Obesity and Future Prostate Cancer Risk among Men after an Initial Benign Biopsy of the Prostate

In 2012, there are expected to be 241,740 new cases of prostate cancer, along with 28,170 deaths from the disease, making prostate cancer the most commonly diagnosed non-skin cancer among men in the United States and the second leading cause of cancer-related death among men (1). The advent of widespread prostate-specific antigen (PSA) testing starting around 1988 resulted in many more men being considered at “increased risk” for prostate cancer (2–6). Approximately 1 million biopsies are conducted annually in the United States due to suspicion of prostate cancer, two-thirds of which reveal benign conditions (7). To better understand prostate cancer etiology and to determine whether subpopulations of men diagnosed with benign conditions can be identified who have the highest risk for future prostate cancer diagnoses, we have been investigating risk factors for prostate cancer after an initial negative biopsy result (8, 9). Here, we report on associations between obesity and subsequent risk of prostate cancer after an initial benign biopsy.

General population studies of obesity and prostate cancer incidence have found positive, negative, and null associations (10–12), whereas studies of prostate cancer mortality have found consistent associations between obesity and mortality (13, 14). A recent meta-analysis found a modest increase in prostate cancer risk with increasing body mass index (BMI) (RR = 1.05; 95% confidence interval (CI) 1.01–1.08] per 5 unit increase in BMI (10). However, more consistent results were found across studies when analyses were conducted by grade and stage of disease at diagnosis; obesity was found to be associated with high-grade prostate cancer and unassociated with or perhaps even protective for localized or low-grade disease (10, 15). It has been noted that complex relationships between obesity and health care use and screening and diagnostic efficacy, may affect the validity and interpretation of studies of obesity and prostate cancer risk (16–20). Because of related comorbidities, obese men may have greater contact with the health care system and thus receive more intense prostate cancer screening (16, 21). At the same time obesity is associated with poverty and lower educational attainment, which themselves are associated with poorer access to health care (22–26). However, the Prostate Cancer Prevention Trial, in which consistent PSA testing and digital rectal examinations (DRE) were part of the protocol, found that obesity was associated with a higher risk of high-grade tumors and was protective against low-grade tumors (27). It has also been noted that obesity and weight gain are associated with lower PSA test scores and PSA velocity measures, and it has been suggested that it is more difficult to conduct a thorough DRE in obese men; thus, obesity may delay initial diagnostic tests (17–20). An issue of particular relevance to the question of obesity as a risk factor for prostate cancer after an initial benign biopsy is the finding that obese men have larger prostates, and the sensitivity of needle biopsies may be lower in men with larger prostates (19, 28, 29).

Research on premalignant lesions in the breast, colon, and esophagus has yielded insights into cancer etiology (30–34). However, there is little information on whether obesity is associated with the presence of prostatic intraepithelial neoplasia (PIN), a precursor lesion of prostate cancer. The presence of PIN in a benign biopsy indicates a higher risk for future diagnosis of prostate cancer (8). One small case–control study found that obesity was associated with PIN in benign biopsy specimens of men with smaller (<40 cm³) but not larger prostates (35). The demonstration of associations between obesity and the presence of PIN in benign prostate specimens would provide additional evidence for the biologic basis for associations between obesity and prostate cancer risk.

Here, using a case–control study nested within a historical cohort of men at the Henry Ford Health System (HFHS) who underwent biopsy or transurethral resection of the prostate (TURP) procedures that yielded benign results, we investigate associations between obesity and future prostate cancer incidence. The design of this study allows for analyses of obesity and the presence of PIN in the initial benign specimen. In addition, the follow-up of these men through a comprehensive medical system allows for collection of data on prostate screening behaviors and adjustment for the intensity of prostate cancer screening undergone by the men during follow-up.

A historical cohort of 6,692 men was identified who had a benign prostate specimen collected by needle core biopsy or TURP between January 1990 and December 2002 at Henry Ford Health System. These men were followed up for prostate cancer incidence to December 2007. The incidence of prostate cancer within this high-risk cohort was approximately twice that of the general Detroit Surveillance Epidemiology and End Results (SEER) population, although the ratio of African American to White cases in the cohort (ratio = 1.62) was similar to that in the overall SEER data (ratio = 1.53). Within this cohort, a nested case–control of 574 case–control pairs was assembled (8). Eligibility criteria included a recorded PSA level within a year of cohort entry and no history of a previous prostate cancer diagnosis. “Date of cohort entry” was defined as the date of initial benign prostate procedure; “date of case diagnosis” was the date of first cancer-positive tissue specimen or the date a clinician first reported a clinical diagnosis of prostate cancer. Patients diagnosed with prostate cancer less than one year from the date of initial benign procedure were ineligible for the study. Incidence density sampling was used to select controls with replacement from all cohort members at risk at the time of case occurrence. Controls were randomly selected from among those cohort members who were free of prostate cancer at a follow-up duration greater than or equal to the time between cohort entry and diagnosis of the matched case. Controls were matched to cases on age at entry into cohort (± 2 years), date of entry into cohort (± 2 years), race (African American or White), and type of initial specimen (biopsy or TURP: 7% of cases had a TURP).

