Pulmonary Embolism: Anticoagulate, Thrombolyse, or Wait?

Three Questions, One Blood Clot

A 58-year-old woman with a recent knee replacement comes in short of breath. Heart rate 118, oxygen saturation 91 percent on room air, a subtle right leg that is a little more swollen than the left. You order the CT angiogram, and while you wait, three questions are forming in your head:

• Do I just start anticoagulation and watch?
• Does this patient need a thrombolytic right now?
• Or is she stable enough to go home on a DOAC today?

Those three questions are the entire practice of pulmonary embolism management. Get the answer wrong in either direction and you hurt the patient: too much thrombolysis and you cause a bleed, too little and you watch her right ventricle fail. The good news is that the framework for answering them just got a major upgrade.

The 2026 AHA/ACC Guideline Killed "Massive" and "Submassive"

For years we sorted PE into three buckets: massive (the patient is in shock), submassive (normotensive but with right ventricular strain), and low risk. The problem with "submassive" is that it lumped together two very different patients. One is hypoxic, diaphoretic, and quietly deteriorating. The other has a mildly elevated troponin, feels fine, and is asking when she can leave. Same label, opposite management needs.

The first formal AHA/ACC clinical practice guideline for acute PE, published in Circulation in March 2026, replaced that language with a five-category system. It is worth learning, because it maps directly onto what you should actually do:

• Category A: Subclinical. An incidental PE found on a CT done for something else, no symptoms. These patients can go home.
• Category B: Symptomatic but low severity score (PESI class I to II, sPESI 0, or Hestia 0). Early discharge is generally recommended.
• Category C: Symptomatic with an elevated severity score (PESI III to V, sPESI 1 or more, Hestia 1 or more). Subcategories C1 through C3 separate patients by right ventricular dysfunction and biomarker elevation.
• Category D: Incipient cardiopulmonary failure. The patient looks "fine" on the blood pressure cuff but has rising lactate and creatinine. This is the normotensive shock that catches people off guard.
• Category E: Cardiopulmonary failure. E1 is persistent hypotension with cardiogenic shock; E2 is refractory shock or cardiac arrest.

A respiratory modifier (R) can be appended to any category when there is hypoxia, tachypnea, or escalating oxygen needs. Categories are assigned by the most severe indicator, and patients can move between categories over time, which is a reminder that PE is a dynamic condition, not a one-time label.

Category A and B: The Art of Sending People Home

Outpatient management of low-risk PE is no longer a radical idea. The guideline gives it a Class 2a recommendation, backed by the HOME-PE trial, which showed that the Hestia rule and sPESI performed equally well for selecting patients who can safely be discharged.

sPESI is the score to know. It is six items, one point each: age over 65, history of cancer, chronic cardiopulmonary disease, heart rate over 110, systolic blood pressure under 100, and oxygen saturation under 90 percent. A score of zero carries about a 1.1 percent 30-day mortality, low enough that outpatient care is reasonable. A score of 1 or more jumps to roughly 8.9 percent, and that patient belongs in the hospital.

Two caveats. First, the score only works after the diagnosis is confirmed; it is a risk stratification tool, not a diagnostic one. Second, "outpatient reasonable" assumes the patient has a safe place to go, reliable follow-up, and a way to take anticoagulation without falling off the face of the earth. The algorithm does not replace judgment.

Category C: The Gray Zone Where Most Decisions Happen

Category C is where the interesting medicine lives, and where the old "submassive" label did the most damage. The 2026 guideline is unusually blunt about the low end of this category: for patients in A through C2, systemic thrombolysis is classified as harmful (Class 3), and catheter-based therapy is not recommended. The evidence says the bleeding risk outweighs the benefit when the patient is not actually failing.

Category C3 is where it gets genuinely uncertain. The PEITHO trial, the landmark study of thrombolysis in intermediate-risk PE, showed that alteplase reduced the rate of cardiovascular collapse, but at the cost of more major bleeding, including intracranial hemorrhage. The guideline calls the role of thrombolysis and catheter-based therapy in C3 "unclear," which in guideline-speak means: this is a shared decision, ideally made with a PE response team, not a solo call at 2 a.m.

That is one of the guideline's quiet big wins: activation of PE response teams (PERT) for categories C through E is now a Class 1 recommendation. If your hospital does not have a PERT, this is the year to build one.

