Prostate Cancer and Prostatic Disease (2014) 17, 233–237 & 2014 Macmillan Publishers Limited All rights reserved 1365-7852/14 www.nature.com/pcan
ORIGINAL ARTICLE
BMI is associated with larger index tumors and worse outcome after radical prostatectomy N Hayashi1, M Matsushima2, M Kido1, T Naruoka1, A Furuta1, N Furuta1, H Takahashi3 and S Egawa1 BACKGROUND: To investigate the impact of body mass index (BMI) on tumor characteristics and biochemical recurrence (BCR) after radical prostatectomy (RP) for prostate cancer (PCa) in Japanese men. METHODS: We evaluated data from consecutive patients who had undergone RP. Data analyzed included age, preoperative serum PSA, prostatic volume, BMI (continuous or categorized (p25 kg/m2) values), clinical and pathological findings including index tumor volume (ITV), and current status in areas such as smoker or nonsmoker and presence or absence of diabetes. We analyzed association between BMI and BCR, especially based on ITV using univariate and multivariate analysis. RESULTS: We analyzed data from a total of 703 patients. The median follow-up time was 38.4 months. BCR was diagnosed in 154 patients (21.9%) at a median of 9.7 months postoperatively. Multivariate linear regression analysis adjusted for preoperative variables showed a significant positive association between BMI and ITV (continuous BMI: P ¼ 0.002; categorical BMI: Po0.001, respectively), especially for higher-grade tumors (Gleason score X 7). Cox proportional hazards analysis showed a significant association between continuous BMI and BCR after surgery (preoperative variables, hazard ratio (HR) 1.09, 95% confidence interval (CI) 1.02–1.16, P ¼ 0.008), independent of clinical and pathological findings. In patients with high-risk cancer, the positive association between BMI and BCR was strengthened (preoperative variables, continuous BMI, HR 1.16, 95% CI 1.07–1.26, Po0.001; categorical BMI, HR 2.11, 95% CI 1.29–3.45, P ¼ 0.003, respectively). CONCLUSIONS: Greater BMI significantly correlates with higher rates of BCR after surgery; BMI is a preoperative variable associated with high-grade ITV. Our results suggest that the biological environment created by greater BMI may contribute to increasing tumor aggressiveness. Prostate Cancer and Prostatic Disease (2014) 17, 233–237; doi:10.1038/pcan.2014.15; published online 20 May 2014
INTRODUCTION Obesity has been associated with increased death rates from many malignant tumors including esophagus, colon, rectum, liver, gallbladder, pancreas, kidney and prostate in a prospectively studied cohort of US adults.1 It has also been related to decreased risk of low-grade disease and increased risk of high-grade and advanced prostate cancer (PCa).2–7 However, data on the link between obesity and PCa risk are mostly from western rather than Asian countries. One study from Korea failed to show such an association;8 the researchers attributed the negative findings partly to their leaner study population, with a lower body mass index (BMI) profile than in the west, and suggested that the impact of BMI may differ between Asian and western ethnicities. We previously reported that the prevalence and mean tumor volume of latent PCa have increased over the last 20 years (1983– 1992 vs 2009–2011), from 20.8% (104/500) to 43.0% (46/107, Po0.0001), and 230.4 to 1111.1 mm3, respectively.2 Although patients showing significant tumor volume (4500 mm3) accounted for only 9.6% of cases in 1983–1992, 26.1% were beyond this size in 2009–2011 (P ¼ 0.017). The cause and interpretation of these autopsy findings in Asia is still unclear. However, as one of the reasons, changes of Asian’s lifestyle cause the recent increase of those with higher BMI and could secondarily affect to patients with PCa.3 We thus hypothesized that the increase in BMI together with other confounding factors 1
such as resultant metabolic syndrome may have had a role in the recent change in the PCa landscape in Asia.4 To further investigate the biological effects of obesity on tumor characteristics and treatment outcome, we examined the relationship between BMI, pathological findings and biochemical recurrence (BCR) after radical prostatectomy (RP) for PCa in our Japanese cohort, especially based on the volume of tumor with higher Gleason score (GS) to be a powerful predictor of long-term outcome.5 MATERIALS AND METHODS Study population and specimens We evaluated data from 793 consecutive patients who had undergone RP (open or laparoscopic) at Jikei University Hospital between June 2002 and December 2009. The study design was reviewed and approved by the institutional review board. Of these, patients with neoadjuvant or adjuvant endocrine therapy and/or with a follow-up time of o6 months were excluded from data analysis. Patients treated with adjuvant radiotherapy (RT) were included in our study.6 A total of 703 patients were thus analyzed in this study. All patients underwent preoperative clinical evaluation, including digital rectal examination, chest radiography, computed tomography or magnetic resonance imaging of the abdomen and pelvis, and bone scanning. All RP samples were histologically examined by a single pathologist (HT),7 and GS was assigned. Clinical tumor stage and extent were determined according to the General Rule for Clinical and Pathological Studies on Prostate Cancer.9,10 Briefly, the prostatic capsule on
Department of Urology, Jikei University School of Medicine, Tokyo, Japan; 2Division of Clinical Research and Development, Jikei University School of Medicine, Tokyo, Japan and Department of Pathology, Jikei University School of Medicine, Tokyo, Japan. Correspondence: Dr N Hayashi, Department of Urology, Jikei University School of Medicine, 3-25-8 Nishi-Shimbashi, Minato-ku, Tokyo 105-8461, Japan. E-mail: [email protected] Received 2 May 2013; revised 20 February 2014; accepted 21 February 2014; published online 20 May 2014 3
Radical prostatectomy analyzed by BMI N Hayashi et al
234 each slide was carefully inspected microscopically for areas of complete cancer penetration through the prostatic surface (extraprostatic extension, EPE). Tumor cells at the inked margin of resection were considered to indicate a positive surgical margin (PSM). Specimens were also assessed for seminal vesicle invasion (SVI) or lymph nodal involvement. The volume of the index tumor (ITV, cm3, largest single focus of the tumor) was calculated using the formula 0.52 length width cross-sectional thickness. Each GS in ITV was also determined by a pathologist. In risk group stratification, patients with clinical stage T1c or T2a, PSA p10 ng ml–1 and biopsy GS p 6; patients with clinical stage T2b, PSA 10.1–20 ng ml–1 and biopsy GS 7; or patients with X T2c, a PSA 420 ng ml–1 or a biopsy GS X8 were defined as low risk, intermediate risk or high risk, respectively.11 BMI (kg/m2) was obtained by self-reported. Diabetes is diagnosed by physicians (history and/or medication). Smoking status was estimated by questionnaire.
Follow-up Patients were generally followed up with PSA testing every 3 months for 1 year, every 6 months for 2 years and then yearly, and physical examinations or other tests were added at intervals as clinically indicated. The median follow-up time was 38.4 months (range: 6–120 months). BCR was defined as any PSA increase to 0.2 ng ml–1.
