Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Effectiveness of a Global Multidisciplinary Supportive and Educational Intervention in Thermal Resort on Anthropometric and Biological Parameters, and the Disease-Free Survival after Breast Cancer Treatment Completion (PACThe)

Effectiveness of a Global Multidisciplinary Supportive and Educational Intervention in Thermal... Hindawi Journal of Oncology Volume 2020, Article ID 4181850, 13 pages https://doi.org/10.1155/2020/4181850 Clinical Study Effectiveness of a Global Multidisciplinary Supportive and Educational Intervention in Thermal Resort on Anthropometric and Biological Parameters, and the Disease-Free Survival after Breast Cancer Treatment Completion (PACThe) 1,2 3,4 2 1 Marie-Paule Vasson , Fabrice Kwiatkowski, Adrien Rossary, Sylvie Jouvency, 5 6 5 Marie-Ange Mouret-Reynier, Martine Duclos, Isabelle Van Praagh-Doreau, 7 3 Armelle Travade, and Yves-Jean Bignon Jean Perrin Comprehensive Cancer Centre, Department of Nutrition, 58 Rue Montalembert, 63011 Clermont-Ferrand, France University of Clermont Auvergne, INRA, UMR 1019 Human Nutrition Unit, CRNH-Auvergne, 28 Place Henri Dunant, 63000 Clermont-Ferrand, France Jean Perrin Comprehensive Cancer Centre, Department of Oncogenetics, 58 Rue Montalembert, 63011 Clermont-Ferrand, France University of Clermont-Auvergne, Laboratory of Mathematics, Probabilities and Applied Statistics, 28 Place Henri Dunant, 63000 Clermont-Ferrand, France Jean Perrin Comprehensive Cancer Centre, Department of Oncology, 58 Rue Montalembert, 63011 Clermont-Ferrand, France Gabriel Montpied University Hospital, Department of Sport Medicine and Functional Explorations, 58 Rue Montalembert, 63000 Clermont-Ferrand, France Centre Republique, Department of Senology, 99 Avenue de La Republique, 63100 Clermont Ferrand, France Correspondence should be addressed to Marie-Paule Vasson; m-paule.vasson@uca.fr Received 18 June 2019; Revised 6 November 2019; Accepted 7 February 2020; Published 5 May 2020 Guest Editor: Cigdem Selli Copyright © 2020 Marie-Paule Vasson et al. .is is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. A growing knowledge highlights the strong benefit of regular physical activity in the management of breast cancer patients, but few studies have considered biological parameters in their outcomes. In the prospective randomised trial after breast cancer treatment completion “PAC.e,” we determined the effects of physical activity and nutritional intervention on the biological and anthropometric status of patients after one year of follow-up, and clarified the link between biomarkers at allocation and disease- free survival. 113 patients from the population of the “PAC.e” study (n � 251) were analysed for biological parameters. Patients were randomized after chemotherapy in two arms: the intervention “SPA” receiving a 2-week session of physical training, dietary education, and physiotherapy (n � 57), and the control “CTR” (n � 56). Diet questionnaire, anthropometric measures, and blood parameters were determined at allocation and one year later. Survival and recurrence were checked over 7 years. Data were considered as a function of BMI, i.e., ≤25 for normal, 25–30 for overweight, and >30 for obese patients. At allocation, the large standard deviation for nutrient-intake values reflected an unbalanced diet for some patients in the three groups. At one-year − 6 − 7 follow-up, we noticed an increase in glucose (p< 10 ), insulin (p< 10 ), and adiponectin (p< 0.022) plasma levels for both intervention arms, which were more accentuated for the >30 groups. Using the Cox model, we demonstrated that the highest testosterone plasma values were linked to an increase of the recurrence risk (HR [CI–95%] � 5.06 [1.66–15.41]; p � 0.004). One- year after a global multidisciplinary supportive and educational intervention, we found few anthropometric and biological changes, mainly related to the patient’s initial BMI. We highlighted the importance of plasma testosterone in the evaluation of patient’s recurrence risk. Future studies would help better understand the mechanisms by which such multidisciplinary in- terventions could interact with breast cancer recurrence and define the most effective modalities. 2 Journal of Oncology determined the effects of PAC.e intervention on the bi- 1. Introduction ological and anthropometric status of patients after one-year Over many years, growing knowledge has indicated the follow-up and the link between the biomarkers and disease- strong benefit of regular physical activity in the management free survival with seven years of follow-up after completion of breast cancer patients [1]. Despite an extensive literature of breast cancer treatment. of clinical trials, data from these studies showed positive but modest effects, which may be underestimated due to great 2. Patients and Methods variability in the intervention strategies and intensity of monitoring [2, 3]. .ese interventions produce short-term 2.1. Participants. Two hundred and fifty-one nonmetastatic changes in physical activity and patient behaviour, but data breast cancer patients were enrolled between 2008 and 2010, are scarce on recurrence and long-term follow-up. Some as previously described [14]. .e main inclusion criteria studies have highlighted long-term barriers to exercise after were notably invasive nonmetastatic breast carcinoma; less diagnosis of breast cancer, including psychological barriers than 9 months after chemotherapy/radiotherapy comple- 2, (e.g., low motivation and dislike of gym), environmental tion, complete remission, 18.5< BMI< 40 kg/m and writ- barriers (e.g., employment priority and low access to fa- ten informed consent. Half of the 251 patients (n � 113) were cilities), and lack of time [4]. Regarding the large variability investigated for biological parameters in the present study. of practice procedures, further research is required to in- vestigate how to sustain positive effects of exercise over time and to determine essential attributes of exercise (mode, 2.2. Study Design. Patients were randomized into two groups: “SPA,” for the group attending the 2-week session in intensity, frequency, duration, and timing) by cancer type and cancer treatment for optimal effects [5]. .e intro- thermal centres, and “CTR,” for the control group. .e 2- week session performed in thermal centres included con- duction of wearable activity monitors into cancer care could improve the understanding of the association between sultations with physicians, nutritionists, and psycho-on- physical activity and patient behaviour, as previously sug- cologists; physical activity supervised by a physiotherapist gested [1]. for 2 h daily with endurance activities, strength training, and Moreover, analyses are needed to provide insight into flexibility/stretching; SPA care consisting of bath, shower, how physical activity interventions work. Such studies and massage for half an hour per day; aesthetic care; and should accelerate the identification of effective behaviour dietary meals with adapted menus, dietary education, and caloric intake limited to 1700–2000 kcal/day. changes and permit the development of evidence-based practice with better standardisation. Currently, the mech- Besides standard oncological follow-up of the patients in the two groups, personal consultations with a dietician were anisms by which physical activity mediates its benefits re- main unclear [6]. Most hypotheses regarding the biological organized to perform anthropometric measurements, pro- pathways have focused on the impact of obesity on breast vide dietary advice, and give encouragement for daily cancer risk and recurrence. In that field, the main research physical activity. Evaluation of survival/recurrence was axes are, first, the implication of sex hormones, including made by patients’ oncologist, with a follow-up period of 7 both oestrogens and androgens (testosterone) [7]; second, years [14]. .e overall protocol design is available in a the implication of metabolic hormones, such as insulin/ supplementary file. insulin-like growth factor (IGF) axis and adipocytokines (leptin and adiponectin) [8]; and third, the implication of 2.3. Data Collection. Before randomization and at one year, inflammatory factors (C reactive protein, CRP) [9]. None of the following analyses were performed on half of the pop- these axes has clearly demonstrated efficiency in clinical ulation (SPA: n � 57; CTR: n � 56): trials, despite evidence of increased quality of life (QoL), reduced body weight in obese patients, and reduced (1) Diet questionnaire recurrence. Dietitians evaluated oral intake based on a 72-h self- .e majority of studies that investigate the benefits of reported diet questionnaire. physical activity and nutritional interventions in breast (2) Body composition cancer focus on weight loss, cardiorespiratory capacity, QoL, and overall well-being [5, 10, 11], but few of them considered Body weight was measured at each personal con- the biological parameters of the patients in their outcomes sultation. Lean body mass (LBM), fat mass (FM), and [12, 13]. total body water were evaluated by multifrequency Taking into account these data and the interactions bioelectrical impedance analysis (Bodystat Quadscan between physical activity and BMI, we performed a pro- 4000) using 5, 50, 100, and 200 kHz. Tricipital skin- spective randomized trial “Programme of Accompanying fold thickness was measured using a skin-fold caliper women after breast Cancer treatment completion in .er- (Harpenden caliper). To assess central fat distribu- mal resorts” (PAC.e) for complete-responder breast tion, the waist circumference (WC) was evaluated to cancer patients after chemotherapy. In this trial, we dem- the nearest 0.5 cm using a standard tape measure onstrated that the 2-week intervention durably influences placed between the lowest rib and the iliac crest, with the QoL of breast cancer patients after both short-term [14] the patient in the standing position. .e hip cir- and long-term treatment [15]. In the present study, we cumference (HC) was estimated using a standard Journal of Oncology 3 tape measure placed horizontally at the widest point patients are referred to hereafter as the biological study on the hip. population. At one year post-inclusion, 13 patients withdrew for familial or professional reasons, and 53 and 47 patients (3) Blood sampling and biological assays remained, respectively, for the SPA and CTR groups. .e Blood samples were collected at allocation and at one main covariates were distributed similarly between the al- year. Plasma levels of biomarkers were determined as location groups (Table 1). Cancer treatments were similar follows: glucose and HDL-cholesterol (colorimetry and standard for invasive tumours. Most patients’ tumours methods), C-reactive protein, and transthyretin were HR positive and treated using hormonotherapy, and a (immunonephelometry) were determined at the few (Her2+ tumours) using targeted therapy. biomedical laboratory of the recruiting centre; in- sulin and testosterone (ELISA) were determined at the hospital biochemistry laboratory (Clermont- 3.1. Diet, Body, and Biological Parameters at Allocation. Ferrand); IGF-1, leptin, and adiponectin (luminex) Results of the biological study population were considered in were determined at the Genotool platform (Tou- function of BMI scale and divided into three subgroups, i.e., louse); and CA 15-3 was determined at the anti- 2 2 ≤25 kg/m for normal BMI, [25–30 kg/m ] for overweight, cancer centre radiobiology laboratory (Clermont- and >30 for obesity (Tables 2 & 3). Overall diet mean results Ferrand). (Table 2) were within adult nutritional recommendations (4) Recurrence follow-up (17.3%± 4.1, 46.7%± 10.4, and 35.5%± 8.6, respectively, for protein, carbohydrate, and lipid intakes). A large dispersion Disease-free interval was computed as months of values was observed, resulting in no significant difference elapsed from date of randomization to documented between BMI subgroups except for total energy intake (TEI) breast cancer recurrence during seven years after (p � 0.038) and lipid intake in gram/day (p � 0.034). .e breast cancer treatment completion. All recurrence large standard deviation for each nutrient-intake value re- types were considered, either local or distant (nodes, flected an unbalance diet for some patients in the three BMI metastatis, and/or contralateral breast cancer). subgroups. All body parameters (Table 2) differed significantly by 2.4. Statistical Considerations. Protocol design consisted of a BMI subgroup (p< 10 − 7). As expected, the lean mass/fat multicentre parallel randomized prospective trial. Data were mass ratio decreased with the BMI due to the expansion of analysed using the intention-to-treat principle. Descriptive the body fat mass, i.e., 2.4± 0.6, 1.7± 0.3, and 1.3± 0.3, statistics are presented with mean± standard deviation (SD) respectively, for normal, overweight, and obese subgroups for Gaussian quantitative variables. Outcomes are shown (p< 10 − 7). with 95% confidence intervals. Categorical variables are As previously noticed, we observed a large dispersion of described using counts by class and frequencies (%). all biological parameter values (Table 3) regardless of BMI Comparison of outcomes per allocation group and per subgroup. Increased plasma levels of CRP (p< 10 − 5), in- BMI class was tested with Student’s t-test, one-way analysis sulin (p< 10 − 4), and leptin (p< 10 − 7) showed dysme- of variance (ANOVA), or the Kruskal-Wallis H-test tabolic disorders associated with overweight/obesity. As depending on homoscedasticity or normality of distribu- expected, the ratio of leptin/adiponectin significantly in- tions. Two-way ANOVA was used to compare longitudinal creased with BMI (0.53± 0.51, 1.26± 1.28, and 3.23± 3.86, variations between allocation groups, but without an in- respectively, for normal, overweight, and obese groups, teraction test because of unequal class sizes. Categorical data p< 10 − 7). Conversely, a significant decrease in HDL-C were compared with chi test. To test the association between level with BMI (p< 10 − 4) was observed. Transthyretin, two quantitative parameters, Pearson’s correlation coeffi- similar between groups, was in the physiological range, cient was used, or Spearman’s rank correlation if distribu- showing no malnutrition disorders in the studied pop- tions were not Gaussian. Survival curves were drawn using ulation. Other parameters (glucose, IGF-1, testosterone, and Kaplan-Meier’s method, and comparison of curves was CA 15-3) were in the normal range, with no difference performed using the Log-rank test. A backward and stepwise between BMI groups except for CA 15-3 (p � 0.014). Cox proportional hazard regression model was used to perform the multivariate analysis of survival. Cutoff values of biological parameters to draw survival curves were chosen 3.2. Changes in Diet, Body, and Biological Parameters One among quartiles of distribution. Year Later. One year after inclusion, Diet consumption, All tests were two-sided and the nominal level of sig- body, and biological parameters of patients were reevaluated nificance was 5%. Randomisation and statistics were per- one year after inclusion. All the raw data are presented by formed using SEM software [16]. BMI subgroups in two supplementary data files: one for the SPA group (Supplementary Table 1) and one for the CTR group (Supplementary Table 2). Variations in each pa- 3. Results rameter between inclusion and one-year follow-up are Biological parameters were evaluated at allocation for half of shown in Tables 4 and 5 and analyzed according to the the 251 patients: n � 57 for the “SPA” experimental group intervention group (SPA effect), one-year follow-up (time and n � 56 for the “CTR” control group (Figure 1). .ese 113 effect), and BMI subgroups (BMI effect). 