Background: Two previously healthy soldiers presented with rapidly evolving adult respiratory distress syndrome (ARDS) within days of their return from the National Training Center at Fort Irwin, California. There have been six similar cases reported in the National Training Center area, five with fatal outcomes. Methods: Both patients were treated in an intensive care unit setting. Extensive diagnostic testing was performed, and patients were treated empirically with doxycycline, ceftriaxone, and ribavirin. Results: Both patients had complete recovery. Intensive care unit course was notable for circulatory collapse, respiratory distress requiring ventilatory support, -10% eosinophils on bronchoalveolar lavage. and dramatic peripheral eosinophilia. All cultures, including serology cultures for hantavirus, were negative. Conclusion: There have been no previously described syndromes comprising "viral" prodrome, rapid progression to ARDS, and eosinophilia. Based on the geographic, temporal, and clinical similarity of these two cases, we believe that they share a common, yet undetermined, cause. Both environmental exposures and any undescribed pathogens warrant further investigation.
This report describes the clinical courses and laboratory findings of a rapidly progressive adult respiratory distress syndrome (ARDS) in two previously healthy individuals whose only connection was training at the Army's National Training Center at Fort Irwin, California, during the month of June 1997. Additional cases similar to these had been reported within the Army medical system, but these have not been as extensively investigated or published (L. Stapleton, personal communication). Those cases had fatal outcomes. Also of interest are two similar cases reported in a nearby civilian community at about the same time that these soldiers presented, one of which had a fatal outcome. I The severity and rapid progression of illness, as well as the geographic and temporal clustering of the cases, prompted us to describe the syndrome. We suggest rapid evaluation and treatment of patients reporting recent travel to this area and symptoms suggestive of a viral syndrome.
Patient 1 was a 19-year-old male active duty Army soldier with no significant past@medical history. He initially presented to his troop physician assistant complaining of 18 hours of nausea, vomiting. diarrhea, headache, and chest tightness. He had no cough or sputum production. His travel history was remarkable for a recent 30-day tour of duty at the National Training Center at Fort Irwin (Fig. 1). He had no known exposures to insects, rodents, or chemicals. He had no pets. He denied any risk factors for human immunodeficiency virus (HM and was negative for HIV by enzyme-linked immunosorbent assay in August 1996. His duties at the National Training Center included driving trucks and field training in the desert. Soldiers slept on cots in tents without floors. The weather during the training session was hot and dry, and the patient reported having inhaled "lots of dust." He reported no ill contacts. A tentative diagnosis of a viral syndrome was made, and he was given prochlorperazine and guaifenesin and released. Twelve hours after his initial presentation, he presented to the emergency department at Madigan Army Medical Center (MAMC) with severe dyspnea, chest tightness, and headache. Initial vital signs were temperature 38.3'C, blood pressure 127/62 mm Hg, pulse 159, respiration 28, and oxygen saturation 96% on 2 L by nasal cannula. Chest X-ray showed multilobar interstitial infiltrates (Fig. 2A). Physical examination was notable for diffuse expiratory wheezes throughout all lung fields. Initial peak flow was 150 L/min. Initial laboratory data were remarkable for hemoconcentration and leukocytosis with left shift (Table 1).
Treatment was initiated for presumed community-acquired pneumonia and dehydration. Initial treatment included normal saline fluid resuscitation, intravenous (IV) erythromycin 500 mg every 6 hours, IV ceftriaxone I g every 24 hours, and albuterol nebulizer every 4 hours. The patient's condition worsened overnight, with remarkable tachypnea. Vital signs were temperature 37.4 deg C, blood pressure 126/57 mm Hg, pulse 124, and respiration 48. His chest X-ray progressed to diffuse interstitial infiltrates involving all lobes, consistent with ARDS (Fig. 213). He was transferred to the intensive care unit (ICU) where, after a brief trial of noninvasive ventilation, he was semielectively intubated.
Fiber-optic bronchoscopy with bronchoalveolar lavage (BAL) was performed. The fluid returned clear and yellow, with the appearance of dilute serum. Cellularity was modest but remarkable for 10% eosinophils. Gram's stain, fungal smear, respiratory syncytial virus (RSV) by direct fluorescent antigen, and silver stain for Pneumocystis carinii (PCP) were negative (Table 1).
