Diamonds Are Forever, Tyrosine Kinase Inhibitors Are Not?

We could say that next to being able to give tyrosine kinase inhibitors to patients with CML to improve their life expectancy, the best thing we can do now is to take them away…sometimes.
— Jorge Cortes, MD

Sir Donald Munger: “You have been on holiday, I understand. Relaxing, I hope?”

James Bond: “Oh, hardly relaxing, but most satisfying.” (Diamonds Are Forever)

As tyrosine kinase inhibitors became the mainstay of therapy for patients with chronic myeloid leukemia (CML), our assumption has been that patients would have a favorable outcome as long as they continued receiving therapy. Up until recently, this was considered a treatment for life. This notion was seriously challenged by Mahon and colleagues, who introduced the concept of discontinuation of tyrosine kinase inhibitor treatment with the Stop Imatinib (STIM) trial. First reported in 2010, with a median follow-up of 17 months, it noted a projected probability of sustained complete molecular remission off therapy of 41% at 12 months.

As reported by Etienne et al1 and reviewed in this issue of The ASCO Post, updated results are now available after a median follow-up of 77 months, confirming that a sizable subset of patients may have durable remissions free of treatment. In this update, the projected probability of molecular relapse–free survival at 60 months is 38%. This represents a remarkable change in our preconceptions about tyrosine kinase inhibitor therapy.

Practical Implications

It is important, however, to understand the practical meaning of these findings. In this study, only patients with undetectable transcript levels with a very sensitive assay (with at least 50,000 ABL copies), measured in an experienced and high-quality laboratory, with such transcript levels sustained for at least 2 years, and with a minimum of 6 data points confirming this response during this period (ie, measured every 6 months as currently recommended by European LeukemiaNet), were eligible for this approach. Then, upon discontinuation of treatment, patients were monitored very closely, at least monthly for the first year and then every 2 months during year 2 and then every 3 months.

Proper selection of patients based on adequate monitoring is critical. It is likely that patients with responses that are not quite as deep or as durable will have a higher risk of relapse. Ongoing studies are precisely investigating whether, for example, patients with only an MR4 (BCR-ABL ≤ 0.01%) or even only a major molecular response, or with responses sustained for only 1 year, could be eligible for treatment discontinuation and what outcome can be expected in such instances. And even then, this all still depends on very high-quality, reproducible polymerase chain reaction testing.

In addition, very close monitoring after treatment discontinuation is of paramount importance to identify early relapses. After all, this and other studies have demonstrated that when a relapse is detected early, nearly all patients may regain their response upon re-start of therapy. Lack of adequate monitoring, however, would risk exposing a patient who had an excellent outcome (ie, a sustained very deep molecular response with undetectable transcripts) to a potentially disastrous outcome (eg, detection of relapse only after transformation).

Questions and Challenges Remain

Still, many questions and challenges remain to make treatment discontinuation a routine recommendation. First, it is an unfortunate reality that most patients are currently not being monitored frequently enough during therapy.2 This not only means we are not providing adequate care, potentially not recognizing promptly needs for changing therapy, or perhaps even changing therapy when it is not indicated. It also means we might not be able to tell in many patients whether they would be eligible for a safe consideration of treatment discontinuation. It also creates a major concern that monitoring might not be adequately done after eventual treatment discontinuation.

Second, follow-up is still short. The current report provides the longest follow-up available for any of these studies, an impressive median of 77 months. However, we need to remember that after stem cell transplant—a treatment modality of unquestionable curative potential in CML— some patients may relapse more than 10 or 15 years afterward despite sustained remissions up until then.3 These relapses are not common, but they occasionally come in the form of a sudden blast phase. It is important to continue following this and other studies closely for many years to better understand the potential for late relapses.

Third, this is an approach that is still successful in only a minority of patients. Even with second-generation tyrosine kinase inhibitors as front-line therapy, only approximately 40% to 45% of patients achieve a sustained MR4.5 (BCR-ABL ≤ 0.0032%).4 And of those patients, the present study tells us that approximately 40% will be free from relapse at 5 years. This means that only 15% to 20% of all patients can successfully discontinue treatment with our current approach.

