Facteurs prédictifs de récidive biologique après prostatectomie radicale dans une population afro-caribéenne

25 juin 2018

Auteurs : L. Brureau, E. Emeville, L. Multigner, P. Blanchet
Référence : Prog Urol, 2018, 8, 28, 442-449



Prostate cancer is the second most common cause of non-cutaneous cancer among men worldwide and the leading cause among men in developed countries [1]. It disproportionately affects individuals of African ancestry, irrespective of where they live, and is less common in Caucasian and Asian populations [2]. The reasons for these ethnicity-based differences in incidence are largely unknown, but probably involve a complex interplay between hormonal, environmental, and genetic factors [2, 3]. Moreover, populations of African ancestry also display more aggressive disease and a higher worldwide mortality rate from prostate cancer than other ethnic populations [4, 5].

Surgical resection of the entire gland, radical prostatectomy (RP), is a common therapeutic option, for curative purposes, for men with organ confined prostate cancer. Predicting who will have biochemical disease recurrence (BCR) after RP remains a challenge. Many studies, mostly among Caucasian and Asian populations, have identified various clinical and pathological features to be significant risk factors of BCR after RP. Few studies have investigated BCR after RP in populations of African ancestry. These studies have been performed in the US among the minority African-American group [6, 7] or in Jamaica among Afro-Caribbeans [8] and compared to Caucasian-Americans, assuming that risk factors of BCR were the same, regardless of the ethnicity.

The French West Indies (FWI) is a French overseas department in the Caribbean where most of the population is of African ancestry. The incidence of prostate cancer is high [9], similar to those found in African-American men in the US, and Afro-Caribbean or African men in the United Kingdom [10, 11]. However, previously acquired information on risk factors of BCR after RP, particularly in Afro-American men, may not be fully applicable to Afro-Caribbean populations, because of likely differences in social, environmental, and medical conditions, including access to health care. Here, we report the pre- and post-operative predictive factors of BCR after RP among 964 consecutive Afro-Caribbean patients in a single centre in Guadeloupe (French West Indies).

Patients and methods

Study population

This study took place in Guadeloupe (French West Indies), a Caribbean archipelago of 410,000 inhabitants where most of the inhabitants (∼90%) are of African ancestry. Between April 1, 2000 and December 31, 2010, 2444 new cases of prostate cancer were diagnosed in the University Hospital of Guadeloupe, corresponding to approximately 60% of new cases in the country. During the same period, among the 2444 new cases, 1051 (43%) underwent RP as initial treatment for localized prostate cancer in our Urology department. Patients treated with neoadjuvant hormonal therapy or radiotherapy (n =10) were excluded from the study, as well those whose PSA values did not return below 0.2ng/ml, six weeks after RP (n =77). After surgery, all patients were followed by serial PSA determinations and clinical visits every six months for the first three years and annually thereafter. BCR was defined as two consecutive (usually four weeks apart) PSA measurements above 0.2ng/ml.

Data collection

For each patient, we collected data on their age at positive biopsy and at surgery, the date of surgery, their height and weight at surgery (to calculate the body mass index, kg/m2), diabetes mellitus type 2 (DT2), treated hypertension, the preoperative PSA value, clinical stage, biopsy Gleason score, length of each prostate biopsy core, and length of positive biopsy within each core (to calculate the percentage of core lengths affected by prostate cancer, the sum of the positive core lengths and then dividing by the sum of core lengths, then multiplied by 100), date of surgery, type of surgery, perioperative blood loss (as recorded in the operative notes), pathological clinical stage, pathological Gleason score, surgical margins, prostate weight, and follow-up PSA. Tumour density, as a percentage, was calculated as the estimated tumour volume relative to prostate weight (assuming a density of 1), multiplied by 100 [12].

Only the medical data was used from the database of all radical prostatectomy performed in our department. The study has been approved by the Ethic Committee of the University Hospital of Pointe-à-Pitre.

Statistical analysis

Five-year BCR-free survival rates were calculated by Kaplan-Meier analyses. Kaplan-Meier survival curves were stratified into low, intermediate, and high-risk groups according to D'Amico classification, CAPRA, and CAPRA S scores, and paired log-rank tests were performed. These analyses were restricted to patients with a follow-up of at least five years.

