Clinical Trials of Predictive Medicine New Challenges and Paradigms - PowerPoint PPT Presentation

Clinical Trials of Predictive MedicineNew Challenges and Paradigms

Richard Simon, D.Sc.

Chief, Biometric Research Branch

National Cancer Institute

http://brb.nci.nih.gov

• Powerpoint presentations
• Reprints
• BRB-ArrayTools software
• Data archive
• Web based Sample Size Planning
• Clinical Trials
• Development of gene expression based predictive classifiers

• Most cancer patients don’t benefit from the systemic treatments they receive
• Being able to predict which patients are likely to benefit would
• Benefit patients
• Control medical costs
• Improve the success rate of clinical drug development

• Measured before treatment to identify who will or will not benefit from a particular treatment
• ER, HER2, KRAS
• Prognostic biomarkers
• Measured before treatment to indicate long-term outcome for patients untreated or receiving standard treatment
• Used to identify who does not require more intensive treatment
• OncotypeDx

• Single gene or protein measurement
• ER protein expression
• HER2 amplification
• KRAS mutation
• Index or classifier that summarizes expression levels of multiple genes
• OncotypeDx recurrence score

• They are developed in unfocused studies not designed to address an intended medical use
• The studies are based on convenience samples of heterogeneous patients for whom tissue is available
• Although they correlate with a clinical endpoint, they have no demonstrated medical utility
• They are not “actionable”

• Analytical validation
• Accuracy compared to gold-standard assay
• Robust and reproducible if there is no gold-standard
• Clinical validation
• Does the biomarker predict what it’s supposed to predict for independent data
• Clinical/Medical utility
• Does use of the biomarker result in patient benefit
• Is it actionable?
• Generally by improving treatment decisions

• Cancers of a primary site are generally composed of a heterogeneous group of diverse molecular diseases
• The molecular diseases vary fundamentally with regard to the oncogenic mutations that cause them, and in their responsiveness to specific drugs

• Based on assumptions that
• Qualitative treatment by subset interactions are unlikely
• “Costs” of over-treatment are less than “costs” of under-treatment
• Have led to widespread over-treatment of patients with drugs to which few benefit

• In the past often studied as exploratory post-hoc subset analyses of RCTs.
• Numerous subsets examined
• Not focused pre-specified hypothesis
• No control of type I error

How Can We Develop New Drugs in a Manner More Consistent With Modern Tumor Biology and Obtain Reliable Information About What Regimens Work for What Kind of Tumors?

• Develop a completely specified genomic classifier of the patients likely to benefit from a new drug
• Establish analytical validity of the classifier
• Use the completely specified classifier to design and analyze a new clinical trial to evaluate effectiveness of the new treatment and how it relates to the classifier

• The data used to develop the classifier must be distinct from the data used to test hypotheses about treatment effect in subsets determined by the classifier
• Developmental studies are exploratory
• Studies on which treatment effectiveness claims are to be based should be definitive studies that test a treatment hypothesis in a patient population completely pre-specified by the classifier

• Restrict entry to the phase III trial based on the binary predictive classifier

Develop Predictor of Response to New Drug

Using phase II data, develop predictor of response to new drug

Patient Predicted Responsive

Patient Predicted Non-Responsive

Off Study

New Drug

Control

• Primarily for settings where the classifier is based on a single gene whose protein product is the target of the drug
• eg trastuzumab
• With a strong biological basis for the classifier, it may be unacceptable to expose classifier negative patients to the new drug
• Analytical validation, biological rationale and phase II data provide basis for regulatory approval of the test

• Simon R and Maitnourim A. Evaluating the efficiency of targeted designs for randomized clinical trials. Clinical Cancer Research 10:6759-63, 2004; Correction and supplement 12:3229, 2006
• Maitnourim A and Simon R. On the efficiency of targeted clinical trials. Statistics in Medicine 24:329-339, 2005.
• reprints and interactive sample size calculations at http://linus.nci.nih.gov

• proportion of patients test positive
• effectiveness of new drug (compared to control) for test negative patients
• When less than half of patients are test positive and the drug has little or no benefit for test negative patients, the targeted design requires dramatically fewer randomized patients

• Metastatic breast cancer
• 234 randomized patients per arm
• 90% power for 13.5% improvement in 1-year survival over 67% baseline at 2-sided .05 level
• If benefit were limited to the 25% assay + patients, overall improvement in survival would have been 3.375%
• 4025 patients/arm would have been required

• http://brb.nci.nih.gov

Develop Predictor of

Response to New Rx

Predicted Responsive

To New Rx

Predicted Non-responsive to New Rx

New RX

Control

New RX

Control

• Having a prospective analysis plan is essential
• “Stratifying” (balancing) the randomization is useful to ensure that all randomized patients have tissue available but is not a substitute for a prospective analysis plan
• The purpose of the study is to evaluate the new treatment overall and for the pre-defined subsets; not to modify or refine the classifier
• The purpose is not to demonstrate that repeating the classifier development process on independent data results in the same classifier

• Compare the new drug to the control overall for all patients ignoring the classifier.
• If poverall 0.03 claim effectiveness for the eligible population as a whole
• Otherwise perform a single subset analysis evaluating the new drug in the classifier + patients
• If psubset 0.02 claim effectiveness for the classifier + patients.

• Test for difference (interaction) between treatment effect in test positive patients and treatment effect in test negative patients
• If interaction is significant at level int then compare treatments separately for test positive patients and test negative patients
• Otherwise, compare treatments overall

• 88 events in test + patients needed to detect 50% reduction in hazard at 5% two-sided significance level with 90% power
• If 25% of patients are positive, when there are 88 events in positive patients there will be about 264 events in negative patients
• 264 events provides 90% power for detecting 33% reduction in hazard at 5% two-sided significance level

• Using int=0.10, the interaction test has power 93.7% when there is a 50% reduction in hazard in test positive patients and no treatment effect in test negative patients
• A significant interaction and significant treatment effect in test positive patients is obtained in 88% of cases under the above conditions
• If the treatment reduces hazard by 33% uniformly, the interaction test is negative and the overall test is significant in 87% of cases

• No
• It is incorrect to require that the overall T vs C comparison be significant to claim that T is better than C for test + patients but not for test – patients
• That requirement has been traditionally used to protect against data dredging. It is inappropriate for focused trials of a treatment with a companion test.

Biomarker Adaptive Threshold Design

Wenyu Jiang, Boris Freidlin & Richard Simon

JNCI 99:1036-43, 2007

• Randomized trial of T vs C
• Previously identified a biomarker score B thought to be predictive of patients likely to benefit from T relative to C
• Eligibility not restricted by biomarker
• No threshold for biomarker determined
• Time-to-event data

• Compare T vs C for all patients
• If results are significant at level .04 claim broad effectiveness of T
• Otherwise proceed as follows

• Test T vs C restricted to patients with biomarker B > b
• Let S(b) be log likelihood ratio statistic
• Repeat for all values of b
• Let S* = max{S(b)}
• Compute null distribution of S* by permuting treatment labels
• If the data value of S* is significant at 0.01 level, then claim effectiveness of T for a patient subset
• Compute point and bootstrap interval estimates of the threshold b

• Have identified K candidate predictive binary classifiers B1 , …, BK thought to be predictive of patients likely to benefit from T relative to C
• Eligibility not restricted by candidate classifiers

• If results are significant at level .04 claim broad effectiveness of T
• Otherwise proceed as follows

• Let S(Bk) be log likelihood ratio statistic for treatment effect in patients positive for Bk (k=1,…,K)
• Let S* = max{S(Bk)} , k* = argmax{S(Bk)}
• Compute null distribution of S* by permuting treatment labels
• If the data value of S* is significant at 0.01 level, then claim effectiveness of T for patients positive for Bk*

Adaptive Signature Design

Boris Freidlin and Richard Simon

Clinical Cancer Research 11:7872-8, 2005

• Compare E to C for all patients at significance level 0.04
• If overall H0 is rejected, then claim effectiveness of E for eligible patients
• Otherwise

• Using only the first half of patients accrued during the trial, develop a binary classifier that predicts the subset of patients most likely to benefit from the new treatment T compared to control C
• Compare T to C for patients accrued in second stage who are predicted responsive to T based on classifier
• Perform test at significance level 0.01
• If H0 is rejected, claim effectiveness of T for subset defined by classifier

Cross-Validated Adaptive Signature Design(to be submitted for publication)

Wenyu Jiang, Boris Freidlin, Richard Simon

• Compare T to C for all patients at significance level overall
• If overall H0 is rejected, then claim effectiveness of T for eligible patients
• Otherwise

• Partition the full data set into K parts
• Form a training set by omitting one of the K parts. The omitted part is the test set
• Using the training set, develop a predictive classifier of the subset of patients who benefit preferentially from the new treatment T compared to control C using the methods developed for the ASD
• Classify the patients in the test set as sensitive (classifier +) or insensitive (classifier -)
• Repeat this procedure K times, leaving out a different part each time
• After this is completed, all patients in the full dataset are classified as sensitive or insensitive

• Generate the null distribution of S by permuting the treatment labels and repeating the entire K-fold cross-validation procedure
• Perform test at significance level 0.05 - overall
• If H0 is rejected, claim effectiveness of T for subset defined by classifier
• The sensitive subset is determined by developing a classifier using the full dataset

• No
• Stratification improves balance of stratification factors in overall comparisons
• Stratification does not improve comparability of treatment (T) and control (C) groups within test positive patients or within test negative patients.
• In a fully prospective trial, stratification of the randomization by the test is only useful for ensuring that all patients have adequate test performed

• This has led to many exciting methodological developments
• p>>n problems in which number of genes is much greater than the number of cases
• Statistics has over-focused on inference. Many of the methods and much of the conventional wisdom of statistics are based on inference problems and not applicable to prediction problems

• Feature selection should be optimized for accurate prediction, not for controlling the false discovery rate
• Standard statistical methods for model building and evaluation are not effective
• e.g. Fisher’s LDA vs diagonal LDA
• Model performance on the training set is extremely misleading for p>n problems and should never be reported

• Statistical significance of regression coefficients or of the model is not a proper measure of predictive accuracy

• Validation does not mean that the model is stable or that using the same algorithm on independent data will give a similar model
• Validation of model prediction does not indicate that the model has medical utility for any intended use

• Using cross-validation we can evaluate new methods for analysis of clinical trials in terms of their intended use which is informing therapeutic decision making

• New biotechnology and knowledge of tumor biology provide important opportunities to improve therapeutic decision making
• Treatment of broad populations with regimens that do not benefit most patients is increasingly no longer necessary nor economically sustainable
• The established molecular heterogeneity of human diseases requires the use new approaches to the development and evaluation of therapeutics

• Make sure that they are solving the right problem
• Focus on the big picture
• Prepare themselves to be full partners with their collaborators

• Boris Freidlin
• Yingdong Zhao
• Aboubakar Maitournam
• Wenyu Jiang