Subsegmental PE: a hands-off approach

A recent research letter suggested that there was significant overtreatment of patients with subsegmental pulmonary emboli, and a pretty strikingly high risk of serious adverse events (see dvt subsegmental PE overtreatment jamaintmed2018 in dropbox, or  doi:10.1001/jamainternmed.2018.2971 )

Details:

-- retrospective review of all CT pulmonary angiograms (CTPAs) in 1 Canadian tertiary care hospital, from 2014 to 2016

-- 223 PEs were classified as proximal, lobar, segmental, or subsegmental pulmonary emboli, based on the most proximal embolus

-- mean age 68, 38% with cancer

-- Doppler ultrasounds of the leg veins when done were performed within 2 weeks of the CTPA

 

Results:

-- 79 (36%) of the PEs were subsegmental

    -- 32 of these (41%) had a Doppler ultrasound done, and 8 (25%) were positive for DVT

    -- 62 of 71 (87%) were systemically anticoagulated [excluding the 8 with DVT who were appropriately anticoagulated] vs 135 of 143 (94%) of those with proximal emboli. [ie, similar numbers anticoagulated with subsegmental vs proximal PEs]

-- Adverse events in the 3 months of follow-up in those who were anticoagulated:

    -- in those with isolated subsegmental PE:

        -- 26 patients (42%) had emergency department visits or were readmitted for reasons unrelated to venous thromboembolism (VTE), 21 (34%) had a decrease of hemoglobin level of 2 g/dL or more and received a blood transfusion, and 10 (16%) died though none of the deaths were related to VTE

    -- in those with proximal PE:

        -- 52 patients (39%) had emergency department visits or were readmitted for reasons unrelated to VTE, 40 (30%) had a decrease of hemoglobin level of 2 g/dL or more and received a blood transfusion, and 13 (10%) died

 Commentary:​

-- Subsegmental PEs are diagnosed quite regularly these days, more than 10% of all PEs. This number has increased dramatically as more CT pulmonary angiography (CTPA) is done, and especially as these CT units become more sophisticated, using multi-detector CTPAs. The clinical question is what is the best therapy??, given that there are clear risks from anticoagulation as noted in the above study.

-- Unfortunately, we do not have a great RCT to define what is best to do when patients have subsegmental PEs. 

    --However, this situation is not so different from the typical clinical conundrum: applying the results of even a really good RCT for a therapy involving predominantly middle-aged white patients who are free of other major medical comorbidities to the individual elderly Latino patient in front of you with chronic kidney disease and heart failure. In clinical care, we are constantly making what we feel is the best guess for the particular patient we are seeing​, based on large studies from dissimilar patient populations with specific inclusion/exclusion criteria .

-- The 2016 guidelines for Anti-thrombotic Therapy for VTE Disease (see dvt guidelines chest2016 in dropbox or https://journal.chestnet.org/article/S0012-3692(15)00335-9/pdf ) state “For subsegmental pulmonary embolism and no proximal DVT, we suggest clinical surveillance over anticoagulation with a low risk of recurrent VTE (Grade 2C), and anticoagulation over clinical surveillance with a high risk (Grade 2C)”. They do have the following provisos:

    -- patients with subsegmental PEs who are not anticoagulated should have negative bilateral ultrasound examinations of their legs (and upper extremities if at risk)

    -- it may be reasonable to do one or more follow-up ultrasound examinations of the legs to detect an evolving proximal DVT; this serial testing has been shown to be a safe strategy for patients suspected of having a PE but have nondiagnostic ventilation-perfusion (V/Q) scans (who, by the way, may have had subsegmental PEs: the old PIOPED study found that 17% of patients with a low probability V/Q scan had subsegmental PEs on pulmonary angiography). In fact, those with low or intermediate V/Q scans for suspected PE were prospectively managed safely without anticoagulation if the pretest probability of a PE were low and compression ultrasonography of the legs was negative. It is important to remember that 50% of patients with a PE have an underlying asymptomatic DVT (ie, one cannot rely on symptoms to diagnose a DVT).

    -- patients who are at higher risk for recurrent or progressive VTE should be anticoagulated, e.g. those with active cancer, recent surgery, low cardiopulmonary reserve or marked symptoms

    -- those at higher risk of bleeding should favor no anticoagulant therapy

    -- the decision to anticoagulate should be sensitive to the patient’s preferences

-- the argument to me for following these guidelines is as follows (also see dvt subsegmental PE jthrombhem2012 in dropbox, or DOI: 10.1111/j.1538-7836.2012.04804.x ):

    -- the interobserver agreement for subsegmental PE by radiologists is not great: one study found that there was only 51% agreement. And, these small subsegmental findings are, not infrequently, felt to be artifacts and NOT real (these subsegmental PEs are typically <2.5 mm in diameter).

