Corticotherapy (the administration of

Transcription

1 Critical Care An Update on the Diagnosis of Adrenal Insufficiency and the Use of Corticotherapy in Critical Illness Zachariah Thomas and Gilles L Fraser Corticotherapy (the administration of exogenous glucocorticoids) 1 in critically ill patients has been debated since 1940 when Perla and Marmorston 2 demonstrated that adrenal extracts were beneficial in infections such as malaria and pneumonia (synthetic glucocorticoids were not available for clinical use until 1948). 3 Sepsis, 4 acute respiratory distress syndrome (ARDS), 5 and severe community-acquired pneumonia (SCAP) 6 are common intensive care unit (ICU) diagnoses associated with increased proinflammatory cytokine production; thus, modulation of the inflammatory process with steroids would seem like a viable treatment approach. In addition, inflammatory cytokines are known to inhibit cortisol release and may predispose patients to adrenal insufficiency. 4 Sixty years of research have not produced a consensus regarding the appropriate adrenal response to critical illness or the identification of patients who are likely to benefit from corticotherapy. In addition, the optimal dose and duration of corticotherapy is yet to be defined. In 2004, Burry and Wax 7 published a comprehensive discussion of corticosteroids that focused on septic shock. This review serves as an update regarding the diagnosis of Author information provided at the end of the text. OBJECTIVE: To examine recent literature regarding corticotherapy in critically ill patients suffering from sepsis, acute respiratory distress syndrome (ARDS), and severe community-acquired pneumonia (SCAP). DATA SOURCES: Literature was identified through MEDLINE (1966 April 2007) using combinations of the key words hydrocortisone, adrenal insufficiency, acute respiratory distress syndrome, pneumonia, sepsis, and cortisol. Bibliographies of relevant articles were reviewed for additional citations. Presentations at recent critical care meetings were also incorporated. STUDY SELECTION AND DATA EXTRACTION: Articles were chosen based upon their relevance to the topics covered. DATA SYNTHESIS: Earlier studies using high-dose corticotherapy in the intensive care unit have shown treatment to be ineffective. Recent studies using extended courses of low-to-moderate doses of steroids have found favorable results; however, these results must be interpreted with caution due to limitations in the data. One trial of steroids in septic shock found a survival benefit in patients who failed to increase their baseline cortisol by greater than 9 µg/dl in response to adrenocorticotropic hormone, but these results were not reproduced in a subsequent Phase 3 trial. Recently, inaccuracies in measuring cortisol have been identified, making interpretation of cortisol concentrations difficult. A large-scale study failed to confirm a previously reported mortality benefit of corticotherapy in late ARDS, but preliminary data suggest a role for steroid treatment in early ARDS. Finally, a pilot study has found that hydrocortisone lowers morbidity and mortality in SCAP. CONCLUSIONS: Corticotherapy may be beneficial to some patients with sepsis. The decision to administer steroids in sepsis cannot be based on biochemical markers of adrenal function; rather, treatment should be considered in septic patients with vasopressor refractory hypotension. Although preliminary evidence suggests a role for steroids in early ARDS and SCAP, there are not enough data to suggest routine administration of steroids in these conditions. Additional studies are needed to assess corticotherapy in the critically ill. KEY WORDS: acute respiratory distress syndrome, adrenal insufficiency, cortisol, hydrocortisone, pneumonia, sepsis. Ann Pharmacother 2007;41: Published Online, 14 Aug 2007, DOI /aph.1K082 THIS ARTICLE IS APPROVED FOR CONTINUING EDUCATION CREDIT ACPE UNIVERSAL PROGRAM NUMBER: H The Annals of Pharmacotherapy 2007 September, Volume 41

2 adrenal insufficiency in sepsis and also examines recent advances in the use of corticotherapy in sepsis, ARDS, and SCAP. Data Sources Literature was identified through MEDLINE (1966 April 2007) using combinations of the key words hydrocortisone, adrenal insufficiency, acute respiratory distress syndrome, pneumonia, sepsis, and cortisol. Additional citations were obtained from bibliographies of relevant articles. Presentations at recent intensive care medicine meetings were also incorporated. Advances and Uncertainties in the Diagnosis of Adrenal Insufficiency Although the term relative adrenal insufficiency can be found in the literature as early as the 1950s, 8 it was not incorporated into the vernacular of medicine until Rothwell et al. 9 demonstrated that a failure to increase basal cortisol levels by greater than 9 µg/dl (nonresponse) after a 250 µg adrenocorticotropic hormone (ACTH) challenge was associated with increased mortality in septic shock. In this study, all nonresponders died. In contrast, the mortality rate was 32% in patients with a normal response to ACTH (p < 0.001). Of note, APACHE II scores and basal cortisol concentrations were not different between survivors and nonsurvivors: 22 versus 25 and 23.7 µg/dl versus 27.5 µg/dl, respectively. Given the normal/elevated basal serum cortisol concentrations (reference range 6 24 µg/dl), 10 Rothwell speculated that patients who fail to respond to an ACTH challenge may have adrenal insufficiency relative to their degree of illness. These findings were corroborated nearly 10 years later in a cohort of 189 patients with refractory septic shock (systolic blood pressure <90 mm Hg for at least 1 h despite fluid resuscitation and vasopressor therapy). 11 Using Rothwell s definition, there were 103 nonresponders in this study. Nonresponders had significantly higher mortality (73%) compared with those with a normal response to ACTH (40%). Despite these data, many have suggested that the ACTH test is not appropriate for the diagnosis of adrenal insufficiency in septic shock A major criticism of the ACTH test is that it assesses adrenal gland function directly, thereby potentially missing the diagnosis of adrenal failure secondary to hypothalamic or pituitary failure (secondary adrenal insufficiency). 12 Until recently, there have been no studies to evaluate the ACTH test in sepsis to a reference standard of hypothalamic pituitary adrenal (HPA) axis integrity such as the metyrapone test. Metyrapone inhibits the enzyme 11β-hydroxylase, preventing the conversion of 11-deoxycortisol to cortisol (Figure 1). 