The Eosinophilia-Myalgia Syndrome: Status of 205 Patients and Results of Treatment 2 Years after Onset

Methods

Physicians and Registry

In 1990, physicians recognized as having a particular interest in the eosinophilia-myalgia syndrome were contacted by one of the authors (PH) about developing a national registry to monitor the course of patients with the syndrome who had received L-tryptophan. It was recognized that federal agencies, including the CDC, the National Institutes of Health, and the Food and Drug Administration, had an important role in initially describing the syndrome. However, because of regulatory constraints, a long-term monitoring program instituted by any government agency appeared unlikely.

Therefore, an application for funding a registry was submitted to the Council for Responsible Nutrition, a trade association representing the nutritional products industry. An initial award was granted to the Public Health Foundation, a nonprofit public health service organization, to develop a registry. Fifteen physicians agreed to participate in the eosinophilia-myalgia syndrome registry with the guidance of a scientific advisory group that includes several authors of this report. Further funding was later obtained through applying to an independent research advisory council established by the Showa Denko Corporation, a manufacturer of L-tryptophan that has provided support for research studies on the eosinophilia-myalgia syndrome. During the summer of 1992, responsibility for administration of the registry was transferred to the Division of Rheumatology at Vanderbilt University on the basis of its experience in developing databases for long-term monitoring of patients with rheumatic diseases [20, 21].

Patients

At first, patients were entered into the registry if they met all four of the following criteria: 1) eosinophil count of 1000 cells/mm³ or greater; 2) presence of any of the following on clinical evaluation: fasciitis, peripheral neuropathy, polyradiculopathy, interstitial pulmonary disease, pulmonary hypertension, or myocardial involvement; 3) a history of L-tryptophan consumption; and 4) absence of other conditions that could account for these findings.

The original CDC criteria for the eosinophilia-myalgia syndrome were not used. These criteria had been developed for surveillance purposes and had not been validated for clinical use; members of the advisory committee agreed that a modest redefinition of the criteria beyond the presence of eosinophilia and myalgia was necessary. In March 1991, the first and third criteria were expanded to include patients who had eosinophil counts of less than 1000 cells/mm³ or no history of receiving L-tryptophan but did have other clinical features and laboratory test results consistent with the eosinophilia-myalgia syndrome. Members of the committee believed that these relaxations would facilitate a more comprehensive study of the disease because it became apparent that a few patients with a typical clinical picture did not have documented eosinophilia or a history of using L-tryptophan. The percentage of patients in the registry who had not consumed L-tryptophan (2%) was similar to that in the national database reported from the CDC and was believed to be representative of the general population of patients with the syndrome.

Structured Follow-up Patient Review Form

During the first half of 1990, a structured patient review form was developed that delineated symptoms, signs, laboratory test results, biopsy findings, and radiographic data at the time of onset and at 18- to 24-month review. Patients who had a symptom or sign at baseline were classified into one of seven categories at review: resolved, better, same, mixed, worse, dead, or unknown. A new onset could be recorded more than 3 months after presentation for patients who did not have a sign or symptom at baseline.

The patient review form included a listing of therapeutic interventions and responses to these interventions. To enhance the consistency of the data collected, specific guidelines and criteria were developed for completion of the forms in relation to descriptions of severity and frequency of each symptom and sign. A modest administrative fee was provided to the participating physicians for each completed patient review form.

Data Analysis

The patient review forms were completed between March and September 1991, 18 to 24 months after the onset of the eosinophilia-myalgia syndrome for 245 patients. Data from 18- to 24-month review were available for 205 patients; 9 patients had died, and no follow-up data were available for 31 patients. After the completed forms were reviewed and checked for inconsistencies, each form was returned twice to participating physicians between December 1992 and June 1993 to obtain missing data and clarify occasional apparently contradictory data (such as a symptom marked as “never present” at onset and “improved” after 2 years).

The data were entered into a Paradox microcomputer database (Borland International, Inc., Scotts Valley, California). We computed descriptive statistics on the frequencies of various symptoms at onset and follow-up review and responses to various therapeutic interventions.

