• prostate cancer;
  • obesity;
  • radiotherapy;
  • body mass index;
  • abdominal fat


  1. Top of page
  2. Abstract


Increased body mass index (BMI) has been associated with more aggressive prostate cancer (PC). The relation among abdominal visceral adipose tissue (VAT), subcutaneous adipose tissue (SAT), waist circumference (WC), and BMI was compared with clinical and pathologic findings in patients treated with radiotherapy for localized PC.


VAT, SAT, WC (all measured by planning abdominopelvic computed tomography scan) and BMI were compared with clinical and pathologic factors using univariate analyses. Cox regression analyses were performed to evaluate whether obesity measures significantly predicted risk for secondary malignancies.


Of 276 analyzed patients, 80 (29%) were obese (BMI ≥ 30 kg/m2). Median BMI at baseline was 27.6 kg/m2 (interquartile range [IQR], 25.1-30.5 kg/m2). Increased SAT and VAT were associated with a higher National Comprehensive Cancer Network (NCCN) PC risk group (P = .0001 and .008, respectively). Greater SAT was associated with a higher Gleason score (GS) (P = .030). Younger age at diagnosis was significantly correlated with higher SAT and BMI, whereas increased prostate size was found in patients with higher BMI, WC, SAT, and VAT. At a median follow-up of 42.3 months (IQR, 32.3-59.9 months), 15 secondary malignancies were observed. On multivariate analysis, VAT was a significant predictor for secondary cancers (adjusted hazards ratio, 1.014; P = .0001).


Measurements of greater abdominal adiposity were strongly associated with adverse pathologic features in patients with localized PC, including higher GS and NCCN PC risk groups. Moreover, VAT was found to be a strong risk factor for secondary malignancies. Cancer 2010. © 2010 American Cancer Society.

Obesity is not only a risk factor for different malignancies, including prostate cancer, but has also been associated with worse outcome.1-3 After radical prostatectomy (RP), patients with a higher body mass index (BMI) have been reported to have an increased risk of biochemical failure4, 5 as well as an increased risk for prostate cancer-related mortality.6 Similarly, several studies of conventional external beam radiotherapy (EBRT), with or without androgen deprivation therapy (ADT) for patients with localized prostate cancer, have found the same association between obesity and prostate-specific antigen (PSA) recurrence.7-10

Assuming these poor clinical outcomes in obese patients after curative treatments, one might expect to observe that the biology of prostate cancer would be more aggressive in this population. Dietary factors and alterations in the hormonal environment that favor the proliferation of androgen-independent prostate cancer have been proposed as possible mechanisms to explain the relation between obesity and increased risk of biochemical disease recurrence or prostate cancer-specific mortality.11-13

However, the literature does not uniformly support this unfavorable association. Indeed, some surgical14, 15 or permanent seed brachytherapy (BT) series16-18 failed to confirm obesity as a risk factor for poor clinical outcome. An increased risk of positive surgical margins after RP,19 or a higher rate of spatial shifts in prostate position and setup errors20, 21 observed during conventional EBRT among patients with higher BMI, suggests that technical problems rather than adverse features of the disease might contribute to poorer results in the treatment of obese patients.

BMI is only an indicator of overall adiposity and does not distinguish between adiposity and lean body mass, particularly in men with greater muscle mass. The measurement of abdominal fat using computed tomography (CT) images provides a more accurate and sophisticated measure of abdominal obesity than BMI.22-24 Moreover, abdominal adiposity, in particular the presence of a large abdominal visceral adipose tissue (VAT) compartment, has been found by recent epidemiologic studies to be an important risk factor for cancer development and has been linked with poor outcomes after curative treatments.25-29

Our hypothesis is that abdominal fat distribution can provide a more objective and precise measure of the relation between obesity and pathologic features of prostate cancer. As a result, we examined the association between abdominal VAT and subcutaneous adipose tissue (SAT) other than BMI and waist circumference (WC), and clinical and pathologic findings in a cohort of patients treated with curative radiotherapy (RT) for localized prostate cancer. We also examined whether obesity was linked in prostate cancer patients to a higher risk for secondary malignancies after RT.


