Starting Tyrosine Kinase Inhibitor Cessation in Chronic-Phase Chronic Myeloid Leukemia Patient

A 48-year-old woman with chronic-phase chronic myeloid leukemia (CP-CML) is being treated with imatinib 400 mg daily. She was initially diagnosed at the age of 43 and has been on imatinib for five years. Her Sokal risk score, at diagnosis, was low. She has met the desired treatment milestones including an early molecular response (BCR-ABL transcripts <10% on the International Scale [IS]) after three months on imatinib, and she achieved a complete cytogenetic response after 12 months on imatinib. She has had peripheral blood BCR-ABL transcripts measured using qRT-PCR with a sensitivity of 4.5 logs every three months since diagnosis. Her BCR-ABL transcripts became undetectable 18 months after starting imatinib and have remained undetectable on each subsequent test. She has no history of advanced-phase CML and has never had evidence of resistance to treatment. She tolerates imatinib well except for periorbital edema and occasional muscle cramps. She has been adherent to treatment and followed her physician’s recommendations consistently since she was diagnosed.

Is this patient a candidate for a trial of tyrosine kinase inhibitor (TKI) cessation?

The development of BCR-ABL TKI revolutionized the management of CML, beginning with the U.S. Food and Drug Administration (FDA) approval of imatinib in 2001. Imatinib, along with the four other FDA-approved TKIs that have followed, changed CML from one of the leading indications for allogeneic stem cell transplantation into a chronic disease that can be managed with lifelong oral therapy.^1-5  Progression into the advanced phases of accelerated or blast-phase CML has become exceedingly rare. BCR-ABL TKIs can induce complete cytogenetic responses and very deep levels of molecular response in most CP-CML patients, leading to a life expectancy that rivals that of the general population.⁶ 

Despite these positive aspects of TKI therapy, there remains a negative side to the current management of CML. For some, the idea of taking an oral anti-cancer agent daily for the remainder of their lives is overwhelming. In many cases, remembering to take a daily drug can be a challenge. Furthermore, TKIs can be associated with significant toxicity. Although some potential adverse effects such as cytopenias and transaminitis are common across the entire class of drugs, each TKI has a slightly different adverse effect profile. Some of the more common toxicities include fatigue, fluid retention, muscle cramps, nausea, vomiting, diarrhea, rash, and musculoskeletal pain. Various studies have looked at the impact of TKI-related adverse events and found that even low-grade adverse events can significantly impair a patient’s health-related quality of life.⁷  In addition to toxicity, TKIs present a significant financial burden to both patients and the health-care system in general. The cost of TKIs is upwards of $100,000 per year in most cases, and even with the introduction of generic imatinib, this cost has not significantly declined.⁸  Some studies have identified a correlation between high patient costs and poor adherence to treatment, as patients attempt to ration their drugs in an effort to save money.⁹ 

TKI Discontinuation Trials

For the reasons mentioned above, many CML investigators began asking the question, “Can we discontinue TKIs in patients with deep levels of molecular remission?” The first study to address this question was the STIM1 trial. This study included 100 patients with CP-CML who had achieved undetectable BCR-ABL transcripts measured by quantitative reverse transcriptase polymerase chain reaction (qRT-PCR) while on imatinib. All patients had been on imatinib for a minimum of three years and had maintained a deep molecular response (DMR) for a minimum of two years. Imatinib was discontinued and patients were monitored off treatment. Molecular recurrence (MR) was defined as loss of major molecular response (MMR) or a 1-log increase in BCR-ABL transcript levels on two consecutive tests. The long-term data from this trial were published in 2017 and showed that with a median follow-up time of 77 months, 38 percent of patients remained in a treatment-free remission (TFR), even though most had low levels of detectable BCR-ABL transcripts. The majority of MRs occurred in the first six months after TKI cessation, with a median time to relapse of 2.5 months. Responses were favorable when relapsed patients resumed TKI therapy. Perhaps more importantly, none of the patients progressed to an advanced phase of CML.¹⁰ 

Additional imatinib discontinuation studies included the Australian TWISTER study, which enrolled 40 patients and reported a 47 percent TFR rate after two years; the A-STIM study, which enrolled 80 patients and reported a 61 percent TFR rate at three years; and the KIDS study, which enrolled 90 patients and reported a TFR rate or 58.5 percent at two years.^11-13  Minor differences exist in eligibility criteria and the definition of MR between studies, which could account for some of the differences in rates of TFR.

