A measurable marker can be useful because it is easier to observe than the outcome researchers ultimately care about. The important question is whether changing that marker tells us enough about the intended outcome. A technically reliable measurement and a valid substitute for an outcome are different achievements.
Distinguish a marker from a substitute endpoint
FDA describes surrogate endpoints as substitutes for clinical outcomes and distinguishes them from direct measures of how people feel, function or survive. It explains that evidence is needed to establish whether a surrogate can reliably predict clinical benefit in its intended setting.FDA Facts — Biomarkers and Surrogate Endpoints (opens in a new tab)
In laboratory work, researchers also use proxy measurements, such as a signal intended to represent an underlying cellular process. That does not automatically make the measurement a validated clinical surrogate. The term should remain tied to the context in which substitution is supported.
| Question | Evidence needed |
|---|---|
| Was the marker measured reliably? | Measurement performance |
| Does it relate to the intended outcome? | Evidence linking marker and outcome |
| Does changing it predict the outcome? | Evidence supporting the proposed substitution |
The first check can succeed while the others remain uncertain. A precisely measured marker is still only the quantity that was actually measured until its broader interpretation is established.
Trace the missing link in a proxy claim
Consider an original hypothetical experiment in which a peptide changes a cellular fluorescence signal. The authors interpret that signal as evidence of improved cellular function. The reader should locate the independent evidence connecting this particular signal change to the stated function.
A prior association between signal and function is relevant, but the intervention could affect the signal through another route. For example, the readout might change while the intended function remains unchanged. This is a possible alternative to examine, not a claim that it happened.
If the study also measures the function directly, compare the two outcomes. Do they change together under the tested conditions, and does the analysis support the proposed relationship? A consistent result adds information without proving that the relationship holds in every context.
Keep validation within its demonstrated setting
FDA cautions that even validated surrogate endpoints can fail to capture the overall benefit and risk of a product in another setting, including effects not reflected by the surrogate. Validation is therefore not a licence to transfer the endpoint’s interpretation without examining context.FDA Facts — Biomarkers and Surrogate Endpoints (opens in a new tab)
For a peptide research summary, compare the biological system, intervention class, outcome definition and time scale behind the supporting evidence. A relationship observed in one model need not have been demonstrated in another.
The direction of change also needs context. A higher marker value may be desirable, undesirable or ambiguous depending on what is being measured and why. Do not substitute a favourable adjective for an explanation of the endpoint.
When a paper calls a marker validated, follow the cited validation evidence. Determine whether it validates the analytical measurement, a biological association or substitution for a specific clinical outcome. Those uses of validated are not interchangeable.
Preserve endpoint distinctions in the conclusion
A functional outcome directly tests a function in the system studied, but that system still sets the boundary. A functional cellular result is not automatically a clinical outcome in people, just as a molecular marker is not automatically a functional result.
A useful evidence note states the measured endpoint first, then the broader outcome being inferred and the evidence supporting that connection. If the connection is not established, say that the study reports a proxy change whose broader significance remains uncertain.
Avoid treating all proxy evidence as worthless. Its value depends on the research question. A marker can answer whether a pathway-associated signal changed even when it cannot answer whether an intervention improves health.
This distinction helps keep promising findings useful and accurately bounded. It lets the reader see which part of a claim comes from direct observation and which part still depends on an evidential bridge.
Sources and further detail
- FDA Facts — Biomarkers and Surrogate Endpoints (opens in a new tab)
Definitions, validation and context/benefit-risk limitations read. No historical approval percentage, treatment recommendation or current regulatory eligibility claim reproduced. The cellular-fluorescence example is original.
Sources checked 19 September 2026. Worked examples are illustrative unless a supplied report is explicitly identified. This article has not undergone independent scientific peer review.