  1. Top of page
  2. Abstract

Patient Characteristics and Treatment

We examined 276 patients with histologically confirmed localized prostate cancer treated at the University Hospital of Montreal with curative RT with or without ADT between February 2002 and February 2009. Tumor staging was based on the American Joint Commission on Cancer (AJCC) classification (sixth edition), and prostate tumor risk classes were defined according to the National Comprehensive Cancer Network (NCCN) Guidelines risk group classification.30 Approximately 9.4%, 55.4%, and 35.1% of patients presented as low-risk, intermediate-risk, and high-risk NCCN disease risk groups, respectively. None of the patients had any regional lymph node involvement (radiologically or histopathologically after lymph node sampling or dissection) or evidence of distant metastases at time of treatment, and all were required to have an Eastern Cooperative Oncology Group Zubrod performance status of 0 or 1. Patients were assessed by detailed medical history and physical examination, including a digital rectal examination and pretreatment serum PSA. The presence of comorbidities including diabetes, cardiovascular diseases (CVD), hypertension, or dyslipidemia was recorded at the time of the first consultation. Supplementary staging included a bone scan for patients with a PSA ≥10 ng/mL and a Gleason score (GS) ≥7 to exclude the presence of distant metastasis. Prostate size was determined by the referring urologist by transrectal ultrasound. Clinical and pathologic characteristics at baseline are summarized in Table 1.

Table 1. Baseline Clinical Characteristics of Observed Patients (N = 276)
  1. IQR indicates interquartile range; PSA, prostate-specific antigen; AJCC, American Joint Committee on Cancer; NCCN, National Comprehensive Cancer Network; PPC, percentage of positive cores at prostate biopsies; CVD, cardiovascular disease.

Age, y
 Median (IQR)70 (64-73) 
PSA at diagnosis, ng/mL  
 Median (IQR)7.5 (5.5-12) 
AJCC tumor classification
Gleason score
NCCN risk groups
Prostate volume, cm3 (n = 214)
 Median (IQR)37.6 (29.4-48) 
Diabetes (n = 272)
Hypertension (n = 272)
CVD (n = 272)
Dyslipidemia (n = 272)

Of the patients, 32 (11.6%) underwent exclusive I125 permanent seeds BT at 144 grays (Gy) for low-risk or intermediate-risk prostate cancer. One patient with high-risk disease was treated with BT at 110 Gy followed by 45 Gy EBRT. A total of 147 patients (53.3%) with intermediate-risk prostate cancer were randomized in a prospective multicenter phase 3 randomized trial of high-dose (79.2 Gy in 44 fractions) versus standard dose (70.2 Gy in 39 fractions) EBRT. Ninety-six patients (34.8%) with high-risk disease were treated with EBRT at 70 Gy after randomization in a Canadian phase 3 trial (PCS IV) comparing short-term (18 months) and long-term (36 months) ADT. ADT was initiated 4 months before EBRT and was comprised of a combination of 1 month of an antiandrogen (bicalutamide) and 18 or 36 months of a luteinizing hormone-releasing agonist (goserelin acetate). For the 2 phase 3 studies, patients provided written informed consent.

Patients treated with EBRT underwent a multidetector CT scan of the pelvis for treatment planning. Conversely, patients treated with BT underwent an abdominopelvic CT scan 30 days after permanent iodine seed implantation for dosimetric purposes. At the end of treatment, subjects underwent routine follow-up every 3 months for the first 2 years, every 6 months for an additional 3 years, and then annually thereafter. All patients were followed until death or to the date of most recent follow-up.

