The latest European LeukemiaNet (ELN) recommendations highlight that maintaining treatment-free remission (TFR) and discontinuation of treatment is particularly crucial for patients < 60 years. Achieving a deep molecular response (DMR) is a prerequisite for attaining TFR. Imatinib, the first-generation tyrosine kinase inhibitor (TKI), has become the standard first-line therapy due to its ability to achieve major molecular responses and its established long-term survival benefits. Research suggests that second-generation TKIs may facilitate achieving DMR more readily; however, their toxicity remains a significant concern compared to imatinib. The characteristics of imatinib and second-generation TKIs warrant reconsideration of treatment optimization for patients after initial imatinib. Therefore, we aimed to develop a prognostic model for DMR with risk stratification to guide decisions on whether to switch or continue first-line imatinib.

This retrospective multicenter analysis enrolled 925 newly diagnosed CML-CP patients from 44 centers receiving initial imatinib therapy. A nomogram incorporating variables identified by Cox proportional hazards models for predicting DMR at 5 years was constructed and validated. Moreover, the cut-off point for risk stratification was selected via the “Surv_cutpoint” function from survival package in R.

The median age at diagnosis of the population was 49 years (IQR: 37.00-61.00), and 59.8% were male. Over a median follow-up period of 19 months (IQR: 11.00-41.00), there were 539 (58.27%) cases of MR4.5 observed. Individuals who achieved MR4.5 had a lower frequency of male and exhibited lower levels of white blood cell (WBC) and blood blasts. They also had higher hemoglobin (HGB), a higher incidence of myelofibrosis (MF) and 3-month early molecular response (EMR) compared to those without MR4.5. The proportion of non-MF patients was significantly higher in MR4.5 patients (81.26% vs 63.99%, p < 0.001).

Subgroup analyses were conducted to assess the predictive value of MF across diverse groups. There was no significant interaction between MF and stratified variables. Intriguingly, an inverse association between MF and MR4.5 was observed only to be statistically significant in age ≤ 60 years group after adjusting for confounding factors (adjusted HR = 0.55, 95% CI: 0.40-0.75, p < 0.001).

Cox analyses revealed that MF, WBC, HGB, PLT, and 3-month EMR were independent predictors in the training cohort. Subsequently, these variables were applied to construct a nomogram for predicting MR4.5-free survival at 5 years, and an example of applying this nomogram to clinical scenarios was given. We further made a risk stratification based on total points calculated from the nomogram. Patients could be divided into two risk groups: high risk (total points < 156) and low risk (total points ≥156), and the K-M curves demonstrated significant differences among the two risk groups (p < 0.001). Particular attention should be paid to patients with total points < 156, as they should be considered for switching to a second-line therapy rather than continuing imatinib therapy.

The ROC curves in both the training and validation sets for predicting MR4.5 at 5 years indicated a good discrimination of the nomogram model (AUC of the training cohort: 0.72 (0.66-0.79), AUC of the validation cohort: 0.74 (0.64-0.85)). Furthermore, the calibration curves in both the training and validation cohorts confirmed that the predicted probabilities closely matched the actual outcomes. Additionally, DCA illustrated the clinical net benefit achievable at different risk thresholds showed that the net benefit is superior to intervening on all patients or not intervening on all patients.

In conclusion, in this retrospective cohort study involving 925 CML individuals followed up to 13 years, we developed an intuitive visual model for evaluating the DMR probability at 5 years to assist clinicians in deciding whether to switch from imatinib to second-line therapies.

Acknowledgement: This research was funded by the Key R&D Program of Zhejiang (No. 2022C03137) and the Zhejiang Medical Association Clinical Medical Research Special Fund Project (No. 2022ZYC-D09).

*Correspondence to: Jian Huang, M.D., Ph.D., Department of Hematology, The First Affiliated Hospital of Zhejiang University School of Medicine, Hangzhou 310003, Zhejiang, China. E-mail: househuang@zju.edu.cn

Disclosures

No relevant conflicts of interest to declare.

This content is only available as a PDF.
2024
Sign in via your Institution