PEDIATRICS Vol. 110 No. 3 September 2002, pp. 627-637
EXPERIENCE AND REASON |
Clinical Profile of 30 Infants With Acute Pulmonary Hemorrhage in Cleveland
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSCLINICAL PROFILE OF CASESCASE SERIESDISCUSSIONCONCLUSIONREFERENCES |
|---|
Between 1993 and 2000, 30 infants were hospitalized with acute pulmonary hemorrhage at Rainbow Babies and Children’s Hospital in Cleveland. Most infants presented with severe pulmonary symptoms requiring intensive support, but a few infants had less severe hemorrhage. Three quarters of the patients required ventilator support and blood transfusions. Eleven patients had transitory hemoglobinuria. Five patients died, but infants who survived did well. There are currently no specific treatment modalities, although we have advised moving to a different home and avoiding environmental tobacco smoke. Subsequently, rebleeding from the lower respiratory tract has decreased from 5 of 7 infants to 1 in 21. On the basis of decreased subsequent fatal hemorrhage, high dose glucocorticoids seem to be of some value. Several patients revealed continued low-grade alveolar hemorrhage for months after their initial bleed, even after removal from their original home environments.
Key Words: pulmonary • hemorrhage • infant • home • environment
Abbreviations: BAL, bronchoalveolar lavage • ETS, environmental tobacco smoke
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSCLINICAL PROFILE OF CASESCASE SERIESDISCUSSIONCONCLUSIONREFERENCES |
|---|
In the fall of 1994, an increased incidence and geographic clustering of acute pulmonary hemorrhage among young infants was noticed in Cleveland.1 The Centers for Disease Control and Prevention and the local and state health departments assisted in the investigation of this cluster.2,3 A case-control study and environmental analyses identified an association with fungal contamination of the infants’ homes.4 Informal surveillance5 has yielded >100 cases of unexplained pulmonary hemorrhage in infants reported by pediatricians in other cities in the United States since January 1993, as both sporadic cases and smaller clusters. Several related case reports have appeared.6–9 There have been 41 cases in the Cleveland metropolitan area; 12 of the infants died. Although the initial investigations of this cluster have been reported,2,4,10 the clinical profiles of the 30 patients cared for at Rainbow Babies and Children’s Hospital are described here.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSCLINICAL PROFILE OF CASESCASE SERIESDISCUSSIONCONCLUSIONREFERENCES |
|---|
Case Definition
A case was defined as an infant who was younger than 1 year and experienced an episode of acute, isolated, diffuse alveolar hemorrhage of unknown cause, after discharge from the newborn nursery, and who was hospitalized at Rainbow Babies and Children’s Hospital during the period January 1993 to June 2000. In all of these infants, nonpulmonary bleeding was excluded. Lower respiratory tract blood was documented by endotracheal suctioning and/or confirmed by bronchoalveolar lavage (BAL) or, in the first case, by biopsy. In all subsequent infants who were not intubated, the diagnosis of acute, distal pulmonary hemorrhage was confirmed by BAL performed per the American Lung Association guidelines11 using 3 aliquots of 1.0 mL/kg body wt12 into one of the middle subsegments of the right lower lobe. Bright blood from the distal airways excluded a gastrointestinal source with aspiration of blood into the lung. Upper airway bleeding was excluded by direct or fiber-optic examination. When no clear respiratory bleeding source was evident, infants underwent upper gastroenteroendoscopic evaluation for esophageal and/or gastric bleeding. Early in the patient series, lung biopsy was also performed in a few infants for additional diagnostic information, but it became apparent that this was overly invasive, especially because more endobronchial information could be obtained by fiber-optic bronchoscopy.
Vigorous attempts were made to exclude known secondary causes of isolated, alveolar hemorrhage in infants (Table 1), eg, cardiac lesions causing increased pulmonary venous pressure were excluded by echocardiography. Coagulation studies were normal in all case infants. All infant families were seen by social services, and any care team suspicion of abuse led to evaluation by the child protection service. A prospective protocol of differential diagnostic evaluation was established at the time of the initial investigation of the first 10 cases; however, the acute care teams did not always complete all components, eg, not all infants received an echocardiogram (Table 2). Infants who were cared for at other hospitals or those with unexplained alveolar hemorrhage noted retrospectively by hemosiderin-laden macrophages on postmortem examination are not included in this report. All data were collected following a protocol approved by the Institutional Review Board of University Hospitals of Cleveland.
|
|
|
CLINICAL PROFILE OF CASES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSCLINICAL PROFILE OF CASESCASE SERIESDISCUSSIONCONCLUSIONREFERENCES |
|---|
For illustrating the spectrum of severity of the clinical profile of acute pulmonary hemorrhage seen in this series, short descriptions of 3 infants are presented, 1 with a typical presentation, 1 with a mild presentation, and 1 with a severe presentation.
Case A: Acute Alveolar Hemorrhage With Typical Presentation (Patient 8)
This black male was the first born of fraternal twins at 35 weeks’ gestation. His birth weight was 2.39 kg. He was hospitalized in a second-level nursery for 8 days for tachypnea and to rule out sepsis because his mother was culture positive for group B streptococcus. He received oxygen and ampicillin/gentamicin for 3 days with all cultures eventually negative. He was discharged with no additional diagnoses.
He was reportedly in good health until 6 weeks of age, when a weak cry was noted along with irregular breathing, pallor, lethargy, and perioral cyanosis. This was followed by frothy/foamy red secretions coming from his mouth and then an emesis of the same character. At the local emergency department, he was found to have poor color, nasal flaring and retractions, a respiratory rate of 106, bilateral lung crackles, and room air pulse oximetry of 84% saturation. An arterial blood gas on 2 L/min fraction of inspired oxygen showed a pH of 7.27, partial pressure of carbon dioxide of 44 mm Hg, and partial pressure of oxygen of 170 mm Hg. A chest radiograph demonstrated bilateral diffuse infiltrates. The infant was intubated and transported to our Pediatric Intensive Care Unit, where fresh bloody secretions were suctioned from the endotracheal tube. Neurologic examination was within normal limits except for disconjugate gaze. A computed tomogram of the head, cerebral spinal fluid analyses, and an ophthalmologic examination were normal.
