Optimal Separation Conditions for Antibody-Drug Conjugates by Hydrophobic Interaction Chromatography

Significance of the topic

The accurate characterization of antibody-drug conjugates (ADCs) is essential for biopharmaceutical development and quality control because the average drug-to-antibody ratio (DAR) and its distribution strongly influence efficacy, stability, pharmacokinetics and safety. Hydrophobic interaction chromatography (HIC) provides a native-condition separation that resolves ADC species by hydrophobicity differences arising from varying numbers and positions of conjugated drug molecules. Optimizing HIC conditions improves resolution of DAR species, repeatability of quantification, and robustness for routine ADC analysis.

Objectives and study overview

This application study aimed to optimize HIC separation conditions for a model cysteine-conjugated ADC (brentuximab vedotin) to enable reproducible DAR distribution analysis. Key variables investigated were the type and concentration of salt in the aqueous mobile phase and the proportion of 2-propanol in the organic modifier. The study compared sodium chloride, sodium sulfate and ammonium sulfate as salting-out agents, varied salt concentrations, and evaluated 2-propanol at 10–30% (v/v). Performance metrics included resolution of DAR species, peak shape, peak area, and intra-day repeatability (%RSD).

Methodology

Summary of the analytical approach and experimental factors:

• Sample: Brentuximab vedotin, 10 mg/mL in ultrapure water (cysteine-linked ADC with isomeric DAR species).
• Separation mode: Hydrophobic interaction chromatography under near-native, non-denaturing conditions.
• Mobile phases: Phosphate buffer (100 mmol/L, pH 7.0) with high concentration of selected salt in pump A; buffer with organic modifier in pump B.
• Salt screening: Initial salt concentration 1.0 mol/L for NaCl, Na2SO4, and (NH4)2SO4; subsequently salt concentrations increased (for NaCl up to 4 M; for ammonium sulfate tested 1.0, 1.5 and 2.0 M).
• Organic modifier: 2-propanol in pump B tested at 10, 15, 20 and 30% (v/v) to modulate elution strength and selectivity.
• Gradient: 0→100% B over 20 min, hold then re-equilibrate (total run ~35 min).
• Detection: Photodiode array at 280 nm; injection 10 µL; column temperature 25 °C.

Used instrumentation

The experiments used a corrosion-resistant UHPLC platform and HIC column adapted for high-salt mobile phases:

• UHPLC system: Nexera XS inert (salt-tolerant flow path).
• Column: Shim-pack Bio HIC Butyl (100 mm × 4.6 mm I.D., 4 µm).
• Sample vials: Shim-vial H glass.
• Detector: PDA set to 280 nm (SPD-M40, UHPLC inert cell).

Main results and discussion

Key experimental findings and interpretation:

• Salt type: At 1.0 M NaCl retention was insufficient; increasing NaCl to 4.0 M improved separation but required very high ionic strength. Both 1.0 M Na2SO4 and 1.0 M (NH4)2SO4 delivered superior peak shape and better separation than NaCl at equivalent concentration. Ammonium sulfate was preferred due to higher aqueous solubility and practical handling at elevated molarities.
• Salt concentration effect: Increasing (NH4)2SO4 from 1.0 to 2.0 M sharpened peaks, increased retention and peak heights, and improved resolution between DAR species. 2.0 M (NH4)2SO4 was selected as optimal for subsequent work.
• Organic modifier effect: 2-propanol at 15% (v/v) in mobile phase B provided the best balance of elution strength and sensitivity, maximizing peak areas for higher-DAR species (DAR 6 and DAR 8). At 20% and especially 30% 2-propanol, overall peak areas decreased and peak shapes deteriorated, indicating too strong elution or loss of selectivity.
• Repeatability and DAR calculation: Six replicate injections under optimized conditions (2.0 M (NH4)2SO4 in A; 15% 2-propanol in B) produced peak area %RSD values <5% for all DAR species (individual %RSDs reported near or below 4.4%). The calculated average DAR was ~4.04, consistent with literature for this ADC, demonstrating both accurate quantitation and intra-day precision.

Benefits and practical applications

Practical implications of the optimized HIC method:

• Robust DAR profiling: The protocol enables baseline-level separation of multiple DAR species under native-like conditions suitable for formulation and stability studies.
• High repeatability: %RSDs below 5% support use in routine QC and method validation workflows for lot release or comparability testing.
• Compatibility with high-salt mobile phases: Use of an inert, salt-resistant UHPLC system extends instrument lifetime and reduces corrosion risk when using molar concentrations of ammonium sulfate.
• Simple implementation: The method uses common reagents (phosphate buffer, ammonium sulfate, 2-propanol) and a commercially available HIC butyl phase, facilitating adoption in biopharma labs.

Future trends and potential applications

Anticipated developments and extension opportunities for HIC in ADC analysis:

• Coupling to native mass spectrometry: Desalting strategies or online buffer-exchange interfaces could enable mass-based confirmation of DAR and positional isomers while retaining HIC selectivity.
• Column and stationary phase innovation: New HIC chemistries and smaller particle UHPLC HIC columns will improve resolution and reduce run times for higher throughput.
• Automation and QC integration: Robust, validated HIC methods can be incorporated into automated QC platforms for release testing and stability monitoring.
• Orthogonal method pairing: Combining HIC with SEC, IEC or peptide mapping strengthens characterization by addressing aggregation, charge variants and site-specific conjugation information.

Conclusion

This application demonstrates that careful optimization of salt type and concentration and of organic modifier content yields high-resolution, reproducible HIC separations of ADC species. Ammonium sulfate at ~2.0 M combined with 15% 2-propanol on a Shim-pack Bio HIC Butyl column and a salt-tolerant Nexera XS inert system provided optimal performance for brentuximab vedotin, delivering precise DAR distribution measurement (DAR ≈ 4.04) with %RSD <5%. The approach is suitable for routine ADC characterization, with considerations for salt handling, instrument compatibility and downstream MS integration where needed.