17,18 Lower levels of cortisol stimulate ACTH release, which results in an increase in cortisol precursors proximal to the site inhibition. The metyrapone test is considered diagnostic of adrenal insufficiency if the 11-deoxycortisol concentration at 0800 hours is less than 7 µg/dl in the presence of a serum cortisol concentration less than 8 µg/dl. 10 A serum cortisol level less than 8 µg/dl ensures that adequate suppression of 11β-hydroxylase by metyrapone was achieved. ACTH concentrations should increase by greater than 150 pg/ml in response to the inhibition of cortisol synthesis. An increase in ACTH less than 150 pg/ml confers a diagnosis of secondary adrenal insufficiency. 19 Annane et al. 20 have published the first study validating the diagnostic value of the 250 µg ACTH stimulation test with the metyrapone test in septic patients. The ACTH test was conducted in a cohort of 61 patients who met the American College of Chest Physicians/Society of Critical Care Medicine 21 criteria for severe sepsis/septic shock upon ICU admission. An overnight metyrapone test was then conducted and cortisol, ACTH, and 11-deoxycortisol concentrations were measured the following morning. Likelihood ratios predictive of adrenal function were determined for baseline, ACTH-stimulated, and delta of total and unbound cortisol levels. Forty-four critically ill patients without sepsis served as controls. The findings of the original cohort were then applied to a second cohort of patients with severe sepsis/septic shock (n = 40) for validation purposes. Approximately 60% of septic patients were diagnosed with adrenal insufficiency using metyrapone, compared with only 7% in the control group. 20 The majority of cases met criteria for secondary adrenal insufficiency. The cortisol values associated with adrenal insufficiency are listed in Table 1. As shown in the table, despite not testing the entire HPA axis, nonresponse to ACTH correlated well with the metyrapone test. Validation of the Adrenocorticotropic Hormone Test with Metyrapone Figure 1. Simplified depiction of cortisol synthesis. Metyrapone inhibits the enzyme 11β-hydroxylase, preventing the conversion of 11-deoxycortisol to cortisol. The Annals of Pharmacotherapy 2007 September, Volume

3 Z Thomas and GL Fraser Hamrahian et al. 22 have suggested the use of free cortisol concentrations to diagnose adrenal insufficiency in critically ill patients. Free cortisol cortisol not bound to albumin or cortisol-binding globulin is thought to be responsible for the physiologic actions of cortisol. Free cortisol is of special interest in ICU patients since hypoalbuminemia is a common occurrence. Using metyrapone, Annane et al. 20 demonstrated that free cortisol levels could be used to diagnose adrenal insufficiency in sepsis. However, despite theoretical advantages, free cortisol was not superior to total cortisol levels. Interpretation of Baseline Cortisol Concentrations The question of what is an adequate baseline cortisol continues to be debated. 23,24 Marik and Zaloga 12,23 have proposed baseline cortisol concentrations greater than 25 µg/dl as indicative of normal adrenal function in sepsis. However, based upon the metyrapone test, this value was not associated with any diagnostic significance (likelihood ratio = 0.95; 95% CI 0.43 to 2.10). 20 Annane et al. 20 suggested that adrenal insufficiency is unlikely if the baseline cortisol or the ACTH-stimulated cortisol is greater than or equal to 44 µg/dl. However, this assertion is based on a small number of patients. In the metyrapone study, approximately 7 patients had baseline cortisol concentrations at that level. One patient with a baseline cortisol of approximately 80 µg/dl was diagnosed with adrenal insufficiency. A post-acth cortisol level of 44 µg/dl or more was associated with a likelihood ratio of This likelihood ratio does not describe a robust relationship and is not generally considered useful in terms of ruling out a disease. 25 Clearly, it is difficult to define normal adrenal function based upon a baseline cortisol level, but it seems that baseline cortisol can be used to diagnose adrenal dysfunction with great accuracy. In the metyrapone study, all patients Table 1. Cortisol Cutoff Values Associated with Diagnostic Significance 17 Cutoff Values LR 95% CI Adrenal insufficiency likely basal total cortisol <10 µg/dl delta total cortisol <9 µg/dl to delta free cortisol <2 µg/dl to 9.54 basal cortisol <10 µg/dl or delta 9 µg/dl to basal free cortisol <0.8 µg/dl or delta <2 µg/dl to Adrenal insufficiency less likely a delta total cortisol 16.8 µg/dl to 0.87 post-acth stimulated total cortisol 44 µg/dl to 0.99 ACTH = adrenocorticotropic hormone; LR = likelihood ratio. a The strongest predictor of normal adrenal function was the combination of a post ACTH-stimulated cortisol concentration 44 µg/dl and a delta cortisol 16.8 µg/dl; however, the likelihood ratio was not provided. with baseline cortisol levels less than 10 µg/dl were found to have adrenal insufficiency. The cortisol values determining adrenal insufficiency in the original septic cohort were able to predict adrenal insufficiency with 83% sensitivity and 88% specificity in the validation cohort. Although patients with severe sepsis were included in the metyrapone study, most (72%) patients had septic shock as evidenced by vasopressor use at baseline. 20 Therefore, the findings of the validation cohort should not be applied to patients with sepsis in the absence of shock. It remains unclear whether the ACTH test is a useful diagnostic tool in all septic patients. High- Versus Low-Dose ACTH Test Since the adrenal glands can be maximally stimulated with 5 10 µg of ACTH, 10 administration of 250 µg of ACTH produces levels of corticotropin that far exceed physiologic levels. 26 This can lead to a diagnosis of normal adrenal function in patients with mild secondary adrenal insufficiency 10,20,26 and is of great concern, since up to 80% of sepsis-induced adrenal dysfunction is due to secondary adrenal failure. 20 Thus, many investigators have proposed that the high-dose (250 µg) ACTH test be replaced with a more physiologic test (1 µg ACTH). 12,23 Using a definition of adrenal insufficiency as a delta cortisol level of 9 µg/dl or less, Siraux et al. 27 have shown that approximately one-third of patients with a normal response to the high-dose ACTH stimulation test will be nonresponders to the 1 µg ACTH test. In this study, patients with discordant results had a trend toward increased mortality (p = 0.058). Currently, there are no data to refute or support use of corticotherapy in patients who are nonresponders to the 1 µg ACTH test or in those who have discordant results between the 1 µg and 250 µg tests. A meta-analysis of primarily non-icu patients failed to show a difference between the low and high dose ACTH tests for the diagnosis of adrenal insufficiency. 