Sure - after it defines them explicitly to be patients "meeting BM and peripheral blood morphologic criteria for CR". An utter waste of text if they were complete CRs since it would be a given that true CRs "met BM and peripheral blood morphologic criteria".
But we shall see what we shall see when they release the fuller dataset. For instance I'll bet that the number of patients meeting "Clinical Criteria" for CI is the same as the number of CIs (3 in the abstract). Not 8 (all true CRs and all true PRs must meet something better than CI "Clinical Criteria" - so if you were right there would be 4 "Clinical criteria" patients from the CRs, 1 from the PR and 3 CIs).