Knowing that tyrosine kinase inhibitors do not eradicate the leukemic stem cell, this finding is not surprising. Much research is ongoing trying to discover ways to make more patients eligible. They include identifying and targeting pathways that might be responsible for the persistence of hematopoietic stem cells such as Janus kinase 2 (JAK2) or peroxisome proliferator-activated receptor gamma (PPAR-γ), to mention a few. The JAK2 inhibitor ruxolitinib (Jakafi) and the PPAR-γ agonist pioglitazone (Actos) are in clinical trials in combination with tyrosine kinase inhibitors. The hypothesis is that through such combinations, patients with a good response but not optimal for treatment discontinuation might become eligible to attempt stopping tyrosine kinase inhibitor therapy and/or patients who stop treatment will be more likely to remain remission-free off therapy if the leukemic stem cell is indeed eradicated.

Fourth, this study defined relapse as positivity of the polymerase chain reaction (PCR), with a sensitivity of 5 logs confirmed in a second sample, or loss of a major molecular response in any one sample. Increasingly, only patients with loss of a major molecular response are being considered as having lost response and thus being offered to re-start therapy.5 It is clear that some patients may have low levels of detectable disease off therapy without clear evidence of clinical disease progression and certainly not with transformation for several years. But the clinical and biologic significance and consequences of a life-long presence of low levels of a malignant clone (as defined by the presence of detectable BCR-ABL) are not known at this time.

Fifth, some recent studies have suggested that quality of life changes little if at all after treatment discontinuation.6 This perhaps highlights the fact that many of the symptoms patients experience while on chronic tyrosine kinase inhibitor therapy, such as fatigue or difficulties with memory, may not be related at all—or perhaps only partially—to the tyrosine kinase inhibitor.

Finally, treatment discontinuation is not for everyone. Many patients feel more comfortable continuing therapy than risking a relapse, even if it is likely manageable. Patients should be informed about the option of treatment discontinuation together with the known outcomes and open questions; then it is our job to support them in their decision.

Still, the findings in this study by Etienne et al are very important. Considering the long-term costs of tyrosine kinase inhibitors, the growing concern for arteriothrombotic events with some of them, and the mere biologic and clinical implications of having this option available to some patients, this study is groundbreaking. We could say that next to being able to give tyrosine kinase inhibitors to patients with CML to improve their life expectancy, the best thing we can do now is to take them away…sometimes. ■

Disclosure: Dr. Cortes has received research support from Ariad, Bristol-Myers Squibb, Novartis, Pfizer, and Teva; and is a consultant for Ariad, Bristol-Myers Squibb, Novartis, and Pfizer.


1. Etienne G, Guilhot J, Rea D, et al: Long-term follow-up of the French Stop Imatinib (STIM1) study in patients with chronic myeloid leukemia. J Clin Oncol. October 31, 2016 (early release online).

2. Saleh MN, Haislip S, Sharpe J, et al: Assessment of treatment and monitoring patterns and subsequent outcomes among patients with chronic myeloid leukemia treated with imatinib in a community setting. Curr Med Res Opin 30:529-536, 2014.

3. Goldman JM, Majhail NS, Klein JP, et al: Relapse and late mortality in 5-year survivors of myeloablative allogeneic hematopoietic cell transplantation for chronic myeloid leukemia in first chronic phase. J Clin Oncol 28:1888-1895, 2010.

4. Hochhaus A, Saglio G, Hughes TP, et al: Impact of treatment with frontline nilotinib vs imatinib on sustained deep molecular response in patients with newly diagnosed chronic myeloid Leukemia in chronic phase. 2015 ASH Annual Meeting. Abstract 2781.

5. Rousselot P, Charbonnier A, Cony-Makhoul P, et al: Loss of major molecular response as a trigger for restarting tyrosine kinase inhibitor therapy in patients with chronic-phase chronic myelogenous leukemia who have stopped imatinib after durable undetectable disease. J Clin Oncol 32:424-430, 2014.

6. Mahon F-X, Boquimpani CM, Takahashi N, et al: Patient-reported quality of life before and after stopping treatment in the ENESTop trial of treatment-free remission for patients with chronic myeloid leukemia in chronic phase. 2016 ASH Annual Meeting. Abstract 1891. Presented December 3, 2016.

Related Articles

Safety of Stopping Imatinib in CML With Undetectable Minimal Residual Disease for at Least 2 Years

With a median follow-up of more than 6 years after treatment discontinuation, the STIM1 study demonstrates that [imatinib] can safely be discontinued in patients with a sustained deep molecular response with no late molecular recurrence.
— Gabriel Etienne, MD, PhD, and colleagues





click me