28 This may be due to the fact that, although the high-dose test produces higher levels of corticotropin ( pg/ml) than the low-dose test ( pg/ml), the results of both tests are supraphysiologic and thus stimulate the adrenals maximally. 26,29 However, since the majority of the data in ICU patients have been synthesized using the 250 µg dose, it would seem prudent to recommend against the use of the 1 µg test until further studies are done. 30,31 Variabilities in Endogenous Cortisol Secretion and Measurement The diagnosis of adrenal insufficiency is further complicated by variability in cortisol secretion. Venkatesh et al. 32 measured serum cortisol hourly for 24 hours in septic patients (n = 21) and then conducted a 1 µg ACTH test. Significant hour-to-hour variability (coefficient of variation 1458 The Annals of Pharmacotherapy 2007 September, Volume 41

4 Corticotherapy in Critical Illness 8 30%) was noted. Consistent with other reports, 12 no discernable circadian rhythm was seen in these critically ill patients. Most (91%) patients were nonresponders to the ACTH test. However, 35% of nonresponders had experienced a spontaneous increase in cortisol greater than 9 µg/dl during the previous 24 hours. Thus, treatment decisions based on a single random cortisol level or ACTHstimulated cortisol levels may be prone to error. These variations may, in part, explain previous reports of the poor reproducibility of the ACTH test in patients with sepsis. 13,33 Spontaneous changes in cortisol secretion make both basal cortisol and stimulated cortisol levels difficult to interpret, but there may be an even greater diagnostic dilemma. Regardless of the biochemical definition used to define a euadrenal response, the assumption is that the measurement of serum cortisol is reliable. Most hospitals use immunoassays to measure cortisol, and data suggest there is significant bias among different assays. 34 In addition to stimulating cortisol secretion, ACTH also causes the release of other steroids, such as 17 α-hydroxyprogesterone and dehydroepiandrosterone. 35 Immunoassays have crosssensitivities for these hormones that vary widely, and this may be a key source of the discrepancy. 36 Cohen et al. 35 examined the accuracy of immunoassays in sepsis. A 250 µg ACTH test was conducted in 9 patients with sepsis. Cortisol was measured using 3 different immunoassays and was also measured using high-pressure liquid chromatography (HPLC). HPLC was considered a reference standard due to its low cross-reactivity with steroids other than cortisol. The limits of agreement between the assays and HPLC were unacceptably high (range 62% to 770%). Agreement regarding responder/ nonresponder classification among the methods was less than 50% and discordance was noted among the different assays themselves. Similarly, harmonization data from the CORTICUS (Corticosteroid Therapy of Septic Shock) study also suggest that immunoassays demonstrate wide variation and are prone to misclassifying patients Another recently discovered limitation of immunoassays is the presence of heterophile antibodies in some patients, which can falsely lower serum cortisol measurements. 40,41 Do these results invalidate the previous work done in the area of adrenal insufficiency since immunoassays were used to measure cortisol in these studies? Data suggest otherwise. There seems to be good correlation between nonresponse to ACTH and clinical adrenal insufficiency. Nonresponse to ACTH in septic patients is associated with refractory hypotension, 42 higher levels of proinflammatory cytokines, 43 and increased mortality. 9,11,20,44-46 Treatment of Adrenal Insufficiency in Sepsis A meta-analysis indicated that low-dose corticotherapy reduces mortality in septic shock and concluded that steroids should be reserved primarily for patients who are nonresponders to the ACTH stimulation test (Table 2) This meta-analysis was heavily influenced by a large, randomized, multicenter, placebo-controlled French trial by Annane et al. 48 Recently, several limitations have been identified in this study and these limitations may have subsequently confounded the results of the meta-analysis. Enrollment in the French study required that patients be on mechanical ventilation and demonstrate vasopressordependent septic shock (systolic blood pressure <90 mm Hg for at least 1 h despite fluid resuscitation and vasopressor/inotrope use). 48 Within 8 hours of the onset of shock, patients underwent a 250 µg ACTH test and were randomized to receive placebo (n = 149) or hydrocortisone 50 mg intravenously every 6 hours with enteral fludrocortisone 50 µg daily (n = 150). The investigational protocol was amended approximately 2 years after study initiation to exclude patients who received etomidate, a drug known to inhibit steroidogenesis, 53 within 6 hours of randomization. Table 2. Effect of Corticotherapy on Mortality in Septic Shock 47 Treatment Control Weight Relative Reference Design Intervention (n) (n) (%) Risk (95% CI) Bollaert P, R, DB, PC, MC hydrocortisone 100 mg iv q8h, or placebo for 5 days 8/22 12/ (0.3 to 1.1) (1998) 49 Briegel P, R, DB, PC, SC hydrocortisone 100 mg iv for 1 dose, followed by 5/20 6/ (0.3 to 2.29) (1999) 50 hydrocortisone infusion of 0.18 mg/h for 6 days, followed by slow taper after resolution of infection, or placebo Chawla P, R, DB, PC, SC hydrocortisone 100 mg iv q8h for 72 h, then 6/23 10/ (0.24 to 1.25) (1999) 51 tapered over 4 days, or placebo Annane P, R, PC,DB, MC hydrocortisone 50 mg iv q6h and fludrocortisone 95/ / (0.77 to 1.07) (2002) µg tablet once daily, or placebo for 7 days Yildiz P, R, DB, PC, SC prednisolone 5 mg iv at 0600 h and 2.5 mg iv at 1800 h, 8/20 12/ (0.35 to 1.27) (2002) 52 or placebo for 10 days SUBTOTAL (0.71 to 0.97) DB = double-blind; MC = multicenter; P = prospective; PC = placebo-controlled; R = randomized; SC = single-center. The Annals of Pharmacotherapy 2007 September, Volume