Because the patient complained of headache, lumbar puncture was performed to help rule out disseminated coccidioidomycosis. India ink, latex agglutinins, and cultures were all negative. The fluid was clear and had normal glucose and protein content, with only 2 leukocytes/mm3.
A Swan-Ganz catheter was placed for hemodynamic monitoring. The patient was initially hyperdynamic, with a cardiac index of 5.6 L/min/M^sup 2^ and pulmonary artery occlusion pressure (PAOP) of 10 min Hg, consistent with early sepsis. He required heavy sedation and paralysis to decrease work of breathing and oxygen demand. He required large volumes of crystalloids as well as vasopressor support to maintain systolic blood pressure > 100 mm Hg, sequestering > 5 L in the first 24 hours. His core temperature increased to >40 deg C, requiring a cooling blanket to keep his temperature
In light of the rapid progression to ARDS in this previously healthy young man, our differential diagnosis included hantavirus pulmonary syndrome (HPS), plague, Chlamydia psittacis strain (TWAR), mycoplasma, other community-acquired pneumonias, and fungal infection (coccidioidomycosis). His treatment regimen was altered to include empiric treatment for HPS with IV ribavirin under an investigational new drug protocol. IV doxycycline was substituted for erythromycin on the second hospital day to provide coverage for plague and TAR. Ceftriaxone was continued. Rapid BAL analysis enabled us to avoid empiric therapy with antifungal agents, trimethoprim-sulfamethoxazole, or corticosteroids.
Aside from circulatory collapse and respiratory distress, there was no other organ involvement.
On ICU day 3, the patient showed marked improvement. He was weaned from vasopressors, ventilatory support was reduced, and he began to spontaneously diurese. On ICU day 6, he was extubated. After extubation, his oxygen requirement was minimal (2 L by nasal cannula). On ICU day 7, he was transferred to the general medicine ward. During the course of his hospitalization, his leukocytosis normalized; however, he showed a marked eosinophilia that peaked at 40% (Fig. 3). All cultures and serology tests were negative. The serology sample was forwarded to the Centers for Disease Control and Prevention for investigation of a possible emerging pathogen.
Patient 2 presented to MANIC on hospital day 5 of patient 1. Patient 2 was a 20-year-old female active duty Army soldier with no significant past medical history. She presented to the MAMC emergency department complaining of shortness of breath first noted as painful breathing while running 1 day before admission. This rapidly progressed to shortness of breath at rest and was associated with fever, chills, myalgias, and nausea without vomiting or diarrhea. Her travel profile was similar to that of patient 1 in that she had recently returned from the National Training Center at Fort Irwin. She visited the same training center as patient 1; however, the two patients were in different units and never came into contact with one another. Her duties and accommodations were similar to those of patient 1. She denied exposure to chemicals, insects, or rodents. She had no ill contacts. She owned a ferret. She denied risk factors for HIV and was negative for HIV by enzyme-linked immunosorbent assay in August 1996. Her initial vital signs were temperature 38. 1 deg C, blood pressure 133/65 mm Hg, pulse 127, respiration 27, and pulse oximetry 96% on room air. Chest X-ray showed diffuse reticulonodular opacities. Physical examination was notable for decreased air movement and scattered wheezes. Initial laboratory data were remarkable for leukocytosis with left shift (Table 1). She was admitted for presumed community-acquired pneumonia and treated empirically with IV doxycycline and ceftriaxone. Her tachypnea persisted, with a respiratory rate of 40. She was transferred to the ICU for closer monitoring.
A Swan-Ganz catheter was placed for hemodynamic monitoring, initially revealing a cardiac index of 4.46 L/min/m^sup 2^ and PAOP of 13 mm Hg. She was semielectively intubated, with subsequent fiber-optic bronchoscopy and BAL. BAL washing contained modest cellularity with 8% eosinophils. Gram's stain, fungal smears, RSV by direct fluorescent antigen, and PCP by silver stain were negative (Table 1).
