Real-world comparison of flumatinib and nilotinib as first-line therapy for patients with chronic phase chronic myeloid leukemia: a multicenter retrospective study

Abstract

Background:

Flumatinib is a novel second-generation tyrosine kinase inhibitor (2G-TKI), which was approved in November 2019 in China. A previous phase III study evaluated the efficacy and safety of flumatinib as a first-line therapy for patients with chronic phase chronic myeloid leukemia (CML-CP). However, randomized trials comparing flumatinib with other 2G-TKIs remain lacking.

Objectives:

To assess the efficacy and safety of flumatinib versus nilotinib as a first-line treatment for CML-CP.

Design:

A multicenter retrospective study.

Methods:

We retrospectively analyzed 101 and 64 patients treated with flumatinib and nilotinib during the same period, respectively.

Results:

Patients in the flumatinib group were significantly older than patients in the nilotinib group (median age, 44 vs 37 years; p = 0.004). The optimal response and treatment failure rates at 24 months were comparable between the two groups. At 12 months, 85.1% and 88.2% of patients in the flumatinib and nilotinib groups, respectively, achieved a major molecular response (MMR; p = 0.648). By 24 months, 9.9% and 12.5% of patients suffered treatment failure in the flumatinib and nilotinib groups, respectively (p = 0.602). At 9, 12, and 24 months, the rate of MR⁴ (a BCR::ABL1 transcript level ⩽0.01%) achievement was significantly higher in patients treated with nilotinib than in those treated with flumatinib (26.0% vs 53.7%, p = 0.007; 40.4% vs 60.8%, p = 0.044; and 41.7% vs 80.8%, p = 0.042, respectively). In addition, elevated alanine aminotransferase or aspartate aminotransferase (ALT/AST), glucose, and serum lipid; hyperbilirubinemia; rash; and alopecia were more frequent among patients receiving nilotinib, whereas diarrhea was more frequent in those receiving flumatinib.

Conclusion:

Flumatinib is a suitable alternative as a first-line treatment for patients with CML-CP to achieve a fast MMR with better tolerability.

Introduction

Chronic myeloid leukemia (CML) is a myeloproliferative neoplasm characterized by the BCR::ABL1 fusion gene. With the widespread use of tyrosine kinase inhibitors (TKIs), CML has become a manageable and potentially curable disease.^1–3 Furthermore, CML treatment has become personalized, aimed at restoring social function to the greatest extent. ⁴ Over the past decade, both clinical trials and real-world data have confirmed the efficacy of second-generation TKIs (2G-TKIs) over imatinib as first-line therapy, with acceptable adverse events (AEs).^5–8 Although imatinib is generally well tolerated, rare but serious adverse effects have been reported in patients. ⁹ Therefore, continuous monitoring and the development of alternative treatment options are crucial. As frontline therapy, 2G-TKIs show no survival benefits over imatinib but achieve faster and deeper molecular responses.^10,11 Nilotinib is the first 2G-TKI approved by the Food and Drug Administration (FDA) in 2010 for newly diagnosed adults with chronic phase (CP) CML. Notably, dasatinib was approved earlier as a second-line treatment for CML. The ENESTnd trial reported higher cumulative major molecular response (MMR) and deep molecular response (DMR) with nilotinib than with imatinib over 5 and 10 years, resulting in a higher incidence of treatment-free remission (TFR).^6,12 Second-generation TKIs have distinct safety profiles. Nilotinib is associated with metabolic complications (elevated glucose and lipid levels) and cardiovascular risk, whereas dasatinib is associated with pleural effusion and pulmonary hypertension.^5,11 These characteristic AEs influence treatment selection in clinical practice.

