Eric Carlier

Alterations of red blood cell shape and sialic acid membrane content in septic patients

ObjectiveTo investigate the relationship between red blood cell (RBC) shape and modifications of RBC membrane protein content in critically ill patients with or without sepsis compared with healthy control volunteers. DesignProspective, observational in vitro study. SettingUniversity-affiliated cell biology laboratory. SubjectsHuman erythrocytes from healthy volunteers and nonseptic and septic intensive care unit patients. InterventionsSialic acid membrane content was measured on isolated RBC membrane proteins by high-performance liquid chromatography. RBC shape, estimated by the spherical index (M2/M1) or by the moment and effect of osmolality on RBC shape, was studied by flow cytometry at 25°C. Glycophorin A content was measured with antiglycophorin antibodies in flow cytometry. Measurements and Main ResultsSialic acid content was lower in the septic than in the nonseptic patients (1.98 ± 0.79, 2.20 ± 0.39 &mgr;g/100 &mgr;g membrane protein, respectively; p = .01) and than in the volunteers (2.71 ± 1.00 &mgr;g/100 &mgr;g membrane protein; p < .001). No significant difference was found in glycophorin A content between septic and nonseptic patients. RBCs from septic patients had a more spherical shape in isotonic solution than those of healthy volunteers, as assessed by a computed spherical index (M2/M1 ratio: 1.68 ± 0.34 vs. 1.95 ± 0.32; p = .001). Only the RBCs of septic patients failed to change their shape in hypo-osmolar solution (M2/M1 ratio: 1.68 ± 0.34 in iso-osmolar, 1.56 ± 0.28 in hypo-osmolar solution; p = .17). There was a significant correlation between the RBC shape evaluated by the spherical index or by the moment of the cytometric histogram and the sialic acid membrane content in all critically ill patients (septic and nonseptic patients) (r2 = .16, p = .01 for the moment, and r2 = .17, p = .01 for the spherical index, respectively). ConclusionsRBCs of septic patients are characterized by a more spherical shape, a decreased capacity of sphericity in hypo-osmolar solution, and a reduction in the sialic acid content of the RBC membrane. These modifications in RBC shape and membrane may contribute to the RBC rheologic abnormalities frequently described in sepsis.

Administration of enoximone in cardiogenic shock

Thirteen patients in severe cardiogenic shock, persisting despite the use of adrenergic agents, were treated with enoximone, a recently available phosphodiesterase inhibitor. Cardiogenic shock was characterized by low cardiac output (less than 2.5 liter.min-1.m-2), elevated pulmonary artery balloon-occluded pressure (greater than or equal to 15 mm Hg), decreased urine output (less than 20 ml.hour-1) and increased blood lactate (greater than or equal to 2.0 mEq.liter-1). Ten patients were mechanically ventilated. A short-term intravenous infusion of 0.5 mg.kg-1 in 20 minutes of enoximone resulted in significant increases in cardiac index (from 1.8 +/- 0.3 to 2.9 +/- 0.3 liter.min-1.m-2, p less than 0.001) and stroke index (from 17.8 +/- 3.3 to 21.9 +/- 5.1 ml.m-2, p less than 0.001) and significant decrease in pulmonary artery balloon-occluded pressure (from 21.7 +/- 5.8 to 19.8 +/- 6.0 mm Hg, p less than 0.01) without a consistent change in mean arterial pressure (from 79 +/- 8 to 76 +/- 9 mm Hg, difference not significant). Enoximone administration decreased arterial oxygen tension (from 108 +/- 42 to 94 +/- 36 mm Hg, p less than 0.01) and increased venous admixture (from 12.8 +/- 6.5 to 16.0 +/- 8.0%, p less than 0.01). In 8 patients, a second infusion of 0.5 mg.kg-1 immediately thereafter amplified these changes. All patients but one survived the episode of cardiogenic shock and 5 patients left the hospital alive. These results indicate that the addition of enoximone to adrenergic agents in the treatment of cardiogenic shock can markedly increase cardiac output and stroke volume without substantial effects on arterial pressure.

