Dear Editor,

The availability of BCR-ABL tyrosine kinase inhibitors (TKIs) has led to a paradigm shift in the management of chronic myeloid leukemia (CML). Patients with chronic phase CML now lead a normal to a nearly normal life [1, 2]. Therefore, the focus has now shifted to treatment free remission (TFR). In fact, TFR is now considered the “new goal” in the management of CML-CP. The STIM trial [3] was one of the landmark studies evaluating the results of treatment free remission. Since then, several other studies also have evaluated the outcomes of TFR [4,5,6,7]. With the availability of several studies, the European Leukemia Net (ELN) [8] and the National Comprehensive Cancer Network (NCCN) [9] have laid down guidelines for patient selection and monitoring of patients who are candidates for TFR. However, these guidelines are often difficult to follow in clinical practice in low- and middle-income resource-constrained countries. We began a structured TFR program in March 2022 at our centre with more stringent inclusion criteria, but with liberal monitoring (as compared to ELN and NCCN criteria; Table 1 highlights the differences between the ELN guidelines, NCCN guidelines and our program). A prospective database of TFR was maintained. This retrospective analysis of prospectively maintained database evaluates outcomes of patients who were initiated on TFR till 31st December 2023. Data was updated till 31st December 2024. The study was approved by the institutional ethics committee. Given the nature of the study, the requirement for a consent was waived off.

Table 1 Comparison of ELN, NCCN criteria and our program for initiation of TFR.
Full size table

Patients required to fulfill all of the following criteria to be initiated on TFR—(a) 1st chronic phase CML, (b) p210 transcript, (c) imatinib as 1st line therapy, (d) daily dose of 400 mg per day or a lesser, (e) minimum duration of therapy 8 years, (f) continuous MMR for 5 years, (g) MR4 for minimum of 3 consecutive years or MR4.5 for 2 consecutive years. However, patients who needed a higher dose because of drug interactions (e.g., tuberculosis patients receiving rifampicin) were included. Patients who had previously discontinued imatinib for any reason and then restarted were not eligible. Similarly, patients with 2nd chronic phase were ineligible. Monitoring with quantitative polymerase chain reaction (Q-PCR) was done every 2 months in the entire 1st year of TFR. In the 2nd and 3rd years of TFR, the frequency of Q-PCR was reduced to 3–4 months and 6 months, respectively.

TFR failure was defined as loss of MMR. This was always confirmed twice (at least 2–4 weeks apart) if it happened after 6 months of stopping imatinib. For loss of MMR before 6 months of TFR, repeat testing was not mandatory (however was done wherever possible). In event of TFR failure, imatinib (400 mg per day) was restarted as soon as possible with subsequent Q-PCR testing at the discretion of the treating clinician.

The primary objective was to determine the molecular relapse free survival (MRFS). The event for MRFS was loss of MMR. Patients remaining in MMR at last follow up were censored on the date of last follow up. Patients who restarted imatinib because of withdrawal phenomenon were censored on the date of the last RQ-PCR, which confirmed MMR. Those patients who died (without molecular relapse) because of unrelated causes were considered as having competing events (as was done in the STIM trial) [3]. The secondary objectives included patterns of molecular relapse, evaluation of compliance, factors predicting TFR failure, PCR kinetics and outcomes post TFR failure. Compliance was determined by the number of PCRs done in the 1st year (in the absence of loss of MMR) or till the time of loss of MMR (in pts who lost MMR), whichever was earlier. Patients with minimum 3 PCRs in the first 6 months and minimum 5 PCRs in the 1st year were considered compliant.

Sixty-nine patients with median age 47 years (range 27–78 years) and 49 (71%) males were included. The median time from diagnosis of CML to TFR was 13 (range 8–21) years, and the median duration of continuous MMR before TFR was 8.1 (range 5–9.7) years. With a median follow up of 15 months, the median MRFS was not reached and the 1-year MRFS was 80% (Fig. 1). Two patients died of unrelated causes (one due to cardiac cause and one due to decompensated liver failure) and were taken to have competing event. One patient restarted imatinib due to withdrawal symptoms while in MMR and was censored.

