Ask The Experts: dried blood spots

Do you believe there’s a viable future for dried sampling for regulated quantitative drug bioanalysis?

Neil: Yes, I do believe that there is a viable future for this technology. I think as all the speakers stated, we just need to understand the limitations of the technology and make sure our validation helps us to understand those limitations thoroughly, and apply this technology where the benefits outweigh the downsides. I think we have seen that there is a path for our data to be accepted, so long as we have done good quality experimentation and understand our assay and what our data actually means. I think there is definitely a way forward for this technology.

Will the FDA accept DBS?

Brian: Yes, in fact we are already seeing some applications in which people are exploring the use of DBS, and I suspect that people in the industry are waiting to see who is going to have the first big success with that, and that will probably lead to a wider use for specific applications within their drug development programs.

What are the major limitations of patient self-sampling at home?

Christophe: I think it is very important that patients are instructed very well so they know how to collect samples in the correct way, and they should also try to avoid the issue of contamination. Not all these things appear feasible. There have been several studies that have done patient sampling at home, and in which good results have been obtained. The intrinsic DBS quality can ?be evaluated by, for example, incurred sample reanalysis: if you look at different spots made from the same patient at the same time point and you get the same results, you can be confident that that result is okay. Apart from that you also need to be sure that the patient’s records also have the correct time of sampling. For example, if you wish to go for a pharmacokinetic approach, that is crucial.

Do you think that the stance of regulatory agencies on DBS for regulated quantitative clinical drug bioanalysis is reasonable?

Neil: Yes I think that it is absolutely reasonable. The regulators first job is patient safety and drug efficacy, and they need to make sure that when we are changing the way that we measure one of the important parameters to look at that, then the data we are generating has to be of at least the same high quality as that which we are all used to seeing. So absolutely, I think they have to take a very conservative approach.

Brian: We are really seeking to understand how we can compare the data, because we need to make sure that we are comparing apples to apples and oranges to oranges, and not apples to oranges.

Do you consider the HTC issue to be solved?

Christophe: In my opinion, the word ‘solved’ is perhaps a bit too strong, but I think we can cope with it. Whether it is trying to avoid it by volumetric application, whether it is by predicting the hematocrit, whether it is by using hematocrit independent filter paper or by using dried plasma spots (if indeed you can call it plasma), I think there are enough ways that allow you to cope with the issue, so I do not think at this time point that it is still a major issue. ?

What criteria will the FDA require for DBS validation?

Brian: It is a bit too early for us to tell, unfortunately. There is a lot of research that has been going on, and researchers in the industry have spent a lot of time looking at various different aspects of it and generated a lot of information. As far as the FDA goes, we do not have any of that data. We are just starting to see people submit requests about using it, and telling us what their plans are, and how they want to deal with things, but we have not received any data that we can evaluate for ourselves yet. Without that, it is hard for us to come up with specific criteria for acceptance, so that’s still a little bit down the road for us.

What about the potential limitations of other non-DBS microsampling techniques? What do they include?

Christophe: You have the impact of possible capillary versus venous difference, so in that respect, DBS are not standing alone. In fact, for any capillary, or any microsampling approach not involving ?venous blood, you also have to take that possible difference into account. There is not always a difference present, but you need to evaluate it. If ?venous concentrations do differ from capillary concentrations, then there is no problem as long as it is consistent.

Has the FDA seen any DBS in its applications so far?

Brian: We have not seen anything that has been approved at this point. Everything that we have seen has been earlier in development. There are a number of people in different places who are trying this out ? some are approaching it in terms of general use, others are using it in selected scenarios such as pediatric studies. We have not received any applications with this data in it yet.

How do you think we should get these new technologies accepted by regulatory authorities?

Brian: I do not think this is really a problem. I think there is a bit of a misunderstanding among people in the bioanalytical industry who seem to be a little bit afraid of the FDA that we are going to tell them they cannot use these kinds of things. There really is no official position – we do not really have a role to tell people what sort of technologies to use. We encourage you to use different kinds of technologies, DBS included. The only thing is that at the end of the day when you come in you are going to have to demonstrate to us that it is going to do what you say it can do. Outside of that, you could argue we are already saying it is okay, given that we have a number of people who are using it in their drug development.

