In Greek mythology Peitho was the Goddess of persuasion and seduction. Please feel free to find irony where you see fit…
I am a sucker for catchy trial names. I swoon in anticipation when reading of the recent publication of a large multi-centered randomized control trial on a topic that is relevant to emergency medicine. I am further enthused when the authors have come up with a creative acronym by which to identify their trial. INTERACT2, IST3, the ARISE trial. How does one not feel better about oneself when reading a trial with such illustrious a title as the ARISE trial. So it is not surprising that I have been waiting with bated breath for the publication of the PEITHO trial since its’ results were announced earlier this year at the American College of Cardiology 2013 Scientific Sessions. As of yet the complete trial data has not been published, but recently the primary, secondary and safety outcomes were all released online in slide form.
Other than an inspirational name the PEITHO trial is the largest trial to date examining whether thrombolytics are beneficial in “submassive” pulmonary embolism. The authors claim their trial a success, as their primary endpoint, all-cause mortality or hemodynamic collapse within 7 days, was statistically significant. There was in fact a 3% absolute difference between the tenecteplase and the placebo group. Unfortunately none of this difference resulted from all-cause mortality, which was statistically equivalent. The entire difference was found in what the authors referred to as, “hemodynamic collapse”. This was defined as need for CPR, a systolic blood pressure less than 90mm Hg or a drop of greater than 40mm Hg for greater than 15 minutes with signs of end organ hypoperfusion, or the need for pressors. All these variables are markers for poor outcomes, especially the need for CPR, but they are all surrogates for what we really care about, death. If the overall mortality rate was the same then it is hard to get worked up about 15 minutes of hypotension or the subjective decision to infuse a little dopamine (though the CPR is still unsettling). In fact, the major difference in groups was primarily in need for pressors and greater than 15 minutes of hypotension. Obviously only patients with prolonged hypotension are started on a pressor, and so these two endpoints are essentially measuring the same thing. They are clearly related variables and thus statistically equivalent. Even more statistically bothersome is the authors state that hemodynamic collapse was responsible for 5 deaths in the control group. If these patients were counted as dead then obviously the worst thing has happened to them and they should not be counted again as having “hemodynamic collapse”(an obvious occurrence in most forms of death). It is already debatable whether 15 minutes of hypotension is a clinically relevant endpoint, but it is appears entirely inconsequential when the endpoint of death has already been accounted for in that particular patient. Treating each of these events as separate independent outcomes will statistically overestimate the significance of the results.
It is hard to imagine what the authors were thinking when they wrote their protocol and it will be interesting to see them justify their power calculation when the trial is finally published. Given the recent registry data published on submassive pulmonary embolism (3) the trial was destined for failure from the beginning. Overall mortality rates in the EMPEROR registry was found to be around 3% and in the PEITHO trial the mortality rate in the placebo group was 1.8% (1,3). Even under the best of circumstances with an incredibly efficacious intervention, it would be very difficult to significantly improve such a low mortality rate. If you actual look at the causes of death in the two groups you find that tenecteplase did in fact prevent deaths from pulmonary embolism and hemodynamic collapse. It just happened to cause as many deaths from hemorrhage, intracranial or otherwise. Six people (1.2%) in the control group died from hemodynamic collapse or recurrent pulmonary embolism at 30 days. In contrast seven people (1.4%) in the tenecteplase group died from stroke or bleeding. The overall mortality rate was equivalent, but the causes of death were entirely different.
One of the most important questions the PEITHO trial resolved was the actual risk of intracranial hemorrhage (ICH), when patients with pulmonary embolisms are treated with thrombolytics. Among the many small studies published to date on this topic. the rate of ICH between thrombolytics and heparin was virtually identical (7). So much so that many experts hypothesized that because of the inherent differences in the patients with pulmonary embolisms compared to those with myocardial infarcts or ischemic strokes, the risk of ICH is negligible. Thanks to PEITHO it is apparent that this hypothesis is untrue and thrombolytics present a clear risk in patients with pulmonary embolism. The number-needed-to-harm is around 50 patients to cause one ICH. Given these risks it and the lack of a mortality benefit, it is hard to justify the utility of such an intervention.