Categories D and E: When the Patient Is Failing

Category D is the patient who has not crashed yet but is on the way. Normotensive shock. Rising lactate. A right ventricle that is visibly dilated on the CT. For these patients, the guideline says advanced therapies "may be considered": systemic thrombolysis, catheter-directed lysis, or mechanical thrombectomy.

Category E is the crash. Persistent hypotension, refractory shock, cardiac arrest. Here, advanced therapies are "reasonable": thrombolysis, catheter-directed lysis, mechanical thrombectomy, or surgical embolectomy. In E2 (refractory shock or arrest), systemic thrombolysis is the reasonable default, and VA-ECMO becomes the bridge when the patient is not a lysis candidate or lysis has failed.

One trial result deserves a highlight because it simplifies a fight that used to consume PERT meetings: PEERLESS randomized 550 high-risk PE patients to catheter-directed lysis versus mechanical thrombectomy and found no significant difference in 30-day mortality or major bleeding. If you are choosing between the two, the evidence says it is an operator and situation decision, not a coin flip with stakes.

Diagnosis: Do Not Image Everyone

Before any of this management, you have to make the diagnosis, and the smartest move is often to avoid the scanner. The classic pathway runs through pretest probability.

Wells criteria is the workhorse. Points for clinical signs of DVT (3), PE being the most likely diagnosis (3), heart rate over 100 (1), immobilization or recent surgery (1), prior VTE (1.5), hemoptysis (1), and active cancer (1). In the two-tier model that guidelines favor, a score of 4 or less is "PE unlikely," where a high-sensitivity D-dimer can rule the diagnosis out. A score above 4 is "PE likely," where you go straight to imaging. The one-tier logic is simple: a negative D-dimer in a low-pretest-probability patient ends the workup, and a positive one does not prove anything, because D-dimer is sensitive but its specificity hovers around 50 percent.

Revised Geneva score is the alternative if you prefer all-objective criteria. It scores age over 65, prior VTE, recent surgery or fracture, active cancer, unilateral leg pain, hemoptysis, heart rate, and leg tenderness with unilateral edema. No subjective "PE is most likely" item, which removes one source of bias. Low is 0 to 3, intermediate 4 to 10, high 11 or more.

Two practical refinements worth knowing. For patients over 50, an age-adjusted D-dimer cutoff (age times 10 ng/mL for FEU assays) is validated and meaningfully improves specificity. And if the pretest probability is truly low, the PERC rule can stop the workup before any blood is drawn.

When imaging is needed, CTPA is the default: sensitivity in the mid-90s percent, immediate availability, and the bonus of showing right ventricular size, which feeds directly into your category assignment. A V/Q scan is the alternative when contrast is a problem, renal function is poor, or you are dealing with a pregnant patient, and a normal V/Q scan rules PE out just as cleanly as a normal CT. In the unstable patient, remember the cardinal rule: never delay resuscitation to get the scan.

Anticoagulation: The Default Answer

For the large majority of PE patients, the answer to all three opening questions is the boring one: anticoagulate. The 2026 guideline makes the drug choices explicit.

DOACs over warfarin for any patient eligible for oral anticoagulation (Class 1). Lower bleeding risk, especially intracranial hemorrhage, no routine monitoring, no dietary gymnastics. LMWH over unfractionated heparin when parenteral anticoagulation is needed (Class 1), with less recurrent VTE, predictable dosing, and a lower risk of heparin-induced thrombocytopenia.

The special populations are where the nuance lives:

• Obesity: DOACs are reasonable even in severe obesity. Meta-analyses show apixaban and rivaroxaban are at least as safe and effective as warfarin in this group, which is a relief for everyone who has ever fumbled with weight-based heparin dosing at 240 kilograms.
• Antiphospholipid syndrome: warfarin still wins. DOACs are associated with more arterial thrombotic events in thrombotic APS, with a narrow exception for single-antibody, low-risk patients without arterial history.
• Pregnancy: LMWH or UFH only. DOACs and warfarin are classified as potentially harmful.
• CKD: DOAC over warfarin in stages 2 to 3 (Class 1, Level A). In stages 4 to 5 or dialysis, apixaban may be considered, but the data are genuinely uncertain.
• Liver disease: DOACs are reasonable in Child-Pugh A and B; avoid them in Child-Pugh C.

