1 Fatigue and Cancer Treatment: A Model for the Study of FatigueCharles S. Cleeland, PhD Chair, Department of Symptom Research The University of Texas M. D. Anderson Cancer Center

2 Main Points Cancer treatment is very often associated with large increases in fatigue – related to the toxicity of treatment The severity of fatigue can be measured by self-report (patient-reported outcomes) Treatment-related fatigue generally follows a similar trajectory, dependent on type of treatment This gives us a “window” or model to look at what produces fatigue and what may prevent or treat it

3 Definition of CRF (One of Many)Cancer-related fatigue is an unusual, persistent, subjective sense of tiredness related to cancer or cancer treatment that interferes with usual functioning – NCCN Fatigue Guidelines Committee, 2000

4 Symptoms at End of Life 100 consecutive hospice admissions Fatigue 81%Ng & von Gunten, J Pain Symptom Manage, 1998 100 consecutive hospice admissions Fatigue 81% Anorexia 70% Dyspnea 61% Cough 58% Pain % Depression %

5 Overview of Treatment-Related SymptomsWith more than 10 million cancer patients and cancer survivors, residual symptoms of both disease and treatment represent a significant and under-recognized public-health burden Cancer is increasingly becoming a “chronic” disease, amplifying the issue of symptom burden Until recently, most symptom treatment has been empiric rather than mechanistic, and symptom prevention is virtually unexplored New developments in immunomodulation present opportunities for reducing or even preventing disease-related and treatment-related symptoms

6 The Impact National policy recommendations for increased symptom research (NCI State of Science, 2003) More knowledgeable consumers asking for symptom control Patients who terminate (or are ineligible for) treatment because of symptoms/toxicities Patients and survivors with symptoms that persist (pain, fatigue, impaired cognitive function) Increase in potential symptom-focused therapies (nonspecific anti-inflammatory agents, antidepressants, immunomodulators)

7 Common Long-Term Disease-Related and Treatment-Related SymptomsNeuropathy – pain and loss of fine motor coordination, difficulty with walking Other pain syndromes (e.g., bone pain) Fatigue – most patients during treatment and a significant percentage of cancer survivors Cognitive deficits – Loss of “executive function” and some types of memory, without IQ loss Persistent sleep disturbance, reflected primarily in sleep quality

8 Why Study Treatment-Related Symptoms?Several biologic mechanisms, including Inflammatory cytokines, are hypothesized to be responsible for both disease-related and treatment-related symptoms The time course of treatment-related symptoms is more predictable, and has a beginning time point, allowing for more precise linkage Knowledge concerning the mechanisms of treatment-related symptoms may provide information about these symptoms in other conditions as well (such as fatigue in aging)

9 BMT Patients: Mean Symptom Ratings

10 Brief Fatigue Inventory (Severity)3. Please rate your fatigue (weariness, tiredness) by circling the one number that best describes your WORST level of fatigue during the past 24 hours. No As bad as Fatigue you can imagine

11 Brief Fatigue Inventory (Interference)7. Circle the number that describes how, during the past 24 hours, fatigue has interfered with your: A. General activity Does not Completely Interfere Interferes

12 Means (SDs) for Fatigue Outcome Measures Across Performance Status Category* Significantly different at each level of ECOG performance status (p<0.001)

13 Relationship Between Fatigue at its Worst and Interference

14

15 Most Common Severe Symptoms During Treatment in M. DMost Common Severe Symptoms During Treatment in M. D. Anderson Outpatients (N=527) 7 or greater on M. D. Anderson Symptom Inventory 0–10 scale Cleeland et al, Cancer, 2000)

16 How Symptom Clusters During TreatmentCleeland et al, Cancer, 2000 Nausea Vomiting Bleeding Diarrhea Mouth sores Shortness of breath Cough Constipation Bloated Remembering Attention Distress Worrying Sad Nervous Irritable Lack of appetite Sick Fatigue Weak Not able to get things done Drowsy Pain Disturbed sleep Dry mouth Numbness/tingling

17 Cytokine Treatment Melanoma Patients: Symptom Severity over One Cycle

18 Postoperative Studies: Symptom SeverityMean Symptom Scores (0-10 scale) Time

19 Symptom Patterns with Aggressive Cancer Therapies Fatigue, Pain, Sleep, Distress, Shortness of Breath NSCLC Thoracic Surgery Weeks NSCLC Chemo-radiation Days AML/MDS Allo-HSCT Days GI Chemo-radiation Days

20 Example: Fatigue Severity Across TimeChemoXRT (n=48) Repeated Measures Benefits Ease of data gathering No need keep track of patients across a treatment regiment High volume of patient data in a limited amount of time Drawbacks Limited in scope Lose impact of symptom between assessments Partial picture of symptom expression XRT only (n=46) 7 6 5 Fatigue Severity 4 3 Example of how repeated measures is not effective in capturing true symptom differences between groups – misses spikes, despite the fact that both groups begin and end at same symptom severity 2 Wilkes Lambda = .71 F(1,93)=.285, p=.845 1 1 2 3 4 5 End 7 8 9 10 of Tx Time Point Burkett, et al., 2007, unpublished data.

