Nadine Ritter is president and senior analytical advisor of Global Biotech Experts, LLC and a long-time member of BioProcess International’s editorial advisory board. At a recent CASSS North American CMC Strategy Forum called “Methods on the Move: Addressing Method Transfer Challenges,” she discussed the biopharmaceutical industry’s logistical challenges of analytical test samples for drug substances and products. At the conference, BPI’s editor in chief Anne Montgomery met with her to discuss some key points of this topic.
Montgomery: What are some main issues for contract laboratories regarding sample testing?
Ritter: One issue for analytical testing programs (contract and in-house) is that samples needing analysis must be transferred from where they’re taken (e.g., from a manufacturing floor or stability chambers) to testing locations. For in-process testing, this usually is not so much of a problem because those laboratories tend to be close enough to manufacturing facilities to allow rapid analysis and turnaround of test results. I see the problem most frequently with drug-substance and drug-product samples that must go to one or more locations to complete their required analytical test sets.
Some of those analyses are physicochemical methods, some are chemical methods, some are bioassay methods, and some are microbial methods. It’s rare that a single laboratory can conduct all necessary analyses in the same facility, especially given that some analytical methods require highly diverse techniques and special instrumentation. But even if every test can be performed at one location, samples must transit from their originating locations to that location.
Shipping or even just carrying biopharmaceutical test samples to one (or more) analytical laboratories is not trivial. The goal is to minimize physical stress to those samples so that analytical results will reflect most accurately the nature of the batches they represent. To do that, each sample should be packaged into an appropriate shipment container at the desired temperature and then sent to a designated testing site. Upon receiving samples, laboratories should hold them at appropriate conditions of temperature and time until analyses are started.
Shipping validation involves showing reliability of a shipping procedure for moving a bulk drug substance (e.g., bottles or bags of a bulk-substance batch) from its manufacturing location to a place where a drug product will be formulated and filled — or moving a drug-product lot from production facilities to a warehouse and beyond. That shipping mechanism typically undergoes a rigorous validation process to ensure that it will maintain the intended storage conditions as shipments transit their chain of custody. However, it is much less common to see validation of a shipping process for analytical samples alone. Moreover, facilities and laboratories often have significant inconsistencies in such practices, even within the same company and even with the same product. I’ve seen samples shipped one way for Batch A but later shipped another way for Batch B of the same product, just because no one knew the samples were being treated differently. I’ve also seen samples from one batch of product shipped one way to Lab X and another way to Lab Y; or upon receipt, Lab X holds samples at one temperature, whereas Lab Y holds them at a different temperature. I’ve seen test samples go to a laboratory site, which then sends or subcontracts some tests to another laboratory site. So the first laboratory would receive all samples, separate and reallocate them, and then ship a subset to somebody else. That may be okay for individually vialed test samples, but I’ve seen large-volume bulk samples thawed, subdivided, refrozen, and shipped to a subsequent testing site.
So how are analytical test samples from disparate, undefined conditions of shipping and handling expected to reflect the quality or stability of the product batch from which they came? Or how do they provide reliable data on manufacturing consistency from product batch to product batch?
Temperature and Agitation Considerations
Montgomery: Would you say shipping of test samples is a high-risk process?
Ritter: Yes, it is. Accurate test results are critical to making decisions on product quality and stability. We biotech analytical folks have enough challenges with optimizing, validating, maintaining, and troubleshooting the performance of a plethora of analytical technologies. We definitely don’t want additional complications from uncertainties about the test samples themselves. But numerous best-practices have been developed to minimize that risk.
First, it is incumbent upon those who are responsible for managing analytical testing to define all aspects of the chain of custody for test samples — from the original sampling source to the bench where each test will be performed. That includes testing for the release of each batch of drug substance or drug product as well as testing required in stability protocols.
It is not uncommon to have ICH stability chambers in a separate location from manufacturing sites and most testing sites. Be sure to document exactly where samples taken from manufacturing and from each stability pull point have to go to ensure that you know what tests are required of each sample. That information also is critical for product regulatory filing because the name and location of each testing laboratory conducting release or stability testing is included in a product submission as a part of the approved facilities.