4 Journal of Oncology Enrollment Assessed for eligibility (n = 450) 199 patients refused to participate: (I) personal reasons (n = 58) (II) health difficulties (n = 45) (III) not interested (n = 36) (IV) familial reasons (n = 28) (V) transport problems (n = 15) (VI) work resumption (n = 12) (VII) want to forget the cancer (n = 5) Randomized (n = 251) Allocation Allocated to intervention (n = 126) Allocated to intervention (n = 125) Received SPA intervention (n = 117) Received CTR intervention (n = 115) Did not received SPA intervention (n = 9) Did not received CTR intervention (n = 10) 6 for personal reasons 6 because randomized to CTR group 3 for professional reasons 4 refused to continue Early exit < 1 year Early exit < 1 year Biology analysis 3 for personal reasons 3 for personal reasons on half of the population ∗ ∗ Biology and diet (n = 57) Biology and diet (n = 56) Allocation Allocation 1-year follow-up (n = 49) 1-year follow-up (n = 55) 2 samples missing 7 samples missing Follow-up Survival (n = 56) Survival (n = 55) Follow-up (years) Follow-up (years) median = 5.2 [0.5-6.9] median = 4.8 [0.3-6.8] 1 lost of view 1 lost of view Figure 1: Allocation diagram and flow chart. Diet, nutritional, and body data collection. No significant difference was observed for diet pa- considering both SPA and time effects. For the SPA and CTR rameters (Table 4) regardless of the intervention group, the >30 BMI subgroups, a reduction in brachial and abdominal time window, or the BMI subgroup, except for the total circumferences tended to correlate with an increase in hip energy intake with time (p � 0.039). For the SPA group, circumference. total energy intake remained stable for BMI subgroups ≤25 No significant SPA effect was observed for biological and [25–30 kg/m ], whereas a strong reduction (−400 kcal/ parameters (Table 5), except for transthyretin (p � 0.041) d) in the BMI >30 subgroup led to both carbohydrate and CA 15-3 (p � 0.04) plasma levels, although these (−21.5%) and lipid (−13.8%) intake decreases without remained in the normal ranges. For the time effect, a sig- change in patients’ weight. For the CTR group, total energy nificant increase in both glucose (p � 0.04) and insulin intake decreased for ≤25 and>30 BMI subgroups due to a (p � 0.035) and a decrease in HDL-C (p � 0.027) plasma reduction in protein, carbohydrate, and lipid intakes. levels were observed. As expected, several parameter vari- However, an increase in the mean body weight of 1 kg was ations were related to BMI in the two groups as previously observed for each BMI subgroup (supplementary data), shown at allocation. Notably, we noticed an increase in which was not significant because of the large dispersion of glucose (p< 10 − 6), insulin (p< 10 − 7), and adiponectin individual values. (p � 0.022) plasma levels regardless of the intervention For body parameters (Table 4), we observed that only the group and more accentuated plasma levels for the >30 BMI − 7 BMI effect was significant (p< 10 ). All the parameters subgroups. Conversely, a decrease in HDL-C plasma levels were significantly related to BMI but remained stable was observed (p � 0.007). Journal of Oncology 5 Table 1: Study population characterization. SPA group (n � 57) CTR group (n � 56) Parameter p value Size or mean± SD (%) or [mini-max] Size or mean± SD (%) or [mini-max] 52.0± 7.2 51.9± 10.6 Patients’ age at allocation 0.97 [36–66] [29–71] Menopausal status Yes � 33 (58%) Yes � 35 (63%) 0.62 25.4± 4.6 25.5± 4.4 BMI—body mass index (kg/m ) 0.92 [18.4–35.9] [18.0–38.7] ≤25 kg/m 30 (53%) 27 (48%) BMI—class 25–30 kg/m 16 (28%) 22 (39%) 0.37 >30 kg/m 11 (19%) 7 (13%) 55 9± 15.2 56.8± 14.0 SF36—global score/100 0.30 [19.0–93.0] [29.0–95.0] Surgery for breast cancer Yes � 57 (100%) Yes � 55 (98%) 0.50 Radiotherapy Yes � 54 (95%) Yes � 54 (96%) 0.98 Hormonotherapy Yes � 43 (75%) Yes � 43 (77%) 0.87 Herceptin Yes � 5 (9%) Yes � 7 (13%) 0.56 Chemotherapies: number of 6.3± 1.1 6.0± 0.8 0.29 cycles [5–15] [3–9] .e main covariates of the studied population at allocation are presented with mean ± standard deviation (SD) for Gaussian quantitative variables. Outcomes are shown with 95% confidence intervals. Categorical variables were described using counts by class and frequencies (%). Comparison of outcomes was tested with Student’s t-test or the Kruskal-Wallis H-test depending on homoscedasticity or normality of distributions. Categorical data were compared with the chi test. All tests were two-sided, and the nominal level of significance was 5%. Table 2: Diet and Body parameters at allocation. BMI (kg/m ) Mean± σ All groups (n � 113) p value of BMI effect ≤25 (n � 57) 25–30 (n � 38) >30 (n � 18) Diet parameters Total energy intake (TEI) (kcal/d) 1492± 450 1540± 358 1325± 378 1689± 678 0.038 Protein intake (g/d) 63.6± 20.2 65.3± 15.1 58.7± 20.0 68.8± 30.1 0.86 (% TEI) 17.3± 4.1 17.2± 3.5 17.9± 5.2 16.4± 3.1 0.71 Carbohydrate intake (g/d) 172.6± 61.5 175.3± 54.1 156.8± 53.7 197.2± 85.1 0.65 (% TEI) 46.7± 10.4 45.4± 9.5 48.1± 12.8 47.8± 6.5 0.75 Lipid intake (g/d) 59.7± 25.4 63.5± 22.3 50.6± 23.6 66.8± 31.8 0.034 % TEI 35.5± 8.6 36.8± 8.4 33.5± 9.8 35.8± 5.0 0.14 Body parameters −7 Body weight (kg) 65.2± 12.5 56.6± 6.4 68.5± 5.8 85.3± 10.7 <10 −7 Lean mass (LM) (kg) 42.1± 5.8 39.6± 4.5 43.0± 4.8 47.9± 6.3 <10 −7 (%) 65.2± 6.8 69.6± 5.3 62.9± 3.7 56.3± 4.5 <10 −7 Fat mass (FM) (kg) 23.0± 7.8 17.2± 3.7 25.5± 3.1 36.2± 5.1 <10 −7 (%) 34.6± 6.7 30.1± 5.0 37.3± 3.8 43.1± 4.4 <10 −7 Ratio LM/FM 2.0± 0.6 2.4± 0.6 1.7± 0.3 1.3± 0.3 <10 −7 Cell mass (kg) 25.0± 4.0 22.8± 2.5 25.5± 3.3 30.7± 3.5 <10 −7 Total water (l) 32.9± 3.9 31.1± 2.6 33.2± 2.7 38.1± 4.6 <10 −7 (%) 51.3± 5.4 55.1± 4.0 48.5± 3.2 44.9± 2.8 <10 −7 Extracellular water (%) 24.3± 3.4 25.7± 1.7 23.1± 2.1 22.8± 6.5 <10 −7 Intracellular water (%) 27.1± 2.4 28.2± 1.8 26.1± 2.8 25.6± 1.2 <10 −7 Tricipital fold thickness (cm) 17.4± 8.6 12.5± 5.2 18.8± 7.2 29.6± 6.4 <10 −7 Arm circumference (cm) 30.2± 3.8 27.7± 2.2 31.1± 1.7 36.4± 3.3 <10 −7 Waist circumference (WC) (cm) 84.0± 13.5 75.4± 7.7 86.8± 9.0 105.5± 8.9 <10 −7 Hip circumference (HC) (cm) 101.1± 9.1 95.0± 4.9 103.5± 5.5 115.7± 5.6 <10 Ratio WC/HC 0.83± 0.09 0.79± 0.07 0.84± 0.09 0.92± 0.08 0.000017 Diet parameters for food intake are expressed in raw value (gram/day) and in % of total energy intake. Body parameters are expressed in raw value (kilogram or liter) and in % of body mass. Comparison of outcomes per BMI group at allocation was tested with one-way analysis of variance (ANOVA). .e test was two-sided, and the nominal level of significance was 5%. We found significant positive correlations in the bio- (p � −0.46, p< 10 − 7). .e leptin/adiponectin ratio was logical study population between leptin/adiponectin ratio strongly correlated with waist circumference (r � 0.67, and insulin (r � 0.46, p< 10 − 7) and CRP (r � 0.46, p< 10 − 7), BMI (r � 0.51, p< 10 − 7), and cell mass (r � 0.46, p< 10 − 7) and a negative correlation with HDL-C p< 10 − 7). Moreover, despite the absence of variation in 6 Journal of Oncology Table 3: Biological parameters at allocation. BMI (kg/m ) Mean± σ All groups (n � 113) p value of BMI effect ≤25 (n � 57) 25–30 (n � 38) >30 (n � 18) Glucose (mmol/l) 5.2± 0.6 5.1± 0. 4 5.2± 0.6 5.6± 0.8 0.25 HDL-cholesterol (mmol/l) 2.13± 1.28 2.35± 1.35 1.98± 1.25 1.70± 0.97 0.0001 Transthyretin (g/l) 0.26± 0.04 0.26± 0.04 0.26± 0.04 0.26± 0.04 0.88 C-reactive protein (mg/l) 2.5± 3. 6 1.3± 1.2 3.2± 4.4 5.2± 4.9 0.000002 Insulin (mUI/l) 6.5± 6.2 4.7± 4.4 6.4± 4.4 12.1± 9.8 0.000013 IGF-1 (μg/l) 96.4± 49.3 95.8± 45.6 103.5± 45.7 84.7± 62.6 0.23 −7 Leptin (μg/l) 5.7± 4.7 3.5± 2.6 6.0± 3.0 12.1± 6.0 <10 Adiponectin (mg/l) 8.1± 5.1 8.9± 5.3 7.6± 4.8 6.6± 4.4 0.072 −7 Leptin/adiponectin ratio 1.22± 2.02 0.53± 0.51 1.26± 1.28 3.23± 3.86 <10 Testosterone (nmol/l) 0.82± 0.36 0.79± 0.29 0.83± 0.42 0.87± 0.38 0.67 CA 15-3 (kU/l) 18.1± 18.7 20.1± 24.5 14.1± 9.0 19.7± 8.4 0.014 Plasma biological parameters are expressed in usual unit per liter. Comparison of outcomes per BMI group at allocation was tested with one-way analysis of variance (ANOVA). .e test was two-sided, and the nominal level of significance was 5%. 2 2 testosterone plasma level with SPA, time, or BMI effects, this ≤25 kg/m for normal BMI, [25–30 kg/m ] for overweight, parameter was significantly associated (i) positively with and >30 for obesity. At allocation, the study population’s body weight (r � +0.15, p � 0.03), cell mass (r � +0.19, repartition into BMI subgroups was similar to that of the p � 0.0072), arm circumference (r � +0.15, p � 0.026), WC/ same-age female French population, as previously described HC ratio (r � +0.15, p � 0.027), and transthyretin (r � +0.15, [17]. .e diet intakes are in accordance with the adult p � 0.028) and (ii) negatively with TEI (r � −0.16, p � 0.022) nutritional recommendations for all groups. We noted no and HDL-C (r � −0.19, p � 0.007). difference between the three subgroups but a great variation in declared intakes, particularly in the obese group, raising doubts as to the reliability of the consumption-data col- 3.3. Biological Parameters and Recurrence Relation. We lection based on a 72-h self-report. tested the association between biomarker plasma levels at At allocation, after the completion of breast cancer allocation expressed in quartiles and the risk of recurrence treatment, the biological and body parameters of the pop- during the seven-year follow-up. Highest HDL-cholesterol ulation were in accordance with the usual observed values values were associated with the best survival without re- for normal, overweight, and obesity status. Considering the currence (p � 0.047). Conversely, the lowest testosterone mean value for each parameter defined as EGIR metabolic and CA 15-3 values were associated with longer disease-free syndrome criteria (glucose> 6.1 mmol/l, HDL-C< 1 mmol/l, survival (p � 0.001 and 0.03, respectively) (Table 6). insulin >18 mUI/l (QR4), and waist circumference> 80 cm), .e survival curves for these three biomarkers were done neither overweight nor obesity subgroups met the three in function of the calculated significant threshold values required criteria [18]. Among these parameters, only the (2.13 mmol/l, 0.9 nmol/l, and 20 kUI/l, respectively, for central criterion of obesity (waist circumference) was above HDL-C, testosterone, and CA 15-3) (Figures 2(a), 2(b), 2(e)). the limit value and emerged as the earliest criterion of For testosterone, two other survival curves were plotted metabolic syndrome under our conditions. However, con- taking into account the hormonotherapy status of patients sidering the large value dispersion of all these parameters, (Figures 2(c), 2(d)). .ese latter showed that testosterone some patients of both overweight and obese groups could was relevant for disease-free survival only in patients treated present a metabolic syndrome. with hormonotherapy (p � 0.012 vs. p � 0.69, respectively, Obesity is well-known to be associated with elevated for patients with and without hormonotherapy). Using the circulating levels of insulin, insulin-like growth factor 1 Cox model, the link between these variables and disease-free (IGF-1), leptin, and inflammation [19]. In our study, we survival was tested and demonstrated that only the highest observed a significant increase in CRP, insulin, leptin testosterone values predicted increased recurrence risk (HR plasma levels, and the ratio leptin/adiponectin in parallel [CI–95%] � 5.06 [1.66–15.41], p � 0.004) (Figure 2(f)). with significantly increased adiposity markers (fat mass, arm, waist, and hip circumferences). As expected, circu- 4. Discussion lating anti-inflammatory adiponectin was decreased, reinforcing the sub-chronic inflammation associated with In the present study, we determined the effects of PAC.e intervention (i.e., medical, nutritional, and psychological obesity and related to the risk of recurrence [20]. Sur- prisingly, no difference was observed for IGF-1 and tes- monitoring; physical activity training; SPA; and aesthetic care) on the biological and anthropometric status of patients tosterone plasma contents, contrary to previous at allocation and after one-year follow-up. observations [8, 13], probably due to the huge variability of As obesity has an impact on biological status and is a risk individual values. .eir plasma concentrations were factor for breast cancer, we chose to discuss the data maintained in the physiological range for the female according to three BMI subgroups defined as follows: population of corresponding age [21, 22]. Journal of Oncology 7 Table 4: Variation in diet and body parameters between one-year follow-up and allocation. SPA arm (n � 55) CTR arm (n � 49) p value effect of ≤25 (n � 29) ]25–30] (n � 15) >30 (n � 11) ≤25 (n � 23) 25–30 (n � 19) >30 (n � 7) SPA Time BMI Diet parameters −41.7± 400.5 +25.5± 556.4 −400.1± 527.6 −227.9± 362.6 +165.6± 410.4 −437.6± 955.1 Total energy intake kcal/d 0.91 0.039 0.15 (-0.02%) (+10.8%) (−18.7%) (-12.3%) (+20.1%) (−10.3%) −4.6± 17.9 +3.2± 25.5 −5.8± 19.0 −1.78± 27.1 +5.1± 21.4 −15.4± 45.4 Protein intake g/d 0.71 0.24 0.35 (-3.9%) (+17.5%) (-3.1%) (+6.0%) (+26.1%) (−1.4%) +8.4± 52.2 +6.1± 63.7 −57.9± 78.3 −28.4± 54.8 +6.7± 51.2 −46.7± 121.1 Carbohydrate intake g/d 0.84 0.10 0.38 (+10.1%) (+16.7%) (−21.5%) (−9.8%) (+7.5%) (−1.5%) −5.5± 26.4 −1.3± 36.4 −12.7± 25.2 −9.4± 19.8 +13.0± 13.0 −24.6± 45.4 Lipid intake g/d 0.89 0.099 0.15 (+0.2%) (+24.0%) (−13.8%) (−7.9%) (+67.3%) (−21.1%) Body parameters −0.10± 2.16 +1.47± 4.00 −0.73± 5.76 −0.24± 2.67 +0.26± 3.86 +0.93± 2.41 –7 Body weight kg 0.56 0.45 <10 (−0.2%) (+2.2%) (−0.3%) (−0.5%) (+0.4%) (+1.1%) +1.03± 3.63 −0.25± 1.66 −0.34± 5.85 0.00± 3.20 −0.02± 3.14 +3.09± 7.07 –7 Lean mass (LM) % 0.18 0.85 <10 (+1.7%) (−0.4%) (−0.0%) (−0.0%) (+0.1%) (+6.0%) −0.85± 3.50 +0.25± 1.66 +0.31± 5.90 −0.00± 3.20 +0.02± 3.14 −1.64± 6.05 –7 Fat mass (FM) % 0.11 0.86 <10 (−2.7%) (+0.6%) (+2.0%) (0.0%) (+0.6%) (-3.6%) +0.10± 