Environmental and occupational disorders. Endotoxin content of standardized allergen vaccines

Transcription

1 Endotoxin content of standardized allergen vaccines Bhavini Trivedi, MD, a Cherry Valerio, BS, b and Jay E. Slater, MD b Bethesda, Md Background: Endotoxin is a ubiquitous and potent proinflammatory agent. Previous limited studies suggest that it is present in allergen vaccines and that this could affect the safety and efficacy of allergen immunotherapy. The endotoxin content of standardized allergen vaccines is unknown. Objective: The purpose of this study was to quantify the amount of endotoxin contained in standardized allergen vaccines. Methods: The endotoxin content of 14 allergen vaccines was measured by using the Limulus amebocyte lysate (LAL) gelclot assay. To account for (1,3)-β-D-glucan and protease interference, vaccines were selectively depleted of endotoxin and then retested with the gel-clot assay. Proteases were also heatinactivated in selected vaccines. Fifty-eight lots of vaccines were tested, including at least two manufacturers per vaccine. Results: The endotoxin content of the 58 vaccines ranged from undetectable to 34,000 EU/mL. Cat pelt (12,735 EU/mL; range, 5177 to 33,805) had significantly more endotoxin activity than cat hair (2883 EU/mL; range, 1 to 16,962), and Dermatophagoides farinae extracts (4619 EU/mL; range, 849 to 8485) had more than Dermatophagoides pteronyssinus (11 EU/mL; range, 1 to 34). Grass (160 EU/mL; range, 3 to 1561) and ragweed pollen (341 EU/mL; range, 8 to 1697) vaccines contained less endotoxin. (1,3)-β-D-glucan interference was significant (>10%) only in three ragweed vaccines and two grass vaccines. Heat inactivation had no effect. There were considerable differences in endotoxin content of the same vaccines made by different manufacturers. Conclusions: The endotoxin content of standardized allergen vaccines is extremely variable. Interference by proteases and (1,3)-β-D-glucans is minimal. The effects of the high levels of endotoxin in some vaccines on the immunomodulatory changes associated with allergen immunotherapy require further study. (J Allergy Clin Immunol 2003;111: ) From the a National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, Md, and the b Center for Biologics Evaluation and Research, US Food and Drug Administration, Bethesda, Md. Supported by the Center for Biologics Evaluation and Research, US Food and Drug Administration. The views expressed in this article are the opinions of the authors and are not the official opinion of the US Food and Drug Administration, the National Institutes of Health, or the Department of Health and Human Services. Received for publication November 7, 2002; revised December 22, 2002; accepted for publication December 23, Reprint requests: Jay E. Slater, MD, Laboratory of Immunobiochemistry (HFM-422), Center for Biologics Evaluation and Research, US Food and Drug Administration, 1401 Rockville Pike, Rockville, MD Mosby, Inc. All rights reserved /2003 $ doi: /mai Key words: Endotoxin, lipopolysaccharide, allergen, allergen vaccine, allergen immunotherapy, Limulus amebocyte lysate, endotoxin-neutralizing protein, standardization, protease Endotoxins pyrogenic lipopolysaccharides (LPS) from the cell walls of Gram-negative bacteria are potent proinflammatory compounds that have been shown to cause both acute and chronic illnesses. Injection of endotoxin can cause acute chills, fever, organ failure, and death. 1,2 Thus, most parenteral drugs and biological materials are required to undergo rigorous testing to ensure that they contain no more than a specified quantity of endotoxin, based on dosing instructions. 3 This testing was originally performed with the use of the rabbit pyrogen test, which was subsequently replaced by more sensitive and reliable tests based on the endotoxin-induced clot formed by a lysate prepared from amebocytes found in the hemolymph of the American horseshoe crab. 4 Certain biological materials, among them allergen vaccines, are not required to undergo evaluation for the presence of pyrogens (21CFR [b]). Thus, the endotoxin content of allergen vaccines is largely unknown. In one previous study, house dust (not house dust mite) vaccines had endotoxin levels of 0.45 to 500 µg/ml (4.5 to 5000 endotoxin units [EU]/mL), as measured by Limulus amebocyte lysate (LAL). 5 In another study, house dust vaccines contained 0.01 to 0.1 µg/ml (0.10 to 1 EU/mL) of endotoxin, whereas pollen, dog, mold, and cockroach vaccines did not have detectable endotoxin levels as assessed by LAL assay. 