Statistical analysis We conducted descriptive frequency analysis to evaluate the clinical and pathological characteristics of patients after surgery. Data included in the analysis were patient age, preoperative PSA, BMI, diabetes status, current smoker, preoperative variables (biopsy GS, clinical stage), postoperative variables (prostate weight, specimen GS, and pathological findings
Table 1.
including EPE, SVI, lymph nodal involvement, PSM and ITV), and adjuvant RT. We used the Student’s t-test or Wilcoxon rank sum test for continuous variables and the w2 test for categorical variables as appropriate for the descriptive statistics. To further identify the variables that could affect all subjects, and those that affected GS p6, GS 7 or GS X8 tumors subjects only, we controlled for confounders such as patient age, PSA, biopsy GS (p7 vs X8), clinical stage (cT1 vs cT2–3), and continuous or categorized (X 25) BMI in multivariate linear regression analysis (Tables 2 and 3). We analyzed correlations between clinicopathological covariates (preoperative variables: adjuvant RT, patient age, PSA, BM, current smoker, diabetes status, clinical stage (cT1 vs cT2–3) and biopsy GS (p7 vs X8); postoperative variables included: adjuvant RT, patient age, PSA, BMI, clinical stage (cT1 vs cT2–3), pathological GS (p7 vs X8), PSM, EPE and SVI) on BCR-free survival after surgery, by calculating adjusted hazard ratios (HRs) and the corresponding 95% confidence interval (CI) using a Cox proportional hazards model (Tables 4 and 5). The Spearman’s rank-order correlation coefficient between continuous and categorized BMI and postoperative variables was assessed. All tests were two-tailed, with Po0.05 considered significant. All statistical analyses were performed using both of SPSS, version 17.0 (Chicago, IL, USA) and STATA version 10.0 (College Station, TX, USA).
RESULTS Patient demographics Table 1 lists the demographic characteristics of our study population of 703 patients stratified by BMI (o25 vs X25 kg/m2). Patient age ranged from 47 to 77 years (mean 64, median 64). PSA levels ranged from 3.2 to 65.9 ng dl–1 (mean 10.0, median 8.2).
Patient characteristics BMI
Patients no. Age (years) median/mean PSA (ng ml–1) median/mean BMI (kg/m2) median/mean Current smokers, no. (%) Diabetes, no. (%) Clinical stage, no (%) T1 T2 T3 Biopsy GS, no. (%) 2–6 7 8–10 Risk groups, no. (%) Low Intermediate High EPE, no. (%) SVI, no. (%) LNI, no. (%) (total, n ¼ 585) PSM, no. (%) Pathological GS, no. (%) 2–6 7 8–10 ITV (cm3) median/mean GS p6 tumors (cm3) median/mean GS 7 tumors (cm3) median/mean GS X8 tumors (cm3) median/mean Prostatic weight (g) median/mean Adjuvant RT, no. (%)
Total
Less than 25
25 or Greater
703 64/64 8.2/10.0 23.5/23.6 116 (16.5) 67 (9.5)
516 65/64.4 8.0/9.7 22.6/22.3 73 (14.1) 44 (8.5)
187 63/63.2 8.5/10.6 26.4/27.0 43 (23.0) 23 (12.3)
508 (72.3) 188 (26.7) 7 (1.0)
365 (70.7) 146 (28.3) 5 (1.0)
143 (76.5) 42 (22.5) 2 (1.0)
259 (36.8) 293 (41.7) 151 (21.5)
196 (38.0) 210 (40.7) 110 (21.3)
63 (33.7) 83 (44.4) 41 (21.9)
187 313 203 240 65 20 242
142 231 143 176 46 14 169
45 82 60 64 19 6 73
P-value 0.024 o0.001 o0.001 0.013 0.143 0.122
0.471
0.268 (26.6) (44.5) (28.9) (34.1) (9.2) (3.4) (34.4)
117 (16.6) 450 (64.0) 136 (19.3) 1.8/3.7 0.5/1.8 2.1/3.8 2.7/5.3 38.0/37.8 51 (7.3)
(27.5) (44.8) (27.7) (34.1) (8.9) (2.7) (32.8)
84 (16.3) 335 (64.9) 97 (18.8) 1.6/3.2 0.5/1.8 1.7/3.2 2.5/4.8 37.5/41.7 29 (5.6)
(24.1) (43.9) (32.1) (34.2) (10.2) (3.2) (39.0)
33 (17.6) 115 (61.5) 39 (20.9) 2.6/4.9 0.8/1.8 3.1/5.3 3.1/6.3 38.5/41.8 22 (11.8)
0.977 0.63 0.683 0.129 0.842
o0.001 o0.001 o0.001 o0.001 0.995 0.018
Abbreviations: BMI, body mass index; EPE, extra-prostatic extension; GS, Gleason score, ITV, index tumor volume; LNI, lymph nodal involvement; PSM, positive surgical margin; RT, radiotherapy; SVI, seminal vesicle invasion. Po0.05: significant.
Prostate Cancer and Prostatic Disease (2014), 233 – 237
& 2014 Macmillan Publishers Limited
Radical prostatectomy analyzed by BMI N Hayashi et al
235 BMI ranged from 16.0 to 37.5 kg/m2 (mean 23.6, median 23.5). Eleven (1.6%) patients had BMI of 30 kg/m2 or greater. ITV ranged from 0 to 25.6 ml (mean 3.7, median 1.8). PCa was classified preoperatively as low risk in 187 patients, intermediate risk in 313 patients and high risk in 203 patients, including 151 patients with biopsy GS X8, 40 patients at clinical stage T2c or higher, and 49 patients with PSA X 20 ng dl–1. Lymph nodal involvement was detected in 20 (3.4%) of the 585 patients who had lymph node dissection. When stratified by BMI, statistical differences were detected for patient age, PSA, current smoker status, ITV, GS p6 tumors, GS 7 tumors, GS X8 tumors and adjuvant RT (Po0.05). Association of BMI with ITV Both continuous and categorical (o 25 vs X25 kg/m2) BMI showed a significant but weak correlation with continuous ITV (Spearman’s rank-order correlation coefficient r ¼ 0.135, Po0.001, and r ¼ 0.136, Po0.001, respectively), and not with continuous PSA, pathological GS p7 vs X8, PSM, EPE or SVI (P ¼ 0.141 and 0.117, 0.661 and 0.466, 0.053 and 0.139, 0.857 and 0.999, or 0.189 and 0.546, respectively). In contrast, continuous ITV was significant with all variables. In multivariate linear regression analyses, continuous or categorical BMI was also significantly associated with continuous ITV, accounting for other preoperative variables (P ¼ 0.002, Po0.001, respectively). In the adjusted model, including the interaction term for continuous BMI and higher grade (X GS 7) tumors for ITV, the interaction was still significant (P ¼ 0.03). We then performed subgroup analysis of ITVs, stratified by each GS, using multivariate linear regression analysis. Continuous and categorical BMI showed significant associations both with GS 7 and GS X8 tumors but no association with GS p6 tumors (Table 2). A significant trend for the association between higher grade (GS X7) tumors and continuous and categorical BMI was similarly seen in patients with pathologically localized (p pT2) cancer or negative surgical margin cancer (Table 3). Multivariate Cox proportional hazards analysis for BCR after RP BCR was encountered in 154 patients (21.9%) at a median of 9.7 months (range 0 to 98.6 months) postoperatively. Higher PSA was associated with increasing HR of BCR (preoperative variables: HR 1.05, 95% CI 1.03–1.06, P ¼ 0.001; postoperative variables included: HR 1.03, 95% CI 1.02–1.05, Po0.001, respectively, Table 4).
Similarly, the adjusted HR for BCR in patients with higher clinical stage (cT2–3), biopsy GS X8 as preoperative variables, and pathological GS X8, PSM, EPE, and SVI as postoperative variables was 2.07, 1.96 and 1.89; 1.71, 1.79 and 2.28 (Po0.001, o0.001, ando0.001, 0.02, 0.003, o0.001, respectively). Continuous BMI was positively associated with increasing HR of BCR (preoperative variables: HR 1.09, 95% CI 1.02–1.16, P ¼ 0.008; postoperative variables included: HR 1.07, 95% CI 1.01–1.14, P ¼ 0.01, respectively). In contrast, categorical BMI failed to show significant association with BCR (o25 vs X25 kg/m2, preoperative variables and postoperative variables included, P40.05). Impact of BMI on BCR according to patient subsets In the adjusted model, including the interaction term for continuous BMI and high-risk cancers, the interaction was still significant (P ¼ 0.011). To further define those patients for whom BMI was most closely associated with BCR, we performed Cox proportional hazards analysis for specific subsets of patients. Both
Table 3. Multivariate linear regression analysis in patients with pathological stage pT2 or negative surgical margin Higher-grade (GS X7) tumors (cm3) b-Coefficient
S.e.
