Why Measure Concentration Before the Next Period?
In a crossover study, each participant receives treatments in successive periods. A sample collected immediately before dosing in a later period can reveal whether drug from an earlier period remains measurable in the circulation. This is useful because a washout interval is a design choice, whereas a predose sample provides information about the participant’s concentration at the actual start of the next period.
A measurable concentration at that point is a warning that the participant has residual circulating analyte before the new dose. It does not, by itself, establish how much the subsequent pharmacokinetic profile will be affected. Conversely, an unquantifiable result does not establish that no drug remains anywhere in the body. Predose concentrations are therefore diagnostic evidence, not a certificate that carryover is either present in a clinically important amount or absent.
The sample must be taken before the current period’s dose; otherwise, the measured concentration cannot be interpreted as a period-start baseline. The record should identify the sampling time, the preceding treatment and dose time, the analyte measured, and the assay’s lower limit of quantification (LLOQ). If an active metabolite is relevant, an assay that measures only the parent drug cannot establish whether that metabolite is present.
What an Elimination Prediction Can—and Cannot—Tell Us
A simple elimination model can help assess whether a measured predose concentration is plausible given the time since the preceding dose. For illustration, suppose the analyte follows first-order elimination in a single, well-mixed compartment after the relevant distribution phase. The concentration \(C(t)\) after time \(t\) then satisfies \(dC/dt=-kC\), where \(k\) is the elimination rate constant. Separating variables and integrating from concentration \(C_0\) at time zero gives \(\ln(C(t)/C_0)=-kt\), and hence:
Here, \(t_{1/2}\) is the half-life under the model. The prediction assumes that the relevant elimination phase is adequately described by a single exponential, the starting concentration is appropriate, and the elapsed time is known. Real concentration-time profiles may not satisfy these assumptions: distribution phases, multiple compartments, variable elimination, active metabolites, or other sources of analyte can make the simple prediction incomplete. A model-based prediction is a reference for interpreting a sample, not a substitute for it.
Worked Example: Predicting a Period-Start Concentration
Suppose a hypothetical participant has a concentration of 8.0 ng/mL at the start of a modeled elimination interval. The estimated elimination rate constant is \(0.20\ \text{h}^{-1}\), and the next period’s predose sample is collected 24 hours later. Assume for this calculation that first-order elimination applies over the interval and that no further drug enters the circulation.
The predicted concentration is:
If the assay LLOQ were 0.05 ng/mL, this prediction would be slightly above the quantification limit. A measured result near that value would be compatible with the simplified model and would indicate quantifiable residual analyte. The prediction is not proof that the participant’s true concentration must be 0.0658 ng/mL: uncertainty in the starting concentration, elimination rate, sampling time, and model assumptions all affect it.
The interpretation of a measured result depends on the assay. A concentration at or above the validated LLOQ can be reported quantitatively, subject to the assay’s performance. A result below the LLOQ should generally be described as below the quantification limit, not as zero. The amount could be zero, or it could be positive but too small to quantify reliably. Near the LLOQ, assay imprecision and bias also make fine distinctions between numerical values especially fragile.
Reading Predose Results in Context
A concentration measured before a later-period dose is most informative when the preceding treatment contains the same analyte and the sample was collected at a documented time after that treatment. In that setting, a quantifiable result is evidence of circulating analyte at the start of the period and is consistent with residual exposure. Attribution is less secure if the analyte could have another source, the sample may have been contaminated, or the timing of the prior dose is uncertain. The measurement alone does not identify the source.
Interpretation also requires a clear distinction between the presence of measurable analyte and its possible influence on later outcomes. A low concentration might be measurable without having a meaningful effect on a subsequent profile; establishing such an effect requires more than observing a predose value. Likewise, absence of a quantifiable result cannot exclude analyte in unmeasured tissues, concentrations below the LLOQ, or relevant metabolites not included in the assay.
Worked Example: Summarizing Quantifiable Period-Start Results
In a hypothetical crossover study, five participants have period-2 predose concentrations of 0.42, 0.31, 0.28, 0.19, and 0.11 ng/mL for the analyte administered in period 1. The samples were collected before period-2 dosing, and the validated assay LLOQ is 0.10 ng/mL. Assume sample identity and timing are verified and there is no known alternative source of this analyte.