Next, determine the appropriate temperature conditions under which to transfer analytical samples to each testing site. It would be wise to use the same temperature as that used during storage of the source material to maintain the nature of the material submitted for analysis. But that is not easy in all cases, particularly with stability studies as discussed below. For shipping batches of product, most biomanufacturers use shipping companies that have validated controlled-temperature units for frozen, refrigerated, or controlled-room temperature (CRT) conditions. But it is less common to see that type of product shipper used for analytical test samples, which typically are packaged in much smaller shipment containers.
For materials stored frozen, the typical approach is to ship test samples on dry ice. Assuming material has been shown to be stable for a number of freeze–thaw cycles, shipping protein solutions in a frozen state is useful in that it confers minimal (if any) distress from agitation.
For materials stored refrigerated as liquid solutions, shipping on wet ice is most common. However, shipping liquids on wet ice poses two challenges: The actual temperature at different locations in a box can vary widely depending on how it is packed, and shipping liquids exposes them to agitation. If high-concentration protein solutions in containers with lots of headspace are subjected to extreme agitation during shipment, you could end up with something that resembles whipped cream on the other end — great for meringue, terrible for biopharmaceutical products. So pay attention to the type of sample vials and volumes used when shipping samples as liquids, and try to minimize headspace.
Shipping lyophilized test samples on wet ice is less problematic, of course, because the material is in a solid state. Also, pay attention to how the samples (liquid or lyophilized) are packed in the shipping box relative to the wet ice. Ensure that all samples are equidistant from the ice in the container for temperature consistency.
Shipping biopharmaceutical products at CRT is similar to shipping on wet ice in the challenges for internal temperature consistency and agitation of liquid solutions. Be certain that you know what the receiving laboratory will do with your samples as they await testing. Will frozen samples be thawed and held for a week before testing? Will CRT samples be placed at 2–8 °C? Some testing laboratories have customary internal practices that may or may not be suitable for your specific test samples. Always confirm what is necessary for your samples, and ensure that every testing site treats them the same way upon receipt.
Finally, it is definitely best practice to include a temperature-monitoring device in every shipment box of analytical test samples. Such devices continually record the temperature from origin to recipient. The data they collect documents the range of temperature the samples experienced en route. I even advocate using them in containers for samples being hand-carried from one location to another in case of a delay or other event that could affect the hold temperature. I’ve audited laboratories in Wisconsin that move test samples from one building to another through three feet of snow — as well as in Arizona when it is 112 °F in blazing sun. Whether traveling thousands of miles or hundreds of feet, if something does go awry, I want to have a temperature data logger to prove it. If so, I can invalidate that sample shipment and request a new one. Or I can rule out sample distress as an assignable cause for an out-of-specification (OOS) result that might arise from testing those samples. Investigation of OOS events should include assessing whether they were a result of distress on test samples rather than the true quality of source materials.
Shipping or transfer logistics can be even more complicated for analytical test samples that come from a stability program. Normally, for batch-release testing one target storage condition applies to every drug substance, and one target storage applies to every drug product. So transport of test samples can be designed for the same targets. But in an ICH stability program, two or three temperature conditions can be used (e.g., if accelerated and stress conditions are in a protocol). That means that at certain time points, a set of two or three stability condition test samples will be pulled and shipped. But at what temperature? Test samples may have come from –25 °C, 5 °C, and 25 °C conditions. Are you going to freeze them all and ship them frozen? Are you going to ship the ones at 25 °C in a box at CRT, then ship the ones that are at 5 °C on wet ice, and the –25 °C on dry ice? And what will a receiving laboratory do when it gets that set of samples: return them to their ICH conditions, hold them all at 2–8 °C, or freeze them all?
Going through customs can add another layer of risk. If test samples must be shipped to another country, then they probably will have to go through a customs inspection. An unpredictable time period for custom inspections can increase risk to sample integrity if shipping temperatures are not maintained long enough. Sometimes samples themselves take up the space of a loaf of bread, but they’re in a box the size of a television because all the rest is dry ice.