Currently, 2G-TKIs, including nilotinib, flumatinib, and dasatinib, are available in the Chinese mainland market. Flumatinib is a novel Chinese 2G-TKI, which was recently launched locally in China but has not yet been reviewed by the FDA or EMA. Similar to nilotinib, flumatinib is a derivative of imatinib, which targets the inactivated conformations of the ABL1 kinase domain. ¹³ FESTnd, a phase III study that randomized patients with newly diagnosed CML-CP into flumatinib and imatinib groups, reported the advantageous effects of flumatinib over imatinib, with grade 1 or 2 AEs in both groups. ¹⁴ Based on the efficacy and safety results of the FESTnd trial, flumatinib was approved as a treatment for CML-CP by the National Medical Products Administration (NMPA) in 2019 and is widely used as a first-line or salvage treatment for CML in China. With the short follow-up duration of flumatinib, data on its efficacy and safety in the real world are limited, and head-to-head comparison with other 2G-TKIs remains unavailable. Therefore, this study aimed to compare the real-world efficacy and safety profiles of flumatinib and nilotinib as first-line treatments for CML-CP in Chinese patients. Specifically, we evaluated and compared the molecular responses (including MMR and DMR) achieved with these two agents. We also analyzed treatment-related AEs and treatment discontinuation or drug switching due to side effects. To address these objectives, we retrospectively collected and analyzed the outcomes of the first-line treatment with flumatinib and nilotinib in patients with CML-CP from two Chinese medical centers.

Materials and methods

Patients

Data were collected from the First Affiliated Hospital of Nanjing Medical University and the Affiliated Union Hospital of Tongji Medical College. In total, 175 patients newly diagnosed with CML-CP were initially screened for eligibility at two participating medical centers between December 2019 and April 2023. Figure 1 illustrates the patient screening, enrollment, and group allocation processes. After applying the inclusion and exclusion criteria, 165 patients were finally included in this study. These patients were subsequently classified into two treatment groups, and 101 and 64 patients received flumatinib and nilotinib, respectively, as first-line therapy. Patient demographics, bone marrow evaluation, laboratory tests, and EUTOS long-term survival (ELTS) scores at diagnosis were obtained from the patient files. Patients harboring atypical BCR::ABL1 transcripts were excluded from the study, as these transcripts could not be assessed using standardized reverse transcription-quantitative polymerase chain reaction (RT-qPCR) assays.

Figure 1.

Patient enrollment and treatment group allocation.

Treatment

As nilotinib and flumatinib are both covered by basic medical insurance for patients with newly diagnosed CML, their monthly costs are comparable. Therefore, the choice of TKI is made by physicians, mainly according to age and comorbidities. Patients with metabolic comorbidities such as fatty liver disease, coronary heart disease, cerebrovascular accidents, or peripheral arterio-occlusive disease were excluded from nilotinib treatment, whereas no specific contraindications to flumatinib are applied in our clinical practice.

All the patients were administered a standard dose. The initial dose of nilotinib was 300 mg orally twice per day and that of flumatinib was 600 mg orally once per day. Dose reduction and withholding due to intolerance were allowed at the investigators’ discretion. Any TKI dose reduction or suspension due to early AEs was limited to a maximum of 14 days per instance, with no more than three adjustments allowed. If patients remained intolerant to the standard dosage, they were considered intolerant to the TKI and advised to switch to another TKI.

Follow-up evaluation

Molecular responses were assessed via RT-qPCR, standardized using the International Scale for BCR::ABL1 transcripts from bone barrows or peripheral blood cells. Cytogenetic analysis was performed using chromosome banding analysis of Giemsa-stained metaphases from bone marrow cells.

All patients were followed up according to the European Leukemia Net (ELN) 2020 recommendations for the treatment of CML. ¹⁵ Data on chromosome banding analyses of Giemsa-stained metaphase cells, RT-qPCR results at each landmark time point, and AEs during TKI treatment were collected. Data after switching to another TKI were not recorded or analyzed.