Each result is above the LLOQ, so each is quantifiable. The arithmetic mean is:
The ordered values are 0.11, 0.19, 0.28, 0.31, and 0.42 ng/mL, so the median is 0.28 ng/mL. These summaries describe the five measured concentrations; they do not imply that every participant had a concentration of 0.262 ng/mL or that the observed values have a particular cause with certainty. Given the verified timing and prior administration of the analyte, the results are evidence of residual circulating analyte at period start. They warrant attention when interpreting the later-period data, but the summaries alone do not quantify any effect on those data.
Worked Example: Interpreting Results Below the LLOQ
In a second hypothetical study, six participants have period-2 predose results reported as “<0.05 ng/mL.” The assay LLOQ is 0.05 ng/mL. For one participant, a simplified model predicts a concentration from a starting value of 10 ng/mL, with a half-life of 6 hours, after a 48-hour washout. Assume the model’s first-order elimination conditions hold for this calculation.
The interval is \(48/6=8\) half-lives. Each half-life halves the modeled concentration, so:
The predicted value is below the LLOQ of 0.05 ng/mL, so an assay result reported as “<0.05 ng/mL” is compatible with the model. It does not confirm the prediction or establish that the concentration is zero. For all six participants, the correct summary is that no concentration was quantifiable at the stated limit; the data do not distinguish zero from positive concentrations below 0.05 ng/mL.
Diagnostics, Alternatives, and Failure Modes
A predose measurement is strongest as a diagnostic when collection and analysis are reliable. Check that the sample was collected before dosing, the time since the previous dose is known, the assay is selective for the relevant analyte, and the reported value is interpreted against the correct LLOQ. Review unexpected values for possible sample misidentification, contamination, or recording errors. A result inconsistent with a simple elimination prediction is a reason to examine those assumptions and records; it is not, on its own, proof that the assay or model is wrong.
Predose sampling can be considered alongside other sources of information, each of which has different limitations:
| Approach | What it contributes | Main limitation |
|---|---|---|
| Predose concentration | Measures analyte in the sampled matrix at the period-start time. | Does not detect unmeasured analytes or quantify values below the LLOQ. |
| Elimination-based prediction | Checks whether a concentration is plausible under stated pharmacokinetic assumptions. | Prediction may be inaccurate if the assumed elimination model or inputs are unsuitable. |
| Postdose concentration-time data | Describe concentrations after the new dose and may provide context for the predose value. | Do not replace the period-start sample and can be difficult to interpret if residual analyte is present. |
| Measurement of relevant metabolites | Can reveal residual exposure not captured by a parent-drug assay. | Requires a suitable assay and a clear rationale for which metabolites matter. |
The main mistake is to treat “below the LLOQ” as equivalent to “no carryover.” That reasoning confuses a limit on what the assay can quantify with proof of absence. The correction is to report the result as below the stated limit and to recognize that the sample cannot rule out lower concentrations or analyte outside the sampled matrix. The opposite overstatement is also a pitfall: a quantifiable predose result establishes measurable analyte in that sample, but does not by itself establish a clinically important effect on the next period.
Check Your Understanding
Answer each question, then compare your reasoning with the solutions below.
- A period-2 predose sample is reported as “<0.08 ng/mL.” What can be concluded, and what cannot be concluded?
- A sample is collected five minutes after the period-2 dose, rather than before it. Why is it not a valid period-start predose measurement?
- Under a first-order model, a concentration of 4.0 ng/mL declines with a half-life of 4 hours. What concentration is predicted after 12 hours, and what model assumptions should be considered?
- Why might a measurable predose parent-drug concentration fail to describe all residual pharmacologically relevant exposure?
Solutions. (1) The concentration was not quantifiable at 0.08 ng/mL; it could be zero or any positive value below that limit. The result does not prove absence. (2) The current dose has already been administered, so the measured value may include drug from that dose and cannot isolate the concentration present at period start. (3) Twelve hours is three half-lives, so \(C(12)=4.0(1/2)^3=0.500\ \text{ng/mL}\). This prediction assumes first-order elimination over the interval, a suitable starting concentration, and no additional input. (4) The assay may not measure active metabolites, and the concentration in the sampled matrix may not capture analyte in other tissues. Thus the parent-drug result alone may be incomplete.