Treatment options for chronic venous occlusions (symptoms for more than 14 days or imaging evidence of chronic thrombus changes such as intraluminal calcifications and prominent venous collaterals) range from surgical bypass grafts to prosthetic graft reconstruction to endovascular angioplasty and stenting (McDevitt et al. 2019). Endovascular treatment is generally considered the first-line therapy in adults and is becoming more widely accepted in pediatrics (Rizvi et al. 2008). Endovascular techniques limit the amount of blood loss, allow for multiple access sites to be re-established during a single procedure, and have favorable long-term patency rates (Sullivan et al. 2018). The importance of pre-procedure planning and high quality pre-procedure imaging cannot be overstated when planning a venous reconstruction. Additionally, even less complex cases of line removals in patients with limited central venous access, should be performed with an intent to preserve and clearly document the current state of SVO with dedicated venograms (Sullivan et al. 2018; Gnannt et al. 2019). Venous access for venous recanalization often requires at least 3 sites of access (range 2–4) and may include lower extremity (femoral or popliteal), upper extremity (basilic, cephalic or brachial), and/or internal jugular venous access (McDevitt et al. 2019).
A central tenet of venous reconstruction recommends operators work from a “normal” beginning point peripherally to a “normal” ending point centrally (van Vuuren et al. 2018). Blunt recanalization techniques involve crossing the occluded vessel length with a hydrophilic, stiff wire and using either the front or back end of the wire to gain purchase centrally (Fig. 3). Vessel size and the chronicity of the occlusion guide the wire choice for recanalization. Microwires used for treating chronic coronary occlusions (examples: 0.014″ Whisper wire, Abbott Vascular Santa Clara, CA and 0.014″ Asahi Confianza, Asahi-Intecc Tustin, CA) have been successful in crossing long lengths of occluded veins with virtually no traumatic insult in the author’s experience (Lawson and Seckeler 2018). While the wire purchase is maintained as centrally as possible, a microcatheter, diagnostic hydrophilic catheter, or small bore sheath is then advanced over the wire. This type of technique is particularly useful in pediatric patients and peripheral venous occlusions. Blunt recanalization can be successful and may make it easier to recanalize the venous outflow through the true lumen of the vessel with reported rates of technical success in the lower extremities and pelvis of up to 86% (McDevitt et al. 2019; Lawson and Seckeler 2018). A peak or cap is often present at the peripheral border of the venous occlusion, marking the antegrade path into the native, occluded vessel. If a venous cap is present on the planning venogram (Fig. 4), efforts for recanalization should be focused on that precise location. If central venous wire access is achieved through collateral vessels in cases of central venous agenesis, the decision to reconstruct through the collateral pathways needs to be carefully considered (Fig. 5). The long term patency rates of long length stent complexes are low, but an emergency situation such as extensive thrombus burden in the lower extremities or severe post thrombotic syndrome may warrant total reconstruction of the IVC in order to provide venous outflow for clearance of thrombus or resolve severe symptoms.
Sharp recanalization usually is performed with small gauge, non-coring needles under US, fluoroscopy, or CBCT guidance. The needle length can be variable depending on patient size and distance from the access site to the occlusion (7 cm, 10 cm, 15 cm, or 65 cm). The planning venogram must be scrutinized for the best approach and direction to cross the occlusion: the length of occlusion, the trajectory, and any at risk adjacent structures while crossing the occluded portion of the vessel (lung, pericardium, bowel, etc). The needle can be guided through a direct percutaneous approach (through the skin into the patent vessel and across the occlusion) or via a catheter coaxially loaded into a vascular sheath (sheath into patent vein, catheter advanced to the site of occlusion, and needle through the catheter and eventually across the occlusion). If the direct percutaneous route is chosen for sharp recanalization, a 0.018″ or 0.014″ wire is directly passed into the central, patent vein (Fig. 6). The percutaneous tract is then serially dilated until a sheath is able to be passed over a wire and the rest of the procedure performed. If the coaxial system is chosen for sharp recanalization, the needle is coaxially loaded through an angled catheter, directed towards the occlusion, and passed across (Fig. 7). Typically, a loop snare is placed in the target vessel and once the needle is passed into the snare, the snare is tightened over the needle. The inner obturator is removed from the needle, a 0.018″ wire is passed through the needle, and the snare is then tightened around the wire. The wire is pulled into the snare catheter thereby establishing “flossed” access across the occlusion. A catheter or long tapered sheath is advanced across the wire through the occlusion, and then the wire can be exchanged to a larger, stiffer wire (Amplatz, Rosen, or Lunderquist Cook Bloomington, IN) if necessary.
