Jorge Cortes, MD:
ASCEMBL is a multicenter, open-label phase III trial that recruited patients with chronic-phase CML (CP-CML) who had received ≥ 2 prior TKIs with resistance or intolerance to the most recent TKI. The patients could not have T315I or V299L alterations.³³ The patients were randomized to receive asciminib or the comparator, bosutinib. Asciminib was administered at 40 mg twice daily—a dose associated with very high response rates in a phase I trial that my colleagues and I³⁷ published in the New England Journal of Medicine. Bosutinib was administered at the standard dose of 500 mg daily in this setting.³⁸ Of importance, the study design permitted patients on the bosutinib arm to cross over to asciminib if they experienced treatment failure.³³

The primary endpoint is the rate of major molecular response (MMR) at 24 weeks, and this is the first analysis of that outcome. I want to reiterate that patients with T315I were excluded, which is important because bosutinib is not active against T315I. In contrast, both in vitro and in vivo asciminib are active in patients with T315I.³⁷

The patient characteristics were generally well balanced between arms, but several imbalances may be meaningful.³³ One is the reason for discontinuation of the last TKI. The bosutinib arm had more patients who discontinued therapy because of resistance compared with intolerance (71.1% vs 28.9%, respectively; in the asciminib arm, 60.5% vs 37.6%). Patients who discontinued due to resistance tend to have more difficult-to-treat disease. Two, the bosutinib arm also had more patients who had received ≥ 3 vs only 2 TKIs (60.5% vs 39.5%, respectively; in the asciminib arm, 47.8% vs 52.2%). Again, that represents a patient group that is more difficult to treat.

Also, the rate of any BCR-ABL1 mutation (12.7% to 17.1%) was slightly lower in this study population than I would have expected in a heavily treated population. Other studies assessing patients with CP-CML have reported a BCR-ABL1 mutation rate of 31.6% in those who had failed TKIs and 36% in those who had developed imatinib resistance.^39,40 That being said, the mutations were centrally assessed and well balanced between the arms of ASCEMBL.

Jorge Cortes, MD:
The study met its primary endpoint of MMR at 24 weeks.³³ Asciminib significantly improved the rate of MMR at 24 weeks vs bosutinib (25.5% vs 13.2%, respectively; common treatment difference: 12.2%; 95% CI: 2.19-22.3; P = .029). Essentially, asciminib doubled the rate of MMR at 24 weeks.

Now, what is the importance of achieving MMR at 24 weeks? For the study, it was important to have an early readout. Of course, it will also be important to see the results that will emerge with continuation of the study, and I anticipate that this difference in MMR rates will continue to expand over time, although the crossover may limit our ability to detect the true difference.

Jorge Cortes, MD:
Subgroup analyses of the 24-week MMR rate indicated a benefit with asciminib across all subsets defined by baseline demographics or disease characteristics.³³ The differences were not as meaningful in several groups (eg, in male patients). The most notable lack of difference was in patients with intolerance to their prior TKI, where the 24-week MMR rate was 33.9% with asciminib vs 31.8% with bosutinib.

These data indicate that asciminib may be more beneficial in patients who have received more TKIs vs those who have received fewer TKIs. Essentially, the more heavily treated the patient was, the more they benefited from asciminib vs bosutinib.

Jorge Cortes, MD:
The investigators also performed an analysis of treatment effect adjusted by major cytogenetic response at randomization.³³ This adjusted analysis significantly favors asciminib vs bosutinib, with an odds ratio of 2.35 (95% CI: 1.08-5.12). When the investigators adjusted by major cytogenetic response and other important variables (eg, intolerance vs failure as the reason for discontinuation of last prior TKI), the benefit with asciminib persisted.

Jorge Cortes, MD:
The other efficacy outcomes all favored asciminib vs bosutinib.³³ Among patients who were not in complete cytogenetic response at baseline, the rate of those achieving complete cytogenetic response at Week 24 was 40.8% with asciminib vs 24.2% with bosutinib (common treatment difference: 17.3%; 95% CI: 3.62-31.0). The MR⁴ and MR^4.5 response rates were relatively low, but bear in mind that these were assessed at 24 weeks, so it was actually rather impressive to have an MR^4.5 rate at this early timepoint of 8.9% with asciminib vs 1.3% with bosutinib.