Fig. 1
figure 1

Molecular relapse free survival.

Full size image

Among all patients, 13 (19%) had TFR failure. Of these, 12 (92%) were in the initial 6 months of TFR. Of the 54 (78%) patients in confirmed MMR at 6 months (12 loss of MMR by 6 months, 2 deaths of unrelated causes, and 1 restarted imatinib due to withdrawal symptoms), only 1 (1.8%) had loss of MMR subsequently. Fifty-seven (83%) were adequately compliant to PCR monitoring. Of the 12 with inadequate compliance, 11 (92%) exhibited non-compliance in the initial 6 months of TFR (i.e., they did not have 3 quantitative PCRs in the first 6 months). Compliance did not affect the incidence of TFR failure [2 of 12 (17%) vs 11 TFR failures (19%) in 57 compliant patients, p = 0.83]. Age, gender, duration of TKI therapy before TFR, and duration of continuous MMR before TFR were not different between those who had TFR failure versus those who did not (Table 1S).

Of the 13 patients with TFR failure, 12 were restarted on imatinib (1 was lost to follow up). As of their last follow up, 11 have re-attained MMR and the remaining 1 patient had 1.97 log reduction in the PCR value at 3 months of restarting imatinib, suggesting adequate response. No patient developed imatinib resistance or progressed to accelerated phase/blast crisis.

A review of the studies on TKI discontinuation studies [3,4,5,6,7] reveals that there are differences in the patient selection criteria, monitoring, management after TFR failure, etc. While most studies of TFR report long-term MRFS of 40–60%, one retrospective study [10] reported a MRFS of 79%; in this study from M D Anderson Cancer Centre, the median duration of TKI use prior to discontinuation was approximately 10 years and 99% of patients were in MR4.5 at the time of TKI discontinuation. Several studies have identified the impact of patient and disease related factors on the long-term TFR outcomes. As summarized in a recent review article [11], some of the factors identified in multi-variate analysis in one or more studies include age, 1st line therapy, treatment duration before TFR, duration of MR4 or MR4.5 before discontinuation, Sokal score, and quantitative PCR value by droplet digital PCR before discontinuation. These findings point out that patient selection criteria for TFR have an impact on the long-term success. This may be especially important in low- and middle-income countries (LMICs) where logistic challenges for intense monitoring abound. In our TFR program, using rigorous enrollment criteria, we were able to have a TFR success of ~80% at 1 year. However, we did not find any impact of duration of TKI therapy, and duration of MMR, likely because of the stringent selection criteria followed in our study.

The timing of molecular relapse also has been consistently similar across most studies with 70–95% of molecular relapse occurring within the first 6 months, and virtually all within the first 1 year. The findings in our cohort are also similar with >90% of all failures occurring within the first 6 months. Although late relapses have been reported [12], we did not have any late relapse in our cohort of patients.

The issue of compliance has not been adequately described in many of the pivotal or prospective studies of TFR [3, 6]. Even in some of the “real-world” studies, the issue of compliance has not been adequately addressed [13, 14]. Therefore, we had to use an arbitrary criterion for defining compliance. Using our definition, we identified that about a fifth of the patients were not adequately compliant to monitoring. This is likely to be one of the key logistic challenges faced in LMICs.

Our study has certain limitations. We do not have information about Sokal/other prognostic scores. We did not capture information on withdrawal syndrome and quality of life. We did not have a rigorous post-failure monitoring plan; this was left to the discretion of treating clinicians. Finally, late events may not have been adequately be captured.

Our TFR program demonstrates excellent outcomes (80% MRFS) using rigorous selection criteria, but with a less stringent monitoring plan. More than 90% failures occurred in the first 6 months of TFR with very few failures (<2%) happening in those who continued to remain in MMR at 6 months. Even though the monitoring schedule was less stringent, only about 80% were compliant to the monitoring, highlighting the practical challenges in LMICs. None-the-less, the long-term outcomes of our TFR program suggest that stringent criteria for initiating TFR with less stringent monitoring policy may be the way forward not only in LMIC countries, but even in resource-intense settings.