Laboratory analyses on admission revealed a hemoglobin of 6.4 g/dL, hematocrit of 18.8%, white blood cell count of 5.2 x 109/L, platelet count of 330 x 109/L, with a normal coagulation profile and fibrin degradation products (prothrombin time 13.5 seconds, activated partial thromboplastin time 49 seconds, fibrinogen 103 mg/dL, D-dimer <0.25 mg/L), a moderately elevated total lactate dehydrogenase at 1435 U/L, a mildly elevated creatine phosphokinase of 334 U/L, and a negative toxicology screen. Urinalysis demonstrated 3+ hemoglobinuria, and the peripheral blood smear revealed the red cells to be 1+ fragmented, tear drop, burr, and ovalocytes. The infant received packed red cells and fresh-frozen plasma and remained intubated for 3 days with slowly decreasing bloody secretions. The alveolar infiltrates on the chest radiograph had cleared significantly by the second day and were absent by the fourth day, when he had been weaned to room air and was transferred out of the intensive care unit. A neurologic consultant examination on day 9 noted head lag, decreased tone in the upper extremities, and increased tone in the lower extremities. Through the remainder of his hospital course, he was noted to be irritable with frequent arching and crying. On day 11, he was discharged from the hospital with a cardiopulmonary monitor.
At 9 weeks of age, he was found with blood in his mouth, tachycardia to 200 bpm (per home monitor), and screaming. On hospital admission, no nasal/oral pharyngeal site of bleeding was found by fiber-optic endoscopy, his hematocrit was 29.7% (it had been 34% 10 days earlier), and a chest radiograph had an alveolar infiltrate in the right lower lobe, which subsequently cleared after 3 days. During the 4 days of hospitalization, he received prednisone 2 mg/kg/d and was discharged on 1 mg/kg/d. He went to a different home environment because of suspicions regarding the water damage and mold in the original home.
At 17 weeks of age, the infant underwent a follow-up BAL. The lavage fluid became increasingly bloody as the BAL proceeded and on cytologic evaluation contained numerous hemosiderin-laden macrophages (iron stain index13 264/300). The hematocrit was 29.2% with a reticulocyte count of 3.2% at this time. The prednisone was increased to 2 mg/kg/d.
At 8 months of age, the infant had a resting respiratory rate of 48 bpm, heart rate of 138 bpm, hematocrit of 33.2%, and reticulocyte count of 0.9%, and his stools had been guaiac negative for several months. Repeat bronchoscopy at 9 months yielded xanthochromic lavage fluid and many hemosiderin-laden macrophages (iron stain index of 270/300). The prednisone was tapered to 1 mg/kg/d.
At 34 months, he persisted with an elevated iron index of 59/300. The prednisone had been tapered and stopped 14 months after the initial hemorrhage. Although a neurologic reevaluation at 6 months was within normal limits, at 24 months of age his mother noted him to be clumsy and slow to climb stairs. His chest radiograph at 45 months had a fine reticular pattern, but he failed to appear for a scheduled high-resolution chest computed tomogram and has subsequently been lost to follow-up.
Case B: Acute Pulmonary Hemorrhage With Mild Presentation (Patient 9)
This is the only patient in this series who did not present with overt hemorrhage. She is the twin of case A described above. Her birth weight was 2.2 kg, and her newborn nursery course was similar to that of her brother. She was reportedly well until 6 weeks of age, when, the day after the twin’s pulmonary hemorrhage, she had a choking/cyanosis episode during a feeding and was hospitalized for an acute life-threatening event evaluation. There was no cough and no congestion, although the respiratory rate varied from 30 to 60 bpm. The chest radiograph was normal. A diaper rash was thought to be a possible explanation for intermittently positive stool guaiac testing. The hemoglobin was 9.6 g/dL, and the hematocrit was 27.4%. The acute life-threatening event was attributed to gastroesophageal reflux, and she was discharged from the hospital with cardiopulmonary monitoring.
At 10 weeks of age, the infant was hospitalized for acute onset of cervical adenitis and group B streptococcal (type III) sepsis. Her fever responded overnight to intravenous penicillin, and the adenitis improved over a few days. The neutrophil bacterial killing kinetics were normal. A peripheral blood eosinophilia peaked at 17% (total count 2.53 x 109/L) on day 8 of hospitalization. There was no pulmonary bleeding noted, although she had some nasal congestion and an occasional cough with light, fine crackles on chest examination. No chest radiograph was obtained. Room air pulse oximetry varied between 93% and 99% saturation. Her hemoglobin was 8.3 g/dL, hematocrit was 25.3%, and reticulocyte count was 8.4%. A urinalysis was normal. Her 14-day hospital course for intravenous penicillin was otherwise unremarkable except for irritability.
At 17 weeks of age, the infant underwent bronchoscopy for BAL because it was recognized that her hematologic profile had been consistent with blood loss and because she had previously been living in the same home environment as her twin brother, who had developed acute pulmonary hemorrhage. The airway mucosa was somewhat friable, but no bleeding was observed. Cytologic iron staining revealed numerous hemosiderin-laden macrophages with an iron index of 260/300. Examination before the bronchoscopy revealed a heart rate of 140 bpm, respiratory rate of 26, normal breath sounds, and oral thrush and a diaper rash with the appearance of a yeast infection. She was begun on daily prednisone at 1 mg/kg.
At 8 months of age, the infant’s hemoglobin was 11.6 g/dL, hematocrit was 33.9%, and reticulocyte count was 0.4%. The bronchoscopy was repeated at 9 months of age. The BAL fluid was noted to be very lightly xanthochromic and contained many hemosiderin-laden macrophages with an iron index of 250/300. At 14 months of age, her iron index was 0/300, but she had acute bleeding and a transitory roentgenographic infiltrate in response to the BAL. The chest radiograph rapidly returned to normal, and the prednisone was tapered and stopped 2 months later. The infant had no additional clinical problems in follow-up to 45 months of age.