Risk factors for infection and molecular typing in patients in the intensive care unit colonized with nosocomial Enterobacter aerogenes

OBJECTIVES To determine the frequency of colonization by Enterobacter aerogenes in patients in the intensive care unit (ICU) for more than 48 hours and to evaluate the risk factors for infection in patients colonized by this bacteria. DESIGN An 8-month prospective study. SETTING A 12-bed medical-surgical ICU in a 450-bed, university-affiliated, tertiary-care hospital in Belgium. METHOD Pulsed-field gel electrophoresis was used to determine the genotypes of E. aerogenes isolates. RESULTS We observed two major clones of E. aerogenes in the ICU. Interestingly, 87.5% of infected patients had the same genomic profile for colonization and infection. Risk factors for infection in this particular population included younger age, prolonged hospital stay, mechanical ventilation, and bronchoscopy. CONCLUSIONS Colonization is a major prerequisite for infection. The identification of risk factors for infection in colonized patients can optimize the quality of treatment in the ICU.

SUCCESSFUL TREATMENT OF SEVERE LEGIONELLA PNEUMONIA WITH IMIPENEM

Sir: Legionella pneumophila remains an important cause of community-acquired pneumonias, with an incidence from 2 to 15%, that require hospitalization [1]. The highest mortality (80 %) is found among immunosuppressed patients [2] but is much lower in healthy patients (20%). In pneumonias complicated by the acute respiratory distress syndrome (ARDS), mortality is over 80% among untreated patients and 50±60 % in treated patients [2]. We describe herein a case of severe Legionella pneumonia successfully treated with imipenem as empirical antibiotic treatment choice. A 55-year-old man was admitted with a 2-day history of fever (40 C), cough and dyspnea. He had adult-onset diabetes mellitus and no smoking habit. On arrival, positive physical findings included subfever (37.6 C) and inspiratory crackles in the base of the right lung on pulmonary auscultation. Blood analysis was characteristic only for renal failure (blood urea nitrogen BUN 46 mg/dl, creatinine 2.5 mg/dl) and inflammatory syndrome (WBC) 14900/ mm3, C-reactive protein 20 mg/dl). Roentgenographic studies revealed pneumonia restricted to the right lower lobe. Sputum showed neutrophils in large numbers but cultures remained negative. He was treated with intravenous amoxicillin-clavulanate (2 g three times daily with diagnosis of community-acquired pneumonia. On day 2, he developed tachycardia, mild hypotension and respiratory distress which required mechanical ventilation. Radiograph of the chest revealed new bilateral infiltrates, right-sided cardiac catheterization and a PaO2/FIO2 ratio < 200, which were compatible with a diagnosis of ARDS. Between days 2 and 5, the patients clinical course worsened with alteration of the PaO2/FIO2 ratio, increase in WBC and renal failure with oliguria requiring hemodialysis. On day 5, antibiotic treatment was substituted for imipenem (1 g twice daily) and amikacin (250 mg/day) with doses in relation with renal failure. New microscopical examination and cultures of induced sputum were negative, but demonstrated great number of neutrophils. All blood culture specimens remained sterile but serology for Legionella pneumophila by immunofluorescence, received a few days later, was positive at 1/16 000 (Gull, Legionella pneumophila Antigen lame; Gull Diagnostics, Louvainla-Neuve, Belgium). A serology control 1 month after this stay in the ICU showed a significant decrease (1/4000). On day 8, the inflammatory syndrome and renal failure decreased. The patient was extubated and treatment with imipenem-amikacin was stopped on day 12. Hemodialysis was discontinued on day 24. Legionella pneumonia is probably the third pathogen responsible for communityacquired pneumonia, but in Belgium the incidence of this infection requiring ICU admission is smaller [3], probably because of the lack of specific clinical presentation [4] and the limited use of diagnostic techniques such as polymerase chain reaction and DNA hybridization. Because Legionella pneumophila is an intracellular gram-negative bacterium, standard in vitro susceptibility testing is unreliable for prediction of drug activity, as it does not measure the ability of the drug to enter the cell [2], but some investigators have reported the effectiveness of imipenem in single case reports [5]. Because imipenem has in vitro activity against all major causes of pneumonia and Legionella pneumophila [5] and because of the lack of specificity of the clinical presentation of Legionella pneumonia [4], we think that imipenem can be used as a possible therapy in severe pneumonia complicated by multiple organ dysfunction syndrome.

