Pulmonary Hemorrhage Complicating a Typical Hemolytic-Uremic SyndromePiastra M.a · Ruggiero A.b · Langer A.a · Caresta E.a · Chiaretti A.a · Pulitanò S.a · Polidori G.a · Riccardi R.b
aPediatric Intensive Care Unit and bPediatric Hematology-Oncology, Catholic University Medical School, Rome, Italy
We describe a case of pulmonary bleeding and subsequent acute respiratory distress syndrome (ARDS) in a 20-month-old female suffering from a typical postdiarrheal hemolytic-uremic syndrome (HUS). Acute renal failure was treated early by peritoneal dialysis. It is of interest to underline that thrombocytopenia or any coagulative impairment was absent when this complication occurred, and spontaneous diuresis recovery was ongoing. All examinations failed to identify a unique site of bleeding in the main stem bronchi or trachea. Complete renal and clinical recovery was obtained in spite of this very uncommon complication. Intensivists should be aware of this rare and potentially fatal complication of typical (D+)HUS.
Copyright © 2004 S. Karger AG, Basel
Hemolytic-uremic syndrome (HUS) is a leading cause of acute renal failure in children, and its clinical characteristics include microangiopathic hemolytic anemia, thrombocytopenia and oligoanuria of variable degree, with frequent multiorgan involvement due to diffuse endothelial damage . Nevertheless, pulmonary vascular impairment in HUS is rarely reported in the literature . Pulmonary hemorrhage in typical postdiarrheal HUS of childhood is nearly absent; in adolescents, only one report exists of fatal HUS presenting with pulmonary hemorrhage in a 18-year-old patient . In adults, some reports exist of patients with mitomycin C-associated HUS developing massive pulmonary bleeding  probably related to angiomatoid vascular changes of the lungs . Pulmonary hemorrhage – though uncommon – confers a poor prognosis to HUS-affected patients . As a whole, HUS mortality rate has decreased from about 100% in the first cases described by Gasser et al.  in 1955 to <10% with modern supportive therapy ; however, in developing countries mortality remains very high (>60%) .
A 20-month-old female was admitted to our Pediatric Intensive Care Unit (PICU) in April 2002 presenting with signs of neurological involvement associated with severe anemia and thrombocytopenia.
Her birth- and past medical history, growth and development were unremarkable. The clinical history begun 4 days before with bloody diarrhea (4–6 episodes/day) with rectal prolapse. No antibiotics or cultures were prescribed on initial clinical observation, and dietary counseling was adopted. However, her general condition worsened and intermittent high fever (39°C), vomiting and inappetence appeared. On admission to the Emergency Department, a hematological work-up showed platelet count 122 × 109/l, WBC 10 × 109/l, hemoglobin level 8.3 g/dl. Neurological signs such as irritability, tremors, convergent strabismus, lower extremity hypertonia were observed. Dehydration signs were pronounced. A diagnosis of HUS was suspected and the child was referred to our PICU. On admission, the child was febrile (38°C) and appeared moderately dehydrated and lethargic (GCS score 8–9). Physical examination revealed tachycardia (165 beats/min), tachypnea (40 breaths/min), blood pressure 122/50 mm Hg, SaO2 93% in room air, with normal capillary refill. Lethargia, rigor nucalis, and hemilateral hypertonia (right side) were observed. CSF analysis was normal. Diffuse petechiae and ecchymoses were noted, but no major or mucosal bleeding was recorded. In the following 24 h, the urinary output eventually stopped and a complete HUS picture developed, including Hb 7.6 g/dl, Ht 22% with schistocytosis on blood film, platelet count 25 × 109/l, WBC 7.01 × 109/l (neutrophils 76%), LDH 4,102, total bilirubin 1.1 mg/dl, haptoglobin <6 mg/dl. A slight hyponatremia was observed (Na 128 mEq/l); total calcemia was 7.8 mg/dl, potassium 4.1 mEq/l, albumin 2.8 g/dl, serum creatinine 1.4 mg/dl, BUN 30 mg/dl. Coagulation testing: prothrombin activity 93%, activated partial thromboplastin time (aPTT) 32 s, fibrinogen 238 mg/dl, ATIII 90.6%. D dimer testing was slightly elevated (336 ng/ml). Invasive monitoring was begun with central and arterial catheters, a bladder catheter was placed, and dopamine was started at a rate of 2– 2.5 μg/kg/min. Filtered packed erythrocyte transfusion was given.
