By Robert J Gianotti, MD
Faculty Peer Reviewed
Spring has arrived. Here in the Northeast we have been showered with continual sunshine and our flowering trees are approaching full bloom. Whether enjoying matzah or marshmallow peeps we were all blessed with an action packed weekend on the sports front. Opening weekend in baseball came and went with some great wins, some impressive long balls, and one record breaking 16 inning game. The spring tradition continued with the first major tournament in professional golf, The Masters. Tiger Woods continued to struggle and a golfer who has never taken a lesson in his life hooked his way to victory in Augusta. In between innings, tee times, and delicious lamb sausage I was still able to work my way through the week in medical literature to bring you my favorites.
Early in the week, the New York Times reported on the Choosing Wisely campaign from the American Board of Internal Medicine Foundation[1,2]. The campaign is the result of a consensus among nine medical subspecialty boards (including ABIM and the American College of Cardiology) in conjunction with Consumer Reports. Motivated by skyrocketing healthcare costs, the panel has made recommendations to promote the conservative ordering of diagnostic tests and implementation of unnecessary therapeutic measures. Examples include limiting the routine use of Head CT in the evaluation of syncope without neurological signs and the common practice of prescribing antibiotics as first line treatment for run of the mill sinusitis. How these recommendations will affect community practice is unknown, but it serves as a foundation for building a conversation between patients and providers in making rational decisions on screening and treatments that are common, but may confer little benefit with substantial risk. With an eye toward rational testing and prescribing, we can now turn our attention to the medical literature and see how it fits into this framework.
Recently, we have seen an impressive surge in the number of non-inferiority studies comparing novel pharmaceuticals to long held standards. Leading the pack are studies comparing the effectiveness of the new generation oral anticoagulants to the old standby’s heparin and coumadin in atrial fibrillation and venous thrombosis. This week in the New England Journal of Medicine we find the results of the EINSTEIN-PE study[3], an open label non inferiority trial comparing oral rivaroxaban (an oral factor Xa inhibitor) to standard therapy with enoxaparin bridge to coumadin in treating acute symptomatic pulmonary embolism. The study randomized 4,833 patients to one of the two treatment arms. The primary outcome was not mortality, but instead was powered to detect symptomatic recurrence of DVT as a continuous variable over one year. Mortality, driven by PE in less than 20% of cases, was low in both study groups with death rates occurring at 2.4% and 2.1% respectively (p=0.53). The primary outcome of recurrent DVT, with a non-inferiority margin set at 2.0, was found to be equivalent in both groups at 2.1% with rivaroxaban and 1.8% with standard treatment (P=0.003 for non-inferiority).Major bleeding events occurred at a significantly lower rate in the rivaroxaban group (1.1% vs. 2.2%, p=0.003). Interestingly, over 90% of patients in the rivaroxaban group received low molecular weight heparin for at least one day prior to confirmation of diagnosis when they were subsequently randomized to one of the two groupings. The authors argue that this brief exposure was unavoidable and did not change the natural history or outcome. I am skeptical of this statement, as we are not certain of the comparative effectiveness of each of these agents very early in the course of thrombosis. It should also be noted that with an open label design the study was subject to inherent bias and the authors acknowledge a slightly higher suspected recurrence rate in the rivaroxaban group (10.2% vs. 9.7%) that was not borne out on repeat diagnostic testing. This study confirms that oral factor Xa inhibitors are likely to be an important and safe tool in the treatment of acute PE in the correct population with the right insurance coverage, but mortality benefit remains to be proven.
The true pinnacle of medical innovation would be to describe a treatment that is safe and effective, yet inexpensive. With this triad in mind, we can now turn our attention to the IMMEDIATE trial published this week in JAMA[4]. This trial is a natural follow up to the CREATE-ECLA[5] study that looked at mortality in ST-elevation myocardial infarction following the administration of glucose-insulin-potassium, a.k.a. GIK. The predecessor to the current trial failed to show a mortality benefit from GIK, yet came under fire for the trial design in which GIK was administered following reperfusion in 68% of patients at a median of 6 hours after symptom onset. In the current trial, the authors proposed to expand the window for GIK to provide metabolic support close to the time of the initial event. To do so, GIK was administered in a randomized, double blind fashion to an ironical 911 patients by trained EMS workers that made the diagnosis of acute coronary syndrome in the field. The primary endpoint was progression to myocardial infarction within 24 hours by both biomarker and ECG criteria. The pre-specified secondary endpoints were 30 day survival and a composite of cardiac arrest or in-house mortality. GIK failed to slow progression to MI after 24 hours with 48.7% of patients progressing to MI compared to 52.6% of those that received placebo (OR=0.88, 95% CI 0.66-1.33, p=0.28). 30 day mortality was also found to be similar between groups (4.4% vs. 6.1%, p=0.27). GIK was beneficial when it came to the composite of cardiac arrest or in-house mortality with 6.1% of the GIK patients meeting this endpoint vs. 14.4% of the placebo group (p=0.01). The authors postulate that this is in line with the pathophysiological model of GIK reducing levels of circulating free-fatty acids that have arrhythmogenic potential. A small subset of patients that had post-event imaging demonstrated smaller infarct size with GIK as a percentage of left ventricular mass (median 2% vs. 10%, p=0.02) and this finding may guide future studies of GIK powered for this endpoint. Further research will hopefully find that early GIK is a cost effective measure that will save lives in the long term by reducing infarct size and delaying progression to heart failure.
