Available online at www.sciencedirect.com
ScienceDirect EJSO 41 (2015) 333e338
www.ejso.com
Sarcopenia is associated with toxicity in patients undergoing neo-adjuvant chemotherapy for oesophagogastric cancer B.H.L. Tan a,*, K. Brammer a, N. Randhawa a, N.T. Welch a, S.L. Parsons a, E.J. James b, J.A. Catton a a b
Department of Surgery, City Hospital Campus, Nottingham University Hospitals, Nottingham NG5 1PB, UK Department of Oncology, City Hospital Campus, Nottingham University Hospitals, Nottingham NG5 1PB, UK Accepted 12 November 2014 Available online 26 November 2014
Abstract Background: Patients with potentially curative oesophago-gastric cancer typically undergo neo-adjuvant chemotherapy prior to surgery. The majority of anti-cancer drugs have a narrow therapeutic index. The aim of this study was to determine if features of body composition, assessed using computed tomography (CT) scans, may be predictive of dose-limiting toxicity (DLT) in patients undergoing neoadjuvant chemotherapy for oesophago-gastric cancer. The influence of sarcopenia and DLT on overall survival was also evaluated. Methods: 89 Patients having potentially curative oesophago-gastric cancer surgery were studied. Patients studied had histologically confirmed oesophago-gastric cancer with no evidence of distant metastasis on pre-operative staging. CT scan was performed in all cases at diagnosis. DLT was defined as toxicity leading to postponement of treatment, a drug dose reduction or definitive interruption of drug administration. Results: DLT occurred in 37 out of 89 patients (41.6%) undergoing chemotherapy. Sarcopenia (odds ratio, 2.95; 95% confidence interval, 1.23e7.09; p ¼ 0.015) was associated with DLT on multivariate analysis. Median overall survival for patients who were sarcopenic was 569 days (IQ range: 357e1230 days) vs. 1013 days (IQ range: 496e1318 days) for patients who were not sarcopenic ( p ¼ 0.04). There was no significant difference in overall survival in patients who experienced DLT compared with those that did not ( p ¼ 0.665). Conclusions: Sarcopenia is a significant predictor of DLT in oesophago-gastric cancer patients undergoing neo-adjuvant chemotherapy. These results raise the potential for use of assessment of skeletal muscle mass using CT scans to predict toxicity and individualize chemotherapy dosing. Ó 2014 Elsevier Ltd. All rights reserved.
Keywords: Chemotherapy toxicity; Oesophago-gastric cancer; Body composition; Sarcopenia
Introduction Oesophago-gastric cancer is the fifth most common cancer in the UK, being responsible for 4.8% of new cancer cases every year.1 Until recently surgery alone has been the treatment of choice for all resectable oesophageal, gastro-oesophageal junction (GOJ) and gastric cancers. However, results from OE02 trial of neoadjuvant chemotherapy vs. surgery alone showed a small but significant survival advantage for neoadjuvant Cisplatin and 5* Corresponding author. Tel.: þ44 (0)1159691169. E-mail address: [email protected] (B.H.L. Tan). http://dx.doi.org/10.1016/j.ejso.2014.11.040 0748-7983/Ó 2014 Elsevier Ltd. All rights reserved.
