Beyond Standard Therapy for Colorectal Cancer: Role of Energy Balance in Treatment of Survivors

Transcription

1 Beyond Standard Therapy for Colorectal Cancer: Role of Energy Balance in Treatment of Survivors Jeffrey Meyerhardt, MD, MPH Dana-Farber Cancer Institute Boston, MA 1

2 Disclosure Ad Board: Genentech Honorarium: Chugai Involvement in IDEA Collaboration Study co-chair of CALGB/SWOG

3 Energy Balance Calorie Intake Modified by ingestion and/or absorption (e.g., changes in microflora, surgery, pharmacologic agents) Body Weight Gain Stable Loss Resting Metabolic Rate Energy needed to maintain body function at rest. Accounts for ~60-75% of expenditure. Modified by lean body mass (age), and external/internal temperature. Physical Activity Energy needed for activity. Usually accounts for 15-30% of expenditure, but most modifiable component Thermic Effect of Food Energy needed for digestion or metabolism. Accounts for <10% of expenditure. Modified somewhat by digestibility, capsaicin, caffeine. Energy Intake (energy in) Total Energy Expenditure (energy out) From Demark-Wahnefried et. al. Cancer Epidemiol Biomarkers Prev Aug;21(8): , 3

4 Physical Activity and Colorectal Cancer Outcomes 4

5 Physical Activity Immediately After Surgery Stretching Core Resistance Ambulation Int J Colorectal Dis May;28(5):

6 Physical Activity Immediately After Surgery Exercise group Usual care Median (IQR) Mean (SD) Median (IQR) Mean (SD) p value Effect size Time to flatus (h) 50.9 ( ) 52.2 ± ( ) 71.9 ± Time to liquid diet intake (h) 69.8 ( ) 76.9 ± ( ) 86.0 ± Int J Colorectal Dis May;28(5):

7 Role of Physical Activity during Active Treatment and Recovery Numerous studies have examined association of physical activity to quality of life and overall health Timing and type of intervention varied but generally show improvement in tolerance to therapy, fitness, fatigue and QOL 7

8 Physical Activity and Colorectal Cancer Cohort study from Australia of 526 colorectal cancer patients with pre-diagnosis physical activity assessment Colorectal cancer specific survival Haydon Gut Jan;55(1):62-7 8

9 CALGB NCI-sponsored adjuvant therapy trial for stage III colon cancer Patients randomized to Roswell Park 5-FU/LV or IFL (bolus 5- FU/LV/Irinotecan) 1264 enrolled between 1999 and 2001 Complete questionnaire Complete questionnaire Patients enroll on adjuvant therapy trial chemotherapy every 3 month f/u Saltz, L. B. et al. J Clin Oncol; 25:

10 CALGB

11 Hazard Ratio Recurrence or Death and Exercise: Disease-Free Survival in Stage III Colon Cancer Survivors Regular Physical Activity (met-hours per week) Meyerhardt, J. A. et al. J Clin Oncol; 24:

12 89803 and Exercise: Disease Free Survival Meyerhardt, J. A. et al. J Clin Oncol; 24:

13 89803 and Exercise: Stratification Meyerhardt, J. A. et al. J Clin Oncol; 24:

14 Statistical Considerations Reverse causality Is the exposure changing outcomes or the outcome changing exposure Restrict to events at least 90 days from exposure Sensitivity analyses to extend restriction to 6 months and 12 months Recall bias The clock starts at time of questionnaire completion all events are prospective beyond the exposure data Limits generalizability data speak to those that get to point of questionnaire 14

15 NHS and Post-diagnosis Physical Activity Meyerhardt, J. A. et al. J Clin Oncol; 24:

16 NHS and Post-diagnosis Physical Activity Meyerhardt, J. A. et al. J Clin Oncol; 24:

17 NHS and Pre-diagnosis Physical Activity Meyerhardt, J. A. et al. J Clin Oncol; 24:

18 NHS and Change in Physical Activity Meyerhardt, J. A. et al. J Clin Oncol; 24:

19 Meta-Analysis of Pre-Diagnosis Physical Activity and Colorectal Cancer Outcomes Je, Jeon, et al Int J Cancer In Press. 19