Definition of responses, survival endpoints, and toxicity

Complete hematologic response, major cytogenetic response, complete cytogenetic response (CCyR), and MMR were defined using conventional criteria. Early molecular response (EMR) was defined as a BCR::ABL1 transcript level ⩽10% at 3 months after the initiation of treatment. MR² was defined as a BCR::ABL1 transcript level ⩽1%. MR⁴ was defined as a BCR::ABL1 transcript level ⩽0.01%. MR^4.5 was defined as a BCR::ABL1 transcript level ⩽0.0032%. DMR was referred to as MR⁴ or deeper. ¹⁵ EMR was defined as BCR::ABL1/ABL1 ⩽10% after 3 months of first-line TKI therapy. Optimal responses and failures at each milestone, expressed as BCR::ABL1, were defined according to the ELN 2020 recommendations. ¹⁵ AEs were graded according to the Common Terminology Criteria for Adverse Events (version 5.0; National Cancer Institute, Bethesda, MD, USA). Dose reduction or interruption owing to intolerance was defined as a dose decrease or treatment discontinuation lasting more than 14 days for any reason.

Patient-reported outcomes

The cancer-specific European Organization for Research and Treatment of Cancer Quality of Life Questionnaire-CML24 (EORTC QLQ-CML24) was used to evaluate health-related quality of life (HRQoL) and symptom burden. ¹⁶ This disease-specific questionnaire comprises six scales: symptom burden, impact on worry/mood, impact on daily life, body image problems, satisfaction with care and information, and satisfaction with social life.

Statistical analysis

Continuous variables are expressed as medians and categorical variables are expressed as frequencies. The Statistical Package for the Social Sciences (SPSS) for Windows v.27.0 (SPSS Inc., Chicago, IL, USA) and GraphPad Prism 9.0 (GraphPad Software, San Diego, CA, USA) were used for statistical analysis. Normally and non-normally distributed quantitative data were compared using the t-test and Mann–Whitney U test, respectively. Qualitative data were analyzed using Chi-square tests. To address potential confounding factors, particularly the significant age difference between treatment groups, we performed propensity score matching (PSM) using a 1:1 nearest-neighbor algorithm without replacement. The matching covariates included age, sex, and ELTS score, with a caliper width of 0.2 standard deviations of the propensity score logit. Differences between cumulative incidence rates of MMR and DMR were assessed using a log-rank test. For HRQoL, the scores of each scale of the EORTC QLQ-CML24 were transformed linearly and ranged from 0 to 100 for standardization. ¹⁷ A higher score reflects better function or a higher level of satisfaction. The Mann–Whitney U test was used to assess group differences. Statistical significance was set at p < 0.05.

Reporting guideline

The reporting of this study conforms to Strengthening the reporting of observational studies in epidemiology (STROBE; Supplemental Table 1). ¹⁸

Results

Baseline characteristics

Of the 165 enrolled patients with CML-CP, 101 and 64 were treated with flumatinib and nilotinib, respectively, as first-line therapy. Table 1 presents the demographic and clinical characteristics of the patients. Patients who received flumatinib were significantly older than those who received nilotinib (p = 0.004). Specifically, 16.7% (17/102) and 3.1% (2/64) of patients were aged >65 years in the flumatinib and nilotinib groups at diagnosis, respectively. No significant differences were observed for sex, blasts in peripheral blood and bone marrow, blood cell counts, or other features. According to the ELTS stratification, 80 (79.2%), 16 (15.8%), and 5 (5.0%) patients in the flumatinib group and 56 (87.5%), 7 (10.9%), and 1 (1.6%) in the nilotinib group were classified as low risk, intermediate risk, and high risk, respectively. The risk stratification of the two groups was not significantly different (p = 0.161).

Table 1.

Baseline patient characteristics at diagnosis.

