National EMS Network Newsletter, February 1998

Current Status of Research on EMS: Gerald J. Gleich, M.D.

By Gerald J. Gleich, M.D.

It has been almost a decade since the eosinophilia-myalgia syndrome (EMS) epidemic. Readers of this publication know full well the nature of this syndrome with its manifold manifestations and its ability to cause prolonged suffering. Readers also know the extensive work performed by the state Departments of Health and the Centers for Disease Control showing an association between ingestion of L-tryptophan and the occurrence of EMS and, subsequently, that tryptophan causing EMS was produced by Showa Denko KK. It is noteworthy that both of these associations have come under strong attack. A summary of these attacks was published in a 1996 supplement to the Journal of Rheumatology (1). The thrust of the critics is that all epidemiological studies of EMS are flawed and, thus, no valid evidence exists that ingestion of L-tryptophan caused the EMS epidemic. Indeed, the very existence of the EMS epidemic is doubted by these critics (2). Responses to these criticisms have been forthcoming from Kilbourne and his colleagues of the Centers for Disease Control (3) and from Belongia and myself (4). Suffice to say that the criticisms of epidemiological studies of EMS and its relationship to L-tryptophan can be answered point by point and readers interested in this debate should consult references 1-4. From the perspective of one concerned with discovering the cause(s) of EMS, I must add that patients with EMS symptoms were common in late 1989 and 1990; now, new patients with these symptoms are rare.

Research into the cause of EMS has continued and has focused on contaminants in L-tryptophan implicated in causing EMS. Careful chemical analyses comparing implicated to nonimplicated L-Tryptophan have revealed several contaminants, including peak E (1,1'-ethylidenebis[tryptophan]), 3- (phenyl amino)alanine and others, including one peak, called AAA, which is especially strongly related to EMS. These studies show us the chemical contaminants in the L-tryptophan, which are associated with EMS. But association does not prove guilt! What we need is a way to prove that a particular contaminant or a combination of contaminants causes EMS. In one sense, the EMS epidemic was a tragic experiment in which we learned that certain contaminants in Showa Denko L-tryptophan cause disease. But which contaminants and how did they cause disease? Presumably, feeding the contaminants to healthy persons would again cause disease. But can we identify a surrogate for EMS in patients? Two approaches to this end have been taken, one, to cause disease in an animal fed L-tryptophan, and the other to develop a test tube analysis. Establishment of either an EMS-like disease in an animal fed implicated L-tryptophan (referred to as an animal model) or a response stimulated in a test tube (referred to as a bioassay) would permit identification of the critical contaminants. Dissection of the animal model would permit an understanding of the disease mechanism. It is worthy of emphasis that this approach also applies to the ingredients in the oil causing the Spanish toxic oil syndrome.

Initial results from bioassay experiments with L-tryptophan appeared promising. Results from several laboratories suggested that a bioassay could be established such that cells from patients (and normal persons) could be stimulated by implicated, but not nonimplicated, L-tryptophan. However, more extensive studies have not corroborated these reports and a reproducible bioassay that consistently and robustly discriminates between implicated and nonimplicated L-tryptophan still eludes us (5). Other studies have tested whether L-tryptophan contaminants are able to stimulate cells, for example, to produce increased collagen, and two studies showed that peak E stimulates human cells to produce collagen (6,7). These results are of interest because increased collagen formation is associated with skin thickening and fibrosis, features characteristic of EMS. However, these findings do not mean that peak E is the cause of EMS: it only establishes that peak E can be associated with an EMS-like characteristic. To date, no bioassay able to consistently and robustly discriminate implicated from nonimplicated L-tryptophan has emerged from research laboratories.

Initial results from animal models of EMS also appeared promising. Two studies in rats showed that skin thickening could be induced by administration of case-associated L-tryptophan or peak E. However, subsequent analyses showed that even nonimplicated tryptophan induced skin thickening (8). In mice, rather comparable results have been shown, and skin inflammation and fibrosis are stimulated by administration of peak E. These assays encouraged hope that an animal model of EMS could be developed. However, additional experiments, summarized by Clauw (9), were not able to reproduce these initial promising results. Because various strains of mice are available and results could differ strikingly depending upon the strain tested, various strains were analyzed and again none of the animals reproducibly developed EMS-like abnormalities. One contaminant in implicated L-tryptophan is 3-(phenylamino)alanine. This compound is remarkably similar to 3-phenylamino-1, 2-propanediol, a substance implicated in the Spanish toxic oil syndrome. The finding of two quite similar molecules from two epidemics with similar manifestations prompted additional studies. In one study, 3-phenylamino-1 ,2-propanediol (in the toxic oil) was incubated with liver cells and was metabolically transformed into 3-(phenylamino)alanine (the L-tryptophan contaminant) (10). This finding suggested a common pathway for EMS and toxic oil syndrome with 3-(phenylamino)alanine as a critical molecule. So, an animal feeding study was conducted in which 3-(phenylamino)alanine was fed to mice. Unfortunately, no clear-cut EMS-like picture developed. However, it is possible that the L-tryptophan contaminants are metabolized differently by different species, such that only primates possess the critical cellular machinery to produce the disease. To test this possibility, L-tryptophan was fed to cynomolgus monkeys. Both case-associated and nonimplicated L-tryptophan were given for 12 weeks; these animals consumed L-tryptophan dosages of 800 mg/kg, which is 10-20 times greater than the dosages consumed by patients who developed EMS. Although the monkeys developed gastrointestinal symptoms (manifested by vomiting), nonetheless no EMS-like abnormalities occurred. Thus, no reproducible animal model of EMS useful for identification of the critical L-tryptophan contaminants and for dissection of the abnormal physiology of EMS has been identified.

