Impact of body composition parameters on clinical outcomes in patients with metastatic castrate-resistant prostate cancer treated with docetaxel

Clinical Nutrition ESPEN 13 (2016) e39ee45

Contents lists available at ScienceDirect

Clinical Nutrition ESPEN journal homepage: http://www.clinicalnutritionespen.com

Original article

Impact of body composition parameters on clinical outcomes in patients with metastatic castrate-resistant prostate cancer treated with docetaxel Samantha J. Cushen a, *, Derek G. Power b, Kevin P. Murphy c, Ray McDermott d, Brendan T. Griffin e, Marvin Lim d, Louise Daly a, Peter MacEneaney f, Kathleen O' Sullivan g, Carla M. Prado h, Aoife M. Ryan a a

School of Food & Nutritional Sciences, University College Cork, Cork, Ireland Department of Medical Oncology, Mercy & Cork University Hospitals, Cork, Ireland c Department of Radiology, Cork University Hospital, Cork, Ireland d Department of Medical Oncology, St. Vincents University Hospital, Dublin, Ireland e School of Pharmacy, University College Cork, Ireland f Department of Radiology, Mercy University Hospital, Cork, Ireland g School of Mathematical Sciences, University College Cork, Ireland h Alberta Institute for Human Nutrition, Department of Agricultural, Food and Nutritional Science, University of Alberta, Edmonton, Canada b

a r t i c l e i n f o

s u m m a r y

Article history: Received 15 May 2015 Accepted 1 April 2016

Background: Body composition may influence clinical outcomes of certain chemotherapeutic agents. We examined the prognostic significance of skeletal muscle mass and adipose tissue on docetaxel toxicity and overall survival in patients with metastatic castrate resistant prostate cancer (mCRPC). Methods: A retrospective review of patients medical records with mCRPC, treated with docetaxel was conducted. Body composition parameters (skeletal muscle mass, muscle attenuation [MA], visceral and subcutaneous adipose tissue) were measured at L3 by computed tomography (CT) and defined using previously established cut points. Toxicity profile was assessed after 3 cycles of the drug and graded according to the National Cancer Institute Common Toxicity Criteria (version 4). Overall survival was analysed. Results: Overall 63 patients, mean age 69 years (SD 8.3), were included. Sarcopenia was present in 47% (n ¼ 30) and of these 26.7% (8/30) were sarcopenic obese. Common toxicities (all grades) observed included fatigue (80.9%), pain (46%), and constipation (34.9%). DLT occurred in 22 (34.9%) patients; of these 10 patients (15.8%) experienced dose reductions and 12 patients (19%) experienced dose terminations. Measurements of adiposity were not predictive of DLT, however 59.1% patients who had a combination of both sarcopenia and low MA experienced DLT compared to 29.3% of patients without sarcopenia and low MA (p ¼ 0.021). Skeletal muscle index and MA were significantly lower in patients who experienced neutropenia (grade IeII) (46.5 cm2/m2 vs. 51.2 cm2/m2, p ¼ 0.005) compared to their counterparts (24.6 HU vs. 32.2 HU, p ¼ 0.044). Neither sarcopenia nor sarcopenic obesity was associated with overall survival. In multivariate analysis, BMI 25 kg/m2 (HR: 0.349, CI: 0.156e0.782, p ¼ 0.010) was a significant predictor of longer overall survival and both visceral fat index  median 58.7 cm2/m2 (HR: 2.266 CI: 1.066e4.814, p ¼ 0.033) and anaemia (HR: 2.81, CI: 1.297e6.091, p ¼ 0.009) were significant predictors of shorter overall survival. Conclusions: Sarcopenia and low MA are associated with neutropenia (grade IeII). Furthermore, presence of anaemia, high volume of visceral fat and BMI <25 kg/m2 are associated with reduced survival in patients with castrate resistant prostate cancer being treated with docetaxel chemotherapy. © 2016 European Society for Clinical Nutrition and Metabolism. Published by Elsevier Ltd. All rights reserved.

Keywords: Chemotherapy Prostate cancer Sarcopenia Docetaxel Body composition

* Corresponding author. Tel.: þ353 214902406. E-mail address: [email protected] (S.J. Cushen). http://dx.doi.org/10.1016/j.clnesp.2016.04.001 2405-4577/© 2016 European Society for Clinical Nutrition and Metabolism. Published by Elsevier Ltd. All rights reserved.

