Pharmaceutical Product Impurities: Considering Beta Glucans

• Bio Products Laboratory Ltd

Introduction

There are many pharmacological impurities which can appear, through unintended routes, in formulated final products. Although undesirable in terms of not meeting the strict requirements of a prepared product, the impact upon the drug formulation and potential impact upon the patient will vary according to the type and concentration of the impurity.

One impurity which can arise is beta glucan [1]. β-Glucans are soluble polysaccharides of glucose that can be produced by many prokaryotic and eukaryotic organisms. Although glucans can be classed as an impurity, this group of compounds is also manufactured for use in human and in veterinary medicine, although this is not the concern here.

As an impurity, there are different sources of glucan in pharmaceutical processing. The most common source derives from the widespread use of cellulosic based filters in pharmaceutical processing. This is because the heterogeneous molecules of beta glucan constitute a major carbohydrate fraction of plant cell walls. Some filter manufacturers use chemical washing techniques to remove the possibility of glucan fibers, of varying length, leaching from the filter face and some pharmaceutical manufacturers have introduced standardized filter washing methods. Other manufacturers have not made any changes.There is a risk here that a process variation, change of filter supplier, or a lack of control during the filter manufacturing process could see the elution of glucan fibers. In this scenario, unless the pharmaceutical manufacturer purposely looks for glucan using a glucan specific assay, or detects something anomalous using the LAL assay, the presence of cellulosic fibers may go undetected.

This article examines the phenomenon of beta glucans, describing what they are and how they can arise. The article also examines methods for the detection of glucans as well as the phenomenon of LAL test assay interference [2]. Like endotoxin, glucans are large polyscaccharides (homopolymers of glucose) and both bacterial endotoxins and (1→3)-ß-D-glucans are considered Pathogen-Associated Molecular Patterns, substances which elicit inflammatory responses in mammals. Therefore the LAL test reacts with two types of polysaccharide: glucan and LPS (endotoxin) [3]. This can be a both a problem (false positives) and a benefit (in terms of detecting glucan in relation to a process change).

What Are Beta Glucans?

The structure of a glucan is as glucose units held in the β anomeric form. β-Glucans are polysaccharides of D-glucose monomers linked by β-glycosidic bonds. The intramolecular hydrogen bonds between the separate glucan chains results in them become very strong fibres. There are many types of (1→3)-ß-D-glucans, and physical and chemical properties of the molecule affect its biological activities.

(1→3)-ß-D-glucans are known to bind to and activate macrophages, neutrophils, monocytes, and NK cells. They are believed to bind to a variety of cells, for example endothelial cells and fibroblasts [4]. Among the biological effects caused by glucans is the production of cytokines, which are important components in inflammation (glucans are therefore considered pro-inflammatory molecules and they are capable of immuno-modulatory activity). Other activities associated with glucans are nitric oxide synthesis, activation of the complement cascade, and activation of lymphocytes and macrophages. There is evidence from animal studies that glucan and endotoxin can act synergistically to increase the inflammation response. In relation to pharmaceuticals (1→3)-ß-D-glucans may be considered a contaminant [5].

Normal human serum contains low levels of (1→3)-ß-D-glucan, typically 10-40 pg/mL (this level relates to commensal yeasts present in the alimentary canal and gastrointestinal tract). Most pathogenic fungi have (1→3)-ß-D-glucan in their cell walls and minute quantities are sloughed into the circulation during the life cycle [6]. Thus, (1→3)-ß-Dglucan appears in the serum in cases of invasive fungal infection (IFI). It follows that monitoring serum glucanemia for evidence of elevated and rising levels provides a convenient surrogate marker for IFI. Levels above 80 pg/mL in at-risk patients are considered positive [7, 8]. The main risk cited is for patients undergoing long-term hemodialysis.

Sources of Glucan in Pharmaceutical Processing

Common sources of glucans in pharmaceutical manufacturing are filters made from cellulose materials, plant-derived raw materials, cotton-containing enclosures, sugars, naturally-derived raw materials, and cellulose products (such as sponges or filters) [9]. A second source is fungi (or yeast hydrolysate) [10].

The microbial glucan contaminants include zymosan (from yeast), laminarin (from algae), lentinan (from fungi) and curdlan (from bacteria). Although fungi can occur in uncontrolled pharmaceutical processing environments, the levels would need to be considerably high for fungi to be regarded as a major source of glucan in pharmaceutical products. Thus with efficient HEPA or ULPA filters and a robust cleaning program, the presence of fungal derived glucan is rare. The detection of fungal glucan is, as indicated above, of great importance for clinical diagnosis [11].