fluorouracil (5-FU) compared to surgery alone in potentially curable oesophageal cancer (2-year survival 43% vs. 34%, 5-year survival of 23% vs. 17%).2,3 This was followed by the MAGIC trial (The MRC Adjuvant Gastric Infusional Chemotherapy) of perioperative Epirubicin, Cisplatin and 5-FU which was the first randomised trial to demonstrate a conclusive overall survival and progression-free disease survival for gastric, GOJ and lower oesophageal adenocarcinoma compared with surgery alone with a 5-year survival of 36% vs. 23% and a 2-year survival of 50% vs. 42%.4 Currently, patients with locally advanced but potentially curative oesophago-gastric cancer at Nottingham
334
B.H.L. Tan et al. / EJSO 41 (2015) 333e338
University Hospitals receive 3-cycles of Epirubicin, Cisplatin, and Capcitabine (ECX) for adenocarcinoma and 2 cycles of Cisplatin and 5-FU (CF) for squamous cell carcinoma, prior to surgery. The majority of anticancer drugs have a narrow therapeutic index, and it is important to determine factors that explain individual variation in efficacy and toxicity. Serum albumin and neutrophil-lymphocyte ratio (NLR) have been previously shown to be associated with chemotherapy toxicity.5 Body composition, in particular proportions of skeletal muscle (and overall lean tissue) in the body, also appear pertinent as several recent results demonstrate an association between skeletal muscle depletion and excess toxicity during treatment with anti-cancer drugs.6e8 The aim of this study was to determine if serum albumin, NLR and features of body composition may be predictive of dose-limiting toxicity (DLT) in patients undergoing neo-adjuvant chemotherapy for oesophago-gastric cancer. The presence of sarcopenia in patients undergoing surgery for various types of cancer has been shown to adversely affect long term survival.9e11 However, the impact of neo-adjuvant chemotherapy toxicity on longterm outcomes post-surgery has not been elucidated. Another aim of the present study was to assess if the presence of DLT has any influence in the post-operative survival of patients with oesophago-gastric cancer.
Neo-adjuvant chemotherapy regimens Squamous cell carcinoma
Day 1: Day 2: Day 3: Day 4: Cycle frequency: Number of cycles:
Cisplatin 80 mg/m2 intravenous infusion 5-Fluorouracil 1000 mg/m2 intravenous infusion 5-Fluorouracil 1000 mg/m2 intravenous infusion 5-Fluorouracil 1000 mg/m2 intravenous infusion 5-Fluorouracil 1000 mg/m2 intravenous infusion Every 21 days 2
Adenocarcinoma Day 1:
Cycle frequency: Number of cycles:
Epirubicin 50 mg/m2 intravenous bolus Cisplatin 60 mg/m2 intravenous infusion Capecitabine 625 mg/m2 twice daily for 21 days (42 doses) Every 21 days 3
Dose limiting toxicity Dose limiting toxicity (DLT) was defined by intolerable toxicities requiring the postponement of treatment, a drug dose reduction, or definitive interruption of drug administration.
Method Body composition measurements The design of this study was approved by the audit department of Nottingham University Hospitals (approval ID 651). The requirement for informed consent from patients was waived because of its retrospective design.
Study population Patients having potentially curative, locally advanced oesophago-gastric cancer without evidence of metastasis presenting to the Nottingham Upper GI MDT between November 2010 and August 2012 were considered for this study. Patients included had histologically confirmed adenocarcinoma or squamous cell carcinoma. All patients were routinely staged with a combination of computed tomography (CT), endoscopic ultrasound (EUS) and laparoscopy according to the International Union Against Cancer (UICC) system.12 Tumours located around the GOJ were classified according to Siewert and Stein13; type I and II tumours were staged as oesophageal cancers and type III as gastric cancers. Those with locally advanced disease (T2 and greater and/or loco-regional lymphadenopathy) and with sufficient physiological reserve to tolerate neoadjuvant chemotherapy were studied. Pre-treatment serum albumin levels and neutrophil-lymphocyte ratio were also determined.