Jorge Cortes, MD:
Regarding safety, there was not much of a difference in AEs with asciminib vs bosutinib.³³ There was slightly more thrombocytopenia with asciminib (28.8% vs 18.4% with bosutinib). Consistent with its established safety profile, bosutinib exhibited higher rates of gastrointestinal toxicity, rash, and elevated liver function tests.³⁸

Jorge Cortes, MD:
The trial also looked at the very important safety issue of arterial occlusive events.³³ Of the second- and third-generation TKIs, bosutinib generally has the lowest rate of arterial occlusive events.⁴¹ In ASCEMBL, there were more arterial occlusive events with asciminib vs bosutinib (3.2% vs 1.3%, respectively).³³ In the asciminib arm, the investigators observed myocardial ischemia, coronary artery disease, ischemic stroke, and mesenteric artery embolism/thrombosis. These are important AEs that must be closely monitored for and managed in this population.

Jorge Cortes, MD:
The investigators also performed an interesting analysis of BCR-ABL1 mutations.³³ Curiously enough, at the end of treatment in the asciminib arm, there was emergence of 2 mutations in the ATP-binding site and 2 mutations in the myristoyl-binding pocket. Preclinical data would predict the emergence of these mutations with asciminib as opposed to the traditional TKIs. We will need to determine the clinical implications of these mutations and how they affect efficacy if a patient moves from asciminib to a traditional TKI.

Jorge Cortes, MD:
This analysis focused on the dose-expansion cohort with T315I-mutated CML in an ongoing phase I trial.⁴⁵ Note that the asciminib dose used in this cohort was higher than the 40 mg twice daily used in the ASCEMBL trial.³³ Patients with T315I-mutated disease need higher concentrations, and the earlier analysis of this study indicated that asciminib was safe at a dose of 200 mg twice daily.³⁷

This analysis included 52 patients with T315I-mutated CML, of whom 48 had chronic-phase and 4 had accelerated‑phase disease.³³ Patients had received a number of prior TKIs, with 59.6% having received ponatinib.

If we separate the patients into those who were ponatinib naive vs ponatinib pretreated, several characteristics stand out. One, the median time from diagnosis was markedly longer for the ponatinib-pretreated patients, perhaps because they had spent some time already on ponatinib. Two, the number of prior therapies was higher among the ponatinib‑pretreated patients. Indeed, 29.0% of these patients had received ≥ 4 prior TKIs, whereas no ponatinib-naive patients had received that many. Three, there was a trend toward higher BCR-ABL1^IS transcript levels at baseline in the ponatinib‑pretreated patients.

Jorge Cortes, MD:
This analysis looked at the rate of MMR, defined as BCR-ABL1^IS ≤ 0.1%.⁴⁵ The overall MMR was 46.9%, which was quite high for such a heavily pretreated population. Most patients achieved a response fairly early, with 40.8% achieving MMR by 24 weeks. The rate of those achieving MMR increased some over time, emphasizing that it is important to continue patients on treatment. That being said, the patients who responded tended to respond quickly, as evidenced by a median time to MMR of 12.1 weeks.

Jorge Cortes, MD:
We also assessed the 24-week MMR rate and depth of response in patients who were ponatinib naive vs ponatinib pretreated.⁴⁵ As expected, the 24-week MMR rate was higher in the ponatinib-naive patients (57.8% vs 28.6% in the ponatinib-pretreated patients). Nonetheless, I am impressed by such a high response rate in the ponatinib-pretreated patients.

We also observed deeper molecular responses in the ponatinib‑naive patients, with 33.3% having achieved MR^4.5 vs 10.7% of those who were ponatinib pretreated. In the overall population, the median time to MR⁴ and MR^4.5 was 20 weeks—a little longer than the median time to MMR, as would be expected, but still relatively quick.

