Log in
Register|
Geokad ☆ Canada, 2009-04-14 20:15 (5875 d 02:17 ago) Posting: # 3521 Views: 6,353 |
|
|
Hi, I have the two following questions: 1-During interference evaluation from OTCs in method validation, do we have to test for their metabolites? 2-Does anyone test the interference coming from contraceptives? if yes which ones and what is the procedure? — Regards, Geokad |
|
Helmut ★★★   Vienna, Austria, 2009-04-14 21:45 (5875 d 00:46 ago) @ Geokad Posting: # 3525 Views: 5,708 |
|
|
Dear Geokad! ❝ [...] two following questions: Very good ones! ❝ 1-During interference evaluation from OTCs in method validation, do we ❝ have to test for their metabolites? I think we have to distinguish whether samples are coming from a controlled phase I PK study (BA/BE, food, interaction, blahblah), from phase II/III or 'the field' (clinical practice, TDM,...). In almost all phase I protocols I have seen intake of OTC drugs was prohibited for a certain time interval before the first dose. OK, that's no guarantee... You can exclude most interferences with your analyte by common sense (comparing lipohilicities, and pK-values). If you have a selective detector (fluorescence, MS, MS/MS) you shouldn't bother that much. On the other hand metabolites are generally more hydrophilic than the parent and appear earlier in an RP-system... Many people use a pragmatic approach: headache is commonly seen in PK studies (not due to the drug, but to the 14 hours fast, the clinical setting, 'bad' sleep, repetitive blood sampling, etc.). Therefore most protocols allow for administration of e.g. paracetamol/acetaminophen. The analyst should test for a potential interference and give his/her OK. If an interference with the analyte is seen another painkiller has to be selected and tested. To be honest, I have never seen anybody testing for metabolites. In practice sometimes large interferring peaks are seen in chromatograms (especially in multiple dose studies, when the saturation phase was done in an out-patient status and volunteers take drugs...). This is a little bit tricky, but often a slight modification of the method + partial revalidation does the job. ❝ 2-Does anyone test the interference coming from contraceptives? if yes ❝ which ones and what is the procedure? I haven't seen that - but again based on the physicochemical properties of the contraceptive(s) and your analyte I would follow common sense. — Dif-tor heh smusma 🖖🏼 Довге життя Україна! ![[image]](https://static.bebac.at/pics/Blue_and_yellow_ribbon_UA.png) Helmut Schütz ![[image]](https://static.bebac.at/img/CC by.png) The quality of responses received is directly proportional to the quality of the question asked. 🚮 Science Quotes |
|
Ohlbe ★★★ France, 2009-04-17 19:49 (5872 d 02:42 ago) @ Helmut Posting: # 3558 Views: 5,523 |
|
|
Dear Helmut and Geokad, In an ideal world, we would have all the time and money to do a great job. You would indeed test not only possible concomitant medications, but also their metabolites. And testing would not be limited to just checking for visible interferences, but also for possible matrix effects if using LC/MS/MS. (yes, concomitant medications can cause ion suppression or enhancement, even just ibuprofen - see for instance Leverence R., Avery M.J., Kavetskaia O, et al. Signal suppression/enhancement in HPLC-ESI-MS/MS from concomitant medications Biomed Chromatogr. 2007 Nov;21(11):1143-50 We don't live in an ideal world... Regards Ohlbe |
|
ElMaestro ★★★ Denmark, 2009-04-19 07:44 (5870 d 14:47 ago) @ Ohlbe Posting: # 3565 Views: 5,475 |
|
|
Dear all, I think this was an interesting question, and I agree with the two responders. But I wonder about a thing: Let's assume that there is an interference between OTC-drug/contraceptive X and the analyte Y. That means that the subjects subjected to X will have measurements of Y that are wrong (or more wrong than they would be in the absence of X). In this situation, who runs the risk, the sponsor or the patient? In BE-studies randomisations are always used, which means the patients subject to factor X will be 'equally' (on average) partitioned into the two sequences. That's one headache eliminated. If we assume the interference is equal in the two periods, the interference should have equal effects on T and R provided that T and R are actually equivalent. So, this interference will just translate into higher variability. Sponsor's risk, I guess. However, I am not able to convince myself that this is also so when T and R are not actually equivalent, as interferences between X and Y typically show a bit of dependence on the amounts of both. Am I wrong? Thank you. EM. |
