Keywords
mesenteric venous thrombosis - sharp recanalization - superior mesenteric vein - occlusion
Introduction
Mesenteric venous thrombosis (MVT) disrupts bowel venous return. Causes include insult
from surgery or inflammation, mass compression, hypercoagulable states, or congestion
from cirrhosis or heart failure. The superior mesenteric vein (SMV) is most affected.
Chronic MVT represents 20 to 40% of MVT cases and presents with nonspecific abdominal
symptoms or variceal gastrointestinal (GI) hemorrhage. Standard diagnosis is via contrast-enhanced
computed tomography (CT).[1] Recanalization of chronically occluded SMV can be challenging. This report describes
a recanalization technique for a moderately long SMV occlusion, utilizing direct transabdominal
puncture of the SMV upstream of the occlusion as a target for retrograde downstream
sharp recanalization.
Case Report
A 72-year-old woman was referred for chronic SMV occlusion following Whipple procedure
with SMV sleeve reconstruction for pancreatic head adenocarcinoma. SMV thrombosis
occurred repeatedly requiring anticoagulation/antiplatelet therapy and retrograde
endovascular recanalization. Eventually chronic occlusion occurred despite anticoagulation
and endovascular treatments. The patient later presented with duodenal variceal bleeding.
CT demonstrated occlusion of the superior segment of SMV and narrowing of the main
portal vein (MPV)–splenic vein (SV) junction. A more aggressive recanalization approach
was undertaken.
The procedure started with transhepatic portal vein access. There was no portal hypertension
in the downstream intrahepatic portal system, as expected by the absence of liver
disease. Pressures across the MPV–SV stenosis revealed a significant gradient of 13 mm
Hg upstream of the stenosis ([Fig. 1A]). Traditional SMV recanalization attempts (catheter, microcatheter, and glidewires)
and attempts for reaching the patent upstream SMV via inferior mesenteric vein (IMV)
collaterals to try an antegrade recanalization failed.
Fig. 1 (A) Venography of the portal venous circulation using a KMP catheter (Cook Medical,
Indiana, United States) demonstrating occlusion of the proximal SMV with a small round
stump (arrow), and stenosis at the confluence of the portal and splenic veins (arrowhead).
Clips in the abdomen are mostly from previous Whipple procedure with SMV grafting
for pancreatic cancer. (B) Direct percutaneous puncture of the SMV distal to the occlusion with a 21-gauge
Chiba needle, followed by advancement of a 2.6 Fr microcatheter (white arrow) bare-back
over a placed guidewire into the SMV, showing complete occlusion just proximal to
the confluence of the main SMV branches (black arrow), and flow diversion through
prominent collateralization (arrowhead). (C) Use of the support catheter (Navicross, Terumo Medical, Tokyo, Japan; arrow) inserted
through the direct SMV puncture as target for distal landing of sharp guidewire tip
advanced through the percutaneous hepatic access sheath/catheter (arrowhead). (D) Successful advancement of the sharp guidewire tip and sharp recanalization past
the occlusion (arrow), with contrast injection verifying intraluminal position. (E) Placement of a 6 × 40 mm S.M.A.R.T. CONTROL Nitinol stent (Cordis, Florida, United
States; arrows demarking inferior and superior ends of the stent) with subsequent
balloon angioplasty, showing adequate flow through the stent, disappearance of the
multiple collaterals, and no evidence of flow obstruction. SMV, superior mesenteric
vein.
An ultrasound-guided direct percutaneous access to the upstream patent SMV was achieved
using a 21-gauge needle, allowing bare-back placement of a 0.018 guidewire and a 2.6
Fr Navicross (Terumo, Tokyo, Japan) microcatheter ([Fig. 1B, C]). Antegrade recanalization attempts from the upstream access also failed. Retrograde
sharp recanalization using the upstream access as a target was then considered as
a last resource.
Through the transhepatic portal venous access catheter, the back end tip of a glidewire
(Terumo)—bent for directionality and sharpened with a blade—was used to pierce through
the occluded SMV segment using the Navicross catheter placed at the inferior occlusion
stump as target. Successful intraluminal crossing without extravasation was confirmed
by contrast injection ([Fig. 1D]). Over a wire, the crossed-occluded segment was angioplastied and bare-stented.
