CASE REPORT

C

hronic Eosinophilic Pneumonia Due to Visceral Larva MigransKoji Inoue, Yoshikazu Inoue, TomArai, Yukifumi Nawa*,YozoKashiwa,

Satoru Yamamotoand Mitsunori Sakatani

AbstractA 38-year-old womanpresented with worsening cough,

blood eosinophilia, and pulmonary infiltrates. Bronchoal-veolar lavage showed 96.4% eosinophils. The diagnosis ofvisceral larva migrans (VLM) was made based on the posi-tive results in enzyme-linked immunosorbent assay forToxocara canis together with clinical symptoms and labo-ratory data. Pulmonary infiltrates due to VLM generallymanifest as a transient form of Lqffler's syndrome or simpleeosinophilic pneumonia mainly in children. Here we reportan adult case of VLM, with pulmonary infiltrates patho-logically proven to be eosinophilic pneumonia, which per-sisted for 7 weeks before anthelmintic treatment withalbendazole and manifested as chronic eosinophilic pneu-

monia.(Internal Medicine 41 : 478-482, 2002)

K

ey words: Toxocara canis, pulmonary infiltrates, enzyme-linked immunosorbent assay, albendazole

Introduction

"Visceral Larva migrans" (VLM) is a term described byBeaver et al in 1952 (1). VLM results mainly from infestationby the commonroundworms of dogs and cats, Toxocara canis(T. canis ) and T. cati, respectively. A large proportion of dogsare infected, and the soil of public playgrounds is often con-taminated with eggs of Toxocara. VLMoccurs in all temperateand tropical areas of the world; it is found predominately inchildren since they frequently play with dogs or cats and some-times eat with contaminated hands after playing in the sandbox of a public playground. VLM can also occur in adults al-

t

hough their route of infection remains unclear.VLMbegins by ingesting embryonated Toxocara eggs. The

larvae which hatch from the eggs in the intestine invade intothe intestinal wall and disseminate via the portal blood stream

to the liver, and then to various organs, including the brain,heart, and lungs. Lung involvement is commonand manifestsas pulmonary infiltrates. Their presumptive mechanism is anallergic response to the larva in the lungs; eosinophilic infil-trates into the lungs. Most larvae seem able to remain dormantfor manyyears. Eventually some larvae are encapsulated anddestroyed by a host response, while others are seemingly pro-tected by being walled off (2). Since humans are not suitablehosts for Toxocaralarvae, they are unable to reach the intes-tine to lay eggs. Therefore, as the definition, the eggs or thematureadults are never seen in VLM. The definite diagnosisof VLMcanbe made only by proving the larvae. However, thismeans of the diagnosis is considerably unrealistic since thelarvae are only 0.02 mmwhich is small a size to be obtainedfrom extensive pulmonary infiltrates or other affected organs.Forthat reason, practically, the clinical diagnosis has been madeby enzyme-linked immunosorbent assay (ELISA) using anantigen of T. earns. This method has been reported to have a78% sensitivity and 92% specificity (3).

Pulmonary infiltrates due to VLM generally manifest as atransient form of Lqffler's syndrome (4) or simple eosinophilicpneumonia (5), and also have been reported to present as anacute form (6). And their histopathological changes may beeosinophilic pneumonia. This has been, however, little provenin the literature. Here we report an adult case of VLM, withpulmonary infiltrates histopathologically proven to be eosino-philic pneumonia, which did not make a spontaneous resolu-tion within 1 month, persisted for 7 weeks before treatment,and manifested as chronic eosinophilic pneumonia (7).

Case Report

A38-year-old womanhad had non-productive cough for 5months and went to another hospital on December 5, 2000. Aradiograph of the chest showed small pulmonary infiltrates inthe right upper field. Codein was administered. She was ad-mitted to the hospital because of worsening cough and pulmo-nary infiltrates on December 28.

On admission to the hospital, a radiograph of the chest

F

rom the Department of Internal Medicine and Department of Pathology, National Kinki-Chuo Hospital for Chest Diseases, Osaka and *Department ofParasitology, Miyazaki Medical College, Miyazaki

Received for publication October 12, 2001 ; Accepted for publication January 21 , 2002

R

eprint requests should be addressed to Dr. Yoshikazu Inoue, the Department of Internal Medicine, National Kinki-Chuo Hospital for Chest Diseases, 8 1 10,Nagasone, Sakai, Osaka 591-8555

