4 or 8 mm collimator shot for multiple metastases treatments?

Hamza Benmakhlouf1, Thomas Kraepelien2, Ernest Dodoo3

1Norsborg, Sweden 2Department of Hospital Physics, Karolinska University Hospital 3Department of Neurosurgery, Karolinska University Hospital

Keywords: brain metastasis, gamma knife, dose planning, Dose distribution, collimator angle

Interactive Manuscript

Ask Questions of this Manuscript:



     One of the most common indications treated with Leksell Gamma Knife Perfexion is multiple metastases. Often a significant amount of metastases are in the range of 2-6 mm in diameter which can be covered with one shot. The question raised is whether to use a 4 or an 8 mm collimator to cover those very small lesions? For a long time it has been argued that the 4 mm collimator should only be used for functional treatments (i.e. trigeminal neuralgia). However, is this really the optimal choice from a physical, technical, radiobiological and oncological point of view? From the physical aspect there are two main things: the dose-distribution of each shot, the uncertainty in the MR or CT images, and the dose algorithm with its approximations of the head as water (neglecting the air cavities present in the scull). The radiobiological and oncological aspect raises the question of the requirement of a marginal (to cover the latent tumour cells), in that case, how large should a margin be? 
     To draw attention to the physical aspect the dose profiles from a 4 and an 8 mm collimator shots will be evaluated and compared to the lesion sizes.
     No patients were evaluated.
     The dose-profile of a 4 mm and an 8 mm collimator was inserted in the same plot (data was taken from LGP with the dose algorithm TMR10); 22 Gy was prescribed to the 50 % isodose for the former and 22 Gy to the 85 % isodose for the latter (a conventional way to minimize the dose marginal). The treatment time was 20 minutes and 10 minutes, respectively. To the same plot, the tumour dimensions (in this case 2, 4 and 6 mm tumours) were inserted to be able to see the marginal, dose-distributions and homogeneity.
     The margins of the 22 Gy isodose for the different sizes were 2, 1 and 0 mm for the 2, 4, and 6 mm lesions, respectively. In the case of the 10 Gy isodose the margin were 3.5, 2.5, and 1.5 mm for the 2, 4, and 6 mm lesions, respectively.
     This concept was not tested in an actual patient study.
     When considering the uncertainties in the MR images and the dosimetrical approximations along with the latent tumour cells around the contrast enhanced part of the lesion it is argued that a margin is needed.  
     The extent of this margin has to be decided taking the previous arguments into account, then the optimal collimator for treatment of small lesions can be chosen.


Project Roles:

H. Benmakhlouf (), T. Kraepelien (), E. Dodoo ()