P-value
Pathological T2 stage or less (n ¼ 463) Continuous BMI 0.107 Categorical BMI 0.712
0.043 0.26
0.013 0.006
Negative surgical margin (n ¼ 461) Continuous BMI 0.064 Categorical BMI 0.537
0.043 0.267
0.138 0.045
Abbreviations: BMI, body mass index; GS, Gleason score. Po0.05: significant.
Table 4.
Multivariate Cox proportional hazards analysis for BCR-free
survival HR (95% CI)
P-value
Preoperative variables Adjuvant RT Continuous age (years) Continuous PSA (ng ml–1) Continuous BMI (kg/m2) Biopsy GS p7 vs X8 Clinical stage T1 vs T2–3 Current smoker Diabetes status
0.92 (0.53–1.61) 1.01 (0.98–1.04) 1.05 (1.03–1.06) 1.09 (1.02–1.16) 1.96 (1.37–2.80) 2.07(1.48–2.90) 0.94 (0.62–1.44) 0.53 (0.26–1.10)
0.78 0.65 0.001 0.008 o0.001 o0.001 0.78 0.09
Postoperative variables included Adjuvant RT Continuous age (years) Continuous PSA (ng ml–1) Continuous BMI (kg/m2) Pathological GS p7 vs X8 PSM EPE SVI
0.54 1.00 1.03 1.07 1.89 1.71 1.79 2.28
0.06 0.67 o0.001 0.01 o0.001 0.02 0.003 o0.001
Variables
Table 2.
Multivariate linear regression analysis between BMI and each
ITV b-Coefficient
S.e.
P-value
0.14 1.13
0.05 0.28
0.002 o0.001
GS p6 tumors (cm3), (n ¼ 136) Continuous BMI 0.04 Categorical BMI 0.01
0.09 0.57
0.65 0.98
GS 7 tumors (cm3), (n ¼ 447) Continuous BMI Categorical BMI
0.17 1.36
0.06 0.34
0.002 o0.001
GS X8 tumors (cm3), (n ¼ 120) Continuous BMI Categorical BMI
0.29 1.96
0.13 0.74
0.027 0.009
3
ITV (cm ) Continuous BMI Categorical BMI
Abbreviations: BMI, body mass index; GS, Gleason score; ITV, index tumor volume. Po0.05: significant.
& 2014 Macmillan Publishers Limited
(0.29–1.03) (0.98–1.03) (1.02–1.05) (1.01–1.14) (1.32–2.70) (1.21–2.42) (1.22–2.63) (1.47–3.55)
Abbreviations: BCR, biochemical recurrence; BMI, body mass index; CI, confidence interval; EPE, extra-prostatic extension; GS, Gleason score; HR, hazard ratio; PSM, positive surgical margin; RT, radiotherapy; SVI, seminal vesicle invasion. Po0.05: significant.
Prostate Cancer and Prostatic Disease (2014), 233 – 237
Radical prostatectomy analyzed by BMI N Hayashi et al
236 Table 5.
Multivariate Cox proportional hazards analysis in patients with each risk group HR (95% CI)
P-value
1.08 (0.91–1.27) 0.59 (0.17–2.03)
0.39 0.40
Intermediate-risk group Continuous BMI Categorical BMI
1.02 (0.92–1.13) 0.84 (0.45–1.56)
0.75 0.58
High-risk group Continuous BMI Categorical BMI
1.16 (1.07–1.26) 2.11 (1.29–3.45)
o0.001 0.003
Variables Preoperative variables Low-risk group Continuous BMI Categorical BMI
Postoperative variables included Low-risk group Continuous BMI 1.10 (0.88–1.39) Categorical BMI 0.62 (0.13–3.03)
0.40 0.55
Intermediate-risk group Continuous BMI Categorical BMI
1.05 (0.93–1.18) 0.99 (0.47–2.08)
0.42 0.98
High-risk group Continuous BMI Categorical BMI
1.15 (1.06–1.25) 1.97 (1.20–3.22)
0.001 0.007
Abbreviations: BMI, body mass index; CI, confidence interval; HR, hazard ratio. Po0.05: significant.
continuous and categorical BMI showed significant association with BCR in patients with high-risk cancer only (preoperative variables: HR 1.16, 95% CI 1.07–1.26, Po0.001, and HR 2.11, 95% CI 1.29–3.45, P ¼ 0.003, and postoperative variables included: HR 1.13, 95% CI 1.04–1.22, P ¼ 0.001, and HR 1.84, 95% CI 1.13–2.98, P ¼ 0.01, respectively, Table 5). No association was found in patients with low- and intermediate-risk cancers (P40.05). DISCUSSION The association between greater BMI and cancer outcome is one of the major problems in health care. Several western studies have shown a significant association between greater BMI and increased total tumor volume, worse pathological findings, especially higher GS, and increased risk of BCR after RP for PCa.12–16 However, insufficient information is available on Asian men. Lee et al. previously found that higher BMI did not significantly enhance the ability to preoperatively predict tumor EPE and was not significantly associated with other adverse pathological findings including GS or BCR in Korean men who had undergone RP.8 They suggest that BMI up to a certain threshold (around 30 kg/m2) is not significant with respect to tumor aggressiveness. These findings remain controversial and have generated numerous hypotheses, predominantly owing to the lack of direct, sound explanations and to the presence of numerous ethnic and lifestyle confounders including aging, diabetes status and smoking habits.17 Over the last 30 years, the Japanese lifestyle has shifted toward a more ‘westernized’ diet of meat and poultry rather than fish and shellfish, and increased consumption of animal fat.18 Counterintuitively, daily energy intake per person declined by approximately 300 kcal day–1 during the same period. This may be due in part to modernization and an aging society.18 Fewer calories are consumed by elderly people and by fashion-conscious young women in pursuit of a slender body. Nevertheless, obesity has Prostate Cancer and Prostatic Disease (2014), 233 – 237
become more common; today around 30% of Japanese male adults are obese or overweight.3 Men with higher BMI tend to have decreased serum levels of testosterone and increased estrogen, and it seems feasible that PCa may be less androgen dependent in such men.13 Adiposity associated with higher BMI may also contribute to aggressive PCa growth by eliciting the secretion of certain hormones or growth factors.19,20 Positive associations between obesity and castration resistance or metastasis after hormone treatment have also been reported.21 In a recent microarray analysis, higher BMI was associated with higher gene expression of many inflammatory transcripts related to the nuclear factor kappa B pathway, possibly encouraging aggressive cancer biology or treatment resistance.8 These findings support the hypothesis that higher BMI provides a favorable biological environment for the survival and growth of aggressive or high-grade tumors. Both continuous and categorical BMI were significantly associated with continuous ITV. However, in multivariate analysis, greater BMI showed significant association only with higher-grade (GS X7) subsets but not with low-grade (GS p6) subsets. A significant trend toward this positive association was also seen in patients with p pT2 disease or negative surgical margin. An association between greater BMI and larger total tumor volume has also been reported in recent studies.22,23 In addition, the trend showed by our results seems to be the positive association with GS X8 strongly 4 GS 7. Therefore, patients classified by risk groups including GS X8 as a variable were analyzed. Patients with greater categorical BMI (X 25 kg/m2) had an increased HR for BCR after surgery in the high-risk group only. Previous findings showed that high triglyceride levels were significantly associated with a higher incidence of biopsy GS X8 in patients aged X60 years.24 Our current findings are in agreement with this, since hypertriglyceridemia is one of the manifestations of metabolic syndrome.24 Although BMI was unrelated to tumor grade, among men with higher grade ITV, those with higher BMI had larger tumors. Our results thus suggest that higher preoperative BMI can predict the tumor volume of higher grade cancer and could be associated with worse outcomes after surgery in high-risk patients. These findings disagree with those from the Korean study, for reasons that remain obscure at present.8 A single pathologist examined all histological slides and assigned GS in this study; the details of pathological assessment were not noted in the Korean study.25 We did not find smoking habits or diabetes to be independent predictors of tumor aggressiveness in this study. The limitations of our study are its retrospective nature and the limited number of patients. Definitive conclusions will require information over a much longer follow-up period, up to metastasis and cancer death. As in the Korean study, we enrolled only a small number of patients with BMI 30 kg/m2 or greater; because of the small number of those patients in our study, we did not evaluate the categorical data from those patients.8 We did not assess the relationship between serum hormones, growth factor levels and BMI. We cannot provide a sound explanation for our previous autopsy findings of the changes in latent tumor volume and prevalence over the last 20 years based on current observations.12 Longitudinal, prospective, well-controlled studies will be needed to elaborate and confirm these findings. In conclusion, this study showed a positive correlation between BMI and high-grade ITV and BCR after surgery. Greater BMI had a significant impact on BCR after surgery in the high-risk subgroup only. The clinical relevance of BMI may thus be enhanced in patients with high-risk cancer. High-risk patients with higher BMI may be candidates for more aggressive curative treatment. CONFLICT OF INTEREST The authors declare no conflict of interest.