Sharp recanalizations of venous occlusions are often successful, but also carry significant risk if the anticipated needle pass cannot avoid crossing through critical structures or the occlusion length is too long (usually more than 3 cm). The use of intravascular ultrasound (IVUS) has lessened the risk of sharp venous recanalizations. The IVUS-assisted technique for chronic venous occlusions was adapted from the well-described use of IVUS in transjugular intrahepatic portosystemic shunt (TIPS) creation (Farsad et al. 2012; Kao et al. 2016). A longitudinal, side-firing IVUS is placed in the central target vein instead of a snare device, and the complementary US system is used to display the visualized vascular structures. The active tip of the IVUS probe is angled until the needle tip is visualized, which allows for real time visualization of the needle pass in relation to the surrounding structures. Once the needle is confirmed to be within the lumen of the target vein, a wire is advanced centrally. The IVUS probe is removed and exchanged for a snare device; the wire is snared and “flossed access” is achieved across the occlusion. Real time visualization allows for path correction as well as making fewer passes to achieve the recanalization.
There are various other endovascular techniques described in the literature to cross refractory chronic venous occlusions; rates of such refractory occlusions are reported to be as high as 18% in adult literature (Keller et al. 2018). One such technique is radiofrequency wire-aided traversal. The radiofrequency wire (RF) is more commonly used in adults than pediatrics; however, the technique may be applicable to adolescent/near-adult sized patients and should be discussed for thoroughness in this review article. RF wires have been reported in the literature for small cohort studies and retrospective reviews (total number of cases is approximately less than 90 cases) (Keller et al. 2018). The RF wire is used to traverse occlusions in the SVC, brachiocephalic, IVC, and iliac veins with multiple sites of venous access on either side of the occlusion for the procedure. The RF wire is often used only after more standard techniques to cross the occlusion have failed (multiple combinations of hydrophilic stiff wires, catheters and sheaths). The RF wire is positioned on one side of the occlusion with a target snare on the opposite side of the occlusion; the RF wire was advanced slowly under fluoroscopy and/or cone beam CT. When the RF wire makes contact with the metal snare, the system is short circuited which serves as the final confirmation the occlusion had been crossed. In a retrospective review, the mean length of occlusion crossed by the RF wire was 10.05 cm (range 0.8–31.7 cm, median 5.2 cm) (Keller et al. 2018). Technical success rates are reported as high as 80% and 91% (Keller et al. 2018; Sivananthan et al. 2015); and while complication rates are low (roughly 2–5%), the inadvertent injuries tend to be major complications such as tracheal or SVC perforation. Given a favorable anatomic location and carefully selected patient, RF wires should remain a consideration for those who are comfortable and experienced in using them in a refractory chronic venous occlusion.
Once the access is achieved across the occlusion, the patient should be heparinized as the following steps of angioplasty and possible stenting lead to a prothrombotic milieu. At the authors’ institution, the typical loading dose of heparin is 100 units/kg up to max dose of 5000 units. With the wire access established, balloon angioplasty is next step to restoring patency. Noncompliant balloons such as Conquest (BD Bard Tempe, AZ), Dorado (BD Bard Tempe, AZ), Atlas (BD Bard Tempe, AZ), and Powerflex (Cordis Santa Clara, CA) are often used; sizing is based on venograms and IVUS measurements of the vessels flanking the occlusion.
Once a patent channel has been established, stenting is usually performed if the child is skeletally mature and the location of the proposed stents is conducive to stenting. If the child is not skeletally mature, stents can still be considered in the supradiaphragmatic region (i.e. brachiocephalic and SVC) as long as the stent complex can be serially ballooned to a larger caliber over time. This is often achieved with placing a self-expanding stent through a balloon mounted stent; the overall stent diameter is thereby constrained by the smaller, balloon-expandable stent. The “hourglass” shaped, constrained stent complex will provide a controlled expansion with follow up angioplasty procedures to expand the total stent diameter as the child grows (Image 7E). If the location of the occlusion is not amenable to stent placement (near a joint or in the extremity) or if the child is not skeletally mature, repeat balloon angioplasty, potentially with drug eluting balloons, and prophylactic anticoagulation is considered the safest treatment (Image 3D).