0.39 −0.01± 0.14 −0.03± 0.37 +0.04± 0.40 −0.01± 0.29 +0.17± 0.39 –7 LM/FM ratio 0.20 0.77 <10 (+5.0%) (−0.6%) (+1.6%) (+1.7%) (+0.9%) (+14.4%) +1.34± 6.57 +0.95 2.21 +0.54± 3.02 −0.83± 4.52 −0.66± 4.74 +2.15± 5.82 –7 Cell mass kg 0.19 0.34 <10 (+6.8%) (+5.2%) (+2.0%) (−2.8%) (−1.7%) (+7.3%) +0.25± 1.62 +0.51± 1.45 +0.13± 2.53 −0.06± 1.02 +0.11± 1.86 +1.07± 5.15 –7 Total water l 0.22 0.40 <10 (+0.9%) (+1.7%) (+0.8%) (−0.1%) (+0.5%) (+3.2%) +0.51± 1.72 −0.67± 1.67 +0.02± 2.33 +1.14± 4.1 0.98± 3.72 −3.97± 10.72 –7 Extracellular water % 0.80 0.86 <10 (+2.0%) (−2.3%) (+0.6%) (+4.9%) (+4.3%) (−7.2%) +2.21± 9.32 +0.61± 1.32 −0.08± 1.99 −0.35± 4.76 −0.59± 4.68 +1.65± 4.55 Intracellular water % 0.28 0.53 0.00012 (+7.7%) (+3.2%) (−0.2%) (−1.3%) (−1.9%) (+7.0%) +0.46± 3.74 −0.57± 7.78 −3.25± 5.41 +0.23± 5.05 +2.98± 5.60 −2.51± 4.95 –7 Tricipital fold thickness cm 0.36 0.69 <10 (+10.7%) (+1.5%) (−7.5%) (+14.1%) (+25.3%) (−6.3%) −0.10± 1.77 −1.00± 1.49 −1.46± 1.94 −0.46± 1.71 −0.32± 2.16 −0.93± 2.01 –7 Arm circumference cm 0.58 0.32 <10 (−0.3%) (−3.1%) (−3.6%) (−1.7%) (−1.1%) (−2.4%) −2.93± 4.41 +0.43± 3.68 −1.09± 5.00 −0.46± 5.65 +1.45± 8.51 −3.36± 6.24 –7 Waist circumference (WC) cm 0.73 0.81 <10 (−3.7%) (−0.4%) (−1.0%) (−0.4%) (+2.7%) (−3.0%) −0.59± 2.80 1.77± 2.83 0.23± 5.85 −1.30± 3.56 −0.13± 3.71 2.79± 5.53 –7 Hip circumference (HC) cm 0.80 0.66 <10 (−0.6%) (+1.7%) (+0.4%) (−1.1%) (−0.1%) (2.5%) −0.03± 00.05 −0.01± 0.02 −0.01± 0.07 +0.01± 0.05 +0.2± 0.08 −0.05± 0.06 –7 WC/HC ratio 0.66 0.61 <10 (−2.9%) (−1.5%) (−1.1%) (+0.9%) (+3.1%) (−4.9%) Variation for each parameter is expressed in raw value (one-year follow-up value minus allocation value) and in percentage of the allocation value: + sign indicates an increase and –sign indicates a decrease. Two- way ANOVA was used to compare longitudinal variations between allocation arms (SPA effect), or one-year follow-up (time effect), or BMI groups (BMI effect), but without interaction test because of unequal class sizes. All tests were two-sided, and the nominal level of significance was 5%. Significant p values are indicated in bold. 8 Journal of Oncology Table 5: Variation in biological parameters between one-year follow-up and allocation. SPA arm (n � 55) CTR arm (n � 49) p value effect of ≤25 25–30 >30 ≤25 25–30 >30 SPA Time BMI (n � 29) (n � 15) (n � 11) (n � 23) (n � 19) (n � 7) 0.007± 0.446 0.459± 0.748 0.749± 1.358 −0.089± 0.326 0.226± 0.394 1.75± 2.775 –6 Glucose mmol/l 0.23 0.04 <10 (0.6%) (8.9%) (15.9%) (−1.6%) (4.4%) (25.9%) −0.438± 1.471 −0.241± 1.34 −0.542± 1.31 −0.436± 1.384 −0.333± 1.635 −0.028± 0.198 HDL-cholesterol mmol/l 0.41 0.027 0.007 (−5.6%) (3.4%) (−13%) (−4.8%) (7%) (−1.3%) 0.001± 0.038 0.009± 0.026 −0.005± 0.027 0.001± 0.033 −0.007± 0.031 0.002± 0.051 Transthyretin g/l 0.041 0.75 0.79 (1%) (3.5%) (−1.5%) (0.8%) (−2%) (3.2%) −0.146± 1.142 0.238± 2.818 −0.127± 1.481 0.264± 1.168 −0.135± 4.736 −1± 5.545 –7 C-reactive protein mg/l 0.11 0.73 <10 (11.1%) (27.3%) (8.9%) (41.8%) (41.1%) (−6.6%) 0.17± 3.79 1.97± 4.59 2.51± 10.19 0.22± 6.58 4.47± 10.51 4.53± 6.24 –7 Insulin mUI/l 0.41 0.035 <10 (25.7%) (115.7%) (58.4%) (36.1%) (50.9%) (78.4%) −0.79± 27.51 −26.61± 27.25 −18.78± 23.9 −13.18± 39.9 −12.41± 34.83 9.13± 16.55 IGF-1 μg/l 0.32 0.072 0.31 (6.3%) (−19.7%) (−24%) (−4.2%) (−12.2%) (16.5%) −0.03± 1.57 −0.02± 2.72 −2.42± 6.24 0.12± 2.38 1.64± 6.28 −0.93± 3.78 –7 Leptin μg/l 0.66 0.81 <10 (15.6%) (9%) (−2.4%) (21.7%) (23.6%) (−7.4%) 2.33± 4.19 1.29± 2.3 0.57± 1.3 0.99± 2.51 1.15± 2.65 0.65± 1.01 Adiponectin mg/l 0.33 0.082 0.022 (32.6%) (17.3%) (18%) (13.9%) (11.5%) (5.6%) −0.1± 0.27 −0.08± 0.52 −1.68± 2.92 0.13± 1.21 1.2± 4.07 −0.27± 0.29 –6 Leptin/adiponectin ratio 0.73 0.91 2 × 10 (−2.9%) (−0.4%) (−18%) (24.2%) (16.6%) (−13.6%) −0.033± 0.299 −0.013± 0.098 0.045± 0.347 −0.051± 0.244 0.015± 0.332 0.029± 0.757 Testosterone nmol/l 0.086 0.83 0.27 (−3.6%) (5%) (10.7%) (−5%) (3.4%) (21.8%) 2.32± 3.08 0.62± 1.78 0.27± 1.71 −5.32± 28.45 1.56± 2.06 2.71± 1.48 CA 15-3 kU/l 0.04 0.68 0.07 (18.2%) (6.8%) (4.7%) (1.4%) (12.5%) (14%) Variation for each parameter is expressed in raw value (one-year follow-up value minus allocation value) and in percentage of the allocation value: + sign indicates an increase and – sign indicates a decrease. Two- way ANOVA was used to compare longitudinal variations between allocation arms (SPA effect), or one-year follow-up (time effect), or BMI groups (BMI effect), but without interaction test because of unequal class sizes. All tests were two-sided, and the nominal level of significance was 5%. Significant p values are indicated in bold. Journal of Oncology 9 Table 6: Prognostic value of biological parameters on disease-free survival over 7 years. .reshold Parameters at allocation (n � 111) Median (quartiles) st rd ≤1 quartile ≤ Median ≤3 quartile (+) Cholesterol-HDL (mmol/l) 1.78 [1.46–2.13] p � 0.64 p � 0.22 p � 0.047 (−) (−) Testosterone (nmol/l) 0.7 [0.7–0.9] ND p � 0.049 p � 0.001 (−) (−) CA 15-3 (kU/l) 14 [10–20] p � 0.28 p � 0.07 p � 0.03 Association of biological parameters at allocation with the recurrence risk was tested using a two-sided chi test. .e nominal level of significance was 5%. + sign indicates that high values are in favour of a better prognosis, while – sign indicates that these high values worsen prognosis. HDL-C ≥ 2.13 nmol/l (N = 31) 100 100 Testo. < 0.9 nmol/l (N = 87) HDL-C < 2.13 nmol/l (N = 80) 75 75 Testo. ≥ 0.9 nmol/l (N = 24) 50 50 25 25 p = 0.001 p = 0.047 0 0 0 years 2 468 0 years 2 468 (a) (b) Testo. ≥ 0.9 nmol/l (N = 3) Testo. < 0.9 nmol/l (N = 64) Testo. < 0.9 nmol/l (N = 23) Testo. ≥ 0.9 nmol/l (N = 21) p = 0.0053 p = 0.76 0 years 2468 0 years 2468 (c) (d) CA 15-3 ≤ 20 kUl/l (N = 84) Covariables  p 01 6 Hazard ratios [CI‐95%]  (n = 111)  HDL cholesterol 0.17 0.29 [0.05 – 1.72] CA 15-3 > 20 kUl/l (N = 27) Testosterone 0.0044 5.06 [1.66 – 15.41] CA 15‐3 0.48 1.01 [0.99 – 1.03] p = 0.03 0 years 2 4 6 8 (e) (f) Figure 2: Survival curves and hazard ratios for HDL-cholesterol, testosterone, and CA 15-3. (a) HDL-cholesterol. (b) Testosterone—all patients. (c) Testosterone—patients without hormonotherapy. (d) Testosterone—patients with hormonotherapy. (e) CA 15-3. (f) Hazard ratios (Cox model). .reshold values for HDL-cholesterol, testosterone, and CA 15-3 at allocation correspond to the 75% percentile values. .ey were used to draw survival curves using Kaplan-Meier’s method. Comparison of curves was performed using the Log-rank test. Backward stepwise Cox proportional hazard regression model was used to perform the multivariate analysis of survival. All tests were two- sided, and the nominal level of significance was 5%. Disease-free survival Disease-free survival Disease-free survival Disease-free survival Disease-free survival 10 Journal of Oncology carbohydrate and lipid intakes, demonstrated the efficacy of Globally, as measurements were performed after completion of breast cancer treatment, body and biological patient’s nutritional information. As described at allocation, BMI was the major factor parameters seemed to be more linked to BMI status than to breast disease. Nevertheless, as previously described conditioning body and biological parameter changes one [23–25], we cannot exclude that the breast cancer therapy year later. For body parameters, we noted high central may be another cause of metabolic disturbances at allo- adiposity (waist and hip circumferences) in the overweight cation. .at may be the reason for the great variability and obese groups. .e same biomarker variations were observed for all parameters regardless of the BMI observed and reinforced for the overweight and obese subgroup. subgroups (i.e., increase in insulin, leptin, and CRP, and decrease in HDL-C). Moreover, these metabolic disorders One year after inclusion, the impact of the SPA inter- vention on diet, body, and biological parameters was induced an increased glycaemia and a decreased adipo- nectinemia in relation to more pronounced insulin resis- evaluated. Only transthyretin and CA 15-3 plasma levels were significantly affected by the SPA intervention. Trans- tance and sub-chronic inflammation [20]. .us, the obese groups presented two EGIR criteria for metabolic syndrome thyretin, one of the thyroid hormone carriers, is recognized as an acute malnutrition marker whose hepatic synthesis is (glucose and waist circumference) one year after breast reduced in case of inflammation [26]. In our study, trans- cancer treatment completion. .is confirms previous studies thyretin levels remained in the normal range and seemed to establishing that breast cancer posttreatment increases the be without biological meaning in regard of their tiny vari- risk of metabolic syndrome [39, 40]. ations and the absence of inflammation and of lean mass Finally, we clarified the link between biological markers changes. Breast cancer is generally not associated with at allocation and disease-free survival over seven years of follow-up after breast cancer treatment completion. We malnutrition or sarcopenia, especially so long after treat- ment [27]. CA 15-3 is frequently used for diagnosis and confirmed the interest of three biomarkers commonly used in the determination of recurrence risk: the highest plasma follow-up of breast cancer [28]. In our study, an a posteriori bias appeared for these biomarker data because the CTR values of HDL-C and the lowest plasma values of testos- terone and CA 15-3 were associated with a reduced risk of group patients presented higher CA 15-3 concentrations than the SPA group at allocation (Supplementary Table 2). recurrence [41–43]. HDL-C is linked to metabolic disorders One year after treatment completion, as none of the patients and is often related to androgen metabolism [44]. Choles- was in recurrence, CA 15-3 values decreased under the terol is clearly demonstrated to be a key regulator of breast threshold of 30 kU/l, confirming the efficacy of the therapy cancer tumours [45]. Favouring liver cholesterol clearance, [29, 30]. In accordance with previous studies showing an increase in HDL-C limits the availability of cholesterol for modest effects on body and biological parameters of physical recurrent cancer stem cells [46]. In our study, patients with the highest circulating HDL-C presented the lowest recur- activity and nutritional interventions [31, 32], our study shows the lack of one-year impact of a 2-week SPA rence risk. However, this protective effect was not retrieved in the multivariate Cox model, limiting the interest of cir- intervention. Some metabolic disorder changes were pointed out at culating HDL-C determination in recurrence monitoring. one-year follow-up (time effect). Despite a decrease in total As previously noted, CA 15-3 is a useful marker for energy intake, patients presented an increase in glucose and breast cancer follow-up: the circulating value is directly insulin plasma levels associated with a decrease in HDL-C. related to the stage and mass of the tumour [29]. In our .ese parameters suggest the development of insulin re- study, although the lowest circulating CA 15-3 values sistance independently of the BMI effect for overweight were associated with the lowest recurrence risk, the patients and the reinforcement of insulin resistance for obese multivariate Cox model did not confirm this observation. patients. .ese observations are in agreement with previous .is is in agreement with the literature, which has established the interest in CA 15-3 for monitoring breast studies which considered breast cancer as a metabolic dis- ease, with insulin resistance, sub-chronic inflammation, and tumour growth, but its poor prognostic value for re- currence risk [28, 30]. dysmetabolism induced by therapy [33, 34]. Moreover, an increased risk for metabolic syndrome and obesity has been In our study, only testosterone presented a significant described in long-term breast cancer survivors [35]. hazard ratio with disease-free survival; that is, the highest If women with breast cancer frequently lose weight circulating values (>0.9 nmol/l) were associated with re- during chemotherapy, a common unwanted long-term effect currence risk multiplied by ≈5 (HR � 5.06 [1.66–15.41]). of this therapy is weight gain, which often ranges 2–6 kg Notably, this link between testosterone and recurrence risk [10, 36] and penalizes mainly patients with adjuvant therapy only applied to patients receiving adjuvant hormonotherapy. [37]. In our study, weight gain was modest (less than 1 kg) .is observation confirms Venturelli’s observation of in- and concerned mainly the overweight BMI groups, of whom creased recurrence risk for testosterone plasma concentra- tion above 0.96 nmol/l with a hazard ratio of 4.68 for the majority were under hormonal adjuvant therapy. .us, weight control and diet intervention are important to im- overweight women but not for obese ones [47]. Testosterone is strongly associated with the androgen hypothesis of breast prove care and control of recurrence risk in posttreatment breast cancer patients [38]. In our study, the reduction in the carcinogenesis, related to the conversion of androgen into total energy intake provided by diet modification, especially oestrogen by aromatase [13]. .is enzymatic activity is Journal of Oncology 11 increased in obese patients due to the expansion of adiposity Additional Points [48]. However, it is not clear whether testosterone per se is directly responsible for promoting breast cancer risk or Highlights. (1) After breast cancer treatment completion, whether it is just a marker of the dysmetabolism linked to changes in anthropometric and biological parameters are overweight and obesity [49]. .is later hypothesis was mainly dependent on the patient’s BMI level. (2) A rein- confirmed in our study by the significant correlation of forcement of insulin resistance is observed in overweight plasma testosterone with several body and biological and obese patients after one-year treatment completion, markers associated with this dysmetabolism (positively with independently of physical activity and nutritional inter- body weight and ratio of WC/HC, and negatively with HDL- vention. (3) Testosterone plasma levels at the time of C). treatment completion are associated with recurrence risk in Our trial suffers from several limitations: patients receiving adjuvant hormonotherapy. (1) First, the small numbers of patients divided into different BMI subgroups limited the reliability of the Ethical Approval statistical analysis. (2) Second, the determination of biological parameters .e protocol was approved by the AFSSAPS (French Agency at one-year follow-up did not permit the charac- for Sanitary Security of Health Products), the regional Ethics terization of the short-term benefits of our 2-week Committee (2008), and the French National Committee SPA intervention. Moreover, the one-year time controlling personal computerized data (CNIL). .is trial window could explain the weak impact of this in- was performed in compliance with the Helsinki declaration tervention on the biological parameters. and registered in ClinicalTrials.gov with the no. NCT01563588. (3) .ird, the mismatches observed between diet con- sumption and weight changes of patients question the reliability of data collection using the 72-h self- Consent reported diet questionnaire. Written informed consent was obtained from all individual Few studies investigating the benefits of physical ac- participants included in the study. tivity and nutritional interventions in cancer survivors have considered the biological status of the patients in Conflicts of Interest their outcomes. Our data demonstrated that the health changes of patients were mainly related to their body .e authors declare that they have no conflicts of interest. condition and highlighted the importance of evaluating biological and anthropometric status in monitoring Authors’ Contributions cancer survivors. M.