- Email: [email protected]
mL
Adult Urology Prostate Cancer Risk with Positive Family History, Normal Prostate Examination Findings, and PSA Less Than 4.0 ng/mL Edith Canby-Hagino, Javier Hernandez, Timothy C. Brand, Dean A. Troyer, Betsy Higgins, Donna Pauler Ankerst, Ian M. Thompson, Robin J. Leach, and Dipen J. Parekh OBJECTIVES
METHODS
RESULTS
CONCLUSIONS
Family history has been identified as a risk factor for prostate cancer. We assessed the risk of prostate cancer in men with a positive family history (at least one first-degree or second-degree relative), normal digital rectal examination (DRE) and a serum prostate-specific antigen (PSA) level of 4.0 ng/mL. A total of 87 volunteers from the San Antonio Center of biomarkers of risk for prostate cancer study enrolled in a prospective study to assess the prevalence of prostate cancer in men with a family history of prostate cancer, normal DRE findings, and a serum PSA level of 4.0 ng/mL or less. All subjects underwent transrectal ultrasound-guided prostate biopsy. Prostate cancer was diagnosed in 22 (25.3%) of the 87 participants. The PSA values were significantly greater statistically in the men with prostate cancer (median 2.1 ng/mL) than in the men without prostate cancer (median 1.2 ng/mL, P ⫽ 0.01), and the odds of prostate cancer nearly doubled for a doubling of the PSA level within this interval (odds ratio 1.97, P ⫽ 0.02). No statistically significant difference was found in age, race, or number and type of affected relatives in men with and without prostate cancer. The results of our study have demonstrated a high frequency of prostate cancer in men with a positive family history but normal DRE findings and a PSA level of 4.0 ng/mL or less. These prospectively collected data raise an important consideration regarding lowering the PSA cutoff values for offering biopsy in patients with a positive family history and normal DRE findings. UROLOGY 70: 748 –752, 2007. © 2007 Elsevier Inc.
P
rostate cancer is the most common visceral malignancy and second leading cause of cancer death among men in the United States, accounting for an estimated 234,460 new cases and 27,350 deaths in 2006.1 Efforts to improve early detection have focused on the development of biomarkers, as well as the identification of other factors that place men at increased risk of the disease. Although others are in development, serum prostatespecific antigen (PSA) is the most widely used biomarker worldwide for the screening and early detection of prostate cancer. In general, increased PSA levels correlate with both increased risk and increased risk of higher
This study was supported by the Clinical and Epidemiologic and Validation Center of the Early Detection Research Network (National Institutes of Health grant U01CA086402), the American Urological Association Foundation, and the San Antonio Cancer Institute. D. J. Parekh is supported by the University of Texas Health Science Center at San Antonio, San Antonio, Texas Institute for Integration of Medicine and Science Mentored Career Development Award. From the Departments of Urology and Cellular and Structural Biology, University of Texas Health Science Center at San Antonio; and Division of Urology, Brooke Army Medical Center, San Antonio, Texas Reprint requests: Dipen J. Parekh, M.D., Department of Urology, c, 7703 Floyd Curl Drive, San Antonio, TX 78229. E-mail: [email protected] Submitted: February 1, 2007, accepted (with revisions): June 26, 2007
748
© 2007 Elsevier Inc. All Rights Reserved
grade and stage of prostate cancer.2– 4 A quantifiable risk of prostate cancer exists at any level of PSA, making it impossible to establish a cutoff for PSA below which the risk of prostate cancer is negligible.5 An alternative strategy for early detection could include a combination of biomarkers with other commonly cited factors associated with an increased risk of prostate cancer, such as prostate examination, age, race, and family history. A family history of the disease has long been considered an important risk factor for the development of prostate cancer. Twin studies have indicated that 42% to 57% of the risk of prostate cancer is the result of heritable factors.6 – 8 However, only 5% to 10% of prostate cancers can be attributed to highly penetrant autosomal dominant genes that exhibit mendelian inheritance.9,10 A positive family history might accompany other factors that give rise to clinical suspicion for prostate cancer, namely abnormal prostate examination findings or an elevated PSA level. For this reason, and because prostate biopsies are typically not recommended in patients with a PSA level less than 4.0 ng/mL and a normal digital rectal examination (DRE), it is difficult to quantify the contribution of family history to prostate cancer risk. 0090-4295/07/$32.00 doi:10.1016/j.urology.2007.06.1105
To evaluate the exclusive contribution of family history to prostate cancer risk, we conducted a prospective study of men with a family history of prostate cancer, a PSA level less than 4.0 ng/mL, and normal DRE findings by offering and performing transrectal ultrasound-guided prostate biopsies.
MATERIAL AND METHODS Study Subjects The study subjects were derived from the prospective cohort of male participants in the San Antonio Center of Biomarkers of Risk of Prostate Cancer (SABOR) study. The institutional review board approved both the SABOR study and the present study. The eligibility criteria were as follows: a family history of prostate cancer (at least one first-degree or second-degree relative), PSA level of 4.0 ng/mL or less, normal DRE findings, no previous history of prostate cancer, and willing and able to undergo prostate biopsy. The eligibility criteria were designed to identify a study population that did not carry clinical suspicion for prostate cancer other than a positive family history. SABOR is 10-year prospective cohort study with a current enrollment of 3575 men and is one of three prostate cancer studies of the Clinical and Epidemiologic and Validation Centers of the Early Detection Research Network, National Cancer Institute. Participation included collection of biologic specimens for genetic, protein, and trace mineral profiling, and extensive demographic and dietary history, medical symptoms, including American Urological Association symptom scores, and information about prostate-related medications and procedures. In addition, participants were queried annually about a family history for prostate cancer. DREs and PSA measurements were performed annually. Before June 2005, PSA measurements were done using Bayer-Cendaur Chemoluminescent assay. After June 2005, the PSA measurements were done using the Beckman ICMA method. The SABOR participants with a family history of a firstdegree or second-degree relative with prostate cancer were identified during their annual screening visits. First-degree relatives were defined as a father, brother, or son. Second-degree relatives were defined as a grandfather or uncle. Second-degree relatives were included to capture participants whose susceptibility to prostate cancer might have been passed through maternal lineage. The study subjects meeting the family history and other eligibility criteria were offered participation and provided informed consent to undergo additional evaluation with transrectal ultrasound-guided prostate biopsies.
