Influence of body mass index on prostate-specific antigen failure after androgen suppression and radiation therapy for localized prostate cancer

Authors

  • Jason A. Efstathiou MD, PhD,

    Corresponding author
    1. Department of Radiation Oncology, Brigham and Women's Hospital and Dana Farber Cancer Institute, Boston, Massachusetts
    • Department of Radiation Oncology, Brigham & Women's Hospital and Dana Farber Cancer Institute, 75 Francis St., L2, Boston, MA 02115
    Search for more papers by this author
    • Fax: (617) 632-2742.

  • Ming-Hui Chen PhD,

    1. Department of Statistics, University of Connecticut, Storrs, Connecticut
    Search for more papers by this author
  • Andrew A. Renshaw MD,

    1. Department of Pathology, Brigham and Women's Hospital and Dana Farber Cancer Institute, Boston, Massachusetts
    Search for more papers by this author
  • Marian J. Loffredo RN, OCN,

    1. Department of Radiation Oncology, Brigham and Women's Hospital and Dana Farber Cancer Institute, Boston, Massachusetts
    Search for more papers by this author
  • Anthony V. D'Amico MD, PhD

    1. Department of Radiation Oncology, Brigham and Women's Hospital and Dana Farber Cancer Institute, Boston, Massachusetts
    Search for more papers by this author

Abstract

BACKGROUND

Increasing body mass index (BMI) is associated with shorter time to prostate-specific antigen (PSA) failure after radical prostatectomy. Whether BMI is associated with time to PSA failure was investigated in men treated with androgen suppression therapy (AST) and radiation therapy (RT) for clinically localized prostate cancer.

METHODS

The observational prospective cohort study consisted of 102 men with clinically localized prostate cancer who received 70 Gy RT with 6 months of AST on a single arm of a randomized trial between December 1995 and April 2001. Height and weight data were available at baseline for 99 (97%) of the men, from which BMI was calculated. Adjusting for age (continuous) and known prognostic factors including PSA level (continuous), Gleason score, and T-category, Cox regression analyses were performed to analyze whether BMI (continuous) was associated with time to PSA failure (PSA >1.0 ng/mL and increasing >0.2 ng/mL on 2 consecutive visits).

RESULTS

Median age and median BMI (interquartile range [IQR]) at baseline was 72 (69.1–74.7) years and 27.4 (24.8–30.7) kg/m,2 respectively. In addition to an increasing PSA level (P = .006) and Gleason 8–10 cancer (P = .024), after a median follow-up (IQR) of 6.9 (5.6–8.5) years, an increasing BMI was also significantly associated with a shorter time to PSA failure (adjusted hazard ratio [HR]: 1.10; 95% confidence interval [CI]: 1.01–1.19; P = .026) after RT and AST.

CONCLUSIONS

After adjusting for known prognostic factors, baseline BMI is significantly associated with time to PSA failure after RT and AST for men with clinically localized prostate cancer. Further study is warranted to assess the impact of an increasing BMI after AST administration on PSA failure, prostate cancer-specific, and all-cause mortality. Cancer 2007. © 2007 American Cancer Society.

Several randomized trials have shown a survival advantage to adjuvant androgen suppression therapy (AST) for patients with locally advanced (ie, extracapsular extension or node-positive)1–3 or intermediate/high-risk localized prostate cancer.4, 5 Based on this observed survival benefit, the use of AST in conjunction with radiation therapy (RT) has appropriately increased for such patients.6

AST exposes patients to a number of potential adverse effects, namely, decreased libido, hot flashes, gynecomastia, cognitive and mood changes, anemia, osteoporosis, as well as adverse changes in body composition,7, 8 metabolism, and the cardiovascular system.9 Particularly in men of advanced age with localized disease and a single high-risk feature who are not likely to die of their disease, the potential risks of AST must be carefully contemplated.

Increasing body mass index (BMI) has been shown to be associated with increased overall mortality,10 more aggressive prostate cancer,11–13 and higher prostate-specific antigen (PSA) failure rates after radical prostatectomy.12, 14–18 Although the data after radiation therapy is limited, 1 recent report showed that BMI is a predictor of PSA failure among patients treated with external beam radiation therapy (EBRT) alone,19 whereas another report suggested that this may not be true after brachytherapy.20

The prognostic significance of BMI with regard to outcome after AST and RT has not been studied. Therefore, we investigated whether BMI is associated with time to PSA recurrence in patients treated with AST and RT for clinically localized prostate cancer.