Case C: Acute Pulmonary Hemorrhage With Severe Presentation (Patient 12)
The most severe patient presented with chronic pulmonary hemosiderosis and died in <24 hours. This black female was hospitalized at 8 months of age for evaluation of failure to thrive, developmental delay, and tachypnea accompanied by cyanosis and grunting. She had had a brief episode of epistaxis 2 days before admission but no other signs of overt bleeding. She progressed to pulmonary failure within the first 12 hours of hospitalization and on intubation was noted to have pulmonary hemorrhage. Urinalysis revealed 3+ hemoglobinuria. The alveolar hemorrhage became massive over the next few hours, eventually precluding effective ventilation, whereupon she died. Post mortem examination revealed extensive alveolar and interstitial hemosiderosis with pulmonary arterial and right ventricular hypertrophy. No secondary explanation for this long-standing alveolar hemorrhaging was found clinically or post mortem.
|
CASE SERIES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSCLINICAL PROFILE OF CASESCASE SERIESDISCUSSIONCONCLUSIONREFERENCES |
|---|
In the following summations, the 30 patients discussed all were cared for at Rainbow Babies and Children’s Hospital between January 1, 1993, and June 30, 2000. A number of infants were excluded from this series. During the same time period, 11 additional infants had unexplained pulmonary hemorrhage in Cleveland, but they were diagnosed retrospectively by reviewing BAL results (2) or post mortem (7) or were cared for at other area hospitals (2). None of the 30 infants in this series had elevated antibodies against cow milk proteins. Computed tomograms of the chest were used to screen for arteriovenous malformations with pulmonary angiography reserved only for infants from families with positive histories of arteriovenous malformation.
Suffocation is the most difficult component of the differential diagnosis to exclude, although the degree of pulmonary hemorrhage seems to be much less14 than most of the cases in this series. Early in this series, any suspicion of abuse led to an objective evaluation including ophthalmologic examination (10 patients), skeletal survey (18 patients), and head imaging (24 patients) and a protection team investigation. More recently, all infants with acute pulmonary hemorrhage have received these evaluations. Two infants were excluded from the series because of apparent abuse. One infant, who initially presented with severe pulmonary hemorrhage and met the case definition, 2 months later presented with acute retinal hemorrhage and a subdural hematoma and was thus excluded from this series. A second infant was excluded only on the basis of the protection team’s suspicion of abuse at the time of initial presentation. Twenty-eight additional infants presented with symptoms suggestive of alveolar hemorrhage but on BAL were found to be negative, ie, no acute bleeding or elevated hemosiderin-laden macrophages (iron index <20/300).
The 23 infants in the series who lived within a 10-zip code area around the hospital in eastern metropolitan Cleveland all were black. Of the 7 infants from outside this geographic area, 5 were white, 1 was Hispanic, and 1 was black. Twenty-three (77%) were boys, and 4 had been born at <35 weeks’ gestation. The mean age of initial pulmonary hemorrhage was 13.2 weeks (11.6 weeks corrected). The shortest interval between nursery discharge and onset of pulmonary hemorrhage was 2 weeks. Five of the infants died; acute massive alveolar hemorrhage was the major component in 4 of these deaths. One additional death occurred with an apparent apneic event related to acute bronchiolitis.
Summary of Clinical Presentation
The presenting signs and symptoms of the case infants are given along with other presenting parameters in Table 3. A prodrome including cough, chest congestion, unusual crying and irritability, and epistaxis was commonly present for hours to a few days before the onset of more severe symptoms. Limpness and dusky pallor lasting for minutes to a few hours often preceded frank respiratory failure. Occasionally, the infants screamed as though in pain and were difficult to console. Hemoptysis often began with throat gurgling before foamy blood issuing from the nose and mouth, with or without a wet cough. Hemoptysis was usually accompanied by lethargy and respiratory distress, including tachypnea and grunting. Respiratory distress commonly progressed to respiratory failure, although not uniformly requiring intubation. One infant had 2 episodes of apnea and bradycardia before a third episode was accompanied by hemoptysis and hematemesis. Overt hemorrhage was not always evident at presentation. Seven infants presented in respiratory failure and were recognized as having pulmonary hemorrhage only after intubation, when bright red blood was suctioned from the endotracheal tubes.
|
The severity of the initial pulmonary hemorrhage was sufficient to cause acute pulmonary distress in 26 of the 30 patients (87%). Of the 4 infants who died from acute pulmonary hemorrhage, 2 patients’ pulmonary hemorrhage resulted in hypoxic brain damage with death 1 to 2 days later. One patient (case C above) could not be adequately ventilated because the hemorrhage was so massive. The fourth death came with the patient’s third hemorrhage, which occurred while the patient slept prone with loose bedding. The cardiopulmonary monitor prescribed was not attached to the infant. The fifth death occurred 2 months after the initial hemorrhage with several contributing factors, including multiple small foci of acute pulmonary bleeding and viral bronchiolitis.
Twenty-two of the patients (73%) required ventilator support (mean of 4.6 days, excluding 1 infant with preexisting chronic lung disease of prematurity who was dependent on the ventilator for 12 months). Fresh blood was suctioned from the endotracheal tube of all of these intubated infants, usually for <1 day, although 4 continued to bleed for up to 2 days. Twenty-two of the patients (73%) received blood transfusions. One patient (case B described above) has not been observed to have overt bleeding except during a BAL.
Of the 7 patients who had respiratory failure before the detection of pulmonary hemorrhage, 4 had known accompanying stresses, eg, induction of general anesthesia for elective surgery, prolonged (40 minutes) seizure with fever caused by Influenza A, hypernatremic dehydration (Na = 157 mM), hyponatremic seizures (Na = 112 mM). The other 3 of these patients had no known specific stressor, and 1 subsequently died from the pulmonary hemorrhage (case C above). There was no obvious accompanying stress at the time of the acute hemorrhage among the 22 infants who presented with hemoptysis. Toxicology screening of 18 patients failed to find any evidence for illicit drugs, including cocaine (Table 4).
|
Four of the case infants (13%) presented without radiographic infiltrates,15 indicating that chest radiographs may not always demonstrate acute hemorrhage. More common, initial parenchymal infiltrates significantly decreased in 1 to 2 days as noted in previous descriptions of alveolar hemorrhage in contrast to those of infectious origin.16 Fifteen patients had electroencephalograms during their hospitalization; 11 were abnormal. Most common, generalized slowing from deep sedation or its residual was reported, although seizures were observed on 3 of the recordings. Echocardiographs and/or electrocardiograms on 18 patients revealed transient pulmonary hypertension in 2 patients (7 and 23). Right ventricular hypertrophy related to chronic hemosiderosis was seen only in the severe case described above (case C [patient 12]). Abdominal ultrasound detected bilateral adrenal hemorrhages in 1 patient (patient 28).