e40

S.J. Cushen et al. / Clinical Nutrition ESPEN 13 (2016) e39ee45

1. Introduction Prostate cancer is the most common solid tumour in men in the developed world and accounts for 27% of male cancers [17]. Ten percent of men diagnosed with prostate cancer develop metastatic disease, and the 5-year survival for these patients is 30% [33]. Androgen deprivation therapy (ADT), i.e. gonadotropin-releasing hormone (GNRH) analogues, is standard of care in locally advanced and metastatic disease [32,20]. ADT can reduce serum prostate-specific antigen (PSA) and slow down the progression of cancer by its direct action of reducing the production of testosterone, the hormonal driver of prostate cancer. Undesirable side effects of ADT include substantial alterations in body composition [1] with significant gains in body fat and losses in skeletal muscle mass [40]. Low skeletal muscle mass (sarcopenia) can be significant, especially in men over 70 years of age [35], and the quality of muscle (muscle attenuation) can also be adversely affected by ADT [8]. Sarcopenia has emerged as a prevalent body composition phenotype in many cancers and is important because it is predictive of reduced functional ability, shorter time to tumour progression, shorter survival, and higher incidence of dose limiting toxicity (DLT) to many cytotoxic chemotherapy drugs [29,38,10]. Despite initial favourable oncologic responses to ADT, predictable and irreversible resistance to ADT will occur in the vast majority of patients within a median of two years, signifying castrate resistant prostate cancer (CRPC), i.e. progression of disease despite castrate levels of testosterone [31,13]. Chemotherapy is often prescribed to those men with CRPC who have increasing symptoms of progressing disease, e.g. bone pain and fatigue. Toxicity however is of concern, especially as most patients are elderly and may have preexisting age- and ADT-related sarcopenia, as well as medical comorbidities. Docetaxel (TAXOTERE®) has been a standard first line therapy for metastatic CRPC (mCRPC) and exhibits wide inter-patient variability in pharmacokinetics, given it is a narrow therapeutic index drug [12]. Variability in clearance has been correlated with toxicity and treatment efficacy in patients treated with docetaxel [30,34]. Body composition influences the pharmacokinetics of certain cytotoxic agents, as hydrophilic drugs will distribute into the lean compartment whereas lipophilic drugs will distribute into the fat compartment. Thus, changes in adipose tissue and skeletal muscle tissue could lead to increased incidence and severity of chemotherapy toxicities. There is growing literature to suggest that skeletal muscle mass may be a better basis for normalising drug dosages in cancer patients, especially of hydrophilic drugs [22,29]. Likewise, increased adipose tissue may increase volume of distribution for highly lipophilic drugs prolonging their elimination halflives and a recent publication by Prado et al. [25] identified that both muscle and adipose tissue may play a role in predicting toxicity for hydrophobic agents. Docetaxel is highly lipophilic, binds to plasma proteins (>90%; albumin, lipoproteins, and a1-acid glycoprotein) and is primarily eliminated from the body via CYP3A4 mediated hepatic metabolism (70e80%) [9]. The lipophilic profile of docetaxel would suggest a larger volume of distribution in adipose tissue however animal models suggest a greater distribution of docetaxel in muscle tissue compared with adipose tissue [36]. In addition to being linked to chemotherapy toxicity, body composition parameters have been linked to prognosis and overall survival in cancer patients. Sarcopenia has been shown to be independently prognostic of reduced survival in cancers of the biliary tract [21], lung [43] and colon [43]. The results concerning adipose tissue are conflicting, especially in prostate cancer. Higher mortality has been observed in prostate cancer patients with a high

body mass index (BMI) [6]. Conversely, some studies have reported that a high BMI [16] and high volume of subcutaneous adipose tissue [3] are prognostically favourable in mCRPC. For the current study we wanted to examine whether body composition was associated with toxicity secondary to docetaxel treatment. Secondary endpoints included the prognostic significance of body mass parameters and other clinical parameters on overall survival. 2. Patients and methods 2.1. Study population and toxicity evaluation Ethical approval was granted by the Clinical Research Ethics Committee of the Cork Teaching Hospitals, according to good clinical practice and applicable laws. We performed a retrospective analysis of patients with mCRPC treated with first line docetaxel-based systemic chemotherapy in 3 Irish University teaching hospitals over a 6 year period (from January 2008 to December 2013). All patients had received ADT prior to commencing chemotherapy for mCRPC. Patients received 75 mg/m2 of docetaxel every 3 weeks or 50 mg/m2 every 2 weeks (during the analysis period a randomized study reported improved toxicity profiles using the 50 mg/m2 schedule [19]). All patients had a histologic diagnosis of prostate adenocarcinoma and a baseline CT scan within 6 weeks of commencing their 1st cycle of chemotherapy. In addition, the following data was recorded: comorbidities, Gleason scores, PSA levels, Eastern Cooperative Oncology Group (ECOG) performance status, C-reactive protein (CRP) and albumin levels. Two age groups were defined for this analysis (<75 years and 75 years), a range that has been used previously in mCRPC age subgroup analyses [18]. Toxicity profiles were obtained for all cycles of docetaxel however tolerance and toxicity were assessed over the first 3 treatment cycles as we had incomplete toxicity data >3 cycles of docetaxel for all patients. Adverse events were classified according to the common terminology criteria for adverse events (CTCAE) version 4.0. For further analyses, toxicity was divided into grade IeII and grade IIIeIV. Dose limiting toxicity (DLT) was defined as any grade III or higher toxicity leading to a dose reduction, temporary or permanent discontinuation of treatment. Reported toxicities were fatigue, pain, neutropenia (neutrophils < 1.5  109/L), anaemia (Hb < 10 g/dL), neurosensory, constipation, diarrhoea, dyspnoea and nausea and vomiting. 2.2. Body composition assessment Body mass index was calculated by dividing the patients weight in kilograms by height (in meters) squared. Patients with a BMI 18.5e24.9 kg/m2 were defined as normal and patients with a BMI 25 kg/m2 were defined as overweight or obese (30 kg/m2) as per WHO criteria [42]. Cross-sectional area of muscle and adipose tissue was averaged from two consecutive axial images within the same series at the third lumbar vertebra (L3), using OsiriX software version 5.0 (Pixmeo, Geneva, Switzerland). Different tissue compartments were manually outlined and segmentation of the tissue of interest was based on Hounsfield Unit (HU) thresholds (from 29 to þ150 for skeletal muscle, 190 to 30 for subcutaneous adipose tissue and 150 to 50 for visceral adipose tissue). Hand adjustment of the selected areas was performed if necessary and the total cross sectional area of the segmented tissue area was calculated automatically (Fig. 1). Muscle area and total adipose tissue area, were normalized for stature in metres squared (m2) and reported as skeletal muscle index (SMI; cm2/m2), adipose tissue index (ATI; cm2/m2) respectively [23,28]. Subcutaneous fat (SAT)