Anthropometric measurements Weight and height were recorded according to standard methods. Weight was measured with a medical balance beam scale and height was measured with a stadiometer. Body mass index (BMI) was calculated [weight (kg)/height (m2)]. BSA was calculated using the Mosteller formula [BSA (m2) ¼ ([height (cm) weight (kg)]/3600)1/2]. Image analysis CT has proven to be accurate for measuring human body composition.14,15 Regional muscle tissue was measured by CT from electronically stored images, which had been done previously for diagnostic purposes. CT scans was performed in all cases at diagnosis prior to commencement of chemotherapy. The third lumbar vertebra (L3) was chosen as a landmark, and two consecutive slices were assessed to measure cross-sectional area of muscle and adipose tissue as described.16,17 The average value of two images was computed for each patient. Images were analyzed using Slice-O-Matic software V4.3 (Tomovision). Crosssectional area for muscle and adipose tissue was normalized for stature (cm2/m2) and reported. Cutoffs for sarcopenia were based on a CT-based sarcopenic obesity study of cancer patients by Prado et al.16 (i.e.,
B.H.L. Tan et al. / EJSO 41 (2015) 333e338
L3 skeletal muscle index: 38.5 cm2/m2 for women and 52.4 cm2/m2 for men). Total LBM was estimated from muscle cross-sectional areas as described in Mourtzakis et al.18: Total body fatfree mass (FFM) (kg) ¼ 0.3 [skeletal muscle at L3 (cm2)] þ 6.06 (r ¼ 0.94). Statistical analysis Data are presented as mean SD unless otherwise stated. Survival was determined from the date of surgery until death or until the censor date of October 21, 2014. Univariate and multivariate analyses and calculation of odds ratios were done using a logistic regression model. Owing to the large number of covariates examined, only those that were significant on univariate analysis were
335
included in multivariate analysis. Receiver-operator characteristic curves were used to select cutoff values for continuous variables. Values with the best combination of sensitivity and specificity were chosen. A backward stepwise procedure was done to derive a final model of the variables that had a significant relationship with survival. To remove a variable from the model, the corresponding p value had to be >0.05. Comparisons between groups of patients were assessed using independent t-test or ManneWhitney U test. Survival curves were constructed using the KaplaneMeier technique. Log-rank test was used to compare survival between groups of patients. p values < 0.05 were regarded as statistically significant. Statistical analysis was done using SPSS 15.0 statistical package (SPSS Inc.). Results
Table 1 Patient demographics. No. of patients (n ¼ 89) Age (years)a
65.8 8.1
Sex M F
67 (75.3) 22 (24.7)
Dose limiting toxicity Present Absent
37 (41.6) 52 (58.4)
Toxicity prevalence Reduction in dose Postponement of treatment Definitive interruption of treatment
24 (27.0) 6 (6.7) 7 (7.9)
Tumour site Oesophageal Gastro-oesophageal junction Gastric
49 (55) 16 (18) 24 (27)
Histology Adenocarcinoma Squamous cell cancer
72 (80.9) 17 (19.1)
Stage I II III
21 (23.6) 27 (30.3) 41 (46.1)
During the study period, 103 consecutive patients were deemed suitable to undergo neo-adjuvant chemotherapy followed by surgery. 14 patients had disease progression during chemotherapy and had their surgery cancelled, and were hence excluded from the study. Demographic data of the remaining 89 patients are shown in Table 1. 37 patients (41.6%) experienced DLT with seven patients (7.9%) failing to complete neo-adjuvant chemotherapy. 44 patients (44.9%) were sarcopenic prior to commencement of chemotherapy. Estimated lean body mass and BSA were moderately related in this patient population (Fig. 1). Sarcopenic patients were significantly older (age of 68.6 7.0 years (mean S.D.) vs. 63.1 8.3 years; p ¼ 0.001), but were otherwise similar with respect to sex, tumour site, histology, stage of disease, pre-
Pre-chemotherapy serum albumin (g/l)a 34.2 5.1 Pre-chemotherapy neutrophil lymphocyte ratioa 2.68 1.86 BMI (kg/m2)a 25.5 4.7 1.86 0.22 BSA (m2)a Lumbar skeletal muscle index (cm2/m2)a 49.1 9.2 Lumbar adipose tissue index (cm2/m2)a 109.6 63.1 Estimated whole body lean mass (kg)a 49.6 10.3 Sarcopenic 44 (49.4) Values are number of patients with percentages in round parentheses unless indicated otherwise. a Values are mean SD.
Figure 1. Relationship between total lean body mass and body surface area in oesophago-gastric cancer patients prior to neo-adjuvant chemotherapy.
