Background: Epigenomes allow the rectification of disordered gene expressions, thereby providing new targets for pharmacological interventions in cancer therapy. The clinical utility of targeting histone H3 lysine tri-methylation (H3K27me3) as an epigenetic hallmark has been demonstrated. However, in actual therapeutic settings, the mechanism by which H3K27me3-targeting therapies exert their effects and the response of tumor cells remain unclear. Results and Discussion: We conducted intensive molecular monitoring of three patients enrolled in the valemetostat first-in-human phase 1 clinical trial (NCT02732275) for adult T-cell leukemia/lymphoma (ATL). Valemetostat was administered orally once daily (200 mg/day) until a sign of disease progression could be observed. The number of abnormal lymphocytes (Ably) in these three patients drastically decreased after one week of treatment. All patients could stay on valemetostat single agent for more than 2 years with acceptable safety and durable response. PBMC samples (n = 55) were collected from these three patients immediately before (Pre) and during treatment, as well as during the progressive disease (PD) stage. Flow cytometry-based histone methylation assay demonstrated that tumor H3K27me3 level was generally high in the Pre-samples, but the valemetostat treatment reduced it to a normal level in all patients. In addition, ChIP sequencing (ChIP-seq) could directly assess how valemetostat affects the tumor epigenome. The pre-treatment tumor cells showed an overall increasing trend in H3K27me3. For all H3K27me3 peaks (51,683 peaks), 85.9% of the peak signals decreased at the average fold-change of 0.742 after the valemetostat treatment. Focusing on the gene loci, an overall decrease could be observed around the TSS and across the gene body. The characteristic H3K27me3 clusters initially observed in the tumor cells were demethylated. We also conducted integrative multilayered single-cell analyses (scATAC-seq and scRNA-seq) and showed that valemetostat abolishes the highly condensed chromatin structure formed by the plastic H3K27me3 and neutralizes multiple gene loci, such as tumor suppressor genes (TSGs), in all patients. These results demonstrated that valemetostat sufficiently restored the epigenome of the tumor cells close to the healthy state, leading to clinical improvements. Although all tested cases showed a durable response for more than 2 years, these responses were eventually interrupted due to the recurrence of ATL. We identified characteristic somatic mutations of EZH2 during clonal repopulation at PD. The identified amino acid substitutions were all located at the interface between EZH2 and valemetostat. Direct evaluation of the H3K27me3 level in the PD sample harboring the acquired EZH2 mutation showed that tumor cells in a hypermethylated state were repopulating. The scATAC-seq analysis also showed strong chromatin compaction of the same gene set occurred in the rapidly propagating resistant clone, indicating that the clonally selected mutations should be responsible for resistance. In patients free of the EZH2 resistant mutations, we detected TET2 mutation or elevated DNMT3A expression in the propagated resistant clones. Genome-wide DNA methylation analysis revealed increased DNA methylation near the TSS in the resistant clones. The genes characterized by the H3K27me3-related chromatin condensation were restored by valemetostat treatment but were again strongly re-condensed by mCpG gain at PD. Consequently, the gene regulation of several important TSGs restored by valemetostat was repressed again. Furthermore, in the single-cell platform, we successfully identified subpopulations with distinct metabolic and gene translation characteristics implicated in primary susceptibility until the acquisition of the heritable (epi)mutations. Conclusion: This study illustrates, for the first time, the molecular and cellular dynamics in patients in response to a molecular-targeting inhibitor designed for histone methyltransferases. Eliminating H3K27me3 leads to the reprogramming of the cancer epigenome, thereby exerting a sustained benefit on tumors. Targeting chromatin homeostasis might provide opportunities for further sustained epigenetic therapies.