S.J. Cushen et al. / Clinical Nutrition ESPEN 13 (2016) e39ee45

e41

years) and the most common site of metastases was bone (95.2%). Twenty four percent (n ¼ 15) of the patients had a normal BMI (18.5e24.9 kg/m2), 40% (n ¼ 25) were overweight (BMI  25e29.9 kg/m2) and 36% (n ¼ 23) were obese (BMI  30 kg/ m2). None of our patients were underweight (BMI < 18.5 kg/m2). Of the 63 evaluable patients, 30 (47.6%) were sarcopenic, 18 (28.5%) were both overweight and sarcopenic and 8 (12.7%) were sarcopenic obese. In accordance with this, the most common co morbidities were dyslipidemia followed by hypertension and cardiovascular disease. Skeletal muscle index was significantly higher in patients who were overweight or obese (BMI  25 kg/m2) (52.6 cm2/m2 vs. 45.8 cm2/m2, p ¼ 0.006) compared to patients with a normal BMI. Fig. 1. An axial CT at the level L3 depicting skeletal muscle (in red).

3.2. Body composition and dose limiting toxicity area and visceral fat (VAT) area were normalised for stature to derive SAT index (cm2/m2) and VAT index (cm2/m2) respectively. Due to lack of standardized threshold values, the median was used as cut points for SAT index (<55.3 cm2/m2 versus 55.3 cm2/m2) and VAT index (<58.7 cm2/m2 versus 58.7 cm2/m2). Patients were considered sarcopenic if they had a lumbar SMI 53 cm2/m2 for men with BMI 25 kg/m2 and 43 cm2/m2 for men with a BMI <25 kg/m2 [43]. We also evaluated mean muscle attenuation (MA), as expressed as the mean HU, which indirectly measures fat infiltration in muscle and has been extensively studied as a correlate of muscle radiodensity [15]. We reported mean muscle attenuation for the entire muscle area at the L3. The Martin et al., 2013 [43] cut points were used to defined low MA (<41 HU for men with BMI < 25 kg/m2 and <33 HU for men with BMI  25 kg/m2) [43]. 2.3. Statistics Data are expressed as means and standard deviations (SD) where appropriate. Assumptions were verified and normality was assessed. Outliers were identified and analysed with appropriate statistical adjustments. Survival analyses were done using the KaplaneMeier and Cox regression methods. Log rank tests were used to compare survival curves. Univariate and multivariate analyses for overall survival were conducted using the cox proportional hazards model; hazards ratios (HR) and corresponding 95% confidence intervals (CI) were obtained. The proportional hazards assumption was assessed using a global test based on the Schoenfeld residuals. Any variable that attained significance of <0.300 was entered using a forward stepwise strategy into multivariate analysis. Median values were used to dichotomise each parameter. The following variables were included in the univariable analysis; age, BMI, bone metastases and non-osseous metastases, anaemia, muscle density, sarcopenia, VAT index and SAT index. Overall survival was measured from the date of the first CT scan to date of death or until March 2014, at which time they were censored at the last date they were documented to have been alive. Statistical analysis was completed using SPSS (SPSS for Windows, version 20.0; IBM, Chicago, Illinois, USA).

The starting dose was 75 mg/m2 in the majority of patients (60.3%). DLT occurred in 22 (35%) patients; 10 patients (15.8%) experienced dose reductions and 12 patients (19%) experienced dose terminations. Most of the toxicities were concentrated in cycle 2, with 17.5% (n ¼ 11) of patients experiencing a DLT, compared to 16% (n ¼ 10) of patients experiencing a DLT in both cycle 1 and cycle 3. Table 2 summarises the toxicities experienced during the first 3 cycles of treatment with docetaxel. The mean number of toxicities experienced by patients was 5 (SD 3.6). Among digestive toxicities, nausea and vomiting were recorded in 20.6% of cases and diarrhoea in 28.5%. Among haematological toxicities, neutropenia or anaemia occurred in 9.5% and 25.3% of cases respectively (see Table 2 for grades of toxicity). Hypoalbuminemia (<35 g/L) was associated with a greater incidence of anaemia (72.7% vs. 23.7%, p ¼ 0.001). There appeared to be no effect of sarcopenia on DLT: 15/30 (50%) sarcopenic patients developed DLT compared to 10/33 (30%) nonsarcopenic patients (p ¼ 0.110). There was no significant difference in the number of patients with sarcopenia at baseline amongst the two dosing regimens (i.e. 75 mg vs. 50 mg). Low muscle attenuation and sarcopenia were not significantly associated with DLT when analysed separately. However, 59.1% (n ¼ 13) patients who had both sarcopenia and low MA experienced DLT compared to 29.3% (n ¼ 12) of patients without sarcopenia and low MA (p ¼ 0.021). Skeletal muscle index and MA were significantly lower in patients who experienced neutropenia (grade IeII) (46.5 cm2/m2 vs. 51.2 cm2/m2, p ¼ 0.005; 24.6 HU vs. 32.2 HU, p ¼ 0.044). Given that docetaxel is a lipophilic drug, the proportions of adipose tissue were expected to influence the volume of distribution and hence contribute to occurrence of DLT. Body surface area and measurements of adiposity (ATI, VAT index and SAT index) were not predictive of DLT in the entire cohort. Subgroup analyses were conducted in each BMI category however no statistically significant differences were noted regarding the occurrence of DLT (p ¼ 0.827) and sarcopenic obesity was not predicative of DLT (p ¼ 0.511). When individual toxicities were considered, patients suffering from fatigue (all grades) had significantly higher VAT index to those who didn't suffer from fatigue (234.8 cm2/m2 vs. 171.4 cm2/m2 p ¼ 0.044, respectively).