Characteristic	Flumatinib (N = 101)	Nilotinib (N = 64)	p Value
Median age (range)	44 (16–84)	37 (16–73)	0.004
Sex (male/female)	63/38	39/25	0.977
ELTS risk group, n			0.161
Low	79.2% (80)	87.5% (56)
Intermediate	15.8% (16)	10.9% (7)
High	5% (5)	1.6% (1)
Median peripheral-blood blasts (range), %	0 (0–8)	0 (0–9)	0.298
Median bone marrow blasts (range), %	0.8 (0–8.9)	0.2 (0–6)	0.233
Additional chromosomal abnormalities, n	10	5	0.643
Median hemoglobin concentration (range), g/L	118 (62–161)	114 (75–154)	0.933
Median platelet count (range), 109/L	468 (117–1899)	483.5 (118–2011)	0.315
Median white-cell count (range), 109/L	77.58 (8.3–630.72)	81.9 (9.29–519.85)	0.875
Basophils (range), %	4.95 (0–19.8)	5.06 (0–15)	0.797
Eosinophils (range), %	1.45 (0–10)	1.5 (0–7)	0.540

ELTS, EUTOS long-term survival.

Treatment responses

We first analyzed the achievement of molecular response milestones based on BCR::ABL1 transcript levels. The optimal molecular responses at 3, 6, and 12 months were 92.3%, 90.5%, and 85.1% in the flumatinib group and 90.2%, 91.2%, and 88.2% in the nilotinib group, respectively, which did not differ significantly (Figure 2). At 12 and 24 months, no significant difference was found in cumulative treatment failure rates between the flumatinib and nilotinib groups (7.9% vs 10.9% at 12 months, p = 0.511 and 9.9% vs 12.5% at 24 months, p = 0.602).

Figure 2.

Optimal response rates at 3, 6, 9, and 12 months: BCR::ABL ⩽10% at 3 months, ⩽1% at 6 months, and ⩽0.1% at 12 months.

FLU, flumatinib; NIL, nilotinib.

Early response in 153 patients was evaluated based on the cytogenetic and/or molecular response at 3 months. In the flumatinib group, the CCyR, EMR, MR², and MMR rates were 90.3% (65/72), 92.3% (84/91), 72.5% (66/91), and 29.7% (27/91), respectively, whereas in the nilotinib group, these were 85.0% (34/40), 90.2% (55/61), 68.9% (42/61), and 27.9% (17/61), respectively. No significant differences were observed between the two groups at 3 months.

We subsequently evaluated the cumulative treatment responses, including MMR and DMR. In flumatinib group, the MMR achievement rates at 6, 9, 12, 18, and 24 months were 63.5% (47/74), 74.0% (37/50), 85.1% (40/47), 80.6% (25/31), and 66.7% (8/12), respectively, whereas in the nilotinib group, these were 64.9% (37/57), 78.0% (32/41), 88.2% (45/51), 91.4% (32/35), and 88.5% (23/26), respectively (p = 0.869, 0.654, 0.648, 0.360, and 0.246, respectively). Although the difference was not statistically significant, the MMR achievement rates in the flumatinib group at 18 and 24 months were lower than those in the nilotinib group (Figure 3(a)). Furthermore, in the flumatinib group, the MR⁴ achievement rates at 6, 9, 12, 18, and 24 months were 32.4% (24/74), 26.0% (13/50), 40.4% (19/47), 45.2% (14/31), and 41.7% (5/12), respectively, whereas in the nilotinib group, these were 35.1% (20/57), 53.7% (22/41), 60.8% (31/51), 60.0% (21/35), and 80.8% (21/26), respectively. MR⁴ achievements at 9, 12, and 24 months were significantly higher in patients in the nilotinib group than those in the flumatinib group (p = 0.007, 0.044, and 0.042, respectively; Figure 3(b)). Moreover, in the flumatinib group, the MR^4.5 achievement rates at 12, 18, and 24 months were 29.8% (14/47), 38.7% (12/31), and 41.7% (5/12), respectively, whereas in the nilotinib group these were 45.1% (23/51), 51.4% (18/35), and 73.1% (19/26), respectively (p = 0.118, 0.300, and 0.133, respectively). Although the difference was not statistically significant, the late molecular response, in terms of MR^4.5 achievement, was lower with flumatinib than with nilotinib.