-P. V., F. K., M. D., and Y.-J. B. contributed in study 5. Conclusion conception and design. M.-A. M.-R., I. V. P.-D., A. T., and S. J. contributed in patient inclusion and follow-up. M.-P. V., To conclude, our study shows that one year after a global F. K., and A. R. contributed in acquisition, analysis, and multidisciplinary supportive and educational intervention, interpretation of the data. M.-P. V., F. K., A. R., and Y.-J. B. few anthropometric and biological changes could be at- drafted the manuscript. All authors gave the final approval. tributed to this intervention. It demonstrates that the one- year changes of patients are mainly related to their body Acknowledgments mass index (BMI) and confirms the importance of taking into account biological markers of metabolic status in the .e authors thank all patients who participated in the trial; follow-up of posttherapy breast disease. Among the tools the contributors of the programme at the Centre Jean Perrin, needed for this monitoring, our study highlights the in- Pole ˆ Sante´ Republique ´ (Clermont-Ferrand); .ermal resorts terest of plasma testosterone in the evaluation of recurrence (Vichy, Le-Mont-Dore, Chatel-Guyon); ˆ and others who risk. .ese observations may help reinforce care recom- supported the study. .e authors especially thank Nancy mendations for cancer survivors but need to be confirmed Uhrhammer for her careful language editing. .e trial was on a large population for a more comprehensive approach. financed by AFRETH (French Association for Hydrother- Future studies would permit a better understanding of the mal Research) and Auvergne Regional Council, Clermont- mechanisms by which such multidisciplinary interventions Communaute, ´ League against Cancer (Puy de Dome ˆ could interact with breast cancer recurrence and help committee CD63). define the most effective modalities. Supplementary Materials Data Availability Table 1: diet, body, and biological parameters for the SPA .e data used to support the findings of this study are arm at allocation and one-year follow-up. Table 2: diet, body, available from the corresponding author upon request. and biological parameters for the CTR arm at allocation and 12 Journal of Oncology [15] F. Kwiatkowski, M.-A. Mouret-Reynier, M. Duclos et al., one-year follow-up. Overall protocol design. (Supplementary “Long-term improvement of breast cancer survivors’ quality Materials) of life by a 2-week group physical and educational inter- vention: 5-year update of the ‘PAC.e’ trial,” British Journal References of Cancer, vol. 116, no. 11, pp. 1389–1393, 2017. [16] F. Kwiatkowski, M. Girard, K. Hacene, and J. Berlie, “Sem: a [1] G. Gresham, J. Schrack, L. M. Gresham et al., “Wearable suitable statistical software adaptated for research in oncol- activity monitors in oncology trials: current use of an ogy,” Bull Cancer, vol. 87, no. 10, pp. 715–721, 2000. emerging technology,” Contemporary Clinical Trials, vol. 64, [17] M. Arnold, M. Touillaud, L. Dossus et al., “Cancers in France pp. 13–21, 2018. in 2015 attributable to high body mass index,” Cancer Epi- [2] J. A. Ligibel and H. D. Strickler, “Obesity and its impact on demiology, vol. 52, pp. 15–19, 2018. breast cancer: tumor incidence, recurrence, survival, and [18] B. Balkau and M. A. Charles, “Comment on the provisional possible interventions,” American Society of Clinical Oncology report from the WHO consultation. European group for the Educational Book, vol. 33, pp. 52–59, 2013. study of insulin resistance (EGIR),” Diabet Medicine, vol. 16, [3] K. E. Post and J. Flanagan, “Web based survivorship inter- no. 5, pp. 442-443, 1999. ventions for women with breast cancer: an integrative review,” [19] D. F. Argolo, C. A. Hudis, and N. M. Iyengar, “.e impact of European Journal of Oncology Nursing, vol. 25, pp. 90–99, obesity on breast cancer,” Current Oncology Reports, vol. 20, no. 6, p. 47, 2018. [4] S. Y. Loh and A. Nadia, “Methods to improve rehabilitation of [20] S.-A. Lee, H. Sung, W. Han, D.-Y. Noh, S.-H. Ahn, and patients following breast cancer surgery: a review of sys- D. Kang, “Serum adiponectin but not leptin at diagnosis as a tematic reviews,” Breast Cancer: Targets and ?erapy, predictor of breast cancer survival,” Asian Pacific Journal of vol. 2015, no. 7, pp. 81–98, 2015. Cancer Prevention, vol. 15, no. 15, pp. 6137–6143, 2014. [5] S. I. Mishra, R. W. Scherer, P. M Geigle et al., “Exercise in- [21] H. Ketha and R. J. Singh, “Clinical assays for quantitation of terventions on health-related quality of life for cancer sur- insulin-like-growth-factor-1 (IGF1),” Methods, vol. 81, vivors,” ?e Cochrane Database of Systematic Reviews, vol. 15, pp. 93–98, 2015. no. 8, Article ID CD007566, 2012. [22] K. M. Wooding, J. A. Hankin, C. A. Johnson et al., “Mea- [6] M. C. De Boer, E. A. Worner, ¨ D. Verlaan, and surement of estradiol, estrone, and testosterone in post- P. A. M. Van Leeuwen, “.e mechanisms and effects of menopausal human serum by isotope dilution liquid physical activity on breast cancer,” Clinical Breast Cancer, chromatography tandem mass spectrometry without deriv- vol. 17, no. 4, pp. 272–278, 2017. atization,” Steroids, vol. 96, pp. 89–94, 2015. [7] G. Secreto, E. Venturelli, E. Meneghini et al., “Testosterone [23] T. Coskun, F. Kosova, Z. Ari, A. Sakarya, and Y. Kaya, “Effect and biological characteristics of breast cancers in postmen- of oncological treatment on serum adipocytokine levels in opausal women,” Cancer Epidemiology Biomarkers & Pre- patients with stage II-III breast cancer,” Molecular and vention, vol. 18, no. 11, pp. 2942–2948, 2009. Clinical Oncology, vol. 4, no. 5, pp. 893–897, 2016. [8] L. Mauro, G. D. Naimo, E. Ricchio, M. L. Panno, and S. Ando, ` [24] A. Alacacioglu, L. Kebapcilar, Z. Gokgoz et al., “Leptin, insulin “Cross-talk between adiponectin and IGF-IR in breast can- and body composition changes during adjuvant taxane based cer,” Frontiers in Oncology, vol. 5, p. 157, 2015. chemotherapy in patients with breast cancer, preliminary [9] N. M. Iyengar, X. K. Zhou, A. Gucalp et al., “Systemic cor- study,” Indian Journal of Cancer, vol. 53, no. 1, pp. 39–42, relates of white adipose tissue inflammation in early-stage breast cancer,” Clinical Cancer Research, vol. 22, no. 9, [25] J. Bao, N. Borja, M. Rao et al., “Impact of weight change pp. 2283–2289, 2016. during neoadjuvant chemotherapy on pathologic response in [10] M. N. Kirshbaum, “A review of the benefits of whole body triple-negative breast cancer,” Cancer Medicine, vol. 4, no. 4, exercise during and after treatment for breast cancer,” Journal pp. 500–506, 2015. of Clinical Nursing, vol. 16, no. 1, pp. 104–121, 2007. [26] A. Tsuboi, M. Terazawa-Watanabe, T. Kazumi, and K. Fukuo, [11] L. L. Peterson and J. A. Ligibel, “Physical activity and breast “Associations of decreased serum transthyretin with elevated cancer: an opportunity to improve outcomes,” Current On- high-sensitivity CRP, serum copper and decreased hemo- cology Reports, vol. 20, no. 7, p. 50, 2018. globin in ambulatory elderly women,” Asia Pacific Journal of [12] C. L. Rock, C. Pande, S. W. Flatt et al., “Favorable changes in Clinical Nutrition, vol. 24, no. 1, pp. 83–89, 2015. serum estrogens and other biologic factors after weight loss in [27] S. Steinbach, T. Hummel, C. Bohner ¨ et al., “Qualitative and breast cancer survivors who are overweight or obese,” Clinical quantitative assessment of taste and smell changes in patients Breast Cancer, vol. 13, no. 3, pp. 188–195, 2013. undergoing chemotherapy for breast cancer or gynecologic [13] D. Coradini, A. Orenti, E. Venturelli, A. Cavalleri, malignancies,” Journal of Clinical Oncology, vol. 27, no. 11, E. Biganzoli, and S. Oriana, “Serum levels of testosterone and pp. 1899–1905, 2009. SHBG in association with body mass index improve the [28] J. G. Santillan-Ben ´ ´ıtez, H. Mendieta-Zeron, ´ L. M. Gomez- ´ predictive capability of consolidate tumor biomarkers in pre- Olivan ´ et al., “.e tetrad BMI, leptin, leptin/adiponectin (L/A) and postmenopausal breast cancer patients,” Japanese Journal ratio and CA 15-3 are reliable biomarkers of breast cancer,” of Clinical Oncology, vol. 48, no. 4, pp. 308–316, 2018. Journal of Clinical Laboratory Analysis, vol. 27, no. 1, [14] F. Kwiatkowski, M. A. Mouret-Reynier, M. Duclos et al., “Long term improved quality of life by a 2-week group pp. 12–20, 2013. [29] Y. Fujimoto, T. Higuchi, A. Nishimukai et al., “High levels of physical and educational intervention shortly after breast cancer chemotherapy completion. Results of the ‘programme serum CA15-3 and residual invasive tumor size are associated with poor prognosis for breast cancer patients with non- of accompanying women after breast cancer treatment completion in thermal resorts’ (PAC.e) randomised clinical pathological complete response after neoadjuvant chemo- therapy,” Journal of Surgical Oncology, vol. 118, no. 1, trial of 251 patients,” European Journal of Cancer, vol. 49, no. 7, pp. 1530–1538, 2013. pp. 228–237, 2018. Journal of Oncology 13 [30] A. Nicolini, A. Carpi, P. Ferrari et al., “An individual reference [45] C. Bobin-Dubigeon, A. Chauvin, V. Brillaud-Meflah, F. Boiffard, M.-P. Joalland, and J.-M. Bard, “Liver X receptor limit of the serum CEA-TPA-CA 15-3 tumor marker panel in the surveillance of asymptomatic women following surgery (LXR)-regulated genes of cholesterol trafficking and breast cancer severity,” Anticancer Research, vol. 37, no. 10, for primary breast cancer,” Cancer Management and Research, vol. 10, pp. 6879–6886, 2018. pp. 5495–5498, 2017. [46] K. Pelton, C. M. Coticchia, A. S. Curatolo et al., “Hyper- [31] R. T. Chlebowski, “Nutrition and physical activity influence cholesterolemia induces angiogenesis and accelerates growth on breast cancer incidence and outcome,” ?e Breast, vol. 22, of breast tumors in vivo,” ?e American Journal of Pathology, no. Suppl 2, pp. S30–S37, 2013. vol. 184, no. 7, pp. 2099–2110, 2014. [32] A.-M. Foucaut, S. E. Berthouze-Aranda, M. Touillaud et al., [47] E. Venturelli, A. Orenti, A. S. C. Fabricio et al., “Observational “Reduction of health risk factors through an adapted physical study on the prognostic value of testosterone and adiposity in activity program in patients with breast cancer,” Supportive postmenopausal estrogen receptor positive breast cancer Care in Cancer, vol. 22, no. 4, pp. 1097–1104, 2014. patients,” BMC Cancer, vol. 18, no. 1, p. 651, 2018. [33] R. Bhandari, G. A. Kelley, T. A. Hartley, and I. R. H. Rockett, [48] C. Gerard ´ and K. A. Brown, “Obesity and breast cancer—role “Metabolic syndrome is associated with increased breast of estrogens and the molecular underpinnings of aromatase cancer risk: a systematic review with meta-analysis,” Inter- regulation in breast adipose tissue,” Molecular and Cellular national Journal of Breast Cancer, vol. 2014, Article ID Endocrinology, vol. 466, pp. 15–30, 2018. 189384, 13 pages, 2014. [49] F. Bertolini, “Adipose tissue and breast cancer progression: a [34] D. T. Dibaba, D. Braithwaite, and T. Akinyemiju, “Metabolic link between metabolism and cancer,” ?e Breast, vol. 22, syndrome and the risk of breast cancer and subtypes by race, pp. S48–S49, 2013. menopause and BMI,” Cancers, vol. 10, no. 9, p. 299, 2018. [35] A. Engin, “Obesity-associated breast cancer: analysis of risk factors,” in Obesity and Lipotoxicity, vol. 960, pp. 571–606, Springer, Berlin, Germany, 2017. [36] M. Sadim, Y. Xu, K. Selig et al., “A prospective evaluation of clinical and genetic predictors of weight changes in breast cancer survivors,” Cancer, vol. 123, no. 13, pp. 2413–2421, [37] A. Ginzac, E. .ivat, M.-A. Mouret-Reynier et al., “Weight evolution during endocrine therapy for breast cancer in postmenopausal patients: effect of initial fat mass percentage and previous adjuvant treatments,” Clinical Breast Cancer, vol. 18, no. 5, pp. e1093–e1102, 2018. [38] M. De Roon, A. M. May, A. McTiernan et al., “Effect of exercise and/or reduced calorie dietary interventions on breast cancer-related endogenous sex hormones in healthy postmenopausal women,” Breast Cancer Research, vol. 20, no. 1, p. 81, 2018. [39] W. K. Cho, D. H. Choi, W. Park et al., “Effect of body mass index on survival in breast cancer patients according to subtype, metabolic syndrome, and treatment,” Clinical Breast Cancer, vol. 18, no. 5, pp. e1141–e1147, 2018. [40] V. Boonyaratanakornkit and P. Pateetin, “.e role of ovarian sex steroids in metabolic homeostasis, obesity, and post- menopausal breast cancer: molecular mechanisms and ther- apeutic implications,” BioMed Research International, vol. 2015, Article ID 140196, 13 pages, 2015. [41] E. R. Nelson, “.e significance of cholesterol and its me- tabolite, 27-hydroxycholesterol in breast cancer,” Molecular and Cellular Endocrinology, vol. 466, pp. 73–80, 2018. [42] A. Micheli, E. Meneghini, G. Secreto et al., “Plasma testos- terone and prognosis of postmenopausal breast cancer pa- tients,” Journal of Clinical Oncology, vol. 25, no. 19, pp. 2685–2690, 2007. [43] P. Kachhawa, K. Kachhawa, D. Agrawal, V. Sinha, P. D. Sarkar, and S. Kumar, “Association of dyslipidemia, increased insulin resistance, and serum CA 15-3 with in- creased risk of breast cancer in urban areas of north and central India,” Journal of Mid-Life Health, vol. 9, no. 2, pp. 85–91, 2018. [44] M. Voisin, P. De Medina, A. Mallinger et al., “Identification of a tumor-promoter cholesterol metabolite in human breast cancers acting through the glucocorticoid receptor,” Pro- ceedings of the National Academy of Sciences, vol. 114, no. 44, pp. E9346–E9355, 2017. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Oncology Hindawi Publishing Corporation

Effectiveness of a Global Multidisciplinary Supportive and Educational Intervention in Thermal Resort on Anthropometric and Biological Parameters, and the Disease-Free Survival after Breast Cancer Treatment Completion (PACThe)

Download PDF
13 pages

Loading next page...