3. Results

3.3. Body composition parameters and survival

3.1. Patient characteristics

At the time of censoring, 37 patients (58.7%) had died. Overall median survival, was 17.3 months (range 14.3e20.4 months). There was no difference between the median survival among sarcopenic patients compared to non-sarcopenic patients (17.8 months vs. 15 months, p ¼ 0.361). There was a trend towards significance in overall survival in those who experienced DLT compared with those that did not (15 months vs. 25 months, p ¼ 0.066). Patients with BMI

Of 82 consecutive patients, 19 were excluded because they lacked an evaluable CT image performed within 6 weeks of chemotherapy. Body composition and demographic characteristics according to sarcopenia status of the 63 included patients are reported in Table 1. The mean age was 69 years (range 46 yearse81

e42

S.J. Cushen et al. / Clinical Nutrition ESPEN 13 (2016) e39ee45

Table 1 Comparison of baseline demographic and anthropometric characteristics between sarcopenic and non-sarcopenic patients treated with docetaxel using Martin et al. (2013) cut points for males. Variable

All

Patients, n (%) Age (years), mean (SD) Height (m) Weight (kg) BMI (kg/m2) Metastatic site, n (%) Lung Liver Bone PSA at diagnosis (ng/mL) Gleason score at diagnosis, mean (SD) ECOG, n (%) 0e1 2e3 Albumin (g/L), mean (SD) CRP (mg/L), mean (SD) Treatment pathway, n (%) Chemotherapy only Chemotherapy & surgery Chemotherapy & radiotherapy Chemotherapy, radiotherapy & surgery Skeletal muscle index (cm2/m2) Total adipose tissue index (cm2/m2) Visceral adipose tissue index (cm2/m2) Subcutaneous adipose tissue index (cm2/m2) Muscle attenuation (HU)

Non-sarcopenic 33 (53) 65.8(8.6) 1.7 (0.53) 86.8 (14) 27.8 (3.9)

Sarcopenic 30 (47.6) 72.7(6.5) 1.7 (0.54) 79.8 (14.3) 27.8 (4.6)

P

63 69 1.71 82.5 27.8

(100) (8.36) (0.56) (14.5) (4.27)

4 4 60 401.8 8

(6.3) (6.3) (95.2) (967.0) (1.1)

1 1 31 436.9 8.1

(3) (3) (93.9) (918) (1)

59 4 37.8 83.9

(93.6) (6.4) (5.5) (83.9)

30 2 37.8 107

(90.9) (4.1) (5.4) (94)

29 1 37.8 54

(94.8) (7.6) (5.7) (60.2)

0.991 0.115

30 5 20 8 50.7 147.9 75.3 72.5 31.6

(47.6) (7.9) (31.7) (12.6) (8.8) (57.7) (32.9) (35.6) (8.8)

18 2 10 2 56 148.1 56.4 50.7 34.3

(54.5) (6.1) (30.3) (6.1) (8.1) (63.8) (24.8) (29.5) (7.4)

12 3 10 6 44.9 147.9 54.8 56.6 28.4

(40) (10) (33.3) (20) (5) (51.4) (24.8) (22.4) (9.3)

0.255 0.571 0.800 0.197 <0.001 0.988 0.805 0.377 0.007

3 (10) 3 (10) 29 (96.7) 363(1032) 7.9 (1.3)

<0.001 0.268 0.397 0.965 0.278 0.278 0.618 0.766 0.456 0.627

BMI, Body Mass Index; CRP, C-reactive protein; ECOG, Eastern Cooperative Oncology Group criteria Performance Status; HU, Hounsfield Units; PSA, Prostate Specific Antigen; SD, Standard Deviation. Results were considered significant at the P < 0.05 level.