Prostate Ultrasound Scan and Biopsy After antibiotic prophylaxis with oral fluoroquinolones for 1 day before biopsy, the subjects underwent transrectal ultrasonography of the prostate. Ten to twelve biopsy cores were taken from the prostate using a spring-loaded core biopsy needle. Specimens from the right and left sides of the prostate were submitted in separate containers and frozen immediately using the technique described by Scott et al.11 A single pathologist (D.T.) examined all biopsy specimens and provided Gleason grading and scoring of any prostate cancer. The timeframe for the biopsies in the present study was August 2003 to March 2006. All subjects also received antibiotic prophylaxis with oral fluoroquinolones for 2 days after their biopsy. UROLOGY 70 (4), 2007
Table 1. Participant characteristics Characteristic
Prostate Cancer Yes No
Men (n) 22 (25.3) 65 (74.7) Age (yr) Median 60.9 57.8 Range 50.1–80.6 39.6–77.5 Race/ethnicity (n) Non-Hispanic white 14 (63.6) 50 (76.9) Hispanic white 6 (27.3) 7 (10.8) African American 2 (9.1) 8 (12.3) PSA (ng/mL) Median 2.1 1.2 Range 0.3–3.7 0.2–4.0 Previous biopsy (n) 0 20 (90.9) 60 (92.3) ⱖ1 2 (9.1) 5 (7.7) Affected relatives per participant (n) Median 1 1 Range 1–5 1–4 Men with affected 20 (90.9) 55 (84.6) first-degree relative (n) Men with affected 6 (27.3) 22 (34.4) brother (n)
P Value 0.12 0.17
0.01 0.81 0.99
0.46 0.38
PSA ⫽ prostate-specific antigen. Data presented as number of patients, with percentages in parentheses, unless otherwise noted.
Statistical Analysis The age and PSA level in subjects with and without prostate cancer were compared with two-sample t tests. Chi-square analysis was used to compare race/ethnicity and previous biopsy between the two groups. Logistic regression analysis was used to assess the predictive strength of PSA level for biopsy outcome. All tests were performed at the two-sided 0.05 significance level. Data were analyzed using Stata, release 9, software.
RESULTS The characteristics of the 87 study participants are listed in Table 1. Prostate cancer was diagnosed in 22 of the subjects (25.3%). No statistically significant difference was found in age, race/ethnicity, or a history of previous negative biopsy findings between patients diagnosed with prostate cancer and participants with negative prostate biopsy findings (P ⬎0.05 for all). In addition, a history of a first-degree relative with prostate cancer was no more likely among men with positive biopsy findings than among men with negative biopsy findings. Finally, a history of a brother with prostate cancer was not a significant predictor of prostate cancer compared with any family history of prostate cancer. Even within the narrow range of 0 to 4.0 ng/mL for these participants, the PSA level was significantly different between men with cancer (median 2.1 ng/mL, range 0.3 to 3.7) and without cancer (median 1.2 ng/mL, range 0.2 to 4.0, P ⫽ 0.01). The risk of prostate cancer increased with increasing PSA level (Table 2), exhibiting the trend observed in the Prostate Cancer Prevention 749
Table 2. Relationship of PSA to prostate cancer prevalence
PSA Level (ng/mL)
Total (n ⫽ 87)
Men with Prostate Cancer (n ⫽ 22)
ⱕ0.5 0.5–1.0 1.0–2.0 2.0–3.0 3.0–4.0
10 23 28 17 9
2 (20.0) 3 (13.0) 5 (17.9) 5 (29.4) 7 (77.8)
PSA ⫽ prostate-specific antigen. Data in parentheses are percentages.
Trial (PCPT), albeit with our sample size too small to confirm a reproduction.5 The odds of prostate cancer nearly doubled for a doubling of PSA within this interval (odds ratio 1.97, 95% confidence interval 1.11 to 3.52, P ⫽ 0.02). Of the 22 men with prostate cancer, 16 (72.7%) had Gleason grade 6 and 6 (27.3%) Gleason grade 7. The small number of high-grade (Gleason grade 7 or greater) cancers prohibited the assessment of a trend with PSA level.
