gk_chin_necrosis.jpg (21959 bytes)

a = right frontal lesion, b = 3 months after treatment (lesion almost gone) c = 14 months (radiation necrosis) d = 22 months (necrosis has resolved) from J Neurosurg 2001 Jun;94(6):899

Complications of Brain Radiosurgery

Read review article here

Because open surgery is not required, the immediate complications occasionally seen with a craniotomy (e.g. hemorrhage and infection) do not occur with radiosurgery. The primary risk is delayed radiation injury, which typically occurs 3 months or more after treatment as radiation necrosis in the white matter. The incidence of complications is  low and more detailed information is included with the studies related to specific sites.

Some general comments are noted below and some of the other studies are cited elsewhere and here

For patients treated with brain metastases if there is a persistent or recurrent mass, it may be recurrent or resistant tumor or radionecrosis, sometimes it is hard to tell the difference even if using MRI spectroscopy, but many patients benefit from a surgical resection (go here).

A number of risk factors have been identified for predicting radiation injury after radiosurgery. The two most important factors are treatment volume and radiation dose. (The larger the volume treated and the higher the dose, the high the risk.   The studies (see graphs below) by Kjellberg defined a dose-volume line that predicted a 1% risk of necrosis and the ILF (integrated logistic formula by Flickinger) predicted a 3% risk In the review by Chin (J Neurosurg 2001 Jun;94(6):899-904 and see below)  they noted the median time to symptomatic necrosis was 4 months (range 2- 14 months) the median time to symptomatic recovery was 7.5 months (range 2 - 16 months) and imaging recovery 10.5 months (range 6 - 16 months.) They found that the 10Gy volume ( if over 10cc) was the best predictor of risk. They found patient re-treated has a much higher risk (23.5% versus 0.4%) and certain tumor types (glioma 17%, metastases 5% and benign tumor 7.5%) They also reviewed the literature and found that the Kjellberg isoeffect line predicted a risk of necrosis that is higher than 1% (more like 3- 11%.

As techniques improve with more accurate imaging (using MRI's) and downward adjustments of the dose (as appropriate) the risk of cranial nerve complications has declined. As the studies below (and on the acoustic neuroma page) demonstrate from Foote (2001) cranial nerve injuries prior to 1994 were 29% and since have fallen to only 2-5%. Niranjon (1999) noted that the  preservation of hearing rose to 100% (from 73%) with lower doses. Flicking reported only 1-2% incidence of cranial neuropathies and the paper  by Miller (1999 Mayo Clinic) showed that the risk of facial nerve neuropathy fell from 38% down to 8% with lower doses..
 
gk_comps_volume.gif (5374 bytes) gk_comps_diameter.gif (5745 bytes)

the risk of necrosis (1% risk from Kjellberg model or 3% risk from ILF data based on the volume or diameter of tissue radiated to a given dose, if you stay below the curves the risk is minimal

Radiation necrosis following gamma knife surgery: a case-controlled comparison of treatment parameters and long-term clinical follow up.

Chin LS, Ma L, DiBiase S.    J Neurosurg 2001 Jun;94(6):899-904

Department of Neurosurgery and Radiation Oncology, University of Maryland School of Medicine, Baltimore, USA. lchin@smail.umaryland.edu

Radiation necrosis is the only significant complication of gamma knife surgery (GKS).  Between September 1994 and December 1998, 286 patients were treated with GKS by the senior author. Of the 243 patients who were suitable for analysis, 17 developed radiation necrosis and were prospectively followed. Concurrently, 17 patients without necrosis were randomly selected as case controls on the basis of histological findings in their lesions. Integral dose-volume histograms (DVHs) were calculated and dose-volume treatment parameters were determined. A comparison was made with both the established Kjellberg and Flickinger isonecrosis risk lines. Clinical outcome was assessed according to time to resolution of symptoms and return to normal radiographic appearance. CONCLUSIONS: Treatment plan variables associated with the risk of necrosis were increased tumor volume (TV) integral dose, increased TV, and increased 10-Gy volume. Other risk factors included repeated radiosurgery to the same lesion and glioma histological findings. The Kjellberg 1% risk line predicted a 5% risk of radiation necrosis and the Flickinger 3% risk line predicted a 3% risk. The median time to development of necrosis was 4 months, and symptomatic and radiographic recovery times were 7.5 and 10.5 months, respectively. The median survival time in patients with necrosis was 30 months. The authors recommend prospective TV determination and DVH calculation for all radiosurgical treatments and the avoidance of repeated radiosurgical treatments to the same lesion when possible.

gk_chin1.gif (3477 bytes)

gk_chin2.gif (2218 bytes)

graphs comparing the Kjellberg 1% (solid line) and Flickinger 3% (dotted line) iso-necrosis risk lines with the data from patients who developed necrosis compared with those who did not

scaterplot of 10-Gy volumes, comparing those who developed necrosis and those who did not

setstats 1

setstats 1

setstats 1