MATERIALS AND METHODS

Patient Selection and Treatment

Between December 1, 1995, and April 15, 2001, 102 of 206 patients from the Harvard outreach (Saint Anne's Hospital, Fall River, Mass; Metro West Medical Canter, Framingham, Mass; Suburban Oncology Center, Quincy, Mass) and central hospitals (Dana-Farber Cancer Institute, Brigham & Women's Hospital, and Beth Israel Deaconess Medical Center) with 1992 American Joint Commission on Cancer21 category T1b to T2b, NX, M0 centrally reviewed adenocarcinoma of the prostate enrolled in a prospective trial and were randomized to receive 70 Gy 3D conformal RT (3D-CRT) in combination with 6 months of AST.4 AST consisted of a combination of a luteinizing hormone-releasing hormone agonist (leuprolide acetate or goserelin) and a nonsteroidal antiandrogen (flutamide).

Eligible patients included those with a PSA of at least 10 ng/mL (maximum, 40 ng/mL) or a Gleason score of at least 7 (range, 5–10) or radiographic evidence using endorectal coil magnetic resonance imaging (MRI) of extracapsular extension or seminal vesical invasion. Patients were considered ineligible if they had a prior history of malignancy except for nonmelanoma skin cancer or any history of hormone therapy use. In addition, all patients were required to have a negative bone scan and pelvic lymph node assessment using MRI or computed tomography (CT) within 6 months of registration. Eligible patients also needed to have an Eastern Cooperative Oncology Group performance status of 0 or 1 (range, 0–4), white blood cell count of at least 3000/μL, hematocrit of more than 30%, platelet count of more than 100 × 103/μL, and a life expectancy of at least 10 years, excluding death related to prostate cancer at study entry. All patients provided written informed consent and the study was approved by the institutional review boards at the Dana-Farber/Harvard Cancer Center, Saint Anne's Hospital, and the Metro West Medical Center. The details of the RT technique, doses, and fields and AST formulation have been described previously.4

Body Mass Index and Patient Characteristics

Height and weight data were available at baseline for 99 of the 102 (97%) men from which BMI (weight in kilograms divided by height in meters squared [kg/m2]) was calculated, using the equation BMI = [weight (pounds)/height (inches)2] × 703.07. A summary of the baseline clinical characteristics of the men is provided in Table 1.

Table 1. Baseline Clinical Characteristics of Analyzable Patients (n = 99)
Characteristic 
  1. IQR indicates interquartile range; BMI, body mass index; PSA, prostate-specific antigen.

Median age, y (IQR)72 (69.1–74.7)
Median baseline BMI, kg/m2 (IQR)27.4 (24.8–30.7)
Median baseline PSA level, ng/mL (IQR)11.1 (7.5–15.9)
Gleason score, no. (%)
 ≤633 (33)
 756 (57)
 8–1010 (10)
Tumor category, no. (%)
 T154 (55)
 T245 (45)

Follow-up Protocol

Follow-up visits were performed after the end of RT every 3 months for 2 years, every 6 months for an additional 3 years, and then annually thereafter. At each follow-up visit a PSA level was obtained before performing the digital rectal examination. PSA failure was defined as a PSA of more than 1.0 ng/mL and increasing by more than 0.2 ng/mL on 2 consecutive measurements to avoid scoring PSA failure from PSA increase due to testosterone rebound that can occur after the AST is discontinued.4 All patients were followed up directly by the site investigators until death or until the dates of the last follow-up, whichever came first. The median follow-up from date of registration was 6.9 years (interquartile range [IQR], 5.6–8.5 years).

