Challenges in Allergen Testing - Spiking and Recoveries

When I started developing immunoassays for the detection of allergens in food, the first thing that struck me was the wide range of different food types or matrices that the assays had to work with. Coming from a medical immunoassay background, there was a limited number of different matrices to work with. In my case, this was blood serum. With food there is an almost infinite range of different sample types, each with their own specific properties.

How do I choose the right test kit?

So how do we ensure that the test kit produced is suitable for use with such a diverse and challenging range of samples? This is where sample validation comes in. The process involves adding a known amount of an allergen of interest to our matrix (spike) and then trying to get that allergen back out again (recovery).
An important thing to remember is that, as the name implies, immunoassays use biological components (antibodies) to achieve the detection of the allergenic proteins of interest. As with all biological systems, the kits are sensitive to extremes.
In the case of foods, the kits may not work as they should in the presence of strong acid or alkali, high salt, high fat, etc. Many of these extremes can be countered during the extraction process. Kits therefore use a buffered system to cope with changes in pH and the addition of the buffer to the sample helps reduce and dilute some of the other problems such as salt and fat.

Is my recovery acceptable?

When it comes to the recovery of a known amount of allergen from a sample matrix, what is deemed acceptable? Before answering this, we need to define where we are starting from. Is it an incurred sample or a spiked one?

Incurred samples are defined as samples in which a known amount of the food allergen has been incorporated during processing, mimicking as closely as possible the actual conditions under which the sample matrix would normally be manufactured.

The subject of incurred samples will be discussed in more depth in a subsequent issue of Spot On. In this article, I will concentrate on outlining a more accessible method of spiking a known amount of allergen into a matrix as received from the supplier or manufacturer and measuring its recovery.    

With regard to recovery, the guidance states that:

“Ideal percent recovery levels would range from 80 to 120%. Recovery levels are affected by both the efficiency of the extraction step and the ELISA procedure.

“With ELISA methods for food allergens, this level of recovery is not always possible, particularly when certain difficult matrixes are analysed. In addition, the recovery from incurred samples can be substantially different from those obtained using spiked samples.

"For this reason, recoveries between 50 and 150% will be considered acceptable so long as they can be shown to be consistent.”

The guidelines were published in 2010 by the Association of Analytical Communities (AOAC) with particular reference to quantitative ELISA (Enzyme Linked Immunosorbent Assay) methods. Many of the key points are also applicable to qualitative or semi-quantitative LFD (Lateral Flow Device) methods.

The “science” behind spiking 

When we receive or encounter a new food type that has not been tested before, we will undertake spike recovery validation to ensure it works as it should with our test kits. We will spike in at three different levels of allergen – low, medium and high – to cover the range of detection of the assay.

The low allergen spike will be close to the Lower Limit of Quantitation, LLOQ, of the ELISA (in this case the lowest value calibrator above 0 ppm) or close to the Limit of Detection, LOD, of a lateral flow device. The medium spike will be in the middle of the ELISA calibration curve, and the high spike will be at or near the Upper Limit of Quantitation, ULOQ (the highest ppm value calibrator). The sample is extracted and tested in accordance with the product insert supplied with the kit.

So for example, if we spike 5 ppm of almond into chocolate, we would expect to see a recovery of 4 ppm to 6ppm. If the result is outside of this range, then there are steps that can be taken to help improve the recovery. From experience, chocolate is one of the most challenging food matrices to test – it is full of tannins and other polyphenols which can bind to any allergenic protein that may be present and form insoluble complexes which are difficult to extract.

Such difficulties can be overcome by adding extra protein to the extraction buffer. The excess protein binds to the polyphenols and makes the allergens available for extraction. My protein of choice is fish gelatine, although other material such as milk powder can be used to improve the extraction efficiency from high polyphenol containing foods. If using milk powder, be careful not to contaminate your laboratory space, especially if you are carrying out milk allergen testing.

Lateral Flow Devices, or strips or dipsticks as they are sometimes referred to, can be validated for spike recovery in a similar way to an allergen ELISA test kit. The thing to be aware of when choosing a high spike level is that although LFDs are capable of detecting very high ppm levels, you can actually overload the device by adding too much allergen. This can occur in amounts greater than 1% of the allergenic food.

Maintaining quality and test precision

It may be necessary for a kit manufacturer to work closely with customers who routinely test challenging food matrixes. It is important to verify that the kit is working as it should and to the customer’s satisfaction. This can be achieved, as detailed above, by undertaking allergen spike recovery experiments into the problematic matrix.

In some cases it may be desirable to modify or change the standard kit method to meet the demands of the sample and/or the customer; this should always be undertaken with the guidance of the kit manufacture to ensure the quality and reproducibility of the test kit.