This study compared weight per volume versus PNU concentrations for birch, ragweed, dog and Alternaria, which are all non-standardized extracts in Canada. Our results show substantial variability when comparing weight per volume and PNU concentrations for both Hollister-Stier and Stallergenes Greer products. The largest variability was observed for Hollister-Stier’s ragweed extract and Stallergenes Greer’s Alternaria extract, while both companies had the smallest variability for their dog extract. The significance of our study is to not only highlight the variable potencies that exist within a sample of non-standardized extracts, but also how these concentrations compare to actual allergen immunotherapy dosing recommendations in Canada and the United States.
There are currently only 19 standardized allergen extracts available in Canada, as most commercial extracts are non-standardized, including birch, ragweed, dog and Alternaria. The extraction process for both standardized and non-standardized products is essentially the same, with quality control measures being the primary difference . Non-standardized extracts are labeled on the basis PNU values, or the weight of the source material extracted with a given volume of extracting fluid (weight per volume). These approaches to labeling concentrations have no established standards for biologic potency, and there are no dose-response studies with non-standardized extracts .
The advantage of standardized extracts over their non-standardized counterparts is the consistency of biological activity. However, recent studies have found that there are significant differences in the composition and content of specific allergen levels among standardized extracts, like house-dust mite (HDM) [10–13]. Since extraction processes differ among manufacturers, standardized extracts may contain different amounts of allergens. In a recent study by Nolte et al. , differences in the content of Der 1 & Der 2 – the major allergens of HDM – were observed despite equivalent concentration labelling. Mean Der 1 to Der 2 ratios of 20.5 and 5.2 were found for two batches of D. farinae from the same manufacturer labelled both as 10000 AU/mL. The mean Der 1 to Der 2-ratio ranged from 0.4 to 20.5 among various manufacturers they examined. A study by Jung et al.  looking at pollen chemistry showed that pollen ranges from 2.5–61% of protein by dry mass. Another study by Roulston et al.  concluded that there was a negative correlation with percentage of protein in pollen grain to pollen grain volume and mass. Furthermore, Schappi et al.  demonstrated that the concentration of birch protein (bet v 1) only represented 0.07% of the total pollen grain mass of 7.85 ng. These collective studies highlight significant variability that exists in protein and allergen content compared to the labelled concentration of the allergen extract.
This lack of consistency in allergenic potency and composition of non-standardized extracts not only affects clinical efficacy, but has important implications for safety . The AAAAI suggests maintenance doses for non-standardized extracts at 3000 to 5000 PNU, or weight per volume 1:100 to 1:200 both at a volume of 0.5 mL . The Canadian Society of Allergy and Clinical Immunology (CSACI) has recommended using 5000 PNU at a volume of 0.5 mL as a recommended maintenance dose . The CSACI has suggested using PNU instead of weight per volume to try to simplify the process of prescribing subcutaneous immunotherapy.
From our study, the corresponding PNU concentration for a weight per volume of 1:100 to 1:200 can be compared to the recommended PNU dosing for non-standardized extracts. We believe that dosing towards a weight per volume of 1:200 prioritizes safety and minimizes the risk of severe reactions with subcutaneous immunotherapy. Birch extract has low PNU concentration (below < 5000) for a weight per volume of 1:200 for both HollisterStier and Stallergenes Greer (Table 1A). In contrast, for both HollisterStier and Stallergenes Greer’s ragweed extract, a weight per volume of 1:200 corresponds to a PNU concentration greater than 5000 (Table 2A). Dog extract for a weight per volume of 1:200, and even for 1:100, corresponds to very low PNUs for both companies (Table 3A). The major allergen content for dog extract is typically too low to allow for effective dosing , and this is likely because the target dose of 5000 PNU may be impossible to reach based on our calculations with both companies (Table 3A). Although not part of our study, it may be possible with acetone precipitated dog extracts to reach therapeutic dosing as lower weight per volume corresponds to a higher PNU . Finally, for Alternaria, corresponding PNU concentrations for HollisterStier is low at only 500 while over 5000 for Stallergenes Greer (Table 4A).
Overall, we suggest selecting a PNU dose that corresponds to a weight per volume of 1:200 as this underscores safety to the patient. Our recommendations for starting PNU dose for the four non-standardized extracts are highlighted in Table 5. If the starting PNU concentration is considerably below 5000 for a weight per volume of 1:200 or proves to be ineffective, such as in birch or Alternaria with HollisterStier, slow up-titration is advised. Conversely, for starting PNU concentrations above 5000 for weight per volume of 1:200, such as ragweed or Alternaria with Stallergenes Greer, we recommend a maximum starting PNU concentration of 5000. Based on our findings and the relative lack of randomized controlled trial data, the efficacy and safety of dog immunotherapy is questionable. We do not recommend dog immunotherapy at this time until further data is available. However, if prescribed, we suggest starting PNU concentrations of 25 for HollisterStier and 1200 for Stallergenes Greer with slow up-titration as needed. Table 6 highlights recommended weight per volume dosing when converting from PNU concentration of 3000. Individual calculations may need to be undertaken for the various non-standardized extracts, as concentrations can vary between extract batches from within the same allergen manufacturer, and certainly between manufacturers. As such, multiple allergen immunotherapy is preferred to be ordered from the same manufacturer.
The efficacy of immunotherapy depends on achieving an optimal therapeutic dose of the allergen extract . Unlike non-standardized extracts, standardized extracts have been extensively studied  and doses used in controlled clinical trials form the basis of the recommended dose ranges . For non-standardized extracts, the therapeutically effective doses must be estimated and individualized . Allergen concentrations that are too low are less likely to be effective, while those that are too high may result in systemic reactions. The variability in biological potency that is present in not only non-standardized extracts, but also standardized extracts, can potentially affect the outcomes of clinical trials trying to prove the effectiveness of allergen immunotherapy. A literature review of trials using non-standardized extracts, including those of birch, dog dander and Alternaria, revealed that extracts are not standardized between studies. Although many clinical trials attempted to individually standardize extracts within their own study, the method of standardization was variable between studies . Various units were used including specific unit (SU) , biological unit (BU) , Radioallergosorbent test (RAST)  units, and weight of the extract , with most units being arbitrarily developed. The maintenance dose also varied greatly between studies, up to more than a 10-fold difference at times [22, 25]. The usefulness of these collective studies is certainly limited by the different biological units used and the variability in dosing, making it difficult to extrapolate to other clinical situations. As we do not have reliable efficacy data for non-standardized immunotherapy extracts, we believe that these extracts should be prescribed in a safer manner by aiming for a 1:200 weight per volume dose. If this is not therapeutically effective, up-titration should be pursued.
Limitations in this study include the small sample of data, the limited number of allergens assessed, and only data from 2 allergen manufacturers were analyzed. However, we believe these trends would be similar among other non-standardized extracts and for other manufacturers. Also, major allergen levels were not available for the allergens assessed in this study.
The data in our study highlights the substantial variability that exists in extract quantification for four non-standardized extracts. More importantly, we observed that that allergen potencies as currently manufactured may not meet the immunotherapy dosing recommendations. From our results, we have demonstrated that the conversion to recommended doses between weight per volume and PNU is variable between lots and between companies for all of the non-standardized allergens studied. Doses of allergen extract should be therapeutically effective, while minimizing the risk of harm. Future research will be necessary to examine larger batches of extracts, and from more manufacturers. Most importantly, randomized controlled trials should be performed to identify safe and clinically effective doses for allergen immunotherapy.