Introduction

Chronic myeloid leukemia (CML) is a myeloproliferative neoplasm that results from aberrant clonal hematopoietic stem cells caused by oncogenic BCRABL fusion protein, which arises from a t(9;22)(q34;q11) chromosomal translocation (Philadelphia chromosome) [1]. BCR-ABL1 tyrosine kinase inhibitors (TKIs) have significantly improved survival outcomes in patients with chronic phase CML (CML-CP), and approximately 50% of patients who achieved a sustained deep molecular response (DMR) can discontinue TKI administration without molecular relapse [2]. However, long-term TKI treatment may be required, with the majority of TKI stop studies showing that approximately 70–100 months of treatment is required before TKI therapy can be discontinued [3,4,5,6,7,8,9,10]. Although TKIs are associated with relatively good safety profiles, adverse events such as renal toxicity and cardiovascular events are often observed with long-term TKI use [11]. In CML-CP clinical trials, one reason for discontinuing TKI treatment is to relieve patients from the adverse effects of long-term TKI use, although it is unclear whether these chronic adverse toxicities are reduced or resolved after discontinuation. Here, a retrospective evaluation has been conducted to determine changes in estimated glomerular filtration rate (eGFR) in patients with CML-CP who discontinued TKI treatment.

Patients and methods

Patients

This retrospective analysis was conducted at Saga University Hospital (Saga, Japan); eGFR was evaluated in patients with CML-CP patients who received imatinib, dasatinib, or nilotinib as first-line treatment and who achieved sustained DMR, subsequently discontinuing TKI therapy between December 2001 and March 2017. Patients who were lost to follow-up, those who had no available laboratory data, and those enrolled in other clinical trials were excluded. The final follow-up date was June 2019. All clinical data were reviewed by two expert hematologists.

The study protocol was approved by the institutional review board of Saga University. All procedures involving human participants were conducted in accordance with the ethical standards of institutional and/or national research committees, and the Declaration of Helsinki. Informed consent was waived because of the retrospective design of the data collection. Information of the research, including the aims of use of specimens and the opportunity to opt out from the research, was made public. No patients objected to this study.

Definitions

Real-time quantitative reverse transcription polymerase chain reaction analysis was performed to detect the BCR-ABL1 mRNA transcript level. DMR was defined as BCR-ABL1 transcript level ≤ 0.01% on the standardized international scale (IS). A major molecular response (MMR) was defined as a BCR-ABL1 IS transcript level ≤ 0.1%. Treatment-free remission (TFR) was defined as sustained DMR in the absence of treatment. Molecular relapse (M-Rel) was defined as the loss of DMR (0.01% > IS) at two consecutive time points or loss of MMR (0.1% > IS) at a single time point; in these cases, the patient was re-treated with the last TKI at the previously effective dose.

Statistical analyses

The eGFR was estimated using the Cockcroft–Gault method at 3 timepoints before and after TKI discontinuation. All continuous variables (age, treatment duration, follow-up duration, complete cell counts, eGFR change or eGFR at TKI discontinuation) were compared between patients who achieved TFR and those exhibiting M-Rel using Mann–Whitney U tests. As patients with M-Rel were re-treated with a TKI immediately after relapse, these subjects were defined as patients who received continuous TKI treatment; all molecular relapses occurred within 6 months of TKI discontinuation. Fisher’s exact tests were used to compare clinical features (sex, hypertension [HT], hyperlipidemia [HL], diabetes mellitus [DM], and first-line TKIs) in the TFR and M-Rel groups. Pearson’s correlation analysis was used to determine the correlation between eGFR and TKI treatment duration, between eGFR and age at diagnosis, and between age at diagnosis and TKI treatment duration. Paired t tests or repeated-measures analysis of variance (ANOVA) were used to detect differences in eGFR across multiple test data. Multiple regression analysis was performed to determine the factors involved in the decline of eGFR at the time of TKI stop (Variables included in the analysis were type of first-line TKIs and TKI treatment duration). Two-sided p values of < 0.05 were considered statistically significant. All statistical analyses were performed using the EZR software package (Saitama Medical Center, Jichi Medical University) [12].

Results

Patient characteristics

Data from 33 patients with CML-CP were included in the analysis (17 males, 16 females). The clinical characteristics of the patients are summarized in Table 1. The median age at diagnosis was 56 years (range, 35–82 years). First-line TKIs were imatinib (IMA, n = 17), dasatinib (DAS, n = 11), and nilotinib (NIL, n = 5); TKIs used prior to discontinuation were IMA (n = 5), DAS (n = 22), and NIL (n = 6). Median dose of IMA, DAS, and NIL per day were 300 mg (range, 150–300 mg), 20 mg (10–100 mg) and 600 mg (all patients received 600 mg of nilotinib), respectively. The median duration of TKI administration was 4.97 years (range, 2.56–13.89 years), and the median follow-up time from diagnosis was 8.47 years (range, 5.11–17.18 years). A total of 21 patients experienced TFR and 12 experienced M-Rel. Fewer male patients experienced M-Rel (male: 2/17, female: 10/16, p = 0.004). Other clinical features did not differ between the TFR and M-Rel groups, including underlying disease-associated chronic kidney disease (CKD) and eGFR level at diagnosis.