& 2014 Macmillan Publishers Limited
Radical prostatectomy analyzed by BMI N Hayashi et al
237 REFERENCES 1 Calle EE, Rodriguez C, Walker-Thurmond K, Thun MJ. Overweight, obesity, and mortality from cancer in a prospectively studied cohort of U.S. adults. N Engl J Med 2003; 348: 1625–1638. 2 Okayasu M, Takahashi H, Furusato B, Egawa S. Comparison of latent prostate cancer detected at autopsy between pre-and post-prostatic specific antigen (PSA) era. Mod Pathol 2012; 25: 231A. 3 Kadota A, Miura K, Ueshima H. [The trend of metabolic syndrome in Japanese National Health and Nutrition Survey]. Nihon Rinsho 2011; 69(Suppl 1): 35–39. 4 Cullen J, Elsamanoudi S, Brassell SA, Chen Y, Colombo M, Srivastava A et al. The burden of prostate cancer in Asian nations. J Carcinog 2012; 11: 7. 5 Albertsen PC. What is the risk posed by prostate cancer? J Natl Cancer Inst Monogr 2012; 2012: 169–174. 6 Raldow A, Hamstra DA, Kim SN, Yu JB. Adjuvant radiotherapy after radical prostatectomy: evidence and analysis. Cancer Treat Rev 2011; 37: 89–96. 7 Hayashi N, Urashima M, Ikemoto I, Kuruma H, Arai Y, Kuwao S et al. Prostate-specific antigen adjusted for total prostatic tumor volume as a predictor for outcome after radical prostatectomy. Prostate Cancer Prostatic Dis 2007; 10: 60–65. 8 Freedman JE, Larson MG, Tanriverdi K, O’Donnell CJ, Morin K, Hakanson AS et al. Relation of platelet and leukocyte inflammatory transcripts to body mass index in the Framingham heart study. Circulation 2010; 122: 119–129. 9 Kanahara (ed.) General Rule for Clinical and Pathological Studies on Prostatic Cancer. 3 edn. Japanese Urological Association and Japanese Pathological Society: Tokyo, 2001. 10 Sbin L, W. C (ed). UICC International Union Against Cancer, 2002. Wiley-Liss: New York. 11 D’Amico AV, Whittington R, Malkowicz SB, Schultz D, Blank K, Broderick GA et al. Biochemical outcome after radical prostatectomy, external beam radiation therapy, or interstitial radiation therapy for clinically localized prostate cancer. JAMA 1998; 280: 969–974. 12 Amling CL, Riffenburgh RH, Sun L, Moul JW, Lance RS, Kusuda L et al. Pathologic variables and recurrence rates as related to obesity and race in men with prostate cancer undergoing radical prostatectomy. J Clin Oncol 2004; 22: 439–445. 13 Freedland SJ, Aronson WJ, Kane CJ, Presti Jr JC, Amling CL, Elashoff D et al. Impact of obesity on biochemical control after radical prostatectomy for clinically localized prostate cancer: a report by the Shared Equal Access Regional Cancer Hospital database study group. J Clin Oncol 2004; 22: 446–453.
& 2014 Macmillan Publishers Limited
14 Freedland SJ, Isaacs WB, Mangold LA, Yiu SK, Grubb KA, Partin AW et al. Stronger association between obesity and biochemical progression after radical prostatectomy among men treated in the last 10 years. Clin Cancer Res 2005; 11: 2883– 2888. 15 Freedland SJ, Terris MK, Presti Jr. JC, Amling CL, Kane CJ, Trock B et al. Obesity and biochemical outcome following radical prostatectomy for organ confined disease with negative surgical margins. J Urol 2004; 172: 520–524. 16 Jayachandran J, Banez LL, Aronson WJ, Terris MK, Presti Jr JC, Amling CL et al. Obesity as a predictor of adverse outcome across black and white race: results from the Shared Equal Access Regional Cancer Hospital (SEARCH) Database. Cancer 2009; 115: 5263–5271. 17 Wilson KM, Giovannucci EL, Mucci LA. Lifestyle and dietary factors in the prevention of lethal prostate cancer. Asian J Androl 2012; 14: 365–374. 18 Nutrition NIoHa. Section of the National Health and Nutrition Survey. National Insutitute of Health and Nutrition. 19 Ma J, Li H, Giovannucci E, Mucci L, Qiu W, Nguyen PL et al. Prediagnostic body-mass index, plasma C-peptide concentration, and prostate cancer-specific mortality in men with prostate cancer: a long-term survival analysis. Lancet Oncol 2008; 9: 1039–1047. 20 Rowlands MA, Holly JM, Gunnell D, Gilbert R, Donovan J, Lane JA et al. The relation between adiposity throughout the life course and variation in IGFs and IGFBPs: evidence from the ProtecT (Prostate testing for cancer and Treatment) study. Cancer Causes Control 2010; 21: 1829–1842. 21 Keto CJ, Aronson WJ, Terris MK, Presti JC, Kane CJ, Amling CL et al. Obesity is associated with castration-resistant disease and metastasis in men treated with androgen deprivation therapy after radical prostatectomy: results from the SEARCH database. BJU Int 2012; 110: 492–498. 22 Capitanio U, Suardi N, Briganti A, Gallina A, Abdollah F, Lughezzani G et al. Influence of obesity on tumour volume in patients with prostate cancer. BJU Int 2012; 109: 678–684. 23 Freedland SJ, Banez LL, Sun LL, Fitzsimons NJ, Moul JW. Obese men have highergrade and larger tumors: an analysis of the duke prostate center database. Prostate Cancer Prostatic Diseases 2009; 12: 259–263. 24 Hayashi N, Matsushima M, Yamamoto T, Sasaki H, Takahashi H, Egawa S. The impact of hypertriglyceridemia on prostate cancer development in patients aged 4/ ¼ 60 years. BJU Int 2012; 109: 515–519. 25 Lee DJ, Ritch C, Desai M, Benson MC, McKiernan JM. The interaction of body mass index and race in predicting biochemical failure after radical prostatectomy. BJU Int 2011; 107: 1741–1747.