A renal ultrasound showed normal sized kidneys with no evidence of renal vein thrombosis. A peritoneal catheter was placed and peritoneal dialysis was started within 24 h. Intravenous fluid intake was strictly regulated on fluid balance, and antihypertensive therapy was added (nifedipine). Vancomycin plus amikacin were started based on presumptions basis before definitive diagnosis; dosages were carefully titrated according to plasma levels. Tracheal intubation and ventilatory support were necessary in the first 48 h because of the hypoventilation related to the lethargic state and abdominal distention. After 4 days of anuria, spontaneous diuresis was obtained (0.4–0.5 ml/kg/h). Hemoglobin level (10.7 g/dl) and platelet count (60 × 109/l) gradually increased (fig. 1). Blood, urine and stool cultures were negative. On day 8, despite clinical and nephrological improvement (spontaneous urine output >0.5 ml/kg/h) and the absence of fluid overload, the child suddenly developed worsening respiratory distress. Tachypnea (RR 35–50/min), intercostal retractions, chest rales and hypoxemia rapidly occurred precipitating into acute respiratory failure. At laryngoscopy, copious amounts of fresh blood were suctioned from the trachea and she was intubated. At this time, platelet count was 85 × 109/l and coagulative status was as follows: prothrombin activity 83%, aPTT 26 s, fibrinogen 191 mg/dl, ATIII 86%, D dimer 308 ng/ml. No mucosal and/or skin bleeding were present. A chest radiogram showed diffuse opacification of both lung fields with basal infiltrates (fig. 2). On echocardiography, no heart enlargement or signs of pulmonary hypertension were observed; the left ventricular ejection fraction was within normal limits (0.72). On subsequent days, an acute respiratory distress syndrome (ARDS) picture developed, and high-level ventilatory support was applied (fig. 3), together with full-dose (intravenous methylprednisolone at 10 mg/kg body weight/daily for 3 days, followed by 5 mg/kg/day until pulmonary improvement was obtained) steroid therapy. An ENT evaluation showed no abnormalities of the upper respiratory tract, and serial fibroscopic examinations of the lower airways and a high-definition chest CT scan failed to identify any bleeding site. Respiratory physical activity was actively performed, including prone positioning, to improve or prevent further atelectasis. After 1 week, the child progressively improved, permitting ventilatory support to be decreased. Tracheal extubation was obtained on day 20: at this time, spontaneous diuresis had fully recovered and the peritoneal catheter could be removed. On discharge from PICU, blood coagulation tests were normal, hemoglobin level and platelet count were stable with a good blood pressure control (BP about 105/75). By day 25, her renal function had normalized (serum creatinine 0.25– 0.5 mg/dl), without electrolyte disorders or metabolic acidosis. For persistent anemia (about 7 g/dl), erythropoietin therapy was introduced. No signs of further hemolysis or iron deficiency were observed. Echocardiogram and EEG were normal. Neurological and lung sequelae were not observed.
Fig. 1. Time course of LDH, platelet count, creatinine and BUN levels during the PICU stay of the patient. Peritoneal dialysis (square) and clinical ARDS (arrow) periods are also evidenced.
Fig. 2. Chest X-ray film of the baby with ARDS showing diffuse bilateral infiltrates following pulmonary hemorrhage (a) and after tracheal extubation (b).
Fig. 3. Time course of ventilatory parameters during the treatment of ARDS: PaO2:FiO2 ratio is indicated on the left scale, whereas positive end-expiratory pressure (PEEP) and peak inspiratory pressure (PIP) levels appear on the right scale (mm Hg). The time course of the oxygenation index (OI) has been added as well.
HUS is one of the leading causes of ARF in children, and one for which early treatment with peritoneal dialysis (PD) may be indicated. Pleural effusions represent the most common respiratory complication of PD, occurring in about 20% of HUS-affected children receiving PD . Regarding pathophysiological mechanisms, it is well accepted that microvascular damage in HUS can lead to loss of capillary integrity: in fact, increased capillary permeability due to microvascular thrombi and endothelial damage represent the common landmark of this disorder. For unexplained reasons, this mechanism is rarely responsible for primary pulmonary damage: although pulmonary microthrombi are often found on autopsy, clinically primary lung involvement as relevant pulmonary edema  or hemorrhage [2, 3] has been unfrequently reported. Primary pulmonary bleeding has probably never been reported as a part of typical (D+)HUS of childhood.
