MSACL 2025 Abstract

Removing Carryover Concerns and Validation of a 6-Analyte Barbiturate Panel on GC-MS by Split Injection

Alison Lightfoot, Ph.D., Chris Thompson, Paul Jannetto, Ph.D., Loralie Langman, Ph.D.
Department of Laboratory Medicine and Pathology, Mayo Clinic, Rochester MN

Alison Lightfoot, Masters in Chemistry with Mathematics, PhD in Materials Chemistry (Presenter) Mayo Clinic

Relevant Financial Disclosures (within past 24 months, reported on Jun 19, 2025) No relevant financial relationship(s) to disclose. Conflict mitigation NOT REQUIRED.

INTRODUCTION:
Barbiturates represent a class of drugs that were originally introduced as sedatives/sleep inducers. Amobarbital is an intermediate-acting barbiturate that has sedative/hypnotic properties used to relieve anxiety and insomnia. It also has analgesic properties. Butalbital is a short-acting barbiturate and is used to control severe headaches and pain. Pentobarbital is a short-acting barbiturate and is used in the treatment of seizures and brain injuries. Phenobarbital is frequently used to control major motor (grand mal) seizures and the treatment of epilepsy. Secobarbital and butabarbital are short-acting barbiturates and are used as a preanesthetic agent and the short-term treatment of insomnia.
Prior to this re-development and validation, an in-house assay was live clinically but had significant problems with carryover leading to inefficient workflows and frequent repeat testing of patients.

OBJECTIVE(S):
The re-development, validation, and implementation of a GC-MS method to quantify the levels of six barbiturates in patient serum samples with a focus on reducing significant carry-over observed in the previous clinical test.

METHODS:
All six barbiturates were extracted from serum using solid phase extraction techniques. The serum was buffered with a pH 6.0 sodium phosphate solution and eluted with 25% hexane in ethyl acetate. The organic phase was then dried and reconstituted in ethyl acetate. Standards and in-house controls were prepared by spiking in bovine serum. Clinical controls were prepared in human serum.

The samples were analyzed using a 1 mcL injection on an Agilent GC-MS (GC model 6890 and MS model 5975). A J&W Scientific DB-5MS, 15m, 0.25mm I.D. capillary column with a film thickness of 1 micron was installed and used for separation in the GC. The GC and MS methods were built using Agilent Chemstation. The following GC method parameters were defined: 260 °C for the inlet temperature, 280 °C for the auxiliary temperature, and an oven ramp from 85 °C to 295 °C over the course of 11 minutes. The samples were injected using split flow injections with a split of 9:1. For the MS, an inert EI (electron-impact) source was installed in the MS and Helium was used as carrier gas. Chemstation was used to build and submit each batch. Agilent Mass Hunter was used as quantification software.

RESULTS:
Carryover was negligible for all six analytes. This is a significant improvement, as the previous assay struggled with significant carryover for phenobarbital and pentobarbital, which caused potential repeat extractions of patient samples as well as requiring a blank injection between every sample on the run. Re-development of the extraction method led to a small reduction in carryover; however, the biggest factor in reducing carryover was changing the GC inlet flow from splitless to split. With the split ratio set at 9:1, less sample was injected onto the column yielding a significant reduction to carryover without sacrificing sensitivity. The split injection also improved analyte chromatography, which previously showed fronting due to overloading on the column, especially at the upper limit of quantitation (ULOQ).

The linearity across the analytical measuring interval (AMI, 0.5-10 mcg/mL) returned a linear regression with R2=1 and slopes between 0.97 and 0.99 for all six analytes (N=5 runs). The precision for all control levels (0.75, 4, and 8 mcg/mL) averaged %CV’s < 3.5% for intraday precision and %CV’s < 4.5% for interday precision (N=20 runs). Accuracy samples for pentobarbital were compared to the current clinical assay and returned a linear regression with R2=0.99 and a slope of 1 (N=19 samples). All six analytes were spiked and compared to theoretical values for accuracy and returned linear regressions with R2=0.99 or 1 and slopes between 0.97 to 1(N=30 samples).

CONCLUSIONS:
Through a full re-development and validation, we have demonstrated a new robust method for the accurate quantitation of six barbiturates in serum that has resolved carry-over concerns and has now been implemented into the clinical laboratory.