478 InternalMedicineVol. 41, No. 6 (June 2002)

Chronic EP Due to Visceral Larva Migrans

showed pulmonary infiltrates in the bilateral upper fields, andhematologic tests revealed leukocytosis and eosinophilia. Hercondition did not improve despite antibiotics. The patient wasreferred to our hospital on January 9, 2001. Except for subacute thyroiditis three years before referral,the patient had been well and never had had asthma or atopicdiseases until she had developed non-productive cough. Shehad lived in an urban area of Osaka, Japan. She had worked ata factory, producing videocassette recorders under industrialdust-free conditions. She had never smoked. She had neverowned a dog nor any pet. She did not remember eating foodcontaminated with soil that could have contained eggs of Toxo-

cara. On admission, the body temperature was 36.6 degrees Cel-sius, the pulse was 76 per minute, and the respirations were 10per minute. The blood pressure was 1 10/70 rnmHg. By auscul-tation, crackles were heard over the upper lungs.

Hematologic tests showed a hematocrit of 3 1.6%, erythrocytesedimentation rate of 83.6 mm/h, platelet count of 326,000/mm3,and white blood cell count of 14,500/mm3 with 58.5% eosino-phils, 23.1% neutrophils, and 14.0% lymphocytes. Analysesof blood chemiatry and enzymes were as follows: urea nitro-gen of 7 mg/dl, creatinine of 0.5 mg/dl, lactate dehydrogenase

of 478 U/Z, aspatate aminotransferase of 16 U/Z, alanine ami-notransferase of 1 1 U/Z, bilirubin of 0.2 mg/dl. Immunologic tests showed an immunoglobulin G (IgG) of1,271 mg/dl, IgA of 156 mg/dl, IgM of 182 mg/dl, and IgE of438 mg/dl. Neither MPO-ANCA nor PR3-ANCA was detectedin her serum. Analysis of arterial blood gas under breathingroom air showed a PaO2 of 81.8 mmHg and PaCO2 of 38.3mmHg. An electrocardiogram revealed no abnormalities. Atuberculin test was negative. A radiogram of the chest on referral to our hospital showedpulmonary infiltrates in the left upper and middle fields (Fig.1 A). High-resolution computed tomography on referral revealedperipheral consolidation and ground-glass opacities in the leftupper and middle fields (Fig. 2). In addition, a chest radiogramjust before treatment on January 25 showed migration of pul-monary infiltrates from the upper and middle fields to the middleand lower fields. Analysis of bronchoalveolar lavage fluid(BALF), obtained from the lingual, showed that its total cellcount was 38.4xlO5/ml with 96.4% eosinophils, 0.8% macroph-ages, 1.8% lymphocytes, and 1.0% mast cells. Specimens oftransbronchial lung biopsy (TBLB) demonstrated eosinophilicpneumonia (Fig. 3). No larvae were identified in the BALF or TBLB specimens.

A

B

Figure 1. A) Chest X-ray on referral to our hospital, showing infiltrates in the left upper and middle field. B) Chest X-ray on the21st day of treatment, demonstrating no evidence of infiltrates.

Internal Medicine Vol. 41, No. 6 (June 2002) 479

Inoue et al

Figure 2. High-resolution computed tomography at the leftmiddle level, on referral to our hospital, revealing peripheralconsolidation and ground-glass opacities.

Figure 3. A transbronchial lung biopsy specimen, demonstrat-ing eosinophilic infiltrates in the alveolar walls and spaces (HEstain, x200).

There were no eggs in her stool. In her serum, anti-IgG to T.cards was detected by a commercial multiple-dot ELISA kit(SRL, Tokyo), which contains 12 different parasite antigensincluding T. canis. The positive results of anti-IgG to T. caniswere further confirmed by microplate ELISA using an excre-tory-secretory antigen derived from T. canis larvae, at the De-partment of Parasitology, Miyazaki Medical College, Japan.This test revealed an A405of 0.35 (cut-off, 0.2). Based on thepositive results of anti-IgG to T. canis and clinical findings,we diagnosed this case as CEP due to VLM.As anthelmintic treatment, albendazole (600 mg per day)was administered on January 25. Her cough had lasted 7 monthsand dissipated 2 weeks after the treatment. Moreover her pul-monary infiltrates that had persisted for 7 weeks cleared 3 weeksafter the treatment (Fig. IB). In addition, her elevated whiteblood cells, blood eosinophils and IgE values fell to within thenormal level, 2 weeks, 6 weeks, 1 week after the treatment,respectively (Table 1 ). Albendazole was continued for 8 weeks.During the 8 weeks of the treatment, slight skin eruptions weretransiently developed as a mild adverse effect.

Discussion

Patients with VLMmost commonlypresent with pulmo-nary symptoms. In a series of 51 children cases with VLM,cough occurred in 80% cases, wheezing in 63%, and fever in

80% (8). The present adult patient presented with cough last-

i

ng 6.5 months.

