Corticosteroids in severe community-acquired pneumonia: the path we choose depends on where we want to get

Severe community-acquired pneumonia is a major cause of admission to intensive care units and its mortality rates remain exceedingly high. In the search for adjunctive therapies, clinicians who were encouraged by available, though limited, evidence prescribed steroids in most patients with severe sepsis or septic shock, including those with community-acquired pneumonia. Current evidence demonstrates that, whereas corticosteroids should not be routinely employed as adjuvant therapy for severe community-acquired pneumonia, there is sufficient equipoise to continue studying the use of corticosteroids.

In this issue of Critical Care, Fernández-Serrano and colleagues [1] evaluate the effi cacy of corticosteroids in the treatment of hospitalized patients with severe community-acquired pneumonia (CAP). Severe CAP is a major cause of admission to intensive care units (ICUs) and, despite optimal antibiotic and supportive treatment, remains associated with exceedingly high mortality rates [2]. For this legitimate reason, intensivists have worked for decades in numerous clinical trials aiming to identify adjunctive strategies for this condition. In this sense, corticosteroids are the prototypical 'ideal candidate' . Corticosteroids present anti-infl ammatory properties that may reduce the intense proinfl ammatory cytokine response and thus modify the occurrence and severity of organ failures [3]. Also, corticosteroids are easily adminis trated and relatively inexpensive and, clinicians believe, do no harm. Given this long list of favorable attributes and the positive results from a multicenter randomized controlled trial (RCT) [4], it is fair of us physicians to want corticosteroids to fi ll the role of 'the ideal adjunctive sepsis therapy' .
In 1958, the Canadian-American economist John Kenneth Galbraith coined the expression 'conventional wisdom' . Th e term was employed to describe concepts that are generally accepted as true without being questioned. Although such ideas are widely held, further examination based on solid evidence proves them to be wrong. Th is may apply to what happened to corticosteroids in sepsis in the last decade.
Physicians encouraged by available evidence and conventional wisdom prescribed steroids in up to 80% of patients with severe sepsis and septic shock [5]. Th e reliance on corticosteroids as the best choice was so strong that it reduced the patient inclusion rate in the Corticus Study and potentially made the study biased for including only patients who, their physicians believed, would not require steroids [6]. Th e willingness to prescribe corticosteroids is disproportionately elevated given the available evidence.
In patients with severe sepsis, clinical trials [6], welldesigned observational studies [7], and data from an international registry [5] could not demonstrate a reduced need for vasopressors or an improvement in organ dysfunctions or survival. In regard to severe CAP, a systematic review performed in 2008 had already pointed out that the available studies do not support the recommendation of corticosteroids as a standard of care [8]. Subsequent observational data [9] and an RCT [10] also failed to show any signifi cant clinical benefi t associated with corticosteroid administration for this subgroup of patients, irrespective of disease severity.
Fernández-Serrano and colleagues [1] present the results of a small (n = 45) single-center RCT evaluating the potential benefi t of a methylprednisolone (MPRD) regimen on outcomes of non-ICU CAP patients, namely

Abstract
Severe community-acquired pneumonia is a major cause of admission to intensive care units and its mortality rates remain exceedingly high. In the search for adjunctive therapies, clinicians who were encouraged by available, though limited, evidence prescribed steroids in most patients with severe sepsis or septic shock, including those with communityacquired pneumonia. Current evidence demonstrates that, whereas corticosteroids should not be routinely employed as adjuvant therapy for severe communityacquired pneumonia, there is suffi cient equipoise to continue studying the use of corticosteroids. the need for invasive or non-invasive mechanical venti lation. MPRD was found to be associated with a faster clinical, radiological, and laboratory resolution of CAP, but with similar mortality. Th e issue of small sample size notwithstanding, fi ve patients (22.7%) from the placebo group needed ICU admission and three required invasive mechanical ventilation, whereas only one from the MPRD arm was admitted to the ICU for non-invasive ventilation, and this resulted in a marked asymmetry between the two groups. Despite these caveats, the authors were able to conclude that MPRD had a positive impact on the clinical course of these very select CAP patients. In our opinion, the results are hard to translate into daily practice because of patient selection and small sample size and should be viewed with a measure of caution.
In view of the current evidence, what is the future of corticosteroids in severe CAP? Is there clinical equipoise to justify new clinical trials in this topic? Given the current studies, it can be said that, although there is insuffi cient evidence to support corticosteroid use as the standard of care in severe CAP, it is still an unresolved matter that deserves further investigation. Future trials should learn from previously published studies to overcome their limitations. Th is does not mean that we should repeat previous studies with larger sample sizes. Instead, we must work on better study designs that make use of a reliable surrogate of outcomes and that represent a true advance in the refi nement of patient selection. Unlike in the examples of severe sepsis and CAP, better patient selection may lead to signifi cant clinical benefi ts for subgroups of patients with severe infections treated with adjunctive corticosteroids. Benefi ts of systemic cortico steroids in septic patients were demonstrated only in well-defi ned infections such as bacterial meningitis in immunocompetent hosts [11] and Pneumocystis jirovecci pneumonia in HIV patients [12]. Nonetheless, despite new scoring systems [13] and biomarkers, these tools are not validated as surrogates of response to corticosteroids [14].
In conclusion, we believe that, while corticosteroids should not be routinely employed as adjuvant therapy for severe CAP, there is suffi cient equipoise to continue studying the use of corticosteroids. Th is is still an open fi eld that may signifi cantly evolve by applying the concept of theragnostics for severe CAP.