Figure 3.

Molecular responses at milestones. (a) MMR rates at 3, 6, 9, 12, 18, and 24 months. (b) MR⁴ rates at 6, 9, 12, 18, and 24 months.

FLU, flumatinib; MMR, major molecular response (BCR::ABL ⩽0.1%); MR⁴, molecular response with a 4-log reduction in BCR::ABL transcripts from baseline (BCR::ABL ⩽0.01%); NIL, nilotinib.

In the flumatinib group, BCR::ABL1 kinase domain mutation analyses were performed in seven patients, six of whom tested negative. One patient harboring the T315I mutation failed to achieve an optimal response at 6 months and required a TKI switch. In the nilotinib group, BCR::ABL1 kinase domain mutation analyses were performed in eight patients. One patient harboring the I432T mutation showed a satisfactory response and achieved a DMR at 6 months. One patient harboring Y253H and E255V mutations failed to achieve MMR at 12 months and required a TKI switch.

In the flumatinib group, 5.9% (6/101) of patients switched to other TKIs during follow-up, which was much lower than that in the nilotinib group (10/64, p = 0.041). In the flumatinib group, TKI was switched in six patients—due to treatment failure in five patients and due to intolerance in one patient. In the nilotinib group, TKI was switched due to resistance in five patients and due to intolerance in five patients.

Treatment responses in the matched cohort

Following PSM, 47 patients with well-balanced baseline features were included in each group. The median age of the patients was 39 (20–76) and 37 (18–73) years in the flumatinib and nilotinib groups, respectively. The cumulative incidence of MMR at 24 months was 86.2% and 94.5% in the flumatinib and nilotinib groups, respectively (p = 0.583; Figure 4(a)). The 24-month cumulative incidence rates of DMR achievement were 51.6% and 86.4% for flumatinib and nilotinib, respectively, with significantly higher rates in the nilotinib group (p = 0.023; Figure 4(b)).

Figure 4.

Cumulative incidence of MMR (a) and DMR (b) in the propensity score-matched cohort.

DMR, deep molecular response; MMR, major molecular response.

Safety

In terms of safety, drug-related AEs associated with both agents were primarily graded 1 or 2 (Table 2). In the flumatinib and nilotinib groups, 15.8% (16/101) and 18.8% (12/64) of patients, respectively, experienced dose reduction or interruption because of AEs. The most common causes for discontinuation of both flumatinib and nilotinib were thrombocytopenia (7/101 vs 7/64) and elevated alanine aminotransferase or aspartate aminotransferase (ALT/AST) levels (3/101 vs 4/64). AEs in terms of AST/ALT elevation and hyperbilirubinemia were significantly lower with flumatinib. Elevated AST/ALT levels were observed in 20.8% (21/101) and 51.6% (33/64) of patients in the flumatinib and nilotinib groups, respectively (p < 0.001). Furthermore, grade 3 or 4 events occurred in 4% (4/101) and 6.3% (4/64) of patients in the flumatinib and nilotinib groups, respectively (p = 0.768). Hyperbilirubinemia was observed in 12.9% (13/101) and 64.1% (41/64) of patients in the flumatinib and nilotinib groups, respectively (p < 0.001). Grade 3 or 4 hyperbilirubinemia occurred in 7.8% (5/64) of patients in the nilotinib group but not in the flumatinib group (p = 0.008). In addition, hyperglycemia (10.9% vs 0%), hyperlipidemia (9.4% vs 0%), rash (34.4% vs 3%), and alopecia (18.8% vs 3%) were more frequent in the nilotinib group than in the flumatinib group. No diarrhea was reported in the nilotinib group, whereas 12 (11.9%) cases were reported in the flumatinib group (p = 0.004). All diarrheal events were grade 1 or 2 and did not lead to dose reduction or interruption. In the flumatinib and nilotinib groups, 47.5% (48/101) and 60.9% (39/64) of patients experienced hematologic abnormalities, including neutropenia (10.9% vs 17.2%), thrombocytopenia (21.8% vs 34.3%), and anemia (30.7% vs 43.7%). No significant difference was observed in the overall incidence of hematologic toxicity between the two groups (p = 0.159).