Venography showed adequate flow through the stent with disappearance of the multiple
collaterals ([Fig. 1E]). The MPV stenosis was also angioplastied with reduction in pressure gradient to
5 mm Hg. Portal venous access hemostasis was achieved using an absorbable gelatin
sponge pledget injected via the sheath into the peripheral intrahepatic tract. No
complications occurred and the patient was discharged after 48 hours.
Three months later, investigation of repeat GI hemorrhage demonstrated re-thrombosis
of the mesenteric stent. Correcting this required transjugular intrahepatic portosystemic
shunt (TIPS) creation and repeat angioplasty of the SMV stent and stenting of the
MPV–SV stenosis ([Fig. 2]). Further re-thrombosis did not occur and GI hemorrhage resolved.
Fig. 2 (A) Axial and (B) sagittal contrast-enhanced CT of the abdomen, showing SMV stent with internal hypodensity
representing developing thrombus (arrows). (C) Eventual occlusion of the SMV stent confirmed by venography. (D) Venography demonstrating improved flow following TIPS (white arrow showing inferior
border of stent), splenoportal junction stenting with 12 × 50 mm Protege stent (Medtronic,
Minnesota, United States; black arrows showing borders) and presence of patent SMV
stent (arrowheads showing superior and inferior borders). (E) Follow-up coronal CT of the abdomen showing patent TIPS stent (white arrows), splenoportal
junction stent (black arrows), and SMV stent (arrowhead showing superior border).
CT, computed tomography; SMV, superior mesenteric vein; TIPS, transjugular intrahepatic
portosystemic shunt.
Discussion
Sharp vascular recanalization, including the use of the back end of a wire, mostly
described in superior vena cava occlusions, carries risks of extravascular perforations.[2] When the vessel to recanalize is straight, such approaches can be done with relative
safety and ease. However, when there are vascular junction points or angles in the
vessel, such as with SMV–MPV junction, the sharp recanalization tool needs to be shaped
appropriately to maneuver through these angles while reducing the risk of extravascular
perforation. In such cases, having a distal target downstream of the angled occlusion
is critical to confirm alignment of the shaped sharp recanalization tool and target.
Our procedure was unique because it was performed in the mesenteric system with the
target placed via a direct puncture of the SMV. Percutaneous SMV access has mainly
been reported in the context of nonsharp main portal vein recanalization–TIPS,[3] and is limited by operator experience, patient body size, and habitus or anatomical
constraints.[4] Other techniques to recanalization of occluded vessels including use of endovascular
radiofrequency (RF) wire or surgical exploration might also have been options. In
this case, difficulties that can occur in directing an RF wire across an angled vascular
segment and hostile abdomen from previous surgery seem to make an argument disfavoring
these approaches.
In the absence of known hepatic disease in our patient, TIPS was initially not thought
to be necessary to maintain SMV stent patency, which was ultimately proven to be a
mistake. In retrospect, a preemptive TIPS at the initial setting would likely have
been useful to improve flow through the SMV stent and maintain primary patency. In
addition, although there was a stenosis at the MPV/SV confluence, this was not involving
the recanalized/stented SMV and was thought to be noncontributary to the SMV occlusion.
In fact, the opinion was that reduced SV to MPV flow may, in fact, redirect flow through
the IMV into the SMV system and improve flow through the recanalized/stented SMV and
help maintain patency. It is unclear whether stenting the MPV–SV stenosis from the
start could have improved primary SMV stent patency.
A rendezvous technique, combining antegrade–retrograde approach, traditionally described
in chronic total occlusion in peripheral artery disease,[5] combined with sharp recanalization, can prove to be a promising avenue to explore
for similar cases of chronic mesenteric vein occlusions.
Conclusion
Sharp recanalization of SMV occlusions is a feasible technique that should be considered
after traditional methods fail. This case showcases the utility of direct transabdominal
upstream SMV puncture, allowing placement of a catheter, the tip of which positioned
upstream of the occlusion as a target for sharp retrograde recanalization combined
with orthogonal fluoroscopic views reduces risks of extravasation and catastrophic
hemorrhage.