The hazard ratio (HR) and 95% confidence intervals (CI) for the association between individual characteristics and BCR were estimated using the Cox proportional-hazards regression model. Time to event was defined as the duration between the date of surgery and the PSA value that defined the recurrence event. Patients who did not relapse were censored at the last normal post-operative PSA measurement before December 31, 2014. Age and prostate weight were analysed as continuous variables (after logarithmic transformation for prostate weight). Log-linearity was not achieved for other continuous variables, so they were categorized according to the median value: pre-operative PSA (<7.5 vs>7.5ng/ml), percentage of length of the positive biopsy core (<4.6 vs>4.6%), and tumour density (<10 vs>10%). The proportional hazards assumption was verified by the log-negative-log survival distribution function of all variables.

For multivariable analysis, we included all characteristics that were significantly associated (P value<0.05) to BCR by univariate analysis (Table 1) and then applied a backward elimination procedure with a P value>0.1. Missing data for selected variables introduced in the multivariable analysis (11 for pre-operative PSA, 14 for pathological stage, 17 for surgical margin, 19 for pathological Gleason score, 28 for biopsy Gleason score, 30 for clinical stage, 37 for DT2, 68 for tumor density, 126 for percentage of length of biopsy positive cores, and 246 for blood loss) were assumed to be missing at random and handled using a missing indicator categorical variable. In addition, a sensitivity analysis restricted to patients with known values of all variables (complete cases) was performed.

We applied a bootstrap technique [13], in the framework of the Cox proportional-hazards regression model, to generate 1000 bootstrap samples by replacement and calculate the percentage of samples for which each variable was associated with BCR with a P value<0.05. We also used logistic regression to explore associated factors for several covariates (perioperative blood loss) restricted to complete cases. All analyses were carried out using SAS software version 9.3 (SAS Institute, Inc., Cary, NC, USA). All tests were two-tailed, and P values<0.05 were considered to be statistically significant.


The study included 964 patients for whom the baseline characteristics, before and after surgery, of the study population are summarized in Table 2. The median age at surgery was 64.6 (range 41.4 to 79.4), respectively. The prevalence of obesity slightly exceeded 10%, whereas the prevalence of DT2 and hypertension were 18.8 and 53.3%, respectively.

The duration of the median follow-up for the overall population was 4.8 years (range 0.3 to 14.3) corresponding to 4,793 person-years, during which 257 men (26.7%) experienced a BCR. The overall five-year BCR-free survival rate was 70.9%: 84%, 60%, and 57% (P <0.001) in patients with low, intermediate, and high risk according to D'Amico classification, respectively (Appendix A; 81.1%, 60.4%, and 27.6% (P <0.001) in patients with low, intermediate, and high risk according to CAPRA scores, respectively (Appendix A); and 81.6%, 56.3%, and 43.7% (p <0.001) in patients with low, intermediate, and high risk according to CAPRA S scores, respectively (Appendix A).

In univariate analysis (Appendix A), DT2, a pre-operative PSA>7.5ng/ml (P =0.048), clinical stage T2 (P =0.0001), percentage of length of biopsy positive cores>4.6% (P <0.0001), a biopsy Gleason score>7 or (4+3) (P <0.0001), pathological stage pT3a (P =0.013) and pT3b (P <0.0001), a pathological Gleason score>7 or (4+3) (P <0.0001), positive surgical margins (P <0.0001), tumour density>10% (P =0.0002), and perioperative blood-loss>2000ml (P =0.025) were significantly associated with increased risk of BCR.

In multivariable analysis, pre-operative risk factors of BCR were DT2 (P =0.038), pre-operative PSA>7.5ng/ml (P =0.002), T2 clinical stage T2 (P =0.0006), biopsy Gleason score>7 or (4+3) (P <0.0001), and percentage of length of prostate biopsy positive core>4.6% (P =0.0006) (Table 1). Results of the sensitivity analysis (complete cases) were comparable with the primary analysis, except for DT2 (P =0.056) which was associated with BCR with borderline statistical significance (Table 1). Significant post-operative predictors of BCR (Table 3) were DT2 (P =0.045), pre-operative PSA>7.5ng/ml (P =0.018), pathological stage T3b (P =0.007), pathological Gleason score>7 or (4+3) (P <0.0001), positive surgical margins (P =0.0001), and intra-operative blood loss>2000ml (P =0.01) (Table 4). All variables were retained as predictive factors in sensitivity analysis restricted to complete cases, except DT2 (Table 4).

