Adjuvant crizotinib did not prolong DFS in resected ALK-positive NSCLC

A hazard ratio of 1.08, with median disease-free survival numerically longer under observation

Among the 153 patients whose tumours were centrally confirmed as ALK-positive, median disease-free survival was 74.6 months (95% CI 71.2 to not assessable) in the crizotinib group and 106.2 months (67.8 to not assessable) in the observation group. The hazard ratio was 1.08, presented with a 90% CI of 0.67 to 1.73 and a 95% CI of 0.61 to 1.90, and p=0.80.

The confidence intervals are wide and the upper bound of the median in each group was not assessable, so the numerical advantage under observation should not be read as harm from crizotinib. A hazard ratio sitting at 1.08 with a p value of 0.80 describes two groups the trial could not separate. Of the 153 patients analysed, 99 (65%) were female, 54 (35%) were male and 121 (79%) were White.

Accrual stopped when adjuvant alectinib was approved, at 166 patients of 168 planned

Enrolment ran from August 2014 to May 2024, and was halted when the US Food and Drug Administration approved adjuvant alectinib for resected ALK-positive NSCLC. By that point 166 patients had been enrolled against a target of 168, so the trial came close to its planned size despite the early stop.

Nearly a decade of accrual for 166 patients describes the practical problem of studying an adjuvant question in a molecular subgroup. The eligible population was narrow by protocol: resected disease of 4 cm or larger or with positive lymph nodes, negative surgical margins, no neoadjuvant therapy and a performance status of 0 or 1. Over the same period the treatment available to those patients changed, and the approval of adjuvant alectinib is what closed enrolment.

A control group that changed from double-blind placebo to observation

Patients were randomly assigned 1:1 by a computer-generated sequence, stratified by stage, previous radiation therapy and sex. The control group was initially a double-blind placebo arm and was changed to observation during the trial. Safety was assessed in all patients whose tumours tested ALK-positive and who received the study drug.

The switch to observation removes blinding from the second half of the trial’s accrual. The abstract does not describe how disease-free survival events were ascertained, nor how many patients were enrolled under each version of the design, so the size of that change cannot be read from it. Adverse events are reported for the crizotinib group only, which is consistent with a control group that received no study drug.

Grade 3 or higher events in 58% of patients treated with crizotinib

Safety was assessed in the patients whose tumours tested ALK-positive and who received the study drug, a smaller group than the 85 assigned to crizotinib. Among them, grade 3 or higher adverse events of any attribution occurred in 46 patients (58%), most commonly diarrhoea in eight (10%), oedema in four (5%) and hypertension in four (5%). One death occurred that was not deemed treatment-related. Serious adverse events occurred in 21 patients (27%), most commonly dyspnoea in three (4%), with abdominal pain, thromboembolic events, diarrhoea, dehydration and hypertension each in two (3%).

These figures describe up to two years of exposure in patients who have no detectable disease, which is the calculation an adjuvant treatment always asks a patient to make. Without a disease-free survival benefit on the other side of it, the toxicity has nothing to purchase.

“Adjuvant crizotinib does not prolong DFS in patients with surgically resected ALK-positive NSCLC.”Gerber et al., The Lancet Respiratory Medicine (2026)

A negative result for crizotinib, and an approval that went to a different ALK inhibitor

Crizotinib is described in the paper as an established first-generation ALK inhibitor approved for advanced ALK-positive NSCLC. The negative adjuvant result here concerns that drug, and the approval that closed the trial was granted to a different one. What E4512 establishes is that an ALK inhibitor’s record in advanced disease does not by itself carry into the adjuvant setting.

The authors state that crizotinib should not be recommended as adjuvant therapy for patients with resected ALK-positive NSCLC. The abstract does not report overall survival, and it gives no subgroup analyses by stage or by receipt of adjuvant chemotherapy, so whether any group fared differently cannot be read from it.