In a series of 17 children cases with VLM, pulmonary infil-trates occurred in 42% cases (9). Pulmonary infiltrates due toVLM are presumed to be an allergic response to the larvaemigrating through the pulmonary circulation; eosinophilic in-filtrates into the lungs. Therefore these pulmonary infiltratesmay be histopathologicaUy consistent with eosinophilic pneu-monia. This has been, however little proven in the literature. Inthe present case, transbronchial lung biopsy demonstrated thatits pulmonary infiltrates were an eosinophilic pneumonia. Itappears that the TBLB-proven eosinophilic pneumonia of ourcase indicates that pulmonary infiltrates due to VLM are histo-

p

athologically compatible with eosinophilic pneumonia.

In 1932, Loffler clinically described "Lqffler's syndrome"in which pulmonary infiltrates were transient and cleared within6 to 12 days (4). Nowadays, most cases of Loffler's syndromeare thought to be due to parasitic etiology, especially VLM. In

Table 1. Hematologic Laboratory Values

F irst 7 th 1 4 th 2 1th 4 2th

treatm ent tre atm ent treatm ent treatm ent treatm en t

d ay d ay d ay d ay day

WBC (/m m 3) 10.600 9.7 00 7.300 7.200 6 .100

E o sin oph ils (% ) 50.8 35.8 22.5 9.9 6.0

IgE * (m g/dl) 4 38 2 28 14 5 117 9 8

WBC: white blood cells. *The normal range for IgE is 15 to 400 mg per deciliter.

480 Internal Medicine Vol. 41 , No. 6 (June 2002)

Chronic EP Due to Visceral Larva Migrans

1952, Crofton et al clinically described "simple eosinophilicpneumonia''' (SEP) with an increased duration to spontaneousresolution up to 1 month (5). Pulmonary infiltrates due to VLMgenerally manifest as a transient form of Loffler's syndrome orSEP, and also have been reported to present as an acute form(6). In 1969, Carrington et al histopathologically described 9cases of "chronic eosinophilic pneumonia" (CEP) which werecontrasted with both Loffler's syndrome and SEP (7). In theirstudy, 8 of the 9 cases were idiopathic. The other case wasprobably caused by a drug. Their study did not precisely men-tion how long pulmonary infiltrates due to CEP persisted be-fore treatment with corticosteroids. We assume that, however,pulmonary infiltrates due to eosinophilic pneumonia whichpersist over 1 month may less frequently make a spontaneousresolution. In the present case, pulmonary infiltrates due toVLM, pathologically proven to be eosinophilic pneumonia, didnot make a spontaneous resolution within 1 month and per-sisted for 7 weeks before treatment, which suggests that theeosinophilic pneumonia of our case manifests as CEP.

In the series of 51 children cases with VLM, leukocytosiswas seen in 100% with blood eosinophilia ranging from 30 to82% eosinophils (8). On the other hand, in the review of 105cases with CEP, leukocytosis was seen in 63% with a meanvalue of 13,360/mm3 ranging from 5,800 to 30,900/mm3, andblood eosinophilia of greater than 6% was seen in 88% with amean value of 26% ranging from 1 to 90% (10). In the presentcase, hematologic tests showed leukocytosis of 14,500/mm3with 58.5% eosinophils, which makes no difference betweenVLM and CEP.

In a series of 5 patients with CEP, a range of bronchoalveolarlavage (BAL) eosinophilia from 14 to 75% has been reported(ll). On the other hand, in a patient with VLM, BAL has beenreported to show marked eosinophilia of 64% (6). In the presentcase, BAL showed 38.4xlO5 cells/ml, which is about 38 timeshigher than in healthy subjects, with 96.4% eosinophils. Theremarkable BAL eosinophilia in our case suggests that pulmo-nary infiltrates in VLM are attributable to eosinophils.