 
/lp/hindawi-publishing-corporation/effectiveness-of-a-global-multidisciplinary-supportive-and-educational-oDXYsKNMta

References (67)

Publisher
Hindawi Publishing Corporation
Copyright
Copyright © 2020 Marie-Paule Vasson et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
ISSN
1687-8450
eISSN
1687-8469
DOI
10.1155/2020/4181850
Publisher site
See Article on Publisher Site

Abstract

Hindawi Journal of Oncology Volume 2020, Article ID 4181850, 13 pages https://doi.org/10.1155/2020/4181850 Clinical Study Effectiveness of a Global Multidisciplinary Supportive and Educational Intervention in Thermal Resort on Anthropometric and Biological Parameters, and the Disease-Free Survival after Breast Cancer Treatment Completion (PACThe) 1,2 3,4 2 1 Marie-Paule Vasson , Fabrice Kwiatkowski, Adrien Rossary, Sylvie Jouvency, 5 6 5 Marie-Ange Mouret-Reynier, Martine Duclos, Isabelle Van Praagh-Doreau, 7 3 Armelle Travade, and Yves-Jean Bignon Jean Perrin Comprehensive Cancer Centre, Department of Nutrition, 58 Rue Montalembert, 63011 Clermont-Ferrand, France University of Clermont Auvergne, INRA, UMR 1019 Human Nutrition Unit, CRNH-Auvergne, 28 Place Henri Dunant, 63000 Clermont-Ferrand, France Jean Perrin Comprehensive Cancer Centre, Department of Oncogenetics, 58 Rue Montalembert, 63011 Clermont-Ferrand, France University of Clermont-Auvergne, Laboratory of Mathematics, Probabilities and Applied Statistics, 28 Place Henri Dunant, 63000 Clermont-Ferrand, France Jean Perrin Comprehensive Cancer Centre, Department of Oncology, 58 Rue Montalembert, 63011 Clermont-Ferrand, France Gabriel Montpied University Hospital, Department of Sport Medicine and Functional Explorations, 58 Rue Montalembert, 63000 Clermont-Ferrand, France Centre Republique, Department of Senology, 99 Avenue de La Republique, 63100 Clermont Ferrand, France Correspondence should be addressed to Marie-Paule Vasson; m-paule.vasson@uca.fr Received 18 June 2019; Revised 6 November 2019; Accepted 7 February 2020; Published 5 May 2020 Guest Editor: Cigdem Selli Copyright © 2020 Marie-Paule Vasson et al. .is is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. A growing knowledge highlights the strong benefit of regular physical activity in the management of breast cancer patients, but few studies have considered biological parameters in their outcomes. In the prospective randomised trial after breast cancer treatment completion “PAC.e,” we determined the effects of physical activity and nutritional intervention on the biological and anthropometric status of patients after one year of follow-up, and clarified the link between biomarkers at allocation and disease- free survival. 113 patients from the population of the “PAC.e” study (n � 251) were analysed for biological parameters. Patients were randomized after chemotherapy in two arms: the intervention “SPA” receiving a 2-week session of physical training, dietary education, and physiotherapy (n � 57), and the control “CTR” (n � 56). Diet questionnaire, anthropometric measures, and blood parameters were determined at allocation and one year later. Survival and recurrence were checked over 7 years. Data were considered as a function of BMI, i.e., ≤25 for normal, 25–30 for overweight, and >30 for obese patients. At allocation, the large standard deviation for nutrient-intake values reflected an unbalanced diet for some patients in the three groups. At one-year − 6 − 7 follow-up, we noticed an increase in glucose (p< 10 ), insulin (p< 10 ), and adiponectin (p< 0.022) plasma levels for both intervention arms, which were more accentuated for the >30 groups. Using the Cox model, we demonstrated that the highest testosterone plasma values were linked to an increase of the recurrence risk (HR [CI–95%] � 5.06 [1.66–15.41]; p � 0.004). One- year after a global multidisciplinary supportive and educational intervention, we found few anthropometric and biological changes, mainly related to the patient’s initial BMI. We highlighted the importance of plasma testosterone in the evaluation of patient’s recurrence risk. Future studies would help better understand the mechanisms by which such multidisciplinary in- terventions could interact with breast cancer recurrence and define the most effective modalities. 2 Journal of Oncology determined the effects of PAC.e intervention on the bi- 1. Introduction ological and anthropometric status of patients after one-year Over many years, growing knowledge has indicated the follow-up and the link between the biomarkers and disease- strong benefit of regular physical activity in the management free survival with seven years of follow-up after completion of breast cancer patients [1]. Despite an extensive literature of breast cancer treatment. of clinical trials, data from these studies showed positive but modest effects, which may be underestimated due to great 2. Patients and Methods variability in the intervention strategies and intensity of monitoring [2, 3]. .ese interventions produce short-term 2.1. Participants. Two hundred and fifty-one nonmetastatic changes in physical activity and patient behaviour, but data breast cancer patients were enrolled between 2008 and 2010, are scarce on recurrence and long-term follow-up. Some as previously described [14]. .e main inclusion criteria studies have highlighted long-term barriers to exercise after were notably invasive nonmetastatic breast carcinoma; less diagnosis of breast cancer, including psychological barriers than 9 months after chemotherapy/radiotherapy comple- 2, (e.g., low motivation and dislike of gym), environmental tion, complete remission, 18.5< BMI< 40 kg/m and writ- barriers (e.g., employment priority and low access to fa- ten informed consent. Half of the 251 patients (n � 113) were cilities), and lack of time [4]. Regarding the large variability investigated for biological parameters in the present study. of practice procedures, further research is required to in- vestigate how to sustain positive effects of exercise over time and to determine essential attributes of exercise (mode, 2.2. Study Design. Patients were randomized into two groups: “SPA,” for the group attending the 2-week session in intensity, frequency, duration, and timing) by cancer type and cancer treatment for optimal effects [5]. .e intro- thermal centres, and “CTR,” for the control group. .e 2- week session performed in thermal centres included con- duction of wearable activity monitors into cancer care could improve the understanding of the association between sultations with physicians, nutritionists, and psycho-on- physical activity and patient behaviour, as previously sug- cologists; physical activity supervised by a physiotherapist gested [1]. for 2 h daily with endurance activities, strength training, and Moreover, analyses are needed to provide insight into flexibility/stretching; SPA care consisting of bath, shower, how physical activity interventions work. Such studies and massage for half an hour per day; aesthetic care; and should accelerate the identification of effective behaviour dietary meals with adapted menus, dietary education, and caloric intake limited to 1700–2000 kcal/day. changes and permit the development of evidence-based practice with better standardisation. Currently, the mech- Besides standard oncological follow-up of the patients in the two groups, personal consultations with a dietician were anisms by which physical activity mediates its benefits re- main unclear [6]. Most hypotheses regarding the biological organized to perform anthropometric measurements, pro- pathways have focused on the impact of obesity on breast vide dietary advice, and give encouragement for daily cancer risk and recurrence. In that field, the main research physical activity. Evaluation of survival/recurrence was axes are, first, the implication of sex hormones, including made by patients’ oncologist, with a follow-up period of 7 both oestrogens and androgens (testosterone) [7]; second, years [14]. .e overall protocol design is available in a the implication of metabolic hormones, such as insulin/ supplementary file. insulin-like growth factor (IGF) axis and adipocytokines (leptin and adiponectin) [8]; and third, the implication of 2.3. Data Collection. Before randomization and at one year, inflammatory factors (C reactive protein, CRP) [9]. None of the following analyses were performed on half of the pop- these axes has clearly demonstrated efficiency in clinical ulation (SPA: n � 57; CTR: n � 56): trials, despite evidence of increased quality of life (QoL), reduced body weight in obese patients, and reduced (1) Diet questionnaire recurrence. Dietitians evaluated oral intake based on a 72-h self- .e majority of studies that investigate the benefits of reported diet questionnaire. physical activity and nutritional interventions in breast (2) Body composition cancer focus on weight loss, cardiorespiratory capacity, QoL, and overall well-being [5, 10, 11], but few of them considered Body weight was measured at each personal con- the biological parameters of the patients in their outcomes sultation. Lean body mass (LBM), fat mass (FM), and [12, 13]. total body water were evaluated by multifrequency Taking into account these data and the interactions bioelectrical impedance analysis (Bodystat Quadscan between physical activity and BMI, we performed a pro- 4000) using 5, 50, 100, and 200 kHz. Tricipital skin- spective randomized trial “Programme of Accompanying fold thickness was measured using a skin-fold caliper women after breast Cancer treatment completion in .er- (Harpenden caliper). To assess central fat distribu- mal resorts” (PAC.e) for complete-responder breast tion, the waist circumference (WC) was evaluated to cancer patients after chemotherapy. In this trial, we dem- the nearest 0.5 cm using a standard tape measure onstrated that the 2-week intervention durably influences placed between the lowest rib and the iliac crest, with the QoL of breast cancer patients after both short-term [14] the patient in the standing position. .e hip cir- and long-term treatment [15]. In the present study, we cumference (HC) was estimated using a standard Journal of Oncology 3 tape measure placed horizontally at the widest point patients are referred to hereafter as the biological study on the hip. population. At one year post-inclusion, 13 patients withdrew for familial or professional reasons, and 53 and 47 patients (3) Blood sampling and biological assays remained, respectively, for the SPA and CTR groups. .e Blood samples were collected at allocation and at one main covariates were distributed similarly between the al- year. Plasma levels of biomarkers were determined as location groups (Table 1). Cancer treatments were similar follows: glucose and HDL-cholesterol (colorimetry and standard for invasive tumours. Most patients’ tumours methods), C-reactive protein, and transthyretin were HR positive and treated using hormonotherapy, and a (immunonephelometry) were determined at the few (Her2+ tumours) using targeted therapy. biomedical laboratory of the recruiting centre; in- sulin and testosterone (ELISA) were determined at the hospital biochemistry laboratory (Clermont- 3.1. Diet, Body, and Biological Parameters at Allocation. Ferrand); IGF-1, leptin, and adiponectin (luminex) Results of the biological study population were considered in were determined at the Genotool platform (Tou- function of BMI scale and divided into three subgroups, i.e., louse); and CA 15-3 was determined at the anti- 2 2 ≤25 kg/m for normal BMI, [25–30 kg/m ] for overweight, cancer centre radiobiology laboratory (Clermont- and >30 for obesity (Tables 2 & 3). Overall diet mean results Ferrand). (Table 2) were within adult nutritional recommendations (4) Recurrence follow-up (17.3%± 4.1, 46.7%± 10.4, and 35.5%± 8.6, respectively, for protein, carbohydrate, and lipid intakes). A large dispersion Disease-free interval was computed as months of values was observed, resulting in no significant difference elapsed from date of randomization to documented between BMI subgroups except for total energy intake (TEI) breast cancer recurrence during seven years after (p � 0.038) and lipid intake in gram/day (p � 0.034). .e breast cancer treatment completion. All recurrence large standard deviation for each nutrient-intake value re- types were considered, either local or distant (nodes, flected an unbalance diet for some patients in the three BMI metastatis, and/or contralateral breast cancer). subgroups. All body parameters (Table 2) differed significantly by 2.4. Statistical Considerations. Protocol design consisted of a BMI subgroup (p< 10 − 7). As expected, the lean mass/fat multicentre parallel randomized prospective trial. Data were mass ratio decreased with the BMI due to the expansion of analysed using the intention-to-treat principle. Descriptive the body fat mass, i.e., 2.4± 0.6, 1.7± 0.3, and 1.3± 0.3, statistics are presented with mean± standard deviation (SD) respectively, for normal, overweight, and obese subgroups for Gaussian quantitative variables. Outcomes are shown (p< 10 − 7). with 95% confidence intervals. Categorical variables are As previously noticed, we observed a large dispersion of described using counts by class and frequencies (%). all biological parameter values (Table 3) regardless of BMI Comparison of outcomes per allocation group and per subgroup. Increased plasma levels of CRP (p< 10 − 5), in- BMI class was tested with Student’s t-test, one-way analysis sulin (p< 10 − 4), and leptin (p< 10 − 7) showed dysme- of variance (ANOVA), or the Kruskal-Wallis H-test tabolic disorders associated with overweight/obesity. As depending on homoscedasticity or normality of distribu- expected, the ratio of leptin/adiponectin significantly in- tions. Two-way ANOVA was used to compare longitudinal creased with BMI (0.53± 0.51, 1.26± 1.28, and 3.23± 3.86, variations between allocation groups, but without an in- respectively, for normal, overweight, and obese groups, teraction test because of unequal class sizes. Categorical data p< 10 − 7). Conversely, a significant decrease in HDL-C were compared with chi test. To test the association between level with BMI (p< 10 − 4) was observed. Transthyretin, two quantitative parameters, Pearson’s correlation coeffi- similar between groups, was in the physiological range, cient was used, or Spearman’s rank correlation if distribu- showing no malnutrition disorders in the studied pop- tions were not Gaussian. Survival curves were drawn using ulation. Other parameters (glucose, IGF-1, testosterone, and Kaplan-Meier’s method, and comparison of curves was CA 15-3) were in the normal range, with no difference performed using the Log-rank test. A backward and stepwise between BMI groups except for CA 15-3 (p � 0.014). Cox proportional hazard regression model was used to perform the multivariate analysis of survival. Cutoff values of biological parameters to draw survival curves were chosen 3.2. Changes in Diet, Body, and Biological Parameters One among quartiles of distribution. Year Later. One year after inclusion, Diet consumption, All tests were two-sided and the nominal level of sig- body, and biological parameters of patients were reevaluated nificance was 5%. Randomisation and statistics were per- one year after inclusion. All the raw data are presented by formed using SEM software [16]. BMI subgroups in two supplementary data files: one for the SPA group (Supplementary Table 1) and one for the CTR group (Supplementary Table 2). Variations in each pa- 3. Results rameter between inclusion and one-year follow-up are Biological parameters were evaluated at allocation for half of shown in Tables 4 and 5 and analyzed according to the the 251 patients: n � 57 for the “SPA” experimental group intervention group (SPA effect), one-year follow-up (time and n � 56 for the “CTR” control group (Figure 1). .ese 113 effect), and BMI subgroups (BMI effect). 