Dear colleagues,

I have a regulatory question with respect to validation of a RP-HPLC method. We have developed a RP method (C4-column) to determine the amount of degradation in a pharmaceutical protein product.
Essentially the amount of degradation is calculated from the relative peak area responses (i.e. the peak area of degs. is compared with total peak area). Although we have extensively validated the method we see that now and then the amount of degradation depends on the column batch used. Furthermore, we cannot prevent this batch-to-batch variation by priming procedures (saturation with protein solutions doesn't help). So, we have tried another brand of column (also C4)and we have noticed that this column performs much better with respect to batch-to-batch variation. Furthermore, the new column meets all system suitability criteria that were set for the old column. So, my question is: "can we use the new column without any further validation or should we perform additional experiments?". Note that in our protocols we refer to a specific column (brand and type) but we also state that equivalent columns are acceptable as well.However, up to now we have never switched to a new type of column without a revalidation.

Thanks in advance for your comments

I think that this is a very gray area. What I believe you need to do is to satisfy yourself, in a scientifically sound and defendable manner, that the new brand of column is as good as, or better, than the old. What this actually entails is up to you, however, since you may be required to defend your reasoning at some point in the future, too much data is probably better than too little. One additional test you may wish to consider is a side by side comparison of the old brand and new brand columns using the same sample preparations and instrument just to confirm that, on average, they really do provide statistically equivalent results.

Anyone have other suggestions? This is a topic that I would expect to generate some interesting discussion.

Best Regards,

Michael Hinsberg

http://www.hinsbarlabs.com

I experimented with an "equivalent" column recently, and if I had chosen to use it I would have felt I had to do a full validation. The new column had the same dimensions, same basic packing material (C18), but the effect on separations was profound. Retention times changed, separation quality changed, and so on. Using the new column would have entailed a change in the mobile phase program. While the new column would have been nice, (read cheaper) it simply would have entailed more work to start using it.

So, in response to the first post, it really depends. If the column causes a change in the chromatography (separations, retention times) you would probably be well advised to at least show that you haven't caused peaks to combine.

regards
Brad

Take a look at Furman, Dorsey, and Snyder, Pharm. Tech, 1998, 22(6), 58..64