The acute clinical manifestations were not limited to the pulmonary system. Generalized seizures were components of the acute course in 7 of the infants (Table 3). Three of the infants presented with failure to thrive, 1 of whom had developmental delays. Eleven infants had frank hemoglobinuria, 5 to the level of 3+ (Table 5). Five of these infants had negative screening tests for glucose-6-phosphate dehydrogenase deficiency. The hemoglobinuria was transitory, and no additional abnormal urine analyses were observed. Normal renal histology was confirmed post mortem in 3 infants. All of the 27 patients whose admission peripheral blood smears were evaluated had at least 1+ fragmented or damaged red cells.
|
Twenty-eight of the patients had blood cultures, only 1 of which was positive (single positive culture for Staphylococcus epidermidis). Cultures of tracheal aspirates and/or BAL samples (obtained without a protected brush) were positive in 10 patients (Table 5). Attempts to culture environmental fungi from these samples were unsuccessful. In neither of these cases or the 1 with presumed sepsis did infection seen to offer an adequate cause for the alveolar hemorrhage in that the radiographic infiltrates cleared rapidly. One infant with marked hemoptysis also had Pneumocystis carinii pneumonia (patient 19). Although his human immunodeficiency virus antigen, culture, and polymerase chain reaction all were negative, a transiently absent T-cell mitogenic response to concanavalin A was found.
Pathology
Three infants had an open lung biopsy performed at the time of their initial presentation. All 3 had fresh alveolar hemorrhage with negative iron (hemosiderin) staining. Two of the 3 also demonstrated mild chronic interstitial pneumonitis, with lymphocytes and histiocytes infiltrating alveolar septa; 1 of these had occasional polymorphonuclear leukocytes present in an otherwise chronic interstitial pneumonia (see Table 6).
|
Autopsies were performed in 4 of the 5 infants who died. Two of them died during their first episode of pulmonary bleeding: 1 had massive pulmonary hemosiderosis and acute hemorrhage, and the other had massive pulmonary hemorrhage, iron stain negative. The other 2 autopsied patients died 1 and 4 months after the initial bleed: 1 had multifocal alveolar hemorrhage with positive hemosiderin staining, and the other showed pulmonary hemosiderosis and extensive pulmonary hemorrhage. No other organs had significant pathologic findings except for infant 12, who had small foci of abnormal central nervous system myelination in deep gray matter structures.
Treatment
When the infants received supportive therapy on a timely basis, the prognosis seemed to be fairly good. Only 1 of our patient deaths occurred with an infant whose terminal course began in the hospital (most severe case described above); the other deaths occurred or reached irreversibility at home. The 3 deaths that occurred after the initial hospitalization were infants who either never received steroids or had them stopped shortly after discharge. Initially, patients were treated with glucocorticoids somewhat empirically, but subsequently the finding of chronic inflammation on lung biopsy and post mortem together with the inflammation seen in animal models (see below) suggested potential efficacy of steroid treatment as used in other inflammatory disorders with alveolar hemorrhage.23 Only 1 of the surviving infants did not receive steroids (patient 1), and only 1 other (patient 19; P carinii pneumonia patient) had steroids discontinued after 1 week. From this experience, we currently treat with methylprednisolone, 1 mg/kg every 6 hours while the patient is in the intensive care unit and transition to maintenance prednisone, 1 mg/kg/d, by the time of discharge from the hospital. A repeat BAL is performed 3 to 6 weeks after the initial pulmonary hemorrhage, and the prednisone is taken to every other day if the iron index is <100/300. The prednisone dose is increased and returned to every day if the infant has a subsequent pulmonary hemorrhage or if there are several positive stool guaiacs accompanied by a reticulocytosis and/or tachypnea. Although hemoglobin, hematocrit, stool guaiacs, and reticulocyte counts are very helpful in following these patients, all of these determinations can transition to the normal range or negative, even though the hemosiderin-laden macrophages present in the BAL are still very significant at iron indices of >100/300. Our usual protocol is to repeat the BAL every 2 to 4 months on the basis of normalcy of other monitoring parameters. When the BAL iron index goes below 50/300, the steroids are transitioned through a 4-week period of physiologic replacement and then stopped.
Although this protocol of BAL iron index to monitor for continued alveolar hemorrhage and thus continued use of steroid treatment was instituted somewhat arbitrarily, it was with some experimental and clinical insight.13,18 We recognized the possible inaccuracy of the suggested time of 2 weeks to clear the alveoli of hemosiderin-laden macrophages after an acute hemorrhage13 and chose 3 to 6 weeks as the first interval before follow-up bronchoscopy. Subsequently reported animal studies19 suggested a significant decrease in hemosiderin-laden macrophages by 30 days but low levels observed out to 60 days. Our use of an iron index of 50/300 to terminate steroid treatment seems reasonable in light of these animal studies; however, extrapolation from mice to human infants is not without risk.
The only specific treatment measure suggested in our center is to remove the infant from the residential environment. Before our recommendation that infants not be returned to the home in which the hemorrhage occurred, 5 of 7 infants had recurrent overt pulmonary hemorrhaging (Table 6). The use of steroid treatment without changing the home environment was not sufficient to preclude rebleeding. Subsequent to the changing of the home environment, only 1 of 21 infants has had overt pulmonary rebleeding. We strongly stress the need for a tobacco smoke-free environment. Our current pathophysiologic concepts (see below) include tobacco smoke as a trigger of the alveolar hemorrhage. One infant who had been in a new environment for 16 months and off of steroids for 3 months had a major, life-threatening pulmonary hemorrhage at 22 months of age, several weeks after the return of environmental tobacco smoke (ETS) to his environment.
Subsequent Course
Most of the infants were found to have continued low-grade alveolar bleeding on the basis of the continued elevated iron index of the BAL (Table 6). Although all but 2 of the infants in this series who survived the initial hospitalization underwent fiber-optic bronchoscopy for BAL, 22 had the procedure 3 to 12 weeks after the acute pulmonary bleed. Hemosiderin-laden macrophages have been thought to be mostly cleared from alveolar spaces by 3 weeks after an acute bleed (reference 13, see above); therefore, finding an elevated iron index remaining >3 weeks later suggests continued bleeding. Eighteen of the 22 patients (82%) had iron indices 35/300 or greater with 15 >100/300 and 6 >200/300. Twelve (40%) of the patients had elevated iron indices for >4 months. This continued pulmonary hemosiderosis occurred regardless of protective environments and prolonged glucocorticoid therapy.