With one exception (see below), there are no rules about the shipping process. It is up to you to determine how you are going to pull analytical samples and ship them for release testing or from stability chambers to every recipient testing site. If you are a contract laboratory, then it is wise to discuss with each sponsor exactly how it will be shipping samples to you for testing as well as what should be done with them after receipt but before testing.
When I directed a contract laboratory, we would outline those procedures with each client, especially for stability studies. Our ICH chambers were off-site in a secured location, so we placed instructions in each stability protocol on how pulled samples would be managed for that product, based on discussions with the client (and data to support it). If data showed no negative impact of at least one or more freeze–thaw cycle, most clients wanted stability pulls from any temperature to be shipped frozen and held frozen before thawing for analysis. Logistically, that was certainly the easiest option. But some clients needed the stability samples shipped and held at ICH temperatures from which they were pulled. That meant a pull from three ICH conditions required three different boxes to put the samples in at three different temperatures. The samples had to be carried back to the laboratory and put back at three different temperatures until they were analyzed. It was a challenging process (e.g., triple documentation) but possible with adequate planning.
Stability-testing logistics can get even more complicated if ICH stability chambers are not all in the same location. Now that I am a consultant, I’ve seen situations in which stability samples at their target ICH temperature are in one facility, but a different facility will have stability chambers at different ICH temperatures. So for every pull point of a single batch of product, stability samples would come from different temperatures and from different chamber locations. And one step further: I’ve seen stability samples pulled from a target storage condition tested in one laboratory, but samples pulled from accelerated conditions tested in a different laboratory — both (allegedly) using the same test methods. And the test results for each temperature at each pull point are expected to align seamlessly for each analysis method.
Analysis and Documentation
Montgomery: So when then the results come back then, how do you know that the samples being tested are all right?
Ritter: Human nature is such that if you get analytical data in line with what is expected, you don’t question anything about the analysis, including the condition of the test sample that was analyzed. (Why should you? It passed, right?) It’s only when you get unexpected test results that you suddenly look for something wrong, particularly if it’s an OOS result. Only then will you begin to track the root cause of the problem.
A laboratory’s OOS investigation should include whether a test sample was distressed during its chain of custody to or in the testing laboratory. If so, it means an invalid result for the analysis, with justification to obtain a new sample that is not distressed. Including data loggers in each shipment of analytical test samples can provide the proof needed to support such a conclusion and justify a retest using new samples.
However, you should not wait until an unexpected result occurs to verify the integrity of each shipped test sample. Have operational practices in place to define visual inspection of test-sample shipments at the time they are sent and at the time they are received. Such practices would include noting
- the physical state of each test sample vial (e.g., frozen solid, intact lyophilized cake, or clear liquid solution)
- the orientation of samples in the shipping container (e.g., vials individually wrapped in bubble wrap or each separated from ice blocks by two sheets of paper)
- the addition of data loggers.
Large boxes with critical temperature ranges might include two data loggers, to bridge the internal space. Senders should conduct such observations, and recipients should repeat them immediately upon receipt when shipping boxes are opened. If a test sample is observed to be — or even suspected to be —altered in condition from the sending state, that should be noted in the receiving records and communicated to the testing team. Discussions should be held about whether the affected samples should be tested or invalidated and replaced. That decision should be justified and documented. It may be possible to photograph the samples before shipping and after receipt to further objectively document the physical condition of materials.
Montgomery: Is it to be expected that those practices be captured in written procedures for sending and receiving test samples?