336
B.H.L. Tan et al. / EJSO 41 (2015) 333e338
chemotherapy serum albumin and NLR levels (Table 2). As expected, BMI and BSA were found to be significantly lower in those patients with sarcopenia compared with those who did not have sarcopenia. Lumbar adipose tissue index was also significantly lower in sarcopenic patients (90.4 47.6 vs. 128.4 70.9 cm2/m2; p ¼ 0.004). The prevalence of DLT was more common in patients with sarcopenia compared with those without sarcopenia (54.5% vs. 28.9%; p ¼ 0.015). BMI, BSA and sarcopenia were associated with DLT on univariate analysis. Using receiver-operator characteristic curves, the cutoff value with the best discriminatory value for BSA was >1.89 m2. A BMI cutoff of
ScienceDirect EJSO 41 (2015) 333e338
www.ejso.com
Sarcopenia is associated with toxicity in patients undergoing neo-adjuvant chemotherapy for oesophagogastric cancer B.H.L. Tan a,*, K. Brammer a, N. Randhawa a, N.T. Welch a, S.L. Parsons a, E.J. James b, J.A. Catton a a b
Department of Surgery, City Hospital Campus, Nottingham University Hospitals, Nottingham NG5 1PB, UK Department of Oncology, City Hospital Campus, Nottingham University Hospitals, Nottingham NG5 1PB, UK Accepted 12 November 2014 Available online 26 November 2014
Abstract Background: Patients with potentially curative oesophago-gastric cancer typically undergo neo-adjuvant chemotherapy prior to surgery. The majority of anti-cancer drugs have a narrow therapeutic index. The aim of this study was to determine if features of body composition, assessed using computed tomography (CT) scans, may be predictive of dose-limiting toxicity (DLT) in patients undergoing neoadjuvant chemotherapy for oesophago-gastric cancer. The influence of sarcopenia and DLT on overall survival was also evaluated. Methods: 89 Patients having potentially curative oesophago-gastric cancer surgery were studied. Patients studied had histologically confirmed oesophago-gastric cancer with no evidence of distant metastasis on pre-operative staging. CT scan was performed in all cases at diagnosis. DLT was defined as toxicity leading to postponement of treatment, a drug dose reduction or definitive interruption of drug administration. Results: DLT occurred in 37 out of 89 patients (41.6%) undergoing chemotherapy. Sarcopenia (odds ratio, 2.95; 95% confidence interval, 1.23e7.09; p ¼ 0.015) was associated with DLT on multivariate analysis. Median overall survival for patients who were sarcopenic was 569 days (IQ range: 357e1230 days) vs. 1013 days (IQ range: 496e1318 days) for patients who were not sarcopenic ( p ¼ 0.04). There was no significant difference in overall survival in patients who experienced DLT compared with those that did not ( p ¼ 0.665). Conclusions: Sarcopenia is a significant predictor of DLT in oesophago-gastric cancer patients undergoing neo-adjuvant chemotherapy. These results raise the potential for use of assessment of skeletal muscle mass using CT scans to predict toxicity and individualize chemotherapy dosing. Ó 2014 Elsevier Ltd. All rights reserved.
Keywords: Chemotherapy toxicity; Oesophago-gastric cancer; Body composition; Sarcopenia
Introduction Oesophago-gastric cancer is the fifth most common cancer in the UK, being responsible for 4.8% of new cancer cases every year.1 Until recently surgery alone has been the treatment of choice for all resectable oesophageal, gastro-oesophageal junction (GOJ) and gastric cancers. However, results from OE02 trial of neoadjuvant chemotherapy vs. surgery alone showed a small but significant survival advantage for neoadjuvant Cisplatin and 5* Corresponding author. Tel.: þ44 (0)1159691169. E-mail address: [email protected] (B.H.L. Tan). http://dx.doi.org/10.1016/j.ejso.2014.11.040 0748-7983/Ó 2014 Elsevier Ltd. All rights reserved.
fluorouracil (5-FU) compared to surgery alone in potentially curable oesophageal cancer (2-year survival 43% vs. 34%, 5-year survival of 23% vs. 17%).2,3 This was followed by the MAGIC trial (The MRC Adjuvant Gastric Infusional Chemotherapy) of perioperative Epirubicin, Cisplatin and 5-FU which was the first randomised trial to demonstrate a conclusive overall survival and progression-free disease survival for gastric, GOJ and lower oesophageal adenocarcinoma compared with surgery alone with a 5-year survival of 36% vs. 23% and a 2-year survival of 50% vs. 42%.4 Currently, patients with locally advanced but potentially curative oesophago-gastric cancer at Nottingham
334
B.H.L. Tan et al. / EJSO 41 (2015) 333e338
University Hospitals receive 3-cycles of Epirubicin, Cisplatin, and Capcitabine (ECX) for adenocarcinoma and 2 cycles of Cisplatin and 5-FU (CF) for squamous cell carcinoma, prior to surgery. The majority of anticancer drugs have a narrow therapeutic index, and it is important to determine factors that explain individual variation in efficacy and toxicity. Serum albumin and neutrophil-lymphocyte ratio (NLR) have been previously shown to be associated with chemotherapy toxicity.5 Body composition, in particular proportions of skeletal muscle (and overall lean tissue) in the body, also appear pertinent as several recent results demonstrate an association between skeletal muscle depletion and excess toxicity during treatment with anti-cancer drugs.6e8 The aim of this study was to determine if serum albumin, NLR and features of body composition may be predictive of dose-limiting toxicity (DLT) in patients undergoing neo-adjuvant chemotherapy for oesophago-gastric cancer. The presence of sarcopenia in patients undergoing surgery for various types of cancer has been shown to adversely affect long term survival.9e11 However, the impact of neo-adjuvant chemotherapy toxicity on longterm outcomes post-surgery has not been elucidated. Another aim of the present study was to assess if the presence of DLT has any influence in the post-operative survival of patients with oesophago-gastric cancer.