Jorge Cortes, MD:
We also looked at the emergence of BCR-ABL1 mutations with asciminib in this cohort.⁴⁵ The left table presents results in patients who did not achieve MMR, of whom 6 of 26 acquired BCR-ABL1 mutations, mostly in the ATP-binding pocket (eg, F359 and M244). There was also a mutation in the myristoyl pocket, E453Q. Only 1 of the 6 acquired mutations occurred in a ponatinib-naive patient.

Among the 2 patients who lost MMR, 1 developed the F359V mutation in the ATP-binding pocket.

Jorge Cortes, MD:
As of data cutoff (April 2, 2020), treatment was ongoing for 67.3% of the patients.⁴⁵ That high rate speaks for the durability of the response. Indeed, the median dose intensity (399 mg/day) was quite close to the intended dose of 400 mg daily. These data suggest that the patients continued to do well on asciminib for a long period of time.

Jorge Cortes, MD:
These safety data again indicate that asciminib is well tolerated.⁴⁵ As in the ASCEMBL trial, we observed some thrombocytopenia, with 17.3% of patients experiencing grade ≥ 3 thrombocytopenia. The other notable AE was lipase elevation, which occurred at grade ≥ 3 in 15.4% of patients. These data underscore the importance of monitoring patients on asciminib.

This was a safe drug overall, with no deaths observed on study. We did observe 3 on-treatment arterial occlusive events, with 1 patient experiencing both ischemic stroke and cerebral ischemia. It is important that all patients receiving any TKIs undergo monitoring and assessment for arterial occlusive events.

Jorge Cortes, MD:
The phase II OPTIC trial recruited 283 patients with CP-CML and ≥ 2 prior TKIs or with a T315I mutation.⁴⁷ Patients were randomized to a starting ponatinib dose of 45 mg daily (the standard dose for ponatinib), 30 mg daily, or 15 mg daily, with patients in the first 2 arms reducing their dose to 15 mg daily upon achieving ≤ 1% BCR-ABL1^IS. The primary endpoint was the rate of patients achieving ≤ 1% BCR-ABL1^IS at 12 months.

Patient characteristics were well balanced across all arms. The trial permitted patients to have risk factors for arterial occlusive events as long as these were controlled at the time of study entry. Regarding cardiovascular risk factors at baseline, arterial hypertension was present in 23% to 28%, hypercholesterolemia in 16% to 20%, and diabetes in 3% to 7%. This was a study population with very highly resistant disease, with 59% to 65% of patients having had, at best, a complete hematologic response but not a cytogenetic response to the last prior TKI.

Regarding BCR-ABL1 mutation status, 21% to 27% of patients had T315I, and 15% to 19% had other mutations. In addition, 51% to 60% of patients had received ≥ 3 TKIs, and 60% to 68% had received ≥ 2 second-generation TKIs.

Jorge Cortes, MD:
The 12-month rate of patients achieving ≤ 1% BCR-ABL1^IS was markedly higher in patients who received the 45-mg starting dose.⁴⁷ The rate was 48% in the 45-mg group, dropping to 35% with 30 mg and then 23% with 15 mg. This dose-response trend is particularly pronounced in patients with a T315I mutation, where the rate was 60% with 45 mg, 25% with 30 mg, and only 6% with 15 mg. Patients with the most refractory disease had the greatest benefit. In those who had only a complete hematologic response to the last prior TKI, the response rate was 43% with 45 mg, 22% with 30 mg, and 15% with 15 mg. Those who had received ≥ 3 prior TKIs exhibited response rates of 49%, 30%, and 22%, respectively.

Overall, the 45-mg starting dose was associated with higher response rates, particularly in those with the most challenging disease.

Jorge Cortes, MD:
Shown here are the very important safety data for this trial.⁴⁷ This study had an independent adjudication committee who reviewed all arterial occlusive events and determined whether they met the criteria for an arterial occlusive event per American College of Cardiology/American Heart Association criteria.⁴⁷ The committee was blinded to the dose, any dose modification, and the investigator’s opinion on causality. Thus, I would consider the data on adjudicated treatment-emergent arterial occlusive events to be quite robust.