Enoximone in low output states following cardiac surgery

Enoximone, a phosphodiesterase F-III inhibitor with inotropic and vasodilating properties, was administered to six patients with a low output state (cardiac index below 2.0 L/min /m2) early after cardiac surgery for valvular replacement (n = 4) or coronary artery bypass grafts (n = 2). Incremental slow (20-minute) intravenous (IV) boluses of enoximone of 0.5 mg/kg in the first three patients and 0.25 mg/kg in the remaining three patients, up to a total dose ranging from 0.25 to 2.5 mg/kg, produced an increase in cardiac index from 1.6 ± 0.1 (mean ± SE) to 2.2 ± 0.1 L/min /m2 (P < .01), an increase in heart rate from 83 ± 7 to 102 ± 13 beats/min (P < .02), a decrease in pulmonary capillary wedge pressure from 20 ± 2 to 15 ± 1 mm Hg (P < .05), and a decrease in mean arterial pressure from 77 ± 4 to 64 ± 6 mm Hg (P < .05). Left ventricular stroke work did not change, 16 ± 1 to 17 ± 3 g/m (P is not significant). Mean arterial pressure decreased to below 60 mm Hg in five patients at various times up to 30 minutes after the last IV enoximone bolus, and this hypotension had to be treated with dopamine alone (n = 3), in combination with 600 mL IV colloids (n = 1), or with norepinephrine alone (n = 1). Arrhythmias occurred in three patients and consisted in sustained ventricular tachycardia (n = 1), ventricular premature beats and a short episode of ventricular tachycardia (n = 1), and supraventricular tachycardia (n = 2). Temporal relationship strongly suggests that both hypotension and arrhythmia were caused by the administration of enoximone in these patients. Thus, enoximone increases cardiac output but appears to be poorly tolerated in low output states following cardiac surgery.

Methylene blue administration in septic shock: A clinical trial

OBJECTIVE A release of nitric oxide has been incriminated in the cardiovascular alterations of septic shock. Since guanylate cyclase is the target enzyme in the endothelium-dependent relaxation mediated by nitric oxide, we studied the acute effects of methylene blue, a potent inhibitor of guanylate cyclase in patients with septic shock. DESIGN Prospective clinical trial. SETTING Medical-surgical intensive care unit in a university hospital. PATIENTS Fourteen patients with severe septic shock requiring adrenergic therapy. INTERVENTIONS Short-term intravenous infusion of methylene blue. MEASUREMENTS AND MAIN RESULTS Hemodynamic measurements were obtained at baseline, and 30, 60, and 90 mins after the infusion of 2 mg/kg of methylene blue. Methylene blue administration was followed by a progressive increase in mean arterial pressure (from 61.1 +/- 7.6 to 71.7 +/- 12.0 mm Hg at 60 mins, p < .01). Pulmonary arterial pressure, cardiac filling pressures, cardiac output oxygen delivery, and oxygen consumption were not significantly affected. Left ventricular stroke work increased from 42.5 +/- 17.9 to 48.9 +/- 14.5 g.m after 60 mins (p < .05). Arterial lactate concentration decreased from 3.4 +/- 1.4 to 2.7 +/- 1.3 mmol/L (p < .05). Since these effects were transient, a second dose of methylene blue was administered 90 mins later to six patients and was followed by a similar response. No adverse effect was observed. CONCLUSIONS In septic shock patients, the administration of methylene blue results in a transient and reproducible increase in arterial pressure, associated with an improvement in cardiac function, but does not increase cellular oxygen availability. The significant reduction in blood lactate concentration is probably related to the reductor effect of methylene blue, rather than to an improvement in tissue oxygenation.