Anthropometric Measures of Obesity

BMI (defined as weight in kilograms divided by height in meters squared; kg/m2) was categorized into 4 different cohorts as previously described by others.8, 31 Patients were categorized as: 1) normal weight (BMI <25 kg/m2), 2) overweight (BMI, 25-29.9 kg/m2), 3) obese (BMI, 30-34.9 kg/m2), and 4) morbidly obese (BMI >35 kg/m2). This categorization was based on the widely recognized World Health Organization classification. Among the 276 subjects, the predominant origin was French-Canadian with none of Asian ethnicity. Weight and height were measured and recorded at the first clinical visit.

Abdominal fat adiposity was measured retrospectively from the radiotherapy planning CT images of the subjects. VAT and SAT volumes were delineated manually by 1 single observer for each subject at the iliac crest level (L4-5 interspace) on 3 slices of 3 mm in thickness using the Eclipse(Varian Associates, Palo Alto, CA) treatment planning system (Fig. 1). The approximate WC was calculated on the planning CT with this formula used to calculate the circumference of an ellipse: 2π sqrt × (latero-lateral abdominal dimension2) + (anterior-posterior abdominal dimension2)/2. Abdominal dimensions were measured at the level of the iliac crests.

thumbnail image

Figure 1. Radiotherapy planning computed tomography images are shown. (Top) Abdominal visceral adipose tissue and (Bottom) subcutaneous adipose tissue volumes were traced manually at the iliac crest level.

Download figure to PowerPoint

Statistical Analysis

Anthropometric measures of abdominal obesity including BMI as well as age, PSA level at diagnosis, and prostate volume were analyzed as continuous variables. Categorical variables included BMI (<25 kg/m2 vs 25-29.9 kg/m2 vs 30-34.9 kg/m2 vs >35 kg/m2); clinical tumor classification (T1 vs T2 vs T3); NCCN risk classes (low-risk vs intermediate-risk vs high-risk NCCN risk group); GS (≤6 vs 7 vs ≥8); and percentage of positive cores at biopsy (PPC), defined as the number of cores with any amount of cancer divided by the number of total cores (<50 vs ≥50%). The relation between obesity measures and the presence at baseline of diabetes, CVD, dyslipidemia, or hypertension was also analyzed. The correlation between BMI and different fat measurements was quantified using Pearson correlation coefficient. Clinical variables were compared across BMI groups and other anthropometric measures of obesity using analysis of variance, Student t test, or Pearson correlation coefficient for continuous variables and chi-square or Fisher exact tests for categorical variables.

Kaplan-Meier analyses were used to calculate the time to diagnosis of secondary malignancies that occurred at least 6 months after the initial diagnosis of prostate cancer. Univariate and multivariate Cox proportional regression analyses evaluating each recorded obesity measurement were performed to evaluate the effect of each variable on the risk of developing secondary cancers. The multivariate model included all covariates of interest that were found to be independently associated with the study endpoint. Statistical tests were 2-sided, and a P <.05 was considered statistically significant. Statistical analyses were performed using SPSS 17.0 statistic software package (SPSS Inc, Chicago, IL).


  1. Top of page
  2. Abstract

Baseline Clinical Characteristics

Among the 276 analyzed subjects, median age and BMI at baseline were 70 years (interquartile range [IQR], 64-73 years) and 27.6 kg/m2 (IQR, 25.1-30.5 kg/m2), respectively. Approximately 24% of patients (n = 66) were considered as having normal weight, 47% were considered as overweight (n = 130), 21% were considered as obese (n = 58), and 8% were considered as morbidly obese (n = 22). Mean (± standard deviation [SD]) VAT, SAT, and WC measures for the population were 86.5 ± 51.2 cm3, 169.5 ± 92.8 cm3, and 93.7 ± 11.1 cm, respectively. BMI values were significantly correlated with VAT (Pearson r coefficient, .621; P = .0001), SAT (Pearson r coefficient, .826; P = .0001), and WC (Pearson r coefficient, .891; P = .0001).

