key: cord-0006895-ehoo6fcr authors: Hasibeder, Walter R.; Dünser, Martin W.; Halabi, Milad; Brinninger, Gabriele title: The relationship between extravascular lung water and oxygenation in three patients with influenza A (H1N1)-induced respiratory failure date: 2010-10-20 journal: Wien Klin Wochenschr DOI: 10.1007/s00508-010-1475-1 sha: f8198633e07d861da5af59261119edd2bad7ead6 doc_id: 6895 cord_uid: ehoo6fcr This case series reports the correlation between extravascular lung water (EVLW) and the partial arterial oxygen pressure/fractional inspiratory oxygen (PaO(2)/FiO(2)) ratio in three patients with severe influenza A (H1N1)-induced respiratory failure. All patients suffered from grave hypoxia (PaO(2), 26–42 mmHg) and were mechanically ventilated using biphasic airway pressure (PEEP, 12–15 mmHg; FiO(2), 0.8–1) in combination with prone positioning at 12 hourly intervals. All patients were monitored using the PICCO(®) system for 8–11 days. During mechanical ventilation, a total of 62 simultaneous determinations of the PaO(2)/FiO(2) ratio and EVLW were performed. A significant correlation between EVLW and the PaO(2)/FiO(2) ratio (Spearman-rho correlation coefficient, –0.852; p < 0.001) was observed. In all patients, a decrease in EVLW was accompanied by an improvement in oxygenation. Serum lactate dehydrogenase levels were elevated in all patients and significantly correlated with EVLW during the intensive care unit stay (Spearman-rho correlation coefficient, 0.786; p < 0.001). In conclusion, EVLW seems increased in patients with severe H1N1-induced respiratory failure and appears to be closely correlated with impairments of oxygenatory function. Summary. Th is case series reports the correlation between extravascular lung water (EVLW) and the partial arterial oxygen pressure/fractional inspiratory oxygen (PaO 2 /FiO 2 ) ratio in three patients with severe infl uenza A (H1N1)-induced respiratory failure. All patients suff ered from grave hypoxia (PaO 2 , 26-42 mmHg) and were mechanically ventilated using biphasic airway pressure (PEEP, 12-15 mmHg; FiO 2 , 0.8-1) in combination with prone positioning at 12 hourly intervals. All patients were monitored using the PICCO® system for 8-11 days. During mechanical ventilation, a total of 62 simultaneous determinations of the PaO 2 / FiO 2 ratio and EVLW were performed. A signifi cant correlation between EVLW and the PaO 2 /FiO 2 ratio (Spearmanrho correlation coeffi cient, -0.852; p < 0.001) was observed. In all patients, a decrease in EVLW was accompanied by an improvement in oxygenation. Serum lactate dehydrogenase levels were elevated in all patients and signifi cantly correlated with EVLW during the intensive care unit stay (Spearman-rho correlation coeffi cient, 0.786; p < 0.001). In In April 2009, a novel infl uenza A virus of swine-origin (H1N1) virus was identifi ed in two separate patients in the United States [1] . Since then, infl uenza A (H1N1) infection has rapidly spread over Central/South America, the United States, Australia and Europe [2] [3] [4] . Th e clinical spectrum of the disease is highly variable ranging from mild self-limited fl u-like symptoms to severe respiratory failure and death. By August 10, 2010, the World Health Organization stated that the infl uenza A (H1N1) outbreak has moved into the post-pandemic period [5] . As of August 1, 2010, at least 18,449 deaths due to infections with H1N1 virus had been reported from 214 countries [6] . Th e relationship between EVLW and oxygenation in three patients wkw 21-22/2010 © Springer-Verlag case report Because of the recent emergence of the disease, pathophysiology of H1N1-associated respiratory failure remains incompletely understood. In this case series, we report on the relationship of extravascular lung water (EVLW) and oxygenation failure in three patients with severe respiratory failure due to infl uenza A (H1N1) infection. All patients or their next of kins gave written informed consent that the course of their disease is published anonymously. Between Nov. 2009 and Feb. 2010, 37 patients with complicated infl uenza A (H1N1) disease were admitted to our hospital. H1N1 infection was confi rmed by nasopharyngeal swabs analyzed with a real-time reverse-transcriptase polymerase chain reaction assay (Artus Infl uenza/H1®; Quiagen, Hilden, Germany). Th ree patients (2 females, 1 male) entered the intensive care unit because of severe hypoxia and were intubated and mechanically ventilated (Table 1 ). Further diagnostic work-up included