COMMENT A family history of prostate cancer represents a complex and poorly understood amalgamation of both genetic and environmental factors. Exploration for new biomarkers has led to the identification of many inherited genetic polymorphisms associated with the risk and prognosis of prostate cancer. The identification of individual polymorphisms has not adequately explained the contribution of family history to prostate cancer risk, because the identified genetic variants might be innocuous constituents in a haplotype that confers increased risk, and many genetic variants do not confer the same level of risk across populations that vary by race or geography.12,13 Most studies linking prostate cancer to a family history of the disease have been case-control studies.14,15 Several additional cohort and meta-analysis studies have also confirmed the association between an increased risk of prostate cancer and a family history of the disease. The reported increased odds of prostate cancer have ranged from 2.2 to 5.6 for relatives of men with a history of prostate cancer compared with relatives of those without.16 –22 The risk also appears to be increased for family members of men with an earlier age of diagnosis, cancer in a brother instead of another relative, and cancer in more than one family member.16 –20,22 The collection and verification of family history data can also be complex. The strength of the association between prostate cancer and family history is affected by recall bias. The study structure in which the association is made, as well as the manner in which the family history and disease status are ascertained, can either reduce or enhance these biases. Case-control studies are prone to recall bias, as are retrospective studies. Ascertainment bias can also confound case-control studies. Studies have 750
demonstrated that men with a positive family history are more likely to perceive vulnerability to developing prostate cancer, resulting in increased intention and actual pursuit of screening for prostate cancer compared with men with no family history of prostate cancer.23,24 In addition, primary care providers recognize family history as a risk factor for prostate cancer and are more likely to offer prostate cancer screening to men with a family history than to men with no family history.25,26 To minimize the effect of recall and ascertainment biases, we conducted a prospective cohort study to evaluate the contribution of family history to prostate cancer development. Our results were derived from family histories that were all collected prospectively, before enrollment in the study, and before prostate biopsy or a diagnosis of cancer, minimizing the effect of the abovementioned biases. Another potential confounder of cancer ascertainment when studying the association between family history and prostate cancer risk is that factors other than family history usually prompt prostate biopsies, such as an elevated PSA level or abnormal DRE findings. Case-control studies investigating the contribution of family history to prostate cancer risk have generally matched cases and controls by age and locality, but not by PSA level or DRE findings. Because factors other than family history have led to the diagnosis of prostate cancer, the contribution of family history to prostate cancer risk might have been misrepresented. Our study attempted to isolate the contribution of family history to prostate cancer risk by excluding factors other than family history that might lead to a cancer diagnosis by limiting the population to men with normal DRE findings and a PSA level of 4.0 ng/mL or less. One limitation of our study was that we did not have an adequate control group of men who had undergone biopsy and had a PSA level of 4.0 ng/mL or less and normal DRE findings, but no family history, in the SABOR cohort with which to directly benchmark the high prostate cancer prevalence in this group. For the most part, these men had not been offered a prostate biopsy because they had no clinical indication (PSA level greater than 4.0 ng/mL or abnormal DRE findings). However, the prevalence rate in this study, 25.3%, exceeded that (21.9%) reported for a large group of 5519 PCPT placebo participants reported by Thompson et al.27 that included both positive and negative family history participants, normal and abnormal DRE results, and a range of PSA values. Thompson et al.27 reported a statistically significant increase in prostate cancer risk with a positive family history. Our study results have indicated that a positive family history plays a role as an independent risk factor for prostate cancer in men with a PSA level less than 4.0 ng/mL and normal DRE findings. Our findings have confirmed that no PSA level exists below which the risk of prostate cancer is negligible, although even within the narrow range of 0 to 4.0 ng/mL, UROLOGY 70 (4), 2007
an increasing PSA level correlated with an increasing risk of cancer. Of interest is the unexpectedly high frequency of cancer in these men, who might not otherwise have been offered a prostate biopsy. The placebo group in the PCPT has yielded the largest body of prospective data on the predictive value and performance characteristics of PSA in the 0 to 4.0 ng/mL range.2,5 The PCPT identified family history as a significant risk factor for prostate cancer at PSA levels within this range. Among the men who underwent end-of-study biopsy, 477 (16.2%) reported a positive family history for prostate cancer, and 94 men (19.7%) within this subgroup were diagnosed with cancer compared with 355 (14.4%) of the 2473 men who reported a negative family history for prostate cancer. In our study, 25.3% of men with a PSA level in this range had prostate cancer. The possible explanations for this discrepancy include the smaller sample size of our study and the increased number of biopsy cores taken in the present study (10 to 12 versus a mean number of biopsy cores per subject of 6 the PCPT). The most recent analysis from the PCPT has led to the development of a risk calculator for the determination of prostate cancer risk at prostate biopsy. In a logistic regression analysis conducted in 5519 men from the placebo group of PCPT, all of whom underwent prostate biopsy, the risk factors that predicted for a risk of prostate cancer and/or high-grade disease included PSA level, age, race/ethnicity, family history of prostate cancer, and history of previous negative prostate biopsy. The investigators found a significant contribution of the risk of a positive biopsy from a family history of prostate cancer (odds ratio 1.31, 95% confidence interval 1.11 to 1.55, P ⫽ 0.002). The risk calculator is available online at http:// www.compass.fhcrc.org/edrnnci/bin/calculator/main.asp. The present study has confirmed these associations of risk and the interactions of PSA and family history of prostate cancer for the risk of the disease. It has been proposed that lower PSA cutoffs be recommended for men with a positive family history to equalize PSA-based screening such that all men of equal risk of cancer be afforded the opportunity for enhanced prostate cancer screening by elective prostate biopsy.27 In men with normal DRE findings and no previous biopsy, to obtain a 25% risk of prostate cancer, the PCPT risk calculator would suggest a PSA cutoff of 2.3 ng/mL for men with no family history, but 1.7 ng/mL for men with a positive family history.27 Although lowering the PSA cutoffs for offering prostate biopsy is in routine clinical practice in some centers,28 one of the criticisms of adopting such a strategy is the risk of an increased incidence of indolent cancer. However, recent studies have suggested a reduced risk of mortality associated with active treatment for low-risk and intermediate-risk prostate cancer.29,30 If PSA cutoffs are going to continue to be the standard for referral to biopsy, our study has confirmed UROLOGY 70 (4), 2007
that lowering the PSA cutoff for men with a positive family history is justified.
