A Clot That Wasn't Idiopathic After All: What an Occult JAK2 Mutation Changes About Stopping Anticoagulation
Finding the cause of her portal vein clot should have made the decision easier — instead, discovering an occult blood cancer underneath it turned a finite-duration question into one about whether it's ever safe to stop.
E.M., a 52-year-old woman who teaches high school chemistry, came to the emergency department with three days of worsening abdominal pain and was found on CT to have an acute portal vein thrombosis, with no cirrhosis, no recent abdominal surgery, no malignancy on initial imaging, and no obvious local cause — an idiopathic-appearing splanchnic vein thrombosis, the kind of finding that, a decade ago, would have prompted anticoagulation and a relatively short workup before settling into a standard finite-duration treatment plan. Her complete blood count on admission looked almost unremarkable — hemoglobin and platelet count both within normal range, easy to read as reassuring evidence against an underlying blood disorder if nobody looked further.
Nobody stopped there, because idiopathic-appearing splanchnic vein thrombosis has a specific, well-documented association: a meaningful fraction of cases, when tested, carry the JAK2 V617F mutation and represent an occult myeloproliferative neoplasm — essential thrombocythemia or early polycythemia vera masked by the thrombosis's own effect on portal pressure and blood counts, rather than the truly idiopathic event it first appears to be. Her JAK2 test came back positive, and a bone marrow biopsy confirmed early-stage essential thrombocythemia she had no symptoms of and would not otherwise have been diagnosed with for years, if ever. The finding changes the entire frame of her original question. She came in asking how long she'd need to stay on a blood thinner for a single clot with an identifiable, now-treatable cause; the actual answer now under discussion is whether an MPN-driven prothrombotic state that doesn't go away just because its trigger has been identified should ever let her stop.
She has no family history of blood clots or blood cancers that she's aware of, and the essential thrombocythemia diagnosis was, by her own account, the single most unexpected piece of news she has ever received from a doctor — arriving, as she put it to her hematologist, 'completely out of nowhere, from a stomach ache.' That reaction is itself part of why the team is being deliberate about walking her through what indefinite anticoagulation actually means in daily terms, rather than assuming the clinical logic alone will make the recommendation feel less disorienting.
In follow-up, once the marrow biopsy confirmed the diagnosis
I'd anticoagulate her indefinitely. An occult MPN isn't a transient trigger the way surgery or a long flight would be — it's a persistent, ongoing prothrombotic state, and both ISTH and British Society for Haematology guidance lean toward indefinite anticoagulation in MPN-associated splanchnic vein thrombosis for exactly that reason. Finding the cause here doesn't mean the cause has stopped being active.
I'd want to separate 'finding the cause' from 'the cause remaining fully active,' though — that's a real distinction, not a semantic one. She's about to start cytoreductive therapy that should normalize her counts and reduce her JAK2 allele burden. Treating this as a fixed-duration provoked event, with a real reassessment once cytoreduction has had time to work, is what we'd do for almost any other identified cause.
Calling this identical to a transient surgical or travel trigger assumes cytoreduction fully resolves the same underlying risk a completed surgery does — that's the actual open question, not a settled equivalence.
I manage her essential thrombocythemia going forward, and I want to be honest about where the evidence actually sits rather than pick a side that sounds more resolved than it is: we don't have dedicated data confidently telling us that a treated allele burden and normalized counts safely lower her recurrence risk enough to stop anticoagulation. Indefinite therapy is the current guideline default specifically because that individualized answer doesn't exist yet, not because anyone has proven cytoreduction doesn't help.
I'd start her on indefinite anticoagulation today, being direct with her that this is the honest, current standard rather than a permanently closed question — and revisit it explicitly if better evidence on treated-MPN recurrence risk becomes available during her care, not on a fixed calendar date.
Not fully agreed on the underlying question of whether treated MPN eventually justifies stopping anticoagulation, but agreed on today's plan: rivaroxaban continued indefinitely, hydroxyurea started for her essential thrombocythemia, and an explicit note in her chart that duration will be revisited if dedicated evidence on treated-MPN recurrence risk becomes available, rather than left as a permanently closed decision.