Prostate Cancer and Prostatic Disease (2014), 233 – 237
ORIGINAL ARTICLE
BMI is associated with larger index tumors and worse outcome after radical prostatectomy N Hayashi1, M Matsushima2, M Kido1, T Naruoka1, A Furuta1, N Furuta1, H Takahashi3 and S Egawa1 BACKGROUND: To investigate the impact of body mass index (BMI) on tumor characteristics and biochemical recurrence (BCR) after radical prostatectomy (RP) for prostate cancer (PCa) in Japanese men. METHODS: We evaluated data from consecutive patients who had undergone RP. Data analyzed included age, preoperative serum PSA, prostatic volume, BMI (continuous or categorized (p25 kg/m2) values), clinical and pathological findings including index tumor volume (ITV), and current status in areas such as smoker or nonsmoker and presence or absence of diabetes. We analyzed association between BMI and BCR, especially based on ITV using univariate and multivariate analysis. RESULTS: We analyzed data from a total of 703 patients. The median follow-up time was 38.4 months. BCR was diagnosed in 154 patients (21.9%) at a median of 9.7 months postoperatively. Multivariate linear regression analysis adjusted for preoperative variables showed a significant positive association between BMI and ITV (continuous BMI: P ¼ 0.002; categorical BMI: Po0.001, respectively), especially for higher-grade tumors (Gleason score X 7). Cox proportional hazards analysis showed a significant association between continuous BMI and BCR after surgery (preoperative variables, hazard ratio (HR) 1.09, 95% confidence interval (CI) 1.02–1.16, P ¼ 0.008), independent of clinical and pathological findings. In patients with high-risk cancer, the positive association between BMI and BCR was strengthened (preoperative variables, continuous BMI, HR 1.16, 95% CI 1.07–1.26, Po0.001; categorical BMI, HR 2.11, 95% CI 1.29–3.45, P ¼ 0.003, respectively). CONCLUSIONS: Greater BMI significantly correlates with higher rates of BCR after surgery; BMI is a preoperative variable associated with high-grade ITV. Our results suggest that the biological environment created by greater BMI may contribute to increasing tumor aggressiveness. Prostate Cancer and Prostatic Disease (2014) 17, 233–237; doi:10.1038/pcan.2014.15; published online 20 May 2014
INTRODUCTION Obesity has been associated with increased death rates from many malignant tumors including esophagus, colon, rectum, liver, gallbladder, pancreas, kidney and prostate in a prospectively studied cohort of US adults.1 It has also been related to decreased risk of low-grade disease and increased risk of high-grade and advanced prostate cancer (PCa).2–7 However, data on the link between obesity and PCa risk are mostly from western rather than Asian countries. One study from Korea failed to show such an association;8 the researchers attributed the negative findings partly to their leaner study population, with a lower body mass index (BMI) profile than in the west, and suggested that the impact of BMI may differ between Asian and western ethnicities. We previously reported that the prevalence and mean tumor volume of latent PCa have increased over the last 20 years (1983– 1992 vs 2009–2011), from 20.8% (104/500) to 43.0% (46/107, Po0.0001), and 230.4 to 1111.1 mm3, respectively.2 Although patients showing significant tumor volume (4500 mm3) accounted for only 9.6% of cases in 1983–1992, 26.1% were beyond this size in 2009–2011 (P ¼ 0.017). The cause and interpretation of these autopsy findings in Asia is still unclear. However, as one of the reasons, changes of Asian’s lifestyle cause the recent increase of those with higher BMI and could secondarily affect to patients with PCa.3 We thus hypothesized that the increase in BMI together with other confounding factors 1
such as resultant metabolic syndrome may have had a role in the recent change in the PCa landscape in Asia.4 To further investigate the biological effects of obesity on tumor characteristics and treatment outcome, we examined the relationship between BMI, pathological findings and biochemical recurrence (BCR) after radical prostatectomy (RP) for PCa in our Japanese cohort, especially based on the volume of tumor with higher Gleason score (GS) to be a powerful predictor of long-term outcome.5 MATERIALS AND METHODS Study population and specimens We evaluated data from 793 consecutive patients who had undergone RP (open or laparoscopic) at Jikei University Hospital between June 2002 and December 2009. The study design was reviewed and approved by the institutional review board. Of these, patients with neoadjuvant or adjuvant endocrine therapy and/or with a follow-up time of o6 months were excluded from data analysis. Patients treated with adjuvant radiotherapy (RT) were included in our study.6 A total of 703 patients were thus analyzed in this study. All patients underwent preoperative clinical evaluation, including digital rectal examination, chest radiography, computed tomography or magnetic resonance imaging of the abdomen and pelvis, and bone scanning. All RP samples were histologically examined by a single pathologist (HT),7 and GS was assigned. Clinical tumor stage and extent were determined according to the General Rule for Clinical and Pathological Studies on Prostate Cancer.9,10 Briefly, the prostatic capsule on
Department of Urology, Jikei University School of Medicine, Tokyo, Japan; 2Division of Clinical Research and Development, Jikei University School of Medicine, Tokyo, Japan and Department of Pathology, Jikei University School of Medicine, Tokyo, Japan. Correspondence: Dr N Hayashi, Department of Urology, Jikei University School of Medicine, 3-25-8 Nishi-Shimbashi, Minato-ku, Tokyo 105-8461, Japan. E-mail: [email protected] Received 2 May 2013; revised 20 February 2014; accepted 21 February 2014; published online 20 May 2014 3
Radical prostatectomy analyzed by BMI N Hayashi et al
234 each slide was carefully inspected microscopically for areas of complete cancer penetration through the prostatic surface (extraprostatic extension, EPE). Tumor cells at the inked margin of resection were considered to indicate a positive surgical margin (PSM). Specimens were also assessed for seminal vesicle invasion (SVI) or lymph nodal involvement. The volume of the index tumor (ITV, cm3, largest single focus of the tumor) was calculated using the formula 0.52 length width cross-sectional thickness. Each GS in ITV was also determined by a pathologist. In risk group stratification, patients with clinical stage T1c or T2a, PSA p10 ng ml–1 and biopsy GS p 6; patients with clinical stage T2b, PSA 10.1–20 ng ml–1 and biopsy GS 7; or patients with X T2c, a PSA 420 ng ml–1 or a biopsy GS X8 were defined as low risk, intermediate risk or high risk, respectively.11 BMI (kg/m2) was obtained by self-reported. Diabetes is diagnosed by physicians (history and/or medication). Smoking status was estimated by questionnaire.
Follow-up Patients were generally followed up with PSA testing every 3 months for 1 year, every 6 months for 2 years and then yearly, and physical examinations or other tests were added at intervals as clinically indicated. The median follow-up time was 38.4 months (range: 6–120 months). BCR was defined as any PSA increase to 0.2 ng ml–1.
Statistical analysis We conducted descriptive frequency analysis to evaluate the clinical and pathological characteristics of patients after surgery. Data included in the analysis were patient age, preoperative PSA, BMI, diabetes status, current smoker, preoperative variables (biopsy GS, clinical stage), postoperative variables (prostate weight, specimen GS, and pathological findings
Table 1.
including EPE, SVI, lymph nodal involvement, PSM and ITV), and adjuvant RT. We used the Student’s t-test or Wilcoxon rank sum test for continuous variables and the w2 test for categorical variables as appropriate for the descriptive statistics. To further identify the variables that could affect all subjects, and those that affected GS p6, GS 7 or GS X8 tumors subjects only, we controlled for confounders such as patient age, PSA, biopsy GS (p7 vs X8), clinical stage (cT1 vs cT2–3), and continuous or categorized (X 25) BMI in multivariate linear regression analysis (Tables 2 and 3). We analyzed correlations between clinicopathological covariates (preoperative variables: adjuvant RT, patient age, PSA, BM, current smoker, diabetes status, clinical stage (cT1 vs cT2–3) and biopsy GS (p7 vs X8); postoperative variables included: adjuvant RT, patient age, PSA, BMI, clinical stage (cT1 vs cT2–3), pathological GS (p7 vs X8), PSM, EPE and SVI) on BCR-free survival after surgery, by calculating adjusted hazard ratios (HRs) and the corresponding 95% confidence interval (CI) using a Cox proportional hazards model (Tables 4 and 5). The Spearman’s rank-order correlation coefficient between continuous and categorized BMI and postoperative variables was assessed. All tests were two-tailed, with Po0.05 considered significant. All statistical analyses were performed using both of SPSS, version 17.0 (Chicago, IL, USA) and STATA version 10.0 (College Station, TX, USA).