Neil: I think it is our role as the scientific community to generate high quality data. When we have these new techniques, we need to make sure that we thoroughly understand the benefits as well as the issues with the technology. Where we see issues, we need to know how to control them, so that when we generate data that we submit to the different agencies, we all understand what that data means. We have to take the time to plan and perform the right experiments, and share the data with our colleagues. I think it has been very good seeing a huge amount of experimental work being done by a lot of laboratories on this. Some of them on their own, some of them together in consortia. When we work together in organizations such as the EBF, bringing together the brains of numerous companies to devise experiments together and publish that data, it has been particularly powerful. So that is our role – when we bring a new technology through, we need to understand its impact (both positive and negative) and make sure we are in control of that.

Is controlling the DBS spot size the only reason for differences in quantitation or are there any other issues?

Neil: Other issues include the fact that recovery can vary with hematocrit for some analytes, often with lower recovery observed at high hematocrit levels.

Christophe: I assume that the question is about the difference between volumetric and non-volumetric spotting, with complete DBS analysis in the former case versus taking a punch in the latter case. Important to note is that in the case of volumetric spotting, there is no real control of DBS size (only control of the volume spotted). When using punches, factors that need to be taken into account, in addition to the hematocrit, are:

• the effect of the volume spotted (larger volumes may give some higher saturation of the paper when compared to smaller volumes, leading to somewhat higher levels ? however, this may be compound dependent).
• possible ‘chromatography’ effect, which may give rise to central-peripheral differences.

Are there any FDA guidelines available for DBS collector development and testing?

Neil: The current draft FDA Guidance for Industry on Bioanalytical Method Validation, refers to DBS sampling. It states that, ‘A comprehensive validation will be essential prior to using DBS in regulated studies. This validation should address, at a minimum, the effects of the following issues: storage and handling temperature, homogeneity of sample spotting, hematocrit, stability, carryover, and reproducibility including ISR. Correlative studies with traditional sampling should be conducted during drug development. Sponsors are encouraged to seek feedback from the appropriate FDA review division early in drug development.’

Christophe: Important to note that this is a draft document. Moreover, nothing is mentioned with respect to acceptance criteria (do parameters need to be evaluated in combination and to what extent do volume and hematocrit need to be examined).

Brian: Agree with Neil and Christophe. Due to a lack of data within the Agency, we are unable to provide acceptance criteria at this point in time. However, the take home messages are to consult the agency early in development and to compare results between DBS and your prior matrix/sampling approach.

Are there available techniques to overcome the issue of the high dilution factor applied to prepare DBS samples for LC-MS/MS quantification?

Neil: The use of the most recent sensitive mass spectrometers helps make up for the small volumes and large dilutions involved when DBS samples are analyzed. In addition, a number of manufacturers (Camag, Agilent SCAP, Spark Holland) have made automated DBS elution systems available that transfer the entirety of the elution volume into the LC-MS system, rather than injecting a sub-aliquot.

Christophe: The high sensitivity offered by the new generation MS/MS systems allows quite straightforward sample prep in many cases: extraction, dilution and injection into the LC-MS/MS. In fact, this high dilution factor can be considered as an advantage rather than an issue, because matrix effects may be virtually absent. Obviously, you need high-end equipment for this purpose.

For the majority of drug development or TDM, would it be most ideal to analyze whole blood or plasma?

Neil: The decision whether to analyze blood, or plasma for the majority of drug development should be based upon careful consideration of the blood:plasma ratio, hematocrit, unbound fraction in plasma and blood cell partitioning. See Emmons & Rowland (2010) Bioanalysis 2, 1791 for a comprehensive discussion on this issue. However, for certain situations, other factors may be important, for example sampling in remote locations, or sampling from very small children, or animals.

Christophe: This is something which needs to be evaluated on a case-by-case basis. For drugs that concentrate in RBC, it makes less sense to quantify in plasma (and vice versa).