Table 2.

AEs and newly occurring or worsening hematologic or biochemical abnormalities detected through laboratory tests.

AEs	All grades			Grade 3 or 4
	Flumatinib, N = 101	Nilotinib, N = 64	p Value	Flumatinib, N = 101	Nilotinib, N = 64	p Value
	Number of patients (%)
Neutropenia	11 (10.9%)	11 (17.2%)	0.246	3 (3.0%)	2 (3.1%)	1.000
Anemia	31 (30.7%)	28 (43.7%)	0.088	5 (5.0%)	3 (4.7%)	1.000
Thrombocytopenia	22 (21.8%)	22 (34.4%)	0.075	8 (7.9%)	9 (14.1%)	0.206
Hematologic abnormality	48 (47.5%)	39 (60.9%)	0.093	14 (13.9%)	9 (14.1%)	0.971
AST/ALT elevation	21 (20.8%)	33 (51.6%)	<0.001	4 (4.0%)	4 (6.3%)	0.768
Hyperbilirubinemia	13 (12.9%)	41 (64.1%)	<0.001	0	5 (7.8%)	0.008
Diarrhea	12 (11.9%)	0	0.004	0	0
Glucose elevation	0	7 (10.9%)	0.001	0	0
Serum lipid elevation	0	6 (9.4%)	0.007	0	0
Rash	3 (3.0%)	22 (34.4%)	<0.001	0	0
Alopecia	3 (3.0%)	12 (18.8%)	<0.001	0	0
QTc prolonged	3 (3.0%)	2 (3.1%)	1.000	2 (1.3%)	0	0.522
Cardiovascular events	2 (1.3%)	4 (6.3%)	0.317	0	1 (1.6%)	0.388
Lipase elevation	5 (5.0%)	5 (7.8%)	0.677	0	0
Creatinine elevation	5 (5.0%)	0	0.158	2 (1.3%)	0	0.522
Fatigue	8 (7.9%)	9 (14.1%)	0.206	0	0
Headache	8 (7.9%)	11 (17.2%)	0.069	0	0
Pruritus	7 (6.9%)	9 (14.1%)	0.131	0	0

AEs, adverse events; ALT, alanine aminotransferase; AST, aspartate aminotransferase; QTc, electrocardiogram-corrected interval.

Disease-specific differences in HRQoL

HRQoL outcomes were assessed in 18 patients from each treatment group. All patients had a stable treatment duration of >12 months. Significant differences were observed in the baseline characteristics between the groups, with the mean age being higher (50.2 vs 43.3 years) and the mean duration of treatment being shorter (19 vs 28 months) in the flumatinib group than in the nilotinib group. No statistically significant differences were observed in the six scales of the EORTC QLQ-CML24 questionnaire between the two groups (Table 3). However, patients treated with nilotinib tend to report worse body image problems.

Table 3.

EORTC QLQ-CML24 scores of flumatinib and nilotinib groups.

TKIs	Symptom burden		Impact on worry/mood		Impact on daily life		Body image problems		Satisfaction with care and information		Satisfaction with social life
	Mean (SD)	p	Mean (SD)	p	Mean (SD)	p	Mean (SD)	p	Mean (SD)	p	Mean (SD)	P
Flumatinib	78.78 (14.89)	0.949	78.70 (23.78)	0.479	79.63 (17.57)	0.414	88.89 (19.80)	0.180	90.74 (16.39)	0.654	90.74 (15.36)	0.230
Nilotinib	80.20 (9.98)		82.87 (16.54)		74.69 (18.59)		79.63 (23.26)		87.04 (21.05)		81.48 (23.49)

A higher score reflects better function or a higher level of satisfaction.

