Key points
- Haemolysis and lipaemia can change matrix effect, extraction recovery, analyte stability and the amount of drug released from red cells.
- ICH M10 expects haemolysed and lipaemic matrix to be included in selectivity and evaluated for matrix effect.
- In a study: record the condition at receipt, analyse if the method covers it, assess the impact if it does not - never reject results silently.
What changes in the sample
Haemolysis is the rupture of red blood cells, mostly during collection or handling. Haemoglobin, enzymes and the cell contents are released into the plasma, which turns pink to red. Lipaemia is a high level of lipids, mainly triglycerides, which makes plasma turbid or milky - typically after a fatty meal. Both are graded visually at sample receipt, ideally against a colour or turbidity chart so that the grading is the same for every technician.
How it affects LC-MS/MS results
- Matrix effect. Haemoglobin, haem, extra phospholipids and lipids can suppress or enhance ionisation differently from normal plasma.
- Red-cell partitioning. For drugs that accumulate in red cells, haemolysis releases drug into the plasma and the measured concentration rises - a true change in the sample, not an analytical error.
- Stability. Enzymes released from red cells can degrade esters and other labile compounds; haemoglobin can catalyse oxidation.
- Extraction. Lipophilic analytes partition into the lipid phase, which can lower recovery. Lipaemic samples form emulsions in liquid-liquid extraction and can clog SPE cartridges or filter plates.
- Pipetting. Very lipaemic samples are not homogeneous - a lipid layer on top changes what an aliquot contains if the sample is not mixed well.
Covering it in validation
ICH M10 asks for selectivity to include lipaemic and haemolysed matrix (at least one source of each) in addition to the six normal sources, and expects matrix effect to be evaluated in relevant special matrices. In practice:
| Matrix | Common preparation | Experiment |
|---|---|---|
| Haemolysed | Blank plasma spiked with lysed whole blood - commonly 2% (v/v), a clearly visible level | Selectivity blank; low and high QCs against a normal calibration curve |
| Lipaemic | Naturally lipaemic donor plasma with a documented triglyceride level - closer to real samples than spiked lipid emulsions | Selectivity blank; low and high QCs against a normal calibration curve |
The QCs should meet the usual criteria: mean accuracy within ±15% of nominal and precision ≤ 15%. If the analyte is unstable in haemolysed blood, add a stability experiment in haemolysed matrix as well. Write the tested level into the validation report - it defines the samples the method covers.
When they arrive in a study
- Record at receipt. Note the condition and grade of each affected sample in the sample receipt record, not later from memory.
- Check the validation. If the method was validated for that degree of haemolysis or lipaemia, analyse as normal.
- If it was not, assess the impact before reporting: a partial validation with matrix at the relevant level, or a documented scientific justification. Many sponsors also want such samples flagged in the data.
- Mix well before aliquoting lipaemic samples, and watch the internal standard response of affected samples against the run.
- Report it. List the affected samples and how they were handled in the bioanalytical report. Excluding a result because of sample condition needs a predefined criterion and a documented reason.
Preventing them at the clinic
Most haemolysis is caused during collection: a needle that is too narrow, a slow draw, vigorous shaking of the tube, delayed centrifugation or freezing whole blood. Clear collection instructions and a short training session for the clinic staff often reduce it more than any laboratory measure. Lipaemia is expected in fed bioequivalence studies, where subjects eat a high-fat breakfast before dosing - plan for it by validating lipaemic matrix before the study starts, not after the first samples arrive.
Educational summary for chromatographic (LC-MS/MS) bioanalysis. Always apply the current ICH M10 text, the expectations of the regulator concerned and your laboratory's SOPs. Ligand-binding assays have wider limits in ICH M10.
ARISENTO Insights is published by Autofai Technologies Private Limited, the makers of ARISENTO bioanalytical software.