To date, 22 surviving patients have transitioned off glucocorticoids (2 lost to follow-up, 1 did not receive prolonged steroid courses, and 5 deaths) with a mean duration of treatment of 7.8 months, but some for prolonged courses >14 months (Table 6). It seems that these infants remain vulnerable to additional pulmonary hemorrhages during this period of continued hemosiderosis. Throughout this period, we follow daily respiratory rates (asleep) and stool guaiacs several times a week. Tachypnea decreases over several weeks, and the stools seldom contain occult blood.
Ten of 23 infants (43%) have subsequently had recurrent wheezing illnesses requiring bronchodilator therapy (Table 6). This is approximately twice the reported incidence in the general population.20 Of the 16 infants who have had chest radiographs >1 year after the initial hemorrhage, only 1 has continued peripheral lung abnormalities (mild reticular interstitial markings). Other than those with clinically judged mild to moderate recurrent wheezing, these patients seem to have normal pulmonary clinical status. Seven of the mothers subsequently reported mild to moderate developmental delay in their children (Table 6). Only 1 of these children (patient 17) had had seizures during the acute course of pulmonary hemorrhage.
Epidemiology
Table 4 delineates the individual patient experiences with potential contributing factors. Environmental cocaine exposure, as detected by urinary determinations, was negative in all 8 cases investigated initially and in a total of 18 cases on screening of stored urine samples collected on initial presentation. Other observations support the conclusions of the initial case-control study3,4 (see Table 4). Twenty-six of 29 infants (89%) lived in water-damaged homes, 25 of 28 (89%) lived in homes containing toxigenic fungi, and 27 of 30 (90%) were exposed to ETS in the home. Environmental sampling and history revealed that the bedroom patient 12 (case C) shared with her mother had been contaminated with Stachybotrys chartarum since the second trimester of gestation. The major finding of the case-control study was that exposure to S chartarum and other fungi was a risk factor for pulmonary hemorrhage.4 Nine of the first 10 patients had been exposed to ETS, which was found to be a covariant with toxigenic fungi in the home.4 None of the initial 10 infants had ever been breastfed, whereas 37% of the control infants had been. Among the subsequent cases, 4 were breastfed for a total of 13% (Table 4). Other potential factors, such as antibodies against cow milk proteins and pesticides, were excluded in the initial case-control study.3 Intentional, atraumatic airway obstruction is difficult to exclude in individual cases, but several arguments have been previously presented21 against this being a major component in this series, most notably the dramatic decrease in rebleeding with home environment change (Table 6) but no change in caregiver.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSCLINICAL PROFILE OF CASESCASE SERIESDISCUSSIONCONCLUSIONREFERENCES |
|---|
Primary, diffuse alveolar hemorrhage and subsequent chronic pulmonary bleeding are uncommon problems in young full-term infants. In young children, pulmonary hemosiderosis has been associated with hyperreactivity to cow milk protein (Heiner’s syndrome22), although pathophysiologic connections have not been clearly delineated.17,23 In older children and adults, alveolar capillaritis can be a source of hemorrhage as seen with immune complex deposition in various collagen vascular disorders23–25 or with antibodies against type IV collagen in the pathogenesis of Goodpasture’s syndrome.26 However, immunologically mediated capillary damage is rare in young children in whom the majority of the cases of primary pulmonary hemosiderosis are idiopathic.24 The possibility of toxic causes was raised by a retrospective study of 30 cases during a 20-year period in Greece,27 and certain chemical agents such as the plasticizer trimellitic anhydride28 are known to produce chronic alveolar bleeding.
Potential Pathophysiology
The association of pulmonary hemorrhage in these infants with exposure to S chartarum and other fungi has been made epidemiologically,4 although this association has generated controversy.29,30 Related animal studies31–35 have supported the plausibility of the association. The toxigenic fungus S chartarum is capable of making several classes of toxins, including macrocyclic trichothecenes,36,37 phenylspirodrimanes,38 a cyclosporin,39 a hemolysin,40,41 and proteinases.42,43 The trichothecenes are potent protein synthesis inhibitors44 that in addition to being immune suppressive can induce inflammatory cytokines at lower doses.45 The phenylspirodrimanes and cyclosporin are also immune suppressive by inhibition of C 5 activation and suppressing T-cell mitogenesis, respectively. Comparison of these toxin activities with the clinical profile of the patients suggests several pathophysiologic possibilities after inhalation of the toxin-containing spores. The trichothecenes and hemolysin are potential sources of acute cell injury and death leading to local destruction of the alveolar capillary wall and thereby alveolar hemorrhage. Such a role in the hemorrhagic pneumonia of neonates has been suggested for the hemolysin of group B streptococcus.46 The release of collagen-degrading proteinases from Stachybotrys spores in animal studies43 suggests the possibility of acute structural damage to the alveolar capillary wall. On a more chronic basis, the synthesis of type IV collagen and other protein components of endothelial basement membrane of capillaries in rapidly growing lungs of young infants is vulnerable to the inhibitory activity of the trichothecenes. The resulting capillary fragility would be at risk for stress failure47 whenever the pulmonary capillary pressure is elevated, eg, sympathetic reaction to asphyxia, unequal hypoxic vasoconstriction from ETS. Transmural pressures insufficient to rupture normal capillaries may be pathogenic under these conditions.47 Although there are no specific therapies or antagonists known for the Stachybotrys toxins, one animal study suggested that the use of glucocorticoids during the acute toxic course may be of some value.48
Although highly speculative, it is worth noting additional potential roles of mycotoxins in other manifestations seen in this series. The observation of hemoglobinuria in more than one third of the patients is unusual for an alveolar hemorrhage disorder without renal involvement. Presumably, the hemoglobinuria reflects a hemolytic process perhaps within the alveoli. Although the trichothecenes are known to be hemolytic,49 it is only at fairly high concentrations. Stachybotrys has recently been noted to produce a hemolysin,40,41 perhaps a source of the hemolysis. The red cell abnormalities seen uniformly on the patients’ blood smears, which could have arisen through acute microangiopathic processes, are less likely to be an adequate explanation. Comparison of the degree of hemoglobinuria with the severity of alveolar hemorrhage as judged by the hematocrit nadir (Table 3) failed to give a correlation.