|
Helmut ★★★ Vienna, Austria, 2009-04-19 16:36 (5870 d 05:56 ago) @ ElMaestro Posting: # 3566 Views: 5,457 |
|
|
¡ ElMaestro ! ❝ Let's assume that there is an interference between OTC-drug/contraceptive ❝ X and the analyte Y. [...] ❝ In this situation, who runs the risk, the sponsor or the patient? Both, if the interference is present in only one occasion. Since CVintra will be higher, the sponsor; but imagine this situation: the study was powered enough to still show BE and the undetected interference shifted the T/R towards 1, then it's only the patient's risk (without 'help' from the interference the product would not be BE)... That's why we have to check for (lack of) interferences in bioanalytics. Randomization does not protect us if the interference occur in one occasion only. And generally they will: OTC drugs are not taken regularily (actually the intake is prohibited, but...). Contraceptives are especially nasty because the hormonal content and composition in the formulation varies from day 1 to day 28 (actually no hormones are contained in the 4th week's 'pills' to trigger menstruation...). If female subjects don't come from the same dormitory of a finishing school (probably periods are synchronised) you will see a lot of different amounts of interferences... But that's another story (NLYW!). — Dif-tor heh smusma 🖖🏼 Довге життя Україна! Helmut Schütz The quality of responses received is directly proportional to the quality of the question asked. 🚮 Science Quotes |
|
ElMaestro ★★★ Denmark, 2009-04-21 03:20 (5868 d 19:11 ago) @ Helmut Posting: # 3570 Views: 5,389 |
|
|
Dear HS, I think you answered a question I did not ask, and I guess that comes down to my lack of language skills. But it stil remains very interesting, if I try to think along your lines. ❝ ❝ In this situation, who runs the risk, the sponsor or the patient? ❝ Both, if the interference is present in only one occasion. One occasion, do you mean in one period? Or do you mean in one patient, both periods? ❝ Since CVintra will be higher, the sponsor; but imagine this situation: ❝ the study was powered enough to still show BE and the undetected ❝ interference shifted the T/R towards 1, then it's only the patient's risk ❝ (without 'help' from the interference the product would not be BE)... It sounds to me like this does not require the interference to be present in just one period, cf above. The example partially illustrates my previous point  . ❝ That's why we have to check for (lack of) interferences in bioanalytics. Or.... it could be a very good reason to think more seriously about period effects. It could come down to interference; for example, naïve patients are known to be more susceptible to distress and require NSAIDS in the first period of a multiperiod study (or the first days of a multi-day study). This phenomenon will increase interference in period 1. ❝ Randomization does not protect us if the interference occur in one ❝ occasion only. And generally they will (...blah blah followed by some ❝ gynecholophystic considerations). Can the term 'generally' in this context be qualified in some way? Could you rephrase the paragraph? I lost it around here, sorry. I agree with the facts about hormones and contraceptives but I am not sure I understand your point. Best regards EM. |
|
Helmut ★★★ Vienna, Austria, 2009-04-21 15:19 (5868 d 07:12 ago) @ ElMaestro Posting: # 3573 Views: 5,472 |
|
|
Dear ElMaestro! ❝ I think you answered a question I did not ask, and I guess that comes down ❝ to my lack of language skills. Oh I think your language skills are better than mine. Occasionally I hit the hay common to German native speakers: writing too long English sentences... ❝ ❝ ❝ In this situation, who runs the risk, the sponsor or the patient? ❝ ❝ Both, if the interference is present in only one occasion. ❝ One occasion, do you mean in one period? Or do you mean in one patient, ❝ both periods? One period. Remember this thread is in the 'Bioanalytics' category. The original question arose from FDA's Bioanalytical Method Validation Guidance ( Selectivity is the ability of an analytical method to differentiate and quantify the analyte in the presence of other components in the sample. [...] Potential interfering substances in a biological matrix include endogenous matrix components, metabolites, decomposition products, and in the actual study, concomitant medication and other exogenous xenobiotics. If an interference goes undetected we are in deep shit; no statistical method whatsoever will rescue us.-- Edit: Link corrected for FDA's new site. [Helmut] — Dif-tor heh smusma 🖖🏼 Довге життя Україна! Helmut Schütz The quality of responses received is directly proportional to the quality of the question asked. 🚮 Science Quotes |