20 Meta-Analysis of Post-Diagnosis Physical Activity and Colorectal Cancer Outcomes Je, Jeon, et al Int J Cancer In Press. 20

21 CHALLENGE: Colon Health and Life-Long Exercise Change trial High risk Stage II or stage III colon cancer - completed adjuvant chemotherapy within 2-6 months REGISTRATION Baseline Testing STRATIFICATION Disease stage high risk III; centre; BMI 27.5 vs. > 27.5; ECOG PS 0 vs. 1 RANDOMIZATION ARM 1 Physical Activity Program + General Good Health Education Material (Intervention Arm) ARM 2 General Health Education Materials (Control Arm) Assessment of disease-free survival every 6 months for first 3 years and annually from years 4-10 Courneya Curr Oncol.2008 Dec;15(6):

22 Physical Activity in Metastatic CRC: CALGB Physical Activity (Total MET h/w) P for < trend Median (range) 0.4 ( ) 5.5 ( ) 13.5 ( ) 34.1 ( ) - OS Event/N 461/ / / /224 Unadjusted 1 (Referent) 0.86 ( ) 0.86 ( ) 0.73 ( ) Adjusted 1 1 (Referent) 0.88 ( ) 0.85 ( ) 0.79 ( ) 0.02 Adjusted 2 1 (Referent) 0.92 ( ) 0.88 ( ) 0.83 ( ) 0.06 PFS Event/N 527/ / / /224 Unadjusted 1 (Referent) 0.97 ( ) 0.81 ( ) 0.79 ( ) Adjusted 1 1 (Referent) 0.99 ( ) 0.80 ( ) 0.82 ( ) 0.01 Adjusted 2 1 (Referent) 1.01 ( ) 0.81 ( ) 0.85 ( ) 0.03 Adjusted 1: age, gender, PS (0 v 1), planned chemotherapy (FOLFOX v FOLFIRI), treatment arm A,B, C), BMI, prior adjuvant therapy Adjusted 2: above plus comorbidities and weight loss prior 6 months Guercio et al GI ASCO

23 Physical Activity and Change in Insulin Markers Alberta Physical Activity and Breast Cancer Prevention Trial RCT in postmenopausal, inactive, cancer-free women Year old aerobic exercise intervention of 225 min/week v control group (n= 320) Insulin (miu/ml) Baseline 6 months 12 months Treatment Effect Between Group P Exercisers ( ) <0.001 Control HOMA Exercisers ( ) <0.001 Control Leptin (ng/ml) Exercisers ( ) <0.001 Control Friedenreich et al Endocr Relat Cancer Jun; 18(3):

24 Association between Physical Activity and Outcomes by PTGS2 (COX2) Status Physical activity may also mediate inflammatory status by reducing PGE2 synthesis Yamauchi M et al. Cancer Epidemiol Biomarkers Prev 2013;22:

25 Weight and Body Composition and Colorectal Cancer Outcomes 25

26 NSABP and Body Mass Index Disease-free and overall survival by body mass index (BMI) category in 4288 patients from National Surgical Adjuvant Breast and Bowel Project randomized clinical trials for Dukes B and C colon cancer Dignam, J. J. et al. J. Natl. Cancer Inst :

27 C-SCANS: Diagnosis BMI and Mortality Outcomes (N=3,409) Kroenke et al JAMA Oncol. Published online May 19,

28 Meta Analyses of Post-Diagnosis BMI in Stage I-III CRC Schlesinger et al. Cancer Causes Control 2014 Oct;25(10):

29 Post-diagnosis BMI and Outcomes Author Years N Outcome Hazard Ratio (95% CI) or P value (compared to normal weight) Meyerhardt ,759 OS 1.11 ( ) for >30 BMI v Meyerhardt ,688 OS 1.09 ( ) for >30 BMI v Dignam ,288 Meyerhardt ,053 CRC OS CRC OS 1.27 ( ) for >35 BMI v ( ) for >35 BMI v ( ) for >35 BMI v ( ) for >35 BMI v Sinicrope ,381 OS 1.19 ( ) for >35 BMI v Baade 2011 CRC 1.34 ( ) for >30 BMI v ,561 increase OS until 0.78 ( ) for >35 BMI v Chin ,765 CRC 1.06 ( ) for >30 BMI v BMI > 35 OS 0.94 ( ) for >30 BMI v Kuiper CRC 0.95 ( ) for >30 BMI v OS 1.09 ( ) for >30 BMI v Campell ,303 CRC OS 1.14 ( ) for >30 BMI v ( ) for >30 BMI v Sinicrope OS 1.11 ( ) for >35 BMI v Kroenke OS Risk doesn t 0.76 ( ) for >30 BMI v ( ) for >35 BMI v