Neo-adjuvant chemotherapy regimens Squamous cell carcinoma
Day 1: Day 2: Day 3: Day 4: Cycle frequency: Number of cycles:
Cisplatin 80 mg/m2 intravenous infusion 5-Fluorouracil 1000 mg/m2 intravenous infusion 5-Fluorouracil 1000 mg/m2 intravenous infusion 5-Fluorouracil 1000 mg/m2 intravenous infusion 5-Fluorouracil 1000 mg/m2 intravenous infusion Every 21 days 2
Adenocarcinoma Day 1:
Cycle frequency: Number of cycles:
Epirubicin 50 mg/m2 intravenous bolus Cisplatin 60 mg/m2 intravenous infusion Capecitabine 625 mg/m2 twice daily for 21 days (42 doses) Every 21 days 3
Dose limiting toxicity Dose limiting toxicity (DLT) was defined by intolerable toxicities requiring the postponement of treatment, a drug dose reduction, or definitive interruption of drug administration.
Method Body composition measurements The design of this study was approved by the audit department of Nottingham University Hospitals (approval ID 651). The requirement for informed consent from patients was waived because of its retrospective design.
Study population Patients having potentially curative, locally advanced oesophago-gastric cancer without evidence of metastasis presenting to the Nottingham Upper GI MDT between November 2010 and August 2012 were considered for this study. Patients included had histologically confirmed adenocarcinoma or squamous cell carcinoma. All patients were routinely staged with a combination of computed tomography (CT), endoscopic ultrasound (EUS) and laparoscopy according to the International Union Against Cancer (UICC) system.12 Tumours located around the GOJ were classified according to Siewert and Stein13; type I and II tumours were staged as oesophageal cancers and type III as gastric cancers. Those with locally advanced disease (T2 and greater and/or loco-regional lymphadenopathy) and with sufficient physiological reserve to tolerate neoadjuvant chemotherapy were studied. Pre-treatment serum albumin levels and neutrophil-lymphocyte ratio were also determined.