SVO which typically require stent placement for long term patency include SVC, IVC, and iliac veins (May Thurner). Common stents used in these locations include Wallstent (Boston Scientific Marlborough, MA), Palmaz (Cordis Baar, Switzerland), Viabahn (Gore Medical Flagstaff, AZ), S.M.A.R.T. Control stent (Cordis Baar, Switzerland), and the recently introduced Venovo (BD Bard Tempe, AZ) and Vici stent (Boston Scientific Marlborough, MA). Only in select circumstances, stents are deployed below the inguinal ligament. Van Vuuren et al. described experience in stenting into the common femoral vein in 79 adult patients. Fourteen of the 79 patients had primary stent placement for post-thrombotic changes that extended peripherally to the femoral confluence, but the deep femoral vein had to provide the majority of venous inflow in order to maintain stent patency (van Vuuren et al. 2018).
Following delivery of the stent, angioplasty is performed to ensure the stent is adequately dilated and the stent wall abuts the vessel wall. Final diameter of the stent is determined based on the prior venograms and IVUS measurements (Haddad et al. 2018). Post stent venogram should show decreased number of adjacent collaterals, in-line flow of the deep venous structures, and no filling defects within the stent complex. There is a well-described phenomenon of apparent narrowing in the external iliac vein after placement of an adjacent common iliac venous stent and is thought to represent an imaging phenomenon with no clear clinical significance (Al-Hakim et al. 2018). IVUS is also helpful in evaluating the stent apposition, diameter, and coverage across the occlusion. Areas of slight intrastent stenosis or small, residual thrombus are easily identified with IVUS and may be missed with venography.
Anticoagulation or antiplatelet therapy should be initiated as soon as the stent complex is in place. After consultation with the Hematology service, the best agent is selected and prescribed for the patient. Despite the importance of maintaining stent patency in venous recanalizations, the literature does not suggest on particular method or agent over another. Various options include low molecular weight heparin (LMWH) with dual anti-platelet agents (enoxaparin, clopidogrel, and aspirin) where clopidogrel was only used for 2 months following stent placement (McDevitt et al. 2019), dual antiplatelet agents (clopidogrel and aspirin) (Lawson and Seckeler 2018), and LMWH and aspirin (Sullivan et al. 2018). Little to no data exists for anticoagulation with direct oral anticoagulants in children with chronic central venous occlusion relieved by stent deployment.
Follow up imaging after recanalization is imperative for long term management. In pediatrics, US or MRI is often preferred over CECT. However, US may under or overestimate stent patency due to operator error and MRI may be unable to answer the question of stent patency or stenosis due to metal artifact disturbances. Thus, if there is considerable concern whether the stent remains patent or not, it is often best answered with a diagnostic venogram or CECT. If IVUS is used during the diagnostic venogram, contrast and radiation doses can further be lowered. Primary patency rates of systemic venous occlusion treated with angioplasty alone are not well reported; data is just beginning to emerge for drug eluting balloon angioplasty in venous stenosis/occlusion and long term patency rates for this therapy have not been corroborated in pediatric cohorts. Primary patency rates for children with SVC syndrome (non-malignant occlusion) and stent placement are not well reported in the literature as there are predominantly case reports with no translatable data for larger patient groups. Primary patency rates in pediatric patients with IVC and iliac vein stents range from 64 to 100% in the first 12 months following stent placement (McDevitt et al. 2019; Sullivan et al. 2018; Lungren et al. 2018). Primary assisted and secondary patency rates were 100% in all studies (McDevitt et al. 2019; Lungren et al. 2018). Primary patency of left common iliac venous stents placed in adolescents for May Thurner have been reported at 79% and 89% for secondary patency (Lungren et al. 2018). However, rates of stent patency are not reliably reported beyond 36 months in most studies whether it is due to loss of patients to follow up or disparate mechanisms for documenting further interventions needed to maintain stent patency. The poor data rates for stent patency is particularly concerning when considering the life of the stent in a pediatric patient. Additional studies to better analyze existing data and propose solutions to routine follow up for stent maintenance are needed.