    -- there were no cases of recurrent VTE in retrospective reports that included 60 patients with subsegmental PEs and no proximal DVT who were not anticoagulated, though this was not found in another study of patients with symptomatic subsegmental PEs.

    -- The incidence of subsegmental PEs has largely increased because of the improved detection through the new scanners with increasing number of detectors: there has been a 5.4% average annual increase in detection of these PEs. BUT there has been no change in the number of central or fatal PEs over time [this is yet another example of a technological "advance" leading to lots more (ie too much) information, but clinically this information may be misleading or not clinically relevant, as with lots of totally asymptomatic patients with lower back MRIs that look pretty bad and could actually indicate the need for surgery, eg see doi.org/10.3174/ajnr.A4173​ ]

    -- another study comparing CTPA with V/Q scan found more PEs diagnosed with CTPA, but the rate of VTE over the next 3 months was similar in untreated patients where PE was excluded by either of these diagnostic strategies (ie, even with V/Q scans which do not typically pick up the subsegmental PEs)

    -- also, the sensitivity of D-dimer is only 76% in those with subsegmental PEs on pulmonary angiography, which suggests that many patients who go the the ED, have d-dimers done which are negative, are never diagnosed with subsegmental PE. and there is no evidence of lots of subsequent VTE

-- another retrospective cohort study found that 5.3% of patients with isolated subsegmental PE on anticoagulation had a major bleeding episode

-- a recent systematic review/meta-analysis of observational studies of outcomes of patients with subsegmental PEs with and without anticoagulation (see dvt subsegmental PE AcadEM2018​ in dropbox, or Bariteau A. Academic Emergency Medicine. 2018; 25: 828.) found:

    --14 studies with 15,563 patients and pooled prevalence of subsegmental PE=4.6%. 589 were treated with anticoagulation and 126 were not

    --frequency of bleeding in those on anticoagulation was 8.1% (no data on those not on anticoagulation, and no definition of "bleeding")

    --frequency of VTE recurrence within 90 days was 5.3% in those treated with anticoagulants vs 3.9% untreated (!!!???!!!)

    --frequency of death was 2.1% if treated vs 3.0% untreated (though with overlapping confidence intervals....)

        --BUT: 
            --no reported rates of concomitant cancer or DVT (and, as above, 50% with PEs have asymptomatic DVTs and SHOULD be anticoagulated)

            --rather surprising that the rate of recurrent VTE was higher in those treated with anticoagulation!!!

            --the risk of bleeding in those anticoagulated was a whopping 8.1% (way higher than in those anticoagulated for regular PEs, typically in the 1% range). ?why ?unknown adverse selection bias by clinicians of those to be treated with anticoagulants ?are those patients who develop subsegmental PEs more likely to have cancer or other bad diseases than those who get the big PEs?

so, my sense of the most appropriate middle-ground in this increasingly common problem of subsegmental PEs, which we see pretty frequently in primary care (it seems that a really high percentage of patients going to the ED with any cardiorespiratory condition get these really sensitive, multi-detector CTPAs), is the following (since we are lacking good studies):

    --check for DVTs with leg ultrasounds (arms if high risk, as with catheter in place). if positive, anticoagulate

    --strongly consider repeating the ultrasound a week or so later to see if proximal progression of a DVT requiring anticoagulation

    --and, then discuss with patient, with bias not to anticoagulate but follow the patient (and patients should know that the data are murky, suggesting a perhaps higher risk of bleeding than with regular PEs; this is also confirmed in the above Canadian study: the death rate may be [nonsignificantly] higher without anticoagulation). and i would lean much more to not anticoagulating if bleeding tendency, risk of falls

    --this is basically the conclusion of the 2016 guidelines for Anti-thrombotic Therapy, though (as they suggest) would be useful to have a randomized controlled trial.  But in the meantime....

geoff

If you would like to be on the regular email list for upcoming blogs, please contact me atgmodest@uphams.org

to get access to blogs since 8/15/17:

1. go to http://gmodestmedblogs.blogspot.com/ to see them in reverse chronological order

2. click on 3 parallel lines top left, if you want to see blogs by category, then click on "labels" and choose a category​

3. or you can just type in a name in the search box and get all the blogs with that name in them

to access older blogs from the BMJ website, from October 2013 until 8/15/17: go to http://blogs.bmj.com/bmjebmspotlight/category/archive/ 

please feel free to circulate this to others. also, if you send me their emails, i can add them to the list​