5 Z Thomas and GL Fraser The majority (77%) of patients failed to increase their baseline cortisol by more than 9 µg/dl and were considered nonresponders. 48 Although corticotherapy failed to reduce mortality on an intent-to-treat basis, in the subgroup of nonresponders, corticotherapy reduced 28 day mortality from 63% to 53% (p = 0.04). This survival benefit is particularly impressive considering that these patients had a predicted mortality of 68% based on their Simplified Acute Physiology Score, a validated measure of disease severity in the ICU. 54,55 However, statistical significance was reached only after correcting for numerous baseline characteristics using a Cox regression model. This complex statistical analysis has led some to cast doubt on the conclusions of this study In the underpowered responder group, corticotherapy led to a nonsignificant increase in mortality from 53% to 61%. In addition, corticotherapy reduced the duration of vasopressor use in nonresponders (p = 0.001), but not in responders. A major issue surrounding the French trial 48 is that at least 72 patients received etomidate prior to the protocol amendment, and over 90% of these patients were found to be nonresponders. 59 It is now known that even a single dose of etomidate can blunt adrenal response to ACTH 45,60,61 and can result in lower serum cortisol concentrations compared with etomidate-naïve patients under similar stress. 62 It has also been demonstrated that the adrenal suppressive effects of etomidate in critically ill patients likely last longer than 6 hours. 45,61 Etomidate use may explain why the baseline cortisol concentrations in this study were unexpectedly low considering the severity of illness of this patient population (vasopressor-refractory shock). 63,64 Among nonresponders who received corticotherapy, the mean basal cortisol was 18 µg/dl (SD 12 µg/dl). This low baseline cortisol concentration coupled with the comparatively wide standard deviation has led some to suggest that these patients did not have relative adrenal insufficiency, but in fact had real adrenal failure. 16 This may explain the survival benefit of corticotherapy in this subgroup. A recent study demonstrated that the incidence of relative adrenal insufficiency in etomidate-naïve patients with vasopressor-refractory septic shock is significantly lower (34%) 65 than that reported in the French trial (77%). 48 These data raise questions about the generalizability of the French trial and the true incidence of adrenal insufficiency in septic shock. It is important to mention that when patients who received etomidate are excluded from the data analysis, corticotherapy was still found to be beneficial. 66 Although not prospectively confirmed, it appears that etomidate use in septic shock is commonly associated with adrenal insufficiency and thus should prompt the consideration of empiric corticotherapy. A second meta-analysis also found that low-dose corticotherapy is of benefit in septic shock. 15 However, contrary to the findings of the first meta-analysis, 47 a consistent benefit regardless of response to ACTH was noted. This conclusion was criticized in an accompanying editorial, since the French study, which accounted for the overwhelming majority of patients in this analysis as well, did not demonstrate a universal survival benefit. 67 The recently completed multinational Phase 3 CORTI- CUS study attempted to clarify the role of steroids in patients with less severe forms of sepsis than previously studied. Patients were enrolled if they met standard criteria for sepsis, 21 had sepsis-induced organ dysfunction, and demonstrated hemodynamic instability. Detailed inclusion/exclusion criteria can be found elsewhere. 68 Of note, unlike the French trial, patients were not required to have vasopressor-refractory shock. In CORTICUS, patients could be enrolled if they had a systolic blood pressure less than 90 mm Hg despite fluid replacement or if they needed vasopressors to maintain a systolic blood pressure greater than 90 mm Hg. After a standard ACTH test, patients (N = 499) were randomized to receive hydrocortisone 50 mg intravenously every 6 hours for 5 days followed by a 6 day tapering regimen or placebo. The primary endpoint was 28 day mortality in nonresponders. The full results of the CORTICUS trial are yet to be published, but preliminary data show that corticotherapy did not provide a mortality benefit on an intent-to-treat basis or when mortality was stratified by response to ACTH. 69 Contrary to the French study, corticotherapy reduced the duration of vasopressor usage in responders, but not in nonresponders. However, corticotherapy did not affect overall shock reversal in any subgroup. A trend toward increased superinfections was noted among patients who received hydrocortisone (OR = 1.27; 95% CI 0.96 to 1.68). Based upon inclusion criteria, CORTICUS targeted a less severely ill patient population. This is reflected by a lower overall mortality rate and a lower incidence of nonresponders compared with the French study: 32% versus 58% and 44% versus 77%, respectively. Although CORTICUS was stopped early due to poor enrollment (target enrollment = 800), it is unlikely that additional patients would have led to statistical significance due to the similar mortality rates among nonresponders who received steroids (37.6%) compared with those who received placebo (35.2%). There are several other differences between CORTI- CUS and the French trial that may explain these discordant results. In CORTICUS, hydrocortisone was given without fludrocortisone. This is significant in light of 2 studies suggesting that the combination of hydrocortisone/fludrocortisone has a greater benefit than hydrocortisone alone. 70,71 In addition, the dose of hydrocortisone differed from that used in the French study (hydrocortisone 50 mg intravenously every 6 hours for 7 days). The clinical significance of this change is unknown since the optimal dose and duration of corticotherapy in sepsis remain undefined. Another important difference between the 2 studies is the timing of enrollment The Annals of Pharmacotherapy 2007 September, Volume 41