6 Although both of these studies concluded that endotoxin content had no apparent effect on the allergenic activity of allergen vaccines (as determined by skin test size or basophil histamine release), other studies suggest that endotoxin may act as an adjuvant that potentiates IgE responses to allergens. 7,8 In these studies, mice given endotoxin along with allergen had enhanced specific IgE and IgG1 responses to the allergen. Although these limited data suggest that endotoxin has no effect on the efficacy of allergen vaccines as diagnostic agents, endotoxin may affect both the safety and efficacy of these products as immunotherapeutic reagents. The specific effects of endotoxin appear to be dose- and time-dependent: In animal studies, low doses given before allergen sensitization appear to enhance T H 1 responses, whereas higher doses administered concurrently with allergen exposure tend to produce toxic and inflammatory responses. 9 The purpose of this study was to quantify the amount of endotoxin contained in the standardized allergen vaccines. 777

2 778 Trivedi, Valerio, and Slater J ALLERGY CLIN IMMUNOL APRIL 2003 FIG 1. Coagulation pathways in horseshoe crabs. Abbreviations used ENP: Endotoxin-neutralizing protein EU: Endotoxin unit LAL: Limulus amebocyte lysate LPS: Lipopolysaccharide To quantify the endotoxin content of complex biological materials such as allergen vaccines, the contributions of at least two nonendotoxin components to LAL clotting must be determined. The first is (1,3)-β-D-glucans common cell wall components present in pollens and mold spores which activate the LAL clotting cascade by an alternative pathway (Fig 1). We addressed this possible confounding factor by measuring the formation of the gel clot before and after the selective adsorption of endotoxin and not (1,3)-β-D-glucan with a highly specific endotoxin-neutralizing protein (ENP). 13 Proteases, which have been measured in several allergen vaccines, can also activate the LAL gel-clot cascade. 18 In addition to selective adsorption with ENP, we controlled for this variable by heat-inactivating endogenous proteases of selected vaccines before assay. METHODS Reagents Pyrotell LAL and END-X B15 endotoxin removal affinity resin were obtained from Associates of Cape Cod (Falmouth, Mass). Endotoxin standard (10,000 EU per vial) was purchased from United States Pharmacopeia (Rockville, Md), reconstituted in pyrogenfree water (Charles River Laboratories, Wilmington, Mass), and stored in aliquots at 20 C. Each aliquot was thawed only once and was used for a maximum of 2 days once thawed. Allergen vaccines Fifty-eight lots of 14 standardized allergen vaccines were tested. This included vaccines derived from 8 grass pollen species, Dermatophagoides farinae, Dermatophagoides pteronyssinus, short ragweed, ragweed mix, cat hair, and cat pelt. Thirty-eight lots were prepared in glycerin (50%, vol/vol) alone, 4 in aqueous solution with phenol (0.4%, wt/vol), and 14 in both glycerin and phenol. The two remaining extracts were obtained lyophilized and were reconstituted in pyrogen-free water before testing. For each allergen vaccine type, extracts from at least two different manufacturers were tested. Each of the allergen manufacturers that market products in the United States was represented in this study. LAL assay The LAL gel-clot assay was performed as previously described. 19 Allergen vaccine samples were tested first in a series of 10-fold dilutions, then with a series of 2-fold dilutions to narrow results. All samples, controls, and standard were tested in triplicate. The sensitivity of the Pyrotell lysate used was 0.06 EU/mL. To calculate the endotoxin content of the sample, the reciprocal end point of dilution was multiplied by the lysate sensitivity of 0.06 EU/mL to give the measure (in EU/mL) for each vaccine. The range reported (EU/mL) is for the last positive and the first negative result for each vaccine. Polystyrene, pyrogen-free, sterile, disposable tubes were used for making all dilutions for each assay. Test tubes used for gel-clot incubations were made of borosilicate glass, depyrogenated by baking at 180 C for 3 hours. Only sterile, pyrogen-free disposable microtiter plates, pipettes, and pipette tips were used throughout. In preliminary experiments, the diluents used for allergen vaccines had no detectable endotoxin activity. Endotoxin depletion To assess the contribution to the gel-clot assay of nonendotoxin components in test materials, the allergen vaccines, diluted 1:100, were depleted of endotoxin by using END-X endotoxin affinity resin, 20 which contains ENP bound to a silica solid phase. 