Of the different sources, cellulosic depth filters remain the primary source due to their ubiquity in pharmaceutical manufacturing [12]. Filters are devices (usually a membrane) designed to physically block certain objects or substances while letting others through, depending on their size. Several types of biological products are clarified using disposable depth filters. Depth filters are the variety of filters that use a porous filtration medium to retain particles throughout the medium (rather than just on the surface of the medium). These filters are commonly used when the fluid to be filtered contains a high load of particles. This is because they can retain a large mass of particles before becoming blocked, relative to other types of filters.

Such filters are commonly manufactured from a mixture of diatomaceous earth (containing metal ions such as aluminum, iron, copper and other metals) and cellulose, to provide the filtration media (this provides two dynamic mechanisms for the removal of contaminants, electrokinetic charge and size exclusion). Cellulose is used to construct the initial filter matrix to which filter aids and other additives are impregnated into, as well as to improve contaminant removal and loading capabilities of the filter sheets. The botanical tissue that is the source of the cellulose typically contains (1→3)-ß-Dglucan [13]. Pearson et al estimate that the molecular weight of filter extracted cellulosic glucan is around 24,000 Daltons [14].

Some filter manufacturers attempted to neutralize the glucan using processes such as contacting the acid-soluble portion of the glucan with an alkali solution under conditions sufficient to dissolve the alkalisoluble glucan. In other cases, pharmaceutical manufacturers will have undertaken rinsing studies using water-for-injections to demonstrably remove glucan prior to the pharmaceutical product being passed through the filter.

However, it stands that in many cases an examination of the filter has not been undertaken and thus during processing glucans can be leached from certain filter types and into the filtrate. Where a change to process conditions occurs, the problem is often only detected through routine testing performed using the LAL assay on the finished product.

Endotoxin testing

The LAL test is the compendia test for the examination of bacterial endotoxin in pharmaceutical products (as described in USP chapter <85> and European Pharmacopeia monograph 2.6.14). For large volume parenteral products the LAL test is normally a mandatory test for finished product release.

The presence of (1→3)-ß-D-glucans can give a falsely higher reading in the Limulus amebocyte lysate (LAL) assay for endotoxins (so termed ‘false positives’). The presence of glucans also has a tendency to cause test interference (that is when the test indicates the presence of endotoxin when none are in fact present). Therefore, interference from glucans could produce an out-of- Specification result [15]. Glucans, alongside some other substances, are commonly termed lAl Reactive Materials or alternatively lAl activators [16].

Glucans share a heat-resistant property with endotoxins, and both endotoxins and glucans will pass through sterilizing-grade filters. However, glucans give a negative reaction when tested using the rabbit pyrogen test, whereas endotoxins cause a positive reaction (depending upon the pyrogenic dose relative to the rabbit’s tolerance). This is because glucans are generally regarded as non-pyrogenic [17].

Interference with the LAL Assay

The reason for glucan interfering with the lAl test relates to the biochemical basis of the test itself. The reagent used in the lAl test is generally produced from the blood of a horseshoe crab (although some recombinant varieties of lysate are now available). The clotting mechanism of the blood of the horseshoe crab is designed to prevent the spread of bacterial contamination throughout the horseshoe crab’s biochemical system. When the endotoxin of Gram-negative bacteria connects with the horseshoe crabs amebocytes, a series of enzymatic reactions begin. The pathway alters amebocyte coagulogen into a fibrinogen-like clottable protein, which forms a coagulin gel. The defense mechanism is also effective against fungi; hence a similar reaction occurs in response to a fungal infection, which triggers the clotting cascade [18]. in the reactions, glucans trigger the protease enzyme Factor G, whereas endotoxin triggers the Factor C enzyme, although the end result – coagulin protein– is the same [19]. This is illustrated in Figure 1.

Figure 1. simplified diagram of the LAL assay clotting cascade

Some manufactures of lysate used in the lAl assay produce variants which will not detect glucan (that is, they are endotoxin-specific). This difference in lysates is considered later.

Studies have shown that the degree of lAl test interference increases with the amount of glucan present (one study showed that by increasing the amount of curdlan the degree of lAl test enhancement increased). However, this itself is unreliable and whilst the lAl assay in the unmodified form will detect endotoxin, more specific tests or variants of the lAl test are required to quantify glucan levels.

Techniques for the Detection of Glucan

There are three approaches to testing for glucan. The first approach involves attempting to destroy any endotoxin in a sample using base hydrolysis. This is through the addition of sodium hydroxide to a sample and holding the sample at 37oC for between twelve and fifteen hours. After adjusting the pH through the addition of hydrochloric acid, the lAl assay is run. The basis for the method is that by mixing the final product with sodium hydroxide, any endotoxin in the sample will be destroyed. if the subsequent lAl assay indicates a positive result, then glucans and not endotoxins are present.

The second approach is the subtraction method. This uses two assays for endotoxin, one of which has a blocking substance added to the lysate to reduce the response of the assay to glucan in the sample; and the other with an unmodified type of lysate. The blocking substance is a buffer, which is used to reconstitute lAl and render the reagent insensitive to (1→3)-β-d-glucan interference by effectively blocking the Factor G pathway of the endotoxin clotting cascade [20]. The difference between the two results is proportional to the amount of glucan present [21].