There are several possible explanations for the hemosiderosis that persists for months beyond the acute hemorrhagic period. The capillaries may remain fragile for a prolonged time with an increased susceptibility to leakage even with comparatively minor stresses. Inflammation, as seen in mice and rats exposed to S chartarum spores,31–35 may be induced by the trichothecenes45 or the ß-glucan of the spore wall.50 Persistence of this inflammatory reaction beyond the acute toxicity period could be a source of alveolar bleeding, either directly or through interaction with any persisting capillary fragility. The hemorrhagic inflammation seen in the lungs of rodents after the administration of Stachybotrys spores31–35 suggests the value of continued anti-inflammatory therapy.
Several of the Stachybotrys toxins are immune suppressive. Bilateral otitis media persisting through multiple courses of different antibiotics for 6 months (case C above) is somewhat unusual, as is group B streptococcal sepsis at 10 weeks of age in an infant who received 3 days of antibiotics as a neonate (mild case B above), as is P carinii pneumonia in an otherwise normal infant. The last infant had a transitory but severe suppression of his T-cell mitogenic response to concanavalin A consistent with exposure to cyclosporin. Although the concomitant seizures may have been a result of acute anoxic damage, the neurotoxic activity of the trichothecenes51 may also have played a role.
Thus, potential roles of Stachybotrys toxins can be suggested. However, additional evidence to support causation is needed. The possibility remains that the presence in the infants’ homes of Stachybotrys, a high water-requiring fungus, may simply be an indicator of water intrusion, and other unknown, related factors (in addition to ETS) may play primary or secondary causative roles. Three additional cases of acute pulmonary hemorrhage linked to Stachybotrys or Trichoderma (also a producer of trichothecene mycotoxins) have been reported from other centers.6,7,52
|
CONCLUSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSCLINICAL PROFILE OF CASESCASE SERIESDISCUSSIONCONCLUSIONREFERENCES |
|---|
The recent Cleveland experience with acute pulmonary hemorrhage in infants includes a spectrum of clinical presentations. Most common is severe pulmonary symptoms requiring intensive support, but a few cases had less severe hemorrhage. The latter cases suggest that there may be infants who have the disorder and escape clinical detection. This is supported by the retrospective observation of six sudden infant death syndrome victims from this same geographic area and time period who had extensive pulmonary hemosiderosis on reexamination of their lung sections with iron staining.2,53 Patients followed closely may continue to have low-grade alveolar hemorrhage for months after their initial bleed, even with the removal of the infants from their home environments. There are currently no specific treatment modalities other than termination of environmental mold exposure and avoidance of ETS. The protection of young infants from moldy environments continues to be a prudent public health approach.54
Pediatric Pulmonary Division
Department of Pediatrics
Rainbow Babies and Children’s Hospital
Case Western Reserve University
Cleveland, Ohio
Department of Pathology
University Hospitals of Cleveland
Case Western Reserve University
Cleveland, Ohio
Division of Community Health
Cuyahoga County Board of Health
Cleveland, Ohio
Division of Respiratory Disease Studies
National Institute for Occupational Safety and Health
Morgantown, West Virginia
Georgia Pediatric Cardiology
Atlanta, Georgia
George Washington University School of Public Health and Health Services
Washington, DC
|
ACKNOWLEDGMENTS |
|---|
This project was supported in part by a grant jointly funded by the Centers for Disease Control and Prevention and the National Institutes of Health, ES08549; by a grant from the NIEHS, NIH, ES09742; by NIH General Clinical Research Center Grant M01 RR00080; and by grants from the Rainbow Board of Trustees, the Ohio Department of Health, The Cleveland Foundation, the Gund Foundation, and the Abington Foundation.
We appreciate the continued discussions with David Miller, Bruce Jarvis, Iwana Yike, Steven Vesper, Thomas Rand, and Martha Miller and the data collection efforts of Margaret Pizzi and Azadeh Frazin.
|
FOOTNOTES |
|---|
Received for publication Mar 26, 2001; Accepted Feb 12, 2002.
Reprint requests to (D.G.D.) Department of Pediatrics, Case Western Reserve University, 11100 Euclid Ave, Cleveland, OH 44106-5000. E-mail: dxd9{at}po.cwru.edu
The contents of this publication are solely the responsibility of the authors and do not necessarily reflect the official views of any of the funding agencies, including the Department of Health and Human Services.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSCLINICAL PROFILE OF CASESCASE SERIESDISCUSSIONCONCLUSIONREFERENCES |
|---|
- CDC. Acute pulmonary hemorrhage/hemosiderosis among infants-Cleveland, January 1993–November 1994. MMWR Morb Mortal Wkly Rep.1994; 43 :881 –883[Medline]
- CDC. Update. Pulmonary hemorrhage/hemosiderosis among infants-Cleveland, Ohio, 1993–1996. MMWR Morb Mortal Wkly Rep.1997; 46 :33 –35[Medline]
- Montana E, Etzel RA, Allan TM, Horgan TE, Dearborn DG. Environmental risk factors associated with pediatric idiopathic pulmonary hemosiderosis in a Cleveland community. Pediatrics.1997; 99(1) . Available at: www.pediatrics.org/cgi/content/full/99/1/e5
- Etzel RA, Montana E, Sorenson WG, Kullman GJ, Allan TM, Dearborn DG. Acute pulmonary hemorrhage in infants associated with exposure to Stachybotrys atra and other fungi.
Arch Pediatr Adolesc Med.1998; 152
:757
–762
[Abstract/Free Full Text] - Etzel RA, Dearborn DG. Pulmonary hemorrhage among infants with exposure to toxigenic molds: an update. In: Johanning E, ed. Bioaerosols, Fungi, and Mycotoxins: Health Effects, Assessment, Prevention and Control. Albany, NY: Boyd Printing Co, Inc; 1999:70–83
- Flappan SM, Portnoy J, Jones P, Barnes C. Infant pulmonary hemorrhage in a suburban home with water damage and mold (Stachybotrys atra). Environ Health Perspect.1999; 107 :927 –930[Web of Science][Medline]
- Novotny WE, Dixit A. Pulmonary hemorrhage in an infant following two weeks of fungal exposure.