4 Journal of Oncology Enrollment Assessed for eligibility (n = 450) 199 patients refused to participate: (I) personal reasons (n = 58) (II) health difficulties (n = 45) (III) not interested (n = 36) (IV) familial reasons (n = 28) (V) transport problems (n = 15) (VI) work resumption (n = 12) (VII) want to forget the cancer (n = 5) Randomized (n = 251) Allocation Allocated to intervention (n = 126) Allocated to intervention (n = 125) Received SPA intervention (n = 117) Received CTR intervention (n = 115) Did not received SPA intervention (n = 9) Did not received CTR intervention (n = 10) 6 for personal reasons 6 because randomized to CTR group 3 for professional reasons 4 refused to continue Early exit < 1 year Early exit < 1 year Biology analysis 3 for personal reasons 3 for personal reasons on half of the population ∗ ∗ Biology and diet (n = 57) Biology and diet (n = 56) Allocation Allocation 1-year follow-up (n = 49) 1-year follow-up (n = 55) 2 samples missing 7 samples missing Follow-up Survival (n = 56) Survival (n = 55) Follow-up (years) Follow-up (years) median = 5.2 [0.5-6.9] median = 4.8 [0.3-6.8] 1 lost of view 1 lost of view Figure 1: Allocation diagram and flow chart. Diet, nutritional, and body data collection. No significant difference was observed for diet pa- considering both SPA and time effects. For the SPA and CTR rameters (Table 4) regardless of the intervention group, the >30 BMI subgroups, a reduction in brachial and abdominal time window, or the BMI subgroup, except for the total circumferences tended to correlate with an increase in hip energy intake with time (p � 0.039). For the SPA group, circumference. total energy intake remained stable for BMI subgroups ≤25 No significant SPA effect was observed for biological and [25–30 kg/m ], whereas a strong reduction (−400 kcal/ parameters (Table 5), except for transthyretin (p � 0.041) d) in the BMI >30 subgroup led to both carbohydrate and CA 15-3 (p � 0.04) plasma levels, although these (−21.5%) and lipid (−13.8%) intake decreases without remained in the normal ranges. For the time effect, a sig- change in patients’ weight. For the CTR group, total energy nificant increase in both glucose (p � 0.04) and insulin intake decreased for ≤25 and>30 BMI subgroups due to a (p � 0.035) and a decrease in HDL-C (p � 0.027) plasma reduction in protein, carbohydrate, and lipid intakes. levels were observed. As expected, several parameter vari- However, an increase in the mean body weight of 1 kg was ations were related to BMI in the two groups as previously observed for each BMI subgroup (supplementary data), shown at allocation. Notably, we noticed an increase in which was not significant because of the large dispersion of glucose (p< 10 − 6), insulin (p< 10 − 7), and adiponectin individual values. (p � 0.022) plasma levels regardless of the intervention For body parameters (Table 4), we observed that only the group and more accentuated plasma levels for the >30 BMI − 7 BMI effect was significant (p< 10 ). All the parameters subgroups. Conversely, a decrease in HDL-C plasma levels were significantly related to BMI but remained stable was observed (p � 0.007). Journal of Oncology 5 Table 1: Study population characterization. SPA group (n � 57) CTR group (n � 56) Parameter p value Size or mean± SD (%) or [mini-max] Size or mean± SD (%) or [mini-max] 52.0± 7.2 51.9± 10.6 Patients’ age at allocation 0.97 [36–66] [29–71] Menopausal status Yes � 33 (58%) Yes � 35 (63%) 0.62 25.4± 4.6 25.5± 4.4 BMI—body mass index (kg/m ) 0.92 [18.4–35.9] [18.0–38.7] ≤25 kg/m 30 (53%) 27 (48%) BMI—class 25–30 kg/m 16 (28%) 22 (39%) 0.37 >30 kg/m 11 (19%) 7 (13%) 55 9± 15.2 56.8± 14.0 SF36—global score/100 0.30 [19.0–93.0] [29.0–95.0] Surgery for breast cancer Yes � 57 (100%) Yes � 55 (98%) 0.50 Radiotherapy Yes � 54 (95%) Yes � 54 (96%) 0.98 Hormonotherapy Yes � 43 (75%) Yes � 43 (77%) 0.87 Herceptin Yes � 5 (9%) Yes � 7 (13%) 0.56 Chemotherapies: number of 6.3± 1.1 6.0± 0.8 0.29 cycles [5–15] [3–9] .e main covariates of the studied population at allocation are presented with mean ± standard deviation (SD) for Gaussian quantitative variables. Outcomes are shown with 95% confidence intervals. Categorical variables were described using counts by class and frequencies (%). Comparison of outcomes was tested with Student’s t-test or the Kruskal-Wallis H-test depending on homoscedasticity or normality of distributions. Categorical data were compared with the chi test. All tests were two-sided, and the nominal level of significance was 5%. Table 2: Diet and Body parameters at allocation. BMI (kg/m ) Mean± σ All groups (n � 113) p value of BMI effect ≤25 (n � 57) 25–30 (n � 38) >30 (n � 18) Diet parameters Total energy intake (TEI) (kcal/d) 1492± 450 1540± 358 1325± 378 1689± 678 0.038 Protein intake (g/d) 63.6± 20.2 65.3± 15.1 58.7± 20.0 68.8± 30.1 0.86 (% TEI) 17.3± 4.1 17.2± 3.5 17.9± 5.2 16.4± 3.1 0.71 Carbohydrate intake (g/d) 172.6± 61.5 175.3± 54.1 156.8± 53.7 197.2± 85.1 0.65 (% TEI) 46.7± 10.4 45.4± 9.5 48.1± 12.8 47.8± 6.5 0.75 Lipid intake (g/d) 59.7± 25.4 63.5± 22.3 50.6± 23.6 66.8± 31.8 0.034 % TEI 35.5± 8.6 36.8± 8.4 33.5± 9.8 35.8± 5.0 0.14 Body parameters −7 Body weight (kg) 65.2± 12.5 56.6± 6.4 68.5± 5.8 85.3± 10.7 <10 −7 Lean mass (LM) (kg) 42.1± 5.8 39.6± 4.5 43.0± 4.8 47.9± 6.3 <10 −7 (%) 65.2± 6.8 69.6± 5.3 62.9± 3.7 56.3± 4.5 <10 −7 Fat mass (FM) (kg) 23.0± 7.8 17.2± 3.7 25.5± 3.1 36.2± 5.1 <10 −7 (%) 34.6± 6.7 30.1± 5.0 37.3± 3.8 43.1± 4.4 <10 −7 Ratio LM/FM 2.0± 0.6 2.4± 0.6 1.7± 0.3 1.3± 0.3 <10 −7 Cell mass (kg) 25.0± 4.0 22.8± 2.5 25.5± 3.3 30.7± 3.5 <10 −7 Total water (l) 32.9± 3.9 31.1± 2.6 33.2± 2.7 38.1± 4.6 <10 −7 (%) 51.3± 5.4 55.1± 4.0 48.5± 3.2 44.9± 2.8 <10 −7 Extracellular water (%) 24.3± 3.4 25.7± 1.7 23.1± 2.1 22.8± 6.5 <10 −7 Intracellular water (%) 27.1± 2.4 28.2± 1.8 26.1± 2.8 25.6± 1.2 <10 −7 Tricipital fold thickness (cm) 17.4± 8.6 12.5± 5.2 18.8± 7.2 29.6± 6.4 <10 −7 Arm circumference (cm) 30.2± 3.8 27.7± 2.2 31.1± 1.7 36.4± 3.3 <10 −7 Waist circumference (WC) (cm) 84.0± 13.5 75.4± 7.7 86.8± 9.0 105.5± 8.9 <10 −7 Hip circumference (HC) (cm) 101.1± 9.1 95.0± 4.9 103.5± 5.5 115.7± 5.6 <10 Ratio WC/HC 0.83± 0.09 0.79± 0.07 0.84± 0.09 0.92± 0.08 0.000017 Diet parameters for food intake are expressed in raw value (gram/day) and in % of total energy intake. Body parameters are expressed in raw value (kilogram or liter) and in % of body mass. Comparison of outcomes per BMI group at allocation was tested with one-way analysis of variance (ANOVA). .e test was two-sided, and the nominal level of significance was 5%. We found significant positive correlations in the bio- (p � −0.46, p< 10 − 7). .e leptin/adiponectin ratio was logical study population between leptin/adiponectin ratio strongly correlated with waist circumference (r � 0.67, and insulin (r � 0.46, p< 10 − 7) and CRP (r � 0.46, p< 10 − 7), BMI (r � 0.51, p< 10 − 7), and cell mass (r � 0.46, p< 10 − 7) and a negative correlation with HDL-C p< 10 − 7). Moreover, despite the absence of variation in 6 Journal of Oncology Table 3: Biological parameters at allocation. BMI (kg/m ) Mean± σ All groups (n � 113) p value of BMI effect ≤25 (n � 57) 25–30 (n � 38) >30 (n � 18) Glucose (mmol/l) 5.2± 0.6 5.1± 0. 4 5.2± 0.6 5.6± 0.8 0.25 HDL-cholesterol (mmol/l) 2.13± 1.28 2.35± 1.35 1.98± 1.25 1.70± 0.97 0.0001 Transthyretin (g/l) 0.26± 0.04 0.26± 0.04 0.26± 0.04 0.26± 0.04 0.88 C-reactive protein (mg/l) 2.5± 3. 6 1.3± 1.2 3.2± 4.4 5.2± 4.9 0.000002 Insulin (mUI/l) 6.5± 6.2 4.7± 4.4 6.4± 4.4 12.1± 9.8 0.000013 IGF-1 (μg/l) 96.4± 49.3 95.8± 45.6 103.5± 45.7 84.7± 62.6 0.23 −7 Leptin (μg/l) 5.7± 4.7 3.5± 2.6 6.0± 3.0 12.1± 6.0 <10 Adiponectin (mg/l) 8.1± 5.1 8.9± 5.3 7.6± 4.8 6.6± 4.4 0.072 −7 Leptin/adiponectin ratio 1.22± 2.02 0.53± 0.51 1.26± 1.28 3.23± 3.86 <10 Testosterone (nmol/l) 0.82± 0.36 0.79± 0.29 0.83± 0.42 0.87± 0.38 0.67 CA 15-3 (kU/l) 18.1± 18.7 20.1± 24.5 14.1± 9.0 19.7± 8.4 0.014 Plasma biological parameters are expressed in usual unit per liter. Comparison of outcomes per BMI group at allocation was tested with one-way analysis of variance (ANOVA). .e test was two-sided, and the nominal level of significance was 5%. 2 2 testosterone plasma level with SPA, time, or BMI effects, this ≤25 kg/m for normal BMI, [25–30 kg/m ] for overweight, parameter was significantly associated (i) positively with and >30 for obesity. At allocation, the study population’s body weight (r � +0.15, p � 0.03), cell mass (r � +0.19, repartition into BMI subgroups was similar to that of the p � 0.0072), arm circumference (r � +0.15, p � 0.026), WC/ same-age female French population, as previously described HC ratio (r � +0.15, p � 0.027), and transthyretin (r � +0.15, [17]. .e diet intakes are in accordance with the adult p � 0.028) and (ii) negatively with TEI (r � −0.16, p � 0.022) nutritional recommendations for all groups. We noted no and HDL-C (r � −0.19, p � 0.007). difference between the three subgroups but a great variation in declared intakes, particularly in the obese group, raising doubts as to the reliability of the consumption-data col- 3.3. Biological Parameters and Recurrence Relation. We lection based on a 72-h self-report. tested the association between biomarker plasma levels at At allocation, after the completion of breast cancer allocation expressed in quartiles and the risk of recurrence treatment, the biological and body parameters of the pop- during the seven-year follow-up. Highest HDL-cholesterol ulation were in accordance with the usual observed values values were associated with the best survival without re- for normal, overweight, and obesity status. Considering the currence (p � 0.047). Conversely, the lowest testosterone mean value for each parameter defined as EGIR metabolic and CA 15-3 values were associated with longer disease-free syndrome criteria (glucose> 6.1 mmol/l, HDL-C< 1 mmol/l, survival (p � 0.001 and 0.03, respectively) (Table 6). insulin >18 mUI/l (QR4), and waist circumference> 80 cm), .e survival curves for these three biomarkers were done neither overweight nor obesity subgroups met the three in function of the calculated significant threshold values required criteria [18]. Among these parameters, only the (2.13 mmol/l, 0.9 nmol/l, and 20 kUI/l, respectively, for central criterion of obesity (waist circumference) was above HDL-C, testosterone, and CA 15-3) (Figures 2(a), 2(b), 2(e)). the limit value and emerged as the earliest criterion of For testosterone, two other survival curves were plotted metabolic syndrome under our conditions. However, con- taking into account the hormonotherapy status of patients sidering the large value dispersion of all these parameters, (Figures 2(c), 2(d)). .ese latter showed that testosterone some patients of both overweight and obese groups could was relevant for disease-free survival only in patients treated present a metabolic syndrome. with hormonotherapy (p � 0.012 vs. p � 0.69, respectively, Obesity is well-known to be associated with elevated for patients with and without hormonotherapy). Using the circulating levels of insulin, insulin-like growth factor 1 Cox model, the link between these variables and disease-free (IGF-1), leptin, and inflammation [19]. In our study, we survival was tested and demonstrated that only the highest observed a significant increase in CRP, insulin, leptin testosterone values predicted increased recurrence risk (HR plasma levels, and the ratio leptin/adiponectin in parallel [CI–95%] � 5.06 [1.66–15.41], p � 0.004) (Figure 2(f)). with significantly increased adiposity markers (fat mass, arm, waist, and hip circumferences). As expected, circu- 4. Discussion lating anti-inflammatory adiponectin was decreased, reinforcing the sub-chronic inflammation associated with In the present study, we determined the effects of PAC.e intervention (i.e., medical, nutritional, and psychological obesity and related to the risk of recurrence [20]. Sur- prisingly, no difference was observed for IGF-1 and tes- monitoring; physical activity training; SPA; and aesthetic care) on the biological and anthropometric status of patients tosterone plasma contents, contrary to previous at allocation and after one-year follow-up. observations [8, 13], probably due to the huge variability of As obesity has an impact on biological status and is a risk individual values. .eir plasma concentrations were factor for breast cancer, we chose to discuss the data maintained in the physiological range for the female according to three BMI subgroups defined as follows: population of corresponding age [21, 22]. Journal of Oncology 7 Table 4: Variation in diet and body parameters between one-year follow-up and allocation. SPA arm (n � 55) CTR arm (n � 49) p value effect of ≤25 (n � 29) ]25–30] (n � 15) >30 (n � 11) ≤25 (n � 23) 25–30 (n � 19) >30 (n � 7) SPA Time BMI Diet parameters −41.7± 400.5 +25.5± 556.4 −400.1± 527.6 −227.9± 362.6 +165.6± 410.4 −437.6± 955.1 Total energy intake kcal/d 0.91 0.039 0.15 (-0.02%) (+10.8%) (−18.7%) (-12.3%) (+20.1%) (−10.3%) −4.6± 17.9 +3.2± 25.5 −5.8± 19.0 −1.78± 27.1 +5.1± 21.4 −15.4± 45.4 Protein intake g/d 0.71 0.24 0.35 (-3.9%) (+17.5%) (-3.1%) (+6.0%) (+26.1%) (−1.4%) +8.4± 52.2 +6.1± 63.7 −57.9± 78.3 −28.4± 54.8 +6.7± 51.2 −46.7± 121.1 Carbohydrate intake g/d 0.84 0.10 0.38 (+10.1%) (+16.7%) (−21.5%) (−9.8%) (+7.5%) (−1.5%) −5.5± 26.4 −1.3± 36.4 −12.7± 25.2 −9.4± 19.8 +13.0± 13.0 −24.6± 45.4 Lipid intake g/d 0.89 0.099 0.15 (+0.2%) (+24.0%) (−13.8%) (−7.9%) (+67.3%) (−21.1%) Body parameters −0.10± 2.16 +1.47± 4.00 −0.73± 5.76 −0.24± 2.67 +0.26± 3.86 +0.93± 2.41 –7 Body weight kg 0.56 0.45 <10 (−0.2%) (+2.2%) (−0.3%) (−0.5%) (+0.4%) (+1.1%) +1.03± 3.63 −0.25± 1.66 −0.34± 5.85 0.00± 3.20 −0.02± 3.14 +3.09± 7.07 –7 Lean mass (LM) % 0.18 0.85 <10 (+1.7%) (−0.4%) (−0.0%) (−0.0%) (+0.1%) (+6.0%) −0.85± 