediately), in 5 at 0.1 to 2.0 years, in 3 at 2.1 to 4.0 years and in 11 at more than 4.0 years of followup. Of the 40 white men with PCa the diagnosis was made in 16 at time 0.0 (immediately), in 8 at years 0.1 to 2.0, in 5 at years 2.1 to 4.0 and in 11 at more than 4.0 years of followup (table 1). In black and white men 50 years old or younger median PSA was 0.7 ng/ml. From initial PSA measurement to diagnosis the followup 75th percentile was 1.5 (range 0.0 to 9.1) and 2.2 years (range 0.0 to 9.2) in black and white men, respectively (table 1).

PCa RR by Initial PSA and Race 

On univariate analysis the PCa RR was not significantly different between the groups with initial PSA 0 to 0.6 and 0.7 to 1.5 ng/ml in black and white men (p = 0.520 and 0.860, respectively). On multivariate analysis including initial PSA and age the age adjusted PCa RR was still not significantly different between the groups with initial PSA 0 to 0.6 and 0.7 to 1.5 ng/ml in black and white men (p = 0.510 and 0.960, respectively). However, when initial PSA increased to 1.5 to 2.5 ng/ml, the PCa RR was 11.1-fold higher in black men and 7.6-fold higher in white men than in men with initial PSA 0.6 ng/ml or less (each p <0.001). The age adjusted PCa RR was also significantly increased 9.3-fold in black men and 6.7-fold in white men (each p <0.001). In the group with initial PSA 2.5 to 4.0 ng/ml the age adjusted RR was 27.1 in black men and 16.6 in white men (each p <0.001). Table 2 lists results.

Table 2. PCa RR in men with different initial PSA by race

		Black RR (96% CI)	p Value	White RR (96% CI)	p Value
	PSA (ng/ml):
	0.0–0.6	Referent		Referent
	0.7–1.4	0.6 (0.2–2.5)	0.520	1.1 (0.4–2.9)	0.860
	1.5–2.4	11.1 (4.3–29.0)	<0.001	7.6 (3.2–18.4)	<0.001
	2.5–4.0	22.8 (7.8–66.5)	<0.001	33.3 (14.3–77.3)	<0.001
	Age adjusted PSA (ng/ml):
	0.0–0.6	Referent		Referent
	0.7–1.4	0.6 (0.2–2.4)	0.51	1.0 (0.4–2.7)	0.960
	1.5–2.4	9.3 (3.5–24.3)	<0.001	6.7 (2.7–16.2)	<0.001
	2.5–4.0	16.6 (5.6–49.0)	<0.001	27.1 (11.7–64.7)	<0.001
	Age	1.2 (1.0–1.4)	<0.001	1.1 (1.0–1.2)	<0.001

PCa by Followup PSA Less Than 1.5 vs 1.5 ng/ml or Greater 

The PCa rate increased gradually and significantly at longer followup in black and white men with baseline PSA 1.5 ng/ml or greater (p <0.001 and 0.017, respectively). Black and white men with initial PSA 1.5 ng/ml or less had a low PCa incidence at all followups. The PCa rate at followup did not change significantly in black and white men with baseline PSA 1.5 ng/ml or less (p = 0.647 and 0.323, respectively, see figure).

View Large Image Download to PowerPoint

With time PCa incidence increased significantly in black (A) and white (B) men with initial PSA 1.5 ng/ml or greater (squares) (p <0.001 and 0.017) but did not change significantly when initial PSA was less than 1.5 ng/ml (circles) (p = 0.647 and 0.323, respectively).

Discussion 

Although the PCa prevalence in men 50 years old or younger is low,7 1/5 in this age group have had a PSA test in the last year.8 Of those in the prostate center database at our institution who had a PSA test 3,612 black (8.4%) and 6,209 white (14.4%) men underwent it at or before age 50 years, an age at which we found a 1.5% and 1.0% PCa incidence, respectively (data not shown). Of them 3,530 black (97.7%) and 6,118 white (98.5%) men had baseline PSA 4 ng/ml or less. The PCa rate in these men was 0.8% and 0.7%, respectively. The clinical significance of detecting PCa in these cases is unclear. Lane et al reported that the PCa prevalence in men 50 years old or younger with threshold PSA 1.5 ng/ml was similar to that in older men with threshold PSA 3.0 ng/ml.7 Others reported that younger men have a higher long-term cancer control rate than older men after radical prostatetomy.9, 10