CONCLUSIONS The results of this prospective evaluation of a family history of prostate cancer as a predictor of prostate cancer in men who would not normally be offered a prostate biopsy are sobering. Our findings have reiterated that no PSA level exists below which the risk of prostate cancer is negligible and have confirmed that a positive family history by itself is a strong predictor of prostate cancer in the absence of other clinical indications for a prostate biopsy. Although we refrain from recommending prostate biopsy to all men with a positive family history, it could be argued that the PSA cutoff values for offering prostate biopsy should be lowered because of the high prevalence rate of prostate cancer in this cohort. At present, although patients and their urologists now have a risk calculator to estimate the risk of prostate cancer and aggressive disease if a biopsy is performed, the tools to predict the natural history of many prostate cancers are not available. Continued accumulation of prospectively collected data with full ascertainment of cancer status will enable construction of multifactorial risk calculators, so that patients and urologists can identify specific and individualized risks and match these against their level of acceptance or aversion to such risk. Acknowledgment. To Rosario Mendez-Meza for technical assistance and John Schoolfield for assistance with data analysis. References 1. Jemal A, Siegel R, Ward E, et al: Cancer statistics, 2006. CA Cancer J Clin 56: 106 –130, 2006. 2. Thompson IM, Ankerst DP, Chi C, et al: Operating characteristics of prostate-specific antigen in men with an initial PSA level of 3.0 ng/ml or lower. JAMA 294: 66 –70, 2005. 3. Kupelian P, Katcher J, Levin H, et al: Correlation of clinical and pathologic factors with rising prostate-specific antigen profiles after radical prostatectomy alone for clinically localized prostate cancer. Urology 48: 249 –260, 1996. 4. Antenor JA, Roehl KA, Eggener SE, et al: Preoperative PSA and progression-free survival after radical prostatectomy for stage T1c disease. Urology 66: 156 –160, 2005. 5. Thompson IM, Pauler DK, Goodman PJ, et al: Prevalence of prostate cancer among men with a prostate-specific antigen level ⬍ or ⫽ 4.0 ng per milliliter. [Erratum appears in N Engl J Med 351: 1470, 2004]. N Engl J Med 350: 2239 –2246, 2004. 6. Lichtenstein P, Holm NV, Verkasalo PK, et al: Environmental and heritable factors in the causation of cancer—analyses of cohorts of twins from Sweden, Denmark, and Finland. N Engl J Med 343: 78 – 85, 2000. 7. Page WF, Braun MM, Partin AW, et al: Heredity and prostate cancer: a study of World War II veteran twins. Prostate 33: 240 –245, 1997. 8. Verkasalo PK, Kaprio J, Koskenvuo M, et al: Genetic predisposition, environment and cancer incidence: a nationwide twin study in Finland, 1976 –1995. Int J Cancer 83: 743–749, 1999. 9. Karan D, Lin MF, Johansson SL, et al: Current status of the molecular genetics of human prostatic adenocarcinomas. Int J Cancer 103: 285–293, 2003.
751
10. Bratt O: Hereditary prostate cancer: clinical aspects. J Urol 168: 906 –913, 2002. 11. Scott KM, Fanta P, Calaluce R, et al: Diagnostic frozen prostate sextant biopsies: an approach for preserving protein and RNA for additional studies. Prostate 44: 296 –302, 2000. 12. Balic I, Graham ST, Troyer DA, et al: Androgen receptor length polymorphism associated with prostate cancer risk in Hispanic men. J Urol 168: 2245–2248, 2002. 13. Hernandez J, Balic I, Johnson-Pais TL, et al: Association between an estrogen receptor alpha gene polymorphism and the risk of prostate cancer in black men. J Urol 175: 523–527, 2006. 14. Morganti G, Gianferrari L, Cresseri A, et al: [Clinico-statistical and genetic research on neoplasms of the prostate.] Acta Genet Stat Med 6: 304 –305, 1956. 15. Woolf CM: An investigation of the familial aspects of carcinoma of the prostate. Cancer 13: 739 –744, 1960. 16. Bruner DW, Moore D, Parlanti A, et al: Relative risk of prostate cancer for men with affected relatives: systematic review and metaanalysis. Int J Cancer 107: 797– 803, 2003. 17. Johns LE, and Houlston RS: A systematic review and meta-analysis of familial prostate cancer risk. BJU Int 91: 789 –794, 2003. 18. Hayes RB, Liff JM, Pottern LM, et al: Prostate cancer risk in U.S. blacks and whites with a family history of cancer. Int J Cancer 60: 361–364, 1995. 19. Zeegers MP, Jellema A, and Ostrer H: Empiric risk of prostate carcinoma for relatives of patients with prostate carcinoma: a meta-analysis. Cancer 97: 1894 –1903, 2003. 20. Spitz MR, Currier RD, Fueger JJ, et al: Familial patterns of prostate cancer: a case-control analysis. J Urol 146: 1305–1307, 1991.
752
21. Schuurman AG, Zeegers MP, Goldbohm RA, et al: A case-cohort study on prostate cancer risk in relation to family history of prostate cancer. Epidemiology 10: 192–195, 1999. 22. Cerhan JR, Parker AS, Putnam SD, et al: Family history and prostate cancer risk in a population-based cohort of Iowa men. Cancer Epidemiol Biomarkers Prev 8: 53– 60, 1999. 23. Jacobsen PB, Lamonde LA, Honour M, et al: Relation of family history of prostate cancer to perceived vulnerability and screening behavior. Psycho-Oncology 13: 80 – 85, 2004. 24. Bock CH, Peyser PA, Gruber SB, et al: Prostate cancer early detection practices among men with a family history of disease. Urology 62: 470 – 475, 2003. 25. Durham J, Low M, and McLeod D: Screening for prostate cancer: a survey of New Zealand general practitioners. N Z Med J 116: U476, 2003. 26. Tudiver F, Guibert R, Haggerty J, et al: What influences family physicians’ cancer screening decisions when practice guidelines are unclear or conflicting? J Fam Pract 51: 760, 2002. 27. Thompson IM, Ankerst DP, Chi C, et al: Assessing prostate cancer risk: results from the Prostate Cancer Prevention Trial. J Natl Cancer Inst 98: 529 –534, 2006. 28. Krumholtz JS, Carvalhal GF, Ramos CG, et al: Prostate-specific antigen cutoff of 2.6 ng/mL for prostate cancer screening is associated with favorable pathologic tumor features. Urology 60: 469 – 474, 2002. 29. Wong YN, Mitra N, Hudes G, et al: Survival associated with treatment vs observation of localized prostate cancer in elderly men. JAMA 296: 2683–2693, 2006. 30. Holmberg L, Bill-Axelson A, Garmo H, et al: Prognostic markers under watchful waiting and radical prostatectomy. Hematol Oncol Clin North Am 20: 845– 855, 2006.