Cost-effectiveness is the driver behind an interesting analysis published this month in the journal Hepatology[6]. As internists, we have grown adept at treating overt hepatic encephalopathy in the hospitalized cirrhotic patient, but we don’t often have the tools to recognize the subtle, yet significant changes that define minimal hepatic encephalopathy (MHE). Importantly, MHE has been shown to be a risk factor for motor vehicle accidents (MVAs) due to sub-clinical visual-spatial and cognitive abnormalities in cirrhotic patients. Bajaj et al. performed a cost effectiveness analysis that shows both empirical treatment for MHE with lactulose or rifaxamin and treatment based on several neuropsychological tests used to define MHE would significantly reduce societal and healthcare costs associated with MVAs. The model predicts that diagnosis of MHE with a simple computerized attention test called the inhibitory control test (ICT) and subsequent treatment with lactulose for five years would cost $24,454 per accident prevented and reduce the cirrhotic accident rate from 0.18 per year to 0.039 per year. Net savings are estimated to fall between $1.7 million and $3.6 million over 5 treatment years. The authors wisely argue that rifaxamin, although less cost effective than lactulose, may be a better long term treatment choice as it has a much better side effect profile. Finally, in the same issue of Hepatology[7] we find a meta-analysis using pooled randomized data that supports the use of albumin infusion following large volume paracentesis. The primary pooled outcomes included prevention of post-paracentesis circulatory dysfunction (defined as >50% increase in plasma renin activity; OR =0.39, 0.27-0.55), reduced episodes of hyponatremia (OR =0.58, 0.39-0.87) and mortality (OR=0.64, 0.41-0.98). This data supports the use of albumin as an effective strategy to prevent complications following large volume paracentesis in decompensated cirrhosis. This comes in combination with a growing body of evidence that albumin may have pleiotropic effects in cirrhotics; promoting endogenous hormone and drug delivery, and modulating micro-circulation and endothelial function[8].
That’s game over for this week’s Primecuts. We’ll see you next week for another round of insightful analysis of the medical literature. Until then, back to fantasy baseball, where complicated statistics rival even the most hardened meta-analysis.
Dr. Robert J Gianotti, NYU Chief Medical Resident, NYU Langone Medical Center
Peer reviewed by Neil Shapiro, MD, Editor-In-Chief, Clinical Correlations
Image courtesy of Wikimedia Commons
References
1. Doctor Panels Recommend Fewer Tests for Patients. New York Times. April 4th, 2012.
3. Buller et al. Oral rivaroxaban for the treatment of symptomatic pulmonary embolism. NEJM 2012;366: 1287-1297. http://www.nejm.org/doi/full/10.1056/NEJMoa1113572
4. Selker et al. Out-of-hospital administration of intravenous glucose-insulin-potassium in patients with suspected acute coronary syndromes. JAMA 2012;307:E7-E9. http://sbhci.org.br/wp-content/uploads/2012/03/IMMEDIATE.pdf
5. Mehta SR et al. Effect of glucose-insulin-potassium infusion on mortality in patients with acute ST-segment elevation myocardial infarction: the CREATE-ECLA randomized controlled trial. JAMA 2005;293:437-446.
6. Bajaj et al. Diagnosis and treatment of minimal hepatic encephalopathy to prevent motor vehicle accidents: a cost effectiveness analysis. Hepatology 2012;55: 1164-1171.
7. Bernardi et al. Albumin infusion in patients undergoing large volume paracentesis: a meta-analysis of randomized trials. Hepatology 2012;55:1172-1181.
8. Arroyo V. Human serum albumin: not just a plasma volume expander. Hepatology 2009;50:355-357.