6 Corticotherapy in Critical Illness Randomization into the French study had to be completed within 8 hours of the onset of shock. CORTICUS allowed inclusion up to 72 hours after shock onset. In light of the evidence of the need for the early recognition and treatment of sepsis, 72 this extended inclusion window may have attenuated any mortality benefit of corticotherapy. There are data, however, to support the use of corticotherapy for late septic shock, at least from a hemodynamic perspective. 49 Initial findings of CORTICUS are available, 73 but it seems that we will be left with more questions than answers. Corticotherapy appears to be of no benefit in fluid and vasopressor responsive shock, and thus should be reserved for those with refractory hypotension. Current surveys indicate that many clinicians are using steroids in patients without shock. 74 This practice should be abandoned since the efficacy of treatment is not universal in sepsis, and the adverse effect profile of corticotherapy is yet to be fully elucidated. Acute Respiratory Distress Syndrome ARDS is an inflammatory pulmonary process characterized by acute hypoxemic respiratory failure. 75 ARDS occurs in both medical and surgical patients and is associated with a mortality rate between 34% and 68%. 76 A comprehensive review of ARDS is beyond the scope of this paper and can be found elsewhere. 76,77 The recommendation to use steroids in ARDS is based primarily upon a small (N = 24) single-center trial that was terminated early. 78 As with many studies terminated early, several questions were unresolved The Acute Respiratory Distress Syndrome Network (ARDSNet) recently published the results of the first large confirmatory study of methylprednisolone in late ARDS. 75 The ARDSNet randomized 180 patients with late ARDS (7 28 days) into a placebo-controlled trial of methylprednisolone. 75 Methylprednisolone 2 mg/kg was administered on study day 1, followed by a scheduled treatment course of 21 days. At the end of 21 days, the drug was tapered over 2 4 days. Methylprednisolone did not reduce mortality compared with placebo (29% vs 29%; p = 1). In the subgroup of patients enrolled between days 14 and 28 after ARDS onset, there was increased mortality in the methylprednisolone group (35% vs 8%; p = 0.02). Compared with patients receiving placebo, patients receiving methylprednisolone had more ventilator-free days (p < 0.001) and were extubated sooner (p = 0.006). However, this benefit was mitigated because 20 patients in the methylprednisolone arm required reinstitution of mechanical ventilation compared with only 6 patients in the placebo arm (p = 0.008). Many of the patients who were reintubated demonstrated clinical signs of neuromyopathy and may have been at increased risk for respiratory failure. Another possible explanation for this high reintubation rate may have been an inadequate steroid tapering regimen. ARDS-related inflammation may persist for several weeks, even after patients have been extubated. 82,83 Patients in the Meduri et al. 78 trial were on mechanical ventilation for a median of 12 days, but these patients received a minimum duration of steroids of 18 days (range 18 32). In the ARDSNet study, patients were maintained on steroids for only 4 days after extubation. Discontinuation of steroids during periods of inflammation can result in increased inflammation and clinical deterioration. 84 This study illustrates the need for large-scale, randomized controlled trials in critically ill patients. Despite these disappointing results, there may still be a role for steroids in ARDS. Emerging data suggest that steroids may be beneficial in early ARDS, an area in which they have historically been found ineffective. 85 Post hoc subgroup analysis 43 of the previously described French trial 48 revealed that 177 (59%) patients had ARDS at enrollment. Of these patients, 92 received placebo and 85 received combination corticotherapy. The 2 groups were well matched for severity of illness and ventilator status. Once again it was found that corticotherapy benefited nonresponders to ACTH. Compared with placebo, corticotherapy reduced mortality in nonresponders from 75% to 53% (p = 0.021). Corticotherapy did not affect mortality among responders to ACTH. Oxygenation, as assessed by the PaO 2 /FiO 2, improved in both groups over time; however, the improvement was greater among nonresponders who received steroids (p 0.001). This corresponded to an increase in the number of days alive and off the ventilator in patients who received corticotherapy (mean ± SD, 5.7 ± 8.6 vs 2.6 ± 6.6 days; p = 0.006). Corticotherapy did not affect the duration of mechanical ventilation in responders (p = 0.22). The authors did not report whether any patients required reintubation. This would be of interest in light of the ARDSNet study, since an extended steroid taper was not used. The results of this study are impressive but should be interpreted cautiously since this study carries all of the welldocumented limitations of post hoc subgroup analysis These results are also supported by a recent study of methylprednisolone in early ARDS (within 72 h of diagnosis) by Meduri et al. 90 Patients were randomized to receive methylprednisolone (n = 63) 1 mg/kg followed by a tapering dose or placebo (n = 28). Treatment was continued up to 28 days. Treatment with methylprednisolone was associated with reduced need for mechanical ventilation (p = 0.01), improved Lung Injury Score (p = 0.002), and decreased ICU stay (p = 0.007). Treated patients also had decreased ICU mortality (p = 0.03), but there was no difference in hospital mortality. Interestingly, patients with adrenal insufficiency seemed to benefit less from steroid therapy, but this must be interpreted with caution since the trial was not powered to assess this. Although these results are promising, clinical equipoise must remain until a larger trial, powered for mortality, is able to reproduce these results. As demonstrated by the ARDSNet study, improvements in surrogate markers do not necessarily correlate with improved mortality The Annals of Pharmacotherapy 2007 September, Volume

7 Z Thomas and GL Fraser Severe Community-Acquired Pneumonia SCAP is associated with increased inflammation as evidenced by elevated levels of cytokines in bronchoalveolar lavage fluid. 91 Prior to the availability of effective antimicrobials, steroids were considered an integral component of the treatment paradigm for pulmonary infections. 