13 After overnight adsorption, supernatants were removed and tested with the LAL gel-clot assay in parallel with unadsorbed samples. Residual gel-clot activity after END-X adsorption is assumed to be due to (1,3)-β-D-glucans. Some vaccines with a geometric mean total gelclot activity <19 EU/mL were not adsorbed and were retested. The mean corrected endotoxin content was the difference between the geometric mean gel-clot activity before and after adsorption. Postadsorption samples that were <6 EU/mL or that were not tested were assumed to have a postadsorption mean gel-clot activity of zero. Heat inactivation of proteases Selected allergen vaccines were heated to inactivate endogenous proteases before LAL gel-clot assay. Heat treatment was performed in a Lab-Line Multi-Block Heater (model 2004) at 95 C for 15 minutes. Samples were cooled to room temperature before assay. Statistical analyses The statistical significance of the differences between the corrected endotoxin activities of the different classes of products was determined by using a 1-tailed Mann-Whitney rank sum analysis on

3 J ALLERGY CLIN IMMUNOL VOLUME 111, NUMBER 4 Trivedi, Valerio, and Slater 779 FIG 2. Percentage of total gel-clot activity attributed to β-glucans for each of 6 classes of allergenic extracts. Only the 47 vaccines tested before and after endotoxin adsorption are represented. Error bars represent ranges. TABLE I. Evaluation of untreated and treated extracts for endotoxin content Extract Manufacturer Gel-clot activity Gel-clot activity (after heat treatment) Ryegrass G 1,200-2,400 1,200-2,400 Short ragweed J 1,200-2,400 1,200-2,400 Cat pelt A 24,000-48,000 24,000-48,000 Cat pelt G 4,800-9,600 4,800-9,600 D farinae D 6,000-12,000 6,000-12,000 D farinae E 4,800-9,600 4,800-9,600 D pteronyssinus G D pteronyssinus J The possibility that proteases in pollen, mite, and cat allergen vaccines activate the LAL gel-clot cascade was evaluated by heat treatment, which would be expected to inactivate proteases. Selected extracts were heated at 95 C for 15 minutes, and the untreated and treated extracts were evaluated simultaneously for endotoxin content by LAL assay. Prism software program (version 3.00, GraphPad Software, Inc). This nonparametric method was used because the activity data are not normally distributed. RESULTS Depletion of endotoxin with the use of endotoxin-affinity resin The LAL gel clot forms in the presence of both endotoxin and (1,3)-β-D-glucans (Fig 1) and may be formed in the presence of proteases. To determine the contribution of these nonendotoxin components to gel-clot formation by allergen vaccines, each vaccine was selectively depleted of endotoxin with the use of an ENP affinity resin and then retested. In preliminary experiments, we were able to remove >99% of the gel-clot forming activity by using the resin beads (not shown). Residual LAL gel-clot activity after endotoxin adsorption indicating the presence of (1,3)-β-D-glucans or proteases was significant in 3 of the ragweed pollen vaccines (32% to 52%), 1 ryegrass pollen vaccine (52%), and 1 timothy pollen vaccine (45%); of note is that the timothy grass vaccine endotoxin level was only 10 EU/mL. Of the remaining 53 vaccines, 42 were absorbed and retested, and 90% to 100% of significant LAL gelclot activity could be attributed to endotoxin and not to (1,3)-β-D-glucans or proteases (Fig 2). Heat inactivation of proteases Selected allergen vaccines were heat-treated to inactivate endogenous protease enzyme activity. The LAL gelclot assay was performed before and after heat treatment. No differences in gel-clot activity were observed in any of the vaccines treated (Table I). Endotoxin content of allergen vaccines The mean corrected endotoxin content of the 58 vaccines tested in this study varied from undetectable to >30,000 EU/mL (Table II and Fig 3). The grass pollen

