ANALYTICAL REPEATABILITY AND PRECISION OF THE XEVO CDMS SYSTEM FOR AAV CAPSID CHARACTERIZATION

Analytical repeatability and precision of the Xevo CDMS system for AAV capsid characterization — Summary

Significance of the topic

Accurate, precise and reproducible quantification of adeno-associated virus (AAV) capsid populations is essential for gene therapy development and manufacturing. Only genome-containing (full) capsids contribute to therapeutic activity; empty, partially filled and overfilled particles are product-related variants that can reduce potency and increase immunogenic risk. Analytical methods that deliver unbiased single-particle mass measurements with high throughput and robust performance across process-relevant concentrations are therefore critical to support process development, lot comparability and CMC decision-making.

Objectives and study overview

This study evaluated the analytical repeatability, intermediate precision and operational robustness of the Waters Xevo CDMS (Charge Detection Mass Spectrometry) system for intact AAV capsid characterization under multi-day, high-throughput conditions and across viral titres representative of upstream and downstream process stages. Key aims were to quantify mass precision, stability of relative peak abundance, and method robustness to concentration changes and injection order.

Methods and experimental design

Samples and preparation:

• Standard AAV8 capsids (CMV-GFP), empty and full, sourced from Virovek Inc. at 2 × 10^13 viral particles (vp)/mL.
• Buffer-exchanged into 200 mM ammonium acetate with 0.01% Pluronic F-68.
• Empty and full capsids combined 1:1 and diluted to 1 × 10^13 vp/mL for repeatability/reproducibility testing (12 injections per day over five consecutive days).
• Additional dilutions to 5 × 10^12 vp/mL (downstream-equivalent) and 1 × 10^11 vp/mL (upstream-equivalent) used to assess robustness to concentration variation; alternating injections of low and high concentration were performed within single-day runs.

CDMS acquisition and data processing:

• Ions produced by positive-mode nano-electrospray ionisation.
• Xevo CDMS instrumentation trapped ions for ~100 ms per event; acquisition periods of ~30 minutes collected roughly 5,000 single-ion events per run.
• Time-domain signals were Fourier transformed to extract oscillation frequency (yielding m/z) and signal magnitude (yielding charge, z); individual ion masses calculated as m = (m/z) × z.
• Signal processing and visualization performed in Waters_Connect software.

Instrumentation used

The CDMS mass analyser comprises a conductive detection cylinder with ion-reflecting mirrors that cause ions to oscillate; induced charge on the cylinder is detected via a low-noise charge-sensitive amplifier and analysed by FFT to provide simultaneous m/z and charge measurements for single particles. The system configuration used positive nano-ESI ionization and trapping parameters tuned to collect ~5,000 ions per run.

Main results and discussion

Repeatability and intermediate precision:

• Relative peak abundance for empty vs full capsids across five days (12 injections per day) showed excellent stability: <2% RSD for peak abundance, with all injections within ±3% of the overall mean.
• Measured capsid masses for empty and full species exhibited very high precision: <1% RSD across the five-day study and all values within ±3% of the mean.

Robustness across concentration and injection order:

• Alternating injections between upstream-equivalent (1 × 10^11 vp/mL) and downstream-equivalent (5 × 10^12 vp/mL) concentrations produced consistent relative quantification for Empty/Partials, Full and Overfull species with RSDs below the predefined 5% acceptance threshold.
• Average measured masses for each species remained highly precise across concentration conditions and injection order, with RSDs well below 1%.

Operational performance metrics:

• First-pass success rate (data acquisition without the need for repeat injections) reported at ~95%, improving throughput and reducing sample consumption.
• Overall usable data rate across runs reached ~99%, indicating reliable data capture for QC and development workflows.

Interpretation:
The direct single-particle mass measurement approach of CDMS eliminates the need for charge-state deconvolution and avoids biases associated with ensemble-based methods. The observed mass precision (<1% RSD) and stable relative quantitation (<2% RSD for peak abundance, <5% across species and concentrations) demonstrate the Xevo CDMS system is fit-for-purpose for robust empty/full distinction and relative quantification in AAV analytics.

Benefits and practical applications

Key practical advantages demonstrated by the study include:

• Accurate single-particle mass measurement enabling unequivocal empty vs full capsid identification and detection of overfilled or partially filled species.
• High precision and repeatability suitable for routine QC, lot release comparability and process development monitoring.
• Robust performance across a wide span of viral titres and against varying injection orders, facilitating integration into upstream and downstream analytical workflows.
• High first-pass and usable data rates that increase throughput and minimize need for reanalysis.

Future trends and potential applications

CDMS is positioned to become a standard orthogonal tool in viral vector analytics as gene therapy programs scale. Expected developments and applications include:

• Integration of CDMS into release testing panels and CMC dossiers as evidence for capsid composition and potency-related attributes.
• Automation and higher-throughput acquisition workflows to support large sample sets in process characterization and stability studies.
• Method extension to other viral vectors and large protein complexes where single-particle mass information provides unique insight.
• Software and data-standard improvements to streamline regulatory reporting and comparability studies.

Conclusion

The Xevo CDMS platform delivered highly repeatable and precise single-particle mass measurements for AAV capsids across multi-day and concentration-variable experiments. Mass precision consistently below 1% RSD, peak abundance variability under 2% RSD, and robust relative quantification (<5% RSD across conditions) support the suitability of CDMS for routine AAV process monitoring, QC applications and CMC decision-making. High first-pass success and usable data rates further reinforce operational robustness and throughput advantages.

References

Christofi E., Haris A., McDonald S., Analytical repeatability and precision of the Xevo CDMS system for AAV capsid characterization, Waters Corporation poster, 2026.