Table 2 Frequencies of toxicities experienced by patients. Toxicity, n (%)

N ¼ 63

Grade IeII

Grade IIIeIV

Fatigue Pain Constipation Neurosensory Diarrhoea Anaemia Dyspnoea Nausea and vomiting Neutropenia

51 29 22 19 18 16 5 13 6

48 27 21 19 17 16 4 13 5

3 (4.7) 2 (3.1) 1 (1.5) e 1 (1.5) e 1 (1.5) e 1 (1.5)

(80.9) (46.0) (34.9) (30.1) (28.5) (25.3) (7.9) (20.6) (9.5)

(76.1) (42.8) (33.3) (30.1) (26.9) (25.3) (6.3) (20.6) (7.9)

25 kg/m2 survived for a median of 19 months versus 13.7 months in those with a BMI <25 kg/m2 (p ¼ 0.038) Fig. 2. Using median cut off values for other body composition parameters we observed no significant differences in overall survival for ATI, VAT index or SAT index. When stratified by BMI categories, low MA was associated with shorter survival (10.7 months; 95% CI: 3.9e17.5) as compared to the high MA group (19.2 months; 95% CI: 10.6e27.9) in normal BMI patients (p ¼ 0.036) but not in patients with BMI 25 kg/m2 (p ¼ 0.209). There was no significant difference in overall survival in patients with serum albumin levels below 35 g/L compared to those with serum albumin levels above 35 g/L (13.9 months [IQ range 8e19.7] vs. 19.2 months [IQ range 9.6e28.8]; p ¼ 0.160). The results of the multivariate Cox analysis (Table 3) including age (75 Vs. <75 years), bone metastases and non-osseous metastases, low MA, sarcopenia, SAT index (<55.3 cm2/m2 versus 55.3 cm2/m2), VAT index (<58.7 cm2/m2 versus 58.7 cm2/m2), anaemia, BMI (25 kg/m2 Vs. <25 kg/m2), showed that VAT index (HR: 2.266, CI: 1.066e4.814, p ¼ 0.003), anaemia (HR: 2.811, CI: 1.297e6.091, p ¼ 0.009) and BMI 25 kg/m2 (HR: 0.349, CI: 0.156e0.782, p ¼ 0.010) were independently associated with overall survival.

Fig. 2. KaplaneMeier curve of patients with BMI <25 and patients with BMI 25.

4. Discussion This retrospective analysis explored the relationship between body composition variables and clinical outcomes (toxicity and survival) in patients treated with docetaxel for mCRPC. Although sarcopenia was not found to be statistically associated with DLT, SMI and MA were significantly lower in patients who experienced

S.J. Cushen et al. / Clinical Nutrition ESPEN 13 (2016) e39ee45

e43

Table 3 Univariate and multivariate analysesa for predictors of overall survival. Variable

No. of patients

Age <75 45 75 18 Bone mets þ other mets No 10 Yes 53 Anaemiac No 47 Yes 16 2 BMI (kg/m ) <25 17 25 46 Low MA No 9 Yes 54 Sarcopenic No 33 Yes 30 2 2 SAT index (cm /m )
Survival (months)b

Univariate

Median

95% CI

HR

95% CI

P

27 10

17.3 17.8

14.6e20 0e43.6

1.596

0.679e3.748

0.283

8 29

16 17.8

12.9e19.1 10.2e25.4

1.260

0.504e3.148

0.621

25 12

23.5 12.3

6.4e40.5 6.2e18.5

3.666

1.617e8.311

13 24

13.7 19.2

2.5e24.9 2.7e35.7

0.492

5 32

15.1 17.7

11e19.2 14.9e20.4

23 14

15.1 25.2

22 15 18 19

No. of deaths

Multivariate HR

95% CI

P

0.002

2.811

1.297e6.091

0.009

0.197e1.228

0.129

0.349

0.156e0.782

0.010

0.393

0.131e1.186

0.097

10.5e19.8 20.8e49.8

0.508

0.224e1.149

0.104

16.8 32

12.5e21.1 4.7e59.3

0.445

0.191e1.038

0.061

17.8 16.9

13.8e21.8 13.7e20.2

2.776

1.178e6.540

0.020

2.266

1.066e4.814

0.033

CI, confidence interval; MA, muscle attenuation; Mets, metastasis; HR, hazard ratio; SE, standard error; VAT, visceral fat; SAT, subcutaneous fat. Results were considered significant at the P < 0.05 level. a Regression coefficient, HR's and P values calculated with Cox proportional hazards model. b Median survival calculated with KaplaneMeier method. c Haemoglobin <13 g/dL, data available on 54 patients only.

neutropenia (grade IeII). Furthermore, the presence of anaemia and high VAT index were associated with a shorter median survival and a heavy body weight (BMI  25 kg/m2) was associated with a longer median survival. Pre-existing ADT-related sarcopenia may compound wasting of skeletal muscle that can occur during the cancer trajectory [39] as well as the loss of kilograms of skeletal muscle that can occur with even a short course of chemotherapy [4], rendering a patient sarcopenic. We showed no association between sarcopenia and DLT, however, when toxicities were considered individually, patients with low SMI experienced significantly more neutropenia compared to non-sarcopenic patients. Previous studies have described contradictory results. Mir et al. [21] reported DLT occurred in 45% of their cohort and sarcopenia was a predictor of DLT in patients with hepatocellular carcinoma treated with sorafenib. Conversely, a study by Parsons et al. [24] reported no significant differences in grade III and IV toxicity between sarcopenic and non-sarcopenic patients with advanced cancer and liver metastases receiving hepatic arterial infusion chemotherapy [24]. The simultaneous loss of skeletal muscle tissue and concurrent gain of adipose tissue is termed sarcopenic obesity and is considered to be a worst case scenario due to the combination of two health related risk factors [26]. In a recently published paper by Anandavadivelan et al. [2] DLT was investigated in 72 patients receiving neo-adjuvant therapy for oesophageal cancer and they showed that sarcopenic obese presented with higher DLT compared to their non-sarcopenic obese counterparts (OR: 5.54, CI: 1.12e27.44). In our study, few sarcopenic obese patients were found (n ¼ 8) and no differences were observed between the occurrence of DLT in sarcopenic obese and non-sarcopenic obese. The prognostic value of sarcopenic obesity is unclear in the present study probably due to the small number of patients and further studies should address this subset of patients in larger cohorts of advanced prostate cancer patients treated with docetaxel.