21 Fatigue Area Under the Curve (AUC): Effectively Combining Severity and TimeChemoXRT (n=48) XRTOnly (n=46) Longitudinal Data Frequent assessment is invaluable Enables tracking symptoms across treatment regimens Provides picture of symptom change as the treatment and recovery progresses Area Under the Curve Single value representing time and symptom severity Compare groups using standard statistical methods Compensates for missing data 60 50 40 30 Average Fatigue AUC 20 AUC is a method of catching differences in symptom expression between groups t (93)=6.41, p = .04 * 10 Baseline Week 2 Week 3 Week 4 Week 5 Week 7 Week 8 Week 9 End of Tx Week 10 Timepoint Burkett, et al., 2007, unpublished data.

22 Ongoing CRF Studies: www.clinicaltrials.govATP American Gingseng Co-enzyme Q10 Etanercept Levocarnitine Modafinil Methyphenidate Zoloft

23 Drug Therapy for CRF: Cochrane Database System Review 2008Type Drug Overall Effect Standardized Mean Difference 95% CI Psychostimulants Methyphenidate [2, N=264] Z=2.40 P=.02 -0.30 -.54 to -.05 Hemopoetic growth factors Erythropoietin [10, N=3735] Z=8.32 P=.008 -.46 to -.29 Darbopoetin [4, N=1650] Z=1.45 P=.05 -0.13 -.27 to .00 Anti-depressants Paroxetine [2, N=645] No superiority over placebo Progestational steroids Megestrol acetate [4, N=587]

24 Sickness Behavior: An Animal Model for Cancer-Related Symptoms?Physiological components – fever, pain, wasting, increased HPA, autonomic activity Behavioral components – somnolence, hyperalgesia, impaired learning, and decreased social interaction, exploration, and eating Inflammatory cytokines/chemokines and neurotransmitters may play central mechanistic role

25 Examples: Cancer- and Treatment-Related Symptoms Associated with InflammationAnorexia/Cachexia: Elevated serum IL-6, TNF-a is confirmed to affect animals; TNF-a is important but not exclusively responsible for anorexic effects of tumor Fatigue: IL-6mAb blocks IL-6 activity in Castleman disease; IL-6 is high in fatigued breast cancer survivors Pain: IL-1b, IL-6 and TNF-a are increased in neuropathic pain, hyperalgesia, and extreme sensitivity to pain Cognitive Dysfunction: Elevated IL-6 is associated with cognitive deficits in AML/MDS and LMD patients Sleep: IL-6, TNF-a are increased in sleep deprivation Paraneoplastic syndrome in SCLC or metastatic RCC: IL-6–related fever, fatigue, weight loss

26

27 Serum IL-6 Fluctuates with Symptom Peak: 100 days of Allo-BMT in AML/MDS PatientsWang et al, Cancer, in press.

28 Random Effects Modeling: Serum Cytokines Predicted Symptom Changes from Baseline to PeakWang et al, Cancer, in press. A. Increase in Symptom Severity (Baseline to Peak) Cytokines Changes (Baseline to Day +8 after allo-HSCT) IL-6 IL-8 IL-10 IL-1RA IL-1b sTNF-RI IL-12p40p70 Estimate‡ 1.112 0.294 0.390 0.147 1.613 1.133 -0.245 Standard error 0.505 0.393 0.357 0.500 0.517 0.915 1.842 P § 0.006 0.411 0.437 0.776 0.383 0.218 0.710 B. Change in Symptom Severity (Baseline to Day +30) Cytokines Changes (Baseline to Day +100 after allo-HSCT) 1.050 0.169 0.610 0.037 -0.135 1.367 -0.454 0.245 0.194 0.369 0.304 0.564 0.650 < .0001 0.100 0.903 0.812 0.036 0.124 Dependent variable is a symptom component score. Age, gender, race, disease status, infusion cell service, conditioning regimen, and infusion dose of CD34+ cells were adjusted in all models.

