Background: Nilotinib induced significantly faster and deeper molecular responses vs imatinib in the ENESTnd trial. Achieving these deeper molecular responses may increase patient eligibility for future TKI discontinuation studies. Methods: CML-CP pts (N = 207) who achieved a complete cytogenetic response but were still BCR-ABL positive by RQ-PCR after ≥ 24 mo on imatinib were randomized 1:1 to receive nilotinib 400 mg BID (n = 104) or to continue their imatinib dose (400 or 600 mg QD [n = 103]). The primary endpoint was confirmed CMR (undetectable BCR-ABL by RQ-PCR with a sample sensitivity of ≥ 4.5 logs in 2 consecutive samples). Other endpoints included molecular responses (MMR ≤ 0.1%^IS, MR⁴ ≤ 0.01%^IS, and MR^4.5 ≤ 0.0032%^IS) and BCR-ABL ratio over time. Results: Rate of confirmed CMR was higher in the nilotinib arm vs imatinib by 12 mo (12.5% vs 5.8%) (Table). Rate of CMR (undetectable BCR-ABL in at least 1 sample) by 12 mo was significantly higher on nilotinib vs imatinib (23.1% vs 10.7%; P = .02). Rates of MMR, MR⁴, MR^4.5, and CMR were also superior in pts switched to nilotinib, and these pts had significantly shorter times to achieve these responses. Imatinib-treated pts had minimal evidence of improvement in molecular response vs a median 0.5-log reduction in BCR-ABL by 12 mo for the nilotinib cohort. With 12-mo follow-up, 84% of pts remained on nilotinib and 96% on imatinib. The nilotinib safety profile was consistent with prior studies. Both drugs were well tolerated. Conclusions: Twice as many pts achieved deeper molecular responses after switching to nilotinib vs staying on imatinib.

	Nilotinib 400 mg BID (n = 104)	Imatinib 400 or 600 mg QD (n = 103)
Molecular response by 12 mo (ITT population), %
Confirmed CMR	12.5 P = .108*	5.8
CMR	23.1 P = .02*	10.7
Molecular response by 12 mo (in pts without the response at baseline), %
MMR	n = 24 75.0 P = .006**	n = 28 35.7
MR4	n = 74 48.6 P = .006**	n = 78 25.6
MR4.5	n = 94 33.0 P = .008**	n = 91 16.5
CMR	n = 101 20.8 P = .03**	n = 100 10.0

*Stratified Cochran-Mantel-Haenszel test. **Stratified log-rank test.