Enhanced Risk Assessment for a Traditional LC Separation Method Using ICH Q14 Regulatory Principles and a Simplified HPLC System

Enhanced Risk Assessment for a Traditional LC Separation Method Using ICH Q14 — Application Note Summary

Significance of the topic

The paper applies modern regulatory expectations (USP <1220> and ICH Q14) and Analytical Quality by Design (AQbD) principles to a legacy liquid chromatography assay. Bringing DoE-driven, risk‑based development to an established method improves understanding of long‑term performance, supports robust method transfer, and reduces the likelihood of failures during routine use or instrument changes. This is relevant for QC laboratories, method owners, and regulatory submissions where objective, data‑rich evidence of method robustness is required.

Objectives and overview of the study

The study sought to augment a traditional metronidazole assay (API plus six impurities) with enhanced knowledge using DoE-based AQbD software and a simplified HPLC platform. Specific goals were to: (1) screen method parameter space broadly (±40%) and then focus (±25%), (2) identify an Acceptable Performance Region (APR) and key failure modes, and (3) generate data to inform method transfer risk assessments and control strategies.

Methodology

• Sample: Metronidazole API with six impurities prepared as a system suitability test (SST) mixture at 0.1% in water.
• DoE strategy: Two sequential experiments — a wide screening matrix covering ±40% of critical method parameters, followed by a focused matrix covering ±25% around the nominal method. Primary factors evaluated were pump flow, gradient time, and column oven temperature.
• Performance metrics: System suitability criteria included peak symmetry (tailing ≤ 1.2), USP resolution ≥ 2.0 for the API, and peak identification by relative retention time (RRT) supported by PDA spectral matching. Visualizations (response surfaces, contour plots, overlay APR plots) and statistical summaries were used to summarize performance.
• Verification: Targeted verification runs spanning ±25% around nominal conditions to confirm model predictions for peak behavior (RRT, tailing, resolution).

Used Instrumentation

• LC system: Alliance iS HPLC System with Photodiode Array (PDA) Detector.
• Column: XSelect CSH C18, 4.6 × 150 mm, 2.5 µm.
• Mobile phases: 0.05% formic acid in water (A) and 0.05% formic acid in acetonitrile (B).
• Operating conditions: Column 40 °C (nominal), flow 1.250 mL/min (nominal), injection 5 µL, equilibration 3.0 min, run time 11.0 min, detection at 315 nm.
• Software: Empower 3 CDS for data acquisition and Fusion QbD Professional for DoE, modeling and APR visualization.

Main results and discussion

• Peak elution order remained consistent across the screening and focused DoE experiments; PDA spectral library enabled reliable peak identification.
• Peak count (number of resolved peaks within the run time) increased in areas combining high flow (up to 2.000 mL/min) with short gradient times (as low as 4.0 min). Response surface and contour plots identified these trends clearly.
• Operation below ~1.100 mL/min produced incomplete elution of all seven peaks within the 11 min run time, representing a critical failure region for method performance.
• API USP resolution remained stable across focused verification runs, indicating no significant risk to assay accuracy for the active ingredient under the tested parameter variations.
• Impurity RRTs shifted by less than approximately two seconds under focused conditions, showing limited retention variability; however, individual impurities had different sensitivity to temperature and flow. Example: impurity B showed decreasing RRT with increased temperature, whereas impurity E exhibited the opposite trend.
• Symmetry factor (tailing) for API and impurities met SST criteria across the focused domain, but contour plots suggested potential further improvement at column temperatures below the nominal 40 °C.
• APR visualization (white region) defined method conditions that consistently met SST criteria; red/grey zones indicated parameter combinations outside acceptable performance.

Benefits and practical applications of the method

• Enhanced DoE-derived data provide objective evidence to justify method boundary choices, inform proven acceptable ranges (PAR) or control strategies, and support regulatory submissions aligned with ICH Q14 and USP <1220> expectations.
• Identification of critical risks (e.g., low flow leading to incomplete elution) allows laboratories to prioritize instrument qualification and component checks during transfer — for example, flagging flow accuracy as a high‑risk attribute.
• Demonstrates that a simplified, user‑friendly HPLC platform (Alliance iS with PDA) can be integrated with modern QbD software to produce actionable method knowledge without requiring high‑complexity hardware.
• Supports development of risk‑based transfer checklists, Ishikawa diagrams, and risk heat maps to guide successful cross‑site implementation.

Method lifecycle considerations

The QbD outputs were used to populate method transfer risk tools: an instrument‑specific transfer checklist, fishbone (Ishikawa) analyses, and heat maps. From the DoE results, flow accuracy was categorized as high risk for transfer (due to sensitivity below 1.100 mL/min), whereas the column heater configuration was judged low risk because temperature changes had limited impact on overall API and impurity performance in the focused domain. These insights can shape acceptance criteria, monitoring plans, and change control decision trees during the method lifecycle.

Future trends and potential uses

• Wider adoption of AQbD and DoE tools for legacy methods will accelerate generation of lifecycle evidence and reduce post‑implementation deviations.
• Integration of robust APR definitions with automated instrument qualification and digital transfer packages will streamline cross‑site method implementation and regulatory submissions.
• Further work could combine QbD-based LC optimization with orthogonal detection (MS) or hyphenated techniques to strengthen impurity identification and control strategies.
• Expanded use of automated visualization and model‑driven decision support will enhance reproducibility and reduce human error in method transfer activities.

Conclusion

Applying ICH Q14 and USP <1220> principles with DoE-based AQbD software to a traditional metronidazole LC assay produced richer, actionable knowledge about method performance. The study identified a clear APR, highlighted flow rate as a primary transfer risk, demonstrated stable API resolution across tested variations, and showed that a straightforward HPLC system can support modern method development workflows. These results support risk‑based method transfer and lifecycle management decisions, improving confidence in routine assay performance.

References

1. ICH Q14 Analytical Procedure Development Guidance for Industry, U.S. FDA and ICH Quality, March 2024.
2. Analytical Procedures Guidance for Industry, ICH Quality, Revision 2, U.S. FDA, March 2024.
3. Fusion QbD Software, LC Method Development User’s Guide. S‑Matrix Corporation.
4. Fadi Alhateeb; P.D. Rainville. Analytical Quality‑by‑Design Based Method Development for the Analysis of Formoterol, Budesonide, and Related Compounds Using UHPLC‑MS, Waters Corporation, October 2019.
5. Mitigating Risk of Validated Analytical Procedure Failures When Upgrading or Replacing LC Assets: Harnessing the Power of Quality by Design (QbD) Principles, Waters Corporation, May 2021.
6. Alliance iS HPLC System: Method Transfer Evaluation Checklist, Waters Corporation, October 2025.