A Prospective Study of Whole Genome Sequencing (ChromoSeq) as an Adjunct to Conventional Genomic Profiling in AML and MDS

Sensitivity of ChromoSeq as measured by total number of recurrent structural variants identified

Time Frame: Through completion of all ChromoSeq tests (estimated to be 15 months)

* As compared to conventional cytogenetics in a real-time clinical setting * The total number of recurrent structural variants will be measured in each sample by ChromoSeq and metaphase cytogenetics yielding a pair of measurements. Each measurement will also be dichotomized into the presence or absence of at least one recurrent structural variant. The hypothesis of no difference in the number of variants detected by each method will be analyzed by a paired-sample t-test. However, if it is determined that the assumptions of a t-test are not tenable then a paired-sample sign test will be used instead. McNemar's test will be used to compare whether or not at least one recurrent structural variant identified is by each method.

Sensitivity of ChromoSeq as measured by total number of copy number alterations identified

Time Frame: Through completion of all ChromoSeq tests (estimated to be 15 months)

* As compared to conventional cytogenetics in a real-time clinical setting * The total number of copy number alterations will be measured in each sample by ChromoSeq and metaphase cytogenetics yielding a pair of measurements. Each measurement will also be dichotomized into the presence or absence of at least one copy number alteration. The hypothesis of no difference in the number of copy number alterations detected by each method will be analyzed by a paired-sample t-test. However, if it is determined that the assumptions of a t-test are not tenable then a paired-sample sign test will be used instead. McNemar's test will be used to compare whether or not at least one copy number alterations is identified is by each method.

Sensitivity of ChromoSeq as measured by number of single nucleotide variants identified

Time Frame: Through completion of all ChromoSeq tests (estimated to be 15 months)

* As compared to high coverage gene panels in a real-time clinical setting * The number of single nucleotide variants will be counted for each sample. Additionally, the data will be dichotomized into the presence or absence of at least one single nucleotide variant. Data will be analyzed by paired-sample t-tests and McNemar's test.

Sensitivity of ChromoSeq as measured by number of insertion-deletions identified

Time Frame: Through completion of all ChromoSeq tests (estimated to be 15 months)

* As compared to high coverage gene panels in a real-time clinical setting * The number of insertion-deletions will be counted for each sample. Additionally, the data will be dichotomized into the presence or absence of at least one insertion-deletion. Data will be analyzed by paired-sample t-tests and McNemar's test.

Determine if risk-stratification using ChromoSeq correlates with overall-survival

Time Frame: Through completion of follow-up for all patients (estimated to be 63 months)

* As compared to metaphase cytogenetics * The relationship of risk-stratification defined by either ChromoSeq or conventional cytogenetics to clinical outcome will be illustrated with Kaplan-Meier survival analyses on overall survival for both ChromoSeq and metaphase cytogenetics. The predictive accuracy of the two methods will be tested by comparing the area under the ROC curves using the method of DeLong et al.

Determine if risk-stratification using ChromoSeq correlates with event-free survival

Time Frame: Through completion of follow-up for all patients (estimated to be 63 months)

* As compared to metaphase cytogenetics * The relationship of risk-stratification defined by either ChromoSeq or conventional cytogenetics to clinical outcome will be illustrated with Kaplan-Meier survival analyses on event-free survival for both ChromoSeq and metaphase cytogenetics. The predictive accuracy of the two methods will be tested by comparing the area under the ROC curves using the method of DeLong et al.

Proportion of cases in which ChromoSeq provides new genetic information to the clinician

Time Frame: Through completion of all ChromoSeq tests (estimated to be 15 months)

* As compared to conventional genomic profiling (cytogenetics, FISH, and next-generation sequencing) that is used for clinical management (such as risk-stratification or institution of targeted gene therapy) * Items in the ChromoSeq Implementation Physician Survey will be used to describe physician evaluation of ChromoSeq with conventional genomic profiling with regard to clinical management. Responses to these items will be presented in frequency tables. For statistical analysis, the values of each item will be recoded from 1-5 to -2 to +2 and one-sample t-tests used to test the null hypothesis that the mean value is 0 (neither agree nor disagree.) In addition, case-reports will be reviewed for qualitative evaluations of physician experience with the two methods.

ChromoSeq turnaround time

Time Frame: Through completion of all ChromoSeq tests (estimated to be 15 months)

-Measured from time of order requisition (hematologic molecular algorithm from Barnes Jewish Hospital) to return of report to the medical record

Proportion of failed ChromoSeq assays

Time Frame: Through completion of all ChromoSeq tests (estimated to be 15 months)

* As compared to failed standard of care genomic profiling assays * Each assay will be categorized as successful or failed and a two-way table constructed displaying ChromoSeq assay status by standard assay status. A Pearson chi-square test will be calculated to test the null hypothesis that assay success is independent of type of assay.