The percentage of inclusion among the 1,000 samples created by the bootstrapping technique (Table 3) confirmed that biopsy Gleason score, clinical stage, percentage of length of positive biopsy core, and pre-operative PSA are the best preoperative risk factors and that pathological Gleason score and surgical margins the best post-operative risk factors for predicting BCR.

We further explored whether blood loss may be explained by patient characteristics using logistic regression. In a multivariable analysis restricted to complete cases, greater age at surgery and obesity appeared to be significantly associated with perioperative blood loss>2000ml (P value=0.026 and 0.016, respectively).


This is the largest study to investigate pre- and post-operative predictive risk factors for BCR after RP for clinically localized prostate cancer in an Afro-Caribbean population. Despite the short follow-up, we confirmed, as shown in other populations, that palpable tumour at digital rectal examination, high Gleason score at biopsy, high pre-operative PSA values, and a high percentage of length of positive biopsy cores are significant pre-operative predictive factors of BCR, and that high pathological Gleason score and positive surgical margins, and to a lesser extent pathological stage (seminal vesicle invasion) and high pre-operative PSA values are significant post-operative predictive factors of BCR after surgery. Although obesity was associated with increased risk of BCR after RP in a racially-mixed US population [14], we did not observe such an association in our study, likely due to the low prevalence of obesity in our population.

We found that perioperative blood loss was significantly associated with increased risk of BCR after RP, confirming findings in a racially-mixed US population [15], despite limitations in estimating blood loss in our study. We agree with these authors that clinical and pathological features, such as pre-operative PSA, pathological stage, pathological Gleason score, and surgical margins, do not explain perioperative blood loss. We used validated classification as D'amico and CAPRA S score to predict the risk of BCR. We observed some difference between D'Amico and CAPRA S score in particular for patients with intermediate and high risk disease. CAPRA S score included more pathological data and could explain the difference between both scores.

In addition, we found that age at surgery and, as observed by Lloyd et al. [16], higher BMI were associated with greater blood loss.

Populations of African descent bear a disproportionate burden of DT2 [17]. We found an association of DT2 with BCR as a pre-operative risk factor in our population, but the association with BCR disappeared in the sensitivity analysis, restricted to complete cases, when DT2 was considered as a post-operative risk factor. Moreover, the very low frequency of association of DT2 in the post-operative bootstrap analysis does not support a major role for this variable in BCR. Our observations are in agreement with a recent meta-analysis based on seven studies on US (racially-mixed), Korean, and European populations that showed no evidence of significant association between DT2 and BCR after RP [18].

Chronic hypertension is also more prevalent in populations of African descent than in those of other ethnic origins [19]. We found no significant association between hypertension and BCR after RP, either as pre- or post-operative risk factor. Most, if not all, subjects in our study with hypertension were treated, although we cannot confirm compliance or the type of medication. Our results are in contrast to those from a US (racially mixed) study where medically recorded hypertension (diagnosed at any point before the prostate cancer diagnosis and without information on treatment) was significantly associated with BCR after RP [20]. A study in Japan has shown that patients with untreated and uncontrolled hypertension has a significant increased risk of BCR, whereas those who are treated and controlled are not at higher risk of BCR [21].


In summary, we have shown that our Afro-Caribbean population share the same major clinical and pathological risk factors of BCR after RP that have been identified in many Caucasian and Asian populations. Although this study was prospectively conducted, the short-term follow-up after surgery and the monocentric design are limits that should be acknowledged. Future studies on larger samples should be undertaken, particularly in the new era of robotic prostate cancer surgery, to confirm the influence of perioperative blood loss and determine the influence of metabolic syndrome on BCR after RP, as well as identify predictive factors of metastasis and long-term survival in the Afro-Caribbean population.

Ethical responsibilities of authors

The manuscript has not been submitted to more than one journal for simultaneous consideration.

The manuscript has not been published previously.

A single study is not split up into several parts to increase the quantity of submissions and submitted to various journals or to one journal over time.

No data have been fabricated or manipulated to support your conclusions.

No data, text, or theories by others are presented as if they were the author's own.

Consent to submit has been received explicitly from all co-authors, as well as from the responsible authorities - tacitly or explicitly - at the institute/organization where the work has been carried out, before the work is submitted.

Authors whose names appear on the submission have contributed sufficiently to the scientific work and therefore share collective responsibility and accountability for the results.

