Article Text

Dissolution of tungsten coils leads to device failure after transcatheter embolisation of pathologic vessels
  1. Department of Pedatric Cardiology and Pediatric Intensive Care
  2. Hannover Medical School
  3. *Institute of Materials Science
  4. University of Hannover
  5. Institute of Inorganic Chemistry
  6. University of Hannover
  7. Germany
  1. Dr med Matthias Peuster, Department of Pediatric Cardiology and Pediatric Intensive Care, Hannover Medical School, 30625 Hannover, Germany; Peuster.Matthias{at}

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Coil embolisation has been the standard transcatheter technique for occlusion of pathologic vessels, aneurysms, and fistulae for more than three decades. Tungsten is an attractive material for use as an endovascular coil, because of its high radiopacity and thrombogenicity.1 Recently, decreasing radiopacity has been reported in patients after implantation of tungsten coils (MDS, Balt Extrusion, Montmorency, France) into cerebral aneurysms > 30 months after implantation.2 Analysis of the coils in an experimental study demonstrated “a clear degree of iron” around the MDS coil occluded arterial aneurysms. The authors concluded: (1) that the coils are not purely tungsten; (2) that there was corrosion of the coils; and (3) that it may be dangerous to use this type of coil in patients.3 To assess whether similar findings could be observed in our patients all transcatheter tungsten coil implantations performed at our institution were reviewed. Furthermore a detailed analysis of the composition of the MDS coil was performed. Out of 104 patients with transcatheter coil occlusion of pathologic vascular connections 87 tungsten coils were implanted in 21 patients. Angiography after the implantation confirmed complete coil occlusion of the vessel in all patients.

Serum tungsten concentrations were analysed by inductive coupled plasma mass spectrometry focusing on the isotope 184 (Labor Schiwara, Hafer Sende, Bremen, Germany); normal values were used from a large cohort study4. The MDS coils were analysed using both inductively coupled plasma atomic emission spectroscopy (ICP-MS; VG Elemental PlasmaQuad II Turbo+) and inductively coupled plasma atomic emission spectroscopy (ICP-AES; Spectro-Flame-EOP) after pressure dissolution of the coil with HF, HNO3, H2O2, HClO4 at 37–200°C and by use of wavelength dispersivex ray spectrometry (WDS). To avoid interference with surface contaminants, the coils were embedded in a conductive embedding medium and ground to achieve representative cross sections.

In 21 patients who received tungsten coils a mean follow up of 97 months (two days to 38 years) was achieved. Fluoroscopy performed in 14/21 patients revealed a decreased radiopacity in 9/14 patients (fig1). Repeat angiography performed in 7/21 patients demonstrated recanalisation of 1–4 of the previously MDS occluded vessels in 5/7 patients; in all patients with recanalised vessels a decreased radiopacity of the coils was observed. Serum concentrations for tungsten were greatly increased in 8/8 patients, ranging from 2.0 μg/l to 14.4 μg/l (mean 6.43 μg/l, normal value < 0.2 μg/l). No unexplained clinical symptoms were reported on follow up examination. The MDS coils we examined are produced from > 99.9 mas.% tungsten.

Figure 1

(A) Fluoroscopy immediately after implantation of tungsten coils in a 22 year old patient with multiple aortopulmonary collaterals. (B) Fluoroscopy six months after implantation. Decreased radiopacity of the tungsten coils (arrows).

Dissolution and device failure of a “permanent” medical implant caused by recanalisation of a previously occluded vessel can occur along with raised serum concentrations of the degradation products. It may be speculated that minimal residual shunting was present in our patients early after coil implantation, thereby accelerating the degradation of tungsten by keeping the pH at a level ⩾ 7.4. However, no adverse clinical effects were identified during follow up of the patients. Although there are no reports published on toxicity or unexplained clinical symptoms in patients receiving tungsten coils, there is a lack of concise data on the toxicity and clearance of tungsten in humans.5 Our own analysis shows that MDS coils are produced from pure tungsten (> 99.9 mas.%). We assume that the analysis performed by others3 was undertaken with an energy dispersive x ray spectrometer (EDX) and the deviation may be caused by incidental superposition of the energy of ultimate and subordinate lines; adequate results can only be obtained with a wavelength dispersive spectrometer because of its superior specificity.

We conclude that tungsten coils can dissolve leading to implant failure and greatly increased serum tungsten concentrations. The clinical use of these coils can no longer be recommended. Although there is lack of evidence for in vitro or in vivo toxicity of tungsten, patients with MDS coils should be followed thoroughly with tests for serum tungsten concentrations, and liver and renal function. Repeat angiography may be warranted if decreased radiopacity is observed.