blood, tra-cheobronchial and urine bacterial cultures, chest X-ray, chest computertomography-scan as well as determination of antinuclear antibodies and antineutrophil cytoplasmatic antibodies. Oseltamivir (Tamifl u®; Hoff mann-La Roche AG; Grenzach-Wyhlen, Germany) was administered at 150 mg bid via a nasogastric tube. Haemodynamic monitoring included an arterial and central venous line. In all patients, cardiac output, intrathoracic blood volume and EVLW were determined using the transpulmonary thermodilution method (PICCO®; PULSION Medical Systems; Munich, Germany). Haemodynamic therapy was guided by an institutional protocol [7] . In all patients, EVLW measurements and blood gas analyses were performed at least 2-4 times daily. Fluid balances were calculated between EVLW measurements. Laboratory investigations including blood cell count, determination of C-reactive protein, procalcitonin, lactate dehydrogenase, liver enzymes, serum creatinine, troponin I were documented once daily. To evaluate the correlation between EVLW and the partial arterial oxygen pressure/fractional inspiratory oxygen concentration (PaO 2 /FiO 2 ) ratio, bivariate correlation analyses applying the Spearman-rho correlation coeffi cient were calculated (SPSS 12.0 Software; SPSS Inc. Chicago, Illinois). Furthermore, correlations between EVLW and lactate dehydrogenase levels, fl uid balance as well as the positive endexpiratory pressure (PEEP) were calculated. P-values <0.05 were considered to indicate statistical signifi cance. Before hospital admission all patients reported high fever, dry cough and sore throat for one week or longer. Patient 2 complained about upper abdominal pain and diarrhoea fi nally leading to dehydration and shock. Two patients were treated with broad spectrum antibiotics, and all patients took non-steroidal antirheumatic medications. None demonstrated particular risk factors suggesting a severe course of H1N1 infection. In all patients, chest radiographs showed bilateral infi ltrates consistent with multilobar pneumonia or the acute respiratory distress syndrome. Chest computertomography displayed diff use interstitial and patchy alveolar infi ltrates involving both lungs with accentuation in dependent regions (Fig. 1 ). Patient characteristics, clinical and laboratory parameters at hospital admission are summarized in Table 1 . At ICU admission, all patients were severely hypoxic with PaO 2 values in the range of 26-42 mmHg. Patient 2 presented with shock, leucopenia, elevated serum procalcitonin and troponin I levels. Patient 3 was in oliguric renal failure necessitating renal replacement therapy. EVLW was increased in all patients (Fig. 2) and significantly correlated with the PaO 2 /FiO 2 ratio during the intensive care unit stay (Fig. 2) . Clinical signs of lung oedema were present in all patients at intensive care unit admission. While EVLW gradually decreased in patients 1 and 3, it initially decreased in patient 2 but escalated again on day 5. Simultaneously, the PaO 2 /FiO 2 ratio deteriorated and refractory multiple organ failure developed in this patient. Serum lactate dehydrogenase levels were elevated in all patients and signifi cantly correlated with EVLW during the intensive care unit stay (r = 0.786, p < 0.001). No correlations between EVLW and PEEP (r = 0.332, p = 0.07) or fl uid balance (r = 0.308, p = 0.12) were observed. [8] . Since our report included only three patients, no conclusions can be made whether EVLW in H1N1-induced respiratory failure is disproportionally higher than in lung failure from other aetiologies. Studies in diff erent critically ill patient populations reported EVLW measurements in the ranges as observed in our patients during the intensive care unit stay [9] . However, EVLW measurements recorded at intensive care unit admission in our patients were higher than those, for example, reported in patients with sepsis-induced acute lung injury or acute respiratory distress syndrome [10, 11] . Considering that PEEP can eff ectively reduce EVLW [12, 13] , one may speculate that the comparably low PEEP values applied in our patients may have contributed to EVLW accumulation in the patients presented. In our patients, EVLW was closely correlated with simultaneous measurements of the PaO 2 /FiO 2 ratio suggesting a potential key role in the impairment of oxygenatory function in patients with acute respiratory failure due to H1N1 infection. A similar though weaker correlation between EVLW and the PaO 2 /FiO 2 ratio was reported in septic shock