3.50 +0.25± 1.66 +0.31± 5.90 −0.00± 3.20 +0.02± 3.14 −1.64± 6.05 –7 Fat mass (FM) % 0.11 0.86 <10 (−2.7%) (+0.6%) (+2.0%) (0.0%) (+0.6%) (-3.6%) +0.10± 0.39 −0.01± 0.14 −0.03± 0.37 +0.04± 0.40 −0.01± 0.29 +0.17± 0.39 –7 LM/FM ratio 0.20 0.77 <10 (+5.0%) (−0.6%) (+1.6%) (+1.7%) (+0.9%) (+14.4%) +1.34± 6.57 +0.95 2.21 +0.54± 3.02 −0.83± 4.52 −0.66± 4.74 +2.15± 5.82 –7 Cell mass kg 0.19 0.34 <10 (+6.8%) (+5.2%) (+2.0%) (−2.8%) (−1.7%) (+7.3%) +0.25± 1.62 +0.51± 1.45 +0.13± 2.53 −0.06± 1.02 +0.11± 1.86 +1.07± 5.15 –7 Total water l 0.22 0.40 <10 (+0.9%) (+1.7%) (+0.8%) (−0.1%) (+0.5%) (+3.2%) +0.51± 1.72 −0.67± 1.67 +0.02± 2.33 +1.14± 4.1 0.98± 3.72 −3.97± 10.72 –7 Extracellular water % 0.80 0.86 <10 (+2.0%) (−2.3%) (+0.6%) (+4.9%) (+4.3%) (−7.2%) +2.21± 9.32 +0.61± 1.32 −0.08± 1.99 −0.35± 4.76 −0.59± 4.68 +1.65± 4.55 Intracellular water % 0.28 0.53 0.00012 (+7.7%) (+3.2%) (−0.2%) (−1.3%) (−1.9%) (+7.0%) +0.46± 3.74 −0.57± 7.78 −3.25± 5.41 +0.23± 5.05 +2.98± 5.60 −2.51± 4.95 –7 Tricipital fold thickness cm 0.36 0.69 <10 (+10.7%) (+1.5%) (−7.5%) (+14.1%) (+25.3%) (−6.3%) −0.10± 1.77 −1.00± 1.49 −1.46± 1.94 −0.46± 1.71 −0.32± 2.16 −0.93± 2.01 –7 Arm circumference cm 0.58 0.32 <10 (−0.3%) (−3.1%) (−3.6%) (−1.7%) (−1.1%) (−2.4%) −2.93± 4.41 +0.43± 3.68 −1.09± 5.00 −0.46± 5.65 +1.45± 8.51 −3.36± 6.24 –7 Waist circumference (WC) cm 0.73 0.81 <10 (−3.7%) (−0.4%) (−1.0%) (−0.4%) (+2.7%) (−3.0%) −0.59± 2.80 1.77± 2.83 0.23± 5.85 −1.30± 3.56 −0.13± 3.71 2.79± 5.53 –7 Hip circumference (HC) cm 0.80 0.66 <10 (−0.6%) (+1.7%) (+0.4%) (−1.1%) (−0.1%) (2.5%) −0.03± 00.05 −0.01± 0.02 −0.01± 0.07 +0.01± 0.05 +0.2± 0.08 −0.05± 0.06 –7 WC/HC ratio 0.66 0.61 <10 (−2.9%) (−1.5%) (−1.1%) (+0.9%) (+3.1%) (−4.9%) Variation for each parameter is expressed in raw value (one-year follow-up value minus allocation value) and in percentage of the allocation value: + sign indicates an increase and –sign indicates a decrease. Two- way ANOVA was used to compare longitudinal variations between allocation arms (SPA effect), or one-year follow-up (time effect), or BMI groups (BMI effect), but without interaction test because of unequal class sizes. All tests were two-sided, and the nominal level of significance was 5%. Significant p values are indicated in bold. 8 Journal of Oncology Table 5: Variation in biological parameters between one-year follow-up and allocation. SPA arm (n � 55) CTR arm (n � 49) p value effect of ≤25 25–30 >30 ≤25 25–30 >30 SPA Time BMI (n � 29) (n � 15) (n � 11) (n � 23) (n � 19) (n � 7) 0.007± 0.446 0.459± 0.748 0.749± 1.358 −0.089± 0.326 0.226± 0.394 1.75± 2.775 –6 Glucose mmol/l 0.23 0.04 <10 (0.6%) (8.9%) (15.9%) (−1.6%) (4.4%) (25.9%) −0.438± 1.471 −0.241± 1.34 −0.542± 1.31 −0.436± 1.384 −0.333± 1.635 −0.028± 0.198 HDL-cholesterol mmol/l 0.41 0.027 0.007 (−5.6%) (3.4%) (−13%) (−4.8%) (7%) (−1.3%) 0.001± 0.038 0.009± 0.026 −0.005± 0.027 0.001± 0.033 −0.007± 0.031 0.002± 0.051 Transthyretin g/l 0.041 0.75 0.79 (1%) (3.5%) (−1.5%) (0.8%) (−2%) (3.2%) −0.146± 1.142 0.238± 2.818 −0.127± 1.481 0.264± 1.168 −0.135± 4.736 −1± 5.545 –7 C-reactive protein mg/l 0.11 0.73 <10 (11.1%) (27.3%) (8.9%) (41.8%) (41.1%) (−6.6%) 0.17± 3.79 1.97± 4.59 2.51± 10.19 0.22± 6.58 4.47± 10.51 4.53± 6.24 –7 Insulin mUI/l 0.41 0.035 <10 (25.7%) (115.7%) (58.4%) (36.1%) (50.9%) (78.4%) −0.79± 27.51 −26.61± 27.25 −18.78± 23.9 −13.18± 39.9 −12.41± 34.83 9.13± 16.55 IGF-1 μg/l 0.32 0.072 0.31 (6.3%) (−19.7%) (−24%) (−4.2%) (−12.2%) (16.5%) −0.03± 1.57 −0.02± 2.72 −2.42± 6.24 0.12± 2.38 1.64± 6.28 −0.93± 3.78 –7 Leptin μg/l 0.66 0.81 <10 (15.6%) (9%) (−2.4%) (21.7%) (23.6%) (−7.4%) 2.33± 4.19 1.29± 2.3 0.57± 1.3 0.99± 2.51 1.15± 2.65 0.65± 1.01 Adiponectin mg/l 0.33 0.082 0.022 (32.6%) (17.3%) (18%) (13.9%) (11.5%) (5.6%) −0.1± 0.27 −0.08± 0.52 −1.68± 2.92 0.13± 1.21 1.2± 4.07 −0.27± 0.29 –6 Leptin/adiponectin ratio 0.73 0.91 2 × 10 (−2.9%) (−0.4%) (−18%) (24.2%) (16.6%) (−13.6%) −0.033± 0.299 −0.013± 0.098 0.045± 0.347 −0.051± 0.244 0.015± 0.332 0.029± 0.757 Testosterone nmol/l 0.086 0.83 0.27 (−3.6%) (5%) (10.7%) (−5%) (3.4%) (21.8%) 2.32± 3.08 0.62± 1.78 0.27± 1.71 −5.32± 28.45 1.56± 2.06 2.71± 1.48 CA 15-3 kU/l 0.04 0.68 0.07 (18.2%) (6.8%) (4.7%) (1.4%) (12.5%) (14%) Variation for each parameter is expressed in raw value (one-year follow-up value minus allocation value) and in percentage of the allocation value: + sign indicates an increase and – sign indicates a decrease. Two- way ANOVA was used to compare longitudinal variations between allocation arms (SPA effect), or one-year follow-up (time effect), or BMI groups (BMI effect), but without interaction test because of unequal class sizes. All tests were two-sided, and the nominal level of significance was 5%. Significant p values are indicated in bold. Journal of Oncology 9 Table 6: Prognostic value of biological parameters on disease-free survival over 7 years. .reshold Parameters at allocation (n � 111) Median (quartiles) st rd ≤1 quartile ≤ Median ≤3 quartile (+) Cholesterol-HDL (mmol/l) 1.78 [1.46–2.13] p � 0.64 p � 0.22 p � 0.047 (−) (−) Testosterone (nmol/l) 0.7 [0.7–0.9] ND p � 0.049 p � 0.001 (−) (−) CA 15-3 (kU/l) 14 [10–20] p � 0.28 p � 0.07 p � 0.03 Association of biological parameters at allocation with the recurrence risk was tested using a two-sided chi test. .e nominal level of significance was 5%. + sign indicates that high values are in favour of a better prognosis, while – sign indicates that these high values worsen prognosis. HDL-C ≥ 2.13 nmol/l (N = 31) 100 100 Testo. < 0.9 nmol/l (N = 87) HDL-C < 2.13 nmol/l (N = 80) 75 75 Testo. ≥ 0.9 nmol/l (N = 24) 50 50 25 25 p = 0.001 p = 0.047 0 0 0 years 2 468 0 years 2 468 (a) (b) Testo. ≥ 0.9 nmol/l (N = 3) Testo. < 0.9 nmol/l (N = 64) Testo. < 0.9 nmol/l (N = 23) Testo. ≥ 0.9 nmol/l (N = 21) p = 0.0053 p = 0.76 0 years 2468 0 years 2468 (c) (d) CA 15-3 ≤ 20 kUl/l (N = 84) Covariables  p 01 6 Hazard ratios [CI‐95%]  (n = 111)  HDL cholesterol 0.17 0.29 [0.05 – 1.72] CA 15-3 > 20 kUl/l (N = 27) Testosterone 0.0044 5.06 [1.66 – 15.41] CA 15‐3 0.48 1.01 [0.99 – 1.03] p = 0.03 0 years 2 4 6 8 (e) (f) Figure 2: Survival curves and hazard ratios for HDL-cholesterol, testosterone, and CA 15-3. (a) HDL-cholesterol. (b) Testosterone—all patients. (c) Testosterone—patients without hormonotherapy. (d) Testosterone—patients with hormonotherapy. (e) CA 15-3. (f) Hazard ratios (Cox model). .reshold values for HDL-cholesterol, testosterone, and CA 15-3 at allocation correspond to the 75% percentile values. .ey were used to draw survival curves using Kaplan-Meier’s method. Comparison of curves was performed using the Log-rank test. Backward stepwise Cox proportional hazard regression model was used to perform the multivariate analysis of survival. All tests were two- sided, and the nominal level of significance was 5%. Disease-free survival Disease-free survival Disease-free survival Disease-free survival Disease-free survival 10 Journal of Oncology carbohydrate and lipid intakes, demonstrated the efficacy of Globally, as measurements were performed after completion of breast cancer treatment, body and biological patient’s nutritional information. As described at allocation, BMI was the major factor parameters seemed to be more linked to BMI status than to breast disease. Nevertheless, as previously described conditioning body and biological parameter changes one [23–25], we cannot exclude that the breast cancer therapy year later. For body parameters, we noted high central may be another cause of metabolic disturbances at allo- adiposity (waist and hip circumferences) in the overweight cation. .at may be the reason for the great variability and obese groups. .e same biomarker variations were observed for all parameters regardless of the BMI observed and reinforced for the overweight and obese subgroup. subgroups (i.e., increase in insulin, leptin, and CRP, and decrease in HDL-C). Moreover, these metabolic disorders One year after inclusion, the impact of the SPA inter- vention on diet, body, and biological parameters was induced an increased glycaemia and a decreased adipo- nectinemia in relation to more pronounced insulin resis- evaluated. Only transthyretin and CA 15-3 plasma levels were significantly affected by the SPA intervention. Trans- tance and sub-chronic inflammation [20]. .us, the obese groups presented two EGIR criteria for metabolic syndrome thyretin, one of the thyroid hormone carriers, is recognized as an acute malnutrition marker whose hepatic synthesis is (glucose and waist circumference) one year after breast reduced in case of inflammation [26]. In our study, trans- cancer treatment completion. .is confirms previous studies thyretin levels remained in the normal range and seemed to establishing that breast cancer posttreatment increases the be without biological meaning in regard of their tiny vari- risk of metabolic syndrome [39, 40]. ations and the absence of inflammation and of lean mass Finally, we clarified the link between biological markers changes. Breast cancer is generally not associated with at allocation and disease-free survival over seven years of follow-up after breast cancer treatment completion. We malnutrition or sarcopenia, especially so long after treat- ment [27]. CA 15-3 is frequently used for diagnosis and confirmed the interest of three biomarkers commonly used in the determination of recurrence risk: the highest plasma follow-up of breast cancer [28]. In our study, an a posteriori bias appeared for these biomarker data because the CTR values of HDL-C and the lowest plasma values of testos- terone and CA 15-3 were associated with a reduced risk of group patients presented higher CA 15-3 concentrations than the SPA group at allocation (Supplementary Table 2). recurrence [41–43]. HDL-C is linked to metabolic disorders One year after treatment completion, as none of the patients and is often related to androgen metabolism [44]. Choles- was in recurrence, CA 15-3 values decreased under the terol is clearly demonstrated to be a key regulator of breast threshold of 30 kU/l, confirming the efficacy of the therapy cancer tumours [45]. Favouring liver cholesterol clearance, [29, 30]. In accordance with previous studies showing an increase in HDL-C limits the availability of cholesterol for modest effects on body and biological parameters of physical recurrent cancer stem cells [46]. In our study, patients with the highest circulating HDL-C presented the lowest recur- activity and nutritional interventions [31, 32], our study shows the lack of one-year impact of a 2-week SPA rence risk. However, this protective effect was not retrieved in the multivariate Cox model, limiting the interest of cir- intervention. Some metabolic disorder changes were pointed out at culating HDL-C determination in recurrence monitoring. one-year follow-up (time effect). Despite a decrease in total As previously noted, CA 15-3 is a useful marker for energy intake, patients presented an increase in glucose and breast cancer follow-up: the circulating value is directly insulin plasma levels associated with a decrease in HDL-C. related to the stage and mass of the tumour [29]. In our .ese parameters suggest the development of insulin re- study, although the lowest circulating CA 15-3 values sistance independently of the BMI effect for overweight were associated with the lowest recurrence risk, the patients and the reinforcement of insulin resistance for obese multivariate Cox model did not confirm this observation. patients. .ese observations are in agreement with previous .is is in agreement with the literature, which has established the interest in CA 15-3 for monitoring breast studies which considered breast cancer as a metabolic dis- ease, with insulin resistance, sub-chronic inflammation, and tumour growth, but its poor prognostic value for re- currence risk [28, 30]. dysmetabolism induced by therapy [33, 34]. Moreover, an increased risk for metabolic syndrome and obesity has been In our study, only testosterone presented a significant described in long-term breast cancer survivors [35]. hazard ratio with disease-free survival; that is, the highest If women with breast cancer frequently lose weight circulating values (>0.9 nmol/l) were associated with re- during chemotherapy, a common unwanted long-term effect currence risk multiplied by ≈5 (HR � 5.06 [1.66–15.41]). of this therapy is weight gain, which often ranges 2–6 kg Notably, this link between testosterone and recurrence risk [10, 36] and penalizes mainly patients with adjuvant therapy only applied to patients receiving adjuvant hormonotherapy. [37]. In our study, weight gain was modest (less than 1 kg) .is observation confirms Venturelli’s observation of in- and concerned mainly the overweight BMI groups, of whom creased recurrence risk for testosterone plasma concentra- tion above 0.96 nmol/l with a hazard ratio of 4.68 for the majority were under hormonal adjuvant therapy. .us, weight control and diet intervention are important to im- overweight women but not for obese ones [47]. Testosterone is strongly associated with the androgen hypothesis of breast prove care and control of recurrence risk in posttreatment breast cancer patients [38]. In our study, the reduction in the carcinogenesis, related to the conversion of androgen into total energy intake provided by diet modification, especially oestrogen by aromatase [13]. .is enzymatic activity is Journal of Oncology 11 increased in obese patients due to the expansion of adiposity Additional Points [48]. However, it is not clear whether testosterone per se is directly responsible for promoting breast cancer risk or Highlights. (1) After breast cancer treatment completion, whether it is just a marker of the dysmetabolism linked to changes in anthropometric and biological parameters are overweight and obesity [49]. .is later hypothesis was mainly dependent on the patient’s BMI level. (2) A rein- confirmed in our study by the significant correlation of forcement of insulin resistance is observed in overweight plasma testosterone with several body and biological and obese patients after one-year treatment completion, markers associated with this dysmetabolism (positively with independently of physical activity and nutritional inter- body weight and ratio of WC/HC, and negatively with HDL- vention. (3) Testosterone plasma levels at the time of C). treatment completion are associated with recurrence risk in Our trial suffers from several limitations: patients receiving adjuvant hormonotherapy. (1) First, the small numbers of patients divided into different BMI subgroups limited the reliability of the Ethical Approval statistical analysis. (2) Second, the determination of biological parameters .e protocol was approved by the AFSSAPS (French Agency at one-year follow-up did not permit the charac- for Sanitary Security of Health Products), the regional Ethics terization of the short-term benefits of our 2-week Committee (2008), and the French National Committee SPA intervention. Moreover, the one-year time controlling personal computerized data (CNIL). .is trial window could explain the weak impact of this in- was performed in compliance with the Helsinki declaration tervention on the biological parameters. and registered in ClinicalTrials.gov with the no. NCT01563588. (3) .ird, the mismatches observed between diet con- sumption and weight changes of patients question the reliability of data collection using the 72-h self- Consent reported diet questionnaire. Written informed consent was obtained from all individual Few studies investigating the benefits of physical ac- participants included in the study. tivity and nutritional interventions in cancer survivors have considered the biological status of the patients in Conflicts of Interest their outcomes. Our data demonstrated that the health changes of patients were mainly related to their body .e authors declare that they have no conflicts of interest. condition and highlighted the importance of evaluating biological and anthropometric status in monitoring Authors’ Contributions cancer survivors. M.