Home Hemochromatosis Iconotech DICOM softwares
 
 
Summary
  • Introduction
  • Technique
  • Indications
  • Results

  •     - .5T
        - 1T
        - 1.5T
  • Spleen...
  • References
  •  
     
     
    Iron, liver and MRI.
    Yves Gandon - Medical Imaging, Rennes, France.

     
       Introduction

    This document explains the way to detect/quantify liver iron overload using MR imaging. 

    It allows you:

    • to become familiar with the principle of the MR quantification of liver iron content and to discover the best indications,
    • to perform a study using your MR equipment by chosing the appropriate MR protocol depending on the field strength of your magnet,
    • and to estimate subjectively, or even calculate on-line, the liver iron content by running a java applet. The calculation algorithm is also depending on the the magnetic field strength of your MR-scan: 0.5T, 1.0T or 1.5T.
     
       Context

    Iron is integrated in the hemoglobin cycle. Iron in excess is predominantly stored in the liver. There is a homeostatic regulation mechanism. A constant small quantity of iron is daily eliminated by the bile. The intestinal absorption is regulated in order to keep the iron store constant.

    Two different mechanisms can produce iron overload:

    • Non-transfusional iron overload is due to an increase of intestinal iron absorption and iron is predominantly stored in the liver, within the hepatocytes where it can exert direct toxicity. This is typically the mechanism occurring in genetic hemochromatosis, by genetic dysregulation of intestinal uptake. This can be also observed in anemia in case of ineffective erythropoïesis. 
    • Iron overload can also be a consequence of multiple transfusions needed to treat severe anemias like ß-thalassemia. Iron is then predominantly deposited in the spleen and in the Kupffer cells where it is less toxic.

    Non-transfusional iron overload is frequent, probably underestimated, and mainly due in western countries to genetic hemochromatosis or dysmetabolic hepatosiderosis. Vital risks, such as cirrhosis and hepatocellular carcinoma can be reduced by an early treatment which consist on repeated phlebotomies (1). Iron overload is rarely evoked on specific clinical signs (melanodermia...) but must often by discovering an elevation of serum iron, transferrin saturation or ferritinemia or by doing family studies (2).

    The diagnosis of genetic hemochromatosis is now based on the specific genetic test (3). In that case the level of iron overload is well correlated to the ferritinemia.
    In the other diseases, serum tests are not a good reflect of the liver iron overload (5).

    On CT-scan images, liver density increases in case of liver iron overload. However this is not enough sensitive, particularly in case of associated steatosis (6). On the other hand, a liver density greater than 80 HU can of course be due to a major overload but also to a long term treatment by amiodarone, a glycogenosis, a Wilson disease… So, CT-scan has no place for the diagnosis or quantification of iron in the liver.

    The gold standard was, up to now, the biopsy with semi-quantitative assessment by histopathology and quantification of liver iron concentration (LIC) by biochemical analysis. Now, MRI is able to replace liver biopsy for this purpose.


     
       Quantification by MR

    The best non-invasive technique to assess the iron liver content is MRI. The superparamagnetic properties of iron stored in the liver explain the decrease of the T2 relaxation times of the liver, which leads to a decrease of the signal intensity of the hepatic parenchyma (7).

    Previous quantitative studies have tried to quantify major overload using spin-echo sequences(7-17). The assessment of liver iron content was one by using liver to muscle ratio (10, 11, 13, 14, 16, 19, 23-27) or by liver T2 relaxation time calculation(12, 16, 18, 20, 26). This parameter is more difficult to obtain and results are probably different from one equipment to another. This T2 calculation gave less sensitive result than liver to fat or liver to muscle ratio obtained from T2-weighted gradient echo sequences (26). 

    This is easy to understand knowing that gradient echo sequences are more sensitive to magnetic susceptibility. The use of T2-weighted gradient echo sequences has permitted to decrease the detection threshold of liver iron overload (26, 27) by comparison to T2-weighted spin echo gradient sequence (14) cancelling previous limitations (22). The use of T2-weighted gradient echo sequence is necessary to detect a slight iron overload (28).

    Using this type of sequences, the liver is usually hyperintense to the muscle. A liver hypointense to the muscle indicates a liver iron overload.

    Normal liver (LIC=20µmol/g, Signa 1.5T)
    GRE "T2+" 
    TR=120 ms, TE=14 ms, PA=20°
    GRE "PD" 
    TR=120 ms, TE=4 ms, PA=20°
    GRE "T1" 
    TR=120 ms, TE=4 ms, PA=90°

     

    In case of slight overload, liver signal intensity decrease can only be seen on T2-weighted gradient echo sequences. The liver is then hypointense to the muscle.
     
     
    Slight overload  (lic=50µmol/g, Signa 1.5T)
    GRE "T2+" 
    TR=120 ms, TE=14 ms, PA=20°
    GRE "PD" 
    TR=120 ms, TE=4 ms, PA=20°
    GRE "T1" 
    TR=120 ms, TE=4 ms, PA=90°

     

    In case of moderate overload the decrease of the liver signal intensity is depicted on all sequence..
     
     
    Moderate overload  (LIC=120µmol/g, Signa 1.5T)
    GRE "T2+" 
    TR=120 ms, TE=14 ms, PA=20°
    GRE "PD" 
    TR=120 ms, TE=4 ms, PA=20°
    GRE "T1" 
    TR=120 ms, TE=4 ms, PA=90°

     
    Major overload cannot be estimated using gradient echo sequences. Above 300µmol/g, liver signal intensity is at the same level that the background noise. A less sensitive sequence (a spin echo sequence with a very short TE around 12 ms) could be used.
     
     
    Major overload  (LIC=350µmol/g, Signa 1.5T)
    GRE "PD" 
    TR=120 ms, TE=4 ms, PA=20°
    GRE "T1" 
    TR=120 ms, TE=4 ms, PA=90°
    GRE "SE "T1" 
    TR=120 ms, TE=14 ms, PA=20°

     
    To get a quantitative data we proposed (as many authors 26, 27) to calculate the ratio of the liver signal intensity obtained by means of ROIs and the signal intensity of paraspinous muscles. A multicentric study which has enrolled 510 patients had permitted to define a simple technique available on all types of equipement. Results obtained allowed to design a calculation software available on this web site.

    If you have the opportunity to check the algorithm proposed in this site by comparison of MR results and biochemical assessment of liver iron concentration (LIC) do not hesitate to contact me.
     

      Details on MR protocol...  
     

    Update: June 10th, 2001 Write to the webmaster