Ritter: Yes. Needless to say, all such practices should be described in operational standard operating procedures (SOPs) or other written instructions to ensure that they are conducted correctly and consistently for every shipment of test samples, both at the sending site and at the receiving site. If samples are intended to be analyzed under good manufacturing practices (GMPs), then documentation is required. GMPs include a statutory requirement for written instructions of all work procedures associated with the manufacturing and testing of products, with documentation to show that such procedures were conducted properly. Several FDA warning letters over the years have included observations that no written procedures exist for pulling, preparing, or shipping GMP analytical samples to their intended testing locations. But beyond documenting such critical routine practices “on the ground,” I recommend that project teams take a step back to outline and document the logistics of their entire testing scheme for product batches. Most project teams do a pretty good job of answering two questions: Where in a process will samples be taken? What testing will be performed on each sample? But far too frequently they miss the next two logical questions: Where will each of those tests be conducted? How will we reliably get the test samples to each testing site? Contract testing laboratories often are highly aware of the differences in those practices. That is because they see many examples of undefined or inconsistent test-sample shipping strategies between — and even within — sponsors’ projects. But of course, when something goes wrong with a product, it is always the testing laboratory’s fault!
Montgomery: That is much more complicated than I realized.
Ritter: Yes, it’s an interesting complication because usually each product has one drug-substance process and one drug-product process. Everything is performed in one facility for drug-substance manufacturing or drug-product formulation and filling. Materials go into their designated location for use in drug substances or drug products. But each drug-substance and drug-product batch may require testing by a dozen different analytical methods at time of batch release and over time on stability for physical, chemical, functional, and microbial attributes. Those analyses (and therefore the test samples on which they are conducted) can be spread across half a dozen different analytical and microbial testing laboratories at distant locations.
The data should provide a high degree of assurance that each result accurately reflects the characteristics of the material the test samples represent. With those results, manufacturers can determine whether a batch or stability pull point truly passes or fails its specifications. The process can require complicated logistics, but it is a vital part of ensuring reliability of the data generated by each testing laboratory. A laboratory can have the most robust, accurate, precise, and validated test methods in the world, but those qualities can’t ensure reliable results if samples being tested do not reflect the quality of the source materials from which they were taken.
Many contract testing laboratories have a dedicated department for test-article receiving. For a contract laboratory, every test sample analyzed comes from somewhere else. But even small or in-house testing laboratories should have defined, documented processes for receiving test samples.
Regardless of laboratory size, it must have clear written procedures on how to inspect test-sample shipping containers and all test sample vials as well as how to document their appearance. Laboratory staff should note in the inspection whether the external condition of the box suggests any physical distress. When the box is opened, conditions inside it should be documented (e.g., whether there is dry ice left, or if wet ice is used then whether items are still cold to the touch). Every vial should be examined to determine whether it is intact, and the general appearance of the contents should be noted (e.g., frozen, partially thawed, liquid, or frothy liquid). A visual inspection checklist can make that process as efficient as possible. Of course, data loggers should be downloaded, examined for excursions, and documented. The more information you capture about the integrity of analytical test sample shipments, the better you can track down and positively identify whether there is a problem.
Montgomery: And you also have to be aware of national holidays when personnel are all off work.
Ritter: Yes, that can definitely happen too. It would be terrible if you planned for two-day delivery when it turns to be a three-day weekend. To minimize scheduling surprises, some companies have policies about which days of the week they will send or receive test sample shipments. Larger companies often have a department of shipping specialists — which is smart when you’ve got millions of dollars of product moving around. Sometimes those departments also handle test-sample shipments, but not always. Test samples themselves aren’t worth millions of dollars, but they can make a million-dollars–worth of product look bad. Those samples are the “quality detectors”; they represent an entire batch. And if they have a problem, you want to be able to quickly determine whether the batch has a problem or the problem is with the test samples alone.
Montgomery: Is there anything else you’d like to add?
Ritter: Bottom line: Define and document. Know what analytical and microbial tests are being performed on which test materials, by whom, and where. Define how release and stability test samples will be prepared and distributed to all designated testing laboratories. Document the specific procedures in written SOPs or instructions of work, then document that these instructions were followed correctly for each shipment. Don’t leave anything to chance; it’s too important to make up procedures as you go along. Yet I see evidence of that all the time.
Maribel Rios is managing editor, and S. Anne Montgomery is editor in chief and cofounder of BioProcess International. Nadine Ritter, PhD, is president and senior analytical advisor at Global Biotech Experts, LLC and a founding member of BPI’s editorial advisory board.