Carbamazepine can be analytically detected in rib bone using a specified extraction and GC–MS procedure. The validation study supporting that conclusion also used mass-normalized response ratios and contained reporting inconsistencies. Its performance data should be evaluated as method evidence, not as proof that a bone concentration directly measures toxicity.
What material was studied?
Fernández-López and colleagues used central portions of the fifth and sixth ribs collected in Murcia, Spain. Specimens were approximately 5 cm long. Three rib specimens from people whose blood was negative for carbamazepine supplied comparison material; this was not a large population of drug-free skeletons.
Surface soft tissue was removed, bone was fragmented, dried overnight at 50°C and pulverized. Powder was kept at −80°C until analysis. A 300 mg portion underwent methanol extraction with ultrasound, followed by solid-phase extraction and GC–MS. Sertraline was listed as the internal standard in the analytical table.
This preparation defines the tested matrix. An intact wet rib, marrow, another bone and dried pulverized rib material should not be assumed to give equivalent results.
How identity was supported
The instrument initially acquired full-scan electron-impact spectra and then used selected-ion monitoring. Table 2 listed carbamazepine retention at 12.9 minutes and ions at m/z 193, 165 and 139. Ion ratios had to remain within 20% of the calibrator averages.
The drug-free pool showed no relevant background after extraction, and no carryover was observed after the highest calibration point. These findings support selectivity under the tested conditions. A retention time near 12 minutes or a single ion alone would not reproduce the complete identification evidence.
What quantitative performance was reported?
Calibration standards were described from the quantitation limit through 500 ng/mg. The determination coefficient was 0.999. The reported detection limit was below 0.1 ng/mg and the quantitation limit was 0.3 ng/mg. Detection and supported numerical quantitation are different claims.
Controls were 10, 150 and 400 ng/mg. Table 3 reported intra-assay relative standard deviations of 1.4%, 2.3% and 0.4%, with inter-assay values of 1.3%, 1.4% and 0.9%. Corresponding absolute errors were 13.6%, 3.3% and 1.9% within assays, and 6.9%, 0.2% and 1.0% between assays.
The prose refers to accuracy results “over 14%,” which conflicts with those table entries. The table values should not be replaced with that inconsistent generalization. The paper also describes a preparation spiked with 50 ng in 0.3 g of bone while other validation quantities use ng/mg. Those descriptions are not interchangeable; the reporting basis needs clarification before reproducing the method from the text alone.
The response-ratio qualification matters
The authors expressed analyte levels through a mass-normalized ratio of analyte peak area to internal-standard peak area, while presenting values in ng/mg. They acknowledged difficulty obtaining recovery information from bone that would provide accurate data. This qualification should remain attached to the numerical results.
Reported extraction recovery was 92.6%, the matrix-effect value 78.8%, and process efficiency 73%. These are different performance measures. None is a biological correction for drug movement after death or a conversion into a blood concentration.
What stability testing covered
The study reported less than 10% concentration change over three freeze–thaw cycles and during its mid-term storage assessment. Methods described monthly triplicate measurements of two samples for six months. These laboratory observations do not validate unchanged drug content during years of burial, environmental exposure or advanced skeletal decomposition.
One authentic positive case demonstrated application, but did not create a population reference interval. The method’s strongest contribution is evidence that a defined bone preparation can produce a supported carbamazepine finding. Questions about poisoning, dose and blood equivalence require additional evidence.
The need to read individual performance results also arises in validation limits within a multianalyte blood assay. A panel or method label cannot replace its actual validation record.
Primary source
Fernández-López L, Mancini R, Rotolo MC, Navarro-Zaragoza J, Hernández del Rincón JP, Falcón M. Carbamazepine Overdose after Psychiatric Conditions: A Case Study for Postmortem Analysis in Human Bone. Toxics. 2022;10:322. doi:10.3390/toxics10060322.
Leave a Reply