BMI (as a continuous and categorical variable) and SAT demonstrated a statistically significant inverse correlation with the age at prostate cancer diagnosis. In morbidly obese patients (BMI ≥ 35 kg/m2), the mean age ± SD at diagnosis was 64.6 ± 5.6 years compared with 68.2 ± 5.7, 69.2 ± 5.8, and 70.5 ± 6.3 years for obese, overweight, and normal weight patients, respectively (P = .0001). For WC, a trend nearing statistical significance toward a lower age at diagnosis was noted (P = .069).

Increased SAT (P = .0001) and VAT (P = .008) volumes were found to be significantly associated with a higher NCCN risk class at diagnosis. Conversely, this association was not found for BMI and WC. Moreover, a greater SAT was associated with a higher GS at biopsy (P = .030). SAT volumes ± SD measured were 155 ± 88 cm3 and 201 ± 120 cm3 in patients presenting with a GS ≤6 and ≥8, respectively. No relations between anthropometric measures of obesity and PSA, clinical T-classification categories, or PPC at biopsy were found.

A statistically significant correlation was found between prostate size and BMI, WC, VAT, and SAT. The mean ± SD prostate volume increased from 37.3 ± 14.8 cm3 in normal weight subjects (BMI <25 kg/m2) to 52.8 ± 41.5 cm3 in morbidly obese patients (≥35 kg/m2; P = .024). Comparisons of clinical and pathologic factors with anthropometric measures of obesity are shown in Table 2.

Table 2. Comparison of Clinical and Pathologic Factors Stratified by Anthropometric Measures of Obesity
VariableBMI, kg/m2P
<25 (n = 66)25-29.9 (n = 130)30-34.9 (n = 58)≥35 (n = 22)CategoricalaContinuousb
  • BMI indicates body mass index; WC, waist circumference; SAT, subcutaneous adipose tissue; VAT, visceral adipose tissue; PSA, prostate-specific antigen; AJCC, American Joint Committee on Cancer; NCCN, National Comprehensive Cancer Network; PPC, percentage of positive cores at prostate biopsy.

  • a

    Values for categorical variables are expressed as percent (the number of patients is shown in parentheses).

  • b

    Values for continuous variables are shown as the mean value ± the standard deviation.

  • c

    Statistically significant values are shown in bold type.

Age, y70.5 ± 6.369.2 ± 5.868.2 ± 5.764.6 ± 5.6.0001c.0001.069.006.672
PSA at diagnosis, ng/mL11.8 ± 12.611.6 ± 9.810.1 ± 7.816.1 ±
AJCC tumor classification
 T148.5 (32)39.2 (51)43.1 (25)63.6 (14).344.294.639.281.608
 T240.9 (27)43.1 (56)37.9 (22)22.7 (5)     
 T310.6 (7)17.7 (23)19 (11)13.6 (3)     
Gleason score
 ≤624.2 (16)26.2 (34)20.7 (12)27.3 (6).988.728.454.030.244
 760.6 (40)58.5 (76)62.1 (36)54.5 (12)     
 ≥815.2 (10)15.4 (20)17.2 (10)18.2 (4)     
NCCN risk groups
 Low10.6 (7)10 (13)8.6 (5)4.5 (1).984.570.392.0001.008
 Intermediate56.1 (37)55.4 (72)55.2 (32)54.5 (12)     
 High33.3 (22)34.6 (45)36.2 (21)40.9 (9)     
 <50%44.6 (30)47.7 (62)37.9 (22)50 (11).620.656.540.322.965
 ≥50%55.4 (36)52.3 (68)62.1 (36)50 (11)     
Prostate size, cm3 (n = 214)37.3 ± 14.840.1 ± 13.842.4 ± 16.652.8 ±

Comorbidities and Secondary Cancer

Diabetes, CVD, hypertension, and dyslipidemia were present at baseline in 18.4% (n = 50), in 35.3% (n = 97), in 47.8% (n = 131), and in 46.7% (n = 128) of patients, respectively. Higher values of BMI, WC, SAT, and VAT were significantly associated with the presence of diabetes at diagnosis (P = .022-.0001). Except for SAT, the same association with anthropometric measures of obesity was also found for the presence of hypertension at baseline. Conversely, CVD and dyslipidemia did not demonstrate any association with the different obesity measures. The relation between comorbidities and obesity parameters are illustrated in Table 3.

