Objective: This experimental study assessed the usefulness of a rat model of corneal angiogenesis for assessing the angiogenic activity of cerebrovascular malformations (CVMs) over time. It is the first investigation to have focused on dynamic quantification of angiogenesis related to CVMs (i.e., assessing changes over time as opposed to one point in time) and the first to have compared different CVMs and other tissues in this way.
Methods: The CVM specimens were from 15 randomly selected arteriovenous malformation (AVM) cases, 15 randomly selected cavernous malformation cases, and 2 venous angioma cases. All the samples came from surgical resections performed in the Marmara University Department of Neurosurgery and the Marmara University Institute of Neurological Sciences between January 1998 and January 2003. Glioblastoma multiforme tissues, normal adult brain tissues, and normal brain artery tissues were used as controls. Tissue from each specimen was implanted in one corneal micropocket in each cornea of a single rat, and the level of angiogenic activity in the cornea was graded (low, moderate, or high) at Days 1 through 9 after tissue implantation. Another set of rats was subjected to the same corneal implantation procedure, and two widely accepted indicators of neovascularization, microvessel counts and vascular endothelial growth factor expression, were assessed at Days 3, 5, 7, and 9 after surgery. The tissue-group results for grade of angiogenic activity, microvessel count, and grade of vascular endothelial growth factor expression were compared.
Results: Of the three CVM types, the AVMs showed the highest angiogenic activity, cavernous malformations exhibited some degree of angiogenic activity (less than AVMs but more than normal brain artery tissue), and angiogenesis induction by venous angiomas was comparable to that of normal brain artery tissue. Corneas implanted with the AVM samples showed Grade 1 angiogenesis on Day 2, Grade 2 angiogenesis on Day 4, and Grade 2.5 angiogenesis on Day 8. Cavernous malformation samples caused Grade 1 angiogenesis after Day 5, and grading remained less than 1.5. Venous angioma samples resulted in angiogenesis graded less than 1.
Conclusion: The results showed that this rat model of corneal angiogenesis is of value for assessing the angiogenic potential of CVMs and for evaluating details of the neovascularization process over time. Further investigation of CVMs with this model might yield information about how angiogenesis can be modulated, and this could lead to development of biological treatments that inhibit this process.