Even severe respiratory compromise in HUS-affected children requiring PD may be caused by pleural effusions (PE), leading to tracheal intubation and mechanical ventilation ; diaphragmatic defects or lymphatic channels from the peritoneal cavity to the pleural space have been suggested. On the contrary, given the extreme rarity, the cause of pulmonary hemorrhage remains unclear: it can occur even in the absence of fibrin thrombus deposition and has been attributed to alveolar wall necrosis . In our case, thin section chest CT scans and serial flexible bronchoscopic examinations failed to identify the site of initial bleeding.
The diagnosis of ARDS was based on severe hypoxemia (PaO2/FiO2 = 70, A-aDO2 = 589), the presence of diffuse infiltrates on the chest film and reduced lung compliance, in the absence of any signs of heart failure or pulmonary hypertension, according to current standards .
The treatment we applied was an ‘open-lung’ ventilatory strategy as for chemically induced ARDS, limiting peak airway pressure and setting PEEP above the lower inflection point, in order to avoid excessive barotrauma. High-dose methylprednisolone therapy was adopted during the first days of mechanical ventilation. No bloody secretions were suctioned from the tracheal tube 72 h after the acute event.
As a whole, acute pulmonary hemorrhage is a rare event in infants and children, in whom the most common etiologies include: cardiac/vascular malformations, infectious processes, mycotoxins, milk protein allergy, immune vasculitis, and trauma. True alveolar hemorrhage should be distinguished from macroscopically bloody fluid caused by laceration of the bronchial mucosa, which is easily detected through flexible bronchoscopy in proximal trachea bronchi. Diffuse pulmonary hemorrhage can lead per se to potentially fatal clinical ARDS: in this setting a therapeutic role has been suggested for high-dose steroids [13, 14]. In our case, a tracheal-bronchial lesion was ruled out as a cause of focal endobronchial bleeding with secondary aspiration. Moreover, the patient had been extubated 4 days before the sudden respiratory failure, and no tracheal decubitus lesion could be demonstrated on bronchoscopy. As seen above, pulmonary hemorrhage may occur in the setting of several vasculitis syndromes (as Wegener’s granulomatosis or microscopic polyangiitis), or systemic lupus erythematosus, rheumatoid arthritis and Goodpasture’s syndrome. Isolated pulmonary capillaritis is a recently described entity in anti-neutrophil cytoplasmic antibody (ANCA)-positive patients with a median age of 30 years: it has been shown to effectively respond to steroid and cyclophosphamide therapy . In our patient, according to the typical features of D(+)HUS, both ANCA, antibasement membrane antibodies and antiphospholipid antibodies were negative, as well as antinuclear antibody (ANA) determinations, both acutely and during a 2-year follow-up. Complement levels (C3, C4, CH50) were normal, thus excluding factor H deficiency as the cause of HUS. Von Willebrand antigen, ristocetin cofactor, protein C and S levels were within normal ranges; activated protein C resistance, prothrombin mutations, ATIII deficiencies were ruled out. The hypothesis of alveolar wall necrosis following vascular thrombosis established in early clinical HUS is intriguing but obviously difficult to prove: prothrombotic events have been recently demonstrated before renal failure in D(+)HUS and are characterized – as in our case – by the coexistence of slightly elevated D dimer and normal PT/PTT and fibrinogen levels .
In spite of the serious pulmonary complication, HUS showed a benign course, with no residual kidney impairment on control clinical assessments; therefore, no renal histology was obtained.
Given the extreme rarity of pulmonary hemorrhage in D(+)HUS, a single case report can only remind this possible – and potentially fatal – complication. As a general rule, children affected by serious and multifaceted illnesses should be treated at tertiary pediatric critical care facilities, in order to reduce mortality and improve general prognosis.
Dr. M. Piastra
Terapia Intensiva Pediatrica
Policlinico Gemelli, L.go Gemelli 8
IT–00168 Roma (Italy)
Tel. +39 06 3015 5203, Fax +39 06 3052 751, E-Mail firstname.lastname@example.org
Received: June 10, 2003
Accepted after revision: October 15, 2003
Number of Print Pages : 5
Number of Figures : 3, Number of Tables : 0, Number of References : 16
Respiration (International Journal of Thoracic Medicine)
Vol. 71, No. 5, Year 2004 (Cover Date: September-October 2004)
Journal Editor: C.T. Bolliger, Cape Town
ISSN: 0025–7931 (print), 1423–0356 (Online)
For additional information: http://www.karger.com/res