With the PEITHO trial debunking the theory that thrombolytics save lives in patients with submassive pulmonary embolisms, and revealing there is a real and measurable risk for harm, we will now address the long-term benefits of the administration of thrombolytics in these patients. This is what I call the NINDSitization of pulmonary embolism. What the 90 day modified Rankin Scale did for thrombolytics in stroke, the 2-year pulmonary artery systolic pressure will attempt to do in pulmonary embolism. Many experts have hypothesized that giving thrombolytics to patients with submassive pulmonary embolisms will result in a decrease in the rate of right ventricular dysfunction these patients will experience in the future (5). Thrombolytics will prevent them from becoming “respiratory cripples”.
This is exactly the hypothesis that the authors of the MOPPET study sought to prove when they published their trial in January 2013 (3). This trial examined whether patients with submassive pulmonary embolisms who were treated with t-PA had a lower rate of secondary pulmonary hypertension than those treated with heparin alone. In this trial they randomized patients with submassive pulmonary embolisms to heparin and half dose t-PA (50 mg), or heparin and placebo. The primary endpoints of this trial were the development of pulmonary hypertension as assessed by echocardiography and the composite endpoint of pulmonary hypertension and recurrent PE. This trial, so underpowered to show anything but extreme differences in the two groups, was shockingly successful. They found a 41% absolute difference between the t-PA and placebo group in frequency of pulmonary hypertension at 28 month follow up. This is seemingly a slam-dunk. Though these results are almost too good to be true. In fact before this study was published rates of chronic thromboembolic pulmonary hypertension after pulmonary embolism were between 4-10% depending upon the study (4,5,6). Why did the MOPPET trial find 56% of their patients to have pulmonary hypertension at 2-years? There are two possible reasons for this discrepancy. The first is that the patients in the MOPPET trial were far sicker than those in the previous trials. In the Pengo trial published in the NEJM in 2004 and in the Becattini trial in CHEST in 2006 the overall in-hospital mortality rates were 8.1% (4,5). This was significantly higher than the 3.3% overall mortality in the MOPPET trial. It is hard to imagine that the patients in the MOPPET trial had more severe pulmonary embolisms and yet died less frequently. A much more believable reason for the ten times higher rate of pulmonary hypertension was a definitional one. By changing the criteria for the diagnosis of chronic thromboembolic pulmonary hypertension to a disease oriented one, the MOPPET authors found that patients after pulmonary embolisms treated with heparin alone have far higher pulmonary systolic pressures than those treated with t-PA and heparin. It is uncertain what clinical relevance this holds.
In both the Pengo and Becattini studies the authors prospectively gathered patients with pulmonary embolisms and followed their cohort for over 3 years (up to ten years in the Pango trial). Patients were only screened for pulmonary hypertension if the presented with symptoms of dyspnea unexplained by other reasons. This sub-group then had an echo to evaluate their right ventricular function. Those that showed right ventricular stress were further evaluated with ventilation/perfusion scans and had their pulmonary pressures measured directly via catheterization. Pulmonary hypertension was considered to be present if the systolic and mean pulmonary artery pressures exceeded 40 and 25 mm Hg, respectively as measured directly via catheterization. Surely these are far more stringent criteria than what was required in the MOPPET trial, and are the primary reasons for the discrepancy in rates of chronic thromboembolic pulmonary hypertension. The most important being that the patients had to present with symptoms and have functional deficits. An elevated pulmonary pressure is clinically meaningless if it has no effect on a patient’s function.
The MOPPET trial on the other hand screened all their patients at 48 hours, 6 months, and 28 months irrespective of whether or not they presented with appropriate symptoms. The definition of pulmonary hypertension was a dichotomous cutoff of a systolic pulmonary blood pressure greater than 40mm Hg as estimated on cardiac echo. This is a perfect example of a disease-oriented outcome and unfortunately since the authors do not tell us if any of the patients were symptomatic, we are unable to determine if it is a clinically relevant outcome. If you examine the 18 individuals diagnosed with chronic thromboembolic pulmonary hypertension in the Pengo trial, the median systolic pulmonary blood pressure was 67.5 and the mean was 71.5(4). Far higher than the 40 mm Hg cutoff the authors of the MOFFET trial used for their definition of pulmonary hypertension. More importantly all 18 of the patients in the Pengo trial were class II or III on the NYHA disability scale, meaning they were all symptomatic.
