Disease-free survival

What it is

Disease-free survival (DFS) is the time from randomisation or curative treatment to cancer recurrence or death from any cause, and in many cancer trials it is reported as a 3-year or 5-year endpoint. DFS is a research outcome used mainly in adjuvant settings, meaning after surgery or radiotherapy given with curative intent, when no visible cancer remains on scans or examination. A trial may report that a treatment improved 3-year DFS, for example, meaning a higher proportion of patients were alive without the cancer returning at 3 years.

DFS is not the same as overall survival (OS). OS measures time until death from any cause, while DFS counts an earlier event: recurrence of cancer or death. Because recurrence usually happens before death, DFS can often be measured sooner than OS and may require fewer patients or shorter follow-up. That is one reason DFS is commonly used in oncology trials, especially for early-stage cancers where patients may live many years after treatment.

A simple way to compare common endpoints is:

Endpoint	What counts as an event	Typical setting
Disease-free survival (DFS)	Recurrence of cancer or death	Adjuvant, curative-intent treatment
Overall survival (OS)	Death from any cause	All cancer settings
Progression-free survival (PFS)	Tumour growth/progression or death	Advanced or metastatic cancer
Event-free survival (EFS)	A broader set of events, often including progression before surgery, recurrence, or death	Often neoadjuvant treatment

In modern oncology usage, DFS is generally preferred for adjuvant trials, while EFS is more often used in neoadjuvant settings because the patient may still have measurable disease at the time treatment starts.

How it works

DFS is not a treatment or a test. It is a time-to-event endpoint used in clinical research. Investigators define in advance what counts as a DFS event, when follow-up starts, and how recurrence will be detected. The starting point is often randomisation in a trial, though some studies describe DFS from surgery or completion of primary treatment.

Patients are then followed over time with clinic visits, imaging, pathology, and other assessments appropriate for that cancer type. If the cancer returns locally, regionally, or at a distant site, that usually counts as a DFS event. Death from any cause also usually counts. Patients who have not had an event by the analysis date are "censored," meaning their available follow-up time is included up to their last disease assessment.

DFS is commonly presented in two ways:

1. A landmark percentage, such as 3-year DFS or 5-year DFS.
2. A hazard ratio (HR) comparing two groups over time.

An HR below 1 suggests fewer DFS events in the treatment group than in the comparison group. For example, an HR of 0.80 would suggest a 20% relative reduction in the rate of recurrence or death over the study period, though the exact meaning depends on the trial design and assumptions.

Evidence and uses

DFS is widely used as a primary endpoint in trials of early-stage solid tumours, especially when the goal is to reduce recurrence after surgery. It has been used in breast cancer, colon cancer, lung cancer, pancreatic cancer, and other settings. Regulators and clinicians pay attention to DFS because preventing recurrence can matter directly to patients, and because waiting for OS results may take many years.

The main advantage of DFS is speed. Compared with OS, DFS can often show whether a treatment is active earlier. This can help move useful adjuvant therapies into practice sooner. In cancers with long post-recurrence survival, OS may be delayed by later-line treatments, crossover, or long natural history, making DFS a practical endpoint.

But DFS is not automatically a perfect substitute for OS. Whether DFS is a good surrogate endpoint depends on the cancer type, stage, treatment setting, and how strongly DFS improvements predict longer survival. Evidence supports DFS as a reasonably strong surrogate for OS in some settings, such as stage III colon cancer, and it has also been used in adjuvant breast and lung cancer trials. In other settings, the link is weaker or still uncertain.

A recent meta-analysis in early-stage pancreatic cancer concluded that DFS is frequently used, but its validity as a surrogate for OS remains uncertain. That is an important limitation for patients reading headlines about a "positive" trial. A DFS benefit may be clinically meaningful, but it does not always guarantee that people live longer overall.

For patients and families in India, DFS often appears in reports of adjuvant therapy trials discussed by oncologists at tumour boards and in guideline-based care. The interpretation is the same as elsewhere: DFS helps estimate how long patients remain free of detectable cancer after curative treatment, but it should be read alongside OS, quality of life, toxicity, and cost.

Diagnosis / how it's measured

DFS is measured in a clinical trial or observational study, not diagnosed in an individual person. To calculate it properly, researchers need:

• a clearly defined start date
• a prespecified definition of recurrence
• a schedule for follow-up and surveillance
• rules for counting death and censoring
• statistical methods such as Kaplan-Meier curves and Cox proportional hazards models

Several details can change the reported DFS result:

Factor	Why it matters
Start point	DFS from randomisation may differ from DFS from surgery
Event definition	Some studies include second primary cancers; others do not
Follow-up intensity	More frequent scans may detect recurrence earlier
Censoring rules	Loss to follow-up can affect estimates
Independent review	Central review may reduce bias in event assessment

This is why DFS numbers from different trials are not always directly comparable. A 5-year DFS rate in one study may reflect a different patient population, surveillance schedule, or event definition than another.

When to see a clinician

If you are reading a pathology report, discharge summary, or news article that mentions DFS, ask your oncologist what it means for your cancer type and stage. The same DFS result can have different importance in colon cancer, breast cancer, lung cancer, or pancreatic cancer.

It is also reasonable to ask:

1. Was DFS the main endpoint or a secondary endpoint?
2. Did the study also improve overall survival or quality of life?
3. What were the side effects and treatment burden?
4. How similar were the trial patients to me?

For an individual patient, follow-up care should not be based on the term DFS alone. Decisions about scans, blood tests, adjuvant therapy, and surveillance schedules should come from your treating team.

Limitations and open questions

DFS is useful, but it has limits. First, it depends on how recurrence is defined and detected. If one trial scans patients more often than another, recurrence may be found sooner even if the biology is similar. Second, DFS can be affected by deaths unrelated to cancer, which may matter more in older or medically complex populations.

Most importantly, a better DFS result does not always mean better overall survival, better quality of life, or better value. Some treatments delay recurrence but add substantial toxicity, cost, or long-term side effects. Others may improve DFS in a narrow subgroup but not in the broader population.

Evidence on DFS as a surrogate for OS is strong in some cancers and mixed in others. Researchers continue to study when DFS is reliable enough for drug approval, guideline changes, and patient decision-making. When reading trial results, DFS is best treated as one important piece of evidence rather than the only measure that matters.