banner ad

During the past decade we have witnessed a tumultuous debate over the disease risks posed by microbes that inhabit metalworking fluid (MWF) systems. Not infrequently, that debate has occurred in the absence of satisfactory data. This paper addresses the author's perspective on what types of data are needed in order to assess the actual disease risks posed by MWF microbes. The approach must be multidisciplinary and coordinated, including stakeholders with expertise in epidemiology, fluid management, immunology, industrial hygiene, microbiology, and public health medicine. Traditional microbiological sampling and test methods must be augmented by new, consensus methods that are adopted by industry stakeholders. Entities performing these tests should be participating in interlaboratory cross-check programs. The author will use Mycobacterium immunogenum to illustrate the general model for this strategy.


Over the course of the past decade, the metalworking industry stakeholders have become increasingly aware of non-infectious disease health risks posed by microbes present as metalworking fluid system contaminants. The author first addressed this topic in 2002 (Passman and Rossmoore (1)). In that paper, Passman and Rossmoore highlighted the microbe-associated issues that represented known and hypothesized health risks to exposed workers. Since that time, interest in the relationship between the bacterium Mycobacterium immunogenum and the disease hypersensitivity pneumonitis (HP) has continued to grow (MMWR (2), Bracker, et al. (3), O'Brien (4), Thorne, et al. (5)). It is possible that the focus on M. immunogenum and HP has been at the expense of adequate attention to other health risks posed by metalworking fluid (MWF) and metalworking system microbial communities. In this paper, the author will present several hypotheses. Drawing on the literature from the metalworking and other industries and indoor environments, the author will argue the case for testing these hypotheses in the metalworking environment.

Hypothesis 1: Endotoxin Exposure Presents a Significant Health Risk to Workers Exposed to MWF Mist as Mist-Associated Bioaerosols

Endotoxin toxicity is well documented (Castellan, et al. (6), Latza, et al. (7), Todar (8); Liebers, Brüning, and Raulf-Heimsoth (9); Rylander (10)). The no observable effect level (NOEL- highest dose that does not cause observable effects) is 9 EU m-3 (Castellan, et al. (6)). Endotoxins are known to cause a range of symptoms from mild fever and respiratory impairment to death (Latza, et al. (7)). Endotoxins (also called pyrogens, since they induce fever) are lipopolysaccharide (LPS) molecules that comprise the outer envelope of Gram-negative bacteria cell walls. Endotoxins are complex amphiphilic molecules approximately 10,000 Daltons (10 kDa) and are comprised of three primary components: Lipid A, Core-polysaccharide, and O-polysaccharide.

The Lipid A portion of the molecule extends from the cell surface into the surrounding environment and is comprised of a phosphorylated N-acetylglucosamine dimer to which typically six unsaturated fatty acids are attached. The structure of the Lipid A moiety is highly conserved among all known gram-negative bacteria. Lipid A reacts at the surface of the macrophages, inducing the release of cytokines (Todar (8)). This immune system response leads directly or indirectly to the symptoms of endotoxin toxicity. Thus, lipid A is the primary toxigenic component of LPS.

The core (R) polysaccharide (R-antigen) is a short sugar chain that is linked to the lipid A component at the 6-carbon position of the N-acetylglucosamine dimer. The sugar 2-keto-2- deoxyoctanoic acid(KDO)is unique to LPS in nature, and is found universally in gram-negative bacteria. Heptose is also present nearly universally in LPS. Other sugars that may be present in the R-antigen include galactose, glucose, and glucosamine. The composition of R-antigen is somewhat conserved, but varies among different genera of gram-negative bacteria.

The innermost portion of the LPS molecule is the Opolysaccharide (O-antigen). Comprised of up to 40 repeating subunits of three to five sugars, the O-antigen of each species (in some cases the O-antigen is strain-specific) is unique.

The toxicological effects of endotoxin are well documented and have been reviewed recently by Leibers, et al. (9). Latza, et al. (7) demonstrated a 5-fold increased risk of wheezing and a 4-fold increased risk of coughing symptoms among textile workers exposed to > 450 EU m-3, as compared to an unexposed control population. Rylander (10) reported that the International Committee on Occupational Health (ICOH) had identified the following ranges for endotoxin health effects:

100 EU m-3 airway inflammation

1,000 EU m-3 systemic effects; acute bronchial restriction

2,000 EU m-3 toxic pneumonitis

10,000 EU m-3 organic dust toxic syndrome

where EU are endotoxin units and 1 EU ≈10 ng endotoxin

Since gram-negative bacteria are nearly ubiquitous in MWF systems, it is reasonable to infer that endotoxin will also be nearly ubiquitous. A survey of MWF facilities (Crook and Swan (11)) reported airborne endotoxin concentrations ranging from 1 EU m-3 to 7,600 EU m-3 among samples collected at machine shops throughout England. Laitinen, et al. (12) surveyed 18 metalworking facilities and reported airborne endotoxin concentrations ranging from <0.4 EU m-3 to 1.4 x 103 EU m-3. Lewis, et al. (13) recovered < 0.05 EU mL-1 to > 1 x 106 EU mL-1 in MWF samples and 0.5 EU m-3 to 2.5 EU m-3 in MWF system aerosols. Park, et al. (14) sampled 140 MWF from small sumps at 19 machine shops. They performed covariance analysis to model the impact of fluid temperature, MWF concentration, pH, tramp oil concentration, formulation type (emulsifiable oil or synthetic) and machining operation on endotoxin concentration. Park and his collaborators determined that tramp oil contamination, elevated temperature, low pH (< 8.5), and fluid type (emulsifiable oils tended to have higher EU mL-1 than did synthetics) contributed to increased endotoxin concentrations. Park and his team did not evaluate MWF formulations that are intentionally contaminated with Gram-negative bacteria (Fluri (15)). Focusing on a single facility Abrams, et al. (16) determined that airborne endotoxin concentration geometric means ranged from 10.8 ± 2.1 EU m-3 in the finished assembly department to 803.6±1.8EUm-3 in the case department. The investigators also reported a strong correlation between endotoxin and total particulates. Zucker, et al. (17) reported airborne endotoxin concentrations of up to 63 EU m-3 and Wang, et al. (18) reported concentrations ranging from 11.6 ± 1.8 EU m-3 near a milling center in one facility to (3.3 ± 0.7) x 104 EUm-3 near a large parts machining center at a second facility.Wang's group recovered 3.4±2.8EUm-3 at a control site. Moreover, they reported a bimodal distribution of airborne endotoxin as a function of aerosol particle size. In the particle size range 1 to 10 µm, EU m-3 covaried with CFU bacteria m-3; peaking at 2.45 µm. Airborne endotoxin concentration had a secondary peak associated with 0.39 µm particles.

Gordon (19) has suggested that endotoxin exposure may play a significant role in the toxicity of used MWF. In a recent survey of MWF microbiology, Simpson, et al. (20) typically recovered > 106 CFU bacteria mL-1 and > 105EU mL-1 from machine sumps. Linnainmaa, et al. (21) reported that at = 500 ppm (active ingredient - a.i.), formaldehyde-condensate microbicides suppressed bulk-fluid endotoxin concentrations; corroborating results that had been reported by Douglas, et al. (22). There is a growing body of literature demonstrating that airborne endotoxin concentrations in the metalworking environment are frequently in the toxic range per the ICOH classifications noted above.

In 2001, ASTM approved a consensus practice for sampling and analyzing bioaerosol-associated endotoxin (ASTM (23)) and in 2002 the society approved a method for testing MWF concentrate for endotoxin (ASTM (24)). Thorne, et al. (25) subsequently evaluated ASTM E2144 against previously reported protocols. (Thorne recommended against using the consensus practice; arguing that ASTM E2144 yielded higher endotoxin background concentrations from filter blanks and greater data variability). Notwithstanding the apparent limitations of the ASTM protocol, Thorne concluded that the results obtained by any of the five methods evaluated did not differ significantly among the methods).

The current situation is that there are consensus methods for determining both bulk fluid and airborne endotoxin concentrations, but insufficient data to model the relationship between MWF and bioaerosol endotoxin concentrations. Moreover, the variables affecting the wide range of airborne endotoxin concentrations reported by Crook and Swan (11), in contrast to the relatively narrow range reported by Lewis, et al. (13), have yet to be examined thoroughly. Airborne endotoxin mapping comparable to total mist particle mapping reported by O'Brien (26) and others is needed in order to quantify the risk posed to people working in machining and metal forming facilities. Additionally, multivariate analysis is needed in order to illuminate the relationships between endotoxin present in bulk, recirculating MWF, and airborne endotoxin concentrations. Data for bulk fluid and airborne endotoxin concentrations need to be coupled with metalworking operations data (fluid chemistry and condition, type of metalworking operation-mist generation dynamics, etc.)-and worker health parameters (for, example respiratory function, antibody titer and endotoxin-exposure related symptoms).

Data for MWF worker exposure remain relatively sparse. However, the existing literature, combined with reports from other industries, substantially confirms Hypothesis 1. Respiratory problems associated with moderate to high endotoxin exposure have been demonstrated unequivocally. Bioaerosol testing at metalworking facilities have documented the presence of endotoxin concentrations well above the 100 EU m-3 ICOH threshold for lung irritation.

Although consensus on the relationship between airborne endotoxin concentration and other MWF bulk fluid and aerosol parameters has yet to be achieved, it is time to pilot improved exposure control strategies. It would also seem prudent for metalworking facilities to incorporate periodic endotoxin bioaerosol mapping surveys into their industrial hygiene surveillance programs.

Hypothesis 2: Mycobacterium immunogenum Is the Microbial Agent Responsible for Hypersensitivity Pneumonitis among Metalworking Industry Workers

. . .Continue to read rest of article (PDF).

Dr. Frederick Passman, PhD is a Certified Metalworking Fluids Specialist with over 35 years experience in Environmental & Industrial Microbiology. His company, Biodeterioration Control Associates, Inc. (BCA) provides clients with unparalleled expertise in Microbial Contamination Control.

©Copyright - All Rights Reserved