In fact as a tool for the diagnosis of pulmonary hypertension, echocardiography proves to be somewhat problematic. The diagnostic accuracy of echocardiography in identifying patients suffering from pulmonary hypertension is questionable. When compared to the gold standard of catheterization the sensitivity and specificity is 83% and 72% respectively (8). Cleary it is not specific enough to be used as the sole criteria to diagnose pulmonary hypertension. When used alone, it will have a false positive rate of close to 30%. Furthermore when used as a screening tool in a population who is asymptomatic it will surely identify more false positives than true positives.
Clearly the discrepancies in the MOPPET trial are a result of workup bias and a disease-oriented outcome that is a poor surrogate for the important clinical outcome of true heart failure. It seems that when patients with submassive pulmonary embolisms are given thrombolytics they will have a lower pulmonary pressure as estimated by echo. Whether this is clinically important is unknown.
Unfortunately the MOPPET trial is unable to answer the question as to whether thrombolytics provide benefit to long-term outcomes in pulmonary embolism. Dr. Jeff Kline’s TOPCOAT study may provide us with more guidance in this direction (9). What is certain from the PEITHO trial is that thrombolytics provide very little benefit in the acute arena we work in every day and with the 2% absolute risk of ICH the potential benefit downstream would have to be quite significant to justify their use at all.
- Pulmonary Embolism Thrombolysis Study PEITHO [PDF]
- Pollack CV, Schreiber D, Goldhaber SZ, Slattery D, Fanikos J, O’Neil BJ, Thompson JR, Hiestand B, Briese BA, Pendleton RC, Miller CD, Kline JA. Clinical characteristics, management, and outcomes of patients diagnosed with acute pulmonary embolism in the emergency department: initial report of EMPEROR (Multicenter Emergency Medicine Pulmonary Embolism in the Real World Registry). J Am Coll Cardiol. 2011 Feb 8;57(6):700-6.
- Sharifi M, Bay C, Skrocki L, Rahimi F, Mehdipour M; “MOPETT” Investigators. Moderate pulmonary embolism treated with thrombolysis (from the “MOPETT” Trial). Am J Cardiol. 2013 Jan 15;111(2):273-7. doi: 10.1016/j.amjcard.2012.09.027. Epub 2012 Oct 24.
- Vittorio Pengo, M.D., Anthonie W.A. Lensing, M.D., Martin H. Prins, M.D., Antonio Marchiori, M.D., Bruce L. Davidson, M.D., M.P.H., Francesca Tiozzo, M.D., Paolo Albanese, M.D., Alessandra Biasiolo, D.Sci., Cinzia Pegoraro, M.D., Sabino Iliceto, M.D., and Paolo Prandoni, M.D. for the Thromboembolic Pulmonary Hypertension Study Group. Incidence of Chronic Thromboembolic Pulmonary Hypertension after Pulmonary Embolism. N Engl J Med 2004; 350:2257-2264
- Kline JA, Steuerwald MT, Marchick MR, Hernandez-Nino J, Rose GA. Prospective evaluation of right ventricular function and functional status 6 months after acute submassive pulmonary embolism: frequency of persistent or subsequent elevation in estimated pulmonary artery pressure. Chest. 2009; 136: 1202–1210.
- Becattini C, Agnelli G, Pesavento R, et al. Incidence of chronic thromboembolic pulmonary hypertension after a first episode of pulmonary embolism. Chest 2006;130(1):172-175.
- B. Tardya, C. Veneta, F. Zenia, M. Coudrota, S. Guyomarc’ha, P. Mismettib. Short term effect of recombinant tissue plasminogen activator in patients with hemodynamically stable acute pulmonary embolism: Results of a meta-analysis involving 464 patients. Thrombosis Research. Volume 124, Issue 6, December 2009, Pages 672–677
- Janda S, Shahidi N, Gin K, Swiston J. Diagnostic accuracy of echocardiography for pulmonary hypertension: a systematic review and meta-analysis. Heart. 2011 Jul;97(13):1112.