Authors are strongly advised to ensure the correct author group, corresponding author, and order of authors at submission.

Only the medical data was used from the database of all radical prostatectomy performed in our department.

The study has been approved by the Ethic Committee of the University Hospital of Pointe-à-Pitre.

Author's contribution

L. Brureau: manuscript writing, project development, data collection.

E. Emeville: data analysis.

L. Multigner: data analysis, manuscript writing.

P. Blanchet: data collection, project development.

Disclosure of interest

The authors declare that they have no competing interest.

Appendix A. Supplementary data

(113 Ko)
(58 Ko)
(50 Ko)
(14 Ko)

Table 1 - Multivariable Cox regression analysis of preoperative predictors of biochemical recurrence after radical prostatectomy.
Characteristics  All casesa (n =964) 
Complete casesb (n =812) 
  HR (95%CI)  P value  HR (95%CI)  P value 
Diabetes Mellitus type 2  
No  1.0    1.0   
Yes  1.37 (1.02-1.85)  0.038  1.36 (0.99-1.87)  0.056 
PSA (ng/ml)  
<7.5  1.0    1.0   
>7.5  1.49 (1.15-1.92)  0.002  1.51 (1.14-2.00)  0.004 
Clinical stage  
T1c  1.0    1.0   
T2  1.55 (1.21-1.98)  0.0006  1.62 (1.23-2.12  0.0006 
Biopsy Gleason score  
<7 or (3+4)  1.0    1.0   
>7 or (4+3)  2.12 (1.54-2.91)  <0.0001  2.13 (1.50-3.02)  <0.0001 
Length of positive biopsy core (%)  
<4.6  1.0    1.0   
>4.6  1.66 (1.24-2.20)  0.0006  1.61 (1.20-2.16)  0.001 

Missing data were handled using a missing indicator categorical variable.
Patients with known values of all variables.

Table 2 - General patient characteristics and clinico-pathological features.
Variables  All
n (%) 
n (%) 
n (%) 
General characteristics        
Age at surgery (years)  64.6 (59.8-68.6)a  64.4 (59.5-68.5)a  64.9 (60.6-69.0)a 
Body Mass Index (BMI)       
< 25  360 (45.1)  264 (45.7)  96 (43.4) 
25-<30  355 (44.4)  255 (44.1)  100 (45.2) 
≥30  84 (10.5)  59 (10.2)  25 (11.3) 
Diabetes Mellitus type 2       
No  753 (81.2)  562 (82.6)  191 (77.3) 
Yes  174 (18.8)  118 (17.4)  56 (22.7) 
No  432 (46.7)  317 (46.7)  115 (46.6) 
Yes  494 (53.3)  362 (53.3)  132 (53.4) 
Preoperative features        
PSA (ng/ml)  7.5 (5.5-10.5)a  7.2 (5.3-9.9)a  8.8 (6.2-12.0)a 
<7.5  471 (49.4)  372 (53.3)  99 (38.8) 
>7.5  482 (50.6)  326 (46.7)  156 (61.2) 
Clinical stage       
T1c  554 (59.3)  437 (63.7)  117 (47.2) 
T2  380 (40.7)  249 (36.3)  131 (52.8) 
Length of positive biopsy core (%)  4.6 (1.9-11.0)a  3.9 (1.6-8.9)a  7.4 (3.3-16.3)a 
<4.6  416 (49.6)  341 (54.9)  75 (34.6) 
>4.6  422 (50.4)  280 (45.1)  142 (65.4) 
Biopsy Gleason score       
<7 or (3+4)  821 (87.7)  622 (90.7)  199 (79.6) 
>7 or (4+3)  115 (12.3)  64 (9.3)  51 (20.4) 
Postoperative features        
Type of surgery       
Open  560 (61.1)  405 (60.2)  155 (63.8) 
Laparoscopic  356 (38.9)  268 (39.8)  88 (36.2) 
Years of surgery       
2000-2005  370 (38.4)  236 (33.4)  134 (52.1) 
2006-2010  594 (61.6)  471 (66.6)  123 (47.9) 
Pathological stage       
pT2  791 (83.3)  609 (87.4)  182 (71.9) 
pT3a  87 (9.2)  55 (7.9)  32 (12.6) 
pT3b  72 (7.6)  33 (4.7)  39 (15.4) 
Pathological Gleason score       
<7 or (3+4)  811 (85.8)  624 (90.0)  187 (74.2) 
>7 or (4+3)  134 (14.2)  69 (10.0)  65 (25.8) 
Surgical margins       
Negative  723 (76.3)  557 (80.0)  166 (66.1) 
Positive  224 (23.7)  139 (20.0)  85 (33.9) 
Prostate weight (g)  39.0 (30-52)a  39 (30-52)a  39 (30-50)a 
Tumour density (%)  10.0 (5-20)a  8.0 (5.0-16.0)a  11.0 (5.0-25.0)a 
<10  443 (49.4)  356 (53.7)  87 (37.3) 
>10  453 (50.6)  307 (46.3)  146 (62.7) 
Blood loss (L)  400 (250-500)a  400 (250-500)a  450 (300-600)a 
< 678 (94.4)  497 (95.4)  181 (91.9) 
1-< 35 (4.9)  23 (4.4)  12 (6.1) 
> 5 (0.7)  1 (0.2)  4 (2.0) 