UROLOGY 70 (4), 2007
METHODS
RESULTS
CONCLUSIONS
Family history has been identified as a risk factor for prostate cancer. We assessed the risk of prostate cancer in men with a positive family history (at least one first-degree or second-degree relative), normal digital rectal examination (DRE) and a serum prostate-specific antigen (PSA) level of 4.0 ng/mL. A total of 87 volunteers from the San Antonio Center of biomarkers of risk for prostate cancer study enrolled in a prospective study to assess the prevalence of prostate cancer in men with a family history of prostate cancer, normal DRE findings, and a serum PSA level of 4.0 ng/mL or less. All subjects underwent transrectal ultrasound-guided prostate biopsy. Prostate cancer was diagnosed in 22 (25.3%) of the 87 participants. The PSA values were significantly greater statistically in the men with prostate cancer (median 2.1 ng/mL) than in the men without prostate cancer (median 1.2 ng/mL, P ⫽ 0.01), and the odds of prostate cancer nearly doubled for a doubling of the PSA level within this interval (odds ratio 1.97, P ⫽ 0.02). No statistically significant difference was found in age, race, or number and type of affected relatives in men with and without prostate cancer. The results of our study have demonstrated a high frequency of prostate cancer in men with a positive family history but normal DRE findings and a PSA level of 4.0 ng/mL or less. These prospectively collected data raise an important consideration regarding lowering the PSA cutoff values for offering biopsy in patients with a positive family history and normal DRE findings. UROLOGY 70: 748 –752, 2007. © 2007 Elsevier Inc.
P
rostate cancer is the most common visceral malignancy and second leading cause of cancer death among men in the United States, accounting for an estimated 234,460 new cases and 27,350 deaths in 2006.1 Efforts to improve early detection have focused on the development of biomarkers, as well as the identification of other factors that place men at increased risk of the disease. Although others are in development, serum prostatespecific antigen (PSA) is the most widely used biomarker worldwide for the screening and early detection of prostate cancer. In general, increased PSA levels correlate with both increased risk and increased risk of higher
This study was supported by the Clinical and Epidemiologic and Validation Center of the Early Detection Research Network (National Institutes of Health grant U01CA086402), the American Urological Association Foundation, and the San Antonio Cancer Institute. D. J. Parekh is supported by the University of Texas Health Science Center at San Antonio, San Antonio, Texas Institute for Integration of Medicine and Science Mentored Career Development Award. From the Departments of Urology and Cellular and Structural Biology, University of Texas Health Science Center at San Antonio; and Division of Urology, Brooke Army Medical Center, San Antonio, Texas Reprint requests: Dipen J. Parekh, M.D., Department of Urology, c, 7703 Floyd Curl Drive, San Antonio, TX 78229. E-mail: [email protected] Submitted: February 1, 2007, accepted (with revisions): June 26, 2007
748
© 2007 Elsevier Inc. All Rights Reserved
grade and stage of prostate cancer.2– 4 A quantifiable risk of prostate cancer exists at any level of PSA, making it impossible to establish a cutoff for PSA below which the risk of prostate cancer is negligible.5 An alternative strategy for early detection could include a combination of biomarkers with other commonly cited factors associated with an increased risk of prostate cancer, such as prostate examination, age, race, and family history. A family history of the disease has long been considered an important risk factor for the development of prostate cancer. Twin studies have indicated that 42% to 57% of the risk of prostate cancer is the result of heritable factors.6 – 8 However, only 5% to 10% of prostate cancers can be attributed to highly penetrant autosomal dominant genes that exhibit mendelian inheritance.9,10 A positive family history might accompany other factors that give rise to clinical suspicion for prostate cancer, namely abnormal prostate examination findings or an elevated PSA level. For this reason, and because prostate biopsies are typically not recommended in patients with a PSA level less than 4.0 ng/mL and a normal digital rectal examination (DRE), it is difficult to quantify the contribution of family history to prostate cancer risk. 0090-4295/07/$32.00 doi:10.1016/j.urology.2007.06.1105
To evaluate the exclusive contribution of family history to prostate cancer risk, we conducted a prospective study of men with a family history of prostate cancer, a PSA level less than 4.0 ng/mL, and normal DRE findings by offering and performing transrectal ultrasound-guided prostate biopsies.
MATERIAL AND METHODS Study Subjects The study subjects were derived from the prospective cohort of male participants in the San Antonio Center of Biomarkers of Risk of Prostate Cancer (SABOR) study. The institutional review board approved both the SABOR study and the present study. The eligibility criteria were as follows: a family history of prostate cancer (at least one first-degree or second-degree relative), PSA level of 4.0 ng/mL or less, normal DRE findings, no previous history of prostate cancer, and willing and able to undergo prostate biopsy. The eligibility criteria were designed to identify a study population that did not carry clinical suspicion for prostate cancer other than a positive family history. SABOR is 10-year prospective cohort study with a current enrollment of 3575 men and is one of three prostate cancer studies of the Clinical and Epidemiologic and Validation Centers of the Early Detection Research Network, National Cancer Institute. Participation included collection of biologic specimens for genetic, protein, and trace mineral profiling, and extensive demographic and dietary history, medical symptoms, including American Urological Association symptom scores, and information about prostate-related medications and procedures. In addition, participants were queried annually about a family history for prostate cancer. DREs and PSA measurements were performed annually. Before June 2005, PSA measurements were done using Bayer-Cendaur Chemoluminescent assay. After June 2005, the PSA measurements were done using the Beckman ICMA method. The SABOR participants with a family history of a firstdegree or second-degree relative with prostate cancer were identified during their annual screening visits. First-degree relatives were defined as a father, brother, or son. Second-degree relatives were defined as a grandfather or uncle. Second-degree relatives were included to capture participants whose susceptibility to prostate cancer might have been passed through maternal lineage. The study subjects meeting the family history and other eligibility criteria were offered participation and provided informed consent to undergo additional evaluation with transrectal ultrasound-guided prostate biopsies.
