Tropical Pulmonary Eosinophilia: A Case Series in a Setting of Nonendemicity

Transcription

1 MAJOR ARTICLE Tropical Pulmonary Eosinophilia: A Case Series in a Setting of Nonendemicity Andrea K. Boggild, 1 Jay S. Keystone, 1,3 and Kevin C. Kain 1,2,3 1 Faculty of Medicine and 2 McLaughlin-Rotman Center for Global Health, McLaughlin Center for Molecular Medicine, University of Toronto, and 3 Tropical Disease Unit, Division of Infectious Diseases, Department of Medicine, UHN Toronto General Hospital, Toronto, Ontario, Canada Background. Tropical pulmonary eosinophilia (TPE) is a rare but serious manifestation of infection with the lymphatic filarial parasites Wuchereria bancrofti and Brugia malayi. Although endemicity is limited to the tropical and subtropical regions of Africa, South America, and Asia, immigration and travel practices have led to the diagnosis of TPE in areas of nonendemicity. Methods. We herein present a case series of all patients with TPE who presented to the Toronto General Hospital during Results. Seventeen individuals presented with TPE during the study period, and all were of South Asian ancestry. All 17 received an incorrect diagnosis at presentation (median number of consultations before diagnosis, 2), the most frequent of which was asthma (76%). Eosinophil count, serum immunoglobulin E levels, and antifilarial antibody titers were elevated in all patients. Ten of 14 patients had an abnormal chest radiograph finding, and 11 of 12 patients had abnormal results of pulmonary function tests. Conclusions. TPE is an important diagnostic consideration in patients with eosinophilia, respiratory symptoms, and history of exposure to this disease. In the untreated individual, TPE can lead to chronic and progressive respiratory compromise and death. Prompt recognition and treatment with diethylcarbamazine is therefore key to minimizing morbidity and mortality. Tropical pulmonary eosinophilia (TPE) is one of a number of syndromes characterized by pulmonary infiltrates and peripheral eosinophilia. TPE is a variant of filariasis that results from a hypersensitivity response to the microfilariae of the lymphatic-dwelling parasites, Wuchereria bancrofti and Brugia malayi [1, 2]. Lymphatic filariasis is transmitted by mosquitoes and is endemic in many of the tropical and subtropical areas of South America, Africa, Asia, and Oceania [1 3]. Approximately 70% of infections are concentrated in India, Nigeria, Bangladesh, and Indonesia, and worldwide, more than 1 billion people are at risk of acquiring lymphatic filariasis [4]. The global epidemiological characteristics of lymphatic filariasis have been altered by the implementation of the Global Program to Eliminate Lymphatic Filariasis, a public-private partnership Received 9 March 2004; accepted 20 May 2004; electronically published 27 September Reprints or correspondence: Dr. Kevin C. Kain, Toronto General Hospital, 200 Elizabeth St., ES-9-412, Toronto, ON M5G 2C4. Clinical Infectious Diseases 2004; 39: by the Infectious Diseases Society of America. All rights reserved /2004/ $15.00 initiated in The regular, mass administration of ivermectin with diethylcarbamazine or albendazole has resulted in the treatment of millions of at-risk individuals [4]. In 2002, 80 million residents of 34 countries were treated, with coverage ranging from!5% (Nigeria, India, Yemen, and Indonesia) to 195% of at-risk individuals (Zanzibar, Tanzania, Egypt, Guyana, Cook Islands, French Polynesia, and Samoa) [4]. Although 130 million people are infected with lymphatic filariasis worldwide,!0.5% of these infections manifest as TPE [1, 2]. The predisposing factors for the development of TPE are not well understood, although there is evidence that host immune response to a filarial antigen, g-glutamyl transpeptidase (ggt), confers an increased risk of developing TPE [5]. Elevated levels of ggt IgG1 and IgE antibodies are overrepresented among individuals with TPE, compared with persons who have other forms of lymphatic filariasis [5]. The differential antibody response has been postulated to play a role in the type of pulmonary inflammation that is characteristic of TPE [5]. Although few cases of TPE are recognized in regions of nonendemicity [6 9], with increasing global travel and immigration, TPE is an important diagnosis to Tropical Pulmonary Eosinophilia CID 2004:39 (15 October) 1123