Duration follows the usual logic: three months for a provoked PE with a reversible risk factor, and a conversation about extended therapy for unprovoked or recurrent events. Here is a recent development that changes that conversation: for extended secondary prevention, the guideline now recommends half-dose apixaban (2.5 mg twice daily) or rivaroxaban (10 mg daily) over full dose, based on the RENOVE and API-CAT trials. Same protection against recurrence, meaningfully less bleeding. If you have a patient on full-dose rivaroxaban for "lifelong" prophylaxis, this is the year to re-dose them.

Thrombolysis: Dosing and Decision-Making

When you have decided that a patient needs systemic thrombolysis, the standard alteplase regimen is 100 mg: a 10 to 20 mg bolus, then the remainder over two hours. In cardiac arrest, many teams give a 50 mg bolus. Tenecteplase is the single-bolus alternative, dosed by weight (30 mg under 60 kg up to 50 mg over 90 kg), and it is the one to reach for if it is what your hospital stocks.

The decision itself should be anchored to the category system and the contraindication list. Recent major surgery, recent intracranial pathology, active bleeding, and uncontrolled severe hypertension are the usual deal-breakers. When a contraindication exists but the patient is still failing, that is the moment catheter-based therapy or surgical embolectomy enters the conversation, ideally with PERT at the table.

Resuscitation Pearls for the Crashing PE

High-risk PE is fundamentally right ventricular cardiogenic shock, and the resuscitation follows that physiology.

• Oxygen first. Maximize FiO2. Oxygen is a pulmonary vasodilator, and a nonrebreather or high-flow cannula is doing real work, not just supporting saturation.
• Epinephrine is the first-line inotrope. The beta-agonist effects dilate the pulmonary vasculature, boost right ventricular contractility, and provide the chronotropy that matters when the terminal rhythm of PE is often brady-asystolic. Titrate to a MAP above 65 and do not fear high doses.
• Vasopressin is second-line, adding systemic vasoconstriction with some pulmonary vasodilation, at roughly 0.04 to 0.08 units per minute.
• Inhaled pulmonary vasodilators (nitric oxide, inhaled epoprostenol, nebulized nitroglycerin or milrinone) are the next rung if the patient is refractory. The iNOPE trial showed inhaled nitric oxide increased the chance of a normal-sized right ventricle in intermediate-risk PE with no adverse events.
• Fluids are rarely helpful. The right ventricle is already dilated against high filling pressures, and more volume worsens the dilation and the interventricular septal bowing that steals from the left side. Give fluid only with clear evidence of hypovolemia, and a small IVC in a "high-risk PE" patient should make you question the diagnosis.

One warning deserves its own paragraph: sedation is dangerous in these patients. The 2026 guideline carries a Class 3 (Harm) recommendation against deep sedation and mechanical ventilation in categories C through E unless clinically necessary. Case series report cardiac arrest rates of 19 to 28 percent after anesthesia induction in PE patients with right ventricular dysfunction, even when they were hemodynamically stable beforehand. The sympathetic tone is what is keeping them alive, and the induction drug takes it away. If intubation is truly required, have vasopressors, inotropes, and VA-ECMO at the bedside before you start.

What Else Is New

A few threads from the last couple of years are worth tracking. Ultrasound-facilitated catheter-directed fibrinolysis, reported in the New England Journal of Medicine in 2025, uses low-dose tPA with ultrasound assistance to try to keep the clot-busting benefit while cutting the bleeding risk, and it is one of the more promising refinements of catheter-based therapy. The PE-TRACT trial is testing catheter-directed therapy against anticoagulation alone in intermediate-high-risk PE, and its results will matter for the C3 category. And on the follow-up side, the guideline now pushes structured post-discharge monitoring, a visit within the first week, a dedicated visit by three months, and screening for at least a year, aimed at catching chronic thromboembolic pulmonary disease, which complicates roughly 3 percent of acute PEs and is treatable if you find it early.

The Short Version

Every PE patient lands in one of five categories, and the category answers your three questions. A and B: anticoagulate and consider sending them home, with sPESI or Hestia in your back pocket. C: anticoagulate, stratify with biomarkers and right ventricular imaging, and keep thrombolysis in the drawer unless you are at C3 with a team behind you. D and E: this is resuscitation plus a real decision about thrombolysis, catheter therapy, or surgery, made with PERT, not alone. Anticoagulation is the default and the DOAC is the drug of choice, with warfarin keeping its throne in antiphospholipid syndrome and LMWH holding pregnancy. And when the patient is crashing, oxygen, epinephrine, caution with fluids, and a very low threshold to avoid the anesthetic machine.

The clot is the same clot it was ten years ago. The map for managing it is finally clear.