Table 3. Comparison of Comorbidities Stratified by Anthropometric Measures of Obesity (N = 272)
VariableBMI, kg/m2P
(n = 65)(n = 127)(n = 58)(n = 22)BMIBMIWCSATVAT
  • BMI indicates body mass index; WC, waist circumference; SAT, subcutaneous adipose tissue; VAT, visceral adipose tissue; CVD, cardiovascular disease.

  • a

    Values for categorical variables are expressed as percent (the number of patients is shown in parentheses).

  • b

    Values for continuous variables are shown as the mean value ± the standard deviation.

  • c

    Statistically significant values are shown in bold type.

 Yes10.8 (7)18.1 (23)22.4 (13)31.8 (7).122.0002c.0001.022.001
 No89.2 (58)81.9 (104)77.6 (45)68.2 (15)     
 Yes29.2 (19)50.4 (64)58.6 (34)59.1 (13).004.001.0001.070.001
 No70.8 (46)49.6 (63)41.4 (24)40.9 (9)     
 Yes35.4 (23)32.3 (41)46.6 (27)22.7 (5).155.749.823.269.077
 No64.6 (42)67.7 (86)53.4 (31)77.3 (17)     
 Yes40 (26)48 (61)50 (29)50 (11).657.261.523.982.351
 No60 (39)52 (66)50 (29)50 (11)     

The median follow-up was 42.3 months (IQR, 32.3-59.9 months). At the time of last follow-up, 15 (5%) secondary malignancies were observed. Median time to diagnosis of second cancer after end of RT was 29.5 months (IQR, 25.2-35.4 months). Metachronous cancer was most likely present in 1 patient diagnosed with a bladder cancer 9.3 months after prostate biopsy. Rectal, bladder, and esophageal carcinomas were diagnosed in 5, 3, and 2 patients, respectively. The remaining patients presented with lung, oropharyngeal, and colon cancer, whereas 1 presented with a hematologic malignancy and 1 with an adenocarcinoma of unknown origin. For the entire population, the estimated 4-year second cancer-free survival rate was 93% ± 1.9%. As listed in Table 4, univariate Cox regression analyses for VAT, WC, and BMI were significantly associated with an increased risk of secondary malignancies. SAT demonstrated only a trend toward a higher risk (P = .063). Conversely, age at diagnosis, use of ADT, and baseline comorbidity factors did not appear to predict the development of secondary cancers. On multivariate analysis, only VAT continued to significantly predict an increased risk of being diagnosed with subsequent malignancy (adjusted hazards ratio, 1.014; P = .0001).

Table 4. Cox Regression Analysis of Second Cancers
VariableUnivariate AnalysisMultivariate Analysis
HR (95% CI)PHR (95% CI)P
  • HR indicates hazard ratio; 95% CI, 95% confidence interval; BMI, body mass index; WC, waist circumference; SAT, subcutaneous adipose tissue; VAT, visceral adipose tissue; ADT, androgen deprivation therapy; CVD, cardiovascular disease.

  • a

    Continuous variable.

  • b

    Statistically significant values are shown in bold type.

  • c

    Categorical variable.

  • d

    HR was not given for a P > .2.