Table 1 Data are presented at median [range] or N (%)
Full size table

Changes in eGFR following TKI discontinuation

At the time of TKI discontinuation, a statistically significant correlation was seen between eGFR and the duration of TKI treatment (r = − 0.478, p = 0.005; Fig. 1). Evaluation of the average eGFR at 3 timepoints (at 3 years before TKI discontinuation, at TKI discontinuation and at 2 years after TKI discontinuation) showed a moderate decline in the 3 years prior to TKI discontinuation, which was similar in both the TFR and M-Rel groups (− 1.46 vs − 2.25 ml/min/1.73m2; p = 0.726; Table 1). A slight increase in mean eGFR was seen in patients with both TFR after TKI discontinuation and in patients with M-Rel (+ 0.70 vs + 0.12 ml/min/1.73m2; p = 1.000; Table 1), whereas the increase was not statistically significant in either group (p = 0.666). In addition, no significant differences in eGFR were seen between the TFR and M-Rel groups (67.95 vs 65.30 ml/min/1.73m2; p = 0.708; Table 1).

Fig. 1
figure 1

Correlation between estimated glomerular filtration (eGFR) at tyrosine kinase inhibitor (TKI) stop and TKI treatment time

Full size image

Changes in eGFR according to first-line TKI

The use of first-generation imatinib or second-generation dasatinib or nilotinib was evaluated to determine any differences in the effect on eGFR. Data showed that eGFR levels in patients who received imatinib as initial treatment were significantly lower than those seen in patients treated with dasatinib or nilotinib (p = 0.027). The time effect on the differences in eGFR was not statistically significant by repeated-measures ANOVA. The duration of TKI treatment was significantly longer in patients receiving imatinib than in patients receiving dasatinib or nilotinib (mean ± standard deviation, 8.13 ± 2.87 years vs. 3.82 ± 1.00 years, respectively; t test, p < 0.001).

Multiple regression analysis on eGFR at the time of TKI discontinuation

We additionally performed multiple regression analysis to determine the factors involved in the decline of eGFR at the time of TKI stop. Multiple regression analysis revealed eGFR at TKI stop was inversely correlated with treatment duration (coefficient − 2.56, p = 0.038), while the type of TKI was not (coefficient 0.835, p = 0.908).

Discussion

Here, we report eGFR changes in patients with CML over a 5-year period before and after TKI discontinuation. The proportion of patients who maintained TFR after cessation of TKIs (64%) was consistent with previous studies [3,4,5,6,7,8,9, 13]. Fewer male patients developed molecular relapse, which is consistent with the results of the ‘Stop Imatinib’ (STIM) trial [3], although male sex has not been seen to be a favorable predictive factor for TFR in other TKI stop studies [3,4,5,6,7,8,9, 13]. Female sex has previously been reported to be an independent favorable factor for DMR [14,15,16]. The existence of sex-specific responses in patients with CML-CP may, therefore, be supported by these clinical observations, but detailed mechanisms have not been fully elucidated, and further investigation is required.

In the present study, TKI treatment duration was seen to correlate with eGFR decline, which is consistent with previous reports [17,18,19,20]. The average eGFR decline until TKI discontinuation was seen to be 1.76 ml/min/1.73 m2/year. Although it is difficult to compare these results with those seen in previous studies, it appears that eGFR levels in the present patient group decreased more rapidly than that seen in healthy Japanese individuals (< 1.0 ml/min/1.73 m2/year) [21, 22]. These results indicate that long-term TKI treatment can induce a decline in eGFR in patients with CML-CP.

Patients who received first-line treatment with imatinib showed lower eGFR levels than patients treated with dasatinib or nilotinib, possibly due to the longer treatment duration associated with imatinib [17, 23]. Imatinib was the first TKI approved for the treatment of CML-CP in Japan in 2001; nilotinib and dasatinib were approved as first-line treatment approximately 10 years later [24]. These second-generation TKIs are stronger inhibitors of BCR-ABL than imatinib, and DMR can be achieved with a shorter treatment duration [25,26,27]. Therefore, the imatinib treatment duration may be significantly longer than dasatinib or nilotinib, leading to a greater incidence of renal toxicities, as observed in this study. The present study showed no significant increase in eGFR at the 2-year timepoint following TKI discontinuation. This follow-up period may be short in comparison with the treatment duration (approximately 5 years) and, therefore, may not be long enough to demonstrate a change in eGFR.

In conclusion, TKI-induced renal toxicities were associated with long-term TKI treatment time, while may be irreversible even following treatment discontinuation.