4 780 Trivedi, Valerio, and Slater J ALLERGY CLIN IMMUNOL APRIL 2003 FIG 3. Corrected endotoxin activity for each of 6 classes of allergenic extracts. All 58 allergen vaccines are included. Error bars represent ranges. (mean content, range = 160 EU/mL, 3 to 1561) and ragweed pollen (341 EU/mL, 8 to 1697) allergen vaccines had a lower apparent endotoxin content than the cat pelt (12,735 EU/mL, 5177 to 33,805), cat hair (2883 EU/mL, 1 to 16,962), and D farinae (4619 EU/mL, 849 to 8485) vaccines. Substantial differences existed in the same allergen vaccine made by different manufacturers. The greatest variability in endotoxin content was found in the cat and dust mite vaccines. All four cat pelt vaccines that we selected had high levels of endotoxin, whereas the cat hair vaccines contained significantly less (P =.02). In the dust mite vaccines there were large differences between the two species: All seven D farinae vaccines had relatively high endotoxin concentrations, whereas all seven D pteronyssinus vaccines contained very little endotoxin (11 EU/mL, 1 to 34) (P =.0003). DISCUSSION In this study, we measured the endotoxin contents of multiple lots of standardized allergen vaccine by LAL gel-clot assay and selective endotoxin adsorption. The lysate for the LAL gel-clot assay which has been used to measure endotoxin concentration since is derived from the amebocytes of the horseshoe crab, Limulus polyphemus. Endotoxin activates a coagulation cascade in the lysate, resulting in clot formation. 22 Although this assay is very sensitive for endotoxin, the enzyme pathway can be affected by other extract components. (1,3)-β-D-glucans common cell wall components present in pollens and mold spores also activate the coagulation cascade, although through a different pathway (Fig 1) In addition, proteases in tested materials can activate the gel-clot cascade, 18 and some allergen vaccines are known to have protease activity To examine possible interference from these sources, we depleted the vaccines of endotoxin and then repeated the LAL assay to assess for the amount of nonendotoxin components contributing to the gel-clot activity of the vaccines. The endotoxin affinity resin that we used consists of silica beads covalently coupled with ENP, an 11.8-kD protein from the amebocytes of the horseshoe crab that binds endotoxin specifically. 13 Therefore, the clotting of samples that have been adsorbed with the resin reflects the protease or (1,3)-β-D-glucan content of the vaccine. Since the gelclot activity in representative allergen vaccines appears to be highly heat-stable (Table I), the nonendotoxin gel-clot activity that we observed (Table II) is more likely to be due to (1,3)-β-D-glucans than to proteases. The endotoxin concentration of different allergen vaccines is highly variable. This variability exists among allergen vaccines derived from different source materials as well as among the allergen vaccines nominally manufactured from the same or similar source materials but produced by different manufacturers. The highest concentrations were found in the dust mite and cat vaccines. However, even in these groups there were some vaccines that contained little to no endotoxin.

5 J ALLERGY CLIN IMMUNOL VOLUME 111, NUMBER 4 Trivedi, Valerio, and Slater 781 TABLE II. Evaluation of endotoxin content of standardized allergen vaccines as measured by LAL gel-clot assay before and after specific depletion of endotoxin Total gel-clot activity Depleted extract gel-clot Mean corrected endotoxin Extract Manufacturer (EU/mL) activity (EU/mL) content (EU/mL) Bermuda A 6-12 <6 8 Bermuda E NT 17 Bermuda F 6-12 <6 8 June grass A <6 17 June grass B June grass J <6 170 Meadow fescue A 6-12 <6 8 Meadow fescue C <6 17 Meadow fescue D <6 34 Meadow fescue F <6 17 Orchard A <6 85 Orchard C <6 17 Redtop E <6 34 Redtop G <6 34 Ryegrass A <6 170 Ryegrass C 600-1, Ryegrass E <6 679 Ryegrass F NT 17 Ryegrass G 1,200-2, ,561 Sweet vernal A <6 3 Sweet vernal D <6 26 Sweet vernal F <6 339 Timothy A Timothy D Ragweed mix A Ragweed mix B Short ragweed A 6-60 NT 8 Short ragweed B Short ragweed C 600-1, Short ragweed D Short ragweed E Short ragweed F 6-60 NT 19 Short ragweed J 1,200-2,400 <6 1,697 Cat pelt A 4,800-6, ,177 Cat pelt A 24,000-48, ,805 Cat pelt G 4,800-6, ,177 Cat pelt G 4,800-9, ,780 Cat hair B <6 339 Cat hair C NT 1 Cat hair E 12,000-24,000 <6 16,962 Cat hair F 600-1,200 <6 849 Cat hair H Cat hair I NT 1 Cat hair J 1,200-2, ,689 D farinae A 6,000-12, ,296 D farinae B 1,200-2,400 <6 1,697 D farinae D 6,000-12,000 <6 8,485 D farinae E 4,800-9,600 <6 6,788 D farinae G 4,800-6, ,367 D farinae I 600-1,200 <6 849 D farinae J 600-1,200 <6 849 D pteronyssinus A 6-12 <6 8 D pteronyssinus B NT 2 D pteronyssinus C 0 NT 1 D pteronyssinus D NT 2 D pteronyssinus G NT 19 D pteronyssinus I 6-12 <6 8 D pteronyssinus J NT 34 The endotoxin content of standardized allergen vaccines was measured by LAL gel-clot assay before and after specific depletion of endotoxin. The gel-clot dilutions after depletion reflect the activity of (1,3)-β-D-glucans or proteases. Corrected endotoxin activity is calculated from the geometric mean activities of the total and depleted samples. NT, Not tested.