30 Definition of the Obesity Paradox Overweight/Obesity is related to increased disease Once people get the disease overweight/obesity is also related to better survival. 30

31 Obesity Paradox Explanations Inverse associations are not real and reflect methodological issues; reverse causation or collider /selection bias Associations are true and plausible; excess energy reserves are important in cancer survivors BMI is not an accurate measure of adiposity, especially among cancer survivors Muscle, which increases with adiposity, may play an important protective role 31

32 Importance of Skeletal Muscle Skeletal muscle is the most abundant tissue in the body Besides contraction, it accomplishes many metabolic functions, modulates insulin resistance, energy expenditure, peptides regulating anabolism and catabolism. Muscle depletion in cancer not just mechanical function loss 32

33 C-SCANS Goal: Examine muscle mass, fat components on CRC survival Hypothesis: Muscle mass is an important independent predictor of survival and help explain the obesity paradox Population: Kaiser Permanente Cancer Registry and Electronic Medical Records CRC Stages I-III, Ages (N=3409), Have CT scan within the window of 2m prior to 3m post dx and before chemo/radiation Have a weight measure at same time Outcomes: Overall and CRC mortality Caan et al CEBP

34 Computerized tomography: Measuring Body Composition Mourtzakis et al. Appl Physiol Nutr Metab Oct;34(5): Prado et al. Curr Opin Support Palliat Care 2009 Dec;3(4): Slice-O-Matic software to assess muscle and fat (Hounsfield units) 34

35 C-SCANS: Body Composition and Mortality Low Medium High # Events Tertile 1 Tertile 2 Tertile 3 P trend All-Cause Mortality 681 Muscle (cm 2 ) Referent 0.78 (0.64, 0.94) 0.66 (0.53, 0.84) Visceral Fat (cm 2 ) Referent 0.82 (0.66, 1.00) 1.01 (0.81, 1.27) 0.89 Subcutaneous Fat (cm 2 ) Referent 0.94 (0.78, 1.15) 0.96 (0.77, 1.21) 0.64 CRC-Specific Mortality 414 Muscle (cm 2 ) Referent 0.75 (0.58, 0.95) 0.60 (0.45, 0.81) Visceral Fat (cm 2 ) Referent 0.68 (0.52, 0.88) 1.03 (0.77, 1.37) 0.87 Subcutaneous Fat (cm 2 ) Referent 1.13 (0.87, 1.46) 1.12 (0.83, 1.50) 0.58 All models adjusted for age at dx (continuous), gender, race, stage, receipt of chemotherapy, colon v rectal, receipt of radiation and BMI with fat and muscle partitioned out Muscle model additionally adjusted for visceral and subcutaneous fat Fat models additionally adjusted for visceral or subcutaneous fat and muscle Caan et al CEBP

36 Survivor function estimate Sarcopenia and All-Cause Mortality Log-rank test p=< Years from diagnosis #At risk: Not Sarcopenic Sarcopenic Caan et al CEBP

37 Rate Ratio Skeletal Muscle and All-Cause Mortality 3 2 Men Women Skeletal Muscle Index (cm2/ht2) Restricted cubic spline adjusted for age, race, stage, site, treatment and partitioned BMI SMI spline further adjusted for SF and VF in quintiles. Caan et al CEBP

38 Rate Ratio Visceral Fat and All-Cause Mortality 3 2 Stat sign. at 80% Stat sign. at 95% Men Women Visceral Fat (cm2) Restricted cubic spline adjusted for age, race, stage, site, treatment and partitioned BMI VF spline further adjusted for SF and muscle in quintiles. Caan et al CEBP

