Download OHSU Cancer PowerPoint Presentation

Login   OR  Register

Share page


Home / Health & Wellness / Health & Wellness Presentations / OHSU Cancer PowerPoint Presentation

OHSU Cancer PowerPoint Presentation

onlinesearch By : onlinesearch

On : Feb 24, 2014

In : Health & Wellness

Embed :

Login / Signup - with account for

  • → Make favorite
  • → Flag as inappropriate
  • → Download Presentation
  • → Share Presentation
  • Slide 1 - U.S. Food and Drug Administration Notice: Archived Document The content in this document is provided on the FDA’s website for reference purposes only. It was current when produced, but is no longer maintained and may be outdated.
  • Slide 2 - Potential Use of Biomarkers or QoL Parameters in Colorectal Cancer Drug Approval Charles D. Blanke, M.D., F.A.C.P. Director, G.I. Malignancies Program Oregon Health & Science University Cancer Institute
  • Slide 3 - Biomarkers and QOL in CRC Drug Approval Biomarkers CEA CA 19-9 + others QoL issues How symptomatic are modern advanced disease patients? Instruments Problems with QoL analysis Recommendations + group discussion
  • Slide 4 - CEA: Background Serum glycoprotein member of the immunoglobulin gene superfamily Intercellular adhesion molecule promoting aggregation of malignant cells; involved in immunity, apoptosis, and metastasis Elevated in a variety of inflammatory diseases Malignant: Breast, lung, gastric, cervical, kidney, bladder carcinomas Melanoma NHL Benign: Cirrhosis, gallstones, emphysema, PUD, diabetes Elevated in ~85% of metastatic CRC patients
  • Slide 5 - CEA Background (cont.) Most, if not all, CEA data precede modern drugs ASCO recommendations for use in metastatic disease: CEA is marker of choice for monitoring CRC Present data are insufficient to use CEA alone to monitor response But, 2 values > baseline = progression, even without corroborating x-rays
  • Slide 6 - Changes in CEA with Cytotoxic Chemotherapy Analysis of sensitivity and specificity for objective response and progression available-see me later Positive predictive value for CEA + true objective response: 53% Negative predictive value for CEA + response: 100% PPV for CEA + progressive disease: 85-100% NPV for CEA + progressive disease: 70%
  • Slide 7 - CEA in CRC: Translation You can have a CEA response without a true ORR You can’t have a true ORR without a CEA response Easier to turn CEA production off than to actually kill cells If you kill cells, CEA falls You can progress without CEA progression You cannot have CEA progression without x-ray PD CEA progression may occur months before change on CT X-ray PD without CEA PD particularly common with growth of anaplastic tumor clones CEA response correlates with survival even in absence of ORR
  • Slide 8 - Additional Caveat: Hepatotoxicity Liver is major site of clearance for CEA Moertel adjuvant trial: 39.6% of patients on 5FU/levamisole sustained hepatotoxicity 19% of patients with hepatotoxicity had CEA rises, in the absence of disease recurrence Does not occur with 5FU alone, but ??? effect of new drugs or combinations Moertel JCO 11:2386, 1993
  • Slide 9 - Quality of Life: A multidimensional construct encompassing complete information on the impact of disease or its treatment on a patient’s usual or expected physical, psychological, and social well-being
  • Slide 10 - How Symptomatic is the Average Advanced Disease CRC Patient? All standard texts reviewed: “Signs and symptoms are usually present” References for that statement never supplied Seer data: No help Trial data: Little help-skewed to less symptomatic patient or those with minor symptoms Review of mito C and IFL trials: 90% asymptomatic Cunningham irinotecan arm 69% “with sx”, but 86% WHO PS 0 or 1 My review of multiple series/trials: < 50% had significant sx With modern early diagnosis, we may have to focus more on delaying symptoms than treating them
  • Slide 11 - How Do We Measure QOL in CRC Patients? Generic instruments Cancer-specific instruments Symptom-focused questionnaires Clinical benefit response
  • Slide 12 - Generic Instruments Multi-item scales used across a wide-range of chronic-disease populations Examples: Sickness Impact Profile, Nottingham Health Profile Allow for comparisons of populations with different health conditions Insensitive for effects of specific interventions