6 782 Trivedi, Valerio, and Slater J ALLERGY CLIN IMMUNOL APRIL 2003 Perhaps the most striking observation of this study is the difference in apparent and corrected endotoxin activities between allergen vaccines made from the two species of house dust mite. This difference is statistically significant and large and was consistent both within and among manufacturers. It cannot be explained by differences in growth media or extraction conditions, which are identical for the two species. Intrinsic differences between the mite species are possible. For example, LAL gel-clot activity may be associated with a nonendotoxin component that binds to ENP and that is present in only one of the mite species. Alternatively, D pteronyssinus may selectively express an endogenous endotoxin-neutralizing activity, which interferes with the gel-clot reaction. There are no data to support one possibility over the others. In preliminary experiments (data not shown), we demonstrated that one of the D farinae vaccines (from manufacturer A) was positive in the rabbit pyrogen test (USP 23) at a dilution of 1:50, confirming that the observed LAL measurement is associated with true endotoxin-like biological activity. In a large European study of the association of indoor allergen levels and asthma, exposure to Der f 1 but not Der p 1 was associated with asthma symptoms. 23 Though not considered by the authors, this could have been due to the presence of intrinsic endotoxin-like activity associated with D farinae but not D pteronyssinus. At this point, the causes and significance of this observation are uncertain and warrant further investigation. Wide ranges in endotoxin concentration also exist among the grass and ragweed vaccines. Interestingly, much of the LAL reactivity of three of the ragweed vaccines tested (ragweed mixes from manufacturers A and B and short ragweed from manufacturer B) as well as one ryegrass pollen vaccine (manufacturer C) and one timothy pollen vaccine (manufacturer D) appeared to be due to (1,3)-β-D-glucans, which is present in plant and fungal cell walls. In each of these cases but the single ryegrass pollen extract, the total gel-clot activity was relatively low (<120 EU/mL). Thus, (1,3)-β-D-glucans do not appear to interfere significantly with the measurement of endotoxins in allergenic extracts by standard techniques. When the LAL gel-clot assay was used, all but one of the vaccines tested contained detectable endotoxin. The physiologic significance of the endotoxin in these vaccines is uncertain. On the one hand, the presence of endotoxin in allergen vaccines may enhance their immunotherapeutic efficacy by deviating the immune response to the allergen from a T H 2- to a T H 1-type response. 9,24 On the other hand, some studies suggest that endotoxin may enhance T H 2 responses, 7,8,25 and endotoxins may increase the frequency and severity of adverse reactions to allergen immunotherapy by causing the release of IL-1 and TNF-α In addition, several studies done have shown that endotoxin in house dust worsens asthma symptoms, 29,30 and patients with asthma are at the highest risk for systemic reactions to immunotherapy. 31,32 In asthmatic persons in particular, the endotoxin content in allergen vaccines may be a serious concern. Several investigators have evaluated the dose-response relation of endotoxin on the inflammatory response in human beings. Many of these studies examined the dose responses with preparations labeled in mass units rather than EU. EUs are determined by comparison of the test sample to the USP reference standard using an LAL-based assay, and the specific activity of endotoxin preparations (in EU/ng) is variable (in USP 23, the range is given as 2 to 50 EU/ng 33 ; other sources suggest 12 EU/ng 3 and 5 EU/ng 4 ). Hence, regulatory controls on the endotoxin content of drugs, biological materials, and devices are based on their EU content, not mass units. To estimate the clinical significance of the endotoxin content of allergen vaccines, we will use a specific activity of 10 EU/ng. In one study, an inhaled dose of 5 µg (50,000 EU) of LPS was followed by a significant increase in sputum neutrophils, and a 50-µg dose (500,000 EU) produced a significant increase in body temperature. 