Arch Pediatr Adolesc Med.2000; 154
:271
–275
[Abstract/Free Full Text] - Tripi PA, Modlin S, Sorenson WG, Dearborn DG. Acute pulmonary hemorrhage in an infant during induction of general anesthesia. Paediatr Anesth.2000; 10 :92 –94
- Vesper SJ, Dearborn DG, Yike I, et al. Evaluation of Stachybotrys chartarum in the house of an infant with idiopathic pulmonary hemorrhage: quantitative assessment before, during, and after remediation. J Urb Health.2000; 77 :66 –85
- Dearborn DG, Yike I, Sorenson WG, Miller MJ, Etzel RA. Overview of investigations into pulmonary hemorrhage among infants in Cleveland, Ohio, Environ Health Perspect.1999; 1073 :495 –499
- Green CG, Eisenberg J, Leong A, Nathanson I, Schnapf BM, Wood RE. Flexible endoscopy of the pediatric airway, Am Rev Respir Dis.1992; 145 :233 –235[Web of Science][Medline]
- Ratjen F, Bruch J Adjustment of bronchoalveolar lavage volume to body weight in children. Pediatr Pulmonol.1996; 21 :184 –188[CrossRef][Web of Science][Medline]
- Sherman JM, Winnie G, Thomassen MJ, Abdul-Karim FW, Boat TF. Time course of hemosiderin production and clearance by human pulmonary macrophages.
Chest.1984; 86
:409
–411
[Abstract/Free Full Text] - Southall DP, Plunkett MCB, Banks MW, Falkov AF, Samuels MP. Covert video recordings of life-threatening child abuse: lessons for child protection.
Pediatrics.1997; 100
:735
–760
[Abstract/Free Full Text] - Berlin S, Sivit C, Dearborn DG. Pulmonary hemorrhage/hemosiderosis. Spectrum of radiographic findings. Annual meeting of the Radiology Society of North America, Chicago 1997. Radiology.1997; 205 :466 (abstr)
- Primack SL, Miller RR, Muller NL. Diffuse pulmonary hemorrhage: clinical, pathologic, and imaging features.
Am J Radiol.1995; 164
:295
–300
[Abstract/Free Full Text] - Stafford HA, Polmar SH, Boat TF. Immunologic studies in cow milk-induced pulmonary hemosiderosis. Pediatr Res.1977; 11 :898 –903[Web of Science][Medline]
- Kahn F, Jones J, England D. Diagnosis of pulmonary hemorrhage in the immunocompromised host. Am J Rev Resir Dis.1987; 186 :155 –160
- Epstein CE, Elidemir O, Colasurdo GN, Fan LL. Time course of hemosiderin production by alveolar macrophages in a murine model.
Chest.2001; 120
:2013
–2020
[Abstract/Free Full Text] - Martinez FD, Morgan WJ, Wright AL, Holberg CJ, Taussig LM. Diminished lung function as a predisposing factor for wheezing respiratory illness in infants. N Engl J Med.1988; 319 :1112 –1117[Abstract]
- Etzel RA, Montana E, Dearborn DG, et al. SIDS or murder?
Pediatrics.1998; 101
:953
–955 (letter)
[Free Full Text] - Heiner DC, Sears JW, Kniker WT. Multiple precipitins to cow’s milk in chronic respiratory disease. Am J Dis Child.1962; 103 :634 –654
- Boat, TF. Pulmonary hemorrhage and hemoptysis. In: Chernick V, Boat TF, eds. Kendig’s Disorders of the Respiratory Tract in Children. 6th Ed. Philadelphia, PA: WB Saunders; 1998:623–633
- Levy J, Wilmott R. Pulmonary hemosiderosis. In: Hilman BC, ed. Pediatric Respiratory Disease: Diagnosis and Treatment. Philadelphia, PA: WB Saunders; 1993:543–549
- McKoy KS. Hemosiderosis. In: Taussig LM, Landau LI, eds. Pediatric Respiratory Medicine. St. Louis, MO: Mosby; 1999:835–841
- Duncan DA, Drummond KN, Michael AF, Vernier RL. Pulmonary hemorrhage and glomerulonephritis. Report of six cases and study of the renal lesion by fluorescent antibody technique and electron microscopy. Ann Intern Med.1965; 62 :920
- Cassimos CD, Chryssanthopoulos C, Panagiotidou C. Epidemiologic observations in idiopathic pulmonary hemosiderosis. J Pediatr.1983; 102 :698 –702[CrossRef][Web of Science][Medline]
- Ahmad D, Patterson R, Morgan WKC, Williams T, Zeiss CR. Pulmonary haemorrhage and haemolytic anemia due to trimellitic anhydride. Lancet.1979; 2 :328 –330[Web of Science][Medline]
- CDC. Update. Pulmonary hemorrhage/hemosiderosis among infants-Cleveland, Ohio, 1993–1996. MMWR Morb Mortal Wkly Rep.2000; 49 :180 –184[Medline]
- Etzel RA, Dearborn DG, Allan TM, et al. Investigator team’s response to MMWR report. Available at: gcrc.meds.cwru.edu/stachy/InvestTeamResponse.html
- Nikulin M, Reijula K, Jarvis BB, Hintikka E-L. Experimental lung mycotoxicosis in mice induced by Stachybotrys atra. Int J Exp Pathol.1996; 77 :213 –218[CrossRef][Web of Science][Medline]
- Nikulin M, Reijula K, Jarvis BB, Veijalainen P, Hintikka E-L. Effects of intranasal exposure to spores of Stachybotrys atra in mice. Fundam Appl Toxicol.1997; 35 :182 –188[CrossRef][Web of Science][Medline]
- Yike I, Miller MJ, Sorenson WG, Walenga R, Tomashefski JF, Dearborn DG. Infant animal model of pulmonary mycotoxicosis induced by Stachybotrys chartarum. Mycopathologia. In press
- Rao CY, Brain JD, Burge HA. Reduction of pulmonary toxicity of Stachybotrys chartarum spores by methanol extraction of mycotoxins.