  • Slide 13 - Cancer-Specific Instruments Address problems specific to a population with cancer Examples: EORTC Quality of Life Questionnaire (QLQ)-C30, Functional Assessment of Cancer Therapy-General version (FACT-G), Functional Living Index-Cancer (FLIC) May have functional, symptoms, global scales +/- single items assessing specific symptoms Have been used in large series which included CRC patients
  • Slide 14 - Symptom-Focused Questionnaires Often used with Cancer Specific Instruments, or other HRQoL questionnaires Some add modules for specific issues not covered by the core instrument tumor-site-specific subscales (FACT-C) More sensitive and specific for clinically important QoL differences EORTC QLQ-CR38 and FACT-Colorectal (FACT-C) designed for colorectal cancer Sub-sub modules being developed for CRC mets to the liver NB: Symptoms have many measurable dimensions, including severity, frequency, functional consequences, and distress perceived by patient
  • Slide 15 - Clinical Benefit Response Composite of status encompassing pain, functional impairment (PS), and weight loss Defined as improvement in at least one parameter, without worsening of others Requires sustained improvement (> 4 weeks) over baseline
  • Slide 16 - Interpretation and Application of HRQoL Scores Mean scores for a group can be compared with “normative values” for a sample of cancer patients Changes in scores within a group can be assessed over time to determine effect of disease or treatment Differences in scores can be compared between two groups given different treatments
  • Slide 17 - Summary of HRQoL Data in CRC Baseline QoL is itself a major prognostic factor Independent predictor of overall survival; stronger than PS! HRQoL is a good global way to balance toxicity vs. therapeutic benefit Review articles claim QoL improves with chemotherapy: Occurs in 25-33% of symptomatic patients overall QoL “maintained” in another 25% In patients without symptoms, chemotherapy not likely to affect QoL, unless adversely Response to chemotherapy does not influence QoL
  • Slide 18 - IFL vs. FL: Front-Line EORTC QLQ v2 used No “significant differences” between arms Smaller decrease in activities of daily living- IFL Saltz NEJM 343:905, 2003
  • Slide 19 - IFL vs. BSC: Second-Line EORTC QLQ-C30 Rate of return 59-62% Later definitive deterioration QoL in “I” arm Cunningham Lancet 355:1041, 2000
  • Slide 20 - FOLFOX vs. LV5FU2: Front-Line EORTC QLQ-C30 v2.0 84% “participated” in analysis Overall QoL comparable Longer time to deterioration of global health-FOLFOX Response did not influence QoL de Gramont JCO 18:2938, 2000
  • Slide 21 - Problem: We Don’t Know that Palliative Chemotherapy Truly Improves QOL CRC Collaborative Group meta-analysis 13 RCT comparing palliative chemotherapy with BSC Chemotherapy included any single drug or combo on any schedule 3 trials: QOL better with chemotherapy; 2 found no differences between chemotherapy + control; 1 (ineffective chemotherapy) found QOL worse with therapy Conclusion: Data on the effect of chemotherapy on QOL are inadequate to draw firm conclusions about the palliative effect CRC Collaborative Group BMJ 321:531, 2000
  • Slide 22 - Problem: Inconsistent Methodologies with HRQoL Assessment Missing data ubiquitous Few patients complete all serial questionnaires May lead to bias, if ill/progressing patients are ones not filling out forms Questions may not be agent/toxicity specific/sensitive (e.g., neuropathy not part of standard CRC questionnaires) Patients filling out QoL as part of studies may not be representative of general CRC population PS 0 patients with high initial QoL will not improve further Reliability depends on whether physician or patient fills out forms
  • Slide 23 - Problems with HRQoL Assessment (cont.): Composite scores usually generated Score on loss of appetite scale may not have same significance as identical score on the pain scale Hard to interpret magnitude of QoL changes between patients May not be sensitive enough to differentiate progression from chemotherapy toxicity Timing of QoL assessment affects results, but there are no rules when to do testing if two regimens are given on different schedules Changes in scores may not reflect important physical/clinical change, if tool is dominated by psychosocial domains (historically less sensitive to treatment-induced effects)