Docetaxel has a high affinity for adipose tissue and it has been reported that there is a significant increase in the volume of distribution of docetaxel in obese patients compared with non-obese patients as well as an increase in the elimination half-life of the terminal phase [44]. In our study measurements of adiposity (ATI, VAT index and SAT index) were not predictive of DLT. This may reflect that patients with sarcopenia display increased levels of a1acid glycoprotein, resulting in increased binding of docetaxel, which leads to reduced hepatic clearance [5]. Prado et al. [25] recently reported the importance of both fat mass and muscle mass in explaining toxicity in a cohort (n ¼ 74) of patients with ovarian cancer treated with a combination of pegylated liposomal doxorubicin (Doxil) and Trabectedin (Yondelis). Lean body mass was not predictive of toxicity but a lower fat mass/lean body mass ratio was the most powerful variable associated with toxicity [25]. There is increasing evidence that obesity defined by BMI may be misleading for predicting clinical outcomes in oncology as BMI does not accurately distinguish between lean and fat tissues, furthermore it is unable to depict the independent prognostic effect of each individual tissue (e.g. VAT and SAT). The relationship between adipose tissue and clinical outcomes is complex in prostate cancer as some studies have shown a high BMI is associated with worse clinical outcomes [6,7]; whereas others have found the an elevated BMI is predictive of longer survival in patients with mCRPC [16]. In our study, patients considered overweight or obese (BMI  25 kg/ m2) showed a better overall survival than patients with a BMI <25 kg/m2. However, patients with a BMI 25 kg/m2 were more muscular compared to patients with a normal BMI (p ¼ 0.006) and this survival advantage is most likely attributed to muscle rather than just larger BMI's. Several theories have been proposed that try to explain the obesity paradox in cancer and one possible explanation is the diagnosis of obesity on the basis of BMI not adiposity [27]. Gonzalez et al. [14] recently reported that the obesity paradox is present in cancer patients but only when obesity is defined by

e44

S.J. Cushen et al. / Clinical Nutrition ESPEN 13 (2016) e39ee45

BMI and not by body composition parameters as measured using bioelectrical impedance analysis. Indeed they found that obesity predicted longer survival rates only when sarcopenia was absent, and, patients with sarcopenic obesity had the poorest prognosis [14]. In our study, high VAT index was significantly associated with shorter survival, indicating that patients with the same BMI may have different quantities of visceral and subcutaneous fat, and VAT index is a body composition parameter that may provide additional information that is not conveyed by BMI. Similar to our results, Wu et al. [41] reported in 333 patients with mCRPC that the presence of high BMI appears to be associated with longer survival whereas the presence of visceral obesity is associated with poorer survival. Likewise, Antoun et al. [3] showed that while neither sarcopenia nor sarcopenic obesity was associated with survival, a high amount of subcutaneous fat is associated with longer overall survival. We noted that VAT and SAT did not behave in a similar manner and we observed a trend towards better overall survival for patients with a high SAT index however the difference was not statistically significant. This discrepancy between BMI, SAT index and VAT index has not yet been thoroughly explored in prostate cancer and requires further evaluation. Our study was limited by its retrospective design and the sample size limited the statistical power for clinically meaningful subgroup analyses. Moreover, a bigger sample size might have enabled us to identify some factors as statistically significant that were found to be borderline significant in the current study. In addition, we were unable to specifically quantify the duration of ADT, prior to commencement of chemotherapy, as a lot of this data had not been recorded in the medical records. Despite the limitations, this study provides an insight into chemotherapeutic treatment of elderly men with mCRPC. This is especially relevant given recent data that has shown the combination of ADT and docetaxel in hormone sensitive metastatic prostate cancer is more effective compared with ADT alone [37]. Therefore it is very likely that men with a high metastatic burden of disease at presentation will receive upfront docetaxel as well as ADT. Prostate cancer patients are surviving longer because of earlier diagnosis and treatment, so predictors of toxicity to chemotherapy are especially relevant.

5. Conclusion In conclusion, we report that sarcopenia and low MA in patients with mCRPC increases the risk of neutropenia (grade IeII) during docetaxel treatment. Furthermore, our study shows that high VAT is associated with poorer survival regardless of BMI, and use of BMI in prostate cancer may be obsolete as it does not accurately describe the prognostic significance of each individual tissue. We speculate that pre-existing changes in body composition, which may be driven by comorbid conditions such as inflammation, are prominent in advanced cancer populations such as ours, and can predict survival and chemotherapy toxicity. Importantly, these observations must be regarded as exploratory and hypothesis generating but should prompt further in-depth evaluation and validation in a prospective study design.

Funding No funding was obtained for this study.

Conflicts of interest None.

Acknowledgements We would like to acknowledge the support of the Health Research Board Clinical Research Facility, Cork (CRF-C).