Prostate Ultrasound Scan and Biopsy After antibiotic prophylaxis with oral fluoroquinolones for 1 day before biopsy, the subjects underwent transrectal ultrasonography of the prostate. Ten to twelve biopsy cores were taken from the prostate using a spring-loaded core biopsy needle. Specimens from the right and left sides of the prostate were submitted in separate containers and frozen immediately using the technique described by Scott et al.11 A single pathologist (D.T.) examined all biopsy specimens and provided Gleason grading and scoring of any prostate cancer. The timeframe for the biopsies in the present study was August 2003 to March 2006. All subjects also received antibiotic prophylaxis with oral fluoroquinolones for 2 days after their biopsy. UROLOGY 70 (4), 2007
Table 1. Participant characteristics Characteristic
Prostate Cancer Yes No
Men (n) 22 (25.3) 65 (74.7) Age (yr) Median 60.9 57.8 Range 50.1–80.6 39.6–77.5 Race/ethnicity (n) Non-Hispanic white 14 (63.6) 50 (76.9) Hispanic white 6 (27.3) 7 (10.8) African American 2 (9.1) 8 (12.3) PSA (ng/mL) Median 2.1 1.2 Range 0.3–3.7 0.2–4.0 Previous biopsy (n) 0 20 (90.9) 60 (92.3) ⱖ1 2 (9.1) 5 (7.7) Affected relatives per participant (n) Median 1 1 Range 1–5 1–4 Men with affected 20 (90.9) 55 (84.6) first-degree relative (n) Men with affected 6 (27.3) 22 (34.4) brother (n)
P Value 0.12 0.17
0.01 0.81 0.99
0.46 0.38
PSA ⫽ prostate-specific antigen. Data presented as number of patients, with percentages in parentheses, unless otherwise noted.
Statistical Analysis The age and PSA level in subjects with and without prostate cancer were compared with two-sample t tests. Chi-square analysis was used to compare race/ethnicity and previous biopsy between the two groups. Logistic regression analysis was used to assess the predictive strength of PSA level for biopsy outcome. All tests were performed at the two-sided 0.05 significance level. Data were analyzed using Stata, release 9, software.
RESULTS The characteristics of the 87 study participants are listed in Table 1. Prostate cancer was diagnosed in 22 of the subjects (25.3%). No statistically significant difference was found in age, race/ethnicity, or a history of previous negative biopsy findings between patients diagnosed with prostate cancer and participants with negative prostate biopsy findings (P ⬎0.05 for all). In addition, a history of a first-degree relative with prostate cancer was no more likely among men with positive biopsy findings than among men with negative biopsy findings. Finally, a history of a brother with prostate cancer was not a significant predictor of prostate cancer compared with any family history of prostate cancer. Even within the narrow range of 0 to 4.0 ng/mL for these participants, the PSA level was significantly different between men with cancer (median 2.1 ng/mL, range 0.3 to 3.7) and without cancer (median 1.2 ng/mL, range 0.2 to 4.0, P ⫽ 0.01). The risk of prostate cancer increased with increasing PSA level (Table 2), exhibiting the trend observed in the Prostate Cancer Prevention 749
Table 2. Relationship of PSA to prostate cancer prevalence
PSA Level (ng/mL)
Total (n ⫽ 87)
Men with Prostate Cancer (n ⫽ 22)
ⱕ0.5 0.5–1.0 1.0–2.0 2.0–3.0 3.0–4.0
10 23 28 17 9
2 (20.0) 3 (13.0) 5 (17.9) 5 (29.4) 7 (77.8)
PSA ⫽ prostate-specific antigen. Data in parentheses are percentages.
Trial (PCPT), albeit with our sample size too small to confirm a reproduction.5 The odds of prostate cancer nearly doubled for a doubling of PSA within this interval (odds ratio 1.97, 95% confidence interval 1.11 to 3.52, P ⫽ 0.02). Of the 22 men with prostate cancer, 16 (72.7%) had Gleason grade 6 and 6 (27.3%) Gleason grade 7. The small number of high-grade (Gleason grade 7 or greater) cancers prohibited the assessment of a trend with PSA level.