The patient's hemodynamic profile became even more hyperdynamic, with cardiac index as high as 6.2 L/min/M^sup 2^ . Her filling pressures remained adequate, with PAOP between 9 and 13 min Hg. She sequestered 10 L of crystalloids, 3 L in the first 24 hours. She did not require vasopressor support. She was sedated and paralyzed to decrease oxygen demand. Her core temperature reached >40 deg C, and a cooling blanket was used. On hospital day 3, ribavirin was started empirically because of the concern for HPS. On hospital day 4, ceftriaxone was discontinued because there was no indication of bacterial pneumonia; ribavirin and doxycycline were continued. On hospital day 5, she showed an increased temperature curve along with a new right lower lobe infiltrate. Blood and pleural fluid cultures were sent. The pleural fluid was exudative without organisms on Gram's stain, consistent with an uncomplicated parapneumonic effusion. Mezlocillin and gentamicin were started empirically to cover presumed nosocomial pneumonia.
Aside from respiratory distress and circulatory collapse, there was no other organ involvement.
On hospital day 6, the patient's filling pressures increased to PAOP 19 nun Hg; she was given I dose of IV furosemide to aid diuresis. With diuresis, lung compliance improved and she was weaned from ventilatory support. She was extubated on hospital day 9 and had minimal subsequent oxygen requirement. After extubation, she had a flat affect along with some dyskinetic movements. These resolved with diphenhydramine given as needed. She was transferred to a general medicine ward on hospital day 10. Her leukocytosis decreased throughout her hospitalization; however, her eosinophil count began to increase during the acute phase of her illness and continued to increase after she began to improve, peaking at 24% (Fig. 4). She was discharged to convalescent leave on hospital day 14. As with patient 1, all cultures were negative. A serum sample was forwarded to the Centers for Disease Control and Prevention for investigation of a possible emerging pathogen.
Our experience with these two young patients not only reinforced the necessity of early recognition and prompt ICU treatment of ARDS but also led to concerns regarding an emerging new pathogen. ARDS has been extensively described in the literature, most often in connection with sepsis syndrome, aspiration of gastric contents. inhaled toxins, or metabolic derangements. In a significant number of patients with ARDS, an inciting factor will never be found. However, in cases such as those described here, with a discrete geographical clustering, high associated mortality (five deaths out of six similar cases that were not extensively investigated), and very brief prodromal phase, we feel that further diagnostic investigation is warranted until an etiologic agent is found. The two patients in our series were evaluated extensively for any likely inciting agent based on their clinical presentation and known environmental exposures.
The clinical presentation common to both of our patients. including nonspecific viral symptoms (myalgias, nausea, vomiting, diarrhea, headache), fever, shortness of breath, tachypnea, tachycardia, and subsequent rapid progression to ARDS, was suggestive of hantavirus pulmonary syndrome. HPS is a well-described syndrome that progresses through three distinct phases: a viral prodromal phase; an acute cardiopulmonary phase; and, for those who survive, a convalescent phase. Laboratory abnormalities include leukocytosis with left shift, atypical lymphocytes, hemoconcentration (hematocrit > 48%), thrombocytopenia, lactic acidosis, elevated low-density lipoprotein, elevated aspartate aminotransferase, prolonged prothrombin time, and occasionally mild proteinuria and/or hematuria. Treatment for HPS has been largely supportive, including mechanical ventilation, vasopressor support, and judicious use of IV fluids. Intravenous ribavirin has proven beneficial in decreasing morbidity and mortality with Hantaan virus, the hantavirus responsible for hemorrhagic fever with renal syndrome. Although initial studies have not proven ribavirin to be beneficial in HPS,1 ribavirin is available for this use as an investigational new drug. Both of our patients were treated empirically with ribavirin until negative serology results for hantavirus were obtained. Each patient received 4 days of treatment.
Other acute pulmonary syndromes that warranted consideration and empiric treatment included community-acquired pneumonia, atypical bacterial pneumonias (Mycoplasma pneumoniae, Chlamydia species, Legionella, pneumonic plague, tularemia, and leptospirosis. These organisms were initially empirically covered with ceftriaxone and doxycycline, Southern California is endemic for coccidioidomycosis; however, early BAL failed to demonstrate any evidence of this or any other fungal infection. Convalescent titers by complement fixation were also negative for coccidioidomycosis. Early BAL also ruled out PCP and RSV in both cases. Other organisms such as Coxiella burnetti and Bacillus anthracis were felt to be less likely in the absence of a history of animal exposure. In addition, profound leukocytosis such as was seen in our patients is not common in Q fever or anthrax, Viral pneumonias, including adenovirus, influenza, parainfluenza, and measles, were also considered in the original differential diagnosis; however, early treatment options for these pneumonias are largely supportive. The fact that these patients were young active duty Army soldiers without known premorbid conditions raises the suspicion of chemical exposures or perhaps environmental toxins, although no history of such exposures was elicited.
