Today was a big day at the European Society of Intensive Care Medicine (ESICM) International Conference. Today was the Hot Topics session with the results from a number of studies being presented, many of which were published online simultaneously with the presentations. I would like to take the opportunity to highlight four such trials presented in this Hot Topics session today: 1) The EMPIRICUS Trial (Empiric Micafungin in critically ill septic patients colonized with candida and with multiple organ dysfunction); 2) The OXYGEN-ICU Trial (Conservative vs. conventional oxygen therapy in ICU patients); 3) LeoPARDS Trial (Levosimendan in Septic Shock); and 4) High Flow Nasal Cannula vs Non-Invasive Ventilation Post-Extubation in High Risk Patients.
The EMPIRICUS trial was published online this morning in JAMA. This is an interesting trial that randomized 260 non-neutropenic, non-immunocompromised septic patients with multisystem organ dysfunction who were colonized with candida to micafungin or placebo in a double-blinded fashion (Timsit JF, et al. JAMA. 2016: Epub ahead of print, October 5, 2016). The micafungin was dosed 100mg once daily for 14 days. The primary endpoint was invasive fungal infection free survival through day 28. Secondary endpoints included invasive fungal infections, 28 and 90 day mortality, organ failures, ventilator-free days, and development of bacterial VAP. Despite decreasing the incidence of invasive fungal infections, daily micafungin did not increase the incidence of invasive fungal infection free survival to day 28, mortality at day 28 or 90, or decrease new organ failures. Given the negative results, it is important to understand which patients were enrolled. In order to be eligible, patients needed to be ventilated for at least 5 days, have candida colonization from at least one (non-rectal swab) site, have one or more organ dysfunctions, have been treated with broad spectrum antibiotics for at least 4 of the previous 7 days, and have a central venous or arterial catheter. Neutropenic and immunosuppressed patients were excluded, along with those already receiving antifungal treatment with an echinocandin. In general, these are patients who I empirically start antifungal coverage on in my practice – and unfortunately, it appears that starting micafungin empirically in these patients is not improving their outcomes overall.
The OXYGEN-ICU Trial (Conservative vs. Conventional Oxygen Therapy in Critically Ill Patients) is another potential practice changing trial presented in the Hot Topics session and published online in JAMA today. A number of observational studies have found an association between hyperoxia and worse outcomes (survival and/or neurological outcomes) in patients with ARDS, after cardiac arrest, and in those treated with therapeutic hypothermia. Unlike the previous observational studies, the OXYGEN-ICU trial investigates the potential effect of hyperoxia in a randomized controlled fashion (Girardis M, et al. JAMA. 2016; Epub ahead of print, October 5, 2016). The single-center, open-label study randomized patients admitted to the ICU and expected to stay 72 hours or longer to receive oxygen therapy targeting a PaO2 between 70 and 100 mm Hg (or SpO2 94-98%) (the conservative group) versus oxygen therapy targeting PaO2 values up to 150 mmHg or SpO2 values 97-100% (conventional control group). The primary endpoint was ICU mortality with incidence of new organ failures and infections after 48 hours of ICU admission as secondary outcomes. Unfortunately, due to low enrollment, the trial was stopped early after 480 of a planned 660 patients were enrolled. Oxygen levels were significantly lower in the conservative group (median PaO2 87 vs 102 mm Hg with FiO2 0.36 vs. 0.39). About 2/3rds of the patients were surgical critically ill patients. 30% of the patients had shock, 15% had renal failure at enrollment, and 67% were mechanically ventilated. The conservative oxygen treatment group had significantly lower ICU mortality (11.6% vs. 20.2%; P=0.01) than the conventional oxygen therapy group. Conservative oxygen therapy also had lower hospital mortality (with a log rank test p-value of 0.02 for the Kaplan Meier analysis), less development of shock and liver failure. In addition, patients treated with conservative oxygen protocol were liberated from the ventilator faster. These results are fascinating to me – they fit my prior belief that hyperoxia may very well be detrimental to our critically ill patients and that we should be considerably more aggressive in weaning oxygen. Beyond that, though, there are a number of interesting aspects of this trial, including that 1/3rd of patients were not ventilated and that it was open-label at a single center. It needs replication, although it is free and fairly easy to implement in clinical practice. I will evaluate this study and the topic of hyperoxia in more detail in a separate blog post in the near future.