2 However, controlled clinical trials of pneumonia have shown that steroids have little effect on clinical outcomes except in Pneumocystis jiroveci pneumonia Confalonieri et al. 95 have published the first randomized controlled trial of extended corticotherapy in SCAP. Patients with clinical and radiographic signs consistent with pneumonia were enrolled if they met 1 major criterion or 2 minor criteria for SCAP as proposed by Ewig et al. 96 Patients (N = 46) were randomized to receive hydrocortisone as a 200 mg intravenous bolus followed by a continuous infusion of 10 mg/h for 7 days or placebo. The primary endpoints were improvement in PaO 2 /FiO 2, change in multiple organ dysfunction score by study day 8, and development of delayed septic shock. HPA axis testing was not done. Despite randomization, there were differences in the groups at baseline. More patients in the control group required invasive mechanical ventilation (p = 0.03). This is an important finding because the need for invasive mechanical ventilation is a known negative prognostic indicator in SCAP. 96 It is therefore surprising that PaO 2 /FiO 2, chest radiograph scores, and C-reactive protein were worse in the hydrocortisone group at baseline. On study day 8, patients who received hydrocortisone were less likely to be on mechanical ventilation (26% vs 65%; p = 0.008). In addition, they had higher PaO 2 /FiO 2 (332 vs 237; p = ), less incidence of delayed septic shock (43% vs 0%; p = 0.001), and greater improvement in multiple organ dysfunction and chest radiograph scores. Hydrocortisone was also associated with improved ICU and hospital survival, but this is difficult to interpret because of the small number of patients in this study. HPA axis testing and serum cortisol concentrations were not assessed in this study. A recent publication demonstrated that serum cortisol is often low in a significant percentage of patients with SCAP. 97 Thus, steroid supplementation may have a physiologic rather than therapeutic mechanism. Despite its limitations, this study serves as proof- of-concept that steroids may have a role in SCAP. Only a large-scale, randomized study that includes assessment of HPA axis integrity can elucidate the role of steroids in SCAP. Summary Our understanding of corticotherapy in sepsis and other inflammatory conditions continues to evolve. It appears that adrenal status plays an important role in determining patient outcomes, but that our ability to assess the adequacy of this gland s function is much less secure. The ACTH test seems capable of diagnosing adrenal insufficiency in sepsis and can identify patients at increased risk of death. However, inaccuracies in the measurement of serum cortisol concentrations preclude its use as a sole determinant of the need for corticosteroid treatment. Rather, steroids should be considered in the subset of septic patients with vasopressor-refractory hypotension. The longstanding practice of steroid administration in late ARDS appears to be futile and, in some cases, potentially harmful. Although limited data suggest a role of steroids in early ARDS, this must be confirmed in an adequately powered prospective study before being adopted into practice. One trial has found steroids to be beneficial in SCAP. It is unknown whether this benefit is related to physiologic replacement of low cortisol concentrations or some other mechanism. Although the reported benefits are impressive, it is clear that additional trials are needed to define the role of steroids in SCAP. Finally, it should be noted that adrenal insufficiency can occur in sepsis, ARDS, and SCAP and that these conditions can occur simultaneously in the same patient. Future studies should try to elucidate whether the effects of corticotherapy are related to the treatment of adrenal insufficiency or the treatment of the underlying disease per se. Zachariah Thomas PharmD, Clinical Assistant Professor of Pharmacy, Ernest Mario School of Pharmacy Rutgers, State University of New Jersey, Piscataway, NJ; Clinical Pharmacist, Hackensack University Medical Center, Hackensack, NJ Gilles L Fraser PharmD FCCM, Professor of Medicine, College of Medicine, University of Vermont, Burlington, VT; Clinical Pharmacist in Critical Care, Maine Medical Center, Portland, ME Reprints: Dr. Thomas, Hackensack University Medical Center, 30 Prospect Ave., Hackensack, NJ 07601, fax 201/ , zachariah. thomas@gmail.com We thank Keri Bicking PharmD and Farooq Bandali PharmD for their thoughtful review of this article and the SCCM/CPP Mentor Mentee Program for support of this work. References 1. Prigent H, Maxime V, Annane D. Clinical review: corticotherapy in sepsis. Crit Care 2004;8: Perla D, Marmorston J. Suprarenal cortical hormone and salt in the treatment of pneumonia and other severe infections. Endocrinology 1940;27: Glyn J. The discovery and early use of cortisone. J R Soc Med 1998; 91: Prigent H, Maxime V, Annane D. Science review: mechanisms of impaired adrenal function in sepsis and molecular actions of glucocorticoids. Crit Care 2004;8: Kollef MH, Schuster DP. The acute respiratory distress syndrome. N Engl J Med 1995;332: Monton C, Torres A, El-Ebiary M, Filella X, Xaubet A, de la Bellacasa JP. Cytokine expression in severe pneumonia: a bronchoalveolar lavage study. Crit Care Med 1999;27: Burry LD, Wax RS. Role of corticosteroids in septic shock. Ann Pharmacother 2004;38: Epub 23 Jan DOI /aph.1C Melby JC, Spink WW. Comparative studies on adrenal cortical function and cortisol metabolism in healthy adults and in patients with shock due to infection. J Clin Invest 1958;37: Rothwell PM, Udwadia ZF, Lawler PG. Cortisol response to corticotropin and survival in septic shock. Lancet 1991;337: The Annals of Pharmacotherapy 2007 September, Volume 41

8 Corticotherapy in Critical Illness 10. Oelkers W. Adrenal insufficiency. N Engl J Med 1996;335: Annane D, Sebille V, Troche G, Raphael JC, Gajdos P, Bellissant E. A 3- level prognostic classification in septic shock based on cortisol levels and cortisol response to corticotropin. JAMA 2000;283: Marik PE, Zaloga GP. Adrenal insufficiency in the critically ill: a new look at an old problem. Chest 2002;122: Loisa P, Uusaro A, Ruokonen E. Single adrenocorticotropic hormone stimulation test does not reveal adrenal insufficiency in septic shock. Anesth Analg 2005;101: Morel J, Venet C, Donati Y, et al. Adrenal axis function does not appear to be associated with hemodynamic improvement in septic shock patients systematically receiving glucocorticoid therapy. Intensive Care Med 2006;32: Minneci PC, Deans KJ, Banks SM, Eichacker PQ, Natanson C. Metaanalysis: the effect of steroids on survival and shock during sepsis depends on the dose. Ann Intern Med 2004;141: Dickstein G. On the term relative adrenal insufficiency or what do we really measure with adrenal stimulation tests? J Clin Endocrinol Metab 2005;90: Arlt W, Allolio B. Adrenal insufficiency. Lancet 2003;361: Schimmer BP, Parker KL. Adrenocorticotropic hormone; adrenocortical steroids and their synthetic analogs; inhibitors of the synthesis and actions of adrenocortical hormones. In: Hardman JG, Limbird LE, eds. Goodman & Gilman s the pharmacological basis of therapeutics. New York: McGraw-Hill, 2001: Berneis K, Staub JJ, Gessler A, Meier C, Girard J, Müller B. Combined stimulation of adrenocorticotropin and compound-s by single dose metyrapone test as an outpatient procedure to assess hypothalamic pituitary adrenal function. J Clin Endocrinol Metab 2002;87: Annane D, Maxime V, Ibrahim F, Alvarez JC, Abe E, Boudou P. Diagnosis of adrenal insufficiency in severe sepsis and septic shock. Am J Respir Crit Care Med 2006;174: Bone RC, Balk RA, Cerra FB, et al. Definitions for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis. The ACCP/SCCM Consensus Conference Committee. American College of Chest Physicians /Society of Critical Care Medicine. Chest 1992;101: Hamrahian AH, Oseni TS, Arafah BM. Measurements of serum free cortisol in critically ill patients. N Engl J Med 2004;350: Marik PE, Zaloga GP. Adrenal insufficiency during septic shock. Crit Care Med 2003;31: Annane D, Fan E, Herridge MS. Pro-con debate: steroid use in ACTH non-responsive septic shock patients with high baseline cortisol levels. Crit Care 2006;10:210. DOI /cc Guyatt GH, Drummond R, eds. Users guides to the medical literature: a manual for evidence-based clinical practice. Chicago: American Medical Association Press; (accessed 2006 Dec 15). 