28 Thus, the minimum airway dose necessary to elicit a physiologic response appears to be well above the doses of endotoxin reached in allergen vaccines. However, the threshold injected doses appear to be considerably lower. An intravenous dose of 4 or 5 ng/kg (40 to 50 EU/kg, or 2800 to 3500 EU for a 70-kg man) elicits a fall in serum iron levels 1 and a rise in temperature, heart rate, and white blood cell count. 1,34 Specific endotoxin limits in drugs and biological materials vary but are based on the maximum human dose (in units or mass units per hour) of the product and a limit of 5.0 EU/kg (0.2 EU/kg for intrathecal products). 3 In our study, the endotoxin content of the allergen vaccines tested was as high as 34,000 EU/mL, and the mean corrected endotoxin content was approximately 1900 EU/mL. Thus, a mix of the vaccines tested here would deliver a dose of 950 EU per 0.5-mL injection at maintenance, and a mix of mite and cat allergen might deliver considerably more. These doses exceed current limits for other products. The dose-response data suggest that these injected doses might result in both local and systemic effects, and immunomodulatory effects might be expected at even lower doses. In conclusion, we have shown that the endotoxin content of allergen vaccines is highly variable and, overall, that the (1,3)-β-D-glucan contribution to measured endotoxin levels is small. We also found that the LAL gel-clot method appears to be the most reliable method for screening allergen vaccines for endotoxin when combined with ENP adsorption to assess (1,3)-β-D-glucan interference. Finally, we found no evidence that protease activity contributed to the high LAL gel-clot activity noted in any of the D farinae and cat pelt vaccines. The clinical consequences of endotoxin in allergen vaccines have not been evaluated. There are no data that indicate that adverse events after allergen immunotherapy are due to the endotoxin content of allergen vaccines, nor are there data to support the possibility that the endotoxin in allergen vaccines enhances the beneficial effects of immunotherapy. However, previous studies of the effects of LPS in human beings suggest that this should

7 J ALLERGY CLIN IMMUNOL VOLUME 111, NUMBER 4 Trivedi, Valerio, and Slater 783 now be examined. Future studies of the efficacy of allergen immunotherapy with standardized allergen vaccines should be controlled for the endotoxin content of the allergens used, and investigators reporting anaphylaxis associated with allergen immunotherapy should consider measuring the endotoxin content of the implicated allergen vaccines. The authors thank Christine Anderson for assistance in the design and implementation of this study. REFERENCES 1. Wolff SM. Biological effects of bacterial endotoxins in man. J Infect Dis 1973;128: Miller MA, Hjelle JT. Endotoxins and their clinical significance. In: Prior RB, editor. Clinical applications of the Limulus amoebocyte lysate test. Boston: CRC Press; p Guideline on the validation of the Limulus Amebocyte Lysate Test as an End-Product Endotoxin Test for Human and Animal Parenteral Drugs, Biological Products and Medical Devices Available at: Accessed March 13, Prior RB. The Limulus amoebocyte lysate test. In: Prior RB, editor. Clinical applications of the Limulus amoebocyte lysate test. Boston: CRC Press; p Siraganian RP, Baer H, Hochstein HD, May JC. Allergenic and biologic activity of commercial preparations of house dust extract. J Allergy Clin Immunol 1979;64: Smith TF, Morrison DC. Lack of interference in skin tests by endotoxin in allergen extracts. Ann Allergy 1986;57; Dannenman PJ, Michael JG. Adjuvant and immunogenic properties of bacterial lipopolysaccharide in IgE and IgG 1 antibody formation in mice. Cell Immunol 1976;22: Slater JE, Paupore EJ, Elwell MR, Truscott W. Lipopolysaccharide augments IgG and IgE responses of mice to the latex allergen Hev b 5. J Allergy Clin Immunol 1998;102: Liu