Results

Seventy-nine percent of patients were women, and 98% were white. The mean age at diagnosis was 50.9 years (range, 20 to 82 years). Ninety-eight percent were reported to have received L-tryptophan before symptoms developed; the drug had been recommended by a health professional in 47% of these instances. The median duration of L-tryptophan use before the onset of symptoms of the eosinophilia-myalgia syndrome was 8 months (mean, 22.4 months; range, less than 1 to 232 months). Ninety-eight percent of the patients had eosinophil counts greater than 1000 cells/mm³ at presentation; the mean eosinophil count was 4705 cells/mm³ (median, 3564 cells/mm³; range, 56 to 41 246 cells/mm³).

More than 75% of the patients had myalgia, severe fatigue, or rash at presentation (Table 1); 51% to 75% had respiratory symptoms, paresthesias, edema, muscle weakness, arthralgias, alopecia, or muscle cramps; and 25% to 50% had weight loss, fever, cognitive changes (impaired memory, impaired concentration, or mood change), weight gain, gastrointestinal symptoms, or dry mouth. Information on the severity of symptoms was not reported for enough patients to provide meaningful data.

Eighteen to 24 months after the onset of the syndrome, physicians reported that myalgia, fatigue, rash, respiratory symptoms, edema, muscle weakness, arthralgia, alopecia, weight loss, fever, weight gain, dry mouth, and oral ulcerations had resolved or improved in more than 60% of the patients. By contrast, physicians reported that cognitive changes had improved in only 27% of patients and had worsened in 32%. The other symptoms reported as worse in at least 10% of patients were muscle weakness (10%), weight gain (10%), and fatigue (11%). New symptoms developing 1 year after the onset of the eosinophilia-myalgia syndrome were unusual. A new onset of cognitive changes was reported in 30 patients (15%), and new onset of muscle cramps was reported in 20 patients (10%); new onset of other symptoms was reported in 10 or fewer patients (5%).

Physical findings at presentation (Table 2) included fasciitis or scleroderma-like skin changes (57%), myopathy (34%), peripheral neuropathy (34%), and interstitial lung disease (14%). After 18 to 24 months, signs detectable on physical examination or other abnormalities associated with each of these major components of illness had resolved or improved in most patients; peripheral neuropathy, which was unchanged in 47% of patients, was only rarely reported as worse. Nine patients (4%) died; older age and ascending neuropathy severe enough to cause quadriplegia were significant predictors of death (data not shown).

Therapeutic interventions are summarized in Table 3. The only medication received by more than 20% of the patients was an oral corticosteroid. Improvement was reported in 79% of the 193 patients who received corticosteroids. Other treatments reported to have resulted in favorable responses in more than 50% of patients included amitriptyline, acetaminophen, aspirin, plasmapheresis, naproxen, methotrexate, diphenhydramine, furosemide, cyclobenzaprine, penicillamine, triazolam, cyclosporine, colchicine, dantrolene, and fluoxetine.

Discussion

The patients described in this series appear to be similar to previously described patients with the eosinophilia-myalgia syndrome [5-15, 17-19]. Analyses done 18 to 24 months after onset indicated that most symptoms in most patients had improved or resolved. Physicians reported that most physical findings present at onset had resolved or improved at 18- to 24-month follow-up; only peripheral neuropathy was unchanged. Although some patients showed unchanged physical findings with residual abnormalities, no evidence of persistent inflammation was seen. These findings are similar to those in a follow-up study of 163 patients who had been included in initial reports of the eosinophilia-myalgia syndrome [22].

The natural history of the eosinophilia-myalgia syndrome appears to be similar to that reported for the toxic oil syndrome, a food-related epidemic that occurred in Spain in 1981 [23-26]. In most patients with the toxic oil syndrome, disease activity is diminished and manifestations regress months to years after the onset of illness [27, 28], although many patients describe persistent fatigue, myalgia, muscle cramps, cognitive changes [28], and functional disability with persistent musculoskeletal and constitutional symptoms [29].