26. Mayenknecht J, Diederich S, Bähr V, Plöckinger U, Oelkers W. Comparison of low and high dose corticotropin stimulation tests in patients with pituitary disease. J Clin Endocrinol Metab 1998;83: Siraux V, De Backer D, Yalavatti G, et al. Relative adrenal insufficiency in patients with septic shock: comparison of low-dose and conventional corticotropin tests. Crit Care Med 2005;33: Dorin RI, Qualls CR, Crapo LM. Diagnosis of adrenal insufficiency. Ann Intern Med 2003;139: Widmer IE, Puder JJ, Konig C, et al. Cortisol response in relation to the severity of stress and illness. J Clin Endocrinol Metab 2005;90: Annane D. Low-dose adrenocorticotropic hormone test is not ready for routine adrenal function testing in the intensive care unit. Crit Care Med 2005;33: Kozyra EF, Wax RS, Burry LD. Can 1 µg of cosyntropin be used to evaluate adrenal insufficiency in critically ill patients? Ann Pharmacother 2005;39: Epub 1 Mar DOI /aph.1E Venkatesh B, Mortimer RH, Couchman B, Hall J. Evaluation of random plasma cortisol and the low dose corticotropin test as indicators of adrenal secretory capacity in critically ill patients: a prospective study. Anaesth Intensive Care 2005;33: Bouachour G, Roy PM, Guiraud MP. The repetitive short corticotropin stimulation test in patients with septic shock. Ann Intern Med 1995; 123: Clark PM, Neylon I, Raggatt PR, Sheppard MC, Stewart PM. Defining the normal cortisol response to the short Synacthen test: implications for the investigation of hypothalamic pituitary disorders. Clin Endocrinol (Oxf) 1998;49: Cohen J, Ward G, Prins J, Jones M, Venkatesh B. Variability of cortisol assays can confound the diagnosis of adrenal insufficiency in the critically ill population. Intensive Care Med 2006;32: Roberts RF, Roberts WL. Performance characteristics of five automated serum cortisol immunoassays. Clin Biochem 2004;37: Briegel J, Vogeser M, Annane D, et al. Correct identification of relative adrenal insufficiency in septic shock (abstract). Intensive Care Med 2005;31:A Vogeser M, Briegel J, Singer M, et al. Measurement of cortisol in septic shock: interlaboratory harmonization (abstract). Intensive Care Med 2005; 31:A Keh D. Corticosteroid therapy in sepsis: where are we? Adv Sepsis 2006;5: Arafah BM. Hypothalamic pituitary adrenal function during critical illness: limitations of current assessment methods. J Clin Endocrinol Metab 2006;91: Bolland MJ, Chiu WW, Davidson JS, Croxson MS. Heterophile antibodies may cause falsely lowered serum cortisol values. J Endocrinol Invest 2005;28: Annane D, Bellissant E, Sebille V, et al. Impaired pressor sensitivity to noradrenaline in septic shock patients with and without impaired adrenal function reserve. Br J Clin Pharmacol 1998;46: Annane D, Sebille V, Bellissant E. Effect of low doses of corticosteroids in septic shock patients with or without early acute respiratory distress syndrome. Crit Care Med 2006;34: Bollaert PE, Fieux F, Charpentier C, Lévy B. Baseline cortisol levels, cortisol response to corticotropin, and prognosis in late septic shock. Shock 2003;19: Malerba G, Romano-Girard F, Cravoisy A, et al. Risk factors of relative adrenocortical deficiency in intensive care patients needing mechanical ventilation. Intensive Care Med 2005;31: Lipiner-Friedman D, Sprung CL, Laterre PF, et al. Adrenal function in sepsis: the retrospective Corticus cohort study. Crit Care Med 2007;35: Annane D, Bellissant E, Bollaert PE, Briegel J, Keh D, Kupfer Y. Corticosteroids for severe sepsis and septic shock: a systematic review and meta-analysis. BMJ 2004;329: Annane D, Sebille V, Charpentier C, et al. Effect of treatment with low doses of hydrocortisone and fludrocortisone on mortality in patients with septic shock. JAMA 2002;288: Bollaert PE, Charpentier C, Levy B, Debouverie M, Audibert G, Larcan A. Reversal of late septic shock with supraphysiologic doses of hydrocortisone. Crit Care Med 1998;26: Briegel J, Forst H, Haller M, et al. Stress doses of hydrocortisone reverse hyperdynamic septic shock: a prospective, randomized, double-blind, single-center study. Crit Care Med 1999;27: Chawla K, Kupfer Y, Tessler S. Hydrocortisone reverses refractory septic shock (abstract). Crit Care Med 1999;27:A Yildiz O, Doganay M, Aygen B, Guven M, Kelestimur F, Tutuu A. Physiological-dose steroid therapy in sepsis [ISRCTN ]. Crit Care 2002;6: Wagner RL, White PF, Kan PB, Rosenthal MH, Feldman D. Inhibition of adrenal steroidogenesis by the anesthetic etomidate. N Engl J Med 1984;310: Le Gall JR, Lemeshow S, Saulnier F. A new Simplified Acute Physiology Score (SAPS II) based on a European/North American multicenter study. JAMA 1993; 270: Erratum in: JAMA 1994;271: Société Française d Anesthésie et de Réanimation. Paris. scores2/saps2.html (accessed 2006 Dec 16). 56. Millo J. Corticosteroids for patients with septic shock (letter). JAMA 2003;289:41. The Annals of Pharmacotherapy 2007 September, Volume