2 consider in patients with an appropriate clinical presentation and exposure history. Although TPE typically has a nonspecific presentation and may mimic a number of conditions, it is most often misdiagnosed as asthma because of pulmonary manifestations, such as paroxysmal cough and dyspnea [10 13]. Other commonly reported symptoms include malaise, fever, and weight loss [10]. Rapid amelioration of signs and symptoms with diethylcarbamazine treatment is a hallmark of TPE. Although the clinical response to diethylcarbamazine is often marked, many patients are left with mild residual pulmonary disease following treatment [14]. Progressive pulmonary fibrosis and, ultimately, respiratory failure are the inevitable sequelae of untreated TPE. The purpose of this study was to characterize the clinical presentation and diagnosis of TPE and to identify barriers to the recognition of this disease in countries in which the disease is not endemic. We present a series of 17 patients with tropical pulmonary eosinophilia in Toronto and describe the demographic characteristics, clinical presentation, and outcome of this syndrome. The cases we present typify the clinical spectrum of TPE, including the potential for adverse outcomes in cases of delayed diagnosis and treatment. METHODS The study was a retrospective case series. Charts of all patients with a diagnosis of TPE who presented to the Tropical Disease Unit (TDU) at Toronto General Hospital during were reviewed for prospectively established variables of interest. The diagnosis of TPE was based on the following clinical and laboratory criteria: (1) clinical history supportive of exposure to lymphatic filariasis, (2) peripheral eosinophilia (eosinophil count, eosinophils/l), (3) elevated serum IgE levels (11000 IU/mL; normal range, IU/mL), (4) increased titers of anti-filarial antibodies, (5) peripheral blood specimen that tested negative for microfilariae, and (6) clinical response to diethylcarbamazine. Demographic, clinical, and laboratory data were extracted from the charts onto a standardized data form and entered into an Excel spreadsheet (Microsoft). Descriptive statistics were performed by means of SigmaStat software, version 2.03 (SPSS). This study was approved by the Research Ethics Board of Toronto General Hospital. Serum specimens were tested for filaria antibodies by the Laboratory of Parasitic Diseases at the National Institutes of Health National Institute of Allergy and Infectious Diseases (Bethesda, MD). During the course of the study period, testing for anti-filarial antibodies was standardized, and presentation of data changed from antibody titers (negative titer,!1:32) to mg/ml of IgG antibodies (as determined by an enzyme immunoassay). An anti-filarial antibody concentration mg/ ml was considered to be a positive result. A negative titer or concentration of anti-filarial antibody was considered to be indicative of a low likelihood of filarial infection. RESULTS During the study period, 17 patients received a diagnosis of TPE. We present several illustrative cases that reflect the spectrums of the disease and the clinical presentations. Patient 1. A 29-year-old man from India presented with a 2-month history of nocturnal cough and occasional wheeze with dyspnea. He had immigrated to Canada 8 months before the onset of symptoms. He also reported a 2-kg weight loss over the preceding 6 months but denied having fever or chills. At the time of his referral to the TDU, he was receiving antibiotics and inhaled b-agonists for presumed asthma. The past medical history was otherwise unremarkable. Physical examination revealed bilateral basilar wheezing. Laboratory investigations revealed eosinophilia (eosinophil count, eosinophils/l) and positive results of filarial serologic tests (titer, 1:4096). His IgE level was 381 IU/mL. Results of stool examination were negative for ova and parasites, and results of strongyloides serological testing were negative. Chest radiography