Yes, you can change an HPLC method without revalidation, as long as the change stays inside the allowable adjustments of USP <621> or Ph. Eur. 2.2.46 and the system suitability test still passes. Since the harmonised text took effect in December 2022, this includes gradient methods: column length and particle size may change as long as the ratio L/dp stays within −25 % to +50 % of the original, with flow rate, injection volume and gradient times scaled by formula. Anything outside these limits is a method modification that needs change control, at least partial revalidation and, for a registered method, a regulatory assessment.
What counts as an adjustment and what as a modification?
An adjustment brings an existing, validated procedure to the current equipment without changing what it measures. A modification changes the procedure itself. The pharmacopoeias define adjustments as a closed list: if the change is not on the list or exceeds the stated limit, it is a modification.
The list covers the chromatographic conditions, not the chemistry. You may not change the stationary phase chemistry (the L designation in USP, the description in Ph. Eur.), the detector principle or the composition of the sample solution.
The main allowable adjustments
| Parameter | Allowed adjustment (summary) |
|---|---|
| Column length L and particle size dp | L/dp within −25 % to +50 % of the original; for a switch between fully porous and superficially porous particles, plate number N within −25 % to +50 % |
| Column internal diameter | may change, with flow rate scaled accordingly |
| Flow rate | scaled to the new dimensions: F2 = F1 × (dc2² / dc1²) × (dp1 / dp2); for isocratic methods a further ±50 % |
| Injection volume | may be scaled with the column volume, as long as precision, linearity and detection limit remain adequate |
| Gradient time segments | scaled so that gradient volume per column volume stays constant |
| Mobile phase pH (aqueous buffer) | ±0.2 units |
| Column temperature | ±10 °C for isocratic methods, where a temperature is specified |
| Isocratic mobile phase composition | minor component ±30 % relative or ±2 % absolute, whichever is larger, no component by more than 10 % absolute |
Check the current text of the chapter that applies to your method before you act on any line; the table is a working summary, not a replacement for it. The detection wavelength, for example, is not on the list of permitted adjustments.
A worked example: modernising a 250 mm column
A gradient method runs on a 250 × 4.6 mm column with 5 µm particles at 1.0 mL/min, 20 µL injection, gradient segment 30 min. The original column is discontinued, and the lab wants a 150 × 3.0 mm column with 2.7 µm superficially porous particles of the same L designation.
1. Is the column within limits?
Original L/dp = 250 mm / 5 µm = 50,000. New L/dp = 150 mm / 2.7 µm = 55,556. Change +11.1 %, inside −25 % to +50 %. (For a fully porous to superficially porous switch, you would also show the plate number stays inside −25 % to +50 %.)
A 100 mm column with the same particles would give L/dp = 37,037, a change of −25.9 %. That is just outside the limit, and the switch would be a modification.
2. Flow rate
F2 = 1.0 × (3.0² / 4.6²) × (5 / 2.7) = 0.79 mL/min.
3. Injection volume
Scaled with column volume: V2 = 20 × (150 × 3.0²) / (250 × 4.6²) = 5.1 µL.
4. Gradient time
t(G2) = t(G1) × (F1 / F2) × (L2 × dc2²) / (L1 × dc1²) = 30 × (1.0 / 0.79) × 0.255 = 9.7 min.
The run time drops from about 30 to about 10 min per gradient segment, solvent consumption by roughly 75 %, and the change is still an adjustment, provided system suitability passes.
The part that is easy to miss: dwell volume
Gradient methods are sensitive to the volume between mixing point and column inlet. Moving a method from a quaternary low-pressure mixing system to a binary high-pressure system can shift dwell volume by a millilitre or more, which on a smaller column means a delay of several column volumes before the gradient arrives. The pharmacopoeias allow adjusting the initial isocratic hold to compensate. Measure the dwell volume of both systems and document the correction; otherwise early-eluting peaks move and resolution criteria fail for reasons that have nothing to do with the column.
When is it no longer an adjustment?
- Different stationary phase chemistry, even if "equivalent" in a column database.
- L/dp or N outside −25 % to +50 %.
- Wavelength, detector type or sample preparation changed.
- Several adjustments combined that together change selectivity, even if each alone is within limits. The pharmacopoeias ask you to consider the cumulative effect.
- System suitability fails after adjustment.
In these cases, the path is change control, a documented assessment of which validation characteristics are affected, the targeted revalidation, and for registered methods the regulatory classification of the change. ICH Q12 established conditions and an approved post-approval change management protocol can make that path predictable.
Limits of the statement
- The allowable adjustments are written for pharmacopoeial methods. Many companies apply them to in-house validated methods through their own procedures; whether that is accepted for a given product depends on the registration and the authority.
- "Within limits" is necessary, not sufficient. The adjusted method must meet all system suitability criteria, and it is good practice to verify critical pairs on real samples, not only on the SST mixture.
- Equivalence of results before and after the change is not proven by the calculation. For release methods, a side-by-side comparison on representative batches is the evidence QA will ask for.
Frequently asked questions
Can I change the HPLC column length without revalidation?
Yes, if the ratio of length to particle size stays within −25 % to +50 % of the original, the stationary phase chemistry stays the same, flow rate and gradient are scaled, and system suitability passes.
Does USP <621> allow adjustments to gradient methods?
Yes. Since the harmonised chapter took effect in December 2022, column dimensions and particle size in gradient methods may be adjusted using the L/dp rule, with gradient times scaled to keep gradient volume per column volume constant.
How do I calculate the new flow rate after a column change?
F2 = F1 × (dc2² / dc1²) × (dp1 / dp2), with dc the internal diameter and dp the particle size. In the example, moving from 4.6 mm and 5 µm to 3.0 mm and 2.7 µm gives 0.79 mL/min from 1.0 mL/min.
Is a change of detection wavelength an allowable adjustment?
No. The detection wavelength is not on the list of permitted adjustments. A different wavelength is a modification of the procedure.
What if two adjustments are each within limits?
They are still allowed, but the combined effect on the separation must be considered and system suitability must pass. If the combination changes selectivity in a way the original validation did not cover, treat it as a modification.