The failure of both the bioassay and the animal feeding experiments to yield robust and reproducible results has been a major disappointment. In retrospect, my initial optimism that we could identify a bioassay should have been tempered by recognition of the failure of all attempts to identify bioassays or animal models of the Spanish toxic oil syndrome. The clinical similarities between these two diseases, EMS and the toxic oil syndrome, suggest that they share a common mechanism of disease or pathophysiology. Failure to identify a means to pursue investigations through a bioassay or an animal model has been disheartening on both sides of the Atlantic Ocean. Thus, one cannot pursue the identification of the individual chemicals causing EMS or the toxic oil syndrome through use of the bioassay nor can one explore the mechanisms by which the disease occurs utilizing the animal model. Thus, the present circumstances regarding EMS (and the toxic oil syndrome) are rather unsatisfactory. In spite of the controversy referred to above, we know that an epidemic occurred. As a consequence of the investigations by the public health authorities, L-tryptophan produced by Showa Denko KK has been implicated as a cause of EMS. Certain contaminants present in the L-tryptophan have been implicated as candidates for causation. However, we do not know the exact structures and without this knowledge one fears that another epidemic will occur at sometime. Recall that the Spanish toxic oil epidemic occurred in the spring of 1981 and the EMS epidemic occurred eight years later. One wonders when the next epidemic of an EMS/toxic oil syndrome-like disease will occur and under what circumstances. Furthermore, our knowledge of the abnormal physiology of EMS and the toxic oil syndrome is remarkably scant. We know that eosinophilia is stimulated in the blood, but only scant information is available concerning the mechanisms by which this occurs. We know that eosinophil degranulation occurs in tissues, but we do not know whether this causes symptoms or simply is a marker for the disease. Activated lymphocytes of the T lymphocyte variety are present in tissues, but we do not know whether these cells are critical for disease. The prolonged course of EMS/toxic oil syndrome suggests a continuing stimulus to the body's immune system, but as yet, we have no markers for this and, thus, no ability to define its occurrence or severity. Because the EMS epidemic is now behind us, funding agencies, such as the National Institutes of Health, are not enthusiastic about granting monies for its investigation. One worthy objective of the Network might be to contact your Representatives and Senators in Congress to alert them to the need for investigation of EMS. Our group at Mayo continues to investigate the contaminants in the toxic oil and in implicated L-tryptophan and to pursue animal models and bioassays. Without such tools, further progress in understanding these diseases will depend on the occurrence of another epidemic, an event which no one wants!

(C) Copyright 1998 Gerald J. Gleich, M.D.


  1. Clauw DJ, Pincus T: Eosinophilia-myalgia syndrome: review and reappraisal of clinical, epidemilogic and animal studies symposium. J Rheumatol 1996; 23 (Suppl 46): 1-110.
  2. Daniels SR, Hudson KI, Horwitz RI: Epidemiology of potential association between L-tryptophan ingestion and eosinophilia-myalgia syndrome. J Clin Epidemiol 1995; 48:1413-27.
  3. Kilbourne EM, Philen RM, Kamb ML, Falk H: Tryptophan produced by Showa Denko and epidemic eosinophilia-myalgia syndrome. J Rheumatol 1996; 23(Suppl 46): 81-8.
  4. Belongia EA, Gleich GJ: The eosinophia-myalgia syndrome revisited. J Rheumatol 1996; 23:1682-85 (Editorial).
  5. Kita H, Mayeno AN, Weyand CM, Goronzy JJ, Weiler DA, Lundy SK, Abrams JS, Gleich GJ: Eosinophil-active cytokine from mononuclear cells cultured with L-tryptophan products: An unexpected consequence of endotoxin contamination. J Allergy Clin Immunol 1995; 95:1261-67.
  6. Tagaki H, Ochoa MS, Zhou L, Helfman T, Murata H, Falanga V: Enhanced collagen synthesis and transcription by peak E, a contaminant of L-tryptophan preparations associated with the eosinophilia-myalgia syndrome. J Clin Invest 1995;96:2120-2125.
  7. Zangrilli JG, Mayeno AN, Vining V, Barga J: 1,1-ethylidenebis(L-tryptophan), an impurity in L-tryptophan associated with eosinophilia-myalgia syndrome, stimulates type I collagen gene expression in human fibroblasts in vitro. Biochem Mol Biol Int 1995;37:925-933.
  8. Love LA, Rader JI, Crofford LJ, et al: Pathological and immunological effects of ingesting L-tryptophan and 1,1-ethylidenebis(L-tryptophan) in Lewis rats. J Clin Invest 1993:91:804-811.
  9. Clauw GJ: Animal models of eosinophilia-myalgia syndrome. J Rhematol 1996;23(Suppl 46):93-8.
  10. Mayeno AN, Benson LM, Naylor S, Colberg-Beers M, Puchalski JT, Gleich GJ; Biotransformation of 3-(Phenylamino)-1,2 propanediol to 3-(Phenylamino)alanine: A chemical link between toxic oil syndrome and eosinophilia-myalgia syndrome. Chem Res Toxicol 1995;8:911-916.

Gerald J. Gleich, M.D. is the
George M. Eisenberg Professor of Medicine and Immunology

at Mayo Clinic & Foundation, Rochester, MN