detected interstitial markings, and pulmonary function testing revealed mild lung restriction and decreased diffusion capacity. The patient was treated with diethylcarbamazine (6 mg/kg per day) for 21 days. Within 3 days after starting therapy, patient 1 noted a dramatic improvement in his symptoms, with complete resolution by the end of treatment. The eosinophil count decreased to eosinophils/l at the completion of therapy. At the 4-month follow-up visit, patient 1 continued to be asymptomatic, and his eosinophil count had further decreased to eosinophils/l. Pulmonary function testing had improved but continued to show borderline restriction and reduced diffusion capacity. Seven months after completion of treatment, patient 1 continued to be asymptomatic, with an eosinophil count of eosinophils/l. Thirteen months after the end of treatment, he remained asymptomatic, and the 9 eosinophil count had normalized to eosinophils/l. Patient 2. A 56-year-old man, originally from Guyana, was admitted to the hospital because of shortness of breath. Patient 2 had immigrated to Canada 3 years before presentation. He described a 2-year history of dyspnea, wheezing, and dry cough, with exacerbation at night. He denied having fever, chills, weight loss, or malaise. During a previous hospital admission for treatment of shortness of breath, patient 2 was assessed by 2 respirologists and received a diagnosis of asthma. Medications included salmeterol, inhaled steroids, terbutaline, and oral prednisone. The past medical history was notable only for a remote history of smoking. Findings on multiple chest radiographs were normal. Pulmonary function testing showed severe obstruction with gas trapping at residual volumes and 1124 CID 2004:39 (15 October) Boggild et al.

3 increased airway resistance. Physical examination revealed wheezes on chest auscultation. Laboratory investigations at the TDU revealed an eosinophil count of eosinophils/l, with an IgE level of 8332 IU/mL. Results of a filarial serological test were positive (380.5 mg/ml). Examination of stool specimens did not detect ova or parasites, and results of a strongyloides serological test were negative. Patient 2 completed 21 days of therapy with diethylcarbamazine (6 mg/kg per day) and noted marked improvement of his respiratory symptoms. The eosinophil count 1 month after the completion of treatment had decreased to eosinophils/l. Pulmonary function testing 6 months following treatment showed moderate airflow limitation with gas trapping and mild reduction of diffusion capacity. The only respiratory complaint at the 6-month follow-up visit was mild shortness of breath with exertion. Patient 3. A 24-year-old woman from Guyana presented to the hospital with a 5-year history of mild intermittent shortness of breath, nocturnal cough, wheezing, and a 4.5-kg weight loss. Patient 3 had received a diagnosis of asthma several years before presentation but did receive specific treatment or undergo testing. She did not smoke and had no allergies or other relevant medical history. She had immigrated to Canada 6 years earlier. At the time of referral to the TDU, patient 3 was 5 months pregnant. Physical examination findings were otherwise unremarkable. Laboratory investigations revealed an eosinophil count of eosinophils/l, an elevated serum IgE level, and positive results of a filarial serologic test (titer, 1:2048). Stool examination did not detect ova or parasites, and results of a strongyloides serologic test were negative. Pulmonary function testing revealed a small reduction in forced vital capacity, a small concentration of trapped gas, and a small reduction in diffusion capacity. Although a diagnosis of TPE was made, chest radiography and treatment with diethylcarbamazine were postponed until after delivery. One month postpartum, patient 3 presented to another hospital with increasing shortness of breath on exertion. Echocardiography showed right axis deviation with right ventricular hypertrophy, and chest radiography revealed pulmonary edema. Two-dimensional echocardiogram with Doppler imaging showed an enlarged right heart with evidence of pulmonary hypertension and an estimated