Taking the Guesswork Out of Stopping TKIs

Multiple trials have now concluded that about half of all patients with chronic myeloid leukemia (CML) who have been taking imatinib (Gleevec; Novartis) long enough to achieve a durable, deep molecular response can successfully stop taking the drug. Those findings prompted the National Comprehensive Cancer Network (NCCN) to update its treatment guidelines for 2017 to include criteria for stopping therapy, which could spare patients both side effects and expensive prescriptions.

However, some experts feel the guidelines leave out many patients who would have a high chance of success if they were to give discontinuation a try. The challenge for the CML community now is to better understand why some patients but not others successfully discontinue tyrosine kinase inhibitor (TKI) therapy, and then harness that knowledge to refine the treatment recommendation and develop more potent induction regimens. “In the future, it's very likely there will be modifications to [the guidelines] as more data become available,” concedes Michael Deininger, MD, PhD, of the University of Utah Huntsman Cancer Institute in Salt Lake City, who vice-chairs the NCCN's CML panel.

About 40% of patients who take imatinib will eventually meet the eligibility guidelines to stop treatment—defined by the NCCN as having sustained BCR–ABL1 transcript levels at least 10,000 times below pretreatment levels for 2 years or longer. Combined with the 50-50 chance of molecular relapse, the overall success rate hovers at around 20%.

One strategy for boosting that rate is to use second-generation TKIs such as nilotinib (Tasigna; Novartis) or dasatinib (Sprycel; Bristol-Myers Squibb). Studies have shown that these drugs produce deep molecular responses in closer to 60% of patients, but it wasn't known whether rates of treatment-free remission (TFR) would be the same as with imatinib.

The results are now in: At the European Hematology Association (EHA) Congress in Madrid, Spain, in May, clinicians presented long-term follow-up results from discontinuation trials of nilotinib and dasatinib showing that about 50% of patients remain in TFR for 96 weeks or longer following drug cessation. “It does suggest that using the second-generation drugs will actually enable us to improve the current success rate,” says Tim Hughes, MD, of the South Australian Health and Medical Research Institute in Adelaide, who presented some of the findings.

One nilotinib discontinuation study also points to the possibility that patients on a second-generation TKI might need to take their drug for only 3 or 4 years, instead of twice that duration for imatinib, to achieve a response deep enough for successful discontinuation. But is 3 years, the minimum recommended by the NCCN, sufficient? “We still don't know the cutoff or the best threshold,” says Francois-Xavier Mahon, MD, PhD, of the University of Bordeaux in France.

It's thought that successful discontinuation rests on the ability of host factors to keep leukemic stem cells in check after TKI therapy is halted. The goal now is to figure out what those factors are.

The most promising biomarkers to date come from the EURO-SKI study, which has implicated natural killer cells, dendritic cells, and a transporter protein involved in drug efflux with the likelihood of relapse. However, the prognostic value of these potential biomarkers has been modest at best, and they remain to be validated in other cohorts or with second-generation TKIs.

Another way to get more patients into TFR might be to combine TKIs with other therapies to achieve ever-deeper molecular responses ahead of drug cessation. A German trial group is testing that idea by adding interferon or placebo to nilotinib. In the United States, researchers plan to launch a placebo-controlled study of ruxolitinib (Jakafi; Incyte), a JAK inhibitor that targets leukemic stem cells in the bone marrow. Kendra Sweet, MD, of the Moffitt Cancer Center in Tampa, Florida, has already tested this strategy in an 11-patient pilot study. “It appears to be very well tolerated,” she says.

The safety of any add-on therapies is paramount because most patients with CML experience only low-grade, manageable side effects with TKIs and the medical risks of discontinuation are minimal. A recent meta-analysis found that only one of 509 patients had progressed to blast crisis 2 years after stopping imatinib, indicating that most patients can simply restart therapy if their BCR–ABL1 levels start to rise (Eur J Cancer 2017;77:48–56). The only major side effect is a withdrawal syndrome marked by muscle and joint pain, a condition that is usually transient.

The psychological hurdles are another matter, though. “It's truly, truly emotionally taxing for the patient,” says Ehab Atallah, MD, of the Medical College of Wisconsin Cancer Center in Milwaukee. Only around half of all surveyed patients with CML say they'd be willing to try discontinuation given that the current likelihood of success is about as good as a coin toss. Most cite the fear of possible disease recurrence as the reason against it.

An alternate strategy to achieve the goal of reduced TKI use while minimizing the risk of relapse might be drug de-escalation. A British team recently found that halving the dose of TKI for 12 months led to recurrence in only 2% of patients who experienced deep molecular responses (Lancet Haematol 2017;4:e310–e316).

The trial investigators then took the remaining 98% of patients off their medications completely. As reported at the EHA Congress, 77% were still in TFR a year later—a success rate that was unexpectedly high given what others had seen in prior discontinuation studies without a de-escalation phase.

“It's a really interesting result, but it's one that we can't explain yet,” says trial investigator Mhairi Copland, MD, PhD, of the University of Glasgow in the UK. Confirmatory studies are needed, she says, adding, “We've now got more questions we need to answer.” –Elie Dolgin