RESULTS Patient demographics Table 1 lists the demographic characteristics of our study population of 703 patients stratified by BMI (o25 vs X25 kg/m2). Patient age ranged from 47 to 77 years (mean 64, median 64). PSA levels ranged from 3.2 to 65.9 ng dl–1 (mean 10.0, median 8.2).
Patient characteristics BMI
Patients no. Age (years) median/mean PSA (ng ml–1) median/mean BMI (kg/m2) median/mean Current smokers, no. (%) Diabetes, no. (%) Clinical stage, no (%) T1 T2 T3 Biopsy GS, no. (%) 2–6 7 8–10 Risk groups, no. (%) Low Intermediate High EPE, no. (%) SVI, no. (%) LNI, no. (%) (total, n ¼ 585) PSM, no. (%) Pathological GS, no. (%) 2–6 7 8–10 ITV (cm3) median/mean GS p6 tumors (cm3) median/mean GS 7 tumors (cm3) median/mean GS X8 tumors (cm3) median/mean Prostatic weight (g) median/mean Adjuvant RT, no. (%)
Total
Less than 25
25 or Greater
703 64/64 8.2/10.0 23.5/23.6 116 (16.5) 67 (9.5)
516 65/64.4 8.0/9.7 22.6/22.3 73 (14.1) 44 (8.5)
187 63/63.2 8.5/10.6 26.4/27.0 43 (23.0) 23 (12.3)
508 (72.3) 188 (26.7) 7 (1.0)
365 (70.7) 146 (28.3) 5 (1.0)
143 (76.5) 42 (22.5) 2 (1.0)
259 (36.8) 293 (41.7) 151 (21.5)
196 (38.0) 210 (40.7) 110 (21.3)
63 (33.7) 83 (44.4) 41 (21.9)
187 313 203 240 65 20 242
142 231 143 176 46 14 169
45 82 60 64 19 6 73
P-value 0.024 o0.001 o0.001 0.013 0.143 0.122
0.471
0.268 (26.6) (44.5) (28.9) (34.1) (9.2) (3.4) (34.4)
117 (16.6) 450 (64.0) 136 (19.3) 1.8/3.7 0.5/1.8 2.1/3.8 2.7/5.3 38.0/37.8 51 (7.3)
(27.5) (44.8) (27.7) (34.1) (8.9) (2.7) (32.8)
84 (16.3) 335 (64.9) 97 (18.8) 1.6/3.2 0.5/1.8 1.7/3.2 2.5/4.8 37.5/41.7 29 (5.6)
(24.1) (43.9) (32.1) (34.2) (10.2) (3.2) (39.0)
33 (17.6) 115 (61.5) 39 (20.9) 2.6/4.9 0.8/1.8 3.1/5.3 3.1/6.3 38.5/41.8 22 (11.8)
0.977 0.63 0.683 0.129 0.842
o0.001 o0.001 o0.001 o0.001 0.995 0.018
Abbreviations: BMI, body mass index; EPE, extra-prostatic extension; GS, Gleason score, ITV, index tumor volume; LNI, lymph nodal involvement; PSM, positive surgical margin; RT, radiotherapy; SVI, seminal vesicle invasion. Po0.05: significant.
Prostate Cancer and Prostatic Disease (2014), 233 – 237
& 2014 Macmillan Publishers Limited
Radical prostatectomy analyzed by BMI N Hayashi et al
235 BMI ranged from 16.0 to 37.5 kg/m2 (mean 23.6, median 23.5). Eleven (1.6%) patients had BMI of 30 kg/m2 or greater. ITV ranged from 0 to 25.6 ml (mean 3.7, median 1.8). PCa was classified preoperatively as low risk in 187 patients, intermediate risk in 313 patients and high risk in 203 patients, including 151 patients with biopsy GS X8, 40 patients at clinical stage T2c or higher, and 49 patients with PSA X 20 ng dl–1. Lymph nodal involvement was detected in 20 (3.4%) of the 585 patients who had lymph node dissection. When stratified by BMI, statistical differences were detected for patient age, PSA, current smoker status, ITV, GS p6 tumors, GS 7 tumors, GS X8 tumors and adjuvant RT (Po0.05). Association of BMI with ITV Both continuous and categorical (o 25 vs X25 kg/m2) BMI showed a significant but weak correlation with continuous ITV (Spearman’s rank-order correlation coefficient r ¼ 0.135, Po0.001, and r ¼ 0.136, Po0.001, respectively), and not with continuous PSA, pathological GS p7 vs X8, PSM, EPE or SVI (P ¼ 0.141 and 0.117, 0.661 and 0.466, 0.053 and 0.139, 0.857 and 0.999, or 0.189 and 0.546, respectively). In contrast, continuous ITV was significant with all variables. In multivariate linear regression analyses, continuous or categorical BMI was also significantly associated with continuous ITV, accounting for other preoperative variables (P ¼ 0.002, Po0.001, respectively). In the adjusted model, including the interaction term for continuous BMI and higher grade (X GS 7) tumors for ITV, the interaction was still significant (P ¼ 0.03). We then performed subgroup analysis of ITVs, stratified by each GS, using multivariate linear regression analysis. Continuous and categorical BMI showed significant associations both with GS 7 and GS X8 tumors but no association with GS p6 tumors (Table 2). A significant trend for the association between higher grade (GS X7) tumors and continuous and categorical BMI was similarly seen in patients with pathologically localized (p pT2) cancer or negative surgical margin cancer (Table 3). Multivariate Cox proportional hazards analysis for BCR after RP BCR was encountered in 154 patients (21.9%) at a median of 9.7 months (range 0 to 98.6 months) postoperatively. Higher PSA was associated with increasing HR of BCR (preoperative variables: HR 1.05, 95% CI 1.03–1.06, P ¼ 0.001; postoperative variables included: HR 1.03, 95% CI 1.02–1.05, Po0.001, respectively, Table 4).
Similarly, the adjusted HR for BCR in patients with higher clinical stage (cT2–3), biopsy GS X8 as preoperative variables, and pathological GS X8, PSM, EPE, and SVI as postoperative variables was 2.07, 1.96 and 1.89; 1.71, 1.79 and 2.28 (Po0.001, o0.001, ando0.001, 0.02, 0.003, o0.001, respectively). Continuous BMI was positively associated with increasing HR of BCR (preoperative variables: HR 1.09, 95% CI 1.02–1.16, P ¼ 0.008; postoperative variables included: HR 1.07, 95% CI 1.01–1.14, P ¼ 0.01, respectively). In contrast, categorical BMI failed to show significant association with BCR (o25 vs X25 kg/m2, preoperative variables and postoperative variables included, P40.05). Impact of BMI on BCR according to patient subsets In the adjusted model, including the interaction term for continuous BMI and high-risk cancers, the interaction was still significant (P ¼ 0.011). To further define those patients for whom BMI was most closely associated with BCR, we performed Cox proportional hazards analysis for specific subsets of patients. Both
Table 3. Multivariate linear regression analysis in patients with pathological stage pT2 or negative surgical margin Higher-grade (GS X7) tumors (cm3) b-Coefficient
S.e.