Data for this study were abstracted from the HFHS medical records. The presence of any notation in the medical record of a family history of prostate cancer in the father or brothers of the subject was used to indicate a positive family history. Data on all PSA tests were abstracted and the PSA test value immediately before the initial benign procedure at the HFHS was used as the baseline PSA level. If a PSA test result immediately before the initial benign procedure was not available (4% of men), the PSA test score from the first PSA test subsequent to the procedure was used. Screening intensity was measured as the number of PSA tests and DRE during follow-up; for cases the period between the initial benign procedure and diagnosis was examined and for controls the period between the initial benign procedure and matching date (corresponding to the time interval between the initial procedure and diagnosis for the matched case) was examined. For men who are diagnosed with prostate cancer, there may be a flurry of PSA tests and DRE that occur as part of the diagnostic process. Therefore, sensitivity analyses were conducted using a measure of screening intensity that removed from the analysis PSA tests and DRE occurring in the month before diagnosis in the cases. Medical record data on height and weight measured as soon after the benign procedure as recoded in the medical record (median 115 days) was used to calculate BMI and overweight (BMI≥ 25 and <30) and obese status (BMI ≥ 30) at the beginning of follow-up. All surgical specimens involved in this study were reviewed for the presence of PIN by a single urological pathologist (O.N. Kryvenko) who was blind to prostate cancer outcomes at the time of review. Pathology data from the tumors were used to classify cases as having high (Gleason score ≥7; 4 primary, 3 secondary) or low-grade tumors. Advanced stage disease was defined as pathologic or clinical stage T3a and higher. Aggressive prostate cancer was defined has having either high-grade or advanced stage disease (36).

Logistic regression models were used to assess whether overweight and obesity status at the time of the initial procedure were associated with the presence of PIN in the benign prostate specimen, after adjustment for age, race, PSA score at initial procedure, and a family history of prostate cancer. Conditional logistic regression models were used to estimate ORs for prostate cancer incidence during follow-up. Associations between body size and prostate cancer incidence were modeled using BMI as a continuous variable and using indicator variables for overweight and obesity status at the initial benign procedure. The matched design accounts for confounding by age, race, and trends in screening and diagnostic procedures during follow-up, and analyses further adjusted for PSA levels at the time of the benign procedure, family history of prostate cancer and the number of PSA tests and DRE during follow-up. Sensitivity analyses were conducted using the screening intensity variables that for cases ignored PSA and DRE tests that occurred in the 30 days before diagnosis. Additional analyses were conducted among cases and their matched controls stratifying cases on median time to diagnosis (1,538 days) and disease aggressiveness.

Of the 574 case–control pairs, 494 pairs had complete data for BMI and the covariates of interest. The primary missing data element was BMI, accounting for almost all of the loss of case–control sets. However, lack of BMI data was not significantly associated with age at enrollment, race, a family history of prostate cancer, or PSA level at the initial procedure. Table 1 documents the distribution of prostate cancer risk factors and screening behaviors in cases and controls and shows that at the initial benign procedure, cases had higher PSA levels and were more likely to have a family history of prostate cancer and had more PSA tests and DRE during follow-up. As would be expected for a cohort of men identified on the basis of having undergone a biopsy or TURP procedure for the suspicion of prostate cancer, mean baseline PSA levels were higher than 4.0 ng/mL.

PIN was found in 11% of the initial benign specimens of the subject, and a finding of PIN was significantly associated with obesity at the time of the procedure (see Table 2). Table 3 shows the results of regression models assessing associations between baseline characteristics and prostate cancer risk overall and for those diagnosed earlier and later in follow-up. Overall, a higher PSA value at the initial procedure and a family history of prostate cancer were associated with prostate cancer incidence, as was the number of PSA tests during follow-up. In the overall analyses, overweight and obesity were associated with a higher risk of prostate cancer; however, the association between obesity and prostate cancer incidence was confined to diagnoses occurring within a shorter time period (<1,538 days—the median duration of follow-up) after the initial benign procedure. Similar results were observed in sensitivity analyses in which PSA and DRE tests conducted in the 30 days before prostate cancer diagnosis were omitted form the measures of screening intensity. When considered as a continuous variable, BMI was marginally associated with prostate cancer incidence overall (OR = 1.15 per 5 unit difference in BMI; 95% CI, 0.98–1.36); the association was stronger for diagnoses occurring earlier during follow-up (OR = 1.26 per 5 unit difference in BMI; 95% CI, 0.99–1.62).