The third technique is the direct measurement of (1→3)-ß-d-glucans [22]. This is a quantitative, specific, and much more accurate approach than the subtraction method. one such method to detect glucan is a specially designed monoclonal antibody test against activated Factor C. The monoclonal antibodies are produced by a fusion with myeloma cells and mouse spleen cells immunized with lpS-Factor C complex. From this, monoclonal antibodies that react with the complex but not with lpS or Factor C are selected. When an appropriate amount of the monoclonal antibody is added to lysate, the lysate will respond to glucan but not to lpS [23].

Assessment

For routine testing, it is important for the quality control laboratory to establish whether it aims to use an endotoxin-specific lysate or not for routine testing. it is generally good advice to use a non-specific lysate so that any reaction can be examined to determine if the reaction is caused by endotoxin or by glucan.

if glucan contamination is suspected within a process, the main way to determine if glucans are present or not is for the laboratory to run two tests on the samples using glucan-specific and glucan-nonspecific (that is endotoxin-specific) lysates. if the resultant reaction is shown to be endotoxin, this can be quantitated using available lAl methods. if it is glucan, then presence-absence can be determined.

The detection of beta glucan in a pharmaceutical product requires careful assessment by medical experts. Unlike bacterial endotoxin, (1→3)-ß-D-glucan is not specifically regulated and the potential impact of the presence of glucan will dependent upon the levels detected, patient physiology and the intended route of administration. Even if the conclusion of the review is that the presence of glucan is not a concern, it is often incumbent upon the manufacturer to identify a root cause and to remove the impurity. Moreover, if glucan is found to appear in the product at a level not detected previously (as with a shift in relation to the LAL assay) this may indicate a process change, such as a change to the way in which a filter has been manufactured or a variation to a process step such as rinsing. This should trigger a root cause investigation.

Given that one primary source of glucan is cellulosic-based, filter manufacturers should ideally test for glucans and certified filters are low in glucan, along with similar assessments for leachables and for endotoxin. If it is not possible for the pharmaceutical manufacturer to use low-glucan filters, then appropriate pre-processing rinsing should be undertaken in order to lower the risk of glucan fibers from being eluted into the product.

Author Biography

Tim Sandle, Ph.D., is the Head of Microbiology at the Bio Products Laboratory. Dr. Sandle is responsible for a range of microbiological tests, batch review, microbiological investigation and policy development. In addition, Dr. Sandle is an honorary consultant with the University of Manchester and is a tutor for the university’s pharmaceutical microbiology MSc course. He also runs an online pharmaceutical microbiology site (www.pharmamicro.com). Dr. Sandle is a chartered biologist and holds a first class honors degree in Applied Biology; a Master degree in education; and a Ph.D. in microbiology. Dr. Sandle can be reached at: [email protected]