COMMENT A family history of prostate cancer represents a complex and poorly understood amalgamation of both genetic and environmental factors. Exploration for new biomarkers has led to the identification of many inherited genetic polymorphisms associated with the risk and prognosis of prostate cancer. The identification of individual polymorphisms has not adequately explained the contribution of family history to prostate cancer risk, because the identified genetic variants might be innocuous constituents in a haplotype that confers increased risk, and many genetic variants do not confer the same level of risk across populations that vary by race or geography.12,13 Most studies linking prostate cancer to a family history of the disease have been case-control studies.14,15 Several additional cohort and meta-analysis studies have also confirmed the association between an increased risk of prostate cancer and a family history of the disease. The reported increased odds of prostate cancer have ranged from 2.2 to 5.6 for relatives of men with a history of prostate cancer compared with relatives of those without.16 –22 The risk also appears to be increased for family members of men with an earlier age of diagnosis, cancer in a brother instead of another relative, and cancer in more than one family member.16 –20,22 The collection and verification of family history data can also be complex. The strength of the association between prostate cancer and family history is affected by recall bias. The study structure in which the association is made, as well as the manner in which the family history and disease status are ascertained, can either reduce or enhance these biases. Case-control studies are prone to recall bias, as are retrospective studies. Ascertainment bias can also confound case-control studies. Studies have 750
demonstrated that men with a positive family history are more likely to perceive vulnerability to developing prostate cancer, resulting in increased intention and actual pursuit of screening for prostate cancer compared with men with no family history of prostate cancer.23,24 In addition, primary care providers recognize family history as a risk factor for prostate cancer and are more likely to offer prostate cancer screening to men with a family history than to men with no family history.25,26 To minimize the effect of recall and ascertainment biases, we conducted a prospective cohort study to evaluate the contribution of family history to prostate cancer development. Our results were derived from family histories that were all collected prospectively, before enrollment in the study, and before prostate biopsy or a diagnosis of cancer, minimizing the effect of the abovementioned biases. Another potential confounder of cancer ascertainment when studying the association between family history and prostate cancer risk is that factors other than family history usually prompt prostate biopsies, such as an elevated PSA level or abnormal DRE findings. Case-control studies investigating the contribution of family history to prostate cancer risk have generally matched cases and controls by age and locality, but not by PSA level or DRE findings. Because factors other than family history have led to the diagnosis of prostate cancer, the contribution of family history to prostate cancer risk might have been misrepresented. Our study attempted to isolate the contribution of family history to prostate cancer risk by excluding factors other than family history that might lead to a cancer diagnosis by limiting the population to men with normal DRE findings and a PSA level of 4.0 ng/mL or less. One limitation of our study was that we did not have an adequate control group of men who had undergone biopsy and had a PSA level of 4.0 ng/mL or less and normal DRE findings, but no family history, in the SABOR cohort with which to directly benchmark the high prostate cancer prevalence in this group. For the most part, these men had not been offered a prostate biopsy because they had no clinical indication (PSA level greater than 4.0 ng/mL or abnormal DRE findings). However, the prevalence rate in this study, 25.3%, exceeded that (21.9%) reported for a large group of 5519 PCPT placebo participants reported by Thompson et al.27 that included both positive and negative family history participants, normal and abnormal DRE results, and a range of PSA values. Thompson et al.27 reported a statistically significant increase in prostate cancer risk with a positive family history. Our study results have indicated that a positive family history plays a role as an independent risk factor for prostate cancer in men with a PSA level less than 4.0 ng/mL and normal DRE findings. Our findings have confirmed that no PSA level exists below which the risk of prostate cancer is negligible, although even within the narrow range of 0 to 4.0 ng/mL, UROLOGY 70 (4), 2007
an increasing PSA level correlated with an increasing risk of cancer. Of interest is the unexpectedly high frequency of cancer in these men, who might not otherwise have been offered a prostate biopsy. The placebo group in the PCPT has yielded the largest body of prospective data on the predictive value and performance characteristics of PSA in the 0 to 4.0 ng/mL range.2,5 The PCPT identified family history as a significant risk factor for prostate cancer at PSA levels within this range. Among the men who underwent end-of-study biopsy, 477 (16.2%) reported a positive family history for prostate cancer, and 94 men (19.7%) within this subgroup were diagnosed with cancer compared with 355 (14.4%) of the 2473 men who reported a negative family history for prostate cancer. In our study, 25.3% of men with a PSA level in this range had prostate cancer. The possible explanations for this discrepancy include the smaller sample size of our study and the increased number of biopsy cores taken in the present study (10 to 12 versus a mean number of biopsy cores per subject of 6 the PCPT). The most recent analysis from the PCPT has led to the development of a risk calculator for the determination of prostate cancer risk at prostate biopsy. In a logistic regression analysis conducted in 5519 men from the placebo group of PCPT, all of whom underwent prostate biopsy, the risk factors that predicted for a risk of prostate cancer and/or high-grade disease included PSA level, age, race/ethnicity, family history of prostate cancer, and history of previous negative prostate biopsy. The investigators found a significant contribution of the risk of a positive biopsy from a family history of prostate cancer (odds ratio 1.31, 95% confidence interval 1.11 to 1.55, P ⫽ 0.002). The risk calculator is available online at http:// www.compass.fhcrc.org/edrnnci/bin/calculator/main.asp. The present study has confirmed these associations of risk and the interactions of PSA and family history of prostate cancer for the risk of the disease. It has been proposed that lower PSA cutoffs be recommended for men with a positive family history to equalize PSA-based screening such that all men of equal risk of cancer be afforded the opportunity for enhanced prostate cancer screening by elective prostate biopsy.27 In men with normal DRE findings and no previous biopsy, to obtain a 25% risk of prostate cancer, the PCPT risk calculator would suggest a PSA cutoff of 2.3 ng/mL for men with no family history, but 1.7 ng/mL for men with a positive family history.27 Although lowering the PSA cutoffs for offering prostate biopsy is in routine clinical practice in some centers,28 one of the criticisms of adopting such a strategy is the risk of an increased incidence of indolent cancer. However, recent studies have suggested a reduced risk of mortality associated with active treatment for low-risk and intermediate-risk prostate cancer.29,30 If PSA cutoffs are going to continue to be the standard for referral to biopsy, our study has confirmed UROLOGY 70 (4), 2007
that lowering the PSA cutoff for men with a positive family history is justified.