Table 4 shows results of analyses in which the case series was stratified by tumor characteristics and cases were compared with their matched controls. Overweight and obese status were significantly associated with higher prostate cancer risk for men with low grade tumors, nonadvanced stage tumors, and nonaggressive tumors as defined by grade and stage. However, these associations for cases with low grade, earlier stage, or less aggressive tumors did not significantly differ from the associations between body size and prostate cancer risk for men with high grade, later stage, or more aggressive tumors.

In this prospective study of men followed-up after an initial prostate procedure yielding a benign result, obesity was associated with the presence of PIN in initial benign specimens and with higher prostate cancer risk during follow-up, although primarily for diagnoses occurring within a shorter time period (<1,538 days) after the initial benign procedure. The association observed here between body size and prostate cancer risk is substantially larger than seen in prior studies (10, 15). This may reflect chance or the composition of the cohort; it has been suggested that obesity reduces the sensitivity of PSA testing and may delay referral for biopsy, therefore it is possible that obesity affects whether a man is referred for a biopsy or TURP (17, 18). Obesity may also reduce the diagnostic efficiency of needle biopsies. Several studies have found that obese men have larger prostates, and the sensitivity of needle biopsies for detecting small tumors may be lower in larger prostates (19, 28, 29). It is possible that part of the observed association between body size and prostate cancer incidence in this cohort reflects associations between body size and larger prostate size which reduced the sensitivity of the original needle biopsy, and tumors missed by the initial biopsy grew and were detected in subsequent biopsies during follow-up (12, 19, 37). The observation that obesity is primarily associated with diagnoses occurring early in the follow-up period is consistent with this interpretation.

In contrast to earlier cohort studies, obesity was not associated with high Gleason grade tumors, in fact, although the differences by grade were not statistically significant, obesity seemed to be more strongly associated with low-grade tumors (10, 15). This difference between the current study and prior studies may reflect the make-up of the cohort–high risk men under medical surveillance or the relatively unique, vertically integrated characteristics of the HFHS. The distribution of tumor grade by obesity status may also relate to the proposition that a subset of the obese men had tumors that were missed by the initial biopsy. If the missed tumors were small and low grade, then they may have still been of lower grade when diagnosed in the first few years after the initial biopsy.

This is one of the first studies to assess associations between obesity and the presence of PIN in benign specimens. PIN is considered to be a precursor lesion for prostate cancer and the presence of PIN in a benign specimen is a strong risk factor for a future diagnosis of prostate cancer (8). Research on risk factors for precursor lesions to esophageal cancer (Barrett esophagus), colon cancer (adenomatous polyps), and breast cancer (ductal carcinoma in situ) has provided important insights into cancer etiology (30–34). Observations that obesity is associated with Barrett esophagus and colonic adenomatous polyps have been interpreted as strengthening the case that obesity causes cancer of these respective organs (38, 39). Here, obesity was observed to be associated with the presence of PIN in the initial benign specimen providing additional evidence that obesity influences prostate carcinogenesis.

The strengths of this study include its prospective design, the availability of initial benign specimens, adjustment for screening intensity during follow-up, the availability of high quality medical records collected within a single integrated health system, and the matching of controls to cases on date of the initial benign procedure and duration of follow-up. Because of this matching, the cases and controls experience the same temporal trends in medical practice, medical technology, and case management, and have the same period of observation in which screening behaviors can occur. The availability of data on PSA testing and DRE during follow-up allowed for control for potential differences in prostate cancer screening intensity by obesity status that may otherwise have introduced bias into the study. A caveat for the study is that it was conducted in a high-risk population who had already undergone a procedure for suspicion of prostate cancer and thus the results may not be generalizable to the general male population. However, approximately 1 million prostate biopsies are conducted annually in the United States, a number which can be expected to rise with the aging American population, two-thirds of these are negative (7). The study is most generalizable to this population of men.

In conclusion, this is one of the first studies to assess associations between obesity and PIN, a precursor lesion for prostate cancer, and the observed association provides further support for a biologic role of obesity in prostate cancer development. Obesity was also associated with a higher incidence of prostate cancer after an initial biopsy or TURP that yielded a benign result; however, this association was only apparent for tumors occurring earlier in follow-up. Obesity may be a factor to consider when planning the intensity of clinical follow-up of men after an initial benign procedure.

No potential conflicts of interest were disclosed.

Conception and design: A. Rundle, B.A. Rybicki

Development of methodology: A. Rundle, D.L. Tang, B.A. Rybicki

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): O. Kryvenko, D.L. Tang, B.A. Rybicki

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): A. Rundle, M. Jankowski, O. Kryvenko, D.L. Tang, B.A. Rybicki

Writing, review, and/or revision of the manuscript: A. Rundle, M. Jankowski, O. Kryvenko, D.L. Tang, B.A. Rybicki

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): A. Rundle, D.L. Tang

Study supervision: A. Rundle, D.L. Tang, B.A. Rybicki

This work was supported by a grant from the National Institute of Environmental Health Sciences (5R01-ES011126).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.