Only cognitive changes were reported to have worsened 2 years after onset in more than 12% of patients with the eosinophilia-myalgia syndrome and to have occurred with a “new onset” in 30 patients. Our data are based on the clinical impressions of the reporting physicians and do not include information on the neurocognitive status of patients before they received L-tryptophan. Further formal neuropsychological testing in some patients has been reported to indicate poor performance in verbal memory and visual search and attention compared with healthy and depressed control patients; this finding is not explained by concurrent use of medications [30, 31]. The specificity and significance of observed neurocognitive abnormalities in patients with the eosinophilia-myalgia syndrome remains uncertain [19]. An ongoing multicenter study is being done to compare the cognitive function in patients who have the eosinophilia-myalgia syndrome with that in control patients who have fibromyalgia or were among the vast majority of patients who received L-tryptophan and did not develop the eosinophilia-myalgia syndrome. The investigators of the study also seek to determine the possible influence of preexisting or coexisting medical and psychiatric illness on cognitive function.

A positive response to corticosteroids was reported in 79% of patients during the acute phase of the eosinophilia-myalgia syndrome. This response included resolution of eosinophilia, which may have influenced the reports of a favorable outcome, but also included resolution of peripheral edema and clearing of pulmonary infiltrates. Although the role of the eosinophil in the pathogenesis of the eosinophilia-myalgia syndrome is unknown, the clinical resolution of symptoms may indicate the potential value of empiric use of corticosteroids in patients with acute unknown conditions associated with eosinophilia. The reporting of favorable responses to other treatments is difficult to interpret because these treatments were used by only a few patients and may partially reflect spontaneous improvement or biases on the part of the patients and physicians. No evidence suggested that treatment with cytotoxic drugs or corticosteroids was valuable in long-term patient management [19].

This database was organized in recognition of a unique opportunity to monitor the natural history and outcomes of a new illness from the time of its initial description. Nonetheless, as the database was developed, several problems that affected our study became apparent. First, most of the patients were selected for inclusion in the registry on the basis of referral to a physician who had a recognized interest in the eosinophilia-myalgia syndrome. Second, although guidelines for collection of clinical data on symptoms, signs, and responses to therapy were developed, they were not rigorously standardized and greatly depended on physician judgment. Third, despite considerable interest of the participating physicians in the eosinophilia-myalgia syndrome and extensive efforts of the database managers to acquire complete data on all patients, residual data were still missing, detracting from an optimal review.

These problems result in part from the fact that standardized data in clinical settings are recorded primarily for laboratory findings but not for symptoms and physical findings. These shortcomings might be partially overcome through the use of standardized patient self-report questionnaires [32, 33]. Such questionnaires are being used in further longitudinal studies of patients with the eosinophilia-myalgia syndrome but are limited by the absence of baseline data.

Nonetheless, we present the largest collection of data on patients with the eosinophilia-myalgia syndrome 2 years after onset, from 15 clinical settings according to a predefined protocol. It is reassuring that only a few further cases of a disease resembling the eosinophilia-myalgia syndrome have been reported since 1989; this may be regarded in part as an important triumph of modern public health surveillance methods. Nonetheless, the possibility remains that a disease similar to the eosinophilia-myalgia syndrome may be triggered in the future by environmental exposures. Further surveillance, including the possibility of a registry for surveillance of eosinophilia of unknown origin, might enhance the early identification of other potential environment-associated diseases.

Disclosure Statement: Phillip A. Hertzman has received consulting fees from attorneys for both plaintiffs and Showa Denko; Daniel J. Clauw has received research grant support from Showa Denko and consulting fees from attorneys for both plaintiffs and Showa Denko; Lee D. Kaufman has received research grant support from Showa Denko and consulting fees from attorneys for plaintiffs; John Varga has received research grant support from Showa Denko and consulting fees from attorneys for both plaintiffs and Showa Denko; Richard M. Silver has received research grant support from Showa Denko and consulting fees from attorneys for plaintiffs; Holly L. Thacker has received consulting fees from attorneys for both plaintiffs and Showa Denko; Phillip Mease has received consulting fees from attorneys for plaintiffs; Luis R. Espinoza has received research grant support from Showa Denko and consulting fees from attorneys for both plaintiffs and Showa Denko; and Theodore Pincus has received grant support from Showa Denko and consulting fees from attorneys for Showa Denko.