On June 15th 2016, I delivered an 8 minute talk arguing that we strongly consider the administration of systemic thrombolytics in patients with submassive pulmonary embolism at SMACCDub. This talk became infamous not for the content but for the masterful counterargument1 given by Iain Beardsell2. For fear of my PTSD re-emerging, I won’t spend too much time on the details of the actual pro-con except to say that lost in the fun, performance and abject dejection of yours truly, was the cogent arguments for and against the treatment. Simon kindly invited me to write up this post detailing my argument but you can also see the video of the actual debate here.1
The central theme in pulmonary embolism or, honestly, any disorder we look after is time. Time is the speciality of the Emergentologist. How much time does the patient have? How much time do I have to make the diagnosis? All ED interventions are there to buy the patient time. This idea rings true for the submassive PE: how much time does the patient have to break that clot down on their own? Do I need to simply watch and wait or do I need to help the process along?
The other central theme, one I discussed in more detail at my other SMACC talk, is that of thin-slicing disease3. The basic concept is that all disease exists on a spectrum from urgent, to emergent to critical. It’s our job to determine where on that spectrum the patient lies because patients on different parts of the spectrum can’t be treated the same. For example, we don’t treat mild hives and itchiness from an allergic reaction the same way we treat airway closure and hypotension. The aggressiveness of our interventions must match the severity of disease.
Pulmonary embolism clearly exists on a spectrum from the urgent end being the sub-segmental clots with mild symptoms to the peri-arrest (or arrested) massive PE. In between there are thousands of shades of severity. We define submassive PE as any embolism with right ventricular strain. Within even this slice of PE, there’s a considerable amount of shades of severity. There’s the 65 year old with multiple segmental PEs and a positive troponin who’s normoxic, normotensive, non-tachycardic and speaks in full sentences but there’s also the 45 year old with RV strain on ultrasound who is visibly dyspneic, hypoxic in the low 90’s and tachycardic. Both of these patients fall within the category of submassive PE since they are hemodynamically stable but their physiology and outcomes are quite different. Applying the same treatment to both patients is non-sensical.
Knowing this, I structured my talk at SMACC to highlight 3 major points:
- Pulmonary embolism patients exist on a spectrum of disease from stable to critically ill. We can’t apply the same treatment at all ends of the spectrum and expect the same outcomes.
- As with most critical care issues, the literature is inadequate and inconclusive. In the absence of robust evidence, we cannot have decision paralysis. In critical patients, time is critical. We must make a decision.
- There is a distinct subset of PE patients who will benefit from lytics. Know the indications, know the dose and pull the trigger.
The submassive PE patient may at first not seem to be at particular risk. After all, hemodynamically, they appear stable. Any hypotension, even episodic, should be highly alarming and make one consider the presence of massive PE. However, the appearance of stability is a false one. The presence of right ventricular strain jeopardizes the left ventricle since the two are in series. 10% of submassive PEs will progress to massive PE and while overall mortality is around 5%, it ranges up to 30% depending on degree of physiologic effect (Wood 2011 4). Both of these numbers are significantly higher than that for PE in general underscoring the potential severity of disease, but still considerably less than the > 50% mortality seen in massive PE. Understanding these mortality differences helps elucidate why the application of systemic lytics isn’t as simple in the submassive group as it is in the massive group.
Despite this complexity, a recent meta-analysis showed an association with decreased mortality in submassive PEs treated with throbmolytics with a number needed to treat of 65 (Chatterjee 20145). This benefit, of course, came with a tradeoff: increased major bleeding events (number needed to harm = 18). Less black and white; more grey.
Because mortality and mortality benefit are difficult to study, a number of researchers have focused on other important outcomes. Patients with submassive PE are at risk of developing RV strain in the short term and resultant pulmonary hypertension in the long term. Pulmonary hypertension is a terrible disease in terms of morbidity. It markedly affects quality of life as it results in chronic shortness of breath, decreased exercise tolerance and, frankly, an inability to participate in meaningful activity. Recent studies demonstrate that seemingly healthy submassive PE survivors were more likely to have decreased exercise tolerance (evidenced by decreased six-minute walk distance), poor heart rate recovery and RV dysfunction (Stevinson 20076, Chow 20147). It is clear that submassive PE, even when it doesn’t kill, has a significant potential to cripple.
Does the application of systemic thrombolytics in submassive PE patients decrease the risk of long-term functional impairment? This is the million dollar question. Unfortunately, the available data doesn’t give us a clear answer. The two studies best constructed to answer this question were the MOPETT and TOPCOAT studies. While both studies were methodologically sound, they asked slightly different questions, looked for different answers and, ultimately, can’t definitively answer the question at hand.