Acute eosinophilic pneumonia (AEP) was also in our differential diagnosis. AEP has been extensively reported in the literature and has been defined by the following criteria: an acute febrile illness of less than 5 days duration with hypoxemic respiratory failure, diffuse alveolar or mixed alveolar-interstitial chest X-ray infiltrates, BAL eosinophils > 25%, absence of parasitic, fungal, or other infection, and prompt, complete, and nonrelapsing response to corticosteroids.1 The cause of AEP is unknown; however, it has been suggested that this syndrome results from an acute hypersensitivity phenomenon to an unidentified inhaled antigen .4 Once the diagnosis of AEP has been established, the recommended treatment is corticosteroids," although improvement with supportive care alone has also been reported.7 AEP is a diagnosis of exclusion after infectious causes have been definitively ruled out. This point is exemplified in the two cases of eosinophilic lung disease described by Lombard et aL, in which the ultimate diagnosis was Coccidioides immitis infection. One of these patients had a fatal outcome after receiving corticosteroids and later Cytoxan for a presumed hypersensitivity reaction8 Additionally, Whitlock and colleagues reported a case of a 41-year-old male with fever and diffuse pulmonary infiltrates. This patient's BAL fluid showed 36% eosinophils. Stains for fungi were negative, but on day 12 a culture of the BAL fluid grew Coccidioides immitis.9 The patient improved with supportive care over a 7-day period. Our patients differed from those with AEP as described above in that they had
BAL is commonly used as a means of rapidly diagnosing and/or ruling out various pulmonary pathogens, including bacterial, viral, and fungal pneumonia, PCR cytomegalovirus, eosinophilic granulomatosis, Mycobacterium tuberculosis, diffuse alveolar hemorrhage, amiodarone-induced lung injury, cytotoxic lung injury, and alveolar proteinosis. Characteristic findings on BAL also help to narrow the differential diagnosis in conditions such as sarcoidosis, hypersensitivity pneumonitis, collagen vascular disease, idiopathic pulmonary fibrosis, and acute or chronic eosinophilic pneumonia. Eosinophilia was seen initially in the bronchial washings of both patients in our series. Peripheral eosinophilia was not seen until well into the acute illness, peaking at 40% in patient I and 25% in patient 2 as the patients recovered, then subsequently normalizing (Figs. 3 and 4). In 1990, Allen et al. examined the diagnostic significance of elevated eosinophils on BAL. They analyzed 1,059 consecutive BAL samples and found that only 48 had >5% eosinophils. The most common diagnoses found in that series of 48 patients with >5% BAL eosinophils were interstitial lung disease (5-51% eosinophils), followed by acquired immunodeficiency syndromeassociated pneumonia (5-43% eosinophils), idiopathic eosinophilic lung disease (21-57% eosinophils), and drug-induced lung injury (5-32% eosinophils) (Table 11). 10
Only two syndromes involving ARDS and eosinophilia have been reported. A single case of ARDS with elevated neutrophils, lymphocytes, and eosinophils on BAL was reported after treatment with a Chinese medicine, Kamisyoyo-san, for seborrheic dermatitis.11 In a retrospective chart review of 20 cases of pulmonary strongyloidiasis, nine patients (45%) developed ARDS, and of these nine patients, eight (89%) had secondary bacterial pneumonias or Gram-negative bacterial sepsis. There was a statistically significant association between steroid use and the development of ARDS in patients with pulmonary strongyloidiasis. All of these ARDS patients had one or more risk factors for severe Strongyloides infection. These include chronic lung disease, age > 65 years, diseases associated with altered cellular immunity, chronic renal failure, alcohol abuse, surgically created intestinal blind loops, and conditions associated with decreased gastric acidity. 12 We have no reason to suspect that our patients were exposed to either Kamisyoyo-san or Strongyloides, which is not endemic to southern California. No other previously described syndromes are characterized by a brief viral prodrome, rapid progression to ARDS, and marked eosinophilia.
