First-line treatment with the third-generation ALK TKI was also associated with higher overall and intracranial response rates in patients with ALK-positive non–small cell lung cancer.
Frontline lorlatinib (Lorbrena) significantly improved progression-free survival (PFS), compared with crizotinib (Xalkori) in patients with ALK-positive non–small cell lung cancer (NSCLC), according to data from a planned interim analysis of the phase 3 CROWN trial that were presented at the 2020 ESMO Virtual Congress.
The third-generation ALK TKI was also associated with higher overall and intracranial response rates.
“These results of the CROWN study support the use of lorlatinib as a highly effective first-line therapy for patients with advanced ALK-positive non–small cell lung cancer,” lead study author Ben Solomon, MD, a medical oncologist in the Lung Service and the Head and Neck service, and group leader of the Molecular Therapeutics and Biomarkers Laboratory in the Research Division, Peter MacCallum Cancer Centre in Australia, said in a presentation during the meeting.
The median PFS by blinded independent central review (BICR) was not estimable (NE; 95% CI, NE-NE) for lorlatinib compared with 9.3 months (95% CI, 7.6-11.1) for crizotinib, leading to a 72% reduction in the risk of disease progression or death (HR, 0.28; 95% CI, 0.19-0.41; 1-sided P <.001).
The intracranial objective response rate (IC-ORR) was 82% in patients with measurable brain metastases, with a complete response (CR) rate of 71% with lorlatinib, along with a significantly prolonged time to IC progression, versus crizotinib.
ALK rearrangements occur in approximately 3% to 5% of patients with NSCLC that result in sensitivity to small-molecule ALK TKIs. Furthermore, resistance to ALK TKIs is a common development and often includes central nervous system (CNS) progression.
Lorlatinib is a highly potent, brain-penetrant, third-generation ALK TKI that is currently approved by the FDA as a treatment for patients with ALK-positive metastatic NSCLC following progression on 1 or more prior ALK-targeted TKI.
The phase 3 CROWN trial randomized 296 patients 1:1 with lorlatinib at 100 mg daily (n = 149) compared with crizotinib (Xalkori) at 250 mg twice daily (n = 147) as a first-line treatment for patients with ALK-positive NSCLC. Patients were stratified by presence of brain metastases and ethnicity (Asian vs non-Asian). No crossover between the 2 arms was permitted, and 5 patients on the crizotinib arm did not receive treatment.
To be eligible for enrollment, patients had treatment-naïve stage IIIB/IV disease, an ECOG performance status of 0 to 2, and at least 1 extracranial measurable target lesion via RECIST v1.1 criteria with no prior radiation required. Asymptomatic treated or untreated CNS metastases were permitted.
All patients underwent restaging chest/abdominal CT scans and brain imaging every 8 weeks. At the interim data cutoff date of March 20, 2020, 103 patients on lorlatinib and 31 patients on crizotinib remained on study treatment.
The primary end point was PFS by BICR; secondary end points were investigator-assessed PFS, BICR- and investigator-assessed ORR, IC-ORR, duration of response (DOR), and IC-DOR via BICR, IC-time to progression by BICR, OS, safety, quality of life.
The baseline characteristics were well balanced between the 2 arms. More than half of patients were female (59%), 48.5% were White, and 54% of patients had an ECOG performance status of 1. Fifty-nine percent of patients were never smokers, 33% were prior smokers, and 7.5% were current smokers. Ninety-three percent of patients had stage IV disease, and 26.5% of patients had brain metastases at baseline; 6.5% of patients had prior brain radiotherapy.
The median duration of follow-up for PFS was 18.3 months and 14.8 months, respectively. Additional results showed that the 12-month PFS rate, assessed by BICR, with lorlatinib was 78% and 39% with crizotinib.
The investigator-assessed median PFS was NE (95% CI, NE-NE) with lorlatinib compared with 9.1 months (95% CI, 7.4-10.9) with crizotinib, leading to a 79% reduction in the risk of disease progression or death (HR, 0.21; 95% CI, 0.14-0.31; 1-sided P <.001). The 12-month PFS rates were 80% and 35% with lorlatinib and crizotinib, respectively.
The PFS benefit by BICR was upheld across prespecified subgroups, including presence of brain metastases (HR, 0.20), ECOG performance status (HR, 0.28), gender (HR, 0.31 for male; HR, 0.26 for female), age (HR, 0.22 for younger than 65 years; HR, 0.35 for older than 65 years), smoking status (HR, 0.24 for never smokers; HR, 0.36 for current/former smokers), and histology (HR, 0.26).
Also by BICR, the ORR was 76% (95% CI, 68-83)with lorlatinib and 58% (95% CI, 49-66) with crizotinib (odds ratio [OR] 2.25; 95% CI, 1.35-3.89). In the lorlatinib arm, there was a 3% CR rate and a 73% partial response rate; 13% of patients had stable disease. In the crizotinib arm, these rates were 0%, 58%, and 28%, respectively. Seven percent and 5% of patients on lorlatinib and crizotinib, respectively, had progressive disease; 4 and 11 patients on each arm were not evaluable. The median DOR was not evaluable with lorlatinib and 11.0 months with crizotinib. The median time to response was 1.8 months in both arms.
IC-OR, via BICR, was 66% in patients on lorlatinib versus 20% with crizotinib with measurable or nonmeasurable brain metastases at baseline (OR, 8.41; 95% CI, 2.59-27.23); the IC-CR rates were 61% and 15%, respectively. The IC-OR was 82% in patients on lorlatinib versus 23% with crizotinib only with measurable brain metastases (OR, 16.83; 95% CI, 1.95-163.23); the IC-CR rates were 71% and 8%, respectively. The median DOR was not evaluable in either lorlatinib group and was 9.4 months in the both measurable/nonmeasurable group and 10.2 months in the only-measurable brain metastases group on crizotinib.
The median time to CNS progression was not estimable with lorlatinib and 16.6 months with crizotinib (HR, 0.07; 95% CI, 0.03-0.17; 1-sided P <.001).
OS data are still immature, Soloman said, but added that the median OS was not estimable in both arms and favored lorlatinib (HR, 0.72; 95% CI, 0.41-1.25).
The median treatment duration was not estimable with lorlatinib and 9.6 months with crizotinib. Regarding safety, grade 3/4 AEs occurred in 72% of lorlatinib-treated patients versus 56% of those on crizotinib; serious AEs were reported in 34% and 27% of patients, respectively. Solomon noted that the majority of grade 3/4 AEs on the lorlatinib arm were laboratory abnormalities regarding cholesterol and triglycerides, that were asymptomatic and readily managed.
AEs occurring in the lorlatinib arm at more than 10% than crizotinib hypocholesterolemia, hypertriglyceridemia, edema, weight gain, peripheral neuropathy, and cognitive effects. AEs more frequently associated with crizotinib included diarrhea, nausea, vision disorder, vomiting, increased aspartate aminotransferase/alanine aminotransferase, and constipation.
There were 7 fatal AEs in each arm; 7% and 49% of patients on lorlatinib had AEs that led to treatment discontinuation and temporary dose interruption, respectively, versus 9% and 47% of patients on crizotinib. Additionally, the change from baseline in global QoL scores was 4.65 (95% CI, 1.14-8.16), favoring lorlatinib.
Solomon B, Bauer T, de Marinis F, et al. Lorlatinib vs crizotinib in the first-line treatment of patients with advanced ALK-positive non–small cell lung cancer: results of the phase 3 CROWN study. Presented at: 2020 ESMO Virtual Congress; September 19-21, 2020; Virtual. Abstract LBA2.