9 Z Thomas and GL Fraser 57. Vincent JL. Is the current management of severe sepsis and septic shock really evidence based? PLoS Med 2006;3:e Rubenfeld GD. When survival is not the same as mortality. Critical Care Alert 2003;10: Annane D, Sebille V. Corticosteroids for patients with septic shock (letter). JAMA 2003;289: Mohammad Z, Afessa B, Finkielman JD. The incidence of relative adrenal insufficiency in patients with septic shock after the administration of etomidate. Crit Care 2006;10:R105. DOI /cc Absalom A, Pledger D, Kong A. Adrenocortical function in critically ill patients 24 h after a single dose of etomidate. Anaesthesia 1999;54: Allolio B, Stuttmann R, Leonhard U, Fischer H, Winkelmann W. Adrenocortical suppression by a single induction dose of etomidate. Klin Wochenschr 1984;62: Jurney TH, Cockrell JL, Lindberg JS, Lamiell JM, Wade CE. Spectrum of serum cortisol response to ACTH in ICU patients. Correlation with degree of illness and mortality. Chest 1987;92: Wade CE, Lindberg JS, Cockrell JL, et al. Upon-admission adrenal steroidogenesis is adapted to the degree of illness in intensive care unit patients. J Clin Endocrinol Metab 1988;67: Jones D, Hayes M, Webb S, French C, Bellomo R. Relative adrenal insufficiency in etomidate-naive patients with septic shock. Anaesth Intensive Care 2006;34: Annane D, Sebille V, Bellissant E. Exploring the role of etomidate in septic shock and acute respiratory distress syndrome (letter). Crit Care Med 2006;34: Luce JM. Physicians should administer low-dose corticosteroids selectively to septic patients until an ongoing trial is completed. Ann Intern Med 2004;141: Hadassah Medical Organization; European Society of Intensive Care Medicine; International Sepsis Forum; The Gorham Foundation. Corticosteroid Therapy of Septic Shock CORTICUS. ct/show/nct (accessed 2007 Apr 20). 69. Sprung CL. CORTICUS Trial. Presented at: Society of Critical Care Medicine 36th Annual Congress, Orlando, FL, February 19, Drawdy S, Allen M, Winterstein AG, Allen M, Jantz MA. Effect of hydrocortisone vs hydrocortisone plus fludrocortisone on mortality in critically ill (abstract). Crit Care Med 2005;33:A Gonzalez H, Nardi O, Annane D. Relative adrenal failure in the ICU: an identifiable problem requiring treatment. Crit Care Clin 2006;22: vii. 72. Rivers E, Nguyen B, Havstad S, et al. Early goal-directed therapy in the treatment of severe sepsis and septic shock. N Engl J Med 2001;345: Corticosteroid therapy of septic shock (CORTICUS). Presented at: 2007 American Thoracic Society International Conference, San Francisco, May 18 23, (accessed 2007 Aug 2). 74. Annane D, Beale R, Reinhart K, Janes J, Nelson DR, Williams MD. Global utilization of low dose corticosteroids in severe sepsis: a report from the PROGRESS Registry (abstract). Crit Care Med 2006;34:A Steinberg KP, Hudson LD, Goodman RB, et al. Efficacy and safety of corticosteroids for persistent acute respiratory distress syndrome. N Engl J Med 2006;354: Ware LB, Matthay MA. The acute respiratory distress syndrome. N Engl J Med 2000;342: Thomas Z, Kimborowicz K. Corticosteroids in ARDS: the never-ending story. Hosp Pharm 2007;42: Meduri GU, Headley AS, Golden E, et al. Effect of prolonged methylprednisolone therapy in unresolving acute respiratory distress syndrome: a randomized controlled trial. JAMA 1998;280: Montori VM, Devereaux PJ, Adhikari NK, et al. Randomized trials stopped early for benefit: a systematic review. JAMA 2005;294: Brun-Buisson C, Brochard L. Corticosteroid therapy in acute respiratory distress syndrome: better late than never? JAMA 1998;280: Wheeler A, Bernard GR, Schoenfeld D, Steinberg K. Methylprednisolone for unresolving ARDS (letter). JAMA 1998;280: Meduri GU, Tolley EA, Chrousos GP, Stentz F. Prolonged methylprednisolone treatment suppresses systemic inflammation in patients with unresolving acute respiratory distress syndrome. Evidence for inadequate endogenous glucocorticoid secretion and inflammation-induced immune cell resistance to glucocorticoids. Am J Respir Crit Care Med 2002;165: Meduri GU, Chrousos GP. Effectiveness of prolonged glucocorticoid treatment in acute respiratory distress syndrome: the right drug, the right way? Crit Care Med 2006;34: Keh D, Boehnke T, Weber-Cartens S, et al. Immunologic and hemodynamic effects of low-dose hydrocortisone in septic shock: a doubleblind, randomized, placebo-controlled, crossover study. Am J Respir Crit Care Med 2003;167: Bernard GR, Luce JM, Sprung CL, et al. High-dose corticosteroids in patients with the adult respiratory distress syndrome. N Engl J Med 1987;317: Dellinger RP. Post hoc analyses in sepsis trials: a formula for disappointment? Crit Care Med 1996;24: Lagakos SW. The challenge of subgroup analyses reporting without distorting. N Engl J Med 2006;354: Erratum in: N Engl J Med 2006;355: Assmann SF, Pocock SJ, Enos LE, et al. Subgroup analysis and other (mis)uses of baseline data in clinical trials. Lancet 2000;355: Smith R, Roberts I. Patient safety requires a new way to publish clinical trials. PLoS Clin Trials 2006;1:e6. DOI /journal.pctr Meduri GU, Golden E, Freire AX, et al. Methylprednisolone infusion in early severe ARDS: results of a randomized controlled trial. Chest 2007; 131: Schutte H, Lohmeyer J, Rosseau S, et al. Bronchoalveolar and systemic cytokine profiles in patients with ARDS, severe pneumonia and cardiogenic pulmonary oedema. Eur Respir J 1996;9: Bozzette SA, Sattler FR, Chiu J, et al. A controlled trial of early adjunctive treatment with corticosteroids for Pneumocystis carinii pneumonia in the acquired immunodeficiency syndrome. California Collaborative Treatment Group. N Engl J Med 1990;323: Marik P, Kraus P, Sribante J, Havlik I, Lipman J, Johnson DW. Hydrocortisone and tumor necrosis factor in severe community-acquired pneumonia. A randomized controlled study. Chest 1993;104: Monton C, Ewig S, Torres A, et al. Role of glucocorticoids on inflammatory response in nonimmunosuppressed patients with pneumonia: a pilot study. Eur Respir J 1999;14: Confalonieri M, Urbino R, Potena A, et al. Hydrocortisone infusion for severe community-acquired pneumonia: a preliminary randomized study. Am J Respir Crit Care Med 2005;171: Ewig S, Ruiz M, Mensa J, et al. Severe community-acquired pneumonia. Assessment of severity criteria. Am J Respir Crit Care Med 1998;158: Salluh JI, Verdeal JC, Mello GW, et al. Cortisol levels in patients with severe community-acquired pneumonia. Intensive Care Med 2006;32: Una Actualización sobre el Diagnóstico de Insuficiencia Adrenal y el Uso de la Terapia Basada en la Utilización de Corticoides en Enfermedades Graves Z Thomas y GL Fraser Ann Pharmacother 2007;41: EXTRACTO OBJETIVO: Examinar la literatura reciente con relación a la terapia en base al uso de corticoides en pacientes gravemente enfermos sufriendo de sepsis, del Síndrome de Dificultad Respiratoria Aguda (SDRA), y Neumonía Severa Adquirida en la Comunidad (SCAP). FUENTES DE INFORMACIÓN: La literatura fue identificada a través de MEDLINE (1966 abril 2007) usando combinaciones de las palabras claves en inglés: hydrocortisone, adrenal insufficiency, acute respiratory distress syndrome, pneumonia, sepsis, y cortisol. Las bibliografías de artículos relevantes fueron revisadas con el propósito de obtener citas 1464 The Annals of Pharmacotherapy 2007 September, Volume 41