MOPETT looked at the reduction in pulmonary hypertension resulting from the use of low dose of alteplase. The results? Pulmonary hypertension was reduced in the thrombolytic group (16% vs. 57%). This gives an absolute difference of 41% and an NNT of 2.5 but, this is a small study (n = 121) looking at a surrogate endpoint. (Sharifi 20138)
TOPCOAT sought to determine if full-dose tenecteplase would decrease the proportion of patients with functional impairment at follow-up. The treatment group was found to have improved 90-day outcomes with an absolute difference of 22% for an NNT of 5. However, the study was terminated early due to the lead investigator (Jeff Kline) changing jobs and the outcome measure isn’t as clean as one would hope. (Kline 20149)
Finally, it’s important to briefly discuss the PEITHO trial as the three studies are often mentioned together (Meyer 201410). The PEITHO study looked at the use of full-dose tenecteplase in reducing hemodynamic decompensation and death; not long-term functional outcomes. While decompensation rates were improved, it came at the cost of increased bleeding and hemorrhagic stroke.
So where does this leave us? Overall, it looks like there’s a modest, but inconsistent reduction in death when systemic lytics are administered to submassive PE patients. There’s a signal in the studies of improved functional outcomes but the data sets are small and flawed. And, the elephant in the room, there is harm with this treatment – increased major bleeding and increased intracranial hemorrhage. Given all of this, how can I justify defending the use of systemic lytics?
It’s all about weighing pros and cons and exploring the details. When we separate out by age we find that major bleeding including intracranial hemorrhage isn’t increased in young (< 65 years) patients. The absence of this harm swings the argument in favor of treatment but not treatment for everyone. Based on my interpretation of the current literature, I think we should consider systemic lysis in the following groups:
- Patients with PE and hemodynamic instability (the massive PE)
- Young patients with good premorbid health with PE, hemodynamic stability but a risk for decompensation or bad functional outcomes evidenced by the presence of right heart strain
- That last one’s a mouth full but that’s for a reason. The decision to thrombolyse someone who’s doesn’t show frank hemodynamic instability is complicated and we shouldn’t try to simplify it. We need to understand the complexity and act accordingly. This is the right group to go after because they’re young, limiting the rate of ICH and they’re active and we want to preserve that function. The flip side of this is that I wouldn’t consider thrombolysis in an older patient with a poor pre-morbid state.
Ideally, you’ve got a PE team that you can discuss these patients with and you’ve got options aside from just systemic thrombolytics to consider (i.e. catheter directed lysis). In the absence of those resources, the decision will be left to us: systemic lytics or AC alone? Of course, the patient should and does have a say and shared decision making can be helpful. Once I’ve selected the patient I think would benefit (see above), I discuss the option with them. I usually frame it thusly:
“You have a disorder called a pulmonary embolism. This means there’s a blood clot in your lung that is making it difficult to breathe and giving you pain. You happen to have a significant clot that’s also making your heart work hard and it’s getting stressed. Although it’s not the norm, these clots can sometimes kill patients. The traditional treatment is to give you a blood thinner which doesn’t break up the clot but limits expansion of the clot while your body breaks it down on it’s own. There is another treatment we can give. It’s a medication that facilitates the breakdown of the clot. Based on the best information we have, this drug is unlikely to change whether you die from this clot or not. However, if you do survive, we believe that this drug will improve your chance of returning to your baseline functional state. This drug increases your risk of bleeding and can even cause bleeding into and around the brain. Because you are young and healthy, this risk is decreased but it still exists. Do you think you want to have this drug administered?”
It’s critical to note that assessing for stability and risk/benefit of thrombolytics is not a one time affair. As the patient’s course progresses, their clinical picture will change. Serial reassessment for the need for thrombolytics is essential.
This is the best we can do given the state of the evidence at this time. We may get better evidence in the upcoming years that changes this risk-benefit analysis but we may not. We are often asked to act with suboptimal information and submassive PE is no exception but we still must act, we must make a decision.
Anand Swaminathan M.D., M.P.H.
Assistant Professor Emergency Medicine
Bellevue/NYU Emergency Department
Editor-in-Chief Core EM coreem.net
Course Director, The Teaching Course flippingmeded.com
Special thanks to Simon Carley, Liz Crowe, Ross Fisher and Dan Horner for their thoughts, commentary and confirmation of my tenuous grasp of the English language.
11 thoughts on “The Argument for Systemic Thrombolytics in Submassive Pulmonary Embolism. Swami at St.Emlyn’s”
Thanks – useful and well written review.
It’s always so hard to consider evidence in the face of our own personal experiences:
First two patients I ever gave tPA to (for MI in 1995) both died of intracranial haemorrhage so I’ve been scared of thrombolytics since. That said 4/5 of those with PE I’ve thrombolysed have done well (and the one who didn’t had arrested mid CTPA).
One point: I’d suggest changing this rather loaded statement within your consent discussion “However, if you do survive, we believe that this drug will improve your chance of returning to your baseline functional state.” The ‘if you do survive’ overstates the mortality risk and would, in my view, make those hearing it think they are going to die – thus favouring the more aggressive intervention (despite the preceding statement). ‘You are likely to survive this event..’ might be a better wording.
What figures do you give the patient in terms of NNT and NNH?
Once again – many thanks
Jon Jones, Leeds
Jon – thanks for your thoughts and I agree, the wording probably needs a bit of a change in the informed consent.