|
NewInPK ☆ 2009-05-13 04:20 (5846 d 18:12 ago) @ Geokad Posting: # 3682 Views: 5,460 |
|
|
Hello, The reality is that no matter how much time and money a lab may have, there are limitations during the validation period because one can never test for all OTCs, their metabolites and the excipients. However the latest "incurred samples re-analysis (ISR)" tests that are being done could help as a continuation of the validation. As far as I know, they are done using the validated conditions, which could generate very reproducible results and not show any hidden problems. However, if we were to apply the same general principles employed during troubleshooting (modifications of the mobile phase, different column, different fragment etc) and perform the ISR tests, chances are some potential issues may be uncovered. Repeating a couple of samples from each subject at some strategically chosen time points, using as small of a change as a very shallow gradient which would increase the retention time by, say, 5 times, may take an extra day, but would give more (albeit not total) confidence in the data. I would even go as far as using a different (maybe milder) extraction for high concentration samples, were sensitivity is not an issue. For instance, a simple protein precipitation for samples normally extracted with SPE, especially if the SPE involves the use of solution with very high or very low pH. Cheers! |
|
Helmut ★★★ Vienna, Austria, 2009-05-16 18:49 (5843 d 03:43 ago) @ NewInPK Posting: # 3688 Views: 5,500 |
|
|
Dear NewInPK! I guess you may be "New in PK" but also an "old salt" in bioanalytics! Your suggestions are the best I've read in the last years! IMHO ISR for a validated method (stability and freeze-thaw given) will confirm original results in most cases. What have we gained? I remember a statement Rudolf E. Kaiser (one of the pioneers of gas chromatography) made in the early 1980ies: 'I consider GLP death of analytical chemistry. The only way one can assess the reliablity of a result - namely applying an alternative analytical methodology - is not allowed.' Since in most cases an alternative method is not available, it's a good idea to challenge the one we have - far beyond rubustness/rugedness...— Dif-tor heh smusma 🖖🏼 Довге життя Україна! Helmut Schütz The quality of responses received is directly proportional to the quality of the question asked. 🚮 Science Quotes |
|
Ohlbe ★★★ France, 2009-05-18 01:09 (5841 d 21:23 ago) @ Helmut Posting: # 3689 Views: 5,299 |
|
|
Dear Helmut and NewInPK, ❝ IMHO ISR for a validated method ❝ (stability and freeze-thaw given) will confirm original ❝ results in most cases. What have we gained? Except for metabolite back-conversion, which is seldom tested during validation ! That's where this ISR business started from, after all. Apart also from lab errors (switching samples), insufficient sample mixing, etc (of course these should not happen if you work properly. But there are some bad labs on this planet too, and many recommendations are made just because of them). IMHO ISR is a bad answer to a real problem, as it will only give you information on between-run precision, but no information on accuracy (your method could be consistently wrong). Unfortunately nobody came up with a better suggestion... apart from NewInPK in his post, which would bring him many enemies in this business if it were to be implemented ! ❝ I consider GLP death of analytical chemistry. I would consider GLP applied by QA personnel/auditors/regulatory inspectors knowing nothing about bioanalysis, to be harmful to analytical chemistry. In France we are lucky to have some GLP inspectors who know also about scientific aspects, and who like study personnel to use their brains, and not just follow the SOPs. ❝ Since in most cases an alternative method is not available, it's a good ❝ idea to challenge the one we have - far beyond robustness/ruggedness... Agreed. But doesn't it mean you have to validate your alternative method too ? Twice the cost and time... And who's going to pay: the lab, or the sponsor ? There are already some discussions between some sponsors and the CROs about ISR, as the sponsors say that ISR is part of validation ! Regards Ohlbe |