In the presnt case, pulmonary infiltrates on chest X-ray wereseen as "the photographic negative of pulmonary edema" ( 12)and they had migrated. Migration of pulmonary infiltratesindicats parasitic etiology but can occur in CEP (13). In ourcase, high-resolution computed tomography revealed consoli-dation and ground-glass opacities in peripheral distribution,indistinguishable from those in CEP (14). VLM is generally self-limited, and the use of anthelminticagents is controversial. We assume that, however, either if thelarvae have not been killed by a human host response and keepmigrating through the lungs or if the larvae have become dor-mant in the lungs but cause overreaction of the human host,pulmonary infiltrates due to VLM may not make a spontane-ous resolution and may persist for a long time. In the presentcase, as the pulmonary infiltrates persisted for 7 weeks we pre-sumed that spontaneous resolution would not occur. Actually,our patient had suffered from chronic cough despite codein.On the other hand, her pulmonary infiltrates cleared 3 weeksafter anthelmintic treatment with albendazole (600 mg per day

for 8 weeks). Moreover, her cough lasting 7 months dissipated2 weeks after the treatment. In addition, her elevated whiteblood cells, blood eosinophils and IgE values were normalizedimmediately after the treatment. In the present study, antibodytiter in the sera of the patient remained unchanged even afterthe successful treatment. We speculate that this might be dueto a gradual release of antigenic components from the destroyedlarvae. During the treatment, skin eruptions were occasionallyrecognized and thought to be a side effect. They were, how-ever not disabling in our patient at all. Hepatic and renal func-tion tests remained normal throughout the course of treatmentand afterwards. In a series of three children with VLM, theydemonstrated rapid clinical and laboratory improvement afterthe use of the anthelmintic drug, thiabendazole (15). We be-lieve that anthelmintic treatment is effective with the view to-wards relieving symptoms and decreasing convalescent time.Nevertheless, it should be determined in a trial whether or notthe use of anthelmintic agents is efficacious.

The case reported here seems to be the first case of CEP dueto VLM. The presence of such a case strongly suggests thatsome cases of idiopathic CEP might include CEP due to VLM.Since immunodiagnosis for VLM can be made without diffi-culty, CEP due to VLM should be excluded before diagnosedas idiopathic CEP.

Refere nces

1) Beaver PC, Snyder CH, Carrera GM, Dent JH, Lafferty JW. Chronic eo- sinophilia due to visceral larva migrans. Report of three cases. Pediatrics 9: 7-19, 1952. 2) Glickman LT, Schantz PM. Epidemiology and pathogenesis of zoonotic toxocariasis. Epidemiol Rev 3: 230-250, 1981. 3) Glickman L, Schantz P, Dombroske R, Cypess R. Evaluation of serodiag- nostic tests for visceral larva migrans. Am J Trop Med Hyg 27: 492-498,

1978.

4) Loffler W. Zur Differential-Diagnose der Lungeninfiltrierungen II. Uber fliichtige Succedan-Infiltrate (mit Eosinophilie) Beitr. z. Klin. d. Tuberk.

79: 368-382, 1932.

5) Crofton JW, Livingstone JL, Oswald NC, Roberts ATM. Pulmonary eo- sinophilia. Thorax 7: 1-35, 1952. 6) Roig J, Romeu J, Riera C, Texido A, Domingo C, Morera J. Acute eo- sinophilie pneumonia due to toxocariasis with bronchoalveolar lavage findings. Chest 102: 294-296, 1992. 7) Carrington CB, Addington WW, GoffAM, et al. Chronic eosinophilie pneumonia. N Engl J Med 280: 787-798, 1969. 8) Huntley CC, Costas MC, Lyerly A. Visceral larva migrans syndrome: clinical characteristics and immunologic studies in 5 1 patients. Pediatrics 36: 523-536, 1965.

9) Snyder CH. Visceral larva migrans. Ten-year's experience. Pediatrics 28: 85-91, 1961.10) Jederlinic PJ, Sicilian L, Gaensler EA. Chronic eosinophilie pneumonia.

A report of 19 cases and a review of the literature. Medicine (Baltimore) 67: 154-162, 1988.1 1) Dejaegher P. Demedts M. Bronchoalveolar lavage in eosinophilie pneu-

monia before and during corticosteroid therapy. Am Rev Respir Dis 129: 631-632, 1984.1 2) Gaensler EA, Carrington CB. Peripheral opacities in chronic eosinophilie

pneumonia: the photographic negative of pulmonary edema. Am J Roentgenol 128: 1-13, 1977.1 3) Marchand E, Reynaud-Gaubert M, Lauque D, et al. Chronic eosinophilie

pneumonia (CEP). A clinical and follow-up study of 61 cases. Am J Respir

Internal Medicine Vol. 41, No. 6 (June 2002) 481

Inoue et al

Crit Care Med 155: A329, 1997. 737-744, 1994.14) Ebara H, Ikezoe J, Johkoh T, et al. Chronic eosinophilic pneumonia: evo- 15) Aur RJ, Pratt CB, Johnson WW.Thiabendazole in visceral larva migrans.

lution of chest radiograms and CT features. J comput Assist Tomogr 18: Am J Dis Child 121: 226-229, 1971.

482 Internal Medicine Vol. 41, No. 6 (June 2002)