As to NNT/NNH, hard to give great numbers based on the small studies. Though NNT/NNH are relatively easy statistical concepts to understand (at least in comparison to sens/spec or even LRs) I don’t know that the patient will really be able to incorporate all this information. Based on MOPPET/TOPCOAT the NNT is pretty low (< 5). Some of the studies show no major bleeding in the young and healthy but I wouldn't say the NNH would be infinite. In the handful of cases I've given lytics in submassive, I haven't given hard numbers for NNT/NNH. I've told the patient, "I think this drug is more likely to help you than to harm you."
Thanks Swami. As ever, providing informed consent in the emergency setting, with unclear evidence, is challenging and so often becomes “what do you recommend doc”. This article will certainly help me try and answer that question in this scenario.
Pingback: Treatment of Submassive Pulmonary Embolism (PE): Full Dose, Half Dose, or No Dose? - R.E.B.E.L. EM - Emergency Medicine Blog
Great post Swami.
Following on from John’s comments about NNT/NNH I sometimes think we don;t explore bleeding risk in enough detail with these patients. There is decent validated data on bleeding risk with thrombolysis (certainly from the MI/Stroke population, with a bit from the PE population – links below) and the balanced decision you talk about should really be a combination of both current/predicted physiological compromise, and personalised bleeding risk for that patient. I think you say this in the text when you talk about young vs old, but not sure it’s quite clear enough. Young people can of course still have a high individual bleeding risk in some circumstances.
I think numbers are really hard in this discussion, as the evidence is not perfect – certainly, if you are talking about full dose thrombolysis you can’t quote NNT/NNH data from MOPPETT for example. I prefer to give the patient my impression of their risk for thromboembolic deterioration and their risk of bleeding.
Tricky – great to see the evidence highlighted though.
My own experience (biased of course) and reading lines up with your excellent summary. I have lysed multiple patients often healthy young females who have big clot loads , but due to their excellent underlying physiology, are not shocked.
I have a collection of before and after echos that show a unbelievably positive improvement in RV function, I have not yet had a significant adverse event – and this includes lysing a guy who had haematuria coming out of bilateral neph tubes.
I have seen 2 young patients, who were not lysed due to strong preference of the accepting physician, who arrested within 24 hrs of admission – normal obs lying still in bed are never a great predictor of how you’ll go when you get up for a wee and knock off another bit of CFV clot.
Looking at the main studies out there , i suspect the reason that they don’t show much benefit is because of patient selection and randomisation ..I know that its akin to blasphemy , but to be honest, I really don’t think that RCTs here have looked at the right group of patients. Its hard to recruit , so they have lots of inpatients and lots of elderly folk. I never lyse older people with PE unless they are grossly unstable and I don’t treat inpatients – most of whom have serious underlying comorbidity which has contributed to their hospital stay and their PE. Randomisation has essentially matched each of my “Yes’s” with a “NO!”.
I only lyse ‘submassive PE’ patients that have been carefully selected for a favourable risk:benefit and who can participate actively in an informed consent process
Having worked through some of the MI lysis era, the serious bleeding rate was extremely low… zero in those <60 for me. We let junior docs lyse AMI – with the NNT for inferiors being really high – without much concern for the whole systemic vs local lysis argument . I suspect there are 2 reasons the evidence was soo much better for AMI than SM PE. First clot acuity – the AMIs have nice fresh clots that haven't turned into a gnarly old fibrinous scar (like a 6 week post #NOF inpatient) – and we picked them up early and treated early. Secondly patients predominately came from home, with their AMI being their only current acute medical problem.
So for me the younger patient with the newly broken leg or just off the international flight, who has lots of clot load and good underlying protoplasm is a great candidate for lysis – Im not sure outside of a massive multi centre trial targeting this demographic – that were going to get definitive evidence – but I for one think we'll be doing these patients a disservice if we wait for it.
Nick – thanks for your thoughts. My experiences mimic yours. I just finished reading the long term PEITHO outcomes as well as Rory Spiegel’s excellent, as always, review (https://emcrit.org/emnerd/em-nerd-case-shadowy-spector/). The article basically shows no benefit to lysis in submassive PE in terms of long term (24+ months) outcomes. However, I think the real take home is exactly what you are saying – we’ve looked at the wrong group. Submassive PE exists on a spectrum and on the severe end, I believe there’s a benefit in short term decompensation and death as well as long term outcomes. This will continue to be a hotly debated area but I like Rory’s take home – not every submassive PE needs lytics but there’s a select group that stands to benefit and that group is picked out based on your bedside evaluation.
Pingback: Lungemboli – submassiv – Mind palace of an ER doc
Pingback: Episode 38 - Do All Submassive PE’s Require Treatment with Thrombolysis? - R.E.B.E.L. EM - Emergency Medicine Blog
Pingback: The Critical Pulmonary Embolism Patient - R.E.B.E.L. EM - Emergency Medicine Blog
Pingback: emDOCs.net – Emergency Medicine EducationR.E.B.E.L. EM - The Critical Pulmonary Embolism Patient - emDOCs.net - Emergency Medicine Education