References [1] Allan CA, Collins VR, Frydenberg M, McLachlan RI, Matthiesson KL. Monitoring cardiovascular health in men with prostate cancer treated with androgen deprivation therapy. Int J Urol Nurs 2012;6:35e41. [2] Anandavadivelan P, Brismar T, Nilsson M, Johar A, Martin L. Sarcopenic obesity: a probable risk factor for dose limiting toxicity during neo-adjuvant chemotherapy in oesophageal cancer patients. Clin Nutr 2015. http:// dx.doi.org/10.1016/j.clnu.2015.05.011 [Epub ahead of print].
s, Fizazi Karim, di [3] Antoun Sami, Bayar Amine, Ileana Ekatarina, Laplanche Agne Palma Mario, et al. High subcutaneous adipose tissue predicts the prognosis in metastatic castration-resistant prostate cancer patients in post chemotherapy setting. Eur J Cancer 2015. http://dx.doi.org/10.1016/j.ejca.2015.07.042 [Epub ahead of print]. [4] Awad S, Tan BH, Cui H, Bhalla A, Fearon KC, Parsons SL, et al. Marked changes in body composition following neoadjuvant chemotherapy for oesophagogastric cancer. Clin Nutr 2012;31:74e7. [5] Bruno R, Vivier N, Vergniol JC, De Phillips SL, Montay G, Sheiner LB. A population pharmacokinetic model for docetaxel (taxotere®): model building and validation. J Pharmacokinet Biopharm 1996;24:153e72. [6] Calle EE, Rodriguez C, Walker-Thurmond K, Thun MJ. Overweight, obesity, and mortality from cancer in a prospectively studied cohort of US adults. N. Engl J Med 2003;348(17):1625e38. [7] Chalfin HJ, Lee SB, Jeong BC, Freedland SJ, Alai H, Feng Z, et al. Obesity and longterm survival after radical prostatectomy. J Urol 2014;192(4):1100e4. [8] Chang D, Joseph DJ, Ebert MA, Galv~ ao DA, Taaffe DR, Denham JW, et al. Effect of androgen deprivation therapy on muscle attenuation in men with prostate cancer. J Med Imaging Radiat Oncol 2014;58:223e8. [9] Clarke SJ, Rivory LP. Clinical pharmacokinetics of docetaxel. Clin Pharmacokinet 1999;36:99e114. [10] Cushen SJ, Power D, Teo MY, Maceneaney P, Maher MM, McDermott R, et al. Body composition by computed tomography as a predictor of toxicity in patients with renal cell carcinoma treated with sunitinib. Am J Clin Oncol 2014. http://dx.doi.org/10.1097/COC.0000000000000061. [12] Franke RM, Carducci MA, Rudek MA, Baker SD, Sparreboom A. Castrationdependent pharmacokinetics of docetaxel in patients with prostate cancer. J Clin Oncol 2010;28:4562e7. [13] Galsky M, Vogelzang N. Docetaxel-based combination therapy for castrationresistant prostate cancer. Ann Oncol 2010;21:2135e44. [14] Gonzalez MC, Pastore CA, Orlandi SP, Heymsfield SB. Obesity paradox in cancer: new insights provided by body composition. Am J Clin Nutr 2014;99(5):999e1005. [15] Goodpaster BH, Kelley DE, Thaete FL, He J, Ross R. Skeletal muscle attenuation determined by computed tomography is associated with skeletal muscle lipid content. J Appl Physiol 2000;89:104e10. [16] Halabi S, Ou SS, Vogelzang NJ, Small EJ. Inverse correlation between body mass index and clinical outcomes in men with advanced castrationerecurrent prostate cancer. Cancer 2007;110:1478e84. [17] Heidenreich A, Bellmunt J, Bolla M, Joniau S, Mason M, Matveev V, et al. EAU guidelines on prostate cancer. Part 1: screening, diagnosis, and treatment of clinically localised disease. Eur Urol 2011;59:61e71. [18] Italiano A, Ortholan C, Oudard S, Pouessel D, Gravis G, Beuzeboc P, et al. Docetaxel-based chemotherapy in elderly patients (age 75 and older) with castration-resistant prostate cancer. Eur Urol 2009;55(6):1368e76. [19] Kellokumpu-Lehtinen PL, Harmenberg U, Joensuu T, McDermott R, Hervonen P, Ginman C, et al. 2-Weekly versus 3-weekly docetaxel to treat castration-resistant advanced prostate cancer: a randomised, phase 3 trial. Lancet Oncol 2013;14(2):117e24. [20] Meng MV, Grossfeld GD, Sadetsky N, Mehta SS, Lubeck DP, Carroll PR. Contemporary patterns of androgen deprivation therapy use for newly diagnosed prostate cancer. Urology 2002;60:7e11. [21] Mir O, Coriat R, Blanchet B, Durand J-P, Boudou-Rouquette P, Michels J, et al. Sarcopenia predicts early dose-limiting toxicities and pharmacokinetics of sorafenib in patients with hepatocellular carcinoma. PLoS One 2012;7: e37563. http://dx.doi.org/10.1371/journal.pone.0037563. [22] Morgan DJ, Bray KM. Lean body mass as a predictor of drug dosage. Implications for drug therapy. Clin Pharmacokinet 1994;26:292e307. [23] Mourtzakis M, Prado CM, Lieffers JR, Reiman T, McCargar LJ, Baracos VE. A practical and precise approach to quantification of body composition in cancer patients using computed tomography images acquired during routine care. Appl Physiol Nutr Metab 2008;33:997e1006. [24] Parsons HA, Tsimberidou AM, Pontikos M, Fu S, Hong D, Wen S, et al. Evaluation of the clinical relevance of body composition parameters in patients with cancer metastatic to the liver treated with hepatic arterial infusion chemotherapy. Nutr Cancer 2012;64:206e17. [25] Prado CM, Baracos VE, Xiao J, Birdsell L, Stuyckens K, Park YC, et al. The association between body composition and toxicities from the combination of

