Cytogenetic Response (CyR) is based on the prevalence of Philadelphia chromosome positive (Ph+) metaphases among cells in metaphase on a bone marrow (BM) aspirate. CCyR was achieved when there was 0 percent (%) Ph+ metaphases among cells in a BM sample when at least 20 metaphases from a BM sample were analyzed, or less than (<) 1% breakpoint cluster region Abelson protooncogene (Bcr-Abl) fusion product among cells in a BM sample or peripheral blood sample when at least 200 cells were analyzed.

Intent-to-treat (ITT) population - included all participants who were randomized to test article.

Molecular response was assessed using Bcr-Abl transcript levels measured by reverse transcriptase polymerase chain reaction (RT-PCR) from peripheral blood. A MMR was defined as a ratio Bcr-Abl/Abl less than or equal to (≤) 0.1% on the international scale (greater than or equal to [≥] 3 log reduction from standardized baseline in ratio of Bcr-Abl to Abl transcripts) with at least 3000 Abl analyzed.

ITT Population

The Kaplan-Meier curve was generated based the time from the first date of CCyR until the first date of confirmed loss of CCyR, objectively documented, for responders only. Participants without confirmed loss of CCyR were censored at the last valid cytogenetic assessment.

CyR is based on the prevalence of Ph+ metaphases among cells in metaphase on a bone marrow sample. CCyR was achieved when there was 0% Ph+ metaphases among cells in a BM sample when at least 20 metaphases from a BM sample were analyzed, or <1% Bcr-Abl fusion product among cells in a BM sample or peripheral blood sample when at least 200 cells were analyzed.

The medians have not been reached in either arm, as such, the premature estimated hazard ratio is provided. Four years rate was displayed since the majority of participants had first CCyR by Year 1.

Subgroup of participants from ITT population who had CCyR.

The Kaplan-Meier curve was generated based on the first date of confirmed CHR until the first date of loss of CHR, objectively documented, for responders only. Participants without confirmed loss of response were censored at the last valid hematologic assessment.

CHR must have been of at least 4 weeks in duration confirmed by 2 assessments at least 4 weeks apart and was defined as follows: white blood cells ≤ institutional upper limit of normal, no peripheral blasts or promyelocytes, myelocytes + metamyelocytes <5% in blood, absolute neutrophil count ≥1.0*10^9/L, platelets ≥100 but <450*10^9/L unless related to therapy, <20% basophils in blood and no extramedually involvement (including hepato- or splenomegaly).

The medians have not been reached in either arm, as such, the premature estimated hazard ratio is provided. Four years rate was displayed since the majority of participants had first CHR by Year 1.

Subgroup of participants from ITT population who had CHR.

The Kaplan-Meier curve was generated based on the first date of MMR until the first date loss of MMR, objectively documented, for responders only. Participants without confirmed loss of response were censored at the last valid molecular assessment.

Molecular response was assessed using Bcr-Abl transcript levels measured by RT-PCR from peripheral blood. MMR is defined as a ratio Bcr-Abl/Abl ≤0.1% on the international scale (≥3 log reduction from standardized baseline in ratio of Bcr-Abl to Abl transcripts) with at least 3000 Abl analyzed.

The medians have not been reached in either arm, as such, the premature estimated hazard ratio is provided. Three years rate was displayed since the majority of imatinib participants had first MMR by Year 2.

Subgroup of participants from ITT population who had MMR.

The cumulative incidence curve was generated based on the time from randomization to the first date of transformation to AP or BP while on study treatment adjusting for the competing risk of treatment discontinuation without transformation, for each participant.

Criteria for transformation to AP: 15 to 29% blasts; ≥30% blasts + promyelocytes; ≥20% basophils in blood or bone marrow; platelets <100*10^9/L (not related to therapy), in blood. Criteria for transformation to BP: ≥30% blasts in blood or bone marrow and extramedullary involvement other than liver or spleen (example: chloromas).

Time to transformation was calculated as weeks = ([date of first documented occurrence of the event - date of randomization] + 1)/7. If transformation was not obtained, censoring was at the last hematologic assessment or death (whichever was earliest). Participants who were not treated contributed time = 1 day/7. 95% confidence interval for the cumulative incidence is from Gray's method.

ITT population