Overall, the rate of adjudicated arterial occlusive events was relatively low (2% to 6%).⁴⁷ I was surprised that there was not much of a difference between the 45-mg and 30-mg dose arms. Looking at patients who had received ≥ 3 TKIs, the rate was 6% with 45 mg, 5% with 30 mg, and 2% with 15 mg. For those less heavily pretreated (≤ 2 TKIs), the rates were 5%, 3%, and 0%, respectively. We observed a slight dose response for this event, but the overall rate is relatively low—perhaps because the mandatory dose reduction once the patient achieved a response was reducing the risk of arterial occlusive events.

Jorge Cortes, MD:
An ongoing challenge in the management of CP-CML is that patients who failed on a second-generation TKI are unlikely to derive benefit from another second-generation TKI.⁴⁸ This pooled analysis of the OPTIC and PACE trials evaluated the efficacy and safety of ponatinib, a third-generation TKI, in patients with CP-CML who had failed ≥ 1 second-generation TKI.⁴⁹ We just discussed the design of the randomized phase II OPTIC trial, so I will focus briefly on the design of the single-arm phase II PACE trial.

The PACE trial enrolled a very high-risk population consisting of 257 adults with CML that was either resistant or intolerant to ≥ 2 TKIs or that had BCR-ABL1 T315I mutation and > 1% BCR-ABL1^IS.⁵⁰ Patients received ponatinib at 45 mg/day. We should note that this pooled analysis included only those patients in the OPTIC trial who received the 45-mg starting dose, which reduced to 15 mg/day upon achievement of ≤ 1% BCR-ABL1^IS.⁴⁹ The pooled analysis assessed the efficacy and safety of ponatinib in 350 patients with CP-CML.

The patient baseline characteristics were very similar between the trials, although the OPTIC trial had a younger patient population (median age: 46 vs 61 years in the PACE trial). The PACE trial had a very high‑risk population that represented the oldest population I am aware of on any TKI study and that had been treated with a greater number of second-generation TKIs. PACE trial patients also had higher rates of cardiovascular risk factors—hypertension, diabetes, and hypercholesterolemia—than did those in the OPTIC trial. Both trials had patient populations with very refractory disease, and 56% of patients in the pooled population had achieved less than or equal to a complete hematologic response as the best response to prior TKI.

Jorge Cortes, MD:
If we look at all patients in each trial, we can see the consistency of the results between trials. For example, the rate of response (defined as ≤ 1% BCR-ABL1^IS) at 12 months was 46% in PACE and 52% in OPTIC.⁴⁹ The 2-year PFS rate was 67% in PACE and 81% in OPTIC. The 2-year OS rate was 85% and 93%, respectively, and was maintained even in patients who had received ≥ 2 TKIs—generally an older population. The results were also consistent in the subgroup who had received 1 TKI.

Although the results were generally consistent between trials, I did notice that response rates were slightly higher in the OPTIC trial. This was likely due to OPTIC’s younger patient population and patients staying on therapy longer because the dose reduction improved tolerability.

Jorge Cortes, MD:
When we look at efficacy in subgroups defined by baseline mutation status, we saw response regardless of mutation status and with consistency between trials.⁴⁹ For example, in patients with T315I mutation, the response rate at 12 months was 56% in PACE and 63% in OPTIC. In patients with no baseline mutations, the response rate at 12 months was 36% and 38%, respectively.

Jorge Cortes, MD:
This analysis also focused on the important safety concern of arterial occlusive events.⁴⁹ The rate of grade 3/4 treatment-emergent AEs that were adjudicated was 16% in PACE and 3% in OPTIC, likely due to the dose reduction and patient selection.

Keeping in mind that the median follow-up was longer in PACE (57 vs 21 months with OPTIC), we can focus on the rates of treatment-emergent arterial occlusive events in the first and second year, for which we have enough data from both trials. The rate of treatment-emergent arterial occlusive events was much lower in OPTIC vs PACE. Indeed, at 1 to < 2 years, the overall rate was 2% in OPTIC vs 15.1% in PACE. This analysis emphasizes that by better selecting patients and adjusting the ponatinib dose, we can mitigate the risk of these events in our patients.