Development, validation and application of a GC-MS method for quantification of glucose and lactate in vitreous humor

Abstract

Develop and apply a GC-MS method for quantifying glucose and lactate in vitreous humor. In routine forensic toxicological screening, glucose and lactate are measured in vitreous humor to evaluate the ante-mortem glycaemic status and to identify possible lactate acidosis. Concentrations for both endogenous compounds are required in the assessment of hypo-, normo- or hyperglycaemic status before death. In our laboratory, no in-house chromatographic method was available and diluted vitreous humor samples had to be analysed by an external laboratory, using a spectrophotometric method validated for serum. Starting from a GC-MS method available for GHB and BHB, we investigated the simultaneous measurement of glucose and lactate to increase sample throughput. An analytical method was developed for analysis of glucose and lactate, based on an available GC-MS method used in-house for other analytes and on literature (Shojaee-Moradie et al., Journal of Mass Spectrometry 1996;31:961–966), employing a combination of two derivatisation reactions. Glucose required both oxime and trimethylsilyl derivatisation to achieve compatibility with GC-MS. For lactate, only the silylation step was necessary. Method validation was carried out using in-house prepared quality control samples. Authentic vitreous humor samples ( n = 100) were analysed in parallel between October 2021 and April 2022, comparing the results of the newly developed method to those from the external laboratory. The derivatisation time and calibration model were optimised. The method was validated according to the guideline on bioanalytical method validation provided by the European Medicines Agency. Linear calibration curves showed R 2 > 0.99 for glucose and lactate and the lower limit of quantification (LLOQ) was 12.5 mg/dL for both compounds. Spearman's ρ showed a strong, statistically significant correlation between the results obtained by the two labs ( P < 0.001; 0.836 for lactate, 0.963 for glucose). Bland–Altman plots were used to assess agreement between the measurements. For glucose, 11% of the analysed samples fell out of the 95% confidence interval; for lactate, only 9%. Deviant samples were re-analysed with the new method, yielding the same results, suggesting the possibility of deviant results by the spectrophotometric method. For lactate, the results of the developed method tended to show a small positive bias compared to those of the clinical lab. The established LLOQ was well below the postmortem cut-off values for determining hypoglycaemia and the measured glucose and lactate concentrations in most authentic samples. Results for the authentic samples showed good agreement between both methods and for both analytes. The developed and validated GC-MS based method has shown to be suitable for accurate measurement of glucose and lactate concentrations in vitreous humor.