EORTC QLQ-CML24, European Organization for Research and Treatment of Cancer Quality of Life Questionnaire-chronic myeloid leukemia 24; SD, standard deviation.

Discussion

Flumatinib (HH-GV-678) is an imatinib derivative enhanced by the addition of a trifluoromethyl group and the replacement of its phenyl ring with a pyridine group. ¹⁹ This structural modification enhances the specificity of BCR::ABL1 inhibition, enabling the compound to overcome resistance mediated by most ABL1 kinase domain mutations in vitro, except for the T315I mutation. ²⁰ The randomized phase III trial FESTnd study confirmed the superior efficacy of flumatinib over imatinib and demonstrated significantly faster and deeper molecular responses in the flumatinib group during the first year of follow-up. ¹⁴ Flumatinib has demonstrated potent BCR::ABL1 inhibition in preclinical studies, showing efficacy against imatinib-resistant mutations. ¹⁸ Although in vitro data suggest that flumatinib may be advantageous over nilotinib in treating some mutation types, real-world evidence remains limited to a single case report on its efficacy against the F359V/C mutation.^19,20 With less than 4 years since NMPA approval and availability only in China, the present study aimed to verify the efficacy and safety of flumatinib in real-world practice. In our study, the MMR achievement rate at 12 months was 85.1% compared with 52.6% in the FESTnd study. ¹⁴ This discrepancy may be attributed to stricter and more frequent AE monitoring in clinical trials, resulting in more dose reduction or discontinuation. Another explanation is that patients in our study who failed to achieve an optimal response tended to switch to another TKI, leading to the exclusion of the missing evaluation data from long-term follow-up, although this was rare. Notably, three patients in the flumatinib group and four in the nilotinib group switched TKI by 12 months.

A recent large-scale real-world study by Zhang et al. ²¹ reported the comparable efficacy of flumatinib and other second-generation TKIs in a large study cohort of patients with CML-CP. However, in our study, the DMR rates in the flumatinib group remained consistently lower than those in the nilotinib group, both in the overall population and after PSM. Our data revealed no statistically significant difference in MMR achievement at any time point between the two groups in the first 2 years. According to the ELN 2020 recommendation, MMR achievement refers to minimal residual leukemia cells and an extremely low risk of disease progression. ¹⁵ Our study revealed a similar potency of flumatinib and nilotinib in this respect. Notably, DMR achievement rates at 9, 12, and 24 months in the flumatinib group were significantly lower than those in the nilotinib group. DMR is widely regarded as the minimum requirement for patients considering TKI discontinuation, with its duration recognized as a key predictive factor for TFR. ²² The PSM analysis, which adjusted for baseline characteristics, including age and risk stratification, confirmed our primary findings regarding comparable MMR rates but significantly different DMR achievement between flumatinib and nilotinib in the entire patient group. The 24-month cumulative DMR rate of 86.4% in the nilotinib group versus 51.6% in the flumatinib group suggested that nilotinib may offer advantages in achieving a DMR, a crucial factor for patients considering TFR strategies. However, the similar MMR rates indicate that both agents are effective in achieving optimal responses. These findings should be interpreted in the context of the distinct safety profile of each drug and the potential need for individualized treatment selection based on patient characteristics and treatment goals. Nevertheless, although flumatinib is widely used in China, the follow-up duration was short in this study, and the proportion of patients followed up for 18 and 24 months in the flumatinib group was much lower than that in the nilotinib group. Moreover, further investigation is needed to assess the ability of this novel TKI to achieve DMR.