In our series median PSA in black and white men 50 years old or younger was 0.7 ng/ml, similar to values in previous studies.4, 5 Initial PSA 0.7 to 1.5 ng/ml did not increase the risk of PCa compared to initial PSA 0 to 0.6 ng/ml. However, initial PSA 1.5 to 2.5 ng/ml increased the age adjusted PCa RR 9.3-fold in black men and 6.7-fold in white men. Even more striking was the finding in the group with initial PSA 2.5 to 4.0 ng/ml that age adjusted PCa RR was 16.6 in black men and 27.1 in white men. These findings are in agreement with those in previous studies showing that increased initial PSA in young men is associated with an increased PCa risk.4, 5, 6, 11 Our results differ from previously published studies indicating that age specific PSA medians should be used as a cutoff for risk stratification to detect PCa.4, 5, 6 Our findings reveal that initial PSA 1.5 ng/ml or greater and not a median of greater than 0.7 ng/ml may be a better cutoff to determine which men 50 years old or younger are at higher risk for PCa in the future and who may benefit from more frequent PSA testing.

Two major differences between our study and previously published series may provide a rationale for some findings. 1) To our knowledge our cohort is the largest in the literature, and the first in which black and white men were evaluated independently. 2) The initial PSA range in our series only included PSA 4.0 ng/ml or less but other studies included PSA 4.0 ng/ml or greater.

In our cohort of men 50 years old or younger with initial PSA 4.0 ng/ml or less the incidence of PCa in black and white men with initial PSA less than 1.5 ng/ml was always less than 0.6% even at up to 9 years of followup. In contrast, the PCa rate in men with initial PSA 1.5 ng/ml or greater was significantly higher than that in men with PSA less than 1.5 ng/ml and it increased with time. At greater than 4-year followup (maximum 9) the PCa incidence in the group with initial PSA greater than 1.5 ng/ml was almost 18.0% in black men and 9.5% in white men (see figure). These findings indicate that PSA 1.5 ng/ml is a better cutoff to stratify PCa risk in men 50 years old or younger.

Results of the Rotterdam section of the European Randomized Study of Screening for Prostate Cancer show that the rate of detected cancer at 8 years of followup with 2 subsequent screening visits was 0.47% in men with PSA 1.0 ng/ml or less, representing 42% of screened men in the 55 to 65-year age range.12 The investigators suggested that PSA screening every 8 years in men with PSA 1.0 ng/ml or less would lead to a considerable decrease in the number of screening visits, and associated costs and stress, and the strategy would carry minimal risk for missing aggressive cancer at a curable stage.12 Loeb et al reported that men 40 to 50 years old with PSA less than the age specific median were at low risk for PCa detection in the short term.6

Initial PSA less than 1.5 ng/ml was associated with a low PCa incidence at different followups in our series, in agreement with previous results. Higher initial PSA also predicted a higher PCa incidence with time in our study. This further confirms that initial PSA 1.5 ng/ml or greater rather than a median of greater than 0.7 ng/ml may be a better cutoff to determine which men 50 years old or younger are at higher risk for PCa in the future. However, prospective studies with long-term followup are warranted.

The biological basis for the relationship between initial PSA and PCa risk remains unclear but theories exist on the cause. 1) Since PCa is generally a gradual process, men with higher initial PSA may be more likely to harbor PCa at an early stage or as a premalignant lesion that can progress and become clinically apparent with time. 2) High grade PIN is 6-fold more common in the biopsies of men with high PSA. Since high grade PIN is considered precancerous, some groups suggest that it may induce higher PSA.4, 13 3) Prostatic inflammation can lead to increased PSA and in young men may result in a higher risk of subsequent PCa.14 4) Insulin-like growth factor-1, and the androgen related genes CYP17, CYP3A4 and SRD5A2 are associated with PCa risk.15, 16

Our series has several limitations. It is retrospective. We did not evaluate PCa family history. We used data from a large tertiary medical center. This is partly counteracted by our cohort size, which may mitigate the limitation. We did not evaluate digital rectal examination results. However, European Randomized Study of Screening for Prostate Cancer results indicate that the overall characteristics of detected PCa cases differ little based on how they are detected, that is by increased PSA, digital rectal examination or transrectal ultrasound.17 Also, our findings may not be applicable to all racial groups since we only included black and white men in our analysis.

Conclusions 

In men 50 years old or younger low initial PSA is associated with a subsequent low PCa rate. Initial PSA less than 1.5 ng/ml was associated with a low PCa rate even at up to 9 years of followup. In contrast, men with initial PSA 1.5 ng/ml or greater were at significantly higher risk for PCa and this incidence increased with time. Based on our findings baseline PSA 1.5 ng/ml is a better risk stratification cutoff than age adjusted median PSA 0.7 ng/ml in black and white men 50 years old or younger. This recommendation is based on 28 and 40 events in black and white men, respectively, in a large study cohort with up to 9 years of followup. Further studies are warranted.