If complete dossier is DBS do we need to do correlation studies? If yes, could you explain the rationale?

Brian: That doesn’t seem necessary if you are developing a NCE. On the other hand this may not be acceptable for an ANDA, if the innovator was developed traditionally with plasma samples. As a general guide, when you switch from one ‘platform’ to another, a comparison of the data will be needed in order to appropriately interpret the data.

There are many divisions of drug review. How do you assure consistency across the board?

Brian: The Guidance is available to all, and we have several internal training and seminar opportunities to disseminate information. Furthermore, reviewers consult one another frequently when dealing with issues that are new to them.

In terms of validation, what additional work ias required beside hematocrit variation test and volume variation test?

Neil: The current draft FDA Guidance for Industry on Bioanalytical Method Validation, refers to DBS sampling. It states that, ‘A comprehensive validation will be essential prior to using DBS in regulated studies. This validation should address, at a minimum, the effects of the following issues: storage and handling temperature, homogeneity of sample spotting, hematocrit, stability, carryover, and reproducibility including ISR. Correlative studies with traditional sampling should be conducted during drug development. Sponsors are encouraged to seek feedback from the appropriate FDA review division early in drug development.’

Christophe: A lot depends how far you wish to go. You may want to evaluate all above-mentioned parameters in isolation or in combination.

Regarding comparison of plasma data to DBS data, isn’t a comparison only relevant if plasma data was previously generated? That is, for a new drug candidate, if you never have generated any plasma data then why would there be a need for a comparison?

Neil: I agree. If plasma data has never been generated, then a DBS to wet plasma comparison should not be required.

Christophe: I completely agree. DBS-blood comparison should suffice in this case. Attention should be paid to a potential capillary – venous difference, however.

Brian: Yes, although there may be some other considerations. I refer you to the response on the earlier question about this issue.

How important is cleaning up the DBS extract (SPE, L;L extraction) before analyzing on LC-MS/MS?

Neil: The experience in our laboratories, where we have analysed large numbers of small molecule analytes by DBS, is that simple aqueous methanolic extraction is sufficient.

Christophe: Also in our experience samples are pretty clean, as long as you don’t include too much water in the extraction solvent. A typical ‘general’ condition might be 80% MeOH/water (+ formic acid), then simple dilution to LC starting conditions and injection into the LC.

Should we compare DBS with plasma data or compare DBS to wet blood?

Neil: It depends on what matrices you are using for other studies and species. If all your work is with DBS, then the comparison should be DBS with wet blood. However, if you have some studies/species in plasma and others in DBS, then you will need to compare DBS with wet plasma.

Christophe: It depends on the context: if all previous data have been generated in plasma, then that can be considered the ‘reference matrix’ to compare with. Otherwise, DBS – wet blood is definitely okay. In fact, this is not something DBS-specific. It actually has to do with blood-plasma comparisons.

Do you think DBS could have greater regulatory acceptance for certain class of drugs (e.g. those exhibiting significant RBC partitioning)?

Christophe: A DBS-blood comparison is easier than a DBS-plasma comparison. In fact, only the ‘technical’ DBS issues have to be dealt with (hematocrit effect, evenly spreading, volume effect etc.).

Brian: From a regulatory perspective, we leave the choice of technologies used to industry. It is up to you to choose when you think DBS is more appropriate than plasma sampling. I don’t foresee any type of scaled response based on drug class at this point.

Plasma is a component of blood, so shouldn’t blood be the preferred matrix for drug measurements? All the manipulations to make plasma from blood could add bias to the data.

Neil: I advise the requestor to read Emmons & Rowland (2010) Bioanalysis 2, 1791 for a comprehensive discussion on this issue.

Christophe: I principally agree. However, there is a long, long history of using plasma that won’t be easily changed.

Do you recommend to assess the comparison between blood and plasma in human even if the PK matrix is blood from the beginning of your development (animals TK & human PK)?

Neil: If all data has been generated in blood, then the only comparison that is needed is between wet and dry blood samples.

View the panel discussion here.