In the present study, flumatinib exhibited excellent tolerability. Aligning with the FESTnd study, diarrhea and liver function abnormalities were the most common non-hematological AEs. Most AEs were transitory and moderate and seldom resulted in drug discontinuation. Furthermore, although both agents demonstrated comparable overall safety profiles, flumatinib exhibited significantly lower rates of hepatic and metabolic AEs, including elevated AST/ALT levels and hyperbilirubinemia. Other organ dysfunctions, such as cardiovascular events, creatinine elevation, and pulmonary toxicity, were quite rare with flumatinib in the current follow-up. In addition, the incidence of hematological AEs in the flumatinib group was lower than that in the nilotinib group, although the difference was not statistically significant. Our study suggests that flumatinib may offer certain advantages in the management of treatment-related complications, which is consistent with our clinical observations. These findings may help explain the age distribution differences between the two groups in real-world practice, as clinicians often consider flumatinib a potentially preferable option over nilotinib for older patients owing to its metabolic and hepatic safety. However, treatment decisions should be individualized and based on comprehensive patient assessments. Although our study provides important insights into the short-term safety profiles of these TKIs, cumulative frequencies of certain AEs may increase with prolonged treatment duration.^12,23,24 Therefore, the follow-up period should be extended to fully elucidate the AE profile of flumatinib. In addition, other nonhematological AEs, such as rash and alopecia, were common in the nilotinib group and were of particular concern to younger patients due to their impact on body image. Flumatinib treatment causes very few changes in appearance, which potentially explains the better body image perception observed in the flumatinib group using the EORTC QLQ-CML24 questionnaire. As most patients with CML-CP have a normal life expectancy, ³ the effect of TKI on appearance is also an important concern in the context of their social function recovery. The results of this study should be interpreted with caution because of the unbalanced treatment duration between the nilotinib and flumatinib groups (28 vs 19 months) and age distribution (43.3 vs 50.2 years), which may have subsequently influenced the interpretation of these patient-reported outcomes. Younger patients and populations with longer treatment exposure are more sensitive to body image changes and treatment-related side effects.

Currently, TKI toxicity is widely attributed to off-target effects. ¹³ For example, the inhibition of PDGFR, which is mainly expressed in pericytes and pulmonary tissues, may cause TKI-related fluid retention and serous cavity effusion. ²⁵ Furthermore, dermatological abnormalities during TKI treatment may be associated with functional changes in KIT, which plays an important role in the development of hematopoietic cells and melanocytes. ²⁶ In addition, the inhibition of vascular endothelial growth factor receptor (VEGFR), which mediates angiogenesis and vascular permeability, and other genes that regulate the function of platelets and vascular endothelial cells may increase the risk of cardiovascular and thrombotic events during TKI treatment. ²⁷ We speculate that the safety and intolerability of flumatinib may be associated with its high selectivity for BCR::ABL1. Furthermore, flumatinib has a lower inhibitory effect on c-KIT and PDGFRβ than imatinib and has no effect on the phosphorylation of epidermal growth factor receptor , VEGFR, cellular SRC, or human epidermal growth factor receptor 2. ²⁰

In conclusion, this study compared the efficacy and safety of flumatinib with those of another 2G-TKI in a relatively large real-world cohort. Our study also had some limitations. The limited follow-up duration and patient scale may not fully demonstrate the durability of responses, treatment resistance, and long-term safety of flumatinib. We could not obtain complete data on treatment responses and AEs due to poor adherence among a small number of patients with unavailable medical records. Nevertheless, our results reveal that flumatinib is a promising alternative for patients with newly diagnosed CML-CP, with good efficacy and safety.

Conclusion

This study provides real-world evidence comparing flumatinib and nilotinib as first-line treatments for CML-CP. Our analysis demonstrated comparable optimal response rates between the two agents, although flumatinib was associated with a lower DMR achievement. Flumatinib exhibited a more favorable hepatic and metabolic safety profile, highlighting it as a preferred option for older patients or those with metabolic disorders.