More recent TKI cessation trials have studied the possibility of discontinuing second-generation TKIs including nilotinib and dasatinib. The EURO-SKI trial has been the largest TKI-stopping study to date and enrolled 868 CML patients treated with imatinib, nilotinib, or dasatinib; two years after TKI cessation, 50 percent of patients remained in TFR.¹⁴  The ENESTfreedom trial studied only patients receiving nilotinib as first-line therapy; the one-year TFR rate was 51.6 percent.¹⁵  The DasFree study enrolled patients receiving dasatinib as first- or second-line treatment, and one year after stopping, 63 percent remained in TFR.¹⁶  Many other examples exist, all of which have strikingly similar results.^17-22 

Predictors of TFR

Many TKI cessation trials have attempted to identify predictors for successful TFR. The majority of studies mandated a minimum of three years of TKI therapy prior to attempting discontinuation, and several have reported that longer TKI duration prior to stopping predicts for higher rates of TFR.^10,14  Although it is clear that achieving a DMR is essential to a successful attempt at TKI cessation, the EURO-SKI trial was the first to report the necessary duration of DMR to predict for better outcomes. Those patients with a DMR for at least 3.1 years prior to TKI discontinuation demonstrated significantly higher TFR rates at six months and two years than those with shorter durations of DMR.¹⁴  The Sokal risk score is another potential predictor of TFR, and many studies have shown that patients with a high risk score have a lower likelihood of remaining in TFR; however; these data have not been reproduced in all TKI cessation studies.^10,11,14 

Criteria for Safely Attempting TKI Discontinuation

Based on the available data, it appears that stopping TKIs is a safe and reasonable endeavor in a subset of patients with sustained deep molecular responses. In 2016, the National Comprehensive Cancer Network (NCCN) CML Panel developed guidelines to identify patients who are appropriate for a trial of TKI cessation. The most pertinent criteria reported in these guidelines include 1) CP-CML without any history of accelerated or blast-phase CML; 2) patient must be on an approved TKI for a minimum of three years; 3) patient must have prior evidence of quantifiable BCR-ABL transcripts; 4) patient must have a stable MR4 documented on at least four tests for at least two years; monthly qRT-PCR monitoring must be done for the first year after stopping TKIs, followed by every-six-week monitoring for the second year and every 12 weeks thereafter; and TKIs must be promptly restarted at the first qRT-PCR result suggesting a loss of MMR (>0.1% IS).

The NCCN guidelines website has a full list of criteria that should be met prior to an attempt at stopping TKIs.²³  Clinical trial data have repeatedly demonstrated that the majority of patients who relapse will do so within the first six months after stopping TKIs.^10,14,15  This accounts for the vital importance of monthly qRT-PCR monitoring using a test with a sensitivity of at least 4.5 logs that reports results using the IS. If this rigorous monitoring is not feasible for any reason, including adherence to visits or testing, then a patient is not an appropriate candidate for stopping treatment.

Defining MR

The definition of MR has differed among various studies; however, most CML investigators feel comfortable using loss of MMR as the trigger for restarting TKIs.¹²  Many patients will have low-level detectable BCR-ABL transcripts below the level of MMR, yet this does not necessarily indicate a need to resume therapy.¹¹  Few, if any, feel comfortable allowing the BCR-ABL transcripts to rise to greater than 0.1 percent IS without reinitiating treatment, as this has been the threshold used to determine MR in the majority of TKI cessation studies.

TKI Withdrawal Syndrome

One of the unexpected phenomena that has emerged in these studies is a TKI withdrawal syndrome. This has been reported in approximately 25 to 30 percent of patients after stopping TKIs and presents as low-grade, diffuse, musculoskeletal pain. It typically appears within the first one to two months after TKI discontinuation and may persist for up to six months. Successful treatment typically includes NSAIDs or steroids, and in most cases, the pain resolves promptly after resuming TKIs. Despite this, resumption of TKIs is not the recommended therapy for TKI withdrawal syndrome in the absence of a MR.^14,24 

What Remains Unknown

Notwithstanding the promising data reported from numerous TKI discontinuation trials, it remains unclear why TKI therapy can be successfully halted in a subset of the population when we know that hematopoietic stem cells are inherently resistant to TKIs. Evidence of low-level, detectable BCR-ABL transcripts is seen in many patients who remain in TFR.¹¹  Thus, it is conceivable that after years of being off therapy, a leukemic stem cell harboring genetic features of a more advanced or resistant CML could emerge. This possibility lends credence to the recommendation for lifelong monitoring by qRT-PCR to measure levels of BCR-ABL transcripts, even in patients who have remained in TFR for many years.

Back to the Case

The woman presented in the case had a low-risk Sokal score when diagnosed with CP-CML. She was on imatinib, without evidence of resistance, for five years. Her BCR-ABL transcripts were undetectable using qRT-PCR with a sensitivity of 4.5 logs for 3.5 years. Based on this information, she met the recommended criteria for an attempt at TKI cessation. After being counseled about the need for monthly qRT-PCR monitoring during the first year off treatment, as well as being educated about the possibility of TKI withdrawal syndrome, the patient chose to stop imatinib. Twelve months after stopping, her qRT-PCR has remained lower than 0.1 percent IS, indicating that she remains in MMR. Although she has had occasional detectable results, she has never crossed the threshold to reinitiate treatment. The patient experienced low-grade musculoskeletal pain in the first six months after stopping her TKI, and this was managed with NSAIDs as needed. The pain resolved within six months. She will continue with frequent qRT-PCR monitoring, yet the available data suggest that she has a very high likelihood of remaining in a TFR for years to follow.