CONCLUSIONS The results of this prospective evaluation of a family history of prostate cancer as a predictor of prostate cancer in men who would not normally be offered a prostate biopsy are sobering. Our findings have reiterated that no PSA level exists below which the risk of prostate cancer is negligible and have confirmed that a positive family history by itself is a strong predictor of prostate cancer in the absence of other clinical indications for a prostate biopsy. Although we refrain from recommending prostate biopsy to all men with a positive family history, it could be argued that the PSA cutoff values for offering prostate biopsy should be lowered because of the high prevalence rate of prostate cancer in this cohort. At present, although patients and their urologists now have a risk calculator to estimate the risk of prostate cancer and aggressive disease if a biopsy is performed, the tools to predict the natural history of many prostate cancers are not available. Continued accumulation of prospectively collected data with full ascertainment of cancer status will enable construction of multifactorial risk calculators, so that patients and urologists can identify specific and individualized risks and match these against their level of acceptance or aversion to such risk. Acknowledgment. To Rosario Mendez-Meza for technical assistance and John Schoolfield for assistance with data analysis. References 1. Jemal A, Siegel R, Ward E, et al: Cancer statistics, 2006. CA Cancer J Clin 56: 106 –130, 2006. 2. Thompson IM, Ankerst DP, Chi C, et al: Operating characteristics of prostate-specific antigen in men with an initial PSA level of 3.0 ng/ml or lower. JAMA 294: 66 –70, 2005. 3. Kupelian P, Katcher J, Levin H, et al: Correlation of clinical and pathologic factors with rising prostate-specific antigen profiles after radical prostatectomy alone for clinically localized prostate cancer. Urology 48: 249 –260, 1996. 4. Antenor JA, Roehl KA, Eggener SE, et al: Preoperative PSA and progression-free survival after radical prostatectomy for stage T1c disease. Urology 66: 156 –160, 2005. 5. Thompson IM, Pauler DK, Goodman PJ, et al: Prevalence of prostate cancer among men with a prostate-specific antigen level ⬍ or ⫽ 4.0 ng per milliliter. [Erratum appears in N Engl J Med 351: 1470, 2004]. N Engl J Med 350: 2239 –2246, 2004. 6. Lichtenstein P, Holm NV, Verkasalo PK, et al: Environmental and heritable factors in the causation of cancer—analyses of cohorts of twins from Sweden, Denmark, and Finland. N Engl J Med 343: 78 – 85, 2000. 7. Page WF, Braun MM, Partin AW, et al: Heredity and prostate cancer: a study of World War II veteran twins. Prostate 33: 240 –245, 1997. 8. Verkasalo PK, Kaprio J, Koskenvuo M, et al: Genetic predisposition, environment and cancer incidence: a nationwide twin study in Finland, 1976 –1995. Int J Cancer 83: 743–749, 1999. 9. Karan D, Lin MF, Johansson SL, et al: Current status of the molecular genetics of human prostatic adenocarcinomas. Int J Cancer 103: 285–293, 2003.
751
10. Bratt O: Hereditary prostate cancer: clinical aspects. J Urol 168: 906 –913, 2002. 11. Scott KM, Fanta P, Calaluce R, et al: Diagnostic frozen prostate sextant biopsies: an approach for preserving protein and RNA for additional studies. Prostate 44: 296 –302, 2000. 12. Balic I, Graham ST, Troyer DA, et al: Androgen receptor length polymorphism associated with prostate cancer risk in Hispanic men. J Urol 168: 2245–2248, 2002. 13. Hernandez J, Balic I, Johnson-Pais TL, et al: Association between an estrogen receptor alpha gene polymorphism and the risk of prostate cancer in black men. J Urol 175: 523–527, 2006. 14. Morganti G, Gianferrari L, Cresseri A, et al: [Clinico-statistical and genetic research on neoplasms of the prostate.] Acta Genet Stat Med 6: 304 –305, 1956. 15. Woolf CM: An investigation of the familial aspects of carcinoma of the prostate. Cancer 13: 739 –744, 1960. 16. Bruner DW, Moore D, Parlanti A, et al: Relative risk of prostate cancer for men with affected relatives: systematic review and metaanalysis. Int J Cancer 107: 797– 803, 2003. 17. Johns LE, and Houlston RS: A systematic review and meta-analysis of familial prostate cancer risk. BJU Int 91: 789 –794, 2003. 18. Hayes RB, Liff JM, Pottern LM, et al: Prostate cancer risk in U.S. blacks and whites with a family history of cancer. Int J Cancer 60: 361–364, 1995. 19. Zeegers MP, Jellema A, and Ostrer H: Empiric risk of prostate carcinoma for relatives of patients with prostate carcinoma: a meta-analysis. Cancer 97: 1894 –1903, 2003. 20. Spitz MR, Currier RD, Fueger JJ, et al: Familial patterns of prostate cancer: a case-control analysis. J Urol 146: 1305–1307, 1991.
752
21. Schuurman AG, Zeegers MP, Goldbohm RA, et al: A case-cohort study on prostate cancer risk in relation to family history of prostate cancer. Epidemiology 10: 192–195, 1999. 22. Cerhan JR, Parker AS, Putnam SD, et al: Family history and prostate cancer risk in a population-based cohort of Iowa men. Cancer Epidemiol Biomarkers Prev 8: 53– 60, 1999. 23. Jacobsen PB, Lamonde LA, Honour M, et al: Relation of family history of prostate cancer to perceived vulnerability and screening behavior. Psycho-Oncology 13: 80 – 85, 2004. 24. Bock CH, Peyser PA, Gruber SB, et al: Prostate cancer early detection practices among men with a family history of disease. Urology 62: 470 – 475, 2003. 25. Durham J, Low M, and McLeod D: Screening for prostate cancer: a survey of New Zealand general practitioners. N Z Med J 116: U476, 2003. 26. Tudiver F, Guibert R, Haggerty J, et al: What influences family physicians’ cancer screening decisions when practice guidelines are unclear or conflicting? J Fam Pract 51: 760, 2002. 27. Thompson IM, Ankerst DP, Chi C, et al: Assessing prostate cancer risk: results from the Prostate Cancer Prevention Trial. J Natl Cancer Inst 98: 529 –534, 2006. 28. Krumholtz JS, Carvalhal GF, Ramos CG, et al: Prostate-specific antigen cutoff of 2.6 ng/mL for prostate cancer screening is associated with favorable pathologic tumor features. Urology 60: 469 – 474, 2002. 29. Wong YN, Mitra N, Hudes G, et al: Survival associated with treatment vs observation of localized prostate cancer in elderly men. JAMA 296: 2683–2693, 2006. 30. Holmberg L, Bill-Axelson A, Garmo H, et al: Prognostic markers under watchful waiting and radical prostatectomy. Hematol Oncol Clin North Am 20: 845– 855, 2006.
UROLOGY 70 (4), 2007