Background
Phase III RCT (CROWN), 5-year analysis. N=296. Untreated advanced ALK-positive NSCLC. First-line lorlatinib vs crizotinib.
Interventions and follow up
Arm A: Lorlatinib 100 mg PO once daily, continuous
Arm B: Crizotinib 250 mg PO BID, continuous
Primary endpoint: PFS
mFollow up: ~5 yr (60.2 mo lorlatinib)
Arm B: Crizotinib 250 mg PO BID, continuous
Primary endpoint: PFS
mFollow up: ~5 yr (60.2 mo lorlatinib)
Results
5-yr PFS: 60% vs 8%, HR 0.19 (95% CI 0.13–0.27; P<.001)
5-yr OS: 77% vs 63%, HR 0.72 (95% CI 0.48–1.09; P=.12, NS)
ORR: 81% vs 63% (CR 10% vs 2%; PR 70% vs 61%)
Median time to intracranial progression: NR vs 16.4 mo (95% CI 12.7–21.9), HR 0.07 (95% CI 0.03–0.12)
Baseline brain mets — mPFS: NR vs 6.0 mo, HR 0.08 (95% CI 0.04–0.19); 5-yr PFS 53% vs not evaluable; intracranial ORR 60% vs 11%
No baseline brain mets — mPFS: NR vs 10.8 mo, HR 0.24 (95% CI 0.16–0.36); 5-yr PFS 63% vs 10%; intracranial ORR 92% vs 33%
5-yr OS: 77% vs 63%, HR 0.72 (95% CI 0.48–1.09; P=.12, NS)
ORR: 81% vs 63% (CR 10% vs 2%; PR 70% vs 61%)
Median time to intracranial progression: NR vs 16.4 mo (95% CI 12.7–21.9), HR 0.07 (95% CI 0.03–0.12)
Baseline brain mets — mPFS: NR vs 6.0 mo, HR 0.08 (95% CI 0.04–0.19); 5-yr PFS 53% vs not evaluable; intracranial ORR 60% vs 11%
No baseline brain mets — mPFS: NR vs 10.8 mo, HR 0.24 (95% CI 0.16–0.36); 5-yr PFS 63% vs 10%; intracranial ORR 92% vs 33%
Adverse events
Constitutional: Edema 57% vs 43%, weight increased 44% vs 13%, fatigue 30% vs 33%, arthralgia 28% vs 14%, back pain 19% vs 14%, myalgia 14% vs 4%, rash 12% vs 8%
Lab changes: Hypercholesterolemia 72% vs 4%, hypertriglyceridemia 66% vs 6%, ALT increased 19% vs 35%, AST increased 16% vs 27%, hyperglycemia 14% vs 4%, anemia 25% vs 10%
Cardiovascular and pulmonary: Hypertension 26% vs 4%, dyspnea 23% vs 18%, overall cardiovascular AEs 28% vs 28%, embolic/thrombotic events 9% vs 11%, pneumonia 11% vs 9%
CNS effects: Peripheral neuropathy 44% vs 16%, cognitive effects 28% vs 7%, mood effects 21% vs 6%, headache 22% vs 20%, vision disorder 19% vs 40%, speech effects 6% vs 0%, psychotic effects 5% vs 0%
Lab changes: Hypercholesterolemia 72% vs 4%, hypertriglyceridemia 66% vs 6%, ALT increased 19% vs 35%, AST increased 16% vs 27%, hyperglycemia 14% vs 4%, anemia 25% vs 10%
Cardiovascular and pulmonary: Hypertension 26% vs 4%, dyspnea 23% vs 18%, overall cardiovascular AEs 28% vs 28%, embolic/thrombotic events 9% vs 11%, pneumonia 11% vs 9%
CNS effects: Peripheral neuropathy 44% vs 16%, cognitive effects 28% vs 7%, mood effects 21% vs 6%, headache 22% vs 20%, vision disorder 19% vs 40%, speech effects 6% vs 0%, psychotic effects 5% vs 0%
Conclusions
At 5 years, lorlatinib delivered unprecedented PFS (60% event-free) and intracranial control vs crizotinib in advanced ALK-positive NSCLC, setting a new 1L benchmark, balanced against a distinctive CNS and metabolic toxicity profile.
Key Limitations
OS not statistically significant (P=.12), confounded by high crossover. Crizotinib comparator now outdated; no head-to-head vs alectinib. CNS/cognitive and metabolic toxicities require active management. Optimal use of biomarker-driven de-escalation unproven.
Clinical Context
FDA approved lorlatinib 1L ALK+ NSCLC (2021); CROWN 5-yr data reinforce preferred 1L positioning. ESMO endorses lorlatinib (and alectinib/brigatinib) as preferred 1L ALK TKIs. Strong CNS activity favors lorlatinib in patients with brain metastases.