References

1. Sandle, T. (2004): “Three Aspects of LAL Testing: Glucans, Depyropgenation and Water System Qualification,” PharMIG News, No. 16, June / July 2004, pp. 3-12.
2. Guy, D. (2003): ‘Endotoxins and Depyrogenation’ in Hodges, N. and Hanlon, G., Industrial Pharmaceutical microbiology: Standards and Controls, Euromed, pp12.1 – 12.15
3. Finkelman M.A. and Tamura H. (2005). Detection and measurement of (103)-beta-Dglucan with Limulus amebocyte lysate-based reagents. In: Young S-H, Castranova V (Eds.). Toxicology of 13- beta-Glucans: Glucans as a Marker for Fungal Exposure. Boca Raton: CRC Press/Taylor & Francis: 179-197
4. Wiederhold, N.P., Najvar, L.K., Vallor, A.C., Kirkpatrick, W.R., Bocanegra, R., Molian, D., Olivo, M., Graybill, J.R., and Patterson, T.F. (2008) Assessment of (1→3)-ß-D-glucan concentration as a measure of disease burden in a murine model of invasive pulmonary aspergillosis Antimicrob. Agents Chemother. 52; 1176-1178
5. Buchacher, A., Krause, D., Wiry, G., and Weinberger, J. (2010). Elevated endotoxin levels in human intravenous immunoglobulin concentrates caused by (1→3)-beta-D-Glucans. PDA J. Pharm. Sci. Technol. 64, 536-544
6. Brown, G.D. and Gordon, S. (2001) “Immune Recognition: A New Receptor for β-Glucans”, Nature 413, 36 – 37
7. Notranicola, S. et al. (2003) ‘Investigation Of A Process Related Contaminant In a Clinical Drug Product’, Recovery of Biological Products XI, Sept 14-19, 2003, Banff Alberta, Canada
8. Nakao, A., Yasui, M., Kawagoe, T., Tamura, H., Tanaka, S., and Tagaki, H. (1997) “False positive endotoxemia derives from gauze glucan after hepatectomy for hepatocellular carcinoma with cirrhosis,” Hepato-Gastroenterology 44: 1413-1418
9. Pearson, F. C., Bohon, J., Lee, W., Bruszer, G., Sagona, M., Jakubowski, G., Dawe, R., Morrison, D., and DinareUo, C., (1984): “Characterization of Limulus Amoebocyte Lysate-Reactive Material from Hollow-Fiber Dialyzers,” Appl. Environ. Microbiol. 48: 1189-1196
10. Persat, F., Ranque, S., Derouin, F., Michel-Nguyen, A., Picot, S., and Sulahian, A. (2008) Contribution of the (1→3)-ß-D-glucan Assay for the Diagnosis of Invasive Fungal Infections. J. Clin. Micro. 36: 1009-1013
11. Hachem, R.Y., Kontoyiannis, D.P., Chemaly, R.F., Jiang, Y., Reitzel, R., and Raad, I. (2008) Utility of galactomannan enzyme immunoassay and (1→3)-beta-D-glucan in diagnosis of invasive fungal infections: low sensitivity for Aspergillus fumigatus infection in hematologic malignancy patients. J. Clin. Micro. 47: 129-133
12. Ohata, A., Usami, M., Horiuchi, t., et al. (2003). Release of (1,3)-beta-D-glucan from depthtype membrane filters and their in vitro effects on proinflammatory cytokine production. Artif. Organs 27, 728-735
13. Pearson, F. C., Caruana, R., Burkart, J., et al. (1987). the use of the limulus amebocyte lysate assay to monitor hemodialyzer-associated soluble cellulosic material (LAL-reactive material). Prog. Clin. Biol. Res. 231, 211-222
14. Pearson, F. C., Bohon, J., lee, W., et al. (1984). Characterization of limulus amoebocyte lysatereactive material from hollowfiber dialyzers. Appl. Environ. Microbiol. 48, 1189-1196
15. Iwanaga, S., Morita, T., Nakamura, T., and Aketagawa, J. (1986). The hemolymph coagulation system in invertebrate animals. J. Protein Chem 5: 255-268
16. Elin, R. J. and Wolff, S. M. (1973): “Nonspecificity of the Limulus Amebocyte Lysate Test: Positive Reactions with Polycnucleotides and Proteins,” The Journal of Infectious Diseases, Vol. 128, No. 3
17. Pool, J., Joharr, G., James, S., Petersen, I. and Bouic, P. (1997): “Differentiation Between Endotoxin and Non-Endotoxin Pyrogens in Human Albumin Solutions Using an Ex Vivo Whole Bood Culture Assay,” Journal of Immunoassay, 20 (1 & 2), pp. 79-89
18. Roslansky, P. F. and Novitsky, T. J. (1991): “Sensitivity of Limulus Amebocyte Lysate (LAL) to LAL-Reactive Glucans,” Journal of Clinical Microbiology, Nov. 1991, pp. 2477-2483
19. Cooper, J. F., Weary, M. E. and Jordan, F. T. (1997): “The Impact of Non-endotoxin LAL-Reactive Materials on Limulus Amebocyte Lysate Analyses,” PDA Journal of Pharmaceutical Science and Technology, Vol. 51, No. 1, January – February, 1997
20. Tanaka, S., Aketagawa, J., Takahashi, S., Shibata, Y., Tsumuraya, Y. and Hashimoto, Y. (1993): “Inhibition of High-Molecular-Weight- (1 β 3)- β -D-Glucan-Dependent Activation of Limulus Coagulation Factor G by Laminaran Oligosaccharides and Curdlan Degradation Products,” Carbohydrate Research, 244, pp. 115-127
21. Pickering, J.W., Sant, H.W., Bowles, C.A., Roberts, W.L., and Woods, GL. (2005) Evaluation of a (1→3)-beta-D-glucan assay for diagnosis of invasive fungal infections. J. Clin. Microbiol. 43:5957-62
22. Miyazaki T, Kohno S, Mitsutake K, et al. (1995). Plasma (1–3)-beta-D-glucan and fungal antigenemia in patients with candidemia, aspergillosis, and cryptococcosis. J Clin Microbiol; 33:3115–3118
23. Yoshida M., Inada K,. et al (1996): “An Assay Method of (1, 3)-ß-D-glucan to Diagnose Invasive Mycosis. A Utilization of Monoclonal Antibody to the Activated Factor C in Blood Coagulation System of Horseshoe Crab.” In Fugal Cells in Biodefense Mechanism (Suzuki, S. & Suzuki, M., eds), Saikon Publishing Co., Ltd. (Tokyo), 1996. pp. 265-271