Median (interquartile range).

Table 3 - Frequency of pre-operative and post-operative predictors of biochemical recurrence after radical prostatectomy in 1000 bootstrap samples at P value<0.05.
Pre-operative predictors 
Post-operative predictors 
Characteristics  Characteristics 
Biopsy Gleason score  99.5  Pathological Gleason score  99.1 
Clinical stage  85.0  Surgical margins  79.7 
Percentage of length of positive biopsy core  83.0  Perioperative blood loss  46.1 
Pre-operative PSA  82.6  Pathological stage  39.7 
Diabetes Mellitus Type 2  46.3  Tumour density  38.4 
Hypertension  9.5  Pre-operative PSA  25.6 
Body Mass Index  9.0  Type of surgery  19.4 
Age at surgery  8.9  Diabetes Mellitus Type 2  16.2 
    Body Mass Index  15.9 
    Age at surgery  13.1 
    Prostate weight  12.7 
    Hypertension  9.0 
    Year of surgery  7.4 

Table 4 - Multivariable Cox regression analysis of postoperative predictors of biochemical recurrence after radical prostatectomy.
Characteristics  All casesa (n =964) 
Complete casesb (n =678) 
  HR (95%CI)  P value  HR (95%CI)  P value 
Diabetes Mellitus type 2          
No  1.0    1.0   
Yes  1.37 (1.01-1.85)  0.045  Not retainedc   
PSA (ng/ml)          
<7.5  1.0    1.0   
>7.5  1.37 (1.06-1.79)  0.018  1.39 (1.02-1.88)  0.037 
Pathological Gleason score          
<7 or (3+4)  1.0    1.0   
>7 or (4+3)  2.36 (1.74-3.19)  <0.0001  2.39 (1.69-3.38)  <0.0001 
Pathological stage          
pT2  1.0    1.0   
pT3a  Not retainedc    Not retainedc   
pT3b  1.68 (1.15-2.45)  0.007  1.72 (1.11-2.66)  0.016 
Surgical margins          
Negative  1.0    1.0   
Positive  1.72 (1.32-2.45)  0.0001  1.73 (1.28-2.36)  0.0004 
Tumour density (%)          
<10  1.0    1.0   
>10  1.26 (0.97-1.63)  0. 090  Not retainedc   
Blood loss (L)          
< 1.0    1.0   
1-< Not retainedc    Not retainedc   
> 3.74 (1.37-10.2)  0.010  3.89 (1.42-10.58)d  0.008 

Missing data were handled using a missing indicator categorical variable.
Patients with known values of all variables.
Variable not retained after backward elimination.