S.J. Cushen et al. / Clinical Nutrition ESPEN 13 (2016) e39ee45

[26]

[27] [28]

[29]

[30]

[31] [32]

[33] [34]

[35]

Doxil and trabectedin in patients with advanced relapsed ovarian cancer. Appl Physiol Nutr Metab 2014;39:693e8. Prado CMM, Wells JCK, Smith SR, Stephan BCM, Siervo M. Sarcopenic obesity: a critical appraisal of the current evidence. Clin Nutr 2012;31(5): 583e601. Prado CM, Gonzalez MC, Heymsfield SB. Body composition phenotypes and obesity paradox. Curr Opin Clin Nutr Metab Care 2015;18(6):535e51. Prado CM, McCargar L, Reiman T, Sawyer M, Martin L, Baracos VE. Prevalence and clinical implications of sarcopenic obesity in patients with solid tumours of the respiratory and gastrointestinal tracts: a population-based study. Lancet Oncol 2008;9:629e35. Prado CMM, Maia YLM, Ormsbee M, Sawyer MB, Baracos VE. Assessment of nutritional status in cancer ? The relationship between body composition and pharmacokinetics. Anticancer Agents Med Chem 2013;12:1197e203. Rudek M, Sparreboom A, Garrett-Mayer E, Armstrong D, Wolff A, Verweij J, et al. Factors affecting pharmacokinetic variability following doxorubicin and docetaxel-based therapy. Eur J Cancer 2004;40:1170e8. Seruga B, Tannock IF. Chemotherapy-based treatment for castration-resistant prostate cancer. J Clin Oncol 2011;29:3686e94. Shahinian VB, Kuo YF, Freeman JL, Orihuela E, Goodwin JS. Increasing use of gonadotropin-releasing hormone agonists for the treatment of localized prostate carcinoma. Cancer 2005;103:1615e24. Siegel R, Naishadham D, Jemal A. Cancer statistics, 2012. CA Cancer J Clin 2012;62:10e29. Slaviero KA, Clarke SJ, McLachlan AJ, Blair EY, Rivory LP. Population pharmacokinetics of weekly docetaxel in patients with advanced cancer. Br J Clin Pharmacol 2004;57:44e53. Smith MR, Saad F, Egerdie B, Sieber PR, Tammela TL, Ke C, et al. Sarcopenia during androgen-deprivation therapy for prostate cancer. J Clin Oncol 2012;30:3271e6.

e45

[36] Strychor S, Eiseman J, Parise R, Joseph E, Egorin M, Zamboni W. Fat and skeletal muscle disposition of docetaxel in SCID mice. Cancer Res 2007;67(9 Suppl.):1553. [37] Sweeney C, Sweeney CJ, Chen YH, Carducci M, Liu G, Jarrard DF, et al. Chemohormonal therapy in metastatic hormone-sensitive prostate cancer. N Engl J Med 2015;373(8):737e46. [38] Tan BHL, Brammer K, Randhawa N, Welch NT, Parsons SL, James EJ, et al. Sarcopenia is associated with toxicity in patients undergoing neo-adjuvant chemotherapy for oesophago-gastric cancer. Eur J Surg Oncol 2015;41: 333e8. [39] Tisdale MJ. Loss of skeletal muscle in cancer: biochemical mechanisms. Front Biosci J Virtual Libr 2001;6:D164e74. [40] Van Londen GJ, Levy ME, Perera S, Nelson JB, Greenspan SL. Body composition changes during androgen deprivation therapy for prostate cancer: a 2-year prospective study. Crit Rev Oncol Hematol 2008;68:172e7. [41] Wu W, Liu X, Chaftari P, Cruz Carreras MT, Gonzalez C, Viets-Upchurch J, et al. Association of body composition with outcome of docetaxel chemotherapy in metastatic prostate cancer: a retrospective review. PLoS One 2015;10. http:// dx.doi.org/10.1371/journal.pone.012204. [42] World Health Organization. Obesity: preventing and managing the global epidemic. Report of a WHO consultation. World Health Organ tech rep ser, vol. 894 (iexii); 2000. p. 1e253. [43] Martin Lisa, Birdsell Laura, MacDonald Neil, Reiman Tony, Clandinin M Thomas, McCargar Linda J, et al. Cancer cachexia in the age of obesity: skeletal muscle depletion is a powerful prognostic factor, independent of body mass index. J Clin Oncol. 2013 JCO-2012. [44] Sparreboom Alex, Wolff Antonio C, Ron H.J. Mathijssen, Chatelut Etienne, Rowinsky Eric K, Verweij Jaap, et al. Evaluation of alternate size descriptors for dose calculation of anticancer drugs in the obese. J Clin Oncol. 2007;25(30): 4707e13.