  • Slide 24 - Problem: HRQoL Reflects Efficacy and Safety (Toxicity) While an advantage for assessing overall impact of drug , leads to same FDA approval problem as using time-to-event In theory, QoL could improve in absence of effective drug (patient “doing something”), especially in an asymptomatic patient QoL could worsen even with major response, with a toxic drug Can we use clinical benefit response (CBR) instead? Worked for gemcitabine in pancreatic cancer It’s easy
  • Slide 25 - Problems with Use of CBR in CRC Does not take into account all disease-related symptoms of importance Would be hard to develop CRC-specific CBR There is only fair consensus among experts which symptoms are important (fever, fatigue, pain, weight loss, diarrhea, abdominal swelling, appetite, constipation, etc.) and/or even common Some symptoms may require surgery to fix In non-pancreatic GI sites, CBR does not correlate with subjective response or QoL anywhere near as well
  • Slide 26 - Additional Problems with Using Symptom Assessment There has been lack of consistency in terminology used to discuss specific symptoms e.g., does fatigue = tiredness, sleepiness, generic decrease in PS? Clinical significance of improvement depends on magnitude of change, but also severity at start e.g., “50%” improvement in pain scale from 2 to 1 would qualify as CBR, but probably not important
  • Slide 27 - Additional Problems with Using Symptom Assessment (cont.) There is no good research definition for what counts as symptom “control” Changing definition alone has changed conclusions of clinical trials (RTOG Osseous Metastasis Trial) Tools not sensitive enough to evaluate all items which may be important: time of onset of relief, duration of relief, how significant relief is to patient, whether severe symptoms are relieved
  • Slide 28 - Colorectal Cancer Drug Approval 1Objective response rate, overall survival 2If enough patients symptomatic to start
  • Slide 29 - Biomarkers: Conclusions CEA (and all markers) overestimate responses and underestimate progressive disease Despite association of fall with improved survival, CEA change is clumsy, and I recommend against use in CRC drug approval ASCO recommends against using CA 19-9, DNA ploidy, flow cytometry characteristics, lipid-associated sialic acid, p53, ras, in monitoring treatment of colorectal cancer Certainly could not use markers to monitor new drugs, unless effect of drugs on LFT fully characterized (Moertel)
  • Slide 30 - HRQoL Assessment:Conclusions Baseline HRQoL does correlate with other important endpoints (e.g., survival) Jury out on whether change in QoL reliably occurs with effective chemotherapy (my bias: probably does) Insensitive for discriminating between two drugs or regimens
  • Slide 31 - HRQoL Assessment:Conclusions (cont.) Improvements in QoL potentially multifactorial (tumor shrinkage, decrease in symptoms, joy at “fighting cancer”) and tools are unable to discriminate reason for benefit Worsening in QoL potentially multifactorial (progression, toxicity, concomitant divorce) Cannot adequately separate efficacy and safety
  • Slide 32 - Clinical Benefit Response: Conclusions The most promising, but needs significant tweaking before useful Must account for all major common symptoms in CRC patients Does not allow assessment of drug in asymptomatic patients (unless we use a “time-to-onset-of-tumor-related -symptoms”, which does not have good supportive data) Would likely be a “stand-alone” factor (i.e., of significant patient advantage by itself), rather than a correlate of other benefit (survival) FDA should query specific QoL experts if found worthy of future exploration (e.g., David Cella)
  • Slide 33 - DISCUSSION

Description : Available OHSU Cancer powerpoint presentation for free download which is uploaded by search an active user in belonging ppt presentation Health & Wellness category.

Tags : OHSU Cancer

Shortcode : Get Shareable link