P-value
Pathological T2 stage or less (n ¼ 463) Continuous BMI 0.107 Categorical BMI 0.712
0.043 0.26
0.013 0.006
Negative surgical margin (n ¼ 461) Continuous BMI 0.064 Categorical BMI 0.537
0.043 0.267
0.138 0.045
Abbreviations: BMI, body mass index; GS, Gleason score. Po0.05: significant.
Table 4.
Multivariate Cox proportional hazards analysis for BCR-free
survival HR (95% CI)
P-value
Preoperative variables Adjuvant RT Continuous age (years) Continuous PSA (ng ml–1) Continuous BMI (kg/m2) Biopsy GS p7 vs X8 Clinical stage T1 vs T2–3 Current smoker Diabetes status
0.92 (0.53–1.61) 1.01 (0.98–1.04) 1.05 (1.03–1.06) 1.09 (1.02–1.16) 1.96 (1.37–2.80) 2.07(1.48–2.90) 0.94 (0.62–1.44) 0.53 (0.26–1.10)
0.78 0.65 0.001 0.008 o0.001 o0.001 0.78 0.09
Postoperative variables included Adjuvant RT Continuous age (years) Continuous PSA (ng ml–1) Continuous BMI (kg/m2) Pathological GS p7 vs X8 PSM EPE SVI
0.54 1.00 1.03 1.07 1.89 1.71 1.79 2.28
0.06 0.67 o0.001 0.01 o0.001 0.02 0.003 o0.001
Variables
Table 2.
Multivariate linear regression analysis between BMI and each
ITV b-Coefficient
S.e.
P-value
0.14 1.13
0.05 0.28
0.002 o0.001
GS p6 tumors (cm3), (n ¼ 136) Continuous BMI 0.04 Categorical BMI 0.01
0.09 0.57
0.65 0.98
GS 7 tumors (cm3), (n ¼ 447) Continuous BMI Categorical BMI
0.17 1.36
0.06 0.34
0.002 o0.001
GS X8 tumors (cm3), (n ¼ 120) Continuous BMI Categorical BMI
0.29 1.96
0.13 0.74
0.027 0.009
3
ITV (cm ) Continuous BMI Categorical BMI
Abbreviations: BMI, body mass index; GS, Gleason score; ITV, index tumor volume. Po0.05: significant.
& 2014 Macmillan Publishers Limited
(0.29–1.03) (0.98–1.03) (1.02–1.05) (1.01–1.14) (1.32–2.70) (1.21–2.42) (1.22–2.63) (1.47–3.55)
Abbreviations: BCR, biochemical recurrence; BMI, body mass index; CI, confidence interval; EPE, extra-prostatic extension; GS, Gleason score; HR, hazard ratio; PSM, positive surgical margin; RT, radiotherapy; SVI, seminal vesicle invasion. Po0.05: significant.
Prostate Cancer and Prostatic Disease (2014), 233 – 237
Radical prostatectomy analyzed by BMI N Hayashi et al
236 Table 5.
Multivariate Cox proportional hazards analysis in patients with each risk group HR (95% CI)
P-value
1.08 (0.91–1.27) 0.59 (0.17–2.03)
0.39 0.40
Intermediate-risk group Continuous BMI Categorical BMI
1.02 (0.92–1.13) 0.84 (0.45–1.56)
0.75 0.58
High-risk group Continuous BMI Categorical BMI
1.16 (1.07–1.26) 2.11 (1.29–3.45)
o0.001 0.003
Variables Preoperative variables Low-risk group Continuous BMI Categorical BMI
Postoperative variables included Low-risk group Continuous BMI 1.10 (0.88–1.39) Categorical BMI 0.62 (0.13–3.03)
0.40 0.55
Intermediate-risk group Continuous BMI Categorical BMI
1.05 (0.93–1.18) 0.99 (0.47–2.08)
0.42 0.98
High-risk group Continuous BMI Categorical BMI
1.15 (1.06–1.25) 1.97 (1.20–3.22)
0.001 0.007
Abbreviations: BMI, body mass index; CI, confidence interval; HR, hazard ratio. Po0.05: significant.
continuous and categorical BMI showed significant association with BCR in patients with high-risk cancer only (preoperative variables: HR 1.16, 95% CI 1.07–1.26, Po0.001, and HR 2.11, 95% CI 1.29–3.45, P ¼ 0.003, and postoperative variables included: HR 1.13, 95% CI 1.04–1.22, P ¼ 0.001, and HR 1.84, 95% CI 1.13–2.98, P ¼ 0.01, respectively, Table 5). No association was found in patients with low- and intermediate-risk cancers (P40.05). DISCUSSION The association between greater BMI and cancer outcome is one of the major problems in health care. Several western studies have shown a significant association between greater BMI and increased total tumor volume, worse pathological findings, especially higher GS, and increased risk of BCR after RP for PCa.12–16 However, insufficient information is available on Asian men. Lee et al. previously found that higher BMI did not significantly enhance the ability to preoperatively predict tumor EPE and was not significantly associated with other adverse pathological findings including GS or BCR in Korean men who had undergone RP.8 They suggest that BMI up to a certain threshold (around 30 kg/m2) is not significant with respect to tumor aggressiveness. These findings remain controversial and have generated numerous hypotheses, predominantly owing to the lack of direct, sound explanations and to the presence of numerous ethnic and lifestyle confounders including aging, diabetes status and smoking habits.17 Over the last 30 years, the Japanese lifestyle has shifted toward a more ‘westernized’ diet of meat and poultry rather than fish and shellfish, and increased consumption of animal fat.18 Counterintuitively, daily energy intake per person declined by approximately 300 kcal day–1 during the same period. This may be due in part to modernization and an aging society.18 Fewer calories are consumed by elderly people and by fashion-conscious young women in pursuit of a slender body. Nevertheless, obesity has Prostate Cancer and Prostatic Disease (2014), 233 – 237
become more common; today around 30% of Japanese male adults are obese or overweight.3 Men with higher BMI tend to have decreased serum levels of testosterone and increased estrogen, and it seems feasible that PCa may be less androgen dependent in such men.13 Adiposity associated with higher BMI may also contribute to aggressive PCa growth by eliciting the secretion of certain hormones or growth factors.19,20 Positive associations between obesity and castration resistance or metastasis after hormone treatment have also been reported.21 In a recent microarray analysis, higher BMI was associated with higher gene expression of many inflammatory transcripts related to the nuclear factor kappa B pathway, possibly encouraging aggressive cancer biology or treatment resistance.8 These findings support the hypothesis that higher BMI provides a favorable biological environment for the survival and growth of aggressive or high-grade tumors. Both continuous and categorical BMI were significantly associated with continuous ITV. However, in multivariate analysis, greater BMI showed significant association only with higher-grade (GS X7) subsets but not with low-grade (GS p6) subsets. A significant trend toward this positive association was also seen in patients with p pT2 disease or negative surgical margin. An association between greater BMI and larger total tumor volume has also been reported in recent studies.22,23 In addition, the trend showed by our results seems to be the positive association with GS X8 strongly 4 GS 7. Therefore, patients classified by risk groups including GS X8 as a variable were analyzed. Patients with greater categorical BMI (X 25 kg/m2) had an increased HR for BCR after surgery in the high-risk group only. Previous findings showed that high triglyceride levels were significantly associated with a higher incidence of biopsy GS X8 in patients aged X60 years.24 Our current findings are in agreement with this, since hypertriglyceridemia is one of the manifestations of metabolic syndrome.24 Although BMI was unrelated to tumor grade, among men with higher grade ITV, those with higher BMI had larger tumors. Our results thus suggest that higher preoperative BMI can predict the tumor volume of higher grade cancer and could be associated with worse outcomes after surgery in high-risk patients. These findings disagree with those from the Korean study, for reasons that remain obscure at present.8 A single pathologist examined all histological slides and assigned GS in this study; the details of pathological assessment were not noted in the Korean study.25 We did not find smoking habits or diabetes to be independent predictors of tumor aggressiveness in this study. The limitations of our study are its retrospective nature and the limited number of patients. Definitive conclusions will require information over a much longer follow-up period, up to metastasis and cancer death. As in the Korean study, we enrolled only a small number of patients with BMI 30 kg/m2 or greater; because of the small number of those patients in our study, we did not evaluate the categorical data from those patients.8 We did not assess the relationship between serum hormones, growth factor levels and BMI. We cannot provide a sound explanation for our previous autopsy findings of the changes in latent tumor volume and prevalence over the last 20 years based on current observations.12 Longitudinal, prospective, well-controlled studies will be needed to elaborate and confirm these findings. In conclusion, this study showed a positive correlation between BMI and high-grade ITV and BCR after surgery. Greater BMI had a significant impact on BCR after surgery in the high-risk subgroup only. The clinical relevance of BMI may thus be enhanced in patients with high-risk cancer. High-risk patients with higher BMI may be candidates for more aggressive curative treatment. CONFLICT OF INTEREST The authors declare no conflict of interest.