The LeoPARDS trial evaluated the effect of Levosimendan on organ failure in patients with septic shock (Gordon AC, et al. NEJM. 2016; Epub ahead of print, October 5, 2016). Levosimendan is a calcium sensitizer which functions as an inotrope. It is commonly used in Europe in patients with septic shock, but is not yet FDA approved for use in the United States. This double blind, placebo-controlled phase IIb trial randomized 516 patients from 34 UK ICUs with resuscitated septic shock, treated with norepiphrine, to Levosimendan or placebo for 24 hours. Adult patients were eligible if they had septic shock and had received norepinephrine for at least 4, but no more than 24, hours. Patients with end stage renal disease or significant liver dysfunction were excluded. Levosimendan was started at 0.1 mcg/kg/min and doubled after 2-4 hours if the patient did not have worsening tachycardia or hypotension. The primary outcome was mean SOFA score in the ICU. Secondary outcomes included 28-day mortality, time to wean from mechanical ventilation, and adverse events. Enrolled patients were sick with mean APACHE II score of 25, 80% mechanically ventilated, mean lactate of 2.2 and norepi dose of about 20 mcg/min. Levosimendan increased heart rate, increased the incidence of SVT, slightly lowered blood pressure, and increased the dose of norepinephrine. It also resulted in higher mean SOFA scores while in the ICU (6.68 vs. 6.06; P=0.053). This did not change in adjusted analysis for age, severity of illness, and type of ICU. Levosimendan worsened the cardiovascular component of the SOFA score (4.43 vs. 4.01) and delayed time to weaning from the ventilator. Overall, 28-day mortality was not significantly different between the groups. These results are disappointing and represent another failure in the number of pharmaceutical agents evaluated for the treatment of sepsis. I don’t know if this study will result in less use of Levosimendan in Europe, but it isn’t going to be helpful in gaining FDA approval in the United States.
The last study presented in the Hot Topics session of the ESICM conference I would like to discuss is a comparison of two different respiratory therapies after extubation (Hernandez G, et al. JAMA. 2016; Epub ahead of print October 5, 2016). This trial enrolled 604 patients from 3 Spanish ICUs in a comparison of high flow nasal cannula (HFNC) versus non-invasive ventilation (NIV) in patients at high risk for reintubation after extubation for treatment of respiratory failure. Patients were considered at high risk if they had any of the following: age greater than 65 years old, APACHE II score higher than 12, BMI above 30, inadequate secretion management, prolonged weaning, more than one comorbidity, heart failure or COPD as etiology of respiratory failure, or airway patency problems. HFNC or NIV using a full-face mask were started at the time of extubation and continued for 24 hours. HFNC was started at 10L/min flow and increased slowly until patients expressed discomfort. Primary outcomes were reintubation within 72 hours and postextubation respiratory failure. The study was designed as a non-inferiority study (non-inferiority boundary of 10% for reintubation) and the results are a bit confusing. HFNC demonstrated non-inferiority to NIV with regards to any reintubation (19.1% in NIV vs. 22.8% in HFNC) and had almost identical respiratory related reintubations. However, more patients experienced respiratory failure in the NIV group compared to the HFNC group (39.8% vs. 26.9%). Median time to reintubation was not different between groups, but median ICU length of stay was lower in the HFNC group (3 vs. 4 days; P=0.048). Honestly, while an important study, I don’t really know what to think of these results. While non-inferior, HFNC did have more reintubations, although not for respiratory failure. NIV had more respiratory failure and longer ICU lengths of stay. Furthermore, I don’t know from this study if either of these are better than conventional oxygen, because it wasn’t one of the treatment arms. While intriguing, I don’t see these study results changing my practice at this time.
Thanks for hanging with me and making it through all the study results. I am interested in hearing any thoughts you might have.