39 Rate Ratio Rate Ratio Mean BMI Associated with Cutpoints (muscle, fat) that Demonstrate Increased Risk 3 BMI <25, low skeletal muscle increases risk of death 2 1 BMI 25.1 BMI 25.3 Men Women Skeletal Muscle Index (cm2/ht2) 3 BMI > 32, adiposity increases risk of death 2 Men Women 1 BMI 32 BMI 35 Caan et al CEBP SubQ + Visceral Fat (cm2) 39

40 89803 and Change in Weight Adjusted Hazard ratio (95% CI) > 5 kg weight loss 1.39 ( ) kg weight loss 1.15 ( ) +/- 2 kg change Referent kg weight gain 1.11 ( ) > 5 kg weight gain 1.19 ( ) Ptrend = 0.13 Ptrend = 0.90 Meyerhardt J Clin Oncol Sep 1;26(25):

41 C-SCANS: Weight Change After Diagnosis Large Loss (>10%) % Weight Change [Hazard Ratio (95% Confidence Interval)] Modest Loss (5-9.9%) Colorectal Cancer Specific-Mortality Stable ( %) Modest Gain (5-9.9%) Large Gain (>10%) P Loss P Gain # Events / At risk Adjusted ** Overall Mortality 65/239 43/ / /453 32/ ( ) 1.58 ( ) Referent 0.84 ( ) 0.93 ( ) < # Events / At risk Adjusted ** 104/239 79/ / /453 68/ ( ) 1.74 ( ) Referent 0.86 ( ) 1.20 ( ) < Meyerhardt J et al CEBP 2016; In Press 41

42 C-SCANS: Weight Change After Diagnosis Meyerhardt J et al CEBP 2016; In Press 42

43 Diet and Colorectal Cancer Outcomes 43

44 Hazard Ratio for Cancer Recurrence or Death CALGB 89803: DFS By Dietary Pattern P, trend < Western diet Prudent diet Quintiles of Dietary Pattern Meyerhardt, J. et al. JAMA (7):

45 CALGB 89803: Dietary Pattern Meyerhardt, J. et al. JAMA 2007;298:2263-a. 45

46 Dietary Patterns Study of 529 CRC patients in Newfoundland Pre-diagnosis diet DFS CRC Colon Rectal Zhu et al BMJ Open 2013 Feb 7;3(2). 46

47 Hazard Ratio for Cancer Recurrence or Death Total Carbohydrates in Colon Cancer Patients Quintiles of Total Carbohydrates Meyerhardt, J. et al JNCI

48 Hazard Ratio for Cancer Recurrence or Death Glycemic Load in Colon Cancer Patients Quintiles of Glycemic Load Meyerhardt, J. et al JNCI

49 Hazard Ratio for Cancer Recurrence or Death Glycemic Load in Colon Cancer Patients BMI < BMI > Quintiles of Glycemic Load Meyerhardt, J. et al JNCI

50 Conclusions / Next Steps Growing data that energy balance factors important in colorectal cancer outcomes Multiple mechanisms may be in place Insulin and insulin-like growth factors Changes in body composition having direct effects Inflammation Tying these observations together will be critical to better understanding mechanism and guiding patients These are all observations will intentional change improve one s condition? Answer is not as simple as losing weight in CRC survivors 50

51 Acknowledgements Dana-Farber Harvard School of Duke Charles Fuchs Public Health Lee Jones Shuji Ogino Ed Giovannucci Kaori Sato Andy Chan Yonsei University Jennifer Chan Walter Willett Justin Jeon Jennifer Ligibel Frank Hu Robert Mayer Kaiser Permanent Kimmie Ng CALGB/ALLIANCE Bette Caan Brian Wolpin Donna Niedzwiecki Erin Weltzein Nadine McCleary Donna Hollis Candyce Kroenke Denise Brady Cynthia Ye Carla Prado Neera Jagirdar Jingjie Xiao Azuka Onye UCSF Adrienne Castillo Devin Wigler Alan Venook Marilyn Kwan Nathalie Fadel Elizabeth Cespedes Laura Shockro MSKCC Nancy Campbell Leonard Saltz ALL PATIENTS and Sui Zhang SUBJECTS who COMPLETED Yale QUESTIONNAIRES and JOINED Melinda Irwin RELATED TRIALS 51