BMI, kg/m2a1.092 (1.014-1.176).020b.853
BMI, kg/m2c
 30-34.94.581 (0.511-41.050).174.745
 ≥3512.965 (1.48-116.06).022.545
WC, cm a1.057 (1.019-1.097).003.695
SAT, cm3a1.004 (1.000-1.008).063  
VAT, cm3a1.014 (1.007-1.021).00011.014 (1.007-1.021).0001
Age at diagnosis, y a.991  
ADT usec.167  
Dyslipidemiac0.506 (0.180-1.425).197  
Diabetesc0.429 (0.146-1.255).122  


  1. Top of page
  2. Abstract

More aggressive cancer biology has been proposed as a possible reason for the increased risk of biochemical disease recurrence and prostate cancer deaths in obese patients treated curatively by either RP or conventional EBRT. However, recent studies addressing this issue using BMI estimates indicate contradictory results. Some prostate biopsy31-34 and prostatectomy series19, 35, 36 have shown more aggressive features among obese patients, whereas no such effect was observed by other authors (Table 5).

Table 5. Comparison of Previous Publications on Anthropometric Measures of Obesity and Clinical and Pathologic Features
StudyNo.aPercent Obese (BMI ≥30 kg/m2)Biopsy or RP SeriesAnalyzed Obesity VariableMore Aggressive Prostate Cancer Features in Obese Versus Nonobese PatientsClinical and Pathologic Features With Significant Association
  • BMI indicates body mass index; RP, radical prostatectomy; PSA, prostate-specific antigen; DRE, digital rectal examination; GS, Gleason score; PSM, positive surgical margins; WC, waist circumference; SAT, subcutaneous adipose tissue; VAT, visceral adipose tissue; NCCN, National Comprehensive Cancer Network.

  • a

    Data in parentheses represent the percentage of patients diagnosed with cancer.

Bittner 200931244 (45.9)34BiopsyBMINoAge
Freedland 20053278732.3BiopsyBMIYesProstate cancer detection, age, PSA, abnormal DRE, prostate size, GS
Gallina 200933181410.6Biopsy + RPBMINoGS at RP, but not at biopsy
Pruthi 200934500 (45)26BiopsyBMINoAge, abnormal DRE, prostate size
Freedland 200935230227.6RPBMIYesAge, PSA, prostate size, GS, PSM, tumor volume
Jayachandran 200819143429.3RPBMIYesAge, PSA, prostate size, GS, PSM
Isbarn 200936153811RPBMINoPSA
Current series27629BiopsyBMI, WC, SAT, VATYesAge, prostate size, GS (SAT), NCCN risk group (SAT, VAT)

Why should prostate cancer in men with higher BMI be more aggressive than in a normal weight population, and why are results in the literature so discordant? These differences might be explained by the fact that BMI is not a reliable estimate of the impact of obesity on prostate cancer patients. A recent large prospective cohort study found that central adiposity was a stronger predictor of prostate cancer than general obesity measured by BMI.37 As observed in 1 population-based study, higher abdominal adiposity was associated with an increased risk of high-grade and advanced prostate cancer, especially among men with low BMI.38 Patients with higher VAT amounts were also at greater risk of being diagnosed with prostate cancer.25

Based on these results, it appears that abdominal fat may more effectively predict aggressive disease biology than BMI. Indeed, adipose tissue produces several hormones and cytokines such as interleukin-6 that have been found to be involved in prostate cancer growth and progression.39-41 Moreover, low testosterone levels in obese patients can contribute to the development of hormone-refractory aggressive prostate cancers.4, 8

To test this hypothesis, we examined the impact of abdominal fat compartments on clinical and histopathologic findings in prostate cancer patients treated with curative RT. To our knowledge, the link between abdominal adiposity measured on CT scan and prostate cancer has yet to be examined. We found that SAT and VAT were significantly associated with a higher NCCN prostate-risk group, whereas no association was found for BMI or WC. Moreover, a higher SAT was found to be associated with a higher GS at diagnosis. It is interesting to note that in our patients, an association with more aggressive characteristics (GS and NCCN prostate-risk group) was observed for SAT and VAT, but not for BMI. This might explain why some previous studies,31, 33, 34, 36 which analyzed only the influence of BMI and not of fat distribution on adverse pathologic features, failed to find such an association (Table 5). The results of the current study highlight the advantage of using more sophisticated anthropometric measures of abdominal adiposity than simple BMI to evaluate the influence of obesity in prostate cancer patients. This is true even more so for older patients, because our patients have a median age of 70 years. It is known that above the age of 60, body weight generally tends to decrease mainly due to loss of muscle mass, whereas body adipose tissue tends to be redistributed toward the abdominal visceral compartment.42