Statistical Methods

Age, PSA level, and BMI were analyzed as continuous variables. Categorical variables included clinical tumor stage (T2 versus T1 [baseline]) and Gleason score (8–10 vs 7 vs 6 or less [baseline]). For the categorical variables, the cutpoints selected were made before the data were examined and were based on established strata. Univariate Cox proportional hazard regression analyses22 were performed to evaluate the solitary effect of each variable on the risk of PSA failure. While adjusting for age and the known prognostic factors including PSA level, Gleason score, and T-category, multivariate Cox regression analyses22 were performed to analyze whether BMI was independently associated with time to PSA failure. Unadjusted and adjusted hazard ratios (HRs) were calculated for all covariates using the Cox proportional hazards model with associated 95% confidence intervals (CI) and P-values for probability estimates of time to PSA failure. The methods of Kaplan and Meier23 were used to estimate time to PSA recurrence stratified by the median value of BMI and comparisons were performed with the log-rank test.24 All statistical tests were 2-sided and a P-value <.05 was considered statistically significant. Statistical Analysis System (SAS) v. 9.1.3 (SAS Institute, Cary, NC) was used for all statistical analyses.

RESULTS

Baseline Clinical Characteristics

As listed in Table 1, the median age and median BMI of the patient cohort at baseline was 72 (IQR 69.1–74.7) years and 27.4 (IQR 24.8–30.7) kg/m2, respectively. Median PSA level at baseline was 11.1 ng/mL (IQR 7.5–15.9). Two-thirds of patients had Gleason score 7 or higher (67%) and approximately one-half had T1 cancer (55%).

Factors Associated With PSA Failure: Univariate Analysis

After a median follow-up of 6.9 years (IQR 5.6–8.5), 25 of the 99 patients developed a PSA-defined recurrence. As shown in Table 2, on univariate analysis, in addition to an increasing PSA (P = .004), an increasing BMI (P = .018) was significantly associated with an increased risk of PSA failure. Age (P = .086), Gleason score 8 to 10 (P = .099), and T2 disease (P = .062) approached statistical significance. Gleason score 7 cancer (P = .777) was not significantly associated with PSA failure.

Table 2. Results of Univariate and Multivariate Analyses of Covariates and Their Association With PSA Failure
CovariateNo. of MenPSA failure
No. EventsUVAMVA
HR [95% CI]PHR [95% CI]P
  1. PSA indicates prostate-specific antigen; UVA, univariate analysis; MVA, multivariate analysis; HR, hazard ratio; CI, confidence interval; BMI, body mass index; GS, Gleason score.

BMI, continuous99251.11 (1.02,1.20).0181.10 (1.01,1.19).026
PSA, continuous99251.06 (1.02,1.11).0041.07 (1.02,1.13).006
Age, y, continuous99250.94 (0.87,1.01).0860.98 (0.91,1.06).624
GS ≤63381.01.0
GS 756120.88 (0.36,2.15).7771.36 (0.46,4.0).580
GS 8 to 101052.56 (0.84,7.85).0994.35 (1.22,15.5).024
T154101.01.0
T245152.14 (0.96,4.78).0621.69 (0.73,3.94).223

Factors Associated With PSA Failure: Multivariate Analysis

Similarly, after adjusting for other covariates on multivariate analysis, an increasing PSA level (adjusted HR: 1.07; 95% CI: 1.02–1.13; P = .006), Gleason 8 to 10 cancer (adjusted HR: 4.35; 95% CI: 1.22–15.5; P = .024), and an increasing BMI (adjusted HR: 1.10; 95% CI: 1.01–1.19; P = .026) were significant independent predictors of PSA failure (Table 2). A significant association with PSA recurrence was not noted for age (P = .624), Gleason score 7 (P = .580), or T2 disease (P = .223).

PSA Recurrence Estimates Based on BMI

As graphically displayed in Figure 1, there was a near statistically significant difference in the Kaplan-Meier estimates of PSA recurrence when stratified about the median value of BMI (P = .076). For the purpose of illustration, the 5-year estimates of PSA recurrence in men with a BMI ≥27.4 kg/m2 (median or more) were 13% higher than in men with a BMI <27.4 kg/m2 (29.03% [95% CI: 18.31, 44.07] vs 16.03% [95% CI: 7.96, 30.78], respectively).

Figure 1.

Kaplan-Meier plot of time to prostate-specific antigen (PSA) recurrence following (radiation therapy) RT and (androgen suppression therapy) AST stratified by median body mass index (BMI). Numbers at risk are listed below the curves.

