ALEX: What the Lung Cancer Trial Found

The question ALEX was built to answer

About one in twenty non-small cell lung cancers carries a rearrangement of a gene called ALK. The gene ends up fused to a neighbor, and the fusion produces a protein that drives the cancer to grow.

Drugs that block that protein already existed. Crizotinib was the standard first one. The problem was where the disease went next. ALK-positive lung cancer has a strong tendency to spread to the brain, and crizotinib crosses into the brain poorly.

Alectinib was designed to get there. ALEX asked whether it should be used first.

Trial at a glance

Field	Detail
Trial	ALEX
Registry number	NCT02075840
Phase	3
Design	Randomized, open-label
Participants	303 adults
Population	Previously untreated advanced ALK-positive NSCLC, including symptom-free brain metastases
Comparison	Alectinib 600 mg twice daily vs crizotinib 250 mg twice daily
Main measure	Investigator-assessed progression-free survival

The amounts are here to describe the comparison that was tested. If you take either drug, use the prescription your own team wrote.

Who took part

The published report describes 303 adults with previously untreated, advanced ALK-positive non-small cell lung cancer, split 152 to alectinib and 151 to crizotinib. People whose cancer had already reached the brain were included, provided it was not causing symptoms — which is unusual, and central to what the trial found.

Median follow-up was 18.6 months in the alectinib group and 17.6 months in the crizotinib group.

What the trial found

Progression or death occurred in 62 of 152 people on alectinib, which is 41%. On crizotinib it was 102 of 151, or 68%.

At twelve months, 68.4% of the alectinib group were alive without their cancer growing, against 48.7% on crizotinib. The hazard ratio was 0.47, with a 95% confidence interval of 0.34 to 0.65 and a p value below 0.001. In plain terms: at any given moment the risk of the cancer growing or the person dying was about half what it was on the older drug. The median time before progression on alectinib had not been reached when the analysis was done.

The brain result was the striking one. Cancer progressed in the brain in 18 alectinib patients — 12% — against 68 crizotinib patients, or 45%.

Tumor shrinkage rates were closer: 82.9% with alectinib and 75.5% with crizotinib, a difference that did not reach statistical significance at p=0.09. Serious side effects, graded 3 to 5, were less common with alectinib at 41% versus 50%.

Why the brain figure mattered so much

For someone living with ALK-positive lung cancer, brain metastases are among the most feared events. They cause headaches, seizures, weakness, and changes in speech and thinking, and treating them has usually meant radiation to the brain with its own lasting effects.

A drug that reaches brain tissue can prevent that rather than react to it. Cutting brain progression from 45% to 12% changed what the years after diagnosis look like, not only how long they last.

When to get checked

ALK-positive lung cancer occurs disproportionately in younger adults and in people who never smoked, which is exactly the group whose symptoms tend to be attributed to something else. Any of these lasting more than three weeks deserves an appointment, whatever your smoking history:

• A cough that will not clear, or a familiar cough that has changed.
• Any blood coughed up, even once.
• Breathlessness on things you used to manage easily.
• Chest or shoulder pain that is present at rest.
• Repeated chest infections settling in the same place.

And if lung cancer is diagnosed, ask whether the tumor has had molecular testing. ALK status cannot be seen on a scan or guessed from the pathology slide. Our overview of lung cancer explains what that testing covers.

Where this sits in the wider picture

SEER, the federal cancer surveillance program, estimates about 229,410 new lung and bronchus cancers in the United States in 2026. ALK-positive disease is a small slice of that, which is why the finding is important to a specific group rather than to everyone with lung cancer.

Across all lung cancers, five-year relative survival is 29.5%, and 51% are found after the disease has already spread. Those are averages over a very mixed population diagnosed years ago, most of it without a targetable mutation. They describe a group and not a person.

What this does not mean

• The trial was open-label. Everyone knew which drug was being given, and the main endpoint was assessed by investigators who also knew. That can bias a judgment about when a scan shows progression.
• Median progression-free survival on alectinib had not been reached, so the true size of the gain was still unknown at this analysis.
• Crizotinib is no longer the comparison that matters. Newer ALK drugs have since been tested against alectinib, so its place has been reopened.
• Participants met specific entry criteria and were well enough to enroll. Results do not automatically transfer to everyone with this cancer.
• A longer time before progression is not the same measure as a longer life. This report did not answer that question.

Questions worth asking

• Has my tumor been tested for ALK, and what did the report say?
• If brain imaging has not been done, should it be?
• What are the side effects of the drug being proposed, and how are they managed?

Our guides to clinical trial phases and finding a clinical trial explain how studies like this are built.

Sources

• N Engl J Med 2017: Alectinib versus Crizotinib in Untreated ALK-Positive NSCLC (ALEX), PMID 28586279 — record retrieved via NCBI eutils
• ClinicalTrials.gov API v2 record for NCT02075840
• NCI PDQ: Non-Small Cell Lung Cancer Treatment (Health Professional Version)
• NCI SEER Cancer Stat Facts: Lung and Bronchus Cancer

How this article was prepared

An AI-assisted editorial system helped prepare this page. No named medical reviewer has reviewed it unless one is listed.

Cancer Explained is published by the National Cancer Information Foundation. It is not medical advice and does not suggest a test or treatment.

See an error, old source, or unclear wording? Tell us.