Appl Environ Microbiol.2000; 66
:2817
–2821
[Abstract/Free Full Text] - Rao CY, Brain JD, Burge HA. The time course of responses to intratracheally instilled toxic Stachybotrys chartarum spores in rats. Mycopathologia.2000; 149 :27 –34[CrossRef][Web of Science][Medline]
- Jarvis BB, Salemme J, Morais A. Stachybotrys toxins 1. Nat Toxins.1995; 3 :10 –16[CrossRef][Medline]
- Jarvis BB, Sorenson WB, Hintikka E-L, et al. Studies of toxin production by isolates of Stachybotrys atra and Memnoniella echinata isolated from homes associated with pulmonary hemosiderosis in infants.
Appl Environ Microbiol.1998; 64
:3620
–3625
[Abstract/Free Full Text] - Miyazaki W, Tamaoka H, Shinohara M, et al. A complement inhibitor produced by Stachybotrys, sp. K-76, a new species of fungi imperfecti. Microbiol Immunol.1980; 24 :1091 –1108[Web of Science][Medline]
- Sakamoto K, Tsujii E, Miyauchi M, et al, M. FR901459, a novel immunosuppressant isolated from Stachybotrys chartarum (atra) No. 19392. J Antibiot (Tokyo).1993; 46 :1788 –1798[Medline]
- Vesper SJ, Dearborn DG, Elidemir O, Haugland RA. Quantification of siderophore and hemolysin from Stachybotrys chartarum strains, including a strain isolated from the lung of a child with pulmonary hemorrhage and hemosiderosis.
Appl Environ Microbiol.2000; 66
:2678
–2681
[Abstract/Free Full Text] - Vesper SJ, Magnuson ML, Dearborn DG, Yike I, Haugland RA. Initial characterization of the hemolysin, stachylysin, from Stachybotrys chartarum.
Infect Immun.2001; 69
:912
–916
[Abstract/Free Full Text] - Kordula T, Banbula A, Macomson J, Travis J. Isolation and properties of Stachyrase A, a chymotrypsin-like serine proteinase from Stachybotrys chartarum.
Infect Immun.2002; 70
:419
–421
[Abstract/Free Full Text] - Yike I, Rand T, Dearborn D. Proteases from the spores of toxigenic fungus Stachybotrys chartarum. Presented at the American Lung Association/American Thoracic Society; May 2002; Atlanta, GA (abstr)
- Feinberg B, MacLaughlin CS. Biochemical mechanism of action of trichothecene mycotoxins. In: Beasley VR, ed. Trichothecene Mycotoxicosis: Pathophysiologic Effects. Vol 1. Boca Raton, FL: CRC Press; 1989:27–36
- Pestka JJ, Bondy GS. Mycotoxin-induced immune modulation. In: Dean JH, Luster MI, Munson AE, Kimber I, eds. Immunotoxicology and Immunopharmacology. New York, NY: Raven Press Ltd; 1994:163–181
- Gibson RL, Nizet V, Rubens CE. Group B streptococcal ß-hemolysin promotes injury of lung microvascular endothelial cells. Pediatr Res.1999; 45 :626 –634[Web of Science][Medline]
- West JB, Mathieu-Costello O. Structure, strength, failure, and remodeling of the pulmonary blood-gas barrier. Annu Rev Physiol.1999; 61 :543 –572[CrossRef][Web of Science][Medline]
- Harvey PW, Healing G, Rees SJ, Everett DJ, Cockburn A. Glucocorticosteroid interactions with natural toxins: a mini review. Nat Toxins.1994; 2 :341 –346[Medline]
- Beasley VR, Bunner DL, Poppenga RH. Conclusions and future directions for research. In: Beasley VR, ed. Trichothecene Mycotoxicosis: Pathophysiological Effects. Vol II. Boca Raton, FL: CRC Press; 1989:169–189
- Rylander R. Indoor air-related effects and airborne (1,3)-ß-D-glucan. Environ Health Perspect.1999; 107(suppl 3) :501 –503
- Feuerstein G, Lorenzana RM, Beasley VR. Effects of trichothecene mycotoxins on the nervous system. In: Beasley VR, ed. Trichothecene Mycotoxicosis: Pathophysiological Effects. Vol II. Boca Raton, FL: CRC Press; 1989:111–122
- Elidemir O, Colasurdo GN, Rossman SN, Fan LL. Isolation of Stachybotrys from the lung of a child with pulmonary hemosiderosis.
Pediatrics.1999; 104
:964
–966
[Abstract/Free Full Text] - Dearborn DG. Pulmonary hemorrhage in infants and children. Curr Opin Pediatr.1997; 9 :219 –224[CrossRef][Medline]
- American Academy of Pediatrics, Committee on Environmental Health. Toxic effects of indoor molds.
Pediatrics.1998; 101
:712
–714
[Abstract/Free Full Text]
PEDIATRICS (ISSN 1098-4275). ©2002 by the American Academy of Pediatrics
CiteULike 
Connotea 
Del.icio.us 
Digg 
Facebook 
Reddit 
Technorati 
Twitter What's this?
This article has been cited by other articles:
|
|
 |
|
  J. J. Pestka, I. Yike, D. G. Dearborn, M. D. W. Ward, and J. R. Harkema Stachybotrys chartarum, Trichothecene Mycotoxins, and Damp Building-Related Illness: New Insights into a Public Health Enigma Toxicol. Sci., July 1, 2008; 104(1): 4 - 26. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. J. Mazur, J. Kim, and the Committee on Environmental Health Spectrum of Noninfectious Health Effects From Molds Pediatrics, December 1, 2006; 118(6): e1909 - e1926. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 O.C. Ioachimescu, S. Sieber, and A. Kotch Idiopathic pulmonary haemosiderosis revisited Eur. Respir. J., July 1, 2004; 24(1): 162 - 169. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. D. Le Cras, R. E. Spitzmiller, K. H. Albertine, J. M. Greenberg, J. A. Whitsett, and A. L. Akeson VEGF causes pulmonary hemorrhage, hemosiderosis, and air space enlargement in neonatal mice Am J Physiol Lung Cell Mol Physiol, July 1, 2004; 287(1): L134 - L142. [Abstract] [Full Text] [PDF]
|
|
eLetters:
Read all eLetters
- Not the Black Mold
- Thivakorn Kasemsri
- Pediatrics Online, 8 Sep 2002 [Full text]
- Re: Not the Black Mold
- Dorr G. Dearborn, et al.
- Pediatrics Online, 22 Sep 2002 [Full text]
- Never say never but
- Thivakorn Kasemsri
- Pediatrics Online, 17 Jul 2008 [Full text]
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