In the current study, morbidly obese subjects (BMI ≥ 35 kg/m2) were younger at the time of diagnosis than patients with a lower BMI. Moreover, SAT volumes were generally found to be inversely correlated with age at diagnosis. These results are in agreement with other published studies,5, 8, 9, 43 suggesting that the development of prostate cancer may start earlier or progress more rapidly in obese patients. Conversely, this analysis did not reveal any association between the different anthropometric measures of obesity and adverse pathologic features of the tumor such as PSA, clinical T classification, or PPC, as observed in other studies.9, 17, 31

RT with concomitant ADT is the mainstay of therapy for men with locally advanced disease.44, 45 However, this treatment has been associated with an increased risk of diabetes and CVD, as observed in 2 large population-based cohorts studies.46, 47 Moreover, in a randomized trial of patients with locally advanced prostate cancer treated with RT and long-term ADT, prevalent diabetes was significantly associated with greater all-cause and nonprostate cancer mortality.48 Within our series, obese patients were more likely to present not only with a higher rate of diabetes and hypertension before curative treatment, but also with more aggressive prostate cancers. Consequently, decisions regarding the use of ADT in obese patients should consider the advantages of improvements in cancer-specific outcomes against the potentially increased risk of serious side effects of ADT. Further studies are warranted to explore whether the benefits of ADT outweigh its risks in obese patients.

This study suggests that obese men have a higher risk of being diagnosed with secondary cancer after an initial diagnosis of prostate cancer. The data from the current study add evidence that obesity, and particularly VAT, may be associated with an increased risk of secondary malignancies in patients previously diagnosed with prostate cancer. Moreover, the relatively short follow-up (ie, 42.3 months) of these patients highlights these results, excluding a confounding action of RT in the development of secondary primaries.49-51 Obese men may need more extensive staging and closer follow-up to detect subsequent new malignancies.

There are inherent limitations in this study that are important to underline. First, our findings are from a single center and patients are mainly (88.4%) from randomized studies; thus, we cannot exclude a potential selection bias in our patients. Therefore, one has to be cautious when drawing inferences regarding the general prostate cancer population. Second, no central pathologic review of the biopsy specimens was performed and therefore, the influence of SAT on the GS must be interpreted with caution. Moreover, the estimation of WC on CT images using a formula to calculate an ellipse warrants further comparison with standard manual measurements in a standing position. WC only had an influence on prostate size, as had all other measures of abdominal adiposity and BMI. We cannot exclude that if WC had been measured with a measuring tape that we would have found an influence on other factors. We did not report the influence of the different measurements of obesity on oncologic outcomes. However, in a subgroup analysis of the patients with intermediate-risk prostate cancer after image-guided radiotherapy techniques, we were unable to demonstrate that obesity or a specific fat distribution predisposes patients toward a higher risk of biochemical failure (unpublished data). In agreement with other authors, we believe that the observed increase in biochemical failure in obese patients in previous EBRT studies may be more closely related to inaccurate dose delivery than to more aggressive disease.

In conclusion, these findings suggest that abdominal adiposity is associated with adverse pathologic features in patients with localized prostate cancer, including higher GS and NCCN risk groups, and an increased risk of secondary malignancies. Further studies are warranted to more accurately assess whether the more aggressive features observed in obese patients translate into adverse clinical outcomes after curative treatments.


  1. Top of page
  2. Abstract