|
Helmut ★★★ Vienna, Austria, 2009-05-18 01:42 (5841 d 20:49 ago) @ Ohlbe Posting: # 3690 Views: 5,346 |
|
|
Dear Ohlbe! ❝ ❝ IMHO ISR for a validated method [...] ❝ Except for metabolite back-conversion, which is seldom tested during ❝ validation ! Very true. ❝ That's where this ISR business started from, after all. Oh, I didn't know that! Thanks for the historical perspective. ❝ [...] ISR is a bad answer to a real problem, as it will only give you ❝ information on between-run precision, but no information on accuracy ❝ (your method could be consistently wrong). Exactly. The concept of accuracy in bioanalytics is current only seen within the framework of a particular method. If I remember it correctly in forensics two different methods had to applied to provide legal acceptable evidence. In the late 1970ies (the dawn of GC-MS) this requirement was relaxed, since everybody started to believe in high selectivity... ❝ Unfortunately nobody came up with a better suggestion... apart from ❝ NewInPK in his post, which would bring him many enemies in this business ❝ if it were to be implemented ! Yes, he/she is courageous - but the arguments are scientifically quite strong. ❝ ❝ I consider GLP death of analytical chemistry. ❝ [...] In France we are lucky to have some GLP inspectors who know also ❝ about scientific aspects, I know at least one. ❝ [...] and who like study personnel to use their brains, and not just follow ❝ the SOPs.  ❝ ❝ Since in most cases an alternative method is not available, it's a good ❝ ❝ idea to challenge the one we have - far beyond robustness/ruggedness... ❝ Agreed. But doesn't it mean you have to validate your alternative method ❝ too ? Only to some extend. Forget about stability, freeze-thaw-cycles, bla bla; a partial validation should do the job. Since we are in the wide field of heresy now, I would even skip batch-to-batch accuracy/precision and only check linearity, LLOQ, and a set of QCs. That's one day work for the validation and another one for NewInPK's samples... ❝ Twice the cost and time... And who's going to pay: the lab, or the ❝ sponsor ? There are already some discussions between some sponsors and the ❝ CROs about ISR, as the sponsors say that ISR is part of validation ! Costs are subjected to negotiations. The question is: are we really interested in getting (more) reliable results? — Dif-tor heh smusma 🖖🏼 Довге життя Україна! Helmut Schütz The quality of responses received is directly proportional to the quality of the question asked. 🚮 Science Quotes |
|
Ohlbe ★★★ France, 2009-05-18 12:22 (5841 d 10:09 ago) @ Helmut Posting: # 3694 Views: 5,274 |
|
|
Dear Helmut, Basically, what NewInPK and you are suggesting is to do ISR, but with a different method  Oh, another possible source of problems, not studied during validation, but which would be found this way: the influence of some excipients, such as Tween 80 or PEG 400. Not found in BE studies, but common in pre-clinical, and possibly in early clinical trials. Can cause matrix effects. Several publications showing these substances to be problematic after IV administration ; possibly also after oral administration. ❝ I know at least one. The French inspector I know you have met is a GCP inspector, not GLP. But I've heard he works closely with his GLP colleagues. Regards Ohlbe |
|
Helmut ★★★ Vienna, Austria, 2009-05-18 13:31 (5841 d 09:01 ago) @ Ohlbe Posting: # 3696 Views: 5,382 |
|
|
Dear Ohlbe! ❝ Basically, what NewInPK and you are suggesting is to do ISR, but with a ❝ different method Essentially yes. BTW along the way it would be nice to change the term. Most analysts simply talk about 'real world samples' to distinguish from spiked ones. I have no opinion about 'incurred samples' - an expression which has no easily understandable meaning for me in the English language. — Dif-tor heh smusma 🖖🏼 Довге життя Україна! Helmut Schütz The quality of responses received is directly proportional to the quality of the question asked. 🚮 Science Quotes |
Ing. Helmut Schütz