& 2014 Macmillan Publishers Limited
Radical prostatectomy analyzed by BMI N Hayashi et al
237 REFERENCES 1 Calle EE, Rodriguez C, Walker-Thurmond K, Thun MJ. Overweight, obesity, and mortality from cancer in a prospectively studied cohort of U.S. adults. N Engl J Med 2003; 348: 1625–1638. 2 Okayasu M, Takahashi H, Furusato B, Egawa S. Comparison of latent prostate cancer detected at autopsy between pre-and post-prostatic specific antigen (PSA) era. Mod Pathol 2012; 25: 231A. 3 Kadota A, Miura K, Ueshima H. [The trend of metabolic syndrome in Japanese National Health and Nutrition Survey]. Nihon Rinsho 2011; 69(Suppl 1): 35–39. 4 Cullen J, Elsamanoudi S, Brassell SA, Chen Y, Colombo M, Srivastava A et al. The burden of prostate cancer in Asian nations. J Carcinog 2012; 11: 7. 5 Albertsen PC. What is the risk posed by prostate cancer? J Natl Cancer Inst Monogr 2012; 2012: 169–174. 6 Raldow A, Hamstra DA, Kim SN, Yu JB. Adjuvant radiotherapy after radical prostatectomy: evidence and analysis. Cancer Treat Rev 2011; 37: 89–96. 7 Hayashi N, Urashima M, Ikemoto I, Kuruma H, Arai Y, Kuwao S et al. Prostate-specific antigen adjusted for total prostatic tumor volume as a predictor for outcome after radical prostatectomy. Prostate Cancer Prostatic Dis 2007; 10: 60–65. 8 Freedman JE, Larson MG, Tanriverdi K, O’Donnell CJ, Morin K, Hakanson AS et al. Relation of platelet and leukocyte inflammatory transcripts to body mass index in the Framingham heart study. Circulation 2010; 122: 119–129. 9 Kanahara (ed.) General Rule for Clinical and Pathological Studies on Prostatic Cancer. 3 edn. Japanese Urological Association and Japanese Pathological Society: Tokyo, 2001. 10 Sbin L, W. C (ed). UICC International Union Against Cancer, 2002. Wiley-Liss: New York. 11 D’Amico AV, Whittington R, Malkowicz SB, Schultz D, Blank K, Broderick GA et al. Biochemical outcome after radical prostatectomy, external beam radiation therapy, or interstitial radiation therapy for clinically localized prostate cancer. JAMA 1998; 280: 969–974. 12 Amling CL, Riffenburgh RH, Sun L, Moul JW, Lance RS, Kusuda L et al. Pathologic variables and recurrence rates as related to obesity and race in men with prostate cancer undergoing radical prostatectomy. J Clin Oncol 2004; 22: 439–445. 13 Freedland SJ, Aronson WJ, Kane CJ, Presti Jr JC, Amling CL, Elashoff D et al. Impact of obesity on biochemical control after radical prostatectomy for clinically localized prostate cancer: a report by the Shared Equal Access Regional Cancer Hospital database study group. J Clin Oncol 2004; 22: 446–453.
& 2014 Macmillan Publishers Limited
14 Freedland SJ, Isaacs WB, Mangold LA, Yiu SK, Grubb KA, Partin AW et al. Stronger association between obesity and biochemical progression after radical prostatectomy among men treated in the last 10 years. Clin Cancer Res 2005; 11: 2883– 2888. 15 Freedland SJ, Terris MK, Presti Jr. JC, Amling CL, Kane CJ, Trock B et al. Obesity and biochemical outcome following radical prostatectomy for organ confined disease with negative surgical margins. J Urol 2004; 172: 520–524. 16 Jayachandran J, Banez LL, Aronson WJ, Terris MK, Presti Jr JC, Amling CL et al. Obesity as a predictor of adverse outcome across black and white race: results from the Shared Equal Access Regional Cancer Hospital (SEARCH) Database. Cancer 2009; 115: 5263–5271. 17 Wilson KM, Giovannucci EL, Mucci LA. Lifestyle and dietary factors in the prevention of lethal prostate cancer. Asian J Androl 2012; 14: 365–374. 18 Nutrition NIoHa. Section of the National Health and Nutrition Survey. National Insutitute of Health and Nutrition. 19 Ma J, Li H, Giovannucci E, Mucci L, Qiu W, Nguyen PL et al. Prediagnostic body-mass index, plasma C-peptide concentration, and prostate cancer-specific mortality in men with prostate cancer: a long-term survival analysis. Lancet Oncol 2008; 9: 1039–1047. 20 Rowlands MA, Holly JM, Gunnell D, Gilbert R, Donovan J, Lane JA et al. The relation between adiposity throughout the life course and variation in IGFs and IGFBPs: evidence from the ProtecT (Prostate testing for cancer and Treatment) study. Cancer Causes Control 2010; 21: 1829–1842. 21 Keto CJ, Aronson WJ, Terris MK, Presti JC, Kane CJ, Amling CL et al. Obesity is associated with castration-resistant disease and metastasis in men treated with androgen deprivation therapy after radical prostatectomy: results from the SEARCH database. BJU Int 2012; 110: 492–498. 22 Capitanio U, Suardi N, Briganti A, Gallina A, Abdollah F, Lughezzani G et al. Influence of obesity on tumour volume in patients with prostate cancer. BJU Int 2012; 109: 678–684. 23 Freedland SJ, Banez LL, Sun LL, Fitzsimons NJ, Moul JW. Obese men have highergrade and larger tumors: an analysis of the duke prostate center database. Prostate Cancer Prostatic Diseases 2009; 12: 259–263. 24 Hayashi N, Matsushima M, Yamamoto T, Sasaki H, Takahashi H, Egawa S. The impact of hypertriglyceridemia on prostate cancer development in patients aged 4/ ¼ 60 years. BJU Int 2012; 109: 515–519. 25 Lee DJ, Ritch C, Desai M, Benson MC, McKiernan JM. The interaction of body mass index and race in predicting biochemical failure after radical prostatectomy. BJU Int 2011; 107: 1741–1747.
Prostate Cancer and Prostatic Disease (2014), 233 – 237