-P. V., F. K., M. D., and Y.-J. B. contributed in study 5. Conclusion conception and design. M.-A. M.-R., I. V. P.-D., A. T., and S. J. contributed in patient inclusion and follow-up. M.-P. V., To conclude, our study shows that one year after a global F. K., and A. R. contributed in acquisition, analysis, and multidisciplinary supportive and educational intervention, interpretation of the data. M.-P. V., F. K., A. R., and Y.-J. B. few anthropometric and biological changes could be at- drafted the manuscript. All authors gave the final approval. tributed to this intervention. It demonstrates that the one- year changes of patients are mainly related to their body Acknowledgments mass index (BMI) and confirms the importance of taking into account biological markers of metabolic status in the .e authors thank all patients who participated in the trial; follow-up of posttherapy breast disease. Among the tools the contributors of the programme at the Centre Jean Perrin, needed for this monitoring, our study highlights the in- Pole ˆ Sante´ Republique ´ (Clermont-Ferrand); .ermal resorts terest of plasma testosterone in the evaluation of recurrence (Vichy, Le-Mont-Dore, Chatel-Guyon); ˆ and others who risk. .ese observations may help reinforce care recom- supported the study. .e authors especially thank Nancy mendations for cancer survivors but need to be confirmed Uhrhammer for her careful language editing. .e trial was on a large population for a more comprehensive approach. financed by AFRETH (French Association for Hydrother- Future studies would permit a better understanding of the mal Research) and Auvergne Regional Council, Clermont- mechanisms by which such multidisciplinary interventions Communaute, ´ League against Cancer (Puy de Dome ˆ could interact with breast cancer recurrence and help committee CD63). define the most effective modalities. Supplementary Materials Data Availability Table 1: diet, body, and biological parameters for the SPA .e data used to support the findings of this study are arm at allocation and one-year follow-up. Table 2: diet, body, available from the corresponding author upon request. and biological parameters for the CTR arm at allocation and 12 Journal of Oncology [15] F. Kwiatkowski, M.-A. Mouret-Reynier, M. Duclos et al., one-year follow-up. Overall protocol design. (Supplementary “Long-term improvement of breast cancer survivors’ quality Materials) of life by a 2-week group physical and educational inter- vention: 5-year update of the ‘PAC.e’ trial,” British Journal References of Cancer, vol. 116, no. 11, pp. 1389–1393, 2017. [16] F. Kwiatkowski, M. Girard, K. Hacene, and J. Berlie, “Sem: a [1] G. Gresham, J. Schrack, L. M. Gresham et al., “Wearable suitable statistical software adaptated for research in oncol- activity monitors in oncology trials: current use of an ogy,” Bull Cancer, vol. 87, no. 10, pp. 715–721, 2000. emerging technology,” Contemporary Clinical Trials, vol. 64, [17] M. Arnold, M. Touillaud, L. Dossus et al., “Cancers in France pp. 13–21, 2018. in 2015 attributable to high body mass index,” Cancer Epi- [2] J. A. Ligibel and H. D. Strickler, “Obesity and its impact on demiology, vol. 52, pp. 15–19, 2018. breast cancer: tumor incidence, recurrence, survival, and [18] B. Balkau and M. A. Charles, “Comment on the provisional possible interventions,” American Society of Clinical Oncology report from the WHO consultation. European group for the Educational Book, vol. 33, pp. 52–59, 2013. study of insulin resistance (EGIR),” Diabet Medicine, vol. 16, [3] K. E. Post and J. Flanagan, “Web based survivorship inter- no. 5, pp. 442-443, 1999. ventions for women with breast cancer: an integrative review,” [19] D. F. Argolo, C. A. Hudis, and N. M. Iyengar, “.e impact of European Journal of Oncology Nursing, vol. 25, pp. 90–99, obesity on breast cancer,” Current Oncology Reports, vol. 20, no. 6, p. 47, 2018. [4] S. Y. Loh and A. Nadia, “Methods to improve rehabilitation of [20] S.-A. Lee, H. Sung, W. Han, D.-Y. Noh, S.-H. Ahn, and patients following breast cancer surgery: a review of sys- D. Kang, “Serum adiponectin but not leptin at diagnosis as a tematic reviews,” Breast Cancer: Targets and ?erapy, predictor of breast cancer survival,” Asian Pacific Journal of vol. 2015, no. 7, pp. 81–98, 2015. Cancer Prevention, vol. 15, no. 15, pp. 6137–6143, 2014. [5] S. I. Mishra, R. W. Scherer, P. M Geigle et al., “Exercise in- [21] H. Ketha and R. J. Singh, “Clinical assays for quantitation of terventions on health-related quality of life for cancer sur- insulin-like-growth-factor-1 (IGF1),” Methods, vol. 81, vivors,” ?e Cochrane Database of Systematic Reviews, vol. 15, pp. 93–98, 2015. no. 8, Article ID CD007566, 2012. [22] K. M. Wooding, J. A. Hankin, C. A. Johnson et al., “Mea- [6] M. C. De Boer, E. A. Worner, ¨ D. Verlaan, and surement of estradiol, estrone, and testosterone in post- P. A. M. Van Leeuwen, “.e mechanisms and effects of menopausal human serum by isotope dilution liquid physical activity on breast cancer,” Clinical Breast Cancer, chromatography tandem mass spectrometry without deriv- vol. 17, no. 4, pp. 272–278, 2017. atization,” Steroids, vol. 96, pp. 89–94, 2015. [7] G. Secreto, E. Venturelli, E. Meneghini et al., “Testosterone [23] T. Coskun, F. Kosova, Z. Ari, A. Sakarya, and Y. Kaya, “Effect and biological characteristics of breast cancers in postmen- of oncological treatment on serum adipocytokine levels in opausal women,” Cancer Epidemiology Biomarkers & Pre- patients with stage II-III breast cancer,” Molecular and vention, vol. 18, no. 11, pp. 2942–2948, 2009. Clinical Oncology, vol. 4, no. 5, pp. 893–897, 2016. [8] L. Mauro, G. D. Naimo, E. Ricchio, M. L. Panno, and S. Ando, ` [24] A. Alacacioglu, L. Kebapcilar, Z. Gokgoz et al., “Leptin, insulin “Cross-talk between adiponectin and IGF-IR in breast can- and body composition changes during adjuvant taxane based cer,” Frontiers in Oncology, vol. 5, p. 157, 2015. chemotherapy in patients with breast cancer, preliminary [9] N. M. Iyengar, X. K. Zhou, A. Gucalp et al., “Systemic cor- study,” Indian Journal of Cancer, vol. 53, no. 1, pp. 39–42, relates of white adipose tissue inflammation in early-stage breast cancer,” Clinical Cancer Research, vol. 22, no. 9, [25] J. Bao, N. Borja, M. Rao et al., “Impact of weight change pp. 2283–2289, 2016. during neoadjuvant chemotherapy on pathologic response in [10] M. N. Kirshbaum, “A review of the benefits of whole body triple-negative breast cancer,” Cancer Medicine, vol. 4, no. 4, exercise during and after treatment for breast cancer,” Journal pp. 500–506, 2015. of Clinical Nursing, vol. 16, no. 1, pp. 104–121, 2007. [26] A. Tsuboi, M. Terazawa-Watanabe, T. Kazumi, and K. Fukuo, [11] L. L. Peterson and J. A. Ligibel, “Physical activity and breast “Associations of decreased serum transthyretin with elevated cancer: an opportunity to improve outcomes,” Current On- high-sensitivity CRP, serum copper and decreased hemo- cology Reports, vol. 20, no. 7, p. 50, 2018. globin in ambulatory elderly women,” Asia Pacific Journal of [12] C. L. Rock, C. Pande, S. W. Flatt et al., “Favorable changes in Clinical Nutrition, vol. 24, no. 1, pp. 83–89, 2015. serum estrogens and other biologic factors after weight loss in [27] S. Steinbach, T. Hummel, C. Bohner ¨ et al., “Qualitative and breast cancer survivors who are overweight or obese,” Clinical quantitative assessment of taste and smell changes in patients Breast Cancer, vol. 13, no. 3, pp. 188–195, 2013. undergoing chemotherapy for breast cancer or gynecologic [13] D. Coradini, A. Orenti, E. Venturelli, A. Cavalleri, malignancies,” Journal of Clinical Oncology, vol. 27, no. 11, E. Biganzoli, and S. Oriana, “Serum levels of testosterone and pp. 1899–1905, 2009. SHBG in association with body mass index improve the [28] J. G. Santillan-Ben ´ ´ıtez, H. Mendieta-Zeron, ´ L. M. Gomez- ´ predictive capability of consolidate tumor biomarkers in pre- Olivan ´ et al., “.e tetrad BMI, leptin, leptin/adiponectin (L/A) and postmenopausal breast cancer patients,” Japanese Journal ratio and CA 15-3 are reliable biomarkers of breast cancer,” of Clinical Oncology, vol. 48, no. 4, pp. 308–316, 2018. Journal of Clinical Laboratory Analysis, vol. 27, no. 1, [14] F. Kwiatkowski, M. A. Mouret-Reynier, M. Duclos et al., “Long term improved quality of life by a 2-week group pp. 12–20, 2013. [29] Y. Fujimoto, T. Higuchi, A. Nishimukai et al., “High levels of physical and educational intervention shortly after breast cancer chemotherapy completion. Results of the ‘programme serum CA15-3 and residual invasive tumor size are associated with poor prognosis for breast cancer patients with non- of accompanying women after breast cancer treatment completion in thermal resorts’ (PAC.e) randomised clinical pathological complete response after neoadjuvant chemo- therapy,” Journal of Surgical Oncology, vol. 118, no. 1, trial of 251 patients,” European Journal of Cancer, vol. 49, no. 7, pp. 1530–1538, 2013. pp. 228–237, 2018. Journal of Oncology 13 [30] A. Nicolini, A. Carpi, P. Ferrari et al., “An individual reference [45] C. Bobin-Dubigeon, A. Chauvin, V. Brillaud-Meflah, F. Boiffard, M.-P. Joalland, and J.-M. Bard, “Liver X receptor limit of the serum CEA-TPA-CA 15-3 tumor marker panel in the surveillance of asymptomatic women following surgery (LXR)-regulated genes of cholesterol trafficking and breast cancer severity,” Anticancer Research, vol. 37, no. 10, for primary breast cancer,” Cancer Management and Research, vol. 10, pp. 6879–6886, 2018. pp. 5495–5498, 2017. [46] K. Pelton, C. M. Coticchia, A. S. Curatolo et al., “Hyper- [31] R. T. Chlebowski, “Nutrition and physical activity influence cholesterolemia induces angiogenesis and accelerates growth on breast cancer incidence and outcome,” ?e Breast, vol. 22, of breast tumors in vivo,” ?e American Journal of Pathology, no. Suppl 2, pp. S30–S37, 2013. vol. 184, no. 7, pp. 2099–2110, 2014. [32] A.-M. Foucaut, S. E. Berthouze-Aranda, M. Touillaud et al., [47] E. Venturelli, A. Orenti, A. S. C. Fabricio et al., “Observational “Reduction of health risk factors through an adapted physical study on the prognostic value of testosterone and adiposity in activity program in patients with breast cancer,” Supportive postmenopausal estrogen receptor positive breast cancer Care in Cancer, vol. 22, no. 4, pp. 1097–1104, 2014. patients,” BMC Cancer, vol. 18, no. 1, p. 651, 2018. [33] R. Bhandari, G. A. Kelley, T. A. Hartley, and I. R. H. Rockett, [48] C. Gerard ´ and K. A. Brown, “Obesity and breast cancer—role “Metabolic syndrome is associated with increased breast of estrogens and the molecular underpinnings of aromatase cancer risk: a systematic review with meta-analysis,” Inter- regulation in breast adipose tissue,” Molecular and Cellular national Journal of Breast Cancer, vol. 2014, Article ID Endocrinology, vol. 466, pp. 15–30, 2018. 189384, 13 pages, 2014. [49] F. Bertolini, “Adipose tissue and breast cancer progression: a [34] D. T. Dibaba, D. Braithwaite, and T. Akinyemiju, “Metabolic link between metabolism and cancer,” ?e Breast, vol. 22, syndrome and the risk of breast cancer and subtypes by race, pp. S48–S49, 2013. menopause and BMI,” Cancers, vol. 10, no. 9, p. 299, 2018. [35] A. Engin, “Obesity-associated breast cancer: analysis of risk factors,” in Obesity and Lipotoxicity, vol. 960, pp. 571–606, Springer, Berlin, Germany, 2017. [36] M. Sadim, Y. Xu, K. Selig et al., “A prospective evaluation of clinical and genetic predictors of weight changes in breast cancer survivors,” Cancer, vol. 123, no. 13, pp. 2413–2421, [37] A. Ginzac, E. .ivat, M.-A. Mouret-Reynier et al., “Weight evolution during endocrine therapy for breast cancer in postmenopausal patients: effect of initial fat mass percentage and previous adjuvant treatments,” Clinical Breast Cancer, vol. 18, no. 5, pp. e1093–e1102, 2018. [38] M. De Roon, A. M. May, A. McTiernan et al., “Effect of exercise and/or reduced calorie dietary interventions on breast cancer-related endogenous sex hormones in healthy postmenopausal women,” Breast Cancer Research, vol. 20, no. 1, p. 81, 2018. [39] W. K. Cho, D. H. Choi, W. Park et al., “Effect of body mass index on survival in breast cancer patients according to subtype, metabolic syndrome, and treatment,” Clinical Breast Cancer, vol. 18, no. 5, pp. e1141–e1147, 2018. [40] V. Boonyaratanakornkit and P. Pateetin, “.e role of ovarian sex steroids in metabolic homeostasis, obesity, and post- menopausal breast cancer: molecular mechanisms and ther- apeutic implications,” BioMed Research International, vol. 2015, Article ID 140196, 13 pages, 2015. [41] E. R. Nelson, “.e significance of cholesterol and its me- tabolite, 27-hydroxycholesterol in breast cancer,” Molecular and Cellular Endocrinology, vol. 466, pp. 73–80, 2018. [42] A. Micheli, E. Meneghini, G. Secreto et al., “Plasma testos- terone and prognosis of postmenopausal breast cancer pa- tients,” Journal of Clinical Oncology, vol. 25, no. 19, pp. 2685–2690, 2007. [43] P. Kachhawa, K. Kachhawa, D. Agrawal, V. Sinha, P. D. Sarkar, and S. Kumar, “Association of dyslipidemia, increased insulin resistance, and serum CA 15-3 with in- creased risk of breast cancer in urban areas of north and central India,” Journal of Mid-Life Health, vol. 9, no. 2, pp. 85–91, 2018. [44] M. Voisin, P. De Medina, A. Mallinger et al., “Identification of a tumor-promoter cholesterol metabolite in human breast cancers acting through the glucocorticoid receptor,” Pro- ceedings of the National Academy of Sciences, vol. 114, no. 44, pp. E9346–E9355, 2017.

Journal

Journal of Oncology – Hindawi Publishing Corporation

Published: May 5, 2020

There are no references for this article.