DISCUSSION

Obesity and prostate cancer are 2 of the most common causes of morbidity and mortality afflicting men in the US.10, 25, 26 Currently, over 30% of American men are obese25 and greater than 230,000 men will have been diagnosed with prostate cancer in 2006.26 Although certain types of cancer have been shown to occur with increased frequency and have a higher likelihood of causing death in obese patients,27, 28 observational studies remain unclear as to the association of obesity with prostate cancer risk.27–34

An increasing BMI, however, has been shown to be associated with higher-grade prostate cancer11–13 and higher PSA recurrence rates after radical prostatectomy.12, 14–18 Yet there has been a paucity of data documenting the prognostic significance of obesity with regard to biochemical outcome after RT alone19, 20 or in men who undergo RT in conjunction with AST for higher-risk prostate cancer.

To our knowledge, this is the first report to show that an increasing baseline BMI is significantly associated with a shorter time to PSA failure after RT and AST in men with clinically localized prostate cancer. Specifically, when stratified about the median BMI, we found that men with a BMI ≥median had a near 2-fold increased risk of PSA recurrence at 5 years (13% absolute risk difference) than men with lower BMIs.

The mechanisms for increased biochemical recurrence are unknown and likely multifactorial. Potential biological explanations take into account alterations in the hormonal milieu. Obesity is associated with changes in serum levels of testosterone, estradiol, insulin, insulin-like growth factors (IGFs), adiponectin, and leptin, all of which may contribute to a microenvironment predisposing to more aggressive disease.35–51 Beyond body mass itself, diet and physical activity may play a role.52–54 Detection and selection biases are also possible. Clinical understaging of the extent of disease can occur in obese men who may have larger prostate sizes55–57 and in whom it may be harder to perform a digital rectal examination.57 However, prostatectomy series that control for adverse pathologic features maintain an association between an increased BMI and risk of biochemical progression.14–16 Another possibility is that treatment may be less effective in obese patients. For example, obese patients have a greater risk of positive surgical margins after radical prostatectomy.12, 14, 57 Technical difficulties can also arise in the accurate delivery of RT to obese patients due to organ motion and set-up error,58–60 and this may be of particular concern with highly conformal techniques that use smaller margins. In addition, obese men may be more resistant to the testosterone-depleting effects of AST.61 Additional research is necessary to assess the relation between obesity, sex steroid levels, and clinical outcomes in men receiving AST.

Our findings must be interpreted within the context of the study design of an observational prospective cohort study. First, our patients were clinically selected on the basis of at least 1 higher-risk factor, treated similarly with RT and AST, and followed as per protocol. In order to generalize these results, they should be confirmed after other management strategies and in men with low-risk clinically localized, as well as locally advanced disease.

Second, our sample size was relatively small and we examined biochemical progression as our outcome measure. With longer follow-up and a larger number of events, the statistical strength of our study should be reinforced and allow us to comment on survival outcomes. Some types of indolent PSA failure may be an inaccurate surrogate endpoint for prostate cancer-specific mortality.62 Assessing the relation between BMI and other prognostic factors such as pretreatment PSA velocity63 and posttreatment PSA doubling time,64 as well as race and diabetes, may prove insightful. In addition, although epidemiologic data28, 29, 65 suggest a relation between obesity and mortality from prostate cancer, 1 study has suggested that an increasing BMI did not independently predict survival after radical prostatectomy.57 Whether obesity influences overall or disease-specific survival in men receiving combined modality treatment with RT and AST is unknown.

Third, as a measure of obesity we reported baseline BMI. Although widely used, BMI may be limited in its representation of body composition. Alternative measurements include waist-to-hip ratio and percentage lean/fat body mass by dual energy x-ray absorptiometry and quantitative computed tomography.8 The kinetics of weight change such as the impact of obesity earlier in life, the duration of obesity, weight changes over time caused by AST, and the impact of weight reduction on the clinical course of disease also deserves further study. AST itself is known to increase fat mass and fasting insulin levels, as well as decrease insulin sensitivity,7, 8, 66 and whether such adverse effects of therapy have an independent effect on cancer control remains to be determined.

In conclusion, an increasing baseline BMI was significantly associated with a shorter time to PSA recurrence after RT and AST in men with clinically localized prostate cancer. Further study is warranted to evaluate the mechanisms for this effect among obese men and to assess the impact of an increasing BMI after AST administration on PSA failure, prostate cancer-specific, and all-cause mortality. Whether lifestyle and dietary changes aimed at losing weight improve cancer control after treatment remains to be determined.

Ancillary