AH. Endotoxin exposure in allergy and asthma: reconciling a paradox. J Allergy Clin Immunol 2002;109: Cooper JF, Weary ME, Jordan FT. The impact of non-endotoxin LALreactive materials on Limulus amebocyte lysate analyses. PDA J Pharm Sci Technol 1997;51: Roslansky PF, Novitsky TJ. Sensitivity of Limulus amebocyte lysate (LAL) to LAL-reactive glucans. J Clin Microbiol 1991;29: Zhang GH, Baek L, Buchardt O, Koch C. Differential blocking of coagulation-activating pathways of Limulus amebocyte lysate. J Clin Microbiol 1994;32: Warren HS, Glennon ML, Wainwright N, Amato SF, Black KM, Kirsch SJ, et al. Binding and neutralization of endotoxin by Limulus antilipopolysaccharide factor. Infect Immunol 1992;60: Iraneta SG, Duschak VG, Rodriguez SM, Seoane MA, Albonico JF, Alonso A. Proteinase and gelatinolytic activities of house dust mite and cockroach extracts. J Investig Allergol Clin Immunol 1999;9: Hewitt CR, Foster S, Phillips C, Horton H, Jones RM, Brown AP, et al. Mite allergens: significance of enzymatic activity. Allergy 1998;53(Suppl 48): Bagarozzi DA Jr, Travis J. Ragweed pollen proteolytic enzymes: possible roles in allergies and asthma. Phytochemistry 1998;47: Esch RE. Role of proteases on the stability of allergenic extracts. Arb Paul Ehrlich Inst Bundesamt Sera Impfstoffe Frankf A M 1992;85: Cooper JF. Evaluation of sterile pharmaceuticals and medical devices: an overview. In: Prior RB, editor. Clinical applications of the Limulus amoebocyte lysate test. Boston: CRC Press; p Adner N, Flink O, Nystrand R, Pihl B, Vegis P. Collaborative study of a control standard endotoxin using the Limulus amebocyte (gel-clot) test. J Parenter Sci Technol 1991;45: Garcia C, Saladino R, Thompson C, Hammer B, Parsonnet J, Wainwright N, et al. Effect of a recombinant endotoxin-neutralizing protein on endotoxin shock in rabbits. Crit Care Med 1994;22: Hochstein HD. Role of the FDA in regulating the Limulus amoebocyte lysate test. In: Prior RB, editor. Clinical applications of the Limulus amoebocyte lysate test. Boston: CRC Press; p Tai JY, Seid RC Jr, Huhn RD, Liu TY. Studies on Limulus amoebocyte lysate, II: purification of the coagulogen and the mechanism of clotting. J Biol Chem 1977;252: Gehring U, Heinrich J, Jacob B, Richter K, Fahlbusch B, Schlenvoigt G, et al. Respiratory symptoms in relation to indoor exposure to mite and cat allergens and endotoxins: Indoor Factors and Genetics in Asthma (INGA) Study Group. Eur Respir J 2001;18: Ebner C. Immunological mechanisms operative in allergen-specific immunotherapy. Int Arch Allergy Immunol 1999;119: Zimmer S, Pollard V, Marshall GD, Garofalo RP, Traber D, Prough D, et al. Effects of endotoxin on the T H 1/T H 2 response in humans. J Burn Care Rehabil 1996;17: Reed CE, Milton DK. Endotoxin-stimulated innate immunity: a contributing factor for asthma. J Allergy Clin Immunol 2001;108: Michel O. Systemic and local airways inflammatory response to endotoxin. Toxicology 2000;152: Michel O, Nagy A-M, Schroeven M, Duchateau J, Neve J, Fondu P, et al. Dose-response relationship to inhaled endotoxin in normal subjects. Am J Respir Crit Care Med 1997;156: Michel O, Kips J, Duchateau J, Vertongen F, Robert L, Collet H, et al. Severity of asthma is related to endotoxin in house dust. Am J Respir Crit Care Med 1996;154: Park JH, Gold DR, Spiegelman DL, Burge HA, Milton DK. House dust endotoxin and wheeze in the first year of life. Am J Respir Crit Care Med 2001;163: Lockey RF, Nicoara-Kasti GL, Theodoropoulos DS, Bukantz SC. Systemic reactions and fatalities associated with allergen immunotherapy. Ann Allergy Asthma Immunol 2001;87: Ragusa FV, Passalacqua G, Gambardella R, Campanari S, Barbieri MM, Scordamaglia A, et al. Nonfatal systemic reactions to subcutaneous immunotherapy: a 10-year experience. J Investig Allergol Clin Immunol 1997;7: Bacterial endotoxins test. United States Pharmacopeia. Rockville, Md: United States Pharmacopeial Convention, Inc, p Michie HR, Manogue KR, Spriggs DR, Revhaug A, O Dwyer S, Dinarello CA, et al. Detection of circulating tumor necrosis factor after endotoxin administration. N Engl J Med 1988;318: