The impact of companion diagnostic device measurement performance on clinical validation of personalized medicine

Abstract

A key component of personalized medicine is companion diagnostics that measure biomarkers, for example, protein expression, gene amplification or specific mutations. Most of the recent attention concerning molecular cancer diagnostics has been focused on the biomarkers of response to therapy, such as V-Ki-ras2 Kirsten rat sarcoma viral oncogene homolog (KRAS) mutations in metastatic colorectal cancer, epidermal growth factor receptor mutations in metastatic malignant melanoma. The presence or absence of these markers is directly linked to the response rates of particular targeted therapies with small-molecule kinase inhibitors or antibodies. Therefore, testing for these markers has become a critical step in the target therapy of the aforementioned tumors. The core capability of personalized medicine is the companion diagnostic devices' (CDx) ability to accurately and precisely stratify patients by their likelihood of benefit (or harm) from a particular therapy. There is no reference in the literature discussing the impact of device's measurement performance, for example, analytical accuracy and precision on treatment effects, variances, and sample sizes of clinical trial for the personalized medicine. In this paper, using both analytical and estimation method, we assessed the impact of CDx measurement performance as a function of positive and negative predictive values and imprecision (standard deviation) on treatment effects, variances of clinical outcome, and sample sizes for the clinical trials.

Precision Medicine - The Golden Gate for Detection, Treatment and Prevention of Alzheimer’s Disease

“…The international blood-based biomarker working group has recently generated the first-ever pre-analytical guidelines for blood-based biomarkers in AD (108); however, CDx introduce the need to fully understand the analytical accuracy and precision on treatment effects, variances, and other aspects of the device performance itself (109). A tremendous effort has been undertaken to move CSF-based AD biomarkers from RUO towards LDTs and IVDs, in recent years.…”

Assessing treatment efficacy in the presence of diagnostic errors

“…Example 1 To illustrate the results described previously, we use a generated data from an example and the related information provided in [14]. The example concerns a randomized, double-blind, controlled clinical trial performed with the goal of evaluating the efficacy of a new drug targeting Alzheimer's disease (AD).…”

“…After a 2-year treatment period, the outcomes measured using a reduction in the Clinical Dementia Rating Scale Sum of Boxes (CDR-SOB) score are compared. The following numerical values were suggested in [14]: the mean reduction of CDR-SOB score for ApoE positive patients are 1.7 and 0.9 for treatment and control arms, respectively, and the mean reduction of CDR-SOB score for ApoE negative patients are 1.01 and 0.75 for treatment and control arms, respectively; variances of the clinical outcome are 2 11 = 1.1, 2 01 = 1.25. The positive predictive value 1 − = 0.9 and negative predictive value 1 − = 0.98 were used for the PCR assay.…”

“…A remotely related work is that of Lin, McClintock, and Williamson, which proposes a sample size adjustment in the situations where the group membership cannot be ascertained until after the outcome is measured; they define a quantity called expected power to adjust sample size in an ad hoc manner [10]. With the exception of the work presented in [11][12][13][14] not much has been performed in terms of evaluating clinical validity in the presence of diagnostic or screening misclassification. The paper [12] discusses continuous outcomes in the framework of enriched randomized placebo-controlled trials -a trial design that focuses solely on the test positive strata.…”

“…Whereas, the papers [11] and [13] are concerned with binary and censored outcomes, respectively. Most recent work [14] addressed the impact of companion diagnostic device performance on the clinical validation of personalized medicine. Their methodological development is limited to unpaired/independent group analysis and depends on the assumption that the true values of the model parameters are known.…”