Anthropometric measurements Weight and height were recorded according to standard methods. Weight was measured with a medical balance beam scale and height was measured with a stadiometer. Body mass index (BMI) was calculated [weight (kg)/height (m2)]. BSA was calculated using the Mosteller formula [BSA (m2) ¼ ([height (cm) weight (kg)]/3600)1/2]. Image analysis CT has proven to be accurate for measuring human body composition.14,15 Regional muscle tissue was measured by CT from electronically stored images, which had been done previously for diagnostic purposes. CT scans was performed in all cases at diagnosis prior to commencement of chemotherapy. The third lumbar vertebra (L3) was chosen as a landmark, and two consecutive slices were assessed to measure cross-sectional area of muscle and adipose tissue as described.16,17 The average value of two images was computed for each patient. Images were analyzed using Slice-O-Matic software V4.3 (Tomovision). Crosssectional area for muscle and adipose tissue was normalized for stature (cm2/m2) and reported. Cutoffs for sarcopenia were based on a CT-based sarcopenic obesity study of cancer patients by Prado et al.16 (i.e.,
B.H.L. Tan et al. / EJSO 41 (2015) 333e338
L3 skeletal muscle index: 38.5 cm2/m2 for women and 52.4 cm2/m2 for men). Total LBM was estimated from muscle cross-sectional areas as described in Mourtzakis et al.18: Total body fatfree mass (FFM) (kg) ¼ 0.3 [skeletal muscle at L3 (cm2)] þ 6.06 (r ¼ 0.94). Statistical analysis Data are presented as mean SD unless otherwise stated. Survival was determined from the date of surgery until death or until the censor date of October 21, 2014. Univariate and multivariate analyses and calculation of odds ratios were done using a logistic regression model. Owing to the large number of covariates examined, only those that were significant on univariate analysis were
335
included in multivariate analysis. Receiver-operator characteristic curves were used to select cutoff values for continuous variables. Values with the best combination of sensitivity and specificity were chosen. A backward stepwise procedure was done to derive a final model of the variables that had a significant relationship with survival. To remove a variable from the model, the corresponding p value had to be >0.05. Comparisons between groups of patients were assessed using independent t-test or ManneWhitney U test. Survival curves were constructed using the KaplaneMeier technique. Log-rank test was used to compare survival between groups of patients. p values < 0.05 were regarded as statistically significant. Statistical analysis was done using SPSS 15.0 statistical package (SPSS Inc.). Results
Table 1 Patient demographics. No. of patients (n ¼ 89) Age (years)a
65.8 8.1
Sex M F
67 (75.3) 22 (24.7)
Dose limiting toxicity Present Absent
37 (41.6) 52 (58.4)
Toxicity prevalence Reduction in dose Postponement of treatment Definitive interruption of treatment
24 (27.0) 6 (6.7) 7 (7.9)
Tumour site Oesophageal Gastro-oesophageal junction Gastric
49 (55) 16 (18) 24 (27)
Histology Adenocarcinoma Squamous cell cancer
72 (80.9) 17 (19.1)
Stage I II III
21 (23.6) 27 (30.3) 41 (46.1)
During the study period, 103 consecutive patients were deemed suitable to undergo neo-adjuvant chemotherapy followed by surgery. 14 patients had disease progression during chemotherapy and had their surgery cancelled, and were hence excluded from the study. Demographic data of the remaining 89 patients are shown in Table 1. 37 patients (41.6%) experienced DLT with seven patients (7.9%) failing to complete neo-adjuvant chemotherapy. 44 patients (44.9%) were sarcopenic prior to commencement of chemotherapy. Estimated lean body mass and BSA were moderately related in this patient population (Fig. 1). Sarcopenic patients were significantly older (age of 68.6 7.0 years (mean S.D.) vs. 63.1 8.3 years; p ¼ 0.001), but were otherwise similar with respect to sex, tumour site, histology, stage of disease, pre-
Pre-chemotherapy serum albumin (g/l)a 34.2 5.1 Pre-chemotherapy neutrophil lymphocyte ratioa 2.68 1.86 BMI (kg/m2)a 25.5 4.7 1.86 0.22 BSA (m2)a Lumbar skeletal muscle index (cm2/m2)a 49.1 9.2 Lumbar adipose tissue index (cm2/m2)a 109.6 63.1 Estimated whole body lean mass (kg)a 49.6 10.3 Sarcopenic 44 (49.4) Values are number of patients with percentages in round parentheses unless indicated otherwise. a Values are mean SD.
Figure 1. Relationship between total lean body mass and body surface area in oesophago-gastric cancer patients prior to neo-adjuvant chemotherapy.
336
B.H.L. Tan et al. / EJSO 41 (2015) 333e338
chemotherapy serum albumin and NLR levels (Table 2). As expected, BMI and BSA were found to be significantly lower in those patients with sarcopenia compared with those who did not have sarcopenia. Lumbar adipose tissue index was also significantly lower in sarcopenic patients (90.4 47.6 vs. 128.4 70.9 cm2/m2; p ¼ 0.004). The prevalence of DLT was more common in patients with sarcopenia compared with those without sarcopenia (54.5% vs. 28.9%; p ¼ 0.015). BMI, BSA and sarcopenia were associated with DLT on univariate analysis. Using receiver-operator characteristic curves, the cutoff value with the best discriminatory value for BSA was >1.89 m2. A BMI cutoff of
