ESMO 2020 | CROWN study: lorlatinib vs crizotinib in 1L NSCLC

At the ESMO meeting, one of the datasets that we’re all eagerly anticipating was the CROWN clinical trial data. There are now several ALK inhibitors, which are either in development or being used, and one of these is lorlatinib. It’s already licensed to be used either after a next-generation inhibitor or crizotinib, and the key issue is, should this be used upfront? And this was the question that was being answered by the CROWN trial. This trial was an international randomized phase three trial of frontline metastatic ALK+ lung cancer patients who were randomized to receive either crizotinib or lorlatinib with the primary endpoint of progression-free survival by blinded independent central review.

Now we’ve all been waiting a long time for this data because many of us have been using lorlatinib in the relapse setting where we find it a highly effective drug. And the key issue is how much better than crizotinib would it be? We all expect it to be better than crizotinib. And what would the comparative efficacy of lorlatinib be against what we’ve already seen with brigatinib in the ALTA-1L trial and alectinib in the ALEX study?

The primary endpoint that was reported from the CROWN trial was progression-free survival by blinded independent review with a hazard ratio of 0.28, which is a very strongly positive trial. The median survival for the crizotinib arm was 9.3 months and it wasn’t reached for the lorlatinib arm. Now, some might say that the crizotinib arm underperformed given that we’ve seen around 10 to 11 months in the brigatinib and alectinib studies, but personally, I can see no particular reason for that because both brigatinib and crizotinib arms were well balanced. Patients couldn’t have had prior chemotherapy going into this study, unlike the ALTA-1L study, and around 25% of patients had CNS metastases going into the study, which is very similar to ALTA-1L at 25% and much less than the ALEX trial of alectinib at around 40, 45%.

Other interesting secondary endpoints included intercranial response. We know that lorlatinib is a highly effective intracranial disease and in those with measurable intracranial brain metastases, lorlatinib had a response of around 82%, which is highly active. I would say that it compares very favorably to what we’ve already seen with ALTA-1L with a measurable disease intracranial response rate of about 78%.

Many of the outpatients that we see have CNS metastases and the key issue is what’s the efficacy of these drugs in patients with CNS disease at baseline? CROWN reported an overall progression-free survival in patients with CNS disease with a hazard ratio of 0.2, which is really very effective. And in those without CNS disease, a hazard ratio of 0.3. Just putting that in context in ALTA-1L: the hazard ratio for brigatinib versus crizotinib for patients with CNS disease was 0.25, very similar to the 0.2 that we see with lorlatinib, but for patients without CNS disease it was 0.65, similarly for alectinib. Whereas what we’re seeing with lorlatinib is 0.32.

For patients without extracranial disease, this looks like a very active drug and it also looks very effective for patients with intercranial disease, but I would also suggest that brigatinib has a role here as well, given the intracranial data in ALTA-1L. One thing that we need to be mindful of is the rather unusual toxicity profile that lorlatinib has. It has a number of adverse event types that it can cause such as hypercholesterolemia, peripheral neuropathy, peripheral edema, but more importantly, neurocognitive effects. And in the trial, a total of around 20% of treatment-related neurocognitive adverse events were reported.

Overall, in summary, this is a strongly positive trial against crizotinib and definitely lorlatinib is a treatment option for patients with upfront disease. It’s very hard to make cross trial comparisons between alectinib, brigatinib, lorlatinib and of course, we also have encartinib with the eXalt3 trial recently reported. The populations in the trials were very subtly slightly different and the endpoints were very different as well, but undoubtedly, lorlatinib represents a treatment option. I guess one of the things that oncologists were all thinking about is the unusual neurocognitive adverse events that we see with lorlatinib and in practice, one would have to balance those against the efficacy that we’ve seen.