Endovenous thermal ablation of the saphenous veins using radiofrequency ablation (RFA) or endovenous laser ablation has become the most common method of treatment for varicose veins and symptomatic venous reflux disease, and is getting wide acceptance. Minor complications such as skin bruising/hematoma, bleeding, transient paresthesias, and skin burns have been reported in 3% to 20% of patients. Major complications which include deep venous thrombosis (DVT) and pulmonary embolism (PE) are rare. Extension of thrombus from the saphenous to the femoral or popliteal veins have been reported to occur in 0% to 6% of patients. The term endovenous heat-induced thrombus (EHIT) was introduced by Kabnick who also classified the level of thrombus from 1 to 4 classes. Hingorani et al reported extension of thrombus in 16% of limbs treated with RFA, and raised caution about the procedure and recommended early postoperative duplex scan evaluation. Most of these thrombi retract or absorb, but they theoretically can also detach and cause a PE.
A 58-year-old man was referred to our office by his primary physician with symptoms of left calf pain and progressive edema which had gradually gotten worse over the previous 3 months. The symptoms were more prominent at the end of the day. There was no family history of thrombophilia and his medical history was negative. He had no history of DVT and no history of smoking. He was taking no medications. On examination, his weight was 134 pounds, height was 5=6, with a body mass index of 21.63. His blood pressure was 110 over 68, his pulse rate was 82/minute, and his respiratory rate was 16/minute. The general physical examination was essentially negative. The right leg had no evidence of varicose veins or stigmata of venous insufficiency. On the left leg, there was an area of skin hyperpigmentation in the distal medial calf. There were also obvious varicosities in the medial and posterior calf regions. His CEAP class was 4a and the Venous Clinical Severity Score was 6. The patient used compression stockings in the past for about 3 months without significant improvement. Duplex ultrasound scan evaluation was performed in our Intersocietal Commission for the Accreditation of Vascular Laboratories which showed reflux in the left great saphenous vein (GSV) with a maximum reflux of 3.5 seconds near the confluence of the saphenous vein. The maximum diameter of the GSV was 11 mm above the knee and 5 mm below the knee. The deep veins were all normal. The patient underwent RFA of the left GSV using ClosureFAST (VNUS MEDICAL Technologies Inc, San Jose, Calif) in a standard technique, with the head down position using tumescent anesthesia totaling 290 ml. The vein was accessed below the knee. The catheter tip was 2.8 cm from the saphenofemoral junction and 2.3 cm from the superficial epigastic vein. No prophylactic anticoagulation or aspirin was given to the patient. Postoperatively, the patient was active and walking, and used elastic stockings. The patient was re-evaluated 3 days later in our vascular laboratory, and the GSV was noted to be completely occluded. There was EHIT class 2 in the left saphenofemoral junction. The common femoral vein was compressible and had flow but there was thrombus protruding into the lumen filling <50% of the lumen (Fig 1). There was no loose tail or floating thrombus. The evaluation was done in a routine fashion and no excessive compression was used. When the saphenofemoral region was re-evaluated after the compression test, the thrombus, which was protruding into the femoral vein, disappeared (Fig 2). The patient was immediately referred to the hospital for PE workup. A computed tomography (CT) scan of the chest with contrast was obtained. This was positive for acute bilateral small segmental pulmonary emboli (Fig 3). The patient had no symptoms of cough, chest pain, or shortness of breath. His vital signs were normal except for a heart rate of 116. The pulse oximetry was 100% on room air. He was admitted to the hospital and treated with enoxaparin sodium (lovenox) 1 mg/kg until the international normalized ratio (INR) was therapeutic. He was discharged from the hospital after 3 days on 7.5 mg of warfarin daily by mouth and followed as an out-patient. Follow-up duplex scans done after 1 week, 1 month, 2 months, and 6 months showed no evidence of DVT and the GSV remained occluded. The oral anticoagulation was discontinued after 4 months. The Venous Clinical Severity Score improved from 6 to 4.
DVT and PE are rare complication of EHIT. Kabnick introduced the term EHIT and noted that this is more benign than the spontaneously occurring thrombosis, in that it is stable and usually regresses or shows complete resolution. He also made the observation that EHIT displays a different sonographic echogenicity and becomes echogenic in 24 hours. The EHIT in our case does show increased echogenicity. He classified EHIT from class 1 to class 4. Another classification system was recently introduced by Lawrence et al6 from level 1 to 6. The International Endovascular Working Group registry shows that DVT/EHIT occurred in 0.27% (10 of 3696 cases) and PE occurred in 0.023% (1 of 3696) after endovenous laser ablation. In a review of 11 articles, Mozes et al7 reported 21 case of DVT and 2 cases of PE after the VNUS Closure (VNUS MEDICAL Technologies Inc) procedure. At the Arizona Heart Institute, with >1000 cases of venous ablation, only 1 case of pulmonary embolism was reported. To decrease the risk of EHIT formation, several suggestions have been offered: the position of the catheter tip should be at least 2 cm from the saphenofemoral or saphenopopliteal junction, reduce the thrombus load by elevation of the leg during ablation, and also by using adequate tumescent anesthesia. A recent article showed that a GSV diameter of >8 mm and history of DVT were associated with EHIT class 2 or greater. But another study which evaluated the influence of procedural factors concluded that there was no difference in catheter tip position or mean diameter of the treated vein between the EHIT and non-EHIT groups. There is no report on association of EHIT with hypercoagulable states. Our patient had no history of DVT or family history of thrombophilia, and the catheter tip was definitely identified at 2.8 cm from the saphenofemoral junction and 2.3 cm from the superficial epigastic vein. The diameter of his GSV was 11 mm, which may have contributed to the EHIT. We hypothesize that the standard technique of compression of the superficial and deep veins used to evaluate for DVT at the saphenofemoral junction may have contributed to the dislodgement of the thrombus causing PE. To our knowledge, there has been no report of PE caused during ultra-sound scan evaluation of EHIT. We recommend that when thrombus protrusion into the femoral or popliteal vein is observed after endovenous ablation, one should not use compression of the femoral vein, to avoid dislodgement and possible PE. Instead, a Valsalva maneuver should suffice. The treatment of asymptomatic PE after venous ablation is controversial, but it may not be necessary when the thrombus load is small. Further clinical study is needed to clarify this point.