29 Random Effects Modeling: Association of Changes Symptom Severity and Cytokines for 7 weeks Concurrent CXRT (N=60) Wang et al, unpublished data. Estimate Std Error DF t value P value Solution for Fixed Effects: Total 15 Items sTNF-R1 1.7644 0.6930 147 2.55 0.0119 Solution for Fixed Effects: Pain + Sore Throat IL-6 0.7977 0.1906 4.19 <.0001 IL-8 0.1698 -2.96 0.0036 Solution for Fixed Effects: Distress + Sadness 2.6074 1.2701 2.05 0.0418 Controlled for age, gender, race, recurrent, chemo, total XRT dose, and XRT technique (3DCRT or IMRT).

30 Random Effects Modeling: Association of Symptom Severity and Cytokines from Baseline to 10 weeks from Start of Concurrent CXRT Wang et al, unpublished data. IL-6 IL-8 IL-10 Estimate SE P value Core items 0.689 0.183 0.0002 0.274 0.173 0.117 0.194 0.332 0.560 Total symptom severity 0.713 < .0001 0.206 0.164 0.213 0.179 0.313 0.570 Pain 1.819 0.327 -0.050 0.317 0.875 0.218 0.522 0.677 Fatigue 1.028 0.328 0.002 0.126 0.315 0.690 -0.506 0.549 0.359 Nausea 0.696 0.294 0.019 0.298 0.358 0.478 0.451 0.291 Lack of appetite 1.317 0.338 0.290 0.331 0.382 0.016 0.536 0.976 Drowsy 1.088 0.329 0.001 0.312 0.322 0.333 0.107 0.524 0.839 Vomiting 0.666 0.240 0.006 -0.400 0.251 0.113 -0.079 0.345 0.819 Sore throat 1.244 0.342 0.0004 -0.276 0.350 0.431 0.429 0.489 0.381

31 Etanercept and Clinical Outcomes, Fatigue, and Depression in Psoriasis: Double-blind Placebo-controlled Randomized Phase III Trial Tyring s et al, Lancet, 2006. With etanercept, a soluble TNF-α receptor, improvements in fatigue were correlated with decreasing joint pain; improvements in depression were less correlated with objective measures of skin clearance or joint pain. Etanercept treatment might relieve fatigue and symptoms of depression associated with this chronic disease.

32 Future Research DirectionsAdditional longitudinal studies of biology of fatigue development and risk for fatigue (genetics) in response to cancer treatment More testing of interventions that modify pathways for inflammation (NF-kB, cytokine inhibition Neuroimaging of treatment-based fatigue Animal models of fatigue for preclinical investigation

33 Continuous Tracking of Activity

34 Fatigue and Aging

35 Symptom Research StrategyClinical Oncology Tested interventions move into community DISCIPLINES Surgery to remove tumor Radiation and/or chemotherapy Clinical Trials * Interventions Phase I and II clinical trials * * Symptom severity differential between patients * = points of interest Protective measures Pre-Surgery TREATMENT TRAJECTORY Symptom Research MDASI MDASI MDASI MDASI Protein Measurement Cytokines Cytokines Cytokines Cytokines Genomics DNA & RNA RNA RNA RNA RNA Neurosensory Sensory testing Sensory testing Biostatistics Longitudinal modeling Neuroimaging fMRI and PET fMRI and PET Neurocognitive Cognitive testing Cognitive testing Emotion Mood assessment Mood assessment Results of interdisciplinary efforts lead to new hypotheses to be tested first in animals Animal Behavior Models of cancer and cancer treatment to test compounds and inform clinical trials

36 Conclusions Increased fatigue is endemic in cancer, and most prominent reported symptom before death Fatigue is conceptually complex, but patient report of tiredness/fatigue represents the state Treatment-related fatigue often follows a similar trajectory, and presents a model for understanding the biology of fatigue Several avenues of investigation of fatigue are now available, and could be applied to this model

37

38 Add slide

39 Examples: Cancer- and Treatment-Related Symptoms Associated with InflammationAnorexia/Cachexia: Elevated serum IL-6, TNF-a is confirmed to affect animals; TNF-a is important but not exclusively responsible for anorexic effects of tumor Fatigue: IL-6mAb blocks IL-6 activity in Castleman disease; IL-6 is high in fatigued breast cancer survivors Pain: IL-1b, IL-6 and TNF-a are increased in neuropathic pain, hyperalgesia, and extreme sensitivity to pain Cognitive Dysfunction: Elevated IL-6 is associated with cognitive deficits in AML/MDS and LMD patients Sleep: IL-6, TNF-a are increased in sleep deprivation Paraneoplastic syndrome in SCLC or metastatic RCC: IL-6–related fever, fatigue, weight loss