Interpreting recovery and matrix effects

The 2003 study by Matuszewski and colleagues separated recovery and matrix effects through experimental comparisons. Its aim was to prevent altered ionization from being interpreted as loss of material during preparation.

The distinction is conceptual: how much analyte passed through the process and how it responded on reaching the detector are different questions. A single comparison against a simple solution may contain both contributions. The article used HPLC-MS/MS bioanalysis and an investigational compound, not every possible peptide. Study.

ICH M10 describes recovery as an evaluation of extraction and processing. The comparison considers material added before and after those stages. The guidance does not require complete recovery, but requires consistency for the analyte and internal standard when one is used.

Incomplete recovery is therefore insufficient to invalidate a method. It must be known whether recovery is characterized and allows the required performance. Nor does apparently high recovery demonstrate absence of response interference. Official guidance.

The 2006 work compared calibration slopes in different biofluid lots. It addressed relative matrix effects: variation among samples that might appear equivalent because they share the label plasma or urine.

Repeating one lot many times estimates a different part of variability from studying different lots. The article also compared internal standards and ionization modes. Its results show why a stable response within one matrix does not automatically demonstrate comparability among matrices. Study.

In the cases examined in 2006, isotope-labeled internal standards showed advantages over certain analogs. That observation does not justify assuming that every labeled standard corrects all interference in every assay.

The relevant comparison is whether analyte and standard represent the stages being controlled sufficiently similarly. The standard's name does not replace method data. Likewise, changing the ionization interface produced differences in the cited studies, but is not a universal solution.

Consider a hypothetical result with weak signal and good repeatability. A reproducible loss would still need to be distinguished from reproducible suppression; both can produce highly consistent numbers. Repeatability does not identify the cause.

To interpret the report, locate the comparison defining recovery, the comparison evaluating matrix effects and the number of different matrices examined. This establishes whether the result describes the complete process or only part of it. These questions allow the evidence to be read without making an isolated percentage a verdict on the method.