Clinical evaluation of the ADVIA Centaur XPT chemiluminescent immunoassay for equine insulin measurement.

Research Overview

• This study evaluated the performance of the ADVIA Centaur XPT chemiluminescent immunoassay (CLIA) for measuring insulin levels in horses.
• The aim was to determine if this analyzer produces reliable results compared to a previously used method (Immulite 2000 XPi) and to establish new insulin decision thresholds for clinical assessment.

Background and Significance

• Insulin dysregulation is a critical factor in equine metabolic syndrome, a condition affecting horses’ health.
• Basal insulin concentrations are commonly measured to assess insulin dysregulation.
• Different laboratory analyzers can produce varying insulin measurements, which complicates comparison and clinical interpretation.
• Having reliable, consistent measurement tools with established decision thresholds is important for accurate diagnosis and management of affected horses.

Aims of the Study

• Evaluate the analytical performance (precision, linearity, dilution recovery) of the ADVIA Centaur XPT CLIA for equine insulin measurement.
• Compare the ADVIA Centaur XPT results to those obtained from the Immulite 2000 XPi analyzer, a previously used CLIA method.
• Develop adapted decision thresholds for clinical classification based on the ADVIA Centaur XPT measurements.

Methods

• Performance testing of the ADVIA Centaur XPT included assessing:
  • Precision – intra-assay (within the same run) and inter-assay (between different runs) variability.
  • Linearity – ability to produce results proportional to insulin concentrations over a broad range.
  • Dilution recovery – accuracy of measurements when samples are diluted.
• Clinical serum samples: 89 equine serum samples submitted for basal insulin testing were analyzed on both analyzers.
• Statistical analyses used to compare results:
  • Deming regression – evaluates relationship and bias between two measurement methods.
  • Bland-Altman analysis – assesses agreement and systematic differences across measurement ranges.
  • Weighted kappa statistics – measures agreement between classification outcomes considering adapted insulin thresholds.
  • Spearman correlation – measures strength of monotonic association between results from the two methods.

Key Results

• Precision:
  • Intra-assay coefficients of variation ranged from approximately 1.92% (high insulin concentrations) to 3.56% (low concentrations), indicating good repeatability.
  • Inter-assay variation was slightly higher but still acceptable, ranging from about 5.19% to 5.78%.
• Comparison between analyzers:
  • Deming regression indicated a proportional bias: the ADVIA Centaur XPT consistently measured lower insulin levels compared to the Immulite 2000 XPi.
  • Bland-Altman plots confirmed this consistent difference across the range of insulin concentrations.
  • Despite bias, the Spearman correlation coefficient was 0.92, showing a strong rank order correlation between the two methods.
• Clinical classification agreement:
  • By adapting decision thresholds specific to the ADVIA Centaur XPT, classification of insulin status showed strong agreement.
  • The weighted kappa statistic was 0.82, indicating strong concordance in classifying insulin dysregulation between the methods.

Conclusions and Implications

• The ADVIA Centaur XPT CLIA demonstrated acceptable analytical precision suitable for clinical use.
• Although it measures slightly lower insulin concentrations than the Immulite 2000 XPi, the strong correlation suggests it reliably ranks insulin levels.
• Using adapted decision thresholds for the ADVIA Centaur XPT enables accurate classification of insulin dysregulation in horses, facilitating consistent clinical interpretation.
• This supports the ADVIA Centaur XPT as a suitable analyzer for measuring equine insulin concentration in the context of metabolic syndrome assessment.

Overall Significance

• Provides veterinarians and laboratories with an alternative CLIA platform for equine insulin measurement.
• Highlights the importance of method-specific decision limits when interpreting insulin test results to ensure correct diagnosis and management.
• Improves standardization and comparability in veterinary endocrinology diagnostics.