Center M.M., Jemal A., Lortet-Tieulent J., Ward E., Ferlay J., Brawley O., et al. International variation in prostate cancer incidence and mortality rates Eur Urol 2012 ;  61 : 1079-1092 [cross-ref]
Hsing A.W., Chokkalingam A.P. Prostate cancer Epidemiology 2006 ;  11 : 1388-1413 [cross-ref]
Damber J.E., Aus G. Prostate cancer Lancet 2008 ;  371 : 1710-1721 [cross-ref]
Ferlay J., Shin H.R., Bray F., Mathers C. Estimates of worldwide burden of cancer in 2008  :  (2010). 2893-2917
Rebbeck T.R., Devesa S.S., Chang B.L., Bunker C.H., Cheng I., Cooney K., et al. Global patterns of prostate cancer incidence, aggressiveness, and mortality in men of african descent Prostate Cancer 2013 ; 56085710.1155/2013/560857
Laudano M.A., Badani K.K., McCann T.R., Mann M.J., Ritch C., Desai M., et al. Significant change in predicted risk of biochemical recurrence after radical prostatectomy more common in black than in white men Urology 2009 ;  74 : 660-664 [inter-ref]
Hamilton R.J., Aronson W.J., Presti J.C., Terris M.K., Kane C.J., Amling C.L., et al. Race, biochemical disease recurrence, and prostate-specific antigen doubling time after radical prostatectomy:results from the SEARCH database Cancer 2007 ;  110 : 2202-2209 [cross-ref]
Ritch C.R., Morrison B.F., Hruby G., Coard K.C., Mayhew R., Aiken W., et al. Pathological outcome and biochemical recurrence-free survival after radical prostatectomy in African-American, Afro-Caribbean (Jamaican) and Caucasian-American men:an international comparison BJU Int 2013 ;  111 : 186-190
Mallick S., Blanchet P., Multigner L. Prostate cancer incidence in guadeloupe, a French Caribbean archipelago Eur Urol 2005 ;  47 (6) : 769-772 [cross-ref]
Ben-Shlomo Y., Evans S., Ibrahim F., Patel B., Anson K., Chinegwundoh F., et al. The risk of prostate cancer amongst black men in the United Kingdom: the PROCESS cohort study Eur Urol 2008 ;  53 : 99-105 [cross-ref]
Laversanne M., Brewster D.H., Gombe Mbalawa C., Kohler B., et al. Cancer incidence in five continents:inclusion criteria, highlights from volume X and the global status of cancer registration Int J Cancer 2015 ;  137 : 2060-2071
Schmid H.P., McNeal J.E. An abbreviated standard procedure for accurate tumor volume estimation in prostate cancer Am J Surg Pathol 1992 ;  16 : 184-191 [cross-ref]
Sauerbrei W., Schumacher M. A bootstrap resampling procedure for model building: application to the Cox regression model Stat Med 1992 ;  11 : 2093-2109 [cross-ref]
Freedland S.J., Terris M.K., Presti J.C., Amling C.L., Kane C.J., Trock B., et al. Obesity and biochemical outcome following radical prostatectomy for organ confined disease with negative surgical margins J Urol 2004 ;  172 (2) : 520-524 [cross-ref]
Lloyd J.C., Bañez L.L., Aronson W.J., Terris M.K., Presti J.C., Amling C.L., et al. Estimated blood loss as a predictor of PSA recurrence after radical prostatectomy: results from the SEARCH database BJU Int 2010 ;  105 (3) : 347-351 [cross-ref]
Lloyd J.C., Bañez L.L., Aronson W.J., Terris M.K., Presti J.C., Amling C.L., et al. Preoperative predictors of blood loss at the time of radical prostatectomy: results from the SEARCH database Prostate Cancer Prostatic Dis 2009 ;  12 (3) : 264-268 [cross-ref]
Werfalli M., Engel M.E., Musekiwa A., Kengne A.P., Levitt N.S. The prevalence of type 2 diabetes among older people in Africa: a systematic review Lancet Diabetes Endocrinol 2016 ;  4 (1) : 72-84 [cross-ref]
Hua Q., Zhu Y., Liu H., Ye X. Diabetes and the risk of biochemical recurrence in patients with treated localized prostate cancer:a meta-analysis Int Urol Nephrol 2016 ;  48 (9) : 1437-1443 [cross-ref]
Ferdinand K.C., Saunders E. Hypertension-related morbidity and mortality in African Americans - why we need to do better J Clin Hypertens 2006 ;  8 : 21-30 [cross-ref]
Asmar R., Beebe-Dimmer J.L., Korgavkar K., Keele G.R., Cooney K.A. Hypertension, obesity and prostate cancer biochemical recurrence after radical prostatectomy Prostate Cancer Prostatic Dis 2013 ;  16 (1) : 62-66 [cross-ref]
Ohwaki K., Endo F., Hattori K. Abdominal obesity, hypertension, antihypertensive medication use and biochemical recurrence of prostate cancer after radical prostatectomy Eur J Cancer 2015 ;  51 (5) : 604-609 [cross-ref]

© 2018 
Elsevier Masson SAS. Tous droits réservés.