patients with acute lung injury or acute respiratory distress syndrome [14] . In addition, EVLW was shown to correlate well with survival and independently predicted outcome in a general critically ill patient population [15] . In line with Th e relationship between EVLW and oxygenation in three patients wkw 21-22/2010 © Springer-Verlag this report, patient 2 in our report exhibited the highest EVLW values (36 mL/kg) and subsequently died. Several main mechanisms might be responsible for high EVLW values and the development of lung oedema in our patients: First, high virus loads in the lower respiratory tract were shown to induce excessive pulmonary infl ammation [16] . A virus-induced "cytokine storm" has been reported for acute respiratory failure due to the infl uenza A (H5N1) virus [17] . Second, infl uenza virus-induced inhibition of amiloride-sensitive sodium channels in respiratory epithelia has specifi cally been reported to impair lung fl uid clearance and promote the formation of lung oedema [18, 19] . Th ird, we observed a close correlation between EVLW and serum levels of lactate dehydrogenase. Elevated lactate dehydrogenase levels have been reported in patients infected with infl uenza A (H1N1) and may refl ect pathogen-induced tissue break-down in the lungs or other tissues such as the musculature [2, 20, 21] . Our results indicate that infl uenza A (H1N1) infection may cause relevant lung tissue destruction thus contributing to lung oedema formation and impaired oxygenation. EVLW in patients with H1N1-induced respiratory failure may serve as a useful surrogate marker of the severity of lung damage. Evolution of EVLW during the clinical course of the disease may guide intensivists in their therapeutic decisions including ventilator settings [9] and lung recruitment [22] , fl uid balance [9, 23] and prone positioning [24] . In conclusion, EVLW seems increased in patients with severe H1N1-induced respiratory failure and appears to be closely correlated with impairments of oxygenatory function. Th is work was supported by institutional funds. No author has a potential confl ict of interest in regards of drugs or techniques discussed in this manuscript. Novel Swine-origin Infl uenza A (H1N1) Virus Investigation Team. Emergence of a novel swine-origin infl uenza A (H1N1) virus in humans Pneumonia and respiratory failure from swine-origin infl uenza A (H1N1) in Mexico Critically ill patients with 2009 infl uenza A (H1N1) infection in Canada Critical care services and 2009 H1N1 infl uenza in Australia and New Zealand WHO Director General. H1N1 in post-pandemic period World Health Organization. Pandemic (H1N1) 2009 -update 112 How to protect the heart in septic shock: a hypothesis on the pathophysiology and treatment of septic heart failure Monitoring extravascular lung water in acute respiratory distress syndrome induced by probable 2009 pandemic infl uenza A (H1N1) virus: report of two cases Bedside assessment of extravascular lung water by dilution methods: temptations and pitfalls Extravascular lung water to blood volume ratios as measures of permeability in sepsis-induced ALI/ARDS Extravascular lung water indexed to predicted body weight is a novel predictor of intensive care unit mortality in patients with acute lung injury Ventilation with positive end-expiratory pressure reduces extravascular lung water and increases lympathic fl ow in hydrostatic pulmonary edema PEEP decreases atelectasis and extravascular lung water but not lung tissue volume in surfactant-washout lung injury Correlation between extravascular lung water and oxygenation in ALI/ARDS patients in septic shock: possible role in the development of atelectasis? Prognostic value of extravascular lung water in critically ill patients Acute lower respiratory tract infection Fatal outcome of human infl uenza A (H5N1) is associated with high viral load and hypercytokinemia Infl uenza virus inhibits amiloride-sensitive Na+ channels in respiratory epithelia Infl uenza virus inhibits EnaC and lung fl uid clearance Infl uenza A pandemics: clinical and organizational aspects: the experience in Chile Clinical characteristics of Korean pediatric patients critically ill with infl uenza A (H1N1) virus Lung recreuitability in ARDS H1N1 patients Comparison of two fl uidmanagement strategies in acute lung injury Infl uence of support on intra-abdominal pressure, hepatic kinetics of indocyanine green and extravascular lung water during prone positioning in patients with ARDS: a randomized crossover study