Within the past 18 months, a cluster of eight cases of an acute respiratory syndrome emerged from the same geographic locale. Six of these cases were poorly described; five had mortal outcomes. Based on their very similar travel profile, BAL findings. clinical course, and temporal relationship, we feel that the two cases described above share a common, yet undetermined, cause. The six less well-defined cases may also have been affected by the same disease entity. These cases occurred independently, arguing against person-to-person transmission. The only known exposure common to both of our patients is dust.
The rapid clinical progression of this syndrome in previously healthy individuals is perhaps its most devastating feature. We believe that prompt recognition of this progression and rapid escalation of care to an ICU setting afforded our patients a favorable outcome. We feel that patients who have been in southern California and present with viral syndromes with marked tachypnea, leukocytosis with left shift, and fever are at risk of rapid progression to ARDS. Our recommendations for treating such patients include early ICU admission, institution of ventilatory and hemodynamic support, broad-spectrum antibiotics, and early bronchoscopy with BAL. BAL is most helpful in reducing the number of empiric therapies used by eliminating some pathogens from consideration. We recommend empiric ribavirin therapy until serology results for hantavirus can be obtained.
Because the syndrome described here includes features of septic shock and ARDS, as well as features of acute eosinophilic pneumonia, the empiric use of corticosteroids seems at best of equivocal value, and we do not recommend this. If these cases represent a form of acute eosinophilic pneumonia, the recoveries we observed without steroids suggest that this treatment may not be necessary. For similar cases without a diagnosis, we urge that serum samples be sent to the Centers for Disease Control and Prevention to help identify the possible emergence of a novel pathogen.
I .Associated Press: Investigators check to see if hantavirus killed woman. Los Angeles Times, June 24, 1997, p 4.
2. Moolenaar R Brennan R, Peters CJ: Hantavirus pulmonary syndrome. Semin Respir Infect 1997; 12: 31-9.
3. Allen J. Davis W: Eosinophilic lung diseases. Am J Respir Crit Care Med 1994: 150:1423-&
4. Badesch D, King T. Schwarz M: Acute eosinophilic pneumonia: a hypersensitivity phenomenon? Am Rev Respir Dis 1989: 139: 249 -52.
5. Allen J, Pacht E, Gadek J, Davis W: Acute eosinophilic pneumonia as a reversible cause of noninfectious respiratory failure. N Eng]. J Med 1989; 321: 569-74.
6. Pope-Harman A, Davis W, Allen E, Christoforidis A. Allen J: Acute eosinophilic pneumonia: a summary of 15 cases and review of the literature. Medicine 1996: 75: 334-42@
7. Hogan T. Riley R, Thomas J: Rapid diagnosis of acute eosinophilic pneumonia in a patient with respiratory failure using bronchoalveolar lavage with calcofluor white staining. J Clin Lab Anal 1997: 11: 202-7.
8. Lombard C. Tazelaar H, Krasne D: Pulmonary eosinophilia in coccidioidal infections. Chest 1987; 91: 734-6.
9. Whitlock W, Dietrich R. Tenholder M: Correspondence to the editor. N Engl J Med 1990;322@635@
10. Allen J, Davis WB, Pacht E: Diagnostic significance of increased bronchoalveolar lavage fluid eosinophils. Am Rev Respir Dis 1990; 142: 642-7.
11. Shiota Y, Wilson JG, Matsumoto H, Munemasa M, Okamura M, Hiyama J, Marukawa M, Ono T, Taniyama K Mashiba H: Adult respiratory distress syndrome induced by a Chinese medicine, Kamisyoyo-san. Intern Med 1996; 35: 494-6.
12. Woodring J, Halfhill H. Berger R. Reed J. Moser N: Clinical and imaging features of pulmonary strongyloidiasis. South Med J 1996; 89: 10 -9.
Guarantor. LTC George N. Giacoppe, MC USA
Contributors: CPT Douglas A. Degler, MC USA; LTC George N. Giacoppe, MC USA
Madigan Army Medical Center, Fort Lewis, WA 98431.
Presented at the National Army American College of Physicians Annual Meeting, Reston, VA, November 1997; the Madigan Army Medical Center Clinical Vignettes, Fort Lewis, WA. November 1997; and the Washington State Chapter of the American College of Physicians Annual Meeting, December 1997.
This manuscript was received for review in June 1998. The revised manuscript was accepted for publication in January 1999.
Copyright Association of Military Surgeons of the United States Dec 1999
Provided by ProQuest Information and Learning Company. All rights Reserved