pulmonary artery systolic pressure of 70 mm Hg. Cardiac catheterization was performed and demonstrated the following: estimated pulmonary artery pressure, 70/40/49 mm Hg; pulmonary artery wedge pressure, 8 mm Hg; left ventricular (LV) end-diastolic pressure, 15 mm Hg; right atrial pressure, 12 mm Hg; and cardiac output, 1.4 L/min. An LV angiogram indicated a grade 2/4 LV (i.e., LV ejection fraction, 40% 60%) Table 1. Selected characteristics of patients with tropical pulmonary eosinophilia who presented to the Tropical Diseases Unit (TDU) at Toronto General Hospital between 1990 and Patient Age, years Sex Country of birth Duration of symptoms, months Duration of Canadian residence, months No. of physicians seen before TDU 1 39 M India F Guyana M Guyana M India M India F Guyana M Guyana F Sri Lanka M India F Guyana F Sri Lanka M Guyana F India M Guyana F Guyana M Sri Lanka M India Overall, median value Tropical Pulmonary Eosinophilia CID 2004:39 (15 October) 1125

4 Table 2. Signs and symptoms at presentation to the hospital for patients with tropical pulmonary eosinophilia. Presenting sign or symptom n/n (%) of patients Cough 15/17 (88) Wheeze 15/17 (88) Dyspnea 15/17 (88) Abnormal chest radiograph findings a 10/14 (71) Fever 5/17 (29) Weight loss 2/17 (12) Fatigue 1/17 (6) Pruritus 1/17 (6) a Documented findings were available for only 14 patients. with diffuse hypokinesis. A 21-day course of diethylcarbamazine (6 mg/kg per day) was started. Five weeks after initiating treatment, patient 3 was clinically unchanged. A second transthoracic echocardiogram showed no improvement. Patient 3 then initiated prednisone (60 mg/kg per day) on a trial basis, with no improvement in symptoms. Her health continued to deteriorate despite intervention, and she died 9 months after TPE was first diagnosed. All patients. Demographic characteristics of the sample are summarized in table 1. The median age was 29 years, with a ratio of male patients to female patients of 1.4:1. All patients were Canadian residents of South Asian ancestry. The region of origin was either Guyana (47%) or the Indian subcontinent (53%). The median interval between immigration to Canada and presentation to the TDU was 18 months (table 1). The majority of patients presented to the hospital after a chronic course of symptoms (median duration of symptoms before presentation, 6 months) (table 1). Common symptoms at presentation were shortness of breath (88%) and nocturnal cough (88%), whereas wheezing was the most common sign (88% of patients) (table 2). Two patients initially denied respiratory or constitutional symptoms and were referred for eosinophilia. However, each of these patients had abnormal physical examination findings (wheezing) or pulmonary function test results. No patient had clinically detectable hepatosplenomegaly, lymphadenopathy, or lymphedema. All 17 patients had been referred to other specialists for diagnosis, consulting a median of 2 physicians each, before being referred to the TDU (table 1). Thirteen patients (76%) had received a previous diagnosis of asthma. Fifteen patients (88%) had received some form of pharmacological intervention for presumed asthma before referral to the TDU, with minimal Table 3. Laboratory parameters of tropical pulmonary eosinophilia cases. Patient Total leukocyte count, 10 9 cells/l a Eosinophil count, 10 9 eosinophils/l b Serum IgE level, IU/mL c Anti-filarial antibody titer d Chest radiograph findings :16,384 Interstitial infiltrates : :4096 Normal e Interstitial infiltrates :4096 Interstitial infiltrates :32, ,100 1:16,384 Interstitial infiltrates :2048 Interstitial infiltrates :2048 Interstitial infiltrates :4096 Interstitial infiltrates ,000 1:16,384 Interstitial infiltrates :4096 Normal :1024 Interstitial infiltrates e Normal ,689 e Normal :32,768 Interstitial infiltrates ,444 1:4096 Overall, median value ,342 a 9 9 Normal range, to leukocytes/l. b Normal range, 9 to eosinophils/l. c Normal range, IU/mL. d Normal titer,!1:32. e For patients 4, 14, and 15, anti-filarial antibody concentrations were 1172, , and 42.4 mg/ml, respectively (normal concentration,!13.0 mg/ml) CID 2004:39 (15 October) Boggild et al.

5 or no improvement in symptoms. b-agonists were the most prescribed medication (76% of patients), and antibiotics (47%), prednisone (41%), inhaled steroids (35%), and ipratropium (12%) were also used. Chest radiograph findings were documented for 14 patients, 10 (71%) of which were reported as interstitial patterning, and 4 (29%) of which were reported as normal. Pulmonary function testing was documented for 12 patients (71%), and of these, results of only 1 test were reported as normal. Pulmonary function testing revealed an isolated restrictive pattern for 3 patients (25%), an isolated obstructive pattern for 1 (8%), an isolated diffusion defect for 2 (16.7%), a combined restrictive pattern with decreased diffusion capacity for 3 (25%), and combined airway obstruction and decreased diffusion capacity for 2 (16.7%). The eosinophil count was elevated in all patients and was eosinophils/l in 16 patients (94%) at presentation to the TDU. The median eosinophil count was eosinophils/l (table 3). The median total leukocyte count was leukocytes/l (table 3). IgE levels and filarial titers were also elevated for all patients. Each patient was treated with diethylcarbamazine (6 mg/kg per day) for at least 21 days. Follow-up data were available for 15 patients, with a mean duration of follow-up of 10.5 months (range, 2 35 months). All but 1 patient who was treated with diethylcarbamazine responded and reported resolution of presenting symptoms. Pulmonary function testing performed after the completion of treated was reported for 4 patients. Pulmonary function parameters returned to normal for only 1 patient, who had initiated a 2-year course of prednisone therapy for his pulmonary symptoms before the diagnosis of TPE was made. DISCUSSION Pulmonary infiltrates with eosinophilia is a broad diagnostic qualifier encompassing a variety of infectious, inflammatory, and allergic etiologies [15]. The differential diagnosis of eosinophilic lung disease includes Loeffler syndrome secondary to helminth infection or drug reaction; allergic bronchopulmonary aspergillosis; chronic eosinophilic pneumonia; vasculitis, such as Churg-Strauss syndrome, polyarteritis nodosa, and Wegener granulomatosis; and idiopathic hypereosinophilic syndrome [13]. Although all of these diseases share similar clinical and laboratory features, there are several criteria that must be met to diagnose TPE, including residence in an area in which filaria is endemic, peripheral eosinophilia, absence of microfilariae in peripheral blood, increased antifilarial antibody titer, elevated serum IgE level, and a favorable clinical response to diethylcarbamazine [10, 16]. A history of paroxysmal nocturnal asthma and radiographic evidence of pulmonary infiltrates are supportive but not diagnostic. It is important to note that chest radiograph findings may be normal in up to 20% of TPE cases [17]. The cases presented above highlight some of the associated findings in TPE, such as interstitial infiltrates (detected by chest radiography) and a restrictive pattern of lung disease with superimposed airway obstruction (revealed by pulmonary function testing), and they reiterate that misdiagnosis is common. Although the initial presentation may closely resemble that of asthma, an irreversibility of airflow limitation with conventional therapies, coupled with marked eosinophilia and exposure or travel history, should raise the suspicion of TPE. In patients with an appropriate exposure history, it is also important to rule out Loeffler syndrome secondary to Ascaris or Strongyloides infection by means of stool examination and serological testing, because this form of nonfilarial TPE is refractory to diethylcarbamazine and may be difficult to distinguish clinically from true filarial TPE [18]. These cases further underscore the necessity of early intervention, because chronic TPE may result in permanent, irreversible deficits in pulmonary function, despite treatment with diethylcarbamazine [14]. In fact, it has recently been shown that pulmonary eosinophils in TPE degranulate and release toxic oxygen radicals long after treatment with diethylcarbamazine [19], increasing damage initiated during the acute stages of disease. When TPE is recognized and treated early, patients often have a dramatic and rapid response to diethylcarbamazine. As illustrated by the second and third cases (patients 2 and 3, respectively), however, the outcome is often linked to the chronicity of TPE. To date, a treatment regimen best suited to patients with chronic TPE has yet to be established. Lung pathological findings in patients with TPE are directly related to the immunologic processes involved in microfilarial clearance, thus intimating a role for anti-inflammatory and immunomodulatory drugs in TPE treatment. IFN-a may hold promise because it has been shown to block IgE-dependent release of eosinophil-derived neurotoxin and IgA-dependent release of eosinophil cationic protein in vitro [20], both of which are cytotoxic and helminthotoxic. Furthermore, IFN-a has clinical use in the treatment of idiopathic hypereosinophilic syndrome [21, 22]. It is also curious that the only patient in our series with normal results of follow-up pulmonary function tests had received a 2-year course of continuous prednisone therapy. It is clear that there is a need for prospectively designed studies that evaluate current and novel treatment regimens for TPE. All of the patients in this case series were of South Asian ancestry and had immigrated to Canada from regions of lymphatic filariasis endemicity, supporting the well-documented geographic preponderance of lymphatic filariasis [3]. In addition, our findings are supportive of results in an expanding body of literature suggesting that disease susceptibility has genetic underpinnings [23]. For instance, immunological hyperreactivity to another filarial nematode, Onchocerca volvulus, has been linked to the Arg110Gln genotypic variant of IL-13 [24], Tropical Pulmonary Eosinophilia CID 2004:39 (15 October) 1127

6 which also confers an IgE-independent risk for asthma and atopy. It is postulated that the Arg110Gln variant leads to enhanced IL-13 signaling, thus heightening the Th2 response [24]. Racial and ethnic biases within disease spectra are also becoming increasingly recognized and clinically relevant [23]. With respect to lymphatic filariasis, it is well documented that the burden of disease is higher in South Asians [10], even in populations in which they are a minority [25]. In addition, it has recently been reported that homozygosity for a variant CHIT1 genotype (encoding chitotriosidase, the putative target for microfilarial chitin), is overrepresented in South Indians who are susceptible to filarial infection, compared with individuals who are not and those who are white or African American [26]. Although the sample in our case series is relatively small, it is of interest that all patients are of the same racial origin, a fact that may well be explained by differential genetic susceptibilities but which requires additional controlled studies. There are several inherent limitations in the present study. Although the study is useful in raising awareness about the diagnosis of TPE in a region of nonendemicity, our ability to draw conclusions is limited by the small sample size and the retrospective design. Nevertheless, our data support currently accepted racial and geographic associations among individuals with TPE and highlight the need for research initiatives in the realm of disease susceptibility and treatment. In summary, TPE is an important diagnostic consideration in patients with an appropriate clinical and exposure history, even in an area in which TPE is not endemic. Individuals from regions in which lymphatic filariasis is endemic who present with dyspnea, cough, wheeze, and persistent eosinophilia should undergoing tests to measure serum IgE and antifilarial antibody levels, in addition to serological testing for Strongyloides stercoralis and stool examination for other helminthes. Individuals with TPE often receive an incorrect diagnosis of asthma, despite their poor response to standard treatments for asthma. Early recognition and treatment of TPE with diethylcarbamazine is key to minimizing morbidity and mortality, because chronic untreated TPE may lead to progressive, irreversible pulmonary fibrosis. Acknowledgments Financial support. Canadian Institutes of Health Research (grant MT ; to K.C.K.) and a Canada Research Chair (to K.C.K.). Conflict of interest. All authors: No conflict. References 1. Lymphatic filariasis: the disease and its control. Fifth report of the WHO Expert Committee on Filariasis. World Health Organ Tech Rep Ser 1992; 821: Lymphatic filariasis: diagnosis and pathogenesis. WHO Expert Committee on Filariasis. Bull World Health Organ 1993; 71: Michael E, Bundy DAP. Global mapping of lymphatic filariasis. Parasitol Today 1997; 13: Molyneux DH, Zagaria N. Lymphatic filariasis elimination: progress in global programme development. Ann Trop Med Parasitol 2002; 96: S Lobos E, Nutman T, Hothersall JS, Moncada S. Elevated immunoglobulin E against recombinant Brugia malayi g-glutamyl transpeptidase in patients with bancroftian filariasis: association with tropical pulmonary eosinophilia or putative immunity. Infect Immun 2003; 71: Choudhri SH, Wong W, Plourde PJ, Lertzman M. Tropical pulmonary eosinophilia in a 63-year-old woman from Guyana. CMAJ 1993; 148: Pelloux H, Moro D, Rejasse C, Ambroise-Thomas P. A case of tropical pulmonary eosinophilia. Clin Infect Dis 1994; 19: Panosian CB, Herman J, Ponsillo MA. A Sri Lankan man with eosinophilia. Rev Infect Dis 1991; 13: McKeeham FR. Tropical eosinophilia in an American college: a report on two cases. J Am Coll Health Assoc 1975; 24: Ong RKC, Doyle RL. Tropical pulmonary eosinophilia. Chest 1998; 113: Obaray A, Khan F, Azueta V, et al. Tropical eosinophilia presenting as acute bronchial asthma: case report with clinical, physiologic, and histologic features before and after treatment. Heart Lung 1982; 11: Jiva TM, Israel RH, Poe RH. Tropical pulmonary eosinophilia masquerading as acute bronchial asthma. Respiration 1996; 63: Jones DA, Pillai DK, Rathbone BJ, et al. Persisting asthma in tropical pulmonary eosinophilia. Thorax 1983; 38: Rom WN, Vijayan VK, Cornelius MJ, et al. Persistent lower respiratory tract inflammation associated with interstitial lung disease in patients with tropical pulmonary eosinophilia following conventional treatment with diethylcarbamazine. Ann Rev Respir Dis 1990; 142: Meeker DP. Pulmonary infiltrates and eosinophilia revisited. Cleve Clin J Med 1989; 56: Ottesen EA, Nutman TB. Tropical pulmonary eosinophilia. Annu Rev Med 1992; 43: Udwaida FE. Tropical eosinophilia. In: Herzog H, ed. Pulmonary eosinophilia: progress in respiration research. Vol 7. Basel, Switzerland: S. Karger, 1975: Rocha A, Dreyer G, Poindexter RW, Ottesen EA. Syndrome resembling tropical pulmonary eosinophilia but of non-filarial aetiology: serological findings with filarial antigens. Trans R Soc Trop Med Hyg 1995; 89: O Bryan L, Pinkston P, Kumaraswami V, et al. Localized eosinophil degranulation mediates disease in tropical pulmonary eosinophilia. Infect Immun 2003; 71: Aldebert D, Lamkhioued B, Desaint C, et al. Eosinophils express a functional receptor for interferon a: inhibitory role of interferon a on the release of mediators. Blood 1996; 87: Fruehauf S, Fiehn C, Haas R, et al. Sustained remission of idiopathic hypereosinophilic syndrome following a interferon therapy. Acta Haematol 1993; 89: Mehlotra RK, Hall LR, Haxhiu MA, Pearlman E. Reciprocal immunomodulatory effects of gamma interferon and interleukin-4 on filariainduced airway hyperresponsiveness. Infect Immun 2001; 69: Burchard EG, Ziv E, Coyle N, et al. The importance of race and ethnic background in biomedical research and clinical practice. N Eng J Med 2003; 348: Hoerauf A, Kruse S, Brattig N, et al. The variant Arg110Gln of human IL-13 is associated with an immunologically hyper-reactive form of onchocerciasis (sowda). Microbes Infect 2002; 4: Teo SK. Tropical pulmonary eosinophilia. Singapore Med J 1991; 32: Choi EH, Zimmerman PA, Foster CB, et al. Genetic polymorphisms